Comparative genomics and transcriptomics of lineages I, II, and III strains of Listeria...

Evolutionary Biology – Concepts, Molecularand Morphological Evolution

Pierre PontarottiEditor

Evolutionary Biology –Concepts, Molecular andMorphological Evolution

EditorDr. Pierre PontarottiUMR 6632Universite d’Aix-Marseille/CNRSLaboratoire Evolution Biologique etModelisation, case 19Place Victor Hugo 313331 Marseille Cedex [email protected]

ISBN 978-3-642-12339-9 e-ISBN 978-3-642-12340-5DOI 10.1007/978-3-642-12340-5Springer Heidelberg Dordrecht London New York

Library of Congress Control Number: 2010933958

# Springer-Verlag Berlin Heidelberg 2010This work is subject to copyright. All rights are reserved, whether the whole or part of the material isconcerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting,reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publicationor parts thereof is permitted only under the provisions of the German Copyright Law of September 9,1965, in its current version, and permission for use must always be obtained from Springer. Violationsare liable to prosecution under the German Copyright Law.The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply,even in the absence of a specific statement, that such names are exempt from the relevant protective laws andregulations and therefore free for general use.

Cover design: WMXDesign GmbH, Heidelberg, Germany

Cover illustration:An antennal tip of a female parasitic wasp (Ichneumonidae: Cryptinae: Latibulus sp.).See Fig. 16.3b

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

Preface

The 13th Evolutionary Biology Meeting was held in Marseille on the 22–25

September 2009. These events aim to gather leading scientists involved in research

on evolutionary biology, promoting an exchange of state-of-the-art knowledge and

the initiation of inter-group collaborations. Over the past years, this has been

rewarded by the publication of several important review articles dealing with this

subject matter. For me personally, the Evolutionary Biology Meeting is a valuable

scientific exchange platform serving as booster for the use of evolutionary-based

approaches not only in biology but also in other scientific fields.

In 2009, some 100 presentations (oral, as well as “fast presentation” and

traditional posters) admirably reflected the epistemological nature of the meeting.

I selected one fifth of the most representative contributions for this book, these 21

articles being organized in different categories: Evolutionary Biology Concepts,

Genome/Molecular Evolution, and Morphological Evolution/Speciation.

I would like to thank the contributors to this book, as well as all other partici-

pants who helped making this meeting such as success, and our sponsors – the

Universite de Provence, CNRS, GDR BIM, Conseil General 13, and Ville de

Marseille. I gratefully acknowledge the support of members of the Association

pour l’Etude de l’Evolution Biologique (AEEB). In addition, I am indebted to the

staff of our publisher, Springer, for their competence and help.

Last but not least, I sincerely wish to thank the AEEB coordinator, Axelle

Pontarotti, for the excellent organization of the meeting and the production of the

book. In terms of collaborative scientific exchange and the publication of this

proceedings, the scientific output of the 13th Marseille meeting reflects the high

quality not only of individual contributions but also of the Marseille way of hosting,

for which Axelle Pontarotti is an outstanding ambassador.

Marseille, France Pierre Pontarotti

May 2010

v

Contents

Part I Evolutionary Biology Concepts

1 Extinct and Extant Reptiles: A Model System for the Study

of Sex Chromosome Evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

Daniel E. Janes

2 Constraints, Plasticity, and Universal Patterns in Genome

and Phenome Evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

Eugene V. Koonin and Yuri I. Wolf

3 Starvation-Induced Reproductive Isolation in Yeast . . . . . . . . . . . . . . . . . 49

Eugene Kroll, R. Frank Rosenzweig, and Barbara Dunn

4 Populations of RNA Molecules as Computational Model

for Evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

Michael Stich, Carlos Briones, Ester Lzaro, and Susanna C. Manrubia

5 Pseudaptations and the Emergence of Beneficial Traits . . . . . . . . . . . . . . 81

Steven E. Massey

Part II Genome/Molecular Evolution

6 Transferomics: Seeing the Evolutionary Forest Using

Phylogenetic Trees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101

John W. Whitaker and David R. Westhead

7 Comparative Genomics and Transcriptomics of Lactation . . . . . . . . . 115

Christophe M. Lefevre, Karensa Menzies, Julie A. Sharp,

and Kevin R. Nicholas

vii

8 Evolutionary Dynamics in the Aphid Genome: Search

for Genes Under Positive Selection and Detection

of Gene Family Expansions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133

Morgane Ollivier and Claude Rispe

9 Mammalian Chromosomal Evolution: From Ancestral States

to Evolutionary Regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143

Terence J. Robinson and Aurora Ruiz-Herrera

10 Mechanisms and Evolution of Dorsal–Ventral Patterning . . . . . . . . . . 159

Claudia Mieko Mizutani and Rui Sousa-Neves

11 Evolutionary Genomics for Eye Diversification . . . . . . . . . . . . . . . . . . . . . . 179

Atsushi Ogura

12 Do Long and Highly Conserved Noncoding Sequences

in Vertebrates Have Biological Functions? . . . . . . . . . . . . . . . . . . . . . . . . . . . 187

Yoichi Gondo

Part III Morphological Evolution/Speciation

13 Male-Killing Wolbachia in the Butterfly Hypolimnas bolina . . . . . . . . 209

Anne Duplouy and Scott L. O’Neill

14 Evolution of Immunosuppressive Organelles from DNA

Viruses in Insects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229

Brian A. Federici and Yves Bigot

15 The Neogastropoda: Evolutionary Innovations of Predatory

Marine Snails with Remarkable Pharmacological Potential . . . . . . . . 249

Maria Vittoria Modica and Mande Holford

16 Antennal Hammers: Echos of Sensillae Past . . . . . . . . . . . . . . . . . . . . . . . . . 271

Nina Laurenne and Donald L.J. Quicke

17 Adaptive Radiation of Neotropical Emballonurid Bats:

Molecular Phylogenetics and Evolutionary Patterns

in Behavior and Morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283

Burton K. Lim

18 Trends in Rhizobial Evolution and Some Taxonomic Remarks . . . . 301

Julio C. Martınez-Romero, Ernesto Ormeno-Orrillo, Marco A. Rogel,

Aline Lopez-Lopez, and Esperanza Martınez-Romero

viii Contents

19 Convergent Evolution of Morphogenetic Processes in Fungi . . . . . . . 317

Sylvain Brun and Philippe Silar

20 Evolution and Historical Biogeography of a Song Sparrow

Ring in Western North America . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329

Michael A. Patten

21 Cave Bear Genomics in the Paleolithic Painted Cave

of Chauvet-Pont d’Arc . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343

Celine Bon and Jean-Marc Elalouf

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357

Contents ix

Contributors

Yves Bigot Laboratoire d’Etude des Parasites GenetiquesParc Grandmont,

Universite de Tours, U.F.R. des Sciences et Techniques, 37200 Tours, France

Celine Bon CEA, IBiTec-S, F-91191, Gif-sur-Yvette cedex, France, celine.bon@

cea.fr

Sylvain Brun UFR des Sciences du Vivant, Universite de Paris 7 – Denis Diderot,

75205 Paris Cedex 13, France; Institut de Genetique et Microbiologie, UMR

CNRS – Universite de Paris 11, UPS Bat. 400, 91405, Orsay cedex, France

Barbara Dunn Department of Genetics, Stanford University, Stanford,

CA 94305, USA

Anne Duplouy School of Biological Sciences, The University of Queensland,

Brisbane, QLD 4072, Australia, [email protected]

Jean-Marc Elalouf CEA, IBiTec-S, F-91191 Gif-sur-Yvette cedex, France

Brian A. Federici Department of Entomology and Interdepartmental Graduate

Programs in Genetics and Microbiology, University of California, Riverside,

CA 92521, USA; Laboratoire d’Etude des Parasites GenetiquesParc Grandmont,

Universite de Tours, U.F.R. des Sciences et Techniques, 37200 Tours, France,

[email protected]

Yoichi Gondo Mutagenesis and Genomics TeamRIKEN BioResource Center,

3-1-1 Koyadai, Tsukuba 305-0074, Japan, [email protected]

Mande Holford York College and Graduate Center, and The American Museum

of Natural History, The City University of New York, NY, USA, mholford@york.

cuny.edu

xi

Daniel E. Janes Department of Organismic and Evolutionary Biology, Harvard

University, Cambridge, MA 02138-3899, USA, [email protected]

Eugene V. Koonin National Center for Biotechnology Information, National

Library of Medicine, National Institutes of Health, Bethesda, MD 20892, USA,

[email protected]

Eugene Kroll Division of Biological Sciences, University of Montana, Missoula,

MT 59812, USA, [email protected]

Nina Laurenne Museum of Natural History, Entomology Division, University

of Helsinki, P.O. Box 17(P. Arkadiankatu 13), 00014, Helsinki, Finland, nina.

[email protected]

ChristopheM. Lefevre Institute for Technology Research and Innovation, Deakin

University, Waurn Ponds, Geelong, VIC 3217, Australia; CRC for Innovative

Dairy Products, Department of Zoology, University of Melbourne, Melbourne,

VIC 3010, Australia; Victorian Bioinformatics Consortium, Monash University,

Clayton, Melbourne, VIC 3080, Australia, [email protected]

Burton K. Lim Department of Natural History, Royal Ontario Museum, 100

Queen’s Park, Toronto, Ontario M5S 2C6, Canada, [email protected]

Aline Lopez-Lopez Centro de Ciencias Genomicas, UNAM, Av. Universidad,

Cuernavaca, Morelos 62210, Mexico

Julio C. Martınez-Romero Centro de Ciencias Genomicas, UNAM,

Av. Universidad, Cuernavaca, Morelos 62210, Mexico

Esperanza Martınez-Romero Centro de Ciencias Genomicas, UNAM,

Av. Universidad, Cuernavaca, Morelos 62210, Mexico, esperanzaeriksson@

yahoo.com.mx

Steven E. Massey Biology Department, University of Puerto Rico – Rio Piedras,

P.O. Box 23360, San Juan, Puerto Rico 00931, USA, [email protected]

Karensa Menzies Institute for Technology Research and Innovation, Deakin



VIC 3010, Australia

Claudia Mieko Mizutani Department of Biology, Case Western Reserve

University, 10900 Euclid Ave, Cleveland, OH 447080, USA Department of

Genetics, Case Western Reserve University, 10900 Euclid Ave, Cleveland,

OH 447080, USA, [email protected]

xii Contributors

Maria Vittoria Modica Sapienza University of Rome, Piazzale Aldo Moro 5,

00185 Rome, Italy, [email protected]

Kevin R. Nicholas Institute for Technology Research and Innovation, Deakin



VIC 3010, Australia

Scott L. O’Neill School of Biological Sciences, The University of Queensland,

Brisbane, QLD 4072, Australia

Atsushi Ogura Division of Advanced Sciences, Ochadai Academic Production,

Ochanomizu University, Ohtsuka 2-1-1, Bunkyo, Tokyo 112-8610, Japan, ogura.

[email protected]

Morgane Ollivier INRA, UMR1099 BiO3P, Domaine de la Motte, F-35653,

Le Rheu, France

Ernesto Ormeno-Orrillo Centro de Ciencias Genomicas, UNAM,

Av. Universidad, Cuernavaca, Morelos 62210, Mexico

Michael A. Patten Oklahoma Biological Survey and Department of Zoology,

University of Oklahoma, 111 E. Chesapeake Street, Norman, OK 73019, USA,

[email protected]

Donald L.J. Quicke Department of Life Sciences, Imperial College London, Sil-

wood Park Campus, Ascot, Berkshire SL5 7PY, UK; Department of Entomology,

Natural History Museum, London, SW7 5BD, UK

Claude Rispe INRA, UMR1099 BiO3P, Domaine de la Motte, F-35653, Le Rheu,

France, [email protected]

Terence J. Robinson Evolutionary Genomics Group, Department of Botany and

Zoology, University of Stellenbosch, Private Bag X1, Matieland 7602, South

Africa, [email protected]

Marco A. Rogel Centro de Ciencias Genomicas, UNAM, Av. Universidad,

Cuernavaca, Morelos 62210, Mexico

R. Frank Rosenzweig Division of Biological Sciences, University of Montana,

Missoula, MT 59812, USA

Aurora Ruiz-Herrera Unitat de Citologia i Histologia, Departament de Biologia

Cel.lular, Fisiologia i Inmunologia, Universitat Autonoma de Barcelona, Campus

Contributors xiii

Bellaterra, 08193, Barcelona, Spain; Institut de Biotecnologia i Biomedicina,

Universitat Autonoma de Barcelona, Campus Bellaterra, 08193 Barcelona, Spain,

[email protected]

Julie A. Sharp Institute for Technology Research and Innovation, Deakin



VIC 3010, Australia

Philippe Silar UFR des Sciences du Vivant, Universite de Paris 7 – Denis Diderot,

75205 Paris Cedex 13, France; Institut de Genetique et Microbiologie, UMR

CNRS – Universite de Paris 11, UPS Bat. 400, 91405 Orsay cedex, France,

[email protected]

Rui Sousa-Neves Department of Biology, Case Western Reserve University,

10900 Euclid Ave, Cleveland, OH 447080, USA

Michael Stich Dpto de Evolucion Molecular, Centro de Astrobiologıa

(CSIC-INTA), Ctra de Ajalvir, km 4, Torrejon de Ardoz, Madrid 28850, Spain,

[email protected]

David R. Westhead Institute of Molecular and Cellular Biology, University of

Leeds, Garstang Building, Leeds LS2 9J, UK, [email protected]

John W. Whitaker Institute of Molecular and Cellular Biology, University of

Leeds, Garstang Building, Leeds, LS2 9J, UK, [email protected]

Yuri I. Wolf National Center for Biotechnology Information, National Library of

Medicine, National Institutes of Health, Bethesda, MD 20892, USA

xiv Contributors

Part IEvolutionary Biology Concepts

Chapter 1

Extinct and Extant Reptiles: A Model System

for the Study of Sex Chromosome Evolution

Daniel E. Janes

Abstract The evolution and functional dynamics of sex chromosomes are focuses

of current biological research. Although common organismal morphologies and

functions of males and females are found among amniotes, underlying sex chromo-

some organizations and sex-determining mechanisms are widely variable. This

chapter investigates the role that reptiles play in the study of sex chromosome

evolution. Reptile studies have described the coevolution of genotypic sex determi-

nation and viviparity, the adaptive significance of sex-determining mechanisms,

and shared ancestry of chromosomes. Novel resources, including whole-genome

sequences and mapped sex-linked markers, have allowed researchers to examine

sex chromosome evolution in reptiles, an important group for this type of study for

their position as the sister group to mammals. Compared with mammals, reptiles

exhibit much more variability in sex chromosome organization, providing raw

material for study of sex chromosome evolution across amniotes.

1.1 Introduction

Embryos develop as either male or female depending on factors that vary widely

among amniotes. Broadly speaking, amniotes can be classified as either genotypi-

cally sex-determined (GSD) or temperature-dependently sex-determined (TSD).

Embryos of GSD species, including all mammals, birds, snakes, and many lizards

and turtles, develop as either male or female depending on chromosomal contribu-

tions from parents at conception. Many, but not all, of these species exhibit detectable

D.E. Janes

Department of Organismic and Evolutionary Biology, Harvard University, Cambridge, MA

02138-3899, USA

e-mail: [email protected]

P. Pontarotti (ed.), Evolutionary Biology – Concepts, Molecularand Morphological Evolution, DOI 10.1007/978-3-642-12340-5_1,# Springer-Verlag Berlin Heidelberg 2010

3

cytogenetic sex differences (i.e., heteromorphic sex chromosomes). The difference

between heteromorphic and homomorphic sex chromosomes could be explained by

the length of the interval since the origin of genotypic sex determination in a species

(Ohno 1967; Janes et al. 2010b). Apparently, sex chromosomes begin to diverge

from each other only after a new GSD system arises (see Sect. 1.3.1). This sex

difference in karyotype is not apparent in individuals of TSD amniotes that develop

as male or female primarily in response to incubation temperature, including all

crocodilians, tuataras, and some turtles and lizards.

In this review, I will describe the variability of sex-determining mechanisms

among amniotes. This variability includes, for example, the temperatures that trigger

male or female development and the timing of temperature’s effect among TSD

species, as well as the presence or absence and type of sex chromosomes in GSD

species. Almost all mammals exhibit male heterogamety in which females carry two

X sex chromosomes of the same size and content, whereas males carry one X sex

chromosome and one smaller, degenerated Y sex chromosome. In birds, females are

heterogametic which means they carry the smaller, degenerated W sex chromosome

and one larger, more gene-rich Z sex chromosome, whereas male birds carry two Z

sex chromosomes. This difference in heterogamety affects the genomics of amniotes

in ways that are discernible from genome sequencing and experimental evidence.

Further, the evolutionary history of sex-determining mechanisms informs the diffe-

rent arrangements of amniotic sex chromosomes that have been studied using

techniques that include phylogenetic inference, cytogenetic mapping, and measure-

ments of population genetics parameters. Recent studies of sex-determining mechan-

isms and, specifically, the evolution of sex chromosomes have focused on extinct and

extant reptiles for two reasons. First, nonavian reptiles exhibit greater variety of sex-

determining mechanisms and sex chromosomes than birds or mammals. Second,

genomic resources for reptiles (including birds) have recently improved to an extent

that previously untestable hypotheses are now open to experimentation and compar-

ative analyses (Janes et al. 2008).

1.2 Sex-Determining Mechanisms

1.2.1 Patterns and Variability

Amniote sex-determining mechanisms are typically described as either GSD or

TSD but within those categories, functional patterns vary. As described above,

GSD species vary in their organization of sex chromosomes [i.e., female hetero-

gamety (ZW system) or male heterogamety (XY system)] (Fig. 1.1a). Phylogenetic

inference and comparative chromosome hybridizations suggest that male and

female heterogamety have evolved more than once among amniotes although the

exact number of independent origins is debated (Ezaz et al. 2009; Organ and Janes

2008). Likewise, the number of independent origins of temperature-dependent sex

4 D.E. Janes

determination is not clear. Although the sex-determining mechanisms of two or

more species may respond to incubation temperature in a similar manner, the

similarity may represent convergence. Three basic patterns of sex-determining

response to incubation temperature (Types Ia, Ib, and II) have been described

(Fig. 1.1b) (Bull 1983). Species that exhibit Type Ia temperature-dependent sex

determination, such as loggerhead (Caretta caretta), green (Chelonia mydas), andleatherback (Dermochelys coriacea) sea turtles, produce more male offspring from

eggs incubated at cooler temperatures (Standora and Spotila 1985). Species with

Type Ib temperature-dependent sex determination, such as all crocodilians, produce

more male offspring from eggs incubated at warmer temperatures (Valenzuela

2004). Species with Type II temperature-dependent sex determination, such as

leopard geckos (Eublepharis macularius), produce a maximal proportion of males

from eggs incubated at an intermediate temperature, whereas cooler or warmer

temperatures yield higher proportions of females (Janes and Wayne 2006; Viets

et al. 1994).

Male Heterogamety

Female Heterogamety

ZZ

AAAA

Z

W

XX

Y

X

No Heterogamety

Incubation Temperature

% M

ale

Offs

prin

g / C

lutc

h

Type Ia TSD

Male HeterogametyFemale HeterogametyNo Heterogamety

Type Ib TSDType II TSDGSD

a b

Fig. 1.1 (a) Pairs of sex chromosomes that consist of either a male-specific Y chromosome and an

X chromosome or a female-specific W chromosome and a Z chromosome. Species that exhibit

these sex chromosomes are described as either male heterogametic (XY system) or female

heterogametic (ZW system). Other GSD species exhibit no detectable heterogameties or sex

differences in karyotype. (b) Influence of incubation temperature on offspring sex ratios among

temperature-dependently (TSD) and genotypically sex-determined (GSD) species. The y-axismodels the proportion of males yielded per clutch of eggs incubated at different points on the

thermal gradient indicated on the x-axis. Sex-determining response to incubation temperature

follows one of three patterns (Type Ia, Ib, or II) in TSD species. GSD species produce similarly

balanced offspring sex ratios regardless of incubation temperature or type of heterogamety

1 Extinct and Extant Reptiles 5

The timing of the effect of temperature on sex-determining response also varies

among TSD reptiles. Shine et al. (2007) tested two TSD lizards for the effects of

fadrozole, a chemical that blocks the bioconversion of testosterone to estrogen,

thereby causing male development in eggs incubated at female-producing tempera-

tures. In this type of experiment, the stage during which fadrozole affects offspring

sex ratios represents the thermally sensitive period when temperature can influence

sex determination. In two TSD reptiles, jacky dragons (Amphibolurus muricatus)and Duperrey’s window-eyed skinks (Bassiana duperreyi), the thermally sensitive

period in which sex could be reversed by fadrozole treatment occurred in the first

half of the postoviposition incubation period. The thermally sensitive period has

been shown to occur slightly later in turtles and tuataras, during only the middle

third of the postoviposition incubation period (Ewert et al. 2004; Mitchell et al.

2006) and occurs even later in crocodilians, during the third quarter of the entire

incubatory period (Lang and Andrews 1994).

GSD amniotes exhibit a similar degree of variability (Organ and Janes 2008). In

birds, snakes, and some turtles and lizards, females are the heterogametic sex. Male

heterogamety is found in some turtles and lizards and throughout mammals (with

exceptions). The mammalian exceptions include, among others, the mole vole

(Ellobius lutescens) in which a Y sex chromosome is absent. Both males and

females of this species carry one X sex chromosome (Just et al. 1995; Vogel

et al. 1998). Within heterogameties, there is variation in the extent of degeneration

of either the male-specific Y sex chromosome or the female-specific W sex chromo-

some. For example, the Z and W sex chromosomes of emus (Dromaius novaehol-landiae) are virtually homomorphic, whereas in chickens (Gallus gallus), the W sex

chromosome is considerably smaller than the Z sex chromosome (Janes et al. 2009;

Solari 1994). Clearly, a single line of demarcation between genotypic and tempera-

ture-dependent sex determination is overly simplistic and does not accurately repre-

sent the evolutionary history of sex-determining mechanisms in amniotes (Sarre et al.

2004).

1.2.2 Adaptive Significance of Sex-Determining Mechanisms

The variability of reptilian sex-determining mechanisms and, among GSD species,

type of heterogamety are difficult to explain. Among agamid lizards, for example,

species within the same genus with no discernible differences in natural history

exhibit different sex-determining mechanisms (Ezaz et al. 2009; Uller et al. 2006).

However, the adaptive significance of both genotypic and temperature-dependent

sex determination has been explored in theory and experimentation. Fisher (1930)

argued that parents should invest equally in sons and daughters. If sons and

daughters represent equivalent parental investment, genotypic sex determination

is expected to balance offspring sex ratios by matching them to the balanced

6 D.E. Janes

probability of inheriting an X or a Y chromosome from a male parent in a male

heterogametic species or the probability of inheriting a Z or a W chromosome from

a female parent in a female heterogametic species. Charnov and Bull (1977)

hypothesized that temperature-dependent sex determination would allow parents

greater control over offspring sex ratios in environments where the costs of sons and

daughters are unequal and fluctuating. However, the Charnov–Bull hypothesis has

not acquired much empirical support. Parents of TSD species do not appear to

control offspring sex ratios by nesting behavior. However, Freedberg and Wade

(2001) suggested that offspring sex ratios are inherited as nest sites, and their

unique exposures to sun and soil temperature are passed matrilineally. Also,

Warner and Shine (2008) demonstrated that incubation temperature can affect

reproductive success in jacky dragons. Male jacky dragons hatched from eggs

incubated at the optimal male-producing temperature had greater lifetime repro-

ductive success than males hatched from eggs incubated at a different temperature

and experimentally masculinized by chemical aromatase inhibition. The same

pattern of greater reproductive success was reported among females incubated at

either the optimal female-producing temperature or a different temperature. This

study provides evidence that, in a TSD species, incubation temperature directly

influences reproductive success in a sex-differential manner. Although this study

supports the Charnov–Bull hypothesis, it does not explain why some species would

benefit from temperature-dependent sex determination but not other closely related

species with similar life history traits.

Reproductive mode, whether a species is oviparous (egg-laying) or viviparous

(live-bearing), is associated with type of sex-determining mechanism. Viviparity

appears to be enabled by genotypic but not temperature-dependent sex determi-

nation. From a sample of 94 extant amniote species for which sex-determining

mechanism, reproductive mode, and phylogenetic position are known, only two,

perhaps three, exhibit both temperature-dependent sex determination and vivi-

parity. The southern water skink (Eulamprus tympanum) and its sister species

(Eulamprus heatwolei) give live birth and exhibit temperature-dependent sex

determination and some evidence suggests that the spotted skink (Niveoscincusocellatus) is also TSD and viviparous (Organ et al. 2009). For TSD species

including these skinks, producing both male and female offspring requires expos-

ing different embryos to one of at least two (optimal male-producing and optimal

female-producing) thermal environments. For viviparous species, this require-

ment entails manipulating maternal body temperature and evidence for maternal

manipulation of body temperature in TSD, viviparous skinks is debated (Allsop

et al. 2006; While and Wapstra 2009). Further, as explained in Sect. 1.4, fluctua-

tions in maternal body temperatures are even less likely in thermally consistent

environments such as deep oceans. Apparently, thermal consistency is not an

issue for oviparous, TSD species such as crocodilians and sea turtles because their

nests experience sufficient thermal variation from top to bottom to explain mixed

sex ratios emerging from clutches of eggs (Georges 1992 but see Warner and

Shine 2009).


1.2.3 Genotype and Environment Interaction

The proximate differences among sex-determining mechanisms remain unclear.

Controlled incubation studies in the laboratory have been used to identify species in

which incubation temperatures may or may not skew offspring sex ratios. These

incubation experiments that measure offspring sex ratios are challenged by the

possibility that a specific temperature that elicits a sex-determining response goes

inadvertently untested. Further, in a tested species, the difference between a tem-

perature that yields a consistent offspring sex ratio and a temperature that yields

lethality may be too small to tease them apart in incubation studies. In the face of

such uncertainty, many experimental characterizations of sex-determining mechan-

isms are considered tentative (Viets et al. 1994).

In addition to results from incubation studies, GSD and TSD species can be

distinguished by the presence or absence of sex chromosomes. If a species has

detectable sex chromosomes, then offspring sex ratios are expected to be defined by

genotype. However, an exception to this rule has been presented by a study of

central bearded dragons (Pogona vitticeps) (Quinn et al. 2007). Central bearded

dragons exhibit clear female heterogamety, yet extreme incubation temperatures

can feminize genotypically male embryos. This result suggests environmental

effects on sex determination in a GSD species. Likewise, genotypic effects have

been reported for leopard geckos (Eublepharis macularius), a reptile that has beenclassified as exhibiting TSD because incubation studies of leopard geckos demon-

strate a clear and repeatable influence of incubation temperature on offspring sex

ratios (Janes et al. 2007; Viets et al. 1993; Wagner 1980). Nonetheless, a quantita-

tive genetic effect on temperature-dependent sex determination is clear from study

of sex-determining response to incubation temperature in different matrilineal lines

of leopard geckos. Janes and Wayne (2006) identified genetically dissimilar

females within a captive-bred colony of leopard geckos. These females were each

mated to fertile males and the resultant offspring were placed randomly within one

of three environmental chambers set to temperatures known to produce either 0%,

�50%, or�70%male offspring. In this species, a 100%male-producing incubation

temperature has not been identified. Although incubation temperature overwhelm-

ingly influenced offspring sex ratios across family lines, a genotype � environment

interaction was detected in the varying offspring sex ratios from different matrilin-

eal lines exposed to the same incubation temperatures. This result suggests that

families vary in their sex-determining response to incubation temperature. Geno-

type � environment interactions also indicate that a studied trait is polygenic

(Falconer and MacKay 1996). Polygenic inheritance is relevant to conservation

of TSD reptiles that may be exceptionally vulnerable to climate change because of

the possibility that they are not exposed to temperatures needed to produce both

sons and daughters (Huey and Janzen 2008). If there is an underlying polygenic

control of sex-determining responses to temperature in TSD reptiles, then there is

opportunity for microevolution and adaptation to changing climates. Recent

modeling has suggested that tuataras (Sphenodon guntheri) occupy a habitat in

8 D.E. Janes

which ambient temperature is expected to change to a degree that could negatively

affect offspring sex ratios within the next century (Huey and Janzen 2008). If sex-

determining responses to temperature do not change adaptively, the remaining

possibilities include extinction or migration to cooler habitats but migration is

unlikely without human intervention considering tuataras’ habitat of small islands

off New Zealand.

1.3 Sex Chromosomes

1.3.1 Origins and Degeneration of Sex Chromosomes

Heteromorphic sex chromosomes arise when one of a pair of sex chromosomes

degenerates to a sufficient degree that cytogenetic differences between the pair are

observable. A number of different causes for this degeneration have been proposed,

including the Hill–Robertson effect, background selection, Muller’s Ratchet,

and hitchhiking of deleterious alleles onto favored mutations (Charlesworth and

Charlesworth 2000; Charlesworth et al. 1987). The Hill–Robertson effect prevents

the repair or elimination of deleterious alleles because of their close linkage to

beneficial alleles and background selection explains rates of elimination or fixation

by the degree to which an allele is either deleterious or beneficial. Mildly deleteri-

ous alleles are more likely to be tolerated than more seriously deleterious alleles

(Charlesworth and Charlesworth 2000). If mildly deleterious alleles are permitted

to accumulate on the Y chromosome as a result of reduced repair via recombination

with the X, then, over time, the mean fitness of the Y chromosome declines. The

accumulation of mildly deleterious alleles, known as Muller’s Ratchet, eventually

causes an allele to become damaged and then eliminated. Following that, the

homologous copy becomes fixed at a rate that is much faster than the fixation rate

for genes that are retained as two copies (Rice 1987). Hitchhiking works in

conjunction with Muller’s Ratchet to hasten the degeneration of the Y chromosome.

Deleterious mutations that hitchhike with favorable alleles on the Y are less likely

to be purged, further reducing the overall fitness of the chromosome. These forces

drive the degeneration of sex chromosomes after an initial event that converts an

ancestral pair of autosomes into sex chromosomes.

Ohno (1967) described the origination of sex chromosomes from ancestral

autosomes. Once a novel sex-determining gene is either exapted from a different

function or transposed to a chromosome from elsewhere in the genome, recombi-

nation ceases in the general vicinity of the gene. This block to recombination

allows parents to pass the sex-determining gene to either sons or daughters,

depending on the nature of the expression of the sex-determining gene. In

mammals, a single-copy gene called the sex-determining region on the Y (Sry)initiates male sexual development (Sinclair et al. 1990). Cessation of recombina-

tion around the Sry or some other ancestral sex-determining gene speeds up


Muller’s Ratchet, causing the degeneration of the mammalian Y chromosome.

The evolution of avian sex chromosomes may have followed a different path. In

chickens, dosage-dependent effects of a Z-linked gene, Dmrt1, appear to drive

male sexual development rather than the absence of a single copy of a W-linked

gene (Smith et al. 2009).

Reptiles provide an excellent model for the process of sex chromosome degen-

eration because of the intermediate stages of chromosomal degeneration found in

the group. For example, the smooth softshell turtle (Apalone mutica) is GSD but sex

chromosomes have not yet been identified, most likely due to a lack of sufficient

heteromorphy (Valenzuela et al. 2006). Further, micro-sex chromosomes have been

found in central bearded dragons (Pogona vitticeps), common snake-necked turtles

(Chelodina longicollis), and Chinese soft-shelled turtles (Pelodiscus sinensis)(Ezaz et al. 2005, 2006; Kawai et al. 2007). The variety of sex chromosome

organizations has been mapped onto phylogenetic trees to investigate the number

of origins of sex chromosomes and types of heterogameties in the group (Janzen

and Krenz 2004; Pokorna and Kratochvil 2009). Parsimony, likelihood, Bayesian,

and stochastic approaches reconstruct temperature-dependent sex determination as

ancestral to archosaurs (turtles, crocodilians, and birds) (Organ and Janes 2008).

Turtles are extraordinarily variable in their organizations of sex chromosomes with

species exhibiting male heterogamety, female heterogamety, no detectable hetero-

gamety, or temperature-dependent sex determination (Organ and Janes 2008).

These results indicate multiple independent origins of sex chromosomes among

archosaurs (Fig. 1.2). Also, Matsubara et al. (2006) demonstrated a lack of sequence

similarity between the female heterogametic sex chromosomes of birds and those of

snakes, indicating at least two independent origins of sex chromosomes. Reptiles,

with such variability and rapidly improving genomic resources, provide tremen-

dous raw material for studies of the causes and consequences of sex chromosome

origination and degeneration.

1.3.2 Detection of Sex Chromosomes

Species for which genotypic sex determination has been ascribed but sex chromo-

somes have not yet been identified are an important focus of research on reptile

genomics (Janes et al. 2010a). For species like the smooth softshell turtle, sex

chromosomes have not been reported but it is unclear if this is because they are

lacking in this species or if current cytogenetic techniques are not yet sufficiently

sensitive to detect them. The cytogenetic technique of C-banding, which stains

the heterochromatic regions of chromosomes, has identified female-specific W sex

chromosomes in central bearded dragons (P. vitticeps) (Ezaz et al. 2005) as well aseastern bearded dragons (Pogona barbata), Nobbi dragons (Amphibolurus nobbi),and Mallee dragons (Ctenophorus fordi) (Ezaz et al. 2009). Comparative genomic

hybridization, Ag–NOR staining, and fluorescent in situ hybridization (FISH)

are also standard techniques for identifying karyotypic sex differences (Kawai

10 D.E. Janes

et al. 2007). As more sex chromosomes are identified, more sex-linked sequences

will be cataloged for reptile species. For example, 18 S–28 S ribosomal RNA genes

are located on both micro-sex chromosomes in the Chinese soft-shelled turtle but in

more copies on the W chromosome than on the Z chromosome (Kawai et al. 2007).

Comparative FISH mapping of sex-linked markers will be useful for supporting or

rejecting hypotheses regarding the evolutionary history of sex-determining

mechanisms. Clearly, snake and bird sex chromosomes have little or no sequence

in common but the similarities and differences of sex chromosomes among birds,

turtles, and possibly TSD reptiles have not yet been characterized (Fig. 1.2) (Janes

et al. 2010b). However, Kawagoshi et al. (2009) identified five Z-linked markers in

the Chinese soft-shelled turtle by FISH mapping cDNA fragments of the genes

GIT2, NF2, SBNO1, SF3A1, and TOP3B. These markers map to chicken chromo-

some 15, suggesting a common origin.

Am

phib

ians

Tur

tles

Cro

codi

lians

Bird

s

Igua

nids

Sna

kes

Lace

rtid

liza

rds

Ski

nks

Gec

kos

Tua

tara

Mam

mal

s

F M

F M

F M

F M

F MF M

F M

F M

0 Mya

100 Mya

200 Mya

300 Mya

F M

F MF M

Fig. 1.2 Presence or absence of male or female heterogamety across amphibians, nonavian and

avian reptiles, and mammals (Organ and Janes 2008). Sex chromosomes have not been reported

for crocodilians or tuataras, both exhibiting temperature-dependent sex determination. Female

heterogamety is exhibited by snakes but is shaded differently in this figure to indicate that snake

sex chromosomes do not share sequence with avian sex chromosomes as the two pairs of sex

chromosomes most likely resulted from independent origins of female heterogamety (Matsubara

et al. 2006). The characterization of similarities or differences between avian sex chromosomes

and female heterogameties found in other reptiles and the estimation of the number of independent

origins of sex chromosomes are focuses of reptilian genomics research (Janes et al. 2010a)


1.3.3 Heterogamety and Dosage Compensation

Hypotheses are emerging about the differences between male and female hetero-

gamety. For example, dosage compensation appears to function differently between

male heterogametic and female heterogametic species. Genes found on the

X chromosome in male heterogametic species and on the Z chromosome in female

heterogametic species occur in different doses between males and females.

Mammals balance gene dosage by inactivating an X chromosome. X-chromosome

inactivation transcriptionally silences genes on one of two X chromosomes in a

female, thereby balancing gene dosage between males and females (Payer and Lee

2008). Birds, however, do not globally inactivate a Z chromosome in males. Rather,

dosage compensation appears to act rarely and on small regions of avian sex

chromosomes (Melamed and Arnold 2007). In fact, global dosage compensation

has only been found in male heterogametic groups, including therian mammals,

fruitflies (Drosophila), and nematodes (Caenorhabditis elegans), whereas local

dosage compensation has been found in female heterogametic groups, including

birds and lepidopterans (Mank 2009). At present, the pattern has only been

described among three male heterogametic groups and two female heterogametic

groups and has yet to be explored among reptiles (but see King and Lawson 1996).

Inactivation or hyper-transcription of sex-linked genes and entire chromosomes

should be compared between closely related male heterogametic and female

heterogametic reptiles, particularly among emydid turtles, chameleons, and geckos

that exhibit differences in heterogamety within families (Organ and Janes 2008).

1.4 Fossil Evidence

Extinct reptiles are relevant to the study of sex chromosome evolution because of

the order in which genotypic sex determination and sex chromosomes evolve. Sex

chromosomes become detectable only after they have been sufficiently affected by

evolutionary forces that arise subsequent to the block to recombination caused by

either the novel function or novel location of a sex-determining gene. Fossils of

extinct reptiles allow us to examine the history of sex-determining mechanisms and

subsequently predict which extinct reptiles exhibited genotypic sex determination.

Organ et al. (2009) used a reversible-jump Markov-chain Monte Carlo algorithm to

establish a Bayesian posterior probability distribution for models of correlated

change between different types of sex-determining mechanisms and reproductive

modes in extant amniotes (see Sect. 1.2.2). Reproductive mode describes the means

by which parents produce young. Among amniotes, species are either viviparous or

oviparous. The Bayesian analysis yielded a significant result for correlated evolu-

tion of genotypic sex determination and viviparity. Oviparity does not effectively

predict a certain sex-determining mechanism but viviparity predicts genotypic sex

determination. As described above, only two, perhaps three, of 94 studied extant

12 D.E. Janes

amniotes are both viviparous and TSD. This correlation permitted a prediction of

genotypic sex determination in extinct species known to be viviparous. In fact,

fossil evidence demonstrates viviparity in several extinct marine reptiles, including

sauropterygians, mosasaurs, and ichthyosaurs. The study predicted sex-determining

mechanisms for seven species for which sex-determining mechanisms were known

but not introduced to the algorithm. This test group included six extant reptiles and

an extinct horse (Propalaeotherium) for which pregnant specimens have been

found in the fossil record. The study showed that genotypic sex determination

could be accurately predicted for viviparous species. All ten marine reptiles exam-

ined in the study were assigned a significant posterior probability of having

genotypic sex determination.

Organ et al. (2009) argued that this result is meaningful for the natural history of

extinct marine reptiles. Oviparity in the open ocean would not have been possible

for amniote species like ichthyosaurs because amniotic eggs require gas-exchange

with the atmosphere (Andrews and Mathies 2000). Extant marine reptiles including

saltwater crocodiles (Crocodylus porosus) and sea turtles nest on land but extinct

marine reptiles like ichthyosaurs did not have a body plan that was likely to allow

terrestrial nesting. Freed by viviparity from the requirement to nest on land, extinct

marine reptiles evolved morphologies that were adaptive to pelagic existence.

These morphologies included fluked tails, dorsal fins, and wing-shaped limbs.

Further, if prerequisite for the evolution of viviparity, genotypic sex determination

may have permitted the adaptive radiation of extinct marine reptiles since viviparity

seems to be a prerequisite for the pelagic existence of those species (Caldwell and

Lee 2001).

1.5 Impact of Genome Projects and Future Directions

The study of sex chromosome evolution has much to gain from current genome

sequencing efforts. At present, only the green anole (Anolis carolinensis) and the

painted turtle (Chrysemys picta) are focuses of genome sequencing projects (Janes

et al. 2008) but the recently announced Genome 10K collection of species that has

been targeted for whole-genome sequencing includes 3,297 nonavian reptiles

(Haussler et al. 2009). In particular, the genome sequences of 140 turtles, 569

iguanids, and 621 geckos that have been targeted for genome sequencing will

provide a window into the variability of sex-determining mechanisms and sex

chromosome organizations found in these three groups. The identities and map

locations of sex-linked markers will support or reject current hypotheses of com-

mon origins of sex chromosomes. For example, Kawai et al. (2009) suggested a

common origin between the sex chromosome pairs of the gecko lizard (Gekkohokouensis) and chicken because they share a linkage group that consists of six

markers. Following the publication of multiple reptile genomes, studies of this kind

will involve more markers in more species, allowing more robust conclusions to be

made regarding the number of independent origins of reptilian sex chromosomes.


Until the sequencing and mapping of sex-linked and sex-differentiating markers

have reached a more advanced stage, studies of reptilian sex chromosomes will be

smaller in scope. Nonetheless, sex-linked markers have been identified in birds

(Backstrom et al. 2006; Hillier et al. 2004), snakes (Matsubara et al. 2006), turtles

(Kawagoshi et al. 2009), and lizards (Kawai et al. 2009). These sequences provide

sufficient raw material for mapping comparisons among pairs of reptilian sex

chromosomes. Comparative mapping studies, in concert with ancestral reconstruc-

tions, will directly inform questions regarding the number of independent origins

of sex chromosomes in reptiles and why sex chromosome systems have higher

turnover in nonavian reptiles than they have in either birds or mammals.

Acknowledgments I would like to thank Miguel Alcaide, Maude Baldwin, Elena Gonzalez, June

Yong Lee, Christopher Organ, and Irene Salicini for their critical reviews of this chapter. This

work has benefited from conversations with Nicole Valenzuela (NV), Scott V. Edwards (SVE),

Tariq Ezaz, Jennifer A.M. Graves, Arthur Georges, and Andrew Sinclair. Support in the laboratory

and valuable discussions were shared by Christopher Balakrishnan, Charles Chapus, and Andrew

Shedlock. Funding for this work was provided by a grant from the United States National Science

Foundation (MCB0817687) to NV and SVE. Last, I would like to thank Pierre Pontarotti for the

invitation to contribute to the 13th Evolutionary Biology Meeting at Marseille where this work was

presented.

References

Allsop DJ, Warner DA, Langkilde T, Du W, Shine R (2006) Do operational sex ratios influence

sex allocation in viviparous lizards with temperature-dependent sex determination? J Evol Biol

19(4):1175–1182

Andrews RM, Mathies T (2000) Natural history of reptilian development: constraints on the

evolution of viviparity. Bioscience 50(3):227–238

Backstrom N, BrandstromM, Gustafsson L, Qvarnstrom A, Cheng H, Ellegren H (2006) Genetic

mapping in a natural population of collared flycatchers (Ficedula albicollis): conservedsynteny but gene order rearrangements on the avian Z chromosome. Genetics 174(1):

377–386

Bull JJ (1983) Evolution of sex determining mechanisms. Benjamin/Cummings, Menlo Park, CA

Caldwell MW, Lee MSY (2001) Live birth in Cretaceous marine lizards (mosasauroids). Proc R

Soc Lond B Biol Sci 268(1484):2397–2401

Charlesworth B, Charlesworth D (2000) The degeneration of Y chromosomes. Phil Trans Roy Soc

Lond B 355(1403):1563–1572

Charlesworth B, Coyne JA, Barton NH (1987) The relative rates of evolution of sex chromosomes

and autosomes. Am Nat 130(1):113–146

Charnov EL, Bull J (1977) When is sex environmentally determined. Nature 266(5605):829–830

Ewert BJ, Etchberger CR, Nelson CE (2004) Turtle sex-determining modes and TSD patterns, and

some TSD pattern correlates. In: Valenzuela N, Lance VA (eds) Temperature-dependent sex

determination in vertebrates. Smithsonian Books, Washington, DC, pp 21–32

Ezaz T, Quinn AE, Miura I, Sarre SD, Georges A, Graves JAM (2005) The dragon lizard Pogonavitticeps has ZZ/ZW micro-sex chromosomes. Chromosome Res 13(8):763–776

Ezaz T, Valenzuela N, Grutzner F, Miura I, Georges A, Burke RL, Graves JAM (2006) An XX/XY

sex microchromosome system in a freshwater turtle, Chelodina longicollis (Testudines:

Chelidae) with genetic sex determination. Chromosome Res 14(2):139–150

14 D.E. Janes

Ezaz T, Quinn AE, Sarre SD, O’Meally D, Georges A, Graves JAM (2009) Molecular marker

suggests rapid changes of sex-determining mechanisms in Australian dragon lizards. Chromo-

some Res 17(1):91–98

Falconer DS, MacKay TFC (1996) Introduction to quantitative genetics. Longmann Press,

London, UK

Fisher RA (1930) The genetical theory of natural selection. Oxford University Press, New York,

USA

Freedberg S, Wade MJ (2001) Cultural inheritance as a mechanism for population sex-ratio bias

in reptiles. Evolution 55(5):1049–1055

Georges A (1992) Thermal characteristics and sex determination in field nests of the pig-nosed

turtle, Carettochelys insculpta (Chelonia, Carettochelydidae), from northern Australia. Aust

J Zool 40(5):511–521

Haussler D, O’Brien SJ, Ryder OA, Barker FK, Clamp M, Crawford AJ, Hanner R, Hanotte O,

Johnson WE, McGuire JA, Miller W, Murphy RW, Murphy WJ, Sheldon FH, Sinervo B,

Venkatesh B, Wiley EO, Allendorf FW, Amato G, Baker CS, Bauer A, Beja-Pereira A,

Bermingham E, Bernardi G, Bonvicino CR, Brenner S, Burke T, Cracraft J, Diekhans M,

Edwards S, Ericson PGP, Estes J, Fjelsda J, Flesness N, Gamble T, Gaubert P, Graphodatsky

AS, Graves JAM, Green ED, Green RE, Hackett S, Hebert P, Helgen KM, Joseph L, Kessing B,

Kingsley DM, Lewin HA, Luikart G, Martelli P, Moreira MAM, Nguyen N, Orti G, Pike BL,

Rawson DM, Schuster SC, Seuanez HN, Shaffer HB, Springer MS, Stuart JM, Sumner J,

Teeling E, Vrijenhoek RC, Ward RD, Warren WC, Wayne R, Williams TM, Wolfe ND,

Zhang YP (2009) Genome 10K: a proposal to obtain whole-genome sequence for 10 000

vertebrate species. J Hered 100(6):659–674

Hillier LW, Miller W, Birney E, Warren W, Hardison RC, Ponting CP, Bork P, Burt DW, Groenen

MAM, Delany ME, Dodgson JB, Chinwalla AT, Cliften PF, Clifton SW, Delehaunty KD,

Fronick C, Fulton RS, Graves TA, Kremitzki C, Layman D, Magrini V, McPherson JD, Miner

TL, Minx P, NashWE, Nhan MN, Nelson JO, Oddy LG, Pohl CS, Randall-Maher J, Smith SM,

Wallis JW, Yang SP, Romanov MN, Rondelli CM, Paton B, Smith J, Morrice D, Daniels L,

Tempest HG, Robertson L, Masabanda JS, Griffin DK, Vignal A, Fillon V, Jacobbson L,

Kerje S, Andersson L, Crooijmans RPM, Aerts J, van der Poel JJ, Ellegren H, Caldwell RB,

Hubbard SJ, Grafham DV, Kierzek AM, McLaren SR, Overton IM, Arakawa H, Beattie KJ,

Bezzubov Y, Boardman PE, Bonfield JK, CroningMDR, Davies RM, Francis MD, Humphray SJ,

Scott CE, Taylor RG, Tickle C, Brown WRA, Rogers J, Buerstedde JM, Wilson SA, Stubbs L,

Ovcharenko I, Gordon L, Lucas S, Miller MM, Inoko H, Shiina T, Kaufman J, Salomonsen J,

Skjoedt K, Wong GKS, Wang J, Liu B, Yu J, Yang HM, Nefedov M, Koriabine M, deJong PJ,

Goodstadt L, Webber C, Dickens NJ, Letunic I, Suyama M, Torrents D, von Mering C,

Zdobnov EM, Makova K, Nekrutenko A, Elnitski L, Eswara P, King DC, Yang S, Tyekucheva

S, Radakrishnan A, Harris RS, Chiaromonte F, Taylor J, He JB, Rijnkels M, Griffiths-Jones S,

Ureta-Vidal A, HoffmanMM, Severin J, Searle SMJ, Law AS, Speed D,Waddington D, Cheng Z,

Tuzun E, Eichler E, Bao ZR, Flicek P, Shteynberg DD, Brent MR, Bye JM, Huckle EJ,

Chatterji S, Dewey C, Pachter L, Kouranov A, Mourelatos Z, Hatzigeorgiou AG, Paterson

AH, Ivarie R, Brandstrom M, Axelsson E, Backstrom N, Berlin S, Webster MT, Pourquie O,

Reymond A, Ucla C, Antonarakis SE, Long MY, Emerson JJ, Betran E, Dupanloup I,

Kaessmann H, Hinrichs AS, Bejerano G, Furey TS, Harte RA, Raney B, Siepel A, Kent WJ,

Haussler D, Eyras E, Castelo R, Abril JF, Castellano S, Camara F, Parra G, Guigo R, Bourque

G, Tesler G, Pevzner PA, Smit A, Fulton LA, Mardis ER, Wilson RK (2004) Sequence and

comparative analysis of the chicken genome provide unique perspectives on vertebrate evolu-

tion. Nature 432(7018):695–716

Huey RB, Janzen FJ (2008) Climate warming and environmental sex determination in tuatara: the

last of the Sphenodontians? Proc R Soc Lond B Biol Sci 275(1648):2181–2183

Janes DE, Wayne ML (2006) Evidence for a genotype � environment interaction in sex-deter-

mining response to incubation temperature in the leopard gecko, Eublepharis macularius.Herpetologica 62(1):56–62


Janes DE, Bermudez D, Guillette LJ, Wayne ML (2007) Estrogens induced male production at a

female-producing temperature in a reptile (Leopard Gecko, Eublepharis macularius) with

temperature-dependent sex determination. J Herpetol 41(1):9–15

Janes DE, Organ C, Valenzuela N (2008) New resources inform study of genome size, content, and

organization in nonavian reptiles. Integr Comp Biol 48(4):447–453

Janes DE, Ezaz T, Graves JAM, Edwards SV (2009) Recombination and nucleotide diversity in

the sex chromosomal pseudoautosomal region of the emu, Dromaius novaehollandiae. J Hered100(2):125–136

Janes DE, Fujita MK, Organ CL, Shedlock AM, Edwards SV (2010a) Genome evolution in

Reptilia, the sister group of mammals. Annu Rev Genom Hum Genet (in press)

Janes DE, Organ CL, Edwards SV (2010b) Variability in sex-determining mechanisms influences

genome complexity in Reptilia. Cytogenet Genome Res 127(2–4):242–248

Janzen FJ, Krenz JG (2004) Phylogenetics: which was first, TSD or GSD? In: Valenzuela N,

Lance VA (eds) Temperature-dependent sex determination in vertebrates. Smithsonian Books,

Washington, DC, pp 121–130

Just W, Rau W, Vogel W, Akhverdian M, Fredga K, Graves JAM, Lyapunova E (1995) Absence

of Sry in species of the vole Ellobius. Nat Genet 11(2):117–118Kawagoshi T, Uno Y, Matsubara K, Matsuda Y, Nishida C (2009) The ZW micro-sex chromo-

somes of the chinese soft-shelled turtle (Pelodiscus sinensis, Trionychidae, Testudines) havethe same origin as chicken chromosome 15. Cytogenet Genome Res 125:125–131

Kawai A, Nishida-Umehara C, Ishijima J, Tsuda Y, Ota H, Matsuda Y (2007) Different origins of

bird and reptile sex chromosomes inferred from comparative mapping of chicken Z-linked

genes. Cytogenet Genome Res 117(1–4):92–102

Kawai A, Ishijima J, Nishida C, Kosaka A, Ota H, Kohno S, Matsuda Y (2009) The ZW sex

chromosomes of Gekko hokouensis (Gekkonidae, Squamata) represent highly conserved

homology with those of avian species. Chromosoma 118(1):43–51

King RB, Lawson R (1996) Sex-linked inheritance of fumarate hydratase alleles in natricine

snakes. J Hered 87:81–83

Lang JW, Andrews HV (1994) Temperature-dependent sex determination in crocodilians. J Exp

Zool 270(1):28–44

Mank JE (2009) The W, X, Y and Z of sex-chromosome dosage compensation. Trends Genet

25(5):226–233

Matsubara K, Tarui H, Toriba M, Yamada K, Nishida-Umehara C, Agata K, Matsuda Y (2006)

Evidence for different origin of sex chromosomes in snakes, birds, and mammals and step-wise

differentiation of snake sex chromosomes. Proc Natl Acad Sci USA 103(48):18190–18195

Melamed E, Arnold AP (2007) Regional differences in dosage compensation on the chicken Z

chromosome. Genome Biol 8(9):R202

Mitchell NJ, Nelson NJ, Cree A, Pledger S, Keall SN, Daugherty CH (2006) Support for a rare

pattern of temperature-dependent sex determination in archaic reptiles: evidence from two

species of tuatara (Sphenodon). Front Zool 3:9Ohno S (1967) Sex chromosomes and sex linked genes. Springer, Berlin

Organ CL, Janes DE (2008) Evolution of sex chromosomes in Sauropsida. Integr Comp Biol 48

(4):512–519

Organ CL, Janes DE, Meade A, Pagel M (2009) Genotypic sex determination enabled adaptive

radiations of extinct marine reptiles. Nature 461(7262):389–392

Payer B, Lee JT (2008) X chromosome dosage compensation: how mammals keep the balance.

Annu Rev Genet 42:733–772

Pokorna M, Kratochvil L (2009) Phylogeny of sex-determining mechanisms in squamate reptiles:

are sex chromosomes an evolutionary trap? Zool J Linn Soc 156(1):168–183

Quinn AE, Georges A, Sarre SD, Guarino F, Ezaz T, Graves JAM (2007) Temperature sex reversal

implies sex gene dosage in a reptile. Science 316(5823):411

Rice WR (1987) Genetic hitchhiking and the evolution of reduced genetic activity of the Y sex

chromosome. Genetics 116(1):161–167

16 D.E. Janes

Sarre SD, Georges A, Quinn A (2004) The ends of a continuum: genetic and temperature-

dependent sex determination in reptiles. Bioessays 26(6):639–645

Shine R, Warner DA, Radder R (2007) Windows of embryonic sexual lability in two lizard species

with environmental sex determination. Ecology 88(7):1781–1788

Sinclair AH, Berta P, Palmer MS, Hawkins JR, Griffiths BL, Smith MJ, Foster JW, Frischauf AM,

Lovell-badge R, Goodfellow PN (1990) A gene from the human sex-determining region

encodes a protein with homology to a conserved DNA-binding motif. Nature 346(6281):

240–244

Smith CA, Roeszler KN, Ohnesorg T, Cummins DM, Fairlie PG, Doran TJ, Sinclair AH (2009)

The avian Z-linked gene DMRT1 is required for male sex determination in the chicken. Nature

461:267–271

Solari AJ (1994) Sex chromosomes and sex determination in vertebrates. CRC Press, Boca

Raton, FL

Standora EA, Spotila JR (1985) Temperature-dependent sex determination in sea turtles. Copeia

3:711–722

Uller T, Mott B, Odierna G, Olsson M (2006) Consistent sex ratio bias of individual female dragon

lizards. Biol Lett 2(4):569–572

Valenzuela N (2004) Introduction. In: Valenzuela N, Lance VA (eds) Temperature-dependent sex

determination in vertebrates. Smithsonian Books, Washington, DC, pp 1–4

Valenzuela N, LeClere A, Shikano T (2006) Comparative gene expression of steroidogenic factor

1 in Chrysemys picta and Apalone mutica turtles with temperature-dependent and genotypic

sex determination. Evol Dev 8(5):424–432

Viets BE, Tousignant A, Ewert MA, Nelson CE, Crews D (1993) Temperature-dependent sex

determination in the leopard gecko, Eublepharis macularius. J Exp Zool 265(6):679–683

Viets BE, Ewert MA, Talent LG, Nelson CE (1994) Sex-determining mechanisms in squamate

reptiles. J Exp Zool 270(1):45–56

Vogel W, Jainta S, Rau W, Geerkens C, Baumstark A, Correa-Cerro LS, Ebenhoch C, Just W

(1998) Sex determination in Ellobius lutescens: the story of an enigma. Cytogenet Cell Genet

80(1–4):214–221

Wagner E (1980) Temperature-dependent sex determination in a gekko lizard. Q Rev Biol 55:21,

appendix

Warner DA, Shine R (2008) The adaptive significance of temperature-dependent sex determina-

tion in a reptile. Nature 451(7178):566–568

Warner DA, Shine R (2009) Maternal and environmental effects on offspring phenotypes in an

oviparous lizard: do field data corroborate laboratory data? Oecologia 161(1):209–220

While GM, Wapstra E (2009) Snow skinks (Niveoscincus ocellatus) do not shift their sex

allocation patterns in response to mating history. Behaviour 146:1405–1422


Chapter 2

Constraints, Plasticity, and Universal Patterns

in Genome and Phenome Evolution

Eugene V. Koonin and Yuri I. Wolf

Abstract Evolutionary genomics identifies multiple constraints that differentially

affect different parts of the genomes of diverse life forms. The selective pressures

that shape the evolution of viral, prokaryotic, and eukaryotic genomes differ

dramatically, and substantial differences exist even between animal and bacterial

lineages. Constraints on protein evolution appear to be more universal and could be

determined by the fundamental physics of protein folding. Some key features of the

molecular phenome such as protein abundance turn out to be unexpectedly con-

served and hence strongly constrained. The constraints that shape the evolution

of genomes and phenomes are complemented by the plasticity and robustness

of genome architecture, expression, and regulation. Several universal “laws” of

genome and phenome evolution were detected, some of which seem to be dictated

by selective constraints and others by neutral process.

2.1 Introduction

In principle, the entire genome of any life form can be perceived as evolving under

constraints (purifying selection) the strength of which varies from 0 (unconstrained

evolution) to 1 (absolute conservation). Moreover, constraints affect evolution at all

levels of biological organization, from genome sequence to genome architecture to

gene expression to molecular interactions to actual organismal phenotypes (Kimura

1983; Lynch 2007c). Generally, constraints on the rates and paths of evolution can

be divided into genomic, those that are manifest at the level of the genome sequence

and architecture, and phenomic, those that pertain to phenotypic characteristics

(although ultimately realized through genomic changes as well). Comparative

E.V. Koonin and Y.I. Wolf

National Center for Biotechnology Information, National Library of Medicine, National Institutes

of Health, Bethesda, MD 20892, USA



19

genomics and systems biology produce massive amounts of diverse data that

provide for previously inconceivable insights into the patterns and processes of

genome and phenome evolution (Kitano 2002; Medina 2005; Koonin and Wolf

2006; Lynch 2007c; Loewe 2009; Yamada and Bork 2009).

Comparative genomics allows us, at least in principle, to measure the strength of

constraints that affect different classes of sites in genomes and to elucidate the

biological nature of these constraints. However, genome comparison does more

than that as it gives us material to address evolutionary constraints beyond the

traditional aspect of sequence conservation to higher level questions such as: how

constrained in evolution are gene repertoires of organisms, genome architecture,

evolution rate itself, and more? The massive influx of data from systems biology

takes the study of evolutionary constraints into new dimensions by allowing

researchers to ask qualitatively new questions: what are the nature and strength of

constraints that affect gene expression, regulatory, and interaction networks, meta-

bolic fluxes and other characteristics of organisms that can be denoted “molecular

phenome”?

In this article, we present a broad overview of the constraints that affect gene

sequences, genome architectures, and molecular phenotypic characteristics such as

gene expression level and the structures of protein–protein interaction and regu-

latory networks. We attempt a genome-wide and organism-wide assessment of

different types of constraints operative at different levels and additionally discuss

the concepts of robustness and plasticity that are intimately linked to constraints.

Of course, the subject we address is vast and cannot be reasonably covered in full

in one, relatively brief review. We leave out some important areas such as deve-

lopmental constraints and only fleetingly touch upon others such as evolution of

regulatory networks. Nevertheless, it is our hope that even such sketchy discussion

reveals some important general aspects of constraints that define evolution at

diverse levels of biological organization.

2.2 Evolutionary Constraints on Sequence Evolution

Across Genomes and Taxa

The origins and characteristic strengths of constraints that affect different classes of

sequences in genomes of different life forms are extremely diverse and certainly are

not yet known in full. Typically, the constraints on sequences encoding proteins and

structural RNAs (such as rRNAs and tRNAs) are stronger than the constraints on

noncoding sequences although, for each type of sequences, there is a broad distri-

bution of constraint strengths, and the ranges of the distributions overlap (Shabalina

and Kondrashov 1999; Margulies et al. 2007). Obviously, constraints that affect a

particular class of sites can be measured only by comparison to another class of sites

that can be construed to evolve neutrally. The choice of an appropriate neutral

model is a major problem in molecular evolution. In the pregenomic era, Motoo

20 E.V. Koonin and Y.I. Wolf

Kimura, the founder of the neutral theory, was the first to come up with the simple

but important idea that pseudogenes that are numerous in vertebrates could be used

as a neutral baseline for assessing selection pressure (Kimura 1983). Despite some

exceptional cases of pseudogene recruitment for specific functions (Khachane and

Harrison 2009), in general, this contention still appears to hold true (Harrison and

Gerstein 2002). Genomics revealed additional sources of (apparently) neutrally

evolving sequences such as introns and intergenic regions in animals (Parsch

et al. 2010; Resch et al. 2007). However, a general difficulty with any attempt to

define a universal baseline of neutral evolution is that different parts of a genome

differ in their mutation rates, and consequently, in the rate of neutral evolution for

which the fixation rate equals the mutation rate (Ellegren et al. 2003). Therefore, for

a reliable estimate of the strength of selection/constraints, the neutral model has to

be derived from the same gene/region for which selection is being measured.

Several such measures have been developed (Nielsen 2005; Charlesworth and

Eyre-Walker 2008; Eyre-Walker and Keightley 2009). The most popular gage of

selection pressure for protein-coding sequences naturally follows from the redun-

dancy and nonrandom structure of the genetic code in which the same amino acid

typically is encoded by codons that differ only in their third (or less commonly first)

positions. This measure, Ka/Ks (dN/dS), is the ratio of the number or rate of

nonsynonymous substitutions (those that change an amino acid in the encoded

protein) to the number or rate of synonymous substitutions (those that occur in

synonymous positions of codons and so do not affect the protein sequence) (Hurst

2002; Ellegren 2008). The assumption that underpins the use of Ka/Ks as a measure

of selection is that synonymous sites evolve neutrally or at least under weak

selection compared with nonsynonymous sites, allowing the use of synonymous

sites as the baseline to measure the constraints on protein evolution. As a crude

approximation, this assumption holds as for the great majority of protein-coding

genes from any organism, Ka/Ks << 1 indicating that, taken as a whole, most

proteins are subject to purifying selection of widely differing strength (Fig. 2.1).

Moreover, the distribution of Ka spans a substantially wider range of values than

the distribution of Ks, indicating that the constraints affecting proteins are qualita-

tively different from and much more diverse than those affecting synonymous sites

(Fig. 2.1). For unconstrained, neutral evolution, Ka ¼ Ks as is the case for most

pseudogenes. For a small subset of protein-coding genes, Ka/Ks > 1, which is

construed as evidence of evolution under positive selection. Genes evolving under

positive selection encode specialized proteins for which rapid change is paramount

for function that typically involves “arms race” between competing agencies

such as hosts and parasites; examples include proteins bacterial surface proteins

(Petersen et al. 2007; Muzzi et al. 2008) and proteins involved in mammalian

spermatogenesis, sperm competition, and sperm–egg interaction (Nielsen et al.

2005; Turner et al. 2008). Of course, evolution under positive selection is not

unconstrained as constraints on the overall protein structure still apply (Worth

et al. 2009) but evolution along the available trajectories proceeds rapidly.

The fact that most protein-coding genes evolve under constraints imposed by

purifying selection by no means implies that all amino acid sites are subject to the

2 Constraints, Plasticity, and Universal Patterns in Genome and Phenome Evolution 21

same constraints. On the contrary, the evolutionary rates of sites and by implication

the strength of constraints affecting different sites are well described by a charac-

teristic skewed Gamma distribution (or more precisely a mixture of Gamma

distribution), with a small fraction of sites that are virtually unconstrained or, in

some cases, subject to positive selection and the majority of the sites subject to

broadly distributed constraints (Kelly and Churchill 1996; Grishin et al. 2000;

Mayrose et al. 2005; Nielsen 2005).

The characteristic strengths of constraints that affect evolution of protein-coding

genes widely differ between organisms. Typically, prokaryotic proteins are sub-

ject to stronger constraints than eukaryotic proteins, especially, those of multi-

cellular forms (plants and animals), with the characteristic median Ka/Ks valuesin the range of 0.01–0.1 and 0.1–0.5, respectively (Fig. 2.1) (Jordan et al. 2002;

0.001 0.01 0.1 1 10

dN/dS ratio for orthologs

Human-Macaque

B.cenocepacia-B.vietnamiensis

Aspergillus-Neosartorya

0.0001 0.001 0.01 0.1 1 10

distance between Human and Macaque orthologs

dN

dS

Fig. 2.1 The distributions of evolutionary rates for nonsynonymous and synonymous sites of

protein-coding genes in primates and the Ka/Ks ratios for three diverse pairs of species (Wolf et al.

2009)


Novichkov et al. 2009b). The values of Ka/Ks and by inference the strength of

constraints widely differ between evolutionary lineages such as diverse lineages of

bacteria and archaea, and seem to be related to the specific lifestyles of the

respective organisms (Novichkov et al. 2009b).

The assumption that synonymous sites in protein-coding genes evolve neutrally

is useful for measuring selection acting at the protein level but in itself is a rough

approximation at best. The universally observed, significant positive correlation

between Ka and Ks (Makalowski and Boguski 1998; Drummond and Wilke 2008,

2009; Ellegren 2008) indicates that evolution of synonymous sites is constrained as

well and suggests that the evolutionary forces that shape the evolution of non-

synonymous and synonymous sites are related (see the section on protein evolution

below).

More accurate and powerful tests for purifying and positive selection affecting

different classes of sites are variations of the classic McDonald–Kreitman test

which compares the patterns of substitutions for within species variation (poly-

morphisms) with those for between species divergences, under the assumption that

the fraction of nonneutral polymorphisms is negligible (Nielsen 2001, 2005).

The overall distributions of constraints across genomes are dramatically different

in life forms with distinct genome architectures, in particular, between viruses and

prokaryotes, on the one hand, with their “wall-to-wall” genomes that consist mostly

of protein-coding and RNA-coding genes, and multicellular eukaryotes in whose

genomes the coding nucleotides are in the minority, on the other hand (Lynch and

Conery 2003; Koonin 2009a) (Fig. 2.2). On a per nucleotide basis, the constraints

affecting compact genomes, particularly, those of prokaryotes are orders of magni-

tude greater than the constraints on the larger genomes of multicellular eukaryotes.

Considering the characteristic low Ka/Ks values indicative of strongly constrained

evolution of protein sequences (Fig. 2.1), there are almost no sequences whose

evolution is (effectively) unconstrained in the compact viral and prokaryotic gen-

omes. The notable exception are pseudogenes that are common in some parasitic

bacteria such as Rickettsia or Mycobacterium leprae (Harrison and Gerstein 2002;

Darby et al. 2007; Monot et al. 2009). In typical genomes of free-living prokaryotes

and especially viruses, noncoding regions constitute only 10–15% of the genome,

and a considerable fraction of these sequences consists of regulatory elements

(promoters, operators, terminators, and translation initiation regions) whose evolu-

tion is variably constrained (Molina and van Nimwegen 2008). The genomes of

most viruses are even more compact than prokaryotic genomes, with nearly all of

the genome sequence taken up by protein-coding genes (Koonin 2009a).

Unicellular eukaryotes resemble prokaryotes in their overall genome architec-

ture (notwithstanding important differences such as the absence of operons and the

presence of varying numbers of introns) and show a roughly similar distribution of

evolutionary constraints although the fraction of apparently unconstrained noncod-

ing sequences in these genomes is somewhat greater. However, the genomes of

multicellular eukaryotes (plants and especially animals) present a stark contrast.

These organisms have intron-rich genomes with long intergenic regions, and a

substantial, albeit variable fraction of these noncoding sequences indeed appear to


undergo unconstrained evolution (Fig. 2.2). Using McDonald–Kreitman-based

approaches, it is possible to estimate the fraction of the nucleotides in a genome

that are subject to evolutionary constraints (Sella et al. 2009). These estimated

fractions substantially differ even between animals: in Drosophila, �70% of the

sites including �65% of the noncoding sites appear to be subject to selection

(including positive selection) (Sella et al. 2009), whereas in mammals, this fraction

is estimated at 5–6% only as determined using repeats ancestral to human and

mouse as a neutral baseline (Waterston et al. 2002). An independent approach based

on the deviations from the expected neutral distribution of insertions and deletions

in mammalian genomes led to an even lower value of�3% of sites under constraint

(Lunter et al. 2006). It is notable, however, that the absolute numbers of sites

subject to selection in these animal genomes of widely different size are quite

close. By contrast, in Arabidopsis, a plant that is comparable toDrosophila in terms

of genome size and overall architecture, the fraction of constrained noncoding sites

appears to be substantially lower.

The estimate of 3–6% for the fraction of constrained sites in mammalian

genomes is remarkable from two opposite standpoints. On the one hand, it appears

that the great majority of the mammalian genomic DNA after all fits the early (and

much maligned) definition of junk (Doolittle and Sapienza 1980). Of course,

0%

20%

40%

60%

80%

100%

ORFs

control elementsintrons

"junk" genome

strong constraints

weak constraints

viru

ses

prok

aryo

tes

unic

ellu

lar

euka

ryot

es

mul

tiice

llula

reu

kary

otes

Fig. 2.2 Approximate distribution of evolutionary constraints across genomes with different

architectures. The fractions of different classes of sequences subject to constraints of varying

strength are shown as rough approximation of the values that are typical of the respective class of

genomes


recruitment of “junk” sequences, such as those of diverse transposable elements,

for various functions is common (Jordan et al. 2003; Bowen and Jordan 2007), so

yesterday’s junk can be today’s essential gene (and vice versa) but at any given

time, most of the primate genome evolves without appreciable constraints. But the

converse aspect of these estimates is that, as protein-coding sequences comprise

only �1.2% of the genome (Waterston et al. 2002), the substantial majority of the

selected sites do not encode amino acids. We still do not know the actual distribu-

tion of the constrained sites among different classes of sequences or the distribution

of selection pressures but some important contributions and their approximate

magnitudes have become clear. In particular, the selective pressure on 50-terminal

and especially long 30-terminal untranslated regions of mammalian genomes is

comparable to that affecting synonymous sites in coding regions if not stronger

(Duret et al. 1993; Shabalina et al. 2004; Drake et al. 2006). An even greater

contribution to the noncoding part of the mammalian “selectome” using the term

in the most general sense as the totality of sites subject to all form of selection as

opposed to the original usage limited to positive selection (Proux et al. 2009) is the

ever-growing compendium of noncoding RNA genes present in vertebrate gen-

omes, the RNome (Costa 2005). A major and currently best characterized part of

the RNome consists of thousands of regulatory microRNAs that are subject to a

broad range of evolutionary constraints (Shabalina and Koonin 2008; Carthew

and Sontheimer 2009). In addition, there are numerous long noncoding (macro)

RNAs the functions of which remain largely unclear although there is striking

anecdotal evidence of roles of these RNAs in gene regulation and development

(Ponting et al. 2009). Approximately 3,000 macroRNAs were found to be con-

served in mammals and are subject to a selective pressure that appears to be

comparable to the constraints affecting protein-coding genes (Ponjavic et al. 2007).

Beyond doubt, the known part of the RNome is the proverbial tip of the iceberg,

especially considering the detection of transcripts from nearly all sequences in

mammalian genomes (Bertone et al. 2004; Johnson et al. 2005). Comparative-

genomic analysis reveals numerous conserved sequences (including the so-called

ultraconserved elements that retained their identity throughout long evolutionary

spans such as the entire course of vertebrate evolution) within introns and intergenic

regions of animals and plant genomes (Dermitzakis et al. 2005; Elgar 2009), but so

far transcription into a specific functional RNA has been demonstrated only for a

few of these (Bejerano et al. 2004; Baira et al. 2008). Nevertheless, it has been

shown that the ultraconserved sequences are subject to “ultraselection” suggesting

key functions that remain to be deciphered (Katzman et al. 2007). On the whole, the

problem of evolutionarily constrained “dark matter” in animal genomes remains

pertinent as the status of the majority of constrained nucleotides is still unclear, at

least, in vertebrates, the organisms with the lowest known gene density. In parti-

cular, the extent of sequence conservation unrelated to transcription but rather

caused by requirements of expression regulation, chromatin structure, and other

factors is still a wide open question.

To succinctly summarize the current understanding of the constraints affecting

different types of sites across the known diversity of the genomes (Fig. 2.2), some


fundamental, straightforward conclusions appear indisputable, in particular, that

nonsynonymous sites in protein-coding sequences and sequences encoding struc-

tural RNAs are among the most strongly constrained and that the characteristic

distributions of constraints critically depend on genome architecture. However,

beyond these basic principles, and perhaps unexpectedly, the evolutionary regimes

seem to widely differ even for rather closely related lineages, and much additional

work in diverse organisms is required to develop a comprehensive picture of the

constraints and pressures that shape genome evolution.

2.3 Evolutionary Constraints on Gene and Genome

Architectures

Beyond sequence evolution, comparative genomics yields massive amounts of

data on the evolution of gene and genome organization, or architecture. An aspect

of gene architecture that is common to all life forms but is particularly prominent

in eukaryotes is the multidomain organization of proteins (Koonin et al. 2000).

Numerous proteins consist of multiple “evolutionary domains” that may or may not

correspond to structural domains but in either case show varying degrees of

evolutionary mobility. The multidomain organization of some key proteins is

conserved through the entire course of evolution of domains of cellular life

(archaea, bacteria, and eukaryotes), as is the case of the association of polymerase

domains with nuclease domains in different families of DNA polymerases (Aravind

and Koonin 1998), to mention just one striking example. More generally, however,

domain rearrangements at all ranges of evolutionary distances form an important

resource of evolutionary plasticity which is particularly remarkable in the case of

so-called promiscuous domains which combine with diverse other domains in

numerous proteins and often provide connections in interaction and regulatory

networks and complexes (Wuchty and Almaas 2005; Basu et al. 2008, 2009).

A feature of gene architecture that is almost fully eukaryote-specific is the

exon–intron organization of protein-coding genes which in eukaryotes consist of

multiple exons separated by introns. A notable discovery of comparative genomics

is the high level of conservation of intron positions over long evolutionary spans:

indeed, up to 25–30% of the intron positions are shared between animals and plants,

with the implication that most of these introns remained in the same positions

throughout eukaryotic evolution (Fedorov et al. 2002; Rogozin et al. 2003; Roy and

Gilbert 2006). Within some of the animals lineages, in particular, vertebrates, there

seems to be almost complete intron stasis, with minimal intron loss and virtually no

gain. In a sharp contrast, evolution of other lineages, such as nematodes, as well as

many groups of unicellular eukaryotes, involves extensive turnover of introns

(Carmel et al. 2007; Roy and Penny 2007). Thus, evolution of eukaryotic gene

architecture shows a complex landscape, with a dynamic evolutionary process in

some lineages but much less change in others.


Genome architecture refers to all aspects of the mapping of genetic elements

onto the genome including gene order, clustering, and co-regulation of genes with

related functions, allocation of genes to individual chromosomes, etc. (Carmel et al.

2007; Lynch 2007c; Roy and Penny 2007; Koonin 2009a). The very first compar-

isons of the order of genes in sequenced bacterial genomes revealed a remarkable

lack of conservation of the long-range gene order which contrasts with the recurrent

presence of partially conserved arrays of co-regulated genes, operons, in diverse

prokaryotes (Mushegian and Koonin 1996a; Dandekar et al. 1998). Subsequent

analysis has shown that the divergence of long-range gene orders in prokaryotes is

roughly proportional to sequence divergence of protein-coding genes but evolution

of gene order is extremely fast such that, for many lineages, no long-range conser-

vation is seen even at very low levels of sequence divergence. Beyond this general

pattern, the rate of gene order decay substantially differs between prokaryotic

lineages (Novichkov et al. 2009b) (Fig. 2.3). The gene order in prokaryotes appears

to be disrupted primarily by inversions centered at the origin of replication the

frequency of which dramatically differs among prokaryotes (Eisen et al. 2000).

Apparently, the origin-centered inversion is a neutral process that is not constrained

(or minimally constrained) by purifying selection and depends primarily on the

activity of the relevant recombination machinery.

In contrast to the lack of conservation of the long-range gene order, prokaryotic

operons are characterized by a combination of evolutionary resilience and plasti-

city, forming overlapping gene arrays that are partially shared by evolutionarily

0

0.05

0.1

0.15

0.2

0.25

0.3

0 0.5 1 1.5 2 2.5

Gen

om

e re

arra

ng

emen

t d

ista

nce

(d

Y)

Sequence distance (dS)

Shewanella baltica

Bacillus anthracis

Burkholderia ambifaria

Yersinia pestis

Fig. 2.3 Divergence of large-scale genome organization vs. protein sequence conservation. The

data are shown for four sets of closely related bacterial strains from the ATGC database (Novichkov

et al. 2009a). The rearrangement distance (dY) is calculated as the fraction of (putative) orthologs

that do not belong to regions of synteny. The dS value of 1 approximately corresponds to 93–97%

identity between the compared sequences (Novichkov et al. 2009b)


distant organisms (Rogozin et al. 2002; Ling et al. 2009). To a large extent, the wide

spread of some operons among prokaryotes (the ribosomal superoperon and mem-

brane transport cassette operons being the prime cases in point) owes to horizontal

gene transfer (HGT) as captured in the selfish operon concept (Lawrence and Roth

1996; Lawrence 1999). When a transferred piece of DNA includes an entire operon

consisting of genes encoding a complete pathway or functional system, the chances

of fixation dramatically increase. The lack of long-range gene order conservation

notwithstanding, the gross architecture of prokaryotic genomes is not entirely

unconstrained: there are substantial biases in gene localization, for instance, the

preferential codirectionality of gene transcription with replication, conceivably, as

a result of selection for minimization of the chance of collision between RNA

polymerase and replication forks (Rocha 2008).

With a few notable exceptions, such as nematodes and trypanosomes, eukaryotes

have no operons; those operons that do exist have nothing to do with prokaryotic

operons and seem to have evolved de novo (Blumenthal 2004; Osbourn and Field

2009). Attempts to identify nonrandomness in the eukaryotic gene order, in the

form of clustering of genes with connected functions, similar expression levels, and

patterns, and other similar characteristics have led to mixed results (Hurst et al.

2004; Koonin 2009a; Osbourn and Field 2009). With some striking exceptions such

as the strict order of the animal Hox genes (Lemons and McGinnis 2006), the trends

in gene clustering tend to be weak, so the gene order can be considered quasi-

random (Koonin 2009a). Evolution of gene order in eukaryotes seems to be

determined, primarily, by random chromosomal breaks, and there are no highly

conserved gene arrays between distantly related forms, such as different animal

phyla, let alone animals and fungi or plants.

On the whole, evolution of genome architecture appears to be shaped by the

interplay of strong constraints that determine the conservation of operons, weak

constraints on other forms of functional clustering and large-scale gene organiza-

tion, and extensive dynamics of genome rearrangements and HGT. This dynamics

both counteracts weak constraints by disrupting gene associations and reinforces

the effect of stronger constraints as in the case of horizontal spread of “selfish”

operons.

2.4 Evolutionary Constraints on Genome Size, Gene

Number, Evolution of Orthologous Gene Lineages,

and Gene Repertoires

The number of protein-coding genes in cellular life forms varies within a surpris-

ingly narrow range compared with the genome size and especially considering the

difference in organizational complexity between prokaryotes and multicellular

eukaryotes. Excluding, on one end of the spectrum, extremely reduced genomes

of some intracellular parasitic bacteria that seem to be on their way to becoming


organelles (Nakabachi et al. 2006) and, on the other end, polyploid plant genomes,

the number of encoded proteins varies only from �500 to �25,000, less than two

orders of magnitude (Koonin 2009a). The largest known bacterial genome contains

only about twofold fewer protein-coding genes than the most complex eukaryotic

genomes. As already mentioned above, the genome architectures are drastically

different between unicellular and multicellular life forms, so that in unicellular

organisms, especially in prokaryotes, the number of encoded proteins closely

correlates with the genome size (roughly constant gene density, around one gene

per kilobase of DNA), whereas in multicellular organisms, especially animals, the

two are decoupled.

What constrains the number of encoded proteins from below and from above?

The low threshold of genomic complexity intuitively relates to a “minimal gene set

for cellular life”, that is, the minimal set of genes sufficient to maintain a functional

cell (in practice, of course, a prokaryotic cell) (Koonin 2003; Moya et al. 2009). The

concept of a minimal gene set is intrinsically linked to the definition of gene

orthology and orthologous gene sets and nonorthologous gene displacement. Ortho-

logs are genes that evolved from a single ancestral gene in the last common ancestor

of the compared genomes in contrast to paralogs, genes that evolved by duplication

(Koonin 2005). For the majority of genes, evolution of orthologous gene lineages is

constrained within a distinct trajectory so that such lineages remain unique and

distinguishable from each other over long evolutionary spans. This evolutionary

distinctness of orthologous lineages provides for the considerable effectiveness of

straightforward methods for identifications of orthologous genes sets based on

“bidirectional best hits” and is key to comparative genomics allowing comprehen-

sive comparison of gene repertoires and delineation of core sets of conserved genes

and putative minimal gene sets (Tatusov et al. 1997; Altenhoff and Dessimoz 2009).

Minimal gene sets for cellular life derived by comparative-genomic and experimen-

tal approaches converge at 250–350 genes and seem to encode most of the essential

cellular functions (Koonin 2003; Moya et al. 2009). However, an apparent paradox

is that a set of 250–350 conserved orthologous genes can be derived only in

comparisons of small sets of genomes of not too diverse organisms as exemplified

by the first analysis of this kind that compared the parasitic bacteria Haemophilusinfluenzae and Mycoplasma genitalium and yielded a hypothetical minimal gene

set of approximately 250 genes (Mushegian and Koonin 1996b). The core set of

ubiquitously conserved genes is continuously shrinking with the addition of new

sequenced genomes and seems to be limited to approximately 30 genes, all encoding

proteins involved in translation and transcription (Charlebois and Doolittle 2004;

Koonin and Wolf 2008). The explanation is nonorthologous gene displacement:

most of the essential cellular functions can be performed by members of more than

one orthologous gene set, and in many cases, genes or systems responsible for the

same function are completely unrelated (Koonin et al. 1996; Koonin 2003). The

relevant concept for defining a minimal genetic complement of a cell – the low

bound of genomic complexity – is not a unique minimal gene set but rather a unique

set of indispensable functional niches that can be filled with diverse collections of

genes. Minimal requirements for specific life styles can be defined similarly, for


instance, the minimal gene complement of an autotrophic organism, which includes

about 1,000 essential functions (Koonin 2003). Thus, the low bound is defined by

the minimal number of functions that are necessary to support a particular life style,

but even at this fundamental level of cellular organization, there is notable plasticity

in terms of specific gene complements supporting these functions.

The nature of the upper bound of genetic complexity is much less clear. However,

the question why, despite the accelerating genome sequencing, the maximum number

of genes practically does not grow, seems pressing, especially, considering the

decoupling of gene number and genome size seen in multicellular prokaryotes. One

attractive hypothesis is the “bureaucratic ceiling of complexity”. It has been noticed

that different functional classes of genes scale differently with the total number of

genes in a genome. Some variation notwithstanding, in prokaryotes, there seem to be

three fundamental exponents that characterizes these dependences: 0, 1, and 2 (van

Nimwegen 2003; Koonin andWolf 2008). Genes for proteins involved in information

processing (translation, transcription, and replication) scale with a 0 exponent, i.e.,

the number of these genes reaches a plateau already in the smallest genomes and

effectively does not depend on the overall genomic complexity; metabolic enzymes

and transport proteins scale roughly proportionally to the total number of genes,

whereas regulators and signal transduction system components scale quadratically

(Fig. 2.4). The characteristic exponents of the three broad functional classes of genes

show remarkably little variation across prokaryotic lineages suggesting that the

differential evolutionary dynamics of genes with different functions reflect funda-

mental “laws” of evolution of cellular organization (Molina and van Nimwegen

2009) or, in other words, distinct, strong constraints on the functional composition

1

10

100

1000

10000

100 1000 10000

Nu

mb

er o

f p

rote

ins

in t

he

clas

s

Total number of proteins in COGs

Transcriptional regulators

Signal transduction

Metabolism

Translation

g= 1.9

g= 1.0

g= 1.9

γ = 0.2

Fig. 2.4 Differential scaling of four broad classes of genes with the total number of genes in

prokaryotic genomes. The data are from (Koonin and Wolf 2008); genes that did not belong to

COGs (typically, 15–20% in each genome) were not taken into account


of genomes. Eukaryotic genes show similar even if less pronounced patterns of power

law gene scaling, with the exponent for the regulatory genes being substantially

greater than one (van Nimwegen 2003).

The deep underlying causes of the superlinear scaling of the regulators remain to

be understood. A simple “toolbox” model of evolution of prokaryotic metabolic

networks seems to be compatible with the quadratic scaling of regulators (Maslov

et al. 2009). Under this model, enzymes for utilizing new metabolites together with

their dedicated regulators are added (primarily, via HGT) to a progressively versa-

tile reaction network, and because of the growing complexity of the preexisting

network that provides enzymes for intermediate reactions, the ratio of regulators to

regulated genes steadily grows. Regardless of the exact underlying mechanisms, the

superlinear scaling of the regulators clearly could determine the upper limit of the

growth of the gene number. At some point (that is not easy to identify precisely),

the cost of adding extra regulation (“inflating bureaucracy”) will inevitably become

unsustainable, curbing the growth of genetic complexity.

The bureaucracy ceiling hypothesis seems particularly plausible in view of the

surprising lack of major gene number expansion in vertebrates where the coupling

between the gene number and genome size is obviously broken (see also below). In

these organisms, the cost of replication can be ruled out as the major factor deter-

mining the upper limit, and the cost of regulation, possibly, along with the cost of

expression, is the most likely candidate for the role of the principal constraint. It is

not by chance, then, that vertebrates evolved other, elaborate means of increasing

the proteomic complexity, such as the pervasive alternative splicing and alternative

transcription (Nilsen and Graveley 2010), and regulatory complexity (the expan-

sive, still under-appreciated regulatory RNome) that do not involve inflation of the

number of protein-coding genes.

A major process of genome evolution that in eukaryotes could be the principal

path to innovation is gene duplication leading to the formation of paralogous gene

families (Ohno 1970; Lespinet et al. 2002). The size distribution of paralogous

families in each studied genome follows a power-law-like function that is repro-

duced, with a high precision, by a simple gene birth and death model conditioned on

the equilibrium (constant size) in genome evolution (Karev et al. 2002; Koonin

et al. 2002). This process seems to underlie a fundamental constraint on gene

demography that is coupled to the constraint on the total number of genes.

Beyond the sheer numbers of genes, comparative genomics yields insights into

the constraints on and plasticity of gene repertoires. In agreement with the findings

on the small and shrinking cores of conserved genes, nonorthologous gene dis-

placement, and extensive redundancy, gene loss has emerged as a major factor

of evolution in all life forms. Gene loss is dominant over other processes in the

evolution of parasites but is extensive in all lineages, in particular, in the evolution

of many animal taxa as illustrated by the high level of orthology between verte-

brates and primitive animals such as sea anemone and trichoplax, in contrast to

much more limited orthologous relationships between vertebrates and arthropods

or nematodes (Putnam et al. 2007; Srivastava et al. 2008). Individual genes show

a broad distribution of propensities for gene loss (PGL) (Krylov et al. 2003), and


moreover, it appears that the observed evolutionary and phenomic features of genes

are compatible with a steady-state model of genome evolution under which the

distribution of PGL as well as the distribution of gene loss rate remain effectively

constant over extended evolutionary spans (Wolf et al. 2009). This distribution

might be another important constraint governing genome evolution.

2.5 The Causes of Evolution of Protein-Coding Genes

Protein-coding genes, at least, the nonsynonymous positions that determine the

amino acid identity, are among the most strongly constrained sequences in all

genomes. However, the distribution of the rates of evolution among orthologous

genes in any pair of compared genomes spans 3–4 orders of magnitude and is much

broader than the distribution of the rates for synonymous sites (Fig. 2.1). Remarkably,

the shapes of the rate distributions for orthologous proteins are highly similar for all

studied cellular life forms, from bacteria to archaea to mammals (Wolf et al. 2009)

(Fig. 2.5). Another universal of genomic and phenomic evolution is the anticorrela-

tion between the rate of evolution of a protein-coding gene and its expression level:

highly expressed genes evolve slowly, a dependence that was invariably observed in

all model organisms for which expression data are available (Pal et al. 2001, 2006;

Krylov et al. 2003; Drummond and Wilke 2008). Given the aforementioned positive

0.01 0.1 1 10

Relative evolution rate

Burkholderia

Salinispora

Methanococcus

Homo

Aspergillus

model

Fig. 2.5 The universal distribution of evolutionary rates across orthologous gene sets. The

evolutionary rates for five pairs of closely related organisms from different branches of life were

calculated as nucleotide distances for the complete sets of orthologous genes (Wolf et al. 2009).

The relative evolution rate for each gene was obtained by dividing its evolution rate by the median

rate for the respective pair of organisms. “Model” refers to estimated transition rates in 134

mutationally connected networks for simulated robustly folding 18-mer protein-like molecules

(Lobkovsky et al. 2010). Original model rates were normalized by their median value and scaled to

standard deviation of 0.25 to match the width of the distributions derived from biological data


correlation between Ka and Ks, it is not surprising that both rates show the same

dependence; more unexpectedly, this anticorrelation with the evolutionary rate was

detected also for 30UTRs but not for 50UTRs (Jordan et al. 2004).

The existence of these universals of genomic evolution and their fundamental

link with phenomic characteristics suggest that the primary causes of protein

evolution could have more to do with fundamental principles of protein folding

than with unique biological functions. It has been proposed that the principal

selective factor underlying the evolution of proteins is robustness to misfolding,

owing to the deleterious effect of misfolded proteins that, in addition to the expen-

diture of energy, can be toxic to the cell (Drummond et al. 2005; Drummond and

Wilke 2008, 2009). Moreover, under this model, evolution of synonymous sites is

constrained, at least, in part, by the same factors as the evolution of proteins owing to

the pressure for the preferential use of optimal codons in highly expressed proteins

and in specific sites that are important for protein folding (Drummond and Wilke

2008; Zhou et al. 2009), and evolution of the 30UTRs could follow the same trend

(Jordan et al. 2004) as these regions are involved in the regulation of translation.

A recent modeling study of misfolding-dominated protein evolution that

employed a simple off-lattice model of protein folding and produced estimates of

evolutionary rates under the assumption that protein misfolding was the only source

of fitness cost (Lobkovsky et al. 2010) reproduced the universal distribution of

protein evolutionary rates as well as the dependence between evolutionary rate and

expression with considerable accuracy (Fig. 2.5). These findings suggest that the

universal rate distribution indeed might be a consequence of fundamental physics of

proteins and provide for a general model of protein evolution under which evolution

of a given protein is determined, primarily, by its intrinsic robustness to misfolding

which also determines the attainable level of translation (Fig. 2.6) (Wolf et al. 2010).

In general, the robustness of a protein to misfolding and accordingly the rate of

evolution are determined by the size of the (nearly) neutral network, that is, the

network of sequences that have approximately the same robustness and accordingly

the same fitness as the original sequence (Wagner 2008). Under the model (Wolf

et al. 2010), the nearly neutral network size is (roughly) inversely proportional to

the robustness of the original sequence, i.e., in the fitness landscape, robust, highly

expressed proteins occupy tall, steep peaks, with small areas of high fitness, hence

slow evolution; in contrast, proteins with lower robustness occupy lower and wider

peaks, with larger areas of high fitness, allowing faster evolution (Fig. 2.6).

The original hypothesis on misfolding-dominated evolution of protein-coding

genes held that misfolding was largely induced by mistranslation of the coding

sequence (Drummond and Wilke 2008, 2009). The latest analysis of the relative

contributions of structural–functional constraints and translation rate to protein

evolution imply that stochastic misfolding of the native sequence could be even

more common and consequential than mistranslation-induced misfolding (Wolf

et al. 2010). Nevertheless, mistranslation (somatic mutation), which is relatively

frequent [10�4–10�5 per codon (Kramer and Farabaugh 2007)], is likely to be an

important factor affecting the instantaneous shape of the robustness landscape by

temporarily expanding the nearly neutral network (Fig. 2.6).


The view of protein evolution under which the primary constraints have to do

more with the maintenance of the native folding as well as intermolecular inter-

actions than with unique protein functions seems to be compatible with the recent

large-scale analysis of protein family evolution (Worth et al. 2009).

2.6 Constraints on Molecular Phenotypes

The advances of systems biology provide for direct evolutionary study of molecular

phenomic variables, such as gene expression, protein abundance, and architecture

of interaction networks. In other words, it is now possible to assess evolutionary

variance and constraints by directly comparing gene expression profiles and net-

works, protein abundances and other features of the molecular phenotype between

different organism and evolutionary lineages.

fold

ing

robu

stne

ss

at low expression(higher evolution rate)

at high expression(lower evolution rate)

fitness

low high

proteinfamily X

sequence space

proteinfamily Y

Fig. 2.6 A conceptual model of misfolding-driven protein evolution. The cartoon schematically

shows the robustness/fitness landscapes for two protein families at high and low expression levels.

The high fitness/robustness area (green) reflects the size of the nearly neutral network in the

sequence space


Molecular phenomic variables, such as gene expression level and number of

interaction partners of a protein, show a distinct structure of dependences among

themselves and with evolutionary variables such as sequence evolution rate and the

rate of gene loss (Wolf et al. 2006). The correlations between phenomic variables

are typically positive, i.e., highly expressed proteins also tend to interact with many

other proteins, to have many paralogs etc., whereas the correlations between the

phenomic and evolutionary variables are negative, for instance, highly expressed

genes on average evolve slower than those expressed at a low level. Thus, as

exemplified by the model of protein evolution discussed above, constraints on the

ranges of phenomic variables, in part, appear to constrain evolution of gene

sequences, gene repertoires, and genome architectures.

Several studies suggested that gene expression in animals is not strongly con-

strained during evolution (Jordan et al. 2004; Khaitovich et al. 2004) or at least has a

major neutral component (Jordan et al. 2004; Khaitovich et al. 2004). However,

subsequent analyzes revealed clear signatures of selective constraints that affect

gene expression (Denver et al. 2005; Jordan et al. 2005; Gilad et al. 2006). Recently,

it has been shown that the abundances of orthologous proteins are strongly corre-

lated even among distantly related animals. A correlation coefficient greater than

0.8 was observed for approximately 3,000 orthologous genes from the nematode

C. elegans and the fly D. melanogaster, a value that is in sharp contrast with the

correlation coefficients in the range of 0.2–0.4 that are typically seen in comparisons

of genomic and molecular phenomic variables (Wolf et al. 2006). Strikingly, the

correlation between protein abundances was found to be substantially greater than

the correlation between mRNA expression rates and between the rates of coding

sequence evolution (measured by comparison of orthologous genes from pairs of

closely related species) within the same set of genes (Schrimpf et al. 2009; Wolf

et al. 2010). Thus, assuming there are no unrecognized biases in the measurements,

protein abundance appears to be constrained during evolution to a substantially

greater extent than gene expression and even stronger than the sequence evolution

itself.

The global architectures of protein interaction and gene coexpression networks

appear to be universal across all life forms, with the characteristic power law

distribution of the network node degree (number of connections) (Barabasi and

Oltvai 2004). Local network structures seem to be much less strongly constrained

and differ even among closely related organisms (Bergmann et al. 2004; Tsaparas

et al. 2006). However, a comparison of gene coexpression networks from the

so-called mutation accumulation lineages of C. elegans, in which the selective

constraints are effectively removed (Denver et al. 2005), with those of the natural

isolate suggests that it is the local wiring of the coexpression network that is

constrained by selection, whereas the global properties are not affected by the

removal of constraints (Jordan et al. 2008). Thus, the similar global network

properties seen in widely different organisms might reflect “neutral” rather than

selective constraints, that is, could have evolved via simple, stochastic, nonselective

processes as exemplified by birth-and-death models of genome and network evolu-

tion (Koonin et al. 2002; Lynch 2007a).


2.7 Constraints on Evolutionary Trajectories: What Happens

When the Tape of Evolution Is Rewound?

An intriguing, deep question in evolutionary biology is how constrained is the

course of evolution itself, or in other words, to what extent the evolutionary process

is free to explore different trajectories between the given initial and end states

(Kassen 2009). In theory, mutational trajectories in sequence space are considered

to be fundamentally stochastic (Mani and Clarke 1990). However, experimental

evolution studies indicate that paths of adaptive evolution are substantially con-

strained by interactions between mutation (epistasis and pleiotropy) although not

to the point of becoming deterministic. A series of experiments on evolution of

bacterial antibiotic resistance resulting from 5 point mutations in the b-lactamase

gene showed that, of the 120 trajectories across the sequence space, 102 were

inaccessible to evolution, and of the remaining 18 trajectories, several had negligi-

ble probability of realization (Weinreich et al. 2006). Even stronger constraints

were identified in a subsequent study that explored a more complex fitness land-

scape by simultaneously evolving resistance to two antibiotics (Novais et al. 2010).

The remarkable long-term study of bacterial evolution under controlled condi-

tions by Lenski and coworkers provides examples of both parallel emergence of the

same mutations under a particular selective pressure and the realization of multiple

trajectories (Barrick et al. 2009; Barrick and Lenski 2009; Kassen 2009; Stanek

et al. 2009). For instance, it has been explicitly shown that evolution of the same,

extremely rare phenotype, the ability to grow on citrate, proceeded along distinct

trajectories in different Escherichia coli populations (Blount et al. 2008).Direct studies of evolutionary trajectories in the sequence space are still very

limited but they have already made it clear that, although historical contingency is

crucial in the evolutionary process (Jacob 1977), the exploration of the sequence

space is strongly constrained so that only a minority of theoretically possible

trajectories are accessible. The extent of these constraints depends on the shape

of the fitness landscape: the more rugged the landscape, the stronger the constraints.

The shape of the landscape itself depends on the nature, strength, and interactions

of the relevant selective factors and evolves with time, which makes it more of a

seascape (Mustonen and Lassig 2009, 2010).

2.8 Robustness, Plasticity, and Evolutionary Constraints

The aspects of evolution that are orthogonal to constraints are the plasticity of

genomic and phenomic characteristics and the robustness of molecular phenotypes

(Wagner 2005). In many groups of organisms, large-scale genome organization

seems to be only weakly constrained so that gene order substantially differs even

between closely related organisms, especially, among prokaryotes (Koonin 2009a;

Novichkov et al. 2009b) (Fig. 2.6). The gene repertoire of many organisms,


especially, prokaryotes shows plasticity that may even exceed the plasticity of

genome architecture as dramatically illustrated by rapid genome reduction in

parasitic bacteria (Darby et al. 2007) and by acquisition of pathogenicity islands

that may comprise over 30% of the recipient genome in bacterial pathogens

(Dobrindt et al. 2003). The plasticity of genome organization and composition is

paralleled by the evolutionary flexibility of regulatory networks and complements

the more strongly constrained evolution of individual genes (Lozada-Chavez et al.

2006; Kazakov et al. 2009).

Evolutionary plasticity and the strength of evolutionary constraints are tightly

linked to robustness of biological systems, that is, resistance of phenotypes to

genetic perturbation (mutations, recombination, etc.). Robustness seems to be an

evolved property as demonstrated by the study of specialized buffering mecha-

nism (for instance, those mediated by molecular chaperones of the HSP90 fam-

ily), the impairment of which (often by environmental stress) reveals hidden

genetic variation and accordingly enhances the evolutionary potential of the

organism (Queitsch et al. 2002; Wagner 2008; Masel and Siegal 2009). Recently,

the concept of variation stabilization has been extended to include numerous

genes that are not molecular chaperones but possess extremely diverse functions;

it seems that stabilization is a general property of interaction networks, so that

disruption of almost any highly connected node reduces robustness of the system

and leads to increased variation (Bergman and Siegal 2003). A comprehensive

study of such “capacitor” properties of yeast mutants revealed approximately 300

genes (about 6% of the total) whose disruption significantly decreased the robust-

ness of yeast to environmental perturbations (Levy and Siegal 2008). Thus,

robustness might be a major, selectable mechanism that counteracts evolutionary

constraints, in particular, those caused by the interaction between mutations, and

enhances plasticity.

2.9 Effective Population Size as the General Determinant

of Evolutionary Constrains and Distinction Between

Constraints and Neutral Conservation

The classic population genetics theory asserts that the effectiveness of purifying

selection is proportional to the effective population size of the given organism

(assuming a uniform mutation rate for simplicity). In other words, only those mut-

ational changes can be fixed or efficiently eliminated during evolution for which

s > 1/Ne, where s is the selection coefficient and Ne is the effective population size(Lynch 2007c). Conversely, mutations with s < 1/Ne are effectively “invisible” to

selection. This simple dependence seems to be an important, possibly, the primary

determinant of the constraints that affect different aspects of genome and phenome

evolution. In particular, differences in Ne seem to underlie the qualitative differ-

ence in the genome architectures of unicellular and multicellular organisms


described above (Lynch and Conery 2003; Lynch 2007b). Substantial genome

expansion seems to be attainable only in organisms with small populations and

the attendant weak selection, such as plants and animals. In these organisms, the

deleterious effect of propagation of nonfunctional sequences is often too small to

allow their “detection” and elimination by purifying selection. Accordingly, evolu-

tionary conservation does not automatically imply that the conserved feature is

constrained by purifying selection but rather, somewhat paradoxically, can reflect

weak purifying selection that is insufficient to eliminate nonadaptive ancestral

features.

Evolution of the exon–intron gene structure in eukaryotes provides an excel-

lent case in point for this population-genetic paradigm. Most of the introns do not

appear to possess a distinct function but do require distinct splicing signals for

transcript maturation to occur accurately. Thus, approximately 25 nucleotides

per intron are subject to purifying selection of varying strength (Lynch 2006a).

Because of the associated cost of selection and also owing to the expenditure of

time and energy on replication and transcription of intronic sequences, function-

less introns are weakly deleterious for the respective organisms. However, a

simple estimate taking into account the characteristic mutation rates in eukaryotes

shows that the deleterious effect of introns is “visible” to purifying selection only

in relatively large populations with Ne on the order of 107 or greater. This is the

characteristic range of effective population sizes of unicellular eukaryotes,

whereas multicellular eukaryotes typically have smaller populations (Lynch and

Conery 2003; Lynch 2006a, 2007c). The effect of these differences on the evo-

lution of genome architecture in eukaryotes is dramatic. Unlike genomes of

unicellular forms that typically contain less than one intron per gene, and in

many case, only a few introns in the entire genome, plants, and animals possess

numerous introns, up to 8 per gene in vertebrates (Roy and Gilbert 2006). The

positions of many introns are conserved in orthologous genes of animals and

plants (see above), that is, most likely, since the time of existence of the last

common ancestor of the extant eukaryotes. However, there seems to be no reason

to claim that, in general, the positions of introns are constrained during evolution.

The conservation of intron positions appears to be due to the weak purifying

selection that precludes efficient elimination of introns in organisms with small

characteristic values of Ne.

Beyond the sheer number of introns, the features of introns themselves drasti-

cally differ: all the introns in intron-poor genomes of unicellular eukaryotes are

short, with tightly controlled lengths and highly conserved, optimized splice signals

at exon–intron junctions (Irimia et al. 2007; Irimia and Roy 2008). By contrast,

introns in intron-rich genomes, such as plants, and animals, are often long (espe-

cially, in vertebrates) and are bounded by relatively weak, suboptimal splice signals

owing to the relatively low selection favoring strong splicing signals (Irimia et al.

2009). The existence of these long introns with weak splice signals, which yield

relatively inaccurate splicing, provides for the evolution of alternative splicing and

nested gene structures, the crucial factors of structural and regulatory diversifica-

tion of proteins and RNAs in multicellular eukaryotes.


The case of intron evolution illustrates the crucial interplay of constraints and

plasticity that is central to the evolution of genomes and molecular phenomes

(Fig. 2.7). Effective population size determines the background strength of purify-

ing selection (constraints). When Ne is small, as in multicellular eukaryotes,

constraints are relatively weak, so plasticity is enhanced such that nonfunctional

genomic elements like introns can be retained, the result being a system that is

relatively inefficient and vulnerable to random factors that can cause extinction, but

also possesses a high potential for evolutionary innovation. Conversely, when Ne is

large, as in most prokaryotes, many aspects of evolution are strongly constrained

although there is still much plasticity in the evolution of these organisms thanks to

dynamic, effectively neutral processes, in particular, HGT.

Its fundamental importance notwithstanding, it is important to keep in mind that

Ne determines the course of evolution only on a coarse grain scale. Thus, a

comparative analysis of the Kn/Ks values among prokaryotic lineages failed to

detect a negative correlation between selective constraints and genome size, as

implied by the straightforward population genetic perspective (Lynch 2006b). On

the contrary, larger genomes tend to evolve under stronger constraints (even when

only free-living microbes are analyzed) suggesting that lifestyle could be a critical

determinant of genome evolution (favoring, in particular, gene acquisition via HGT

in variable environments) independent of Ne (Jordan et al. 2002; Novichkov et al.

2009b).

strong

weak

level of organization

molecular structure anddynamics

local genome context genome architecture molecular phenomics

functional andfolding-critical sites

disordered segments

intron donor andacceptor sites

"junk" genome

typical regulatorysites

genome-scale geneorder

operons and geneclusters

synonymous sites inCDS

introns

gene neighborhoods

typical protein sites

functional andregulatory networks

protein abundance

mRNA abundance

gene islands andsuperoperons

protein function

cons

trai

mts

plas

ticity

low

high

Fig. 2.7 Genomic and phenomic constraints operative at different levels of biological organi-

zation. The scales are rough approximations


2.10 Conclusions: Selective and Neutral Constraints

and Evolutionary Universals

The prevailing theme that emerges from the recent advances of evolutionary

genomics and evolutionary systems biology is the plurality of constraints that affect

the evolution of different types of sequences in any genome, genome architectures,

and molecular phenomes (Fig. 2.7) along with major differences of evolutionary

regimens between taxa. Nevertheless, beyond this diversity, comparative-genomic

and molecular phenomic analysis reveals universal patterns that at least in some

cases are compatible with relatively simple and general models of evolution. As

discussed here, such models start to suggest simple, fundamental causes underlying

important aspects of evolution such as the constraints on evolution of proteins and

evolution of gene repertoire (Table 2.1). In this context, it seems appropriate to

expand the notion of constraints to include not only selective but also “neutral”

constraints that are determined by nonselective, stochastic properties of biological

systems and are often amenable to modeling using techniques borrowed from

statistical physics (Table 2.1) (Frank 2009; Koonin 2009b).

Evolutionary trajectories in the sequence space seem to be strongly constrained,

thus substantially limiting the “tinkering potential” of evolution, using the famous

metaphor of Jacob (Jacob 1977). The evolutionary process thus appears to be a

compromise “between design and bricolage” (Wilkins 2007), the design aspect

Table 2.1 Universals of genome and molecular phenome evolution

Universal pattern Putative underlying

process/model

Nature of

relevant

constraints

References

Approximately log-normal

distribution of

evolutionary rates of

protein-coding genes

Protein folding Selective: protein

robustness to

misfolding

(Wolf et al. 2009;

Lobkovsky et al.

2010)

Anticorrelation between

evolution rate and

expression level

(translation rate) of

protein-coding genes

Protein folding Selective: protein

robustness to

misfolding

dependent on

translation

rate

(Drummond and

Wilke 2008,

2009; Wolf et al.

2010)

Distinct scaling laws for

different functional

classes of genes

“Toolbox”-like growth

of metabolic

networks

Neutral (van Nimwegen

2003; Maslov

et al. 2009;

Molina and van

Nimwegen 2009)

Power law like distribution

of paralogous gene family

size

Birth and death

process of gene

evolution

Neutral (Karev et al. 2002;

Koonin et al.

2002)

Power law like distribution

of node degree in

interaction and

coexpression networks

Network evolution by

preferential

attachments

Neutral (Barabasi and Oltvai

2004; Tsaparas

et al. 2006)


brought about by constraints (certainly having nothing to do with any intelligence)

and the bricolage stemming from the evolved robustness and the ensuing plasticity

of evolving organisms.

Comparative genomics and systems approaches transform evolutionary biology

into a much more complex but also more precise, quantitative field than it was in the

twentieth century. Next generation sequencing, quantitative proteomics, and other

systemic approaches, combined with more specific approaches of experimental

evolution, can be expected to reveal the specific, precise constraints affecting

diverse aspects of genome and phenome evolution.

References

Altenhoff AM, Dessimoz C (2009) Phylogenetic and functional assessment of orthologs inference

projects and methods. PLoS Comput Biol 5:e1000262

Aravind L, Koonin EV (1998) Phosphoesterase domains associated with DNA polymerases of

diverse origins. Nucleic Acids Res 26:3746–3752

Baira E, Greshock J, Coukos G, Zhang L (2008) Ultraconserved elements: genomics, function and

disease. RNA Biol 5:132–134

Barabasi AL, Oltvai ZN (2004) Network biology: understanding the cell’s functional organization.

Nat Rev Genet 5:101–113

Barrick JE, Lenski RE (2009) Genome-wide mutational diversity in an evolving population of

Escherichia coli. Cold Spring Harb Symp Quant Biol 16:345–355

Barrick JE, Yu DS, Yoon SH, Jeong H, Oh TK, Schneider D, Lenski RE, Kim JF (2009) Genome

evolution and adaptation in a long-term experiment with Escherichia coli. Nature

461:1243–1247

Basu MK, Carmel L, Rogozin IB, Koonin EV (2008) Evolution of protein domain promiscuity in

eukaryotes. Genome Res 18:449–461

Basu MK, Poliakov E, Rogozin IB (2009) Domain mobility in proteins: functional and evolution-

ary implications. Brief Bioinform 10:205–216

Bejerano G, Pheasant M, Makunin I, Stephen S, Kent WJ, Mattick JS, Haussler D (2004)

Ultraconserved elements in the human genome. Science 304:1321–1325

Bergman A, Siegal ML (2003) Evolutionary capacitance as a general feature of complex gene

networks. Nature 424:549–552

Bergmann S, Ihmels J, Barkai N (2004) Similarities and differences in genome-wide expression

data of six organisms. PLoS Biol 2:E9

Bertone P, Stolc V, Royce TE, Rozowsky JS, Urban AE, Zhu X, Rinn JL, Tongprasit W, Samanta

M, Weissman S, Gerstein M, Snyder M (2004) Global identification of human transcribed

sequences with genome tiling arrays. Science 306:2242–2246

Blount ZD, Borland CZ, Lenski RE (2008) Historical contingency and the evolution of a key

innovation in an experimental population of Escherichia coli. Proc Natl Acad Sci USA

105:7899–7906

Blumenthal T (2004) Operons in eukaryotes. Brief Funct Genomic Proteomic 3:199–211

Bowen NJ, Jordan IK (2007) Exaptation of protein coding sequences from transposable elements.

Genome Dyn 3:147–162

Carmel L, Rogozin IB, Wolf YI, Koonin EV (2007) Patterns of intron gain and conservation in

eukaryotic genes. BMC Evol Biol 7:192

Carthew RW, Sontheimer EJ (2009) Origins and mechanisms of miRNAs and siRNAs. Cell

136:642–655

Charlebois RL, Doolittle WF (2004) Computing prokaryotic gene ubiquity: rescuing the core from

extinction. Genome Res 14:2469–2477


Charlesworth J, Eyre-Walker A (2008) The McDonald–Kreitman test and slightly deleterious

mutations. Mol Biol Evol 25:1007–1015

Costa FF (2005) Non-coding RNAs: new players in eukaryotic biology. Gene 357:83–94

Dandekar T, Snel B, Huynen M, Bork P (1998) Conservation of gene order: a fingerprint of

proteins that physically interact. Trends Biochem Sci 23:324–328

Darby AC, Cho NH, Fuxelius HH, Westberg J, Andersson SG (2007) Intracellular pathogens go

extreme: genome evolution in the Rickettsiales. Trends Genet 23:511–520

Denver DR, Morris K, Streelman JT, Kim SK, Lynch M, Thomas WK (2005) The transcriptional

consequences of mutation and natural selection in Caenorhabditis elegans. Nat Genet

37:544–548

Dermitzakis ET, Reymond A, Antonarakis SE (2005) Conserved non-genic sequences – an

unexpected feature of mammalian genomes. Nat Rev Genet 6:151–157

Dobrindt U, Agerer F, Michaelis K, Janka A, Buchrieser C, Samuelson M, Svanborg C, Gottschalk

G, Karch H, Hacker J (2003) Analysis of genome plasticity in pathogenic and commensal

Escherichia coli isolates by use of DNA arrays. J Bacteriol 185:1831–1840

Doolittle WF, Sapienza C (1980) Selfish genes, the phenotype paradigm and genome evolution.

Nature 284:601–603

Drake JA, Bird C, Nemesh J, Thomas DJ, Newton-Cheh C, Reymond A, Excoffier L, Attar H,

Antonarakis SE, Dermitzakis ET, Hirschhorn JN (2006) Conserved noncoding sequences are

selectively constrained and not mutation cold spots. Nat Genet 38:223–227

Drummond DA, Wilke CO (2008) Mistranslation-induced protein misfolding as a dominant

constraint on coding-sequence evolution. Cell 134:341–352

Drummond DA, Wilke CO (2009) The evolutionary consequences of erroneous protein synthesis.


Drummond DA, Bloom JD, Adami C, Wilke CO, Arnold FH (2005) Why highly expressed

proteins evolve slowly. Proc Natl Acad Sci USA 102:14338–14343

Duret L, Dorkeld F, Gautier C (1993) Strong conservation of non-coding sequences during

vertebrates evolution: potential involvement in post-transcriptional regulation of gene expres-

sion. Nucleic Acids Res 21:2315–2322

Eisen JA, Heidelberg JF, White O, Salzberg SL (2000) Evidence for symmetric chromosomal

inversions around the replication origin in bacteria. Genome Biol 1(6):RESEARCH0011

Elgar G (2009) Pan-vertebrate conserved non-coding sequences associated with developmental

regulation. Brief Funct Genomic Proteomic 8:256–265

Ellegren H (2008) Comparative genomics and the study of evolution by natural selection. Mol

Ecol 17:4586–4596

Ellegren H, Smith NG, Webster MT (2003) Mutation rate variation in the mammalian genome.

Curr Opin Genet Dev 13:562–568

Eyre-Walker A, Keightley PD (2009) Estimating the rate of adaptive molecular evolution in the

presence of slightly deleterious mutations and population size change. Mol Biol Evol

26:2097–2108

Fedorov A, Merican AF, Gilbert W (2002) Large-scale comparison of intron positions among

animal, plant, and fungal genes. Proc Natl Acad Sci USA 99:16128–16133

Frank SA (2009) The common patterns of nature. J Evol Biol 22:1563–1585

Gilad Y, Oshlack A, Rifkin SA (2006) Natural selection on gene expression. Trends Genet

22:456–461

Grishin NV, Wolf YI, Koonin EV (2000) From complete genomes to measures of substitution rate

variability within and between proteins. Genome Res 10:991–1000

Harrison PM, Gerstein M (2002) Studying genomes through the aeons: protein families, pseudo-

genes and proteome evolution. J Mol Biol 318:1155–1174

Hurst LD (2002) The Ka/Ks ratio: diagnosing the form of sequence evolution. Trends Genet

18:486

Hurst LD, Pal C, Lercher MJ (2004) The evolutionary dynamics of eukaryotic gene order. Nat Rev

Genet 5:299–310


Irimia M, Roy SW (2008) Evolutionary convergence on highly-conserved 3’ intron structures in

intron-poor eukaryotes and insights into the ancestral eukaryotic genome. PLoS Genet 4:

e1000148

Irimia M, Penny D, Roy SW (2007) Coevolution of genomic intron number and splice sites.

Trends Genet 23:321–325

Irimia M, Roy SW, Neafsey DE, Abril JF, Garcia-Fernandez J, Koonin EV (2009) Complex

selection on 5’ splice sites in intron-rich organisms. Genome Res 19:2021–2027

Jacob F (1977) Evolution and tinkering. Science 196:1161–1166

Johnson JM, Edwards S, Shoemaker D, Schadt EE (2005) Dark matter in the genome: evidence of

widespread transcription detected by microarray tiling experiments. Trends Genet 21:93–102

Jordan IK, Rogozin IB, Wolf YI, Koonin EV (2002) Microevolutionary genomics of bacteria.

Theor Popul Biol 61:435–447

Jordan IK, Rogozin IB, Glazko GV, Koonin EV (2003) Origin of a substantial fraction of human

regulatory sequences from transposable elements. Trends Genet 19:68–72

Jordan IK, Marino-Ramirez L, Wolf YI, Koonin EV (2004) Conservation and coevolution in the

scale-free human gene coexpression network. Mol Biol Evol 21:2058–2070

Jordan IK, Marino-Ramirez L, Koonin EV (2005) Evolutionary significance of gene expression

divergence. Gene 345:119–126

Jordan IK, Katz LS, Denver DR, Streelman JT (2008) Natural selection governs local, but

not global, evolutionary gene coexpression networks in Caenorhabditis elegans. BMC Syst

Biol 2:96

Karev GP, Wolf YI, Rzhetsky AY, Berezovskaya FS, Koonin EV (2002) Birth and death of protein

domains: a simple model of evolution explains power law behavior. BMC Evol Biol 2:18

Kassen R (2009) Toward a general theory of adaptive radiation: insights from microbial experi-

mental evolution. Ann N Y Acad Sci 1168:3–22

Katzman S, Kern AD, Bejerano G, Fewell G, Fulton L, Wilson RK, Salama SR, Haussler D (2007)

Human genome ultraconserved elements are ultraselected. Science 317:915

Kazakov AE, Rodionov DA, Alm E, Arkin AP, Dubchak I, Gelfand MS (2009) Comparative

genomics of regulation of fatty acid and branched-chain amino acid utilization in proteobac-

teria. J Bacteriol 191:52–64

Kelly C, Churchill GA (1996) Biases in amino acid replacement matrices and alignment scores

due to rate heterogeneity. J Comput Biol 3:307–318

Khachane AN, Harrison PM (2009) Assessing the genomic evidence for conserved transcribed

pseudogenes under selection. BMC Genomics 10:435

Khaitovich P, Weiss G, Lachmann M, Hellmann I, Enard W, Muetzel B, Wirkner U, Ansorge W,

Paabo S (2004) A neutral model of transcriptome evolution. PLoS Biol 2:E132

Kimura M (1983) The neutral theory of molecular evolution. Cambridge University Press,

Cambridge

Kitano H (2002) Computational systems biology. Nature 420:206–210

Koonin EV (2003) Comparative genomics, minimal gene-sets and the last universal common

ancestor. Nat Rev Microbiol 1:127–136

Koonin EV (2005) Orthologs, paralogs, and evolutionary genomics. Annu Rev Genet 39:309–338

Koonin EV (2009a) Evolution of genome architecture. Int J Biochem Cell Biol 41:298–306

Koonin EV (2009b) Darwinian evolution in the light of genomics. Nucleic Acids Res

37:1011–1034

Koonin EV, Wolf YI (2006) Evolutionary systems biology: links between gene evolution and

function. Curr Opin Biotechnol 17:481–487

Koonin EV, Wolf YI (2008) Genomics of bacteria and archaea: the emerging dynamic view of the

prokaryotic world. Nucleic Acids Res 36(21):6688–6719

Koonin EV, Mushegian AR, Bork P (1996) Non-orthologous gene displacement. Trends Genet

12:334–336

Koonin EV, Aravind L, Kondrashov AS (2000) The impact of comparative genomics on our

understanding of evolution. Cell 101:573–576


Koonin EV, Wolf YI, Karev GP (2002) The structure of the protein universe and genome

evolution. Nature 420:218–223

Kramer EB, Farabaugh PJ (2007) The frequency of translational misreading errors in E. coli is

largely determined by tRNA competition. RNA 13:87–96

Krylov DM, Wolf YI, Rogozin IB, Koonin EV (2003) Gene loss, protein sequence divergence,

gene dispensability, expression level, and interactivity are correlated in eukaryotic evolution.

Genome Res 13:2229–2235

Lawrence J (1999) Selfish operons: the evolutionary impact of gene clustering in prokaryotes and

eukaryotes. Curr Opin Genet Dev 9:642–648

Lawrence JG, Roth JR (1996) Selfish operons: horizontal transfer may drive the evolution of gene

clusters. Genetics 143:1843–1860

Lemons D, McGinnis W (2006) Genomic evolution of Hox gene clusters. Science 313:1918–1922

Lespinet O, Wolf YI, Koonin EV, Aravind L (2002) The role of lineage-specific gene family

expansion in the evolution of eukaryotes. Genome Res 12:1048–1059

Levy SF, Siegal ML (2008) Network hubs buffer environmental variation in Saccharomyces

cerevisiae. PLoS Biol 6:e264

Ling X, He X, Xin D (2009) Detecting gene clusters under evolutionary constraint in a large

number of genomes. Bioinformatics 25:571–577

Lobkovsky AE, Wolf YI, Koonin EV (2010) Universal distribution of protein evolution rates as a

consequence of protein folding physics. Proc Natl Acad Sci USA 107(7):2983–2988, doi:

10.1073/pnas.0910445107

Loewe L (2009) A framework for evolutionary systems biology. BMC Syst Biol 3:27

Lozada-Chavez I, Janga SC, Collado-Vides J (2006) Bacterial regulatory networks are extremely

flexible in evolution. Nucleic Acids Res 34:3434–3445

Lunter G, Ponting CP, Hein J (2006) Genome-wide identification of human functional DNA using

a neutral indel model. PLoS Comput Biol 2:e5

Lynch M (2006a) The origins of eukaryotic gene structure. Mol Biol Evol 23:450–468

Lynch M (2006b) Streamlining and simplification of microbial genome architecture. Annu Rev

Microbiol 60:327–349

Lynch M (2007a) The evolution of genetic networks by non-adaptive processes. Nat Rev Genet

8:803–813

Lynch M (2007b) The frailty of adaptive hypotheses for the origins of organismal complexity.

Proc Natl Acad Sci USA 104(Suppl 1):8597–8604

Lynch M (2007c) The origins of genome architecture. Sinauer Associates, Sunderland, MA

Lynch M, Conery JS (2003) The origins of genome complexity. Science 302:1401–1404

Makalowski W, Boguski MS (1998) Synonymous and nonsynonymous substitution distances are

correlated in mouse and rat genes. J Mol Evol 47:119–121

Mani GS, Clarke BC (1990) Mutational order: a major stochastic process in evolution. Proc R Soc

Lond B Biol Sci 240:29–37

Margulies EH, Cooper GM, Asimenos G, Thomas DJ, Dewey CN, Siepel A, Birney E, Keefe D,

Schwartz AS, Hou M, Taylor J, Nikolaev S, Montoya-Burgos JI, Loytynoja A, Whelan S,

Pardi F, Massingham T, Brown JB, Bickel P, Holmes I, Mullikin JC, Ureta-Vidal A, Paten B,

Stone EA, Rosenbloom KR, Kent WJ, Bouffard GG, Guan X, Hansen NF, Idol JR, Maduro

VV, Maskeri B, McDowell JC, Park M, Thomas PJ, Young AC, Blakesley RW, Muzny DM,

Sodergren E, Wheeler DA, Worley KC, Jiang H, Weinstock GM, Gibbs RA, Graves T, Fulton

R, Mardis ER, Wilson RK, Clamp M, Cuff J, Gnerre S, Jaffe DB, Chang JL, Lindblad-Toh K,

Lander ES, Hinrichs A, Trumbower H, Clawson H, Zweig A, Kuhn RM, Barber G, Harte R,

Karolchik D, Field MA, Moore RA, Matthewson CA, Schein JE, Marra MA, Antonarakis SE,

Batzoglou S, Goldman N, Hardison R, Haussler D, Miller W, Pachter L, Green ED, Sidow A

(2007) Analyses of deep mammalian sequence alignments and constraint predictions for 1%

of the human genome. Genome Res 17:760–774

Masel J, Siegal ML (2009) Robustness: mechanisms and consequences. Trends Genet 25:395–403

Maslov S, Krishna S, Pang TY, Sneppen K (2009) Toolbox model of evolution of prokaryotic

metabolic networks and their regulation. Proc Natl Acad Sci USA 106:9743–9748


Mayrose I, Friedman N, Pupko T (2005) A gamma mixture model better accounts for among site

rate heterogeneity. Bioinformatics 21(Suppl 2):ii151–ii158

Medina M (2005) Genomes, phylogeny, and evolutionary systems biology. Proc Natl Acad Sci

USA 102(Suppl 1):6630–6635

Molina N, van Nimwegen E (2008) Universal patterns of purifying selection at noncoding

positions in bacteria. Genome Res 18:148–160

Molina N, van Nimwegen E (2009) Scaling laws in functional genome content across prokaryotic

clades and lifestyles. Trends Genet 25:243–247

Monot M, Honore N, Garnier T, Zidane N, Sherafi D, Paniz-Mondolfi A, Matsuoka M, Taylor GM,

Donoghue HD, Bouwman A, Mays S, Watson C, Lockwood D, Khamispour A, Dowlati Y,

Jianping S, Rea TH, Vera-Cabrera L, Stefani MM, Banu S, Macdonald M, Sapkota BR,

Spencer JS, Thomas J, Harshman K, Singh P, Busso P, Gattiker A, Rougemont J, Brennan PJ,

Cole ST (2009) Comparative genomic and phylogeographic analysis of Mycobacterium leprae.

Nat Genet 41:1282–1289

Moya A, Gil R, Latorre A, Pereto J, Pilar Garcillan-Barcia M, de la Cruz F (2009) Toward minimal

bacterial cells: evolution vs. design. FEMS Microbiol Rev 33:225–235

Mushegian AR, Koonin EV (1996a) Gene order is not conserved in bacterial evolution. Trends

Genet 12:289–290

Mushegian AR, Koonin EV (1996b) A minimal gene set for cellular life derived by comparison of

complete bacterial genomes [see comments]. Proc Natl Acad Sci USA 93:10268–10273

Mustonen V, Lassig M (2009) From fitness landscapes to seascapes: non-equilibrium dynamics

of selection and adaptation. Trends Genet 25:111–119

Mustonen V, Lassig M (2010) Fitness flux and ubiquity of adaptive evolution. Proc Natl Acad Sci

USA 107(9):4248–4253

Muzzi A, Moschioni M, Covacci A, Rappuoli R, Donati C (2008) Pilus operon evolution in

Streptococcus pneumoniae is driven by positive selection and recombination. PLoSONE 3:e3660

Nakabachi A, Yamashita A, Toh H, Ishikawa H, Dunbar HE, Moran NA, Hattori M (2006) The

160-kilobase genome of the bacterial endosymbiont Carsonella. Science 314:267

Nielsen R (2001) Statistical tests of selective neutrality in the age of genomics. Heredity

86:641–647

Nielsen R (2005) Molecular signatures of natural selection. Annu Rev Genet 39:197–218

Nielsen R, Bustamante C, Clark AG, Glanowski S, Sackton TB, Hubisz MJ, Fledel-Alon A,

Tanenbaum DM, Civello D, White TJ, Sninsky JJ, Adams MD, Cargill M (2005) A scan for

positively selected genes in the genomes of humans and chimpanzees. PLoS Biol 3(6):e170

Nilsen TW, Graveley BR (2010) Expansion of the eukaryotic proteome by alternative splicing.

Nature 463:457–463

Novais A, Comas I, Baquero F, Canton R, Coque TM, Moya A, Gonzalez-Candelas F, Galan JC

(2010) Evolutionary trajectories of beta-lactamase CTX-M-1 cluster enzymes: predicting

antibiotic resistance. PLoS Pathog 6(1):e1000735

Novichkov PS, Ratnere I, Wolf YI, Koonin EV, Dubchak I (2009a) ATGC: a database of

orthologous genes from closely related prokaryotic genomes and a research platform for

microevolution of prokaryotes. Nucleic Acids Res 37:D448–D454

Novichkov PS, Wolf YI, Dubchak I, Koonin EV (2009b) Trends in prokaryotic evolution revealed

by comparison of closely related bacterial and archaeal genomes. J Bacteriol 191:65–73

Ohno S (1970) Evolution by gene duplication. Springer-Verlag, Berlin-Heidelberg-New York

Osbourn AE, Field B (2009) Operons. Cell Mol Life Sci 66:3755–3775

Pal C, Papp B, Hurst LD (2001) Highly expressed genes in yeast evolve slowly. Genetics

158:927–931

Pal C, Papp B, Lercher MJ (2006) An integrated view of protein evolution. Nat Rev Genet

7:337–348

Parsch J, Novozhilov S, Saminadin-Peter SS, Wong KM and Andolfatto P (2010) On the utility of

short intron sequences as a reference for the detection of positive and negative selection in

Drosophila. Mol Biol Evol [Epub ahead of print]


Petersen L, Bollback JP, Dimmic M, Hubisz M, Nielsen R (2007) Genes under positive selection

in Escherichia coli. Genome Res 17:1336–1343

Ponjavic J, Ponting CP, Lunter G (2007) Functionality or transcriptional noise? Evidence for

selection within long noncoding RNAs. Genome Res 17:556–565

Ponting CP, Oliver PL, Reik W (2009) Evolution and functions of long noncoding RNAs. Cell

136:629–641

Proux E, Studer RA, Moretti S, Robinson-Rechavi M (2009) Selectome: a database of positive

selection. Nucleic Acids Res 37:D404–D407

Putnam NH, Srivastava M, Hellsten U, Dirks B, Chapman J, Salamov A, Terry A, Shapiro H,

Lindquist E, Kapitonov VV, Jurka J, Genikhovich G, Grigoriev IV, Lucas SM, Steele RE,

Finnerty JR, Technau U, Martindale MQ, Rokhsar DS (2007) Sea anemone genome reveals

ancestral eumetazoan gene repertoire and genomic organization. Science 317:86–94

Queitsch C, Sangster TA, Lindquist S (2002) Hsp90 as a capacitor of phenotypic variation. Nature

417:618–624

Resch AM, Carmel L, Marino-Ramirez L, Ogurtsov AY, Shabalina SA, Rogozin IB, Koonin EV

(2007) Widespread positive selection in synonymous sites of mammalian genes. Mol Biol Evol

24:1821–1831

Rocha EP (2008) The organization of the bacterial genome. Annu Rev Genet 42:211–233

Rogozin IB, Makarova KS, Murvai J, Czabarka E, Wolf YI, Tatusov RL, Szekely LA, Koonin EV

(2002) Connected gene neighborhoods in prokaryotic genomes. Nucleic Acids Res

30:2212–2223

Rogozin IB, Wolf YI, Sorokin AV, Mirkin BG, Koonin EV (2003) Remarkable interkingdom

conservation of intron positions and massive, lineage-specific intron loss and gain in eukaryotic

evolution. Curr Biol 13:1512–1517

Roy SW, Gilbert W (2006) The evolution of spliceosomal introns: patterns, puzzles and progress.


Roy SW, Penny D (2007) Patterns of intron loss and gain in plants: intron loss-dominated evolution

and genome-wide comparison of O. sativa and A. thaliana. Mol Biol Evol 24:171–181

Schrimpf SP, Weiss M, Reiter L, Ahrens CH, Jovanovic M, Malmstrom J, Brunner E, Mohanty S,

Lercher MJ, Hunziker PE, Aebersold R, von Mering C, Hengartner MO (2009) Comparative

functional analysis of the Caenorhabditis elegans and Drosophila melanogaster proteomes.

PLoS Biol 7:e48

Sella G, Petrov DA, Przeworski M, Andolfatto P (2009) Pervasive natural selection in the

Drosophila genome? PLoS Genet 5:e1000495

Shabalina SA, Kondrashov AS (1999) Pattern of selective constraint in C. elegans and C. briggsae

genomes. Genet Res 74:23–30

Shabalina SA, Koonin EV (2008) Origins and evolution of eukaryotic RNA interference. Trends

Ecol Evol 23:578–587

Shabalina SA, Ogurtsov AY, Rogozin IB, Koonin EV, Lipman DJ (2004) Comparative analysis of

orthologous eukaryotic mRNAs: potential hidden functional signals. Nucleic Acids Res

32:1774–1782

Srivastava M, Begovic E, Chapman J, Putnam NH, Hellsten U, Kawashima T, Kuo A, Mitros T,

Salamov A, Carpenter ML, Signorovitch AY, Moreno MA, Kamm K, Grimwood J, Schmutz J,

Shapiro H, Grigoriev IV, Buss LW, Schierwater B, Dellaporta SL, Rokhsar DS (2008) The

Trichoplax genome and the nature of placozoans. Nature 454:955–960

Stanek MT, Cooper TF, Lenski RE (2009) Identification and dynamics of a beneficial mutation in a

long-term evolution experiment with Escherichia coli. BMC Evol Biol 9:302

Tatusov RL, Koonin EV, Lipman DJ (1997) A genomic perspective on protein families. Science

278:631–637

Tsaparas P, Marino-Ramirez L, Bodenreider O, Koonin EV, Jordan IK (2006) Global similarity

and local divergence in human and mouse gene co-expression networks. BMC Biol 6:70

Turner LM, Chuong EB, Hoekstra HE (2008) Comparative analysis of testis protein evolution in

rodents. Genetics 179:2075–2089


van Nimwegen E (2003) Scaling laws in the functional content of genomes. Trends Genet

19:479–484

Wagner A (2005) Robustness, evolvability, and neutrality. FEBS Lett 579:1772–1778

Wagner A (2008) Neutralism and selectionism: a network-based reconciliation. Nat Rev Genet

9:965–974

Waterston RH, Lindblad-TohK, Birney E, Rogers J, Abril JF, Agarwal P, Agarwala R, Ainscough R,

Alexandersson M, An P, Antonarakis SE, Attwood J, Baertsch R, Bailey J, Barlow K, Beck S,

Berry E, Birren B, Bloom T, Bork P, Botcherby M, Bray N, Brent MR, Brown DG, Brown SD,

Bult C, Burton J, Butler J, Campbell RD, Carninci P, Cawley S, Chiaromonte F, Chinwalla AT,

Church DM, Clamp M, Clee C, Collins FS, Cook LL, Copley RR, Coulson A, Couronne O,

Cuff J, Curwen V, Cutts T, Daly M, David R, Davies J, Delehaunty KD, Deri J, Dermitzakis ET,

Dewey C, Dickens NJ, Diekhans M, Dodge S, Dubchak I, Dunn DM, Eddy SR, Elnitski L,

Emes RD, Eswara P, Eyras E, Felsenfeld A, Fewell GA, Flicek P, Foley K, Frankel WN, Fulton

LA, Fulton RS, Furey TS, Gage D, Gibbs RA, Glusman G, Gnerre S, Goldman N, Goodstadt L,

Grafham D, Graves TA, Green ED, Gregory S, Guigo R, Guyer M, Hardison RC, Haussler D,

Hayashizaki Y, Hillier LW, Hinrichs A, HlavinaW,Holzer T, Hsu F, Hua A, Hubbard T, Hunt A,

Jackson I, Jaffe DB, Johnson LS, Jones M, Jones TA, Joy A, Kamal M, Karlsson EK, Karolchik

D, Kasprzyk A, Kawai J, Keibler E, Kells C, Kent WJ, Kirby A, Kolbe DL, Korf I, Kucherlapati

RS, Kulbokas EJ, Kulp D, Landers T, Leger JP, Leonard S, Letunic I, Levine R, Li J, LiM, Lloyd

C, Lucas S, Ma B, Maglott DR, Mardis ER, Matthews L, Mauceli E, Mayer JH, McCarthy M,

McCombie WR, McLaren S, McLay K, McPherson JD, Meldrim J, Meredith B, Mesirov JP,

Miller W,Miner TL, Mongin E, Montgomery KT, MorganM, Mott R, Mullikin JC, Muzny DM,

NashWE, Nelson JO, NhanMN, Nicol R, Ning Z, NusbaumC,O’ConnorMJ, Okazaki Y, Oliver

K, Overton-Larty E, Pachter L, Parra G, Pepin KH, Peterson J, Pevzner P, Plumb R, Pohl CS,

Poliakov A, Ponce TC, Ponting CP, Potter S, Quail M, Reymond A, Roe BA, Roskin KM, Rubin

EM, Rust AG, Santos R, Sapojnikov V, Schultz B, Schultz J, Schwartz MS, Schwartz S, Scott C,

Seaman S, Searle S, Sharpe T, Sheridan A, Shownkeen R, Sims S, Singer JB, Slater G, Smit A,

Smith DR, Spencer B, Stabenau A, Stange-Thomann N, Sugnet C, Suyama M, Tesler G,

Thompson J, Torrents D, Trevaskis E, Tromp J, Ucla C, Ureta-Vidal A, Vinson JP, Von

Niederhausern AC, Wade CM, Wall M, Weber RJ, Weiss RB, Wendl MC, West AP, Wetter-

strand K, Wheeler R, Whelan S, Wierzbowski J, Willey D, Williams S, Wilson RK, Winter E,

Worley KC, Wyman D, Yang S, Yang SP, Zdobnov EM, Zody MC, Lander ES (2002) Initial

sequencing and comparative analysis of the mouse genome. Nature 420:520–562

Weinreich DM, Delaney NF, Depristo MA, Hartl DL (2006) Darwinian evolution can follow only

very few mutational paths to fitter proteins. Science 312:111–114

Wilkins AS (2007) Between “design” and “bricolage”: genetic networks, levels of selection, and

adaptive evolution. Proc Natl Acad Sci USA 104(Suppl 1):8590–8596

Wolf YI, Carmel L, Koonin EV (2006) Unifying measures of gene function and evolution. Proc

Biol Sci 273:1507–1515

Wolf YI, Novichkov PS, Karev GP, Koonin EV, Lipman DJ (2009) The universal distribution of

evolutionary rates of genes and distinct characteristics of eukaryotic genes of different apparent

ages. Proc Natl Acad Sci USA 106:7273–7280

Wolf YI, Gopich IV, Lipman DJ, Koonin EV (2010) Relative contributions of intrinsic structural-

functional constraints and translation rate to the evolution of protein-coding genes. Genome

Biol Evol 2010:190–199

Worth CL, Gong S, Blundell TL (2009) Structural and functional constraints in the evolution of

protein families. Nat Rev Mol Cell Biol 10:709–720

Wuchty S, Almaas E (2005) Evolutionary cores of domain co-occurrence networks. BMC Evol

Biol 5:24

Yamada T, Bork P (2009) Evolution of biomolecular networks: lessons from metabolic and

protein interactions. Nat Rev Mol Cell Biol 10:791–803

Zhou T, Weems M, Wilke CO (2009) Translationally optimal codons associate with structurally

sensitive sites in proteins. Mol Biol Evol 26:1571–1580


Chapter 3

Starvation-Induced Reproductive Isolation

in Yeast

Eugene Kroll, R. Frank Rosenzweig, and Barbara Dunn

Abstract Speciation in eukaryotes is one of the central issues in evolutionary

biology. Retrospective studies of existing species may not reveal the molecular

events underlying speciation, as it is frequently impossible to distinguish changes

which preceded speciation from those which happened after speciation has

occurred. We propose a model for experimental speciation using a well-studied

Eukaryotic organism, the yeast Saccharomyces cerevisiae, and starvation as an

agent of speciation. Starvation can be viewed as a general and widespread conse-

quence of catastrophic environmental change that leads to a decrease in survival or

reproductive success. We find that yeast populations subjected to a month-long

starvation exhibit a drastic increase in genomic rearrangements compared with a

modest increase in point mutation. We subsequently find that starved yeast popula-

tions become reproductively isolated from their ancestor, which we attribute to

chromosomal abnormalities in the starved clones’ genomes. Our model provides

direct molecular evidence – that speciation can rapidly occur without the precondi-

tion of geographic separation or divergent selection.

3.1 Continuing Uncertainty over Species Definitions

Among the Eukarya

Two central questions in eukaryotic evolutionary biology are: how do new species

emerge and how are they perpetuated? We can provisionally define a species as

group of organisms that shares a complex genetic network of interacting alleles and

E. Kroll, and R.F. Rosenzweig

Division of Biological Sciences, University of Montana, 32 campus dr., Missoula, MT 59812,

USA


B. Dunn

Department of Genetics, Stanford University, Stanford, CA 94305, USA


49

preserves its integrity by restricting the exchange of genetic material with other

such networks (Mayr 1966). The processes by which new networks emerge, i.e.,

speciation, appear to be diverse and their relative contributions remain the subject

of considerable controversy. While Darwin explicitly linked the process of spe-

ciation to the adaptation of organisms to novel environments (Darwin 1859, Ch. 4),

neo-Darwinists have emphasized the role of interpopulation isolation (Fisher 1930;

Dobzhansky 1937; Muller 1940; Mayr 1942). Uncertainty persists as to which of

these emphases is correct (Lande 1989; Vulic et al. 1999; Orr and Presgraves 2000;

Schilthuizen 2000; Turelli et al. 2001; Sinervo and Svensson 2002; Herrmann et al.

2003), largely due to the dearth of knowledge about the specific molecular mechan-

isms that underlie eukaryotic speciation and the fact that species can be defined in

various ways.

The most widely used definition for speciation is based on the biological

species concept, i.e., the cessation of gene flow between groups of organisms,

or “reproductive isolation” (Dobzhansky 1937; Mayr 1942, 1996; Lande 1989;

Coyne and Orr 1998). Though this definition is not universally accepted

(Darwin 1859, Ch. 8; Schilthuizen 2000 and refs. therein), it is, due to its

inherently measurable nature, the most amenable framework for experimental

investigation. Relative reproductive isolation between two species is a quanti-

tative trait that can be measured as a ratio between fertilities of interspecific

hybrids and their conspecific parentals. Importantly, “relative reproductive

isolation” can be used as a proxy to assess divergence between closely related

organisms.

Reproductive isolation in sexual species can be both pre and postzygotic. To

date, efforts to explain incipient speciation in eukaryotes have focused on pre-

zygotic isolation mechanisms such as spatial/temporal or behavioral separation

(Orr and Presgraves 2000). In many cases, the former arises in allopatry, whereas

the latter is viewed as a reinforcing mechanism. That said, much theoretical

and experimental work now indicates that postzygotic mechanisms, e.g., the

inviability or infertility of interspecific hybrids, can play crucial roles in initiating

reproductive isolation, and that prezygotic mechanisms might therefore evolve at

a later stage (Lande 1989; Schliewen et al. 1994; Dieckmann and Doebeli 1999;

Schilthuizen 2000; Turelli et al. 2001; Via 2001). Hence, it may be difficult to

uncover, using existing species as evidence, the important transformative events

that initiate speciation, as genetic divergence following reproductive isolation is

likely to obscure initial steps in the process. In other words, most of the geneticdifferences that separate contemporary species by enforcing isolation may not bethe differences that originally caused speciation. Overall, our research goal is to

elucidate the exact molecular mechanisms that bring about speciation in a model

eukaryote and to do so in real time under controlled laboratory conditions. We

contend that an experimental rather than a comparative approach is more likely

to enable us to clarify the role of postzygotic mechanisms in the initial stages of

speciation.

50 E. Kroll et al.

3.2 The Nature of Postzygotic Reproductive Isolation

in Eukaryotes

Postzygotic reproductive isolation manifests in the inviability or infertility of

hybrid progeny (Orr and Presgraves 2000). While hybrid inviability can be caused

by developmental incompatibilities or dysgenesis (Hartl et al. 1997), hybrid infer-

tility is likely a consequence of defective hybrid meiosis. Establishment of post-

zygotic reproductive isolation in eukaryotes has been explained by one of two

competing theories. One, “the chromosomal theory of speciation,” holds that

chromosomal changes (genomic rearrangements) disrupt recombination and segre-

gation of homologues in meiosis I, and/or fine-scale mutations disrupt meiotic

recombination via the action of mismatch repair (White 1978; King 1993; Radman

and Wagner 1993; Chambers et al. 1996; Searle 1998; Britton-Davidian et al. 2000;

Rieseberg 2001). The other, “the genic theory of speciation” (“speciation genes”)

holds that genic changes, e.g., functional incompatibilities between diverged

alleles, result in lower hybrid fitness (Bateson 1909; Dobzhansky 1937; Muller

1940; Coyne and Orr 1998). A third possibility, the idea that postzygotic isolation

occurs due to a combination of these two theories described above, has also been

proposed (Henikoff et al. 2001; Noor et al. 2001; Rieseberg 2001).

The chromosomal theory of speciation did not gain acceptance during the early

studies on postzygotic reproductive isolation due to two reasons: first, pioneering

experiments in Drosophila by Dobzhansky appeared to demonstrate the genic

nature of postzygotic isolation (Dobzhansky 1933); second, chromosomal specia-

tion appears to be incompatible due to the “underdominance” effect, wherein an

individual is rendered less fertile if it sustains a chromosomal rearrangement, and it

thus would not be able to form a new species (Livingstone and Rieseberg 2004).

The archetypal experiments by T. Dobzhansky first demonstrated that the

sterility of male hybrids formed as a result of interbreeding between two races

of Drosophila pseudoobscura distinguished by several chromosomal rearrange-

ments was due to mis-segregation of homologous chromosomes in meiosis I

(Dobzhansky 1933). Dobzhansky further noted that in rare instances when tetra-

ploid spermatogonia were found in these interracial hybrids, chromosomes also

had mis-segregated in meiosis I. From this observation, Dobzhansky deduced

that because every chromosome in tetraploid hybrid meioses is furnished with its

exact homologue, tetraploidization should have restored faithful segregation of

homologues if mis-segregation had been caused by chromosomal rearrangements

and not by genic incompatibilities. Thus, he concluded that genic incompatibi-

lities, not chromosomal changes, caused hybrid sterility to occur in the male

hybrids of two races of D. pseudoobscura. The ensuing rush for “speciation

genes” or, rather, incompatible alleles, did render some tangible results, notably

from the cloning of Odysseus, a gene encoding a homeobox protein responsible

for interspecific incompatibilies in Drosophila (Ting et al. 1998; Greenberg et al.

2003) and several more genes that control hybrid infertility (Lee et al. 2008;

Phadnis and Orr 2009).

3 Starvation-Induced Reproductive Isolation in Yeast 51

However, there is absolutely no way to make certain that such incompatible

alleles were the actual reason for speciation and not merely the product of species

divergence; in other words, finding speciation genes is not in fact a proof that

speciation ultimately has a genic nature. Intriguingly, in a footnote to his pioneering

paper on the genic nature of reproductive isolation mentioned above, Dobzhansky

acknowledges that he did not report the results of the reciprocal cross, which is “dif-

ferent in many important details” and would be “published elsewhere” (Dobzhansky

1933).

In stark contrast to the aforementioned studies of Dobzhansky, Noor and collea-

gues used the very same species of Drosophila to directly implicate large chromo-

somal inversions in the reproductive isolation between sympatric D. pseudoobscuraand D. persimilis populations (Noor et al. 2001). Indeed, inversions and other small

rearrangements that may have a deleterious effect on meiosis have been shown to

be abundant between related species in many species of yeast (Seoighe et al. 2000;

Kellis et al. 2003; Fischer et al. 2006), as well as in roundworms (Hutter et al.

2000), mice (Hauffe and Searle 1998), plants (Blanc et al. 2000), and a variety of

other organisms (for a review: Eichler and Sankoff 2003).

Experiments on tetraploidization in several species of plants showed that certain

types of chromosomal rearrangements were responsible for postzygotic reproductive

isolation (Anderson 1949; White 1978; Searle 1998; Pialek et al. 2001; Rieseberg

2001). Chromosomal rearrangements have also been implicated in human evolution,

acting to decrease gene flow in the chromosomal regions that harbor inversions

(Navarro and Barton 2003). In Saccharomyces cerevisiae, chromosomal inversions

have been shown to directly and efficiently impair the progression of meiosis

(Dresser et al. 1994; Jinks-Robertson et al. 1997; Chen and Jinks-Robertson 1999).

As for the concept of underdominance – a decrease in, or lack of, the ability to go

through meiosis due to one or more heterozygous rearrangement – overshadowing

the chromosomal speciation theory, it is fair to say that different genomic rearran-

gements may have very different effects on meiosis, ranging from irrelevant to

prohibitive, with all shades in between. Clearly, an organism that contains a chro-

mosomal rearrangement that abrogates meiosis is not going to form a new species;

however, a partial restriction of gene flow resulting from a rearrangement could

allow for faster rates of sequence and functional divergence (Lande 1989; Noor

et al. 2001; Rieseberg 2001; Navarro and Barton 2003), increasing the probability

of speciation. Finally, using the same logic that is used for epistasis in speciation

genes (Bateson 1909), genomic rearrangements can also form incompatible pairs,

further destabilizing meiosis in hybrid organisms.

Assuming that the experimental observations supporting both theories of post-

zygotic isolation are correct, should one conclude that these opposing results reflect

variation in experimental techniques, or are they more readily explained as varia-

tions between diverse taxa? And is it then reasonable to assume that both the genic

and chromosomal models of speciation (acting in concert or separately in different

taxa) can act in the process of speciation? To address these questions experimen-

tally, we have developed a laboratory assay using the yeast S. cerevisiae to isolate

reproductively separated clones during the course of prolonged starvation.

52 E. Kroll et al.

3.3 A Starvation-Based Experimental Model May Help

Resolve Uncertainties Concerning the Molecular

Basis for Speciation

Comparative analyzes of existing species may poorly discriminate between

changes that cause speciation and those that arise secondarily (Schilthuizen

2000). However, experimental evidence obtained under conditions physiologi-

cally close to optimal may be difficult to acquire as these conditions typically

result in low and constant mutation rates (Drake et al. 1998) that make speciation

less likely to occur (Rice and Hostert 1993). We have therefore developed an

experimental laboratory model to study speciation that uses prolonged starvation

as a proxy for sudden and severe environmental change. This treatment effectively

disrupts normal living conditions, disintegrating a population’s niche, over time

diminishing its mean fitness, measured as both survivorship and reproductive

capacity.

Furthermore, starvation is a condition that virtually all species experience and

that many contend with regularly in the wild (Koch 1971; Death and Ferenci

1994). All manner of environmental change, such as wildfire, flood, sudden trans-

fer to a new habitat, or even the invasion of a competitive species can bring about

starvation. We hypothesize that because starvation is universally experienced in

the wild owing to a plethora of circumstances, natural selection has brought

about mechanisms that respond to this generic signal in ways that may increase

population diversity via increased mutations, including large-scale genome

rearrangements.

3.4 Starvation-Responses That Could Increase Population

Genetic Diversity

Escherichia coli’s SOS system activates multiple responses to DNA damage,

nutrient starvation, and low temperature that are both mutagenic and recombino-

genic (Witkin and Wermundsen 1979; Dri and Moreau 1993; Friedberg et al. 1995;

McKenzie et al. 2000). Following activation of the SOS system, bacteria sustain a

high frequency of random mutation, rearrangements, and transposition (Radman

1975; Witkin 1976; Petit et al. 1991; Guerin et al. 2009), revealing a genetic link

between stress caused by highly challenging environmental conditions and varia-

bility (Taddei et al. 1997). In fact, it has been shown that starvation-induced muta-

genesis in bacteria is directly controlled by the SOS system (Taddei et al. 1995;

Hastings et al. 2000; McKenzie et al. 2000; Finkel 2006; He et al. 2006) as well as

by global stress response (Zinser and Kolter 1999; Bjedov et al. 2003; Lombardo

et al. 2004).

Eukaryotes possess a combination of genetic pathways that may be functionally

analogous to those of bacteria, such as checkpoint adaptation, translesion synthesis,


stress signaling, and others (Toczyski et al. 1997; Kai and Wang 2003; Smets et al.

2010). However, although the causal connection between environmental stress and

an increase in adaptively significant variation has been well studied, the molecular

basis for such connection in eukaryotes remains obscure. By employing starvation

to mimic severe stress, we hope to model conditions in nature with which all

populations must contend (Death and Ferenci 1994) and to discover molecular

mechanisms that link catastrophic environmental change with the types of genetic

variation that could lead to speciation.

3.5 Advantages of Using Yeast as Model to Study

Speciation in Real Time

Several factors contributed to our choice of S. cerevisiae as a model organism.

S. cerevisiae is a well-studied organism that possesses most of the major signal

transduction (Smets et al. 2010) and DNAmaintenance pathways (San Filippo et al.

2008) found in other eukaryotes. Also, the genomes of multiple strains of

S. cerevisiae and more than ten-related species have been sequenced. Lastly,

yeast genetics, especially as it relates to DNA maintenance, cell cycle, checkpoints

and stress resistance, is well understood. In S. cerevisiae, as in higher eukaryotes,

the controlled occurrence of DNA double-strand breaks early in meiotic prophase is

essential for the maturation of the synaptonemal complex as well as for chiasmata

formation in diplotene and for faithful homologue segregation at anaphase I

(Peoples et al. 2002; Page and Hawley 2003). This dependence is reinforced by

the pachytene checkpoint (Roeder and Bailis 2000), which ensures that meiotic

recombination and homologue synapsis are completed before cells proceed to

metaphase I. In contrast, the chromosomes in another well-studied yeast species,

Schizosaccharomyces pombe, do not form synaptonemal complexes in meiosis

(Davis and Smith 2003); while in the popular multicellular model organisms

C. elegans and D. melanogaster, double-strand breaks are not required for chromo-

some synapsis to occur (Dernburg et al. 1998; Jang et al. 2003). Moreover,

heterogametic (male) meioses in Drosophila and other Diptera and Lepidoptera

occur in the complete absence of recombination (Hawley 2002). Thus, among

favored models systems, the processes of meiosis in S. cerevisiae most resemble

those found within meioses of mouse and human spermatocytes (Lichten 2001;

Page and Hawley 2003).

Finally, in S. cerevisiae, reproductive isolation manifests as a quantitative trait

that can be scored as the efficiency of producing viable spores or spore yield

(a combination of sporulation efficiency and spore viability). We chose an S288c

strain [BY4743 (Brachmann et al. 1998)] for our speciation studies because this

diploid, unlike other laboratory strains, does not spontaneously sporulate when

starved, and thus starved diploids that have not gone through meiosis can be reliably

obtained.

54 E. Kroll et al.

3.6 Three Modes of PostZygotic Isolation in Yeast – Sequence,

Chromosome, Breakpoint-Recombination

The six nonhybrid species that comprise the sensu stricto group of Saccharomyces(S. cerevisiae, S. paradoxus, S. mikatae, S. cariocanus, S. kudriavzevii, andS. bayanus) show large genomic rearrangements relative to each other, as detected

by pulsed-field gel analysis, with the exception of S. cerevisiae and S. paradoxuswhich are almost identical. Fischer et al. showed that these rearrangements did not

in fact correspond to a phylogenetic tree based on sequence divergence of rRNA

(Fischer et al. 2000, 2006), and thus concluded that genomic rearrangements were

unimportant in the speciation of yeast. Interestingly, the restoration of the colinear-

ity of gene order between two sensu stricto species, S. cerevisiae and S. mikatae, didlead to a partial restoration of the interspecific hybrid fertility (Delneri et al. 2003),

indicating that genomic rearrangements are important for the maintenance of the

postzygotic reproductive isolation in yeast.

Mutational load and the action of the mismatch repair system also affect, albeit

partially, reproductive isolation between S. cerevisiae and S. paradoxus (Chambers

et al. 1996; Chen and Jinks-Robertson 1999), as crossing-over in yeast is dependent on

sequence homology between homeologous chromosomes (Hunter et al. 1996). How-

ever, experiments suggesting these possibilities were conducted with extant species,

where genetic changes such as sequence divergence – proposed as a possible cause for a

reproductive barrier – may actually have occurred after the speciation event and thus

might not be a reason for the initial reproductive barrier. Additionally, dominant

epistatic incompatibilities between two sensu stricto species of Saccharomyces havebeen shown not to be important for speciation by either tetraploidization experiments

(Greig et al. 2002) or directly checking for speciation genes (Greig and Leu 2009).

Although one pair of incompatible alleles has been recently identified between

S. cerevisiae and S. bayanus (Lee et al. 2008), it is again unclear whether this incompat-

ibility was a driving force, or a secondary consequence, of the initial speciation event.

Chromosome rearrangements are plentiful in yeast genomes. Genomic rearran-

gements, such as reciprocal translocations, transpositions, insertions, deletions, and

inversions, are ubiquitous features of even closely related species. Studies using

pulsed-field gel analysis and hybridization, such as Fischer et al. (Fischer et al.

2000), identified only a small subset of all rearrangements and inversions among

the sensu stricto species – as shown by subsequent whole genome sequencing –

because smaller rearrangements and inversions simply cannot be resolved by

pulsed-field gels. Remarkably, of all the syntenic breakpoints between S. cerevisiaeand S. bayanus, less than 10% are large-scale rearrangements (Fischer et al. 2001).

Sequence data from the S. bayanus, S. mikatae, and S. paradoxus genomes have

revealed many more genomic rearrangements than were previously known, espe-

cially at chromosome ends (Kellis et al. 2003). The nine inversions that exist

between the genomes of these three species and S. cerevisiae are flanked by

tRNA genes, usually of the same isoacceptor type (Kellis et al. 2003). This finding

suggests that inversions and perhaps other rearrangements that have accumulated in


the genomes of the Saccharomyces spp. arose via homologous recombination. An

alternative hypothesis is that rearrangements may have been caused by yeast retro-

transposons (Ty), as the tRNA genes are hotspots for Ty1, 3, and 5 transposition

(Natsoulis et al. 1989). In addition, nonhomologous end-joining may have played a

role in creating some of the rearrangements, as has been observed among flor yeast

used in fortified winemaking (Infante et al. 2003).

Thus, in our opinion, certain genomic rearrangements that include small and large

inversions, small translocations, and small insertion–deletions that escape detection

by pulsed-field gel analysis (but discovered later by sequencing) may be a ubiquitous

feature of evolving genomes. We further suggest that such rearrangements may play

a key role in incipient speciation among yeasts and other Eukaryotes.

3.7 Starved Yeast Cultures Sustain High Frequencies

of Genomic Rearrangements

In extant species of Saccharomyces yeast, the rates of genomic rearrangements

are highly variable (Fischer et al. 2006). We contend that starvation as a result of

environmental change can affect the rates of genomic variation. Moreover, we have

already shown that a champagne strain, DB146, sustains a massive amount of

change in genomic architecture after prolonged starvation (Coyle and Kroll 2008).

To appraise the effect of prolonged starvation on genomic change, we starved

multiple random clones of the laboratory yeast diploid BY4743 (Brachmann et al.

1998), essentially as described (Coyle and Kroll 2008). During a 1-month-long

starvation treatment, and accounting for diminished viability, the starving cultures

underwent an average of ten generations. At no point did we observe sporulating

cells in starving cultures. For comparison, we established a control by growing

BY4743 cells in rich medium for approximately twice the number of generations

that starved cultures underwent. Because, strictly speaking, the cells obtained at the

end of these �20 generations are neither ancestral nor “wild-type” to the starved

cultures, we chose to call them “nonstarved” cultures.

3.7.1 Starved Cultures Sporulate at Lower LevelThan the Nonstarved Cultures

Genomic rearrangements may create a reproductive barrier between two popula-

tions, as discussed previously. If a reproductive barrier existed between our starved

and ancestral populations, it would manifest as decreased fertility of starved cultures

in backcrosses between haploid progeny of the starved and ancestral populations

when compared with the values for nonstarved to ancestral backcrosses. Both

efficiency of sporulation (the frequency at which yeast cells form gametes or spores)

and spore viability (colony-forming units per number of spores plated) could be

56 E. Kroll et al.

expected to affect hybrid fertility. Generally, only a partial measure of fertility –

spore viability – is measured in crosses between separate yeast species (Naumov

et al. 2000). Since different species usually require different conditions for optimal

sporulation, sporulation efficiency of the interspecific hybrid lacks an obvious

control. However, in our case, we used only one ancestral strain, and thus we

were able to assess both sporulation efficiency and spore viability of the backcross

hybrids. To score these traits, we incubated the cells overnight in fresh rich medium

to minimize the fraction of dead cells in starved cultures, then sporulated them using

conditions optimized for the ancestral strain, We scored sporulation efficiency and

the viability of the resultant spores. For all comparisons we used nonparametric

statistical tests, as we could not assume normal distribution for our data.

Nonstarved diploid cultures sporulated at the efficiency characteristic of the

BY4743 ancestor and spore viability was nearly 100%. In contrast, starved

BY4743 cultures sporulated about at half the frequency of the nonstarved cultures,

even after prolonged sporulation (Coyle et al. in preparation). Nevertheless, spore

viability among sporulated starved cultures was almost as high as that of spores

derived from the nonstarved cultures (Fig. 3.1).

The fact that starved cultures exhibited significantly lower sporulation efficiency

than nonstarved control suggests the possibility that accumulated changes in the

genomes of starved cultures alter their fertility. Viable spores derived from such

starved cells might be wholly or partially reproductively isolated from each other

and from the ancestral population.

3.7.2 A Subset of Starved Backcrosses Show Lower FertilityThan the Nonstarved Backcrosses

To test how reproductive isolation was distributed within starved cultures we

assessed the fertility of the backcrossed hybrids. We isolated rare spores from

Fig. 3.1 Starved and nonstarved cultures of BY4743 sporulated for 2 days. Arrows denote sporesacks (asci) that contain three or four spores. (a) Starved diploid culture. Only one misshapen spore

sack (ascus) is shown (arrow). (b) Nonstarved culture. The majority of cells have formed asci


1 month starved cultures, germinated those spores into haploid strains, or “starved

isolates” and performed backcrosses. We then sporulated the resultant backcross

hybrids and measured their sporulation efficiency and spore viability; finally, we

compared their hybrid fertility with that of the nonstarved isolates.

The results recapitulate the previous findings for starved diploids: multiple back-

crossed hybrids exhibited significantly lower average sporulation efficiency than the

nonstarved backcrosses (Mann–Whitney U test). Specifically, about one-third of

starved isolates used for the backcross analysis showed a sporulation efficiency that

was significantly lower than those of their respective nonstarved intercrosses (Coyle

et al. in preparation). In contrast to sporulation efficiency, spore viability in all cases

was indistinguishable from the ancestral (Coyle et al. in preparation).

3.7.3 Starved Isolates Reproductively Isolated from theAncestral Population Are Self-Fertile

Complete inability to undergo meiosis would prevent the establishment of a new

species. This might be caused either by mutations in genes important for meiosis or

by a chromosome aberration that prohibits meiosis. To ensure that the starved isolates

could have found a new lineage, capable of sexual reproduction, we selfed starved

isolates that exhibited lower fertility in backcrosses. To do this, we made haploid

progeny of those starved isolates homothallic and isolated their selfed diploid prog-

eny. After sporulating these selfed diploids we found that their sporulation efficiency

was significantly higher than the fertility of the backcross hybrid (Coyle et al. in

preparation). We concluded that starved isolates reproductively isolated from the

ancestral population were self-fertile and able to form new sexually reproducing

lineages, that is, new biological species. These results confirm bona fide incipient

speciation arising in a yeast population within a 1-month period of starvation.

3.7.4 Molecular Basis of Reproductive Barrier in a Starved Isolate

To discover the molecular mechanism of reproductive isolation, we further studied

several of the reproductively isolated starved isolates. Our experiments showed that

forward mutation frequency increased only two times in starved populations com-

pared with the nonstarved control, which could not account for the widespread

reproductive isolation. In contrast, pulsed-field gel analysis revealed a 6.6% total

frequency of new chromosomal variants in the starved BY4743 cultures, with no

rearrangements detected in nonstarved cultures (Coyle et al. in preparation). This

frequency is orders of magnitude higher than can be estimated for a typical labo-

ratory yeast strain (Schmidt et al. 2006). Finally, using microarray-based compara-

tive genomic hybridization (Dunn et al. 2005) we showed that all starved isolates

contained deletions and additions of genomic DNA (Coyle et al. in preparation).

58 E. Kroll et al.

In particular, one isolate contained duplication of the whole Chromosome I (Coyle

et al. in preparation).We decided to examine this disomic haploid isolate further todetermine whether chromosomal abnormalities which arose during starvation could

explain this strain’s reproductive isolation.

As has been reasoned before, in tetraploid hybrid meioses every chromosome is

furnished with its exact homologue (Dobzhansky 1933), therefore tetraploidization

should restore faithful segregation of homologues if mis-segregation in the diploid

hybrid were caused by chromosomal rearrangements and not by genic incompa-

tibilities. In our case, when we crossed the disomic starved isolate to its haploid

ancestor, we obtained a diploid hybrid with trisomy for Chromosome I (two copies

of the chromosome from the starved isolate and one from the ancestor). If the

Chromosome I trisomy were responsible for the lowered fertility of the backcross

hybrid, because there was no homologue furnished for the extra Chromosome I,

we would expect tetraploidization of this hybrid to restore its fertility. If the fertility

of the backcross hybrid were not restored then we would have to assume that an

epistatic interaction between incompatible alleles underlies reproductive barrier

between this isolate and its ancestor.

To test for this possibility, we obtained tetraploid versions of the trisomic

backcross hybrid by deleting one of the two MAT loci in the hybrid. We identified

hybrids expressing either MATa or MATalpha and crossed such strains using a

micromanipulator to produce several independent tetraploid versions of the back-

cross hybrid. We repeated this procedure with the nonstarved isolates to obtain

control tetraploids. After tetraploidy was confirmed by tetrad dissection, we sporu-

lated the resulting diploid hybrids and their tetraploid derivatives and measured the

sporulation efficiency as before. The results are shown in Fig. 3.2.

100

90

80

70

60

50

40

30

20

10

0

a b c d

Fig. 3.2 Relative sporulation efficiency of (a) Diploid starved backcross hybrid with extra

Chromosome I, (b) tetraploid starved backcross hybrid with extra Chromosome I, (c) diploid

nonstarved backcross hybrid, (d) tetraploid backcross hybrid. Ancestral sporulation efficiency is

assumed to be 100%. Spore viability in all strains was indistinguishable from the ancestral


Independently obtained tetraploid derivatives of the trisomic hybrid showed a

dramatic increase in sporulation efficiency compared with the diploid hybrid using

Mann–Whitney U test (Coyle et al. in preparation). In contrast, the increase in

sporulation efficiency of the nonstarved tetraploidized backcross hybrids was indis-

tinguishable from that of the nonstarved diploid backcross, indicating that tetra-

ploidization does not generally result in increased sporulation efficiency in the

nonstarved clones. Our results indicate that reproductive isolation in the starved

disomic isolate cannot be a consequence of the allelic incompatibilities between

the disomic isolate and the nonstarved ancestor. Rather, these results support the

hypothesis that chromosomal rather than genic differences underlie reduced ferti-

lity of the starved isolate.

3.8 Conclusions

The experiments described here provide insight into the phenomenon of starvation-

associated genomic rearrangements and its possible role in establishing reproduc-

tive isolation. Starvation is a condition that most natural organisms frequently

contend within the wild. Because a variety of changes in the external milieu can

result in starvation, we contend that starvation is a generic “interpreter” of catastro-

phic environmental change. Organisms that evolved mechanisms to harness star-

vation as signal to increase population diversity could be expected to leave more

descendants in the wake of such catastrophes. These mechanisms represent an

alternative population-level evolutionary response to the many individual-level

responses that enable organisms to persist under severe stress (e.g., spores, hiber-

nation, aestivation, extreme desiccation resistance, etc.).

Eukaryotes possess genetic mechanisms able to respond to stressful conditions;

however, no connection between starvation, starvation-induced genetic variation,

and speciation has been experimentally established in eukaryotes. Our experiments

provide evidence for this connection by showing that starved yeast populations

sustain genomic rearrangements at a dramatically higher frequency than nonstarved

populations, and that certain clones that survive starvation are reproductively

isolated from their ancestors. These newly evolved clones may represent incipient

species.

Genomic rearrangements have been shown to occur in yeast during chemical

treatment (Hughes et al. 2000) and growth in nutrient-limiting conditions (Adams

et al. 1992; Dunham et al. 2002). In fact, Dunham et al. note that several of their

parallel cultures grown in continuous culture under glucose limitation failed tosporulate, a phenomenon similar to the one observed here (Dunham et al. 2002).

This phenotype arose after 250–500 generations of continuous growth, unlike our

cultures which only underwent �10 generations during the course of starvation.

Recently, another study has shown that adaptation to diverse environments leads to

incipient speciation in yeast (Dettman et al. 2007), echoing the classic experiments

in Drosophila (Rice and Hostert 1993). The authors attempted to examine the

60 E. Kroll et al.

molecular nature of de novo speciation, using correlation between hybrid fitness

and fertility. Interestingly, in contrast to findings in extant yeast species (Greig

2009), their yeast hybrids, like ours, retained almost 100% of spore viability but

exhibited lower sporulation efficiency (Dettman et al. 2007).

We contend that genomic rearrangements arising during starvation may contrib-

ute to reproductive isolation, supporting the chromosomal theory of speciation

(White 1978). When the rate of genomic rearrangements is very low and the

effective population size is high, the chromosomal theory of speciation cannot

plausibly explain the process of speciation (Rieseberg 2001). However, the stress

of complete starvation circumvents these problems by dramatically increasing the

rate of chromosomal rearrangements in starving populations and simultaneously

decreasing the effective population size (because of the lower chances of having

enough resources to mate and also because of lower viability). Thus, environmental

conditions leading to starvation may favor the establishment of small, reproduc-

tively isolated, inbred subpopulations that harbor restructured genomes poised to

undergo rapid speciation without a requirement for any other type of prezygotic

isolation.

Acknowledgments We would like to acknowledge technical help from S. Coyle. This work was

supported by NSF grant 0134648 to E.K., NASA grant NNX07AJ28G grant to R.F.R. and NSF

ADVANCE grant DBI-0340856 to BD

References

Adams J, Puskas-Rozsa S, Simlar J, Wilke CM (1992) Adaptation and major chromosomal

changes in populations of Saccharomyces cerevisiae. Curr Genet 22:13–19

Anderson E (1949) Introgressive hybridization. Chapman & Hall, London

Bateson W (1909) Heredity and variation in modern lights. Darwin and modern science.

Cambridge University Press, Cambridge, UK

Bjedov I, Tenaillon O, Gerard B, Souza V, Denamur E, Radman M, Taddei F, Matic I (2003)

Stress-induced mutagenesis in bacteria. Science 300:1404–1409

Blanc G, Barakat A, Guyot R, Cooke R, Delseny M (2000) Extensive duplication and reshuffling

in the Arabidopsis genome. Plant Cell 12:1093–1101

Brachmann CB, Davies A, Cost GJ, Caputo E, Li J, Hieter P, Boeke JD (1998) Designer deletion

strains derived from Saccharomyces cerevisiae S288C: a useful set of strains and plasmids for

PCR-mediated gene disruption and other applications. Yeast 14:115–132

Britton-Davidian J, Catalan J, da Graca Ramalhinho M, Ganem G, Auffray JC, Capela R, Biscoito

M, Searle JB, da Luz Mathias M (2000) Rapid chromosomal evolution in island mice. Nature

403:158

Chambers SR, Hunter N, Louis EJ, Borts RH (1996) The mismatch repair system reduces meiotic

homeologous recombination and stimulates recombination-dependent chromosome loss. Mol

Cell Biol 16:6110–6120

Chen W, Jinks-Robertson S (1999) The role of the mismatch repair machinery in regulating

mitotic and meiotic recombination between diverged sequences in yeast. Genetics

151:1299–1313

Coyle S, Kroll E (2008) Starvation induces genomic rearrangements and starvation-resilient

phenotypes in yeast. Mol Biol Evol 25:310–318

Coyle S, Dunn B, Rosenzweig RF, Kroll E (in preparation) The molecular basis of starvation-

associated reproductive isolation in yeast


Coyne JA, Orr HA (1998) The evolutionary genetics of speciation. Philos Trans R Soc Lond B

Biol Sci 353:287–305

Darwin C (1859) On the origin of species by means of natural selection, or the preservation of

favoured races in the struggle for life. J. Murray, London

Davis L, Smith GR (2003) Nonrandom homolog segregation at meiosis I in Schizosaccharomyces

pombe mutants lacking recombination. Genetics 163:857–874

Death A, Ferenci T (1994) Between feast and famine: endogenous inducer synthesis in the

adaptation of Escherichia coli to growth with limiting carbohydrates. J Bacteriol

176:5101–5107

Delneri D, Colson I, Grammenoudi S, Roberts IN, Louis EJ, Oliver SG (2003) Engineering

evolution to study speciation in yeasts. Nature 422:68–72

Dernburg AF, McDonald K, Moulder G, Barstead R, Dresser M, Villeneuve AM (1998) Meiotic

recombination in C. elegans initiates by a conserved mechanism and is dispensable for

homologous chromosome synapsis. Cell 94:387–398

Dettman JR, Sirjusingh C, Kohn LM, Anderson JB (2007) Incipient speciation by divergent

adaptation and antagonistic epistasis in yeast. Nature 447:585–588

Dieckmann U, Doebeli M (1999) On the origin of species by sympatric speciation. Nature

400:354–357

Dobzhansky T (1933) On the sterility of the interracial hybrids in Drosophila pseudoobscura. ProcNatl Acad Sci USA 19:397–403

Dobzhansky T (1937) Genetics and the origin of species. Columbia Press, New York

Drake JW, Charlesworth B, Charlesworth D, Crow JF (1998) Rates of spontaneous mutation.

Genetics 148:1667–1686

Dresser ME, Ewing DJ, Harwell SN, Coody D, Conrad MN (1994) Nonhomologous synapsis and

reduced crossing over in a heterozygous paracentric inversion in Saccharomyces cerevisiae.

Genetics 138:633–647

Dri AM, Moreau PL (1993) Phosphate starvation and low temperature as well as ultraviolet

irradiation transcriptionally induce the Escherichia coli LexA- controlled gene sfiA. Mol


Dunham MJ, Badrane H, Ferea T, Adams J, Brown PO, Rosenzweig F, Botstein D (2002)

Characteristic genome rearrangements in experimental evolution of Saccharomyces cerevisiae.

Proc Natl Acad Sci USA 99:16144–16149

Dunn B, Levine RP, Sherlock G (2005) Microarray karyotyping of commercial wine yeast strains

reveals shared, as well as unique, genomic signatures. BMC Genomics 6(1):53–57

Eichler EE, Sankoff D (2003) Structural dynamics of eukaryotic chromosome evolution. Science

301:793–797

Finkel SE (2006) Long-term survival during stationary phase: evolution and the GASP phenotype.

Nat Rev Microbiol 4:113–120

Fischer G, James SA, Roberts IN, Oliver SG, Louis EJ (2000) Chromosomal evolution in

Saccharomyces. Nature 405:451–454

Fischer G, Neuveglise C, Durrens P, Gaillardin C, Dujon B (2001) Evolution of gene order in the

genomes of two related yeast species. Genome Res 11:2009–2019

Fischer G, Rocha EP, Brunet F, Vergassola M, Dujon B (2006) Highly variable rates of genome

rearrangements between hemiascomycetous yeast lineages. PLoS Genet 2:e32

Fisher RA (1930) The Genetical theory of natural selection. Oxford, UK

Friedberg E, Walker G, Siede W (1995) DNA repair and mutagenesis. Am Soc Microbiol,

Washington, DC

Greenberg AJ, Moran JR, Coyne JA, Wu CI (2003) Ecological adaptation during incipient

speciation revealed by precise gene replacement. Science 302:1754–1757

Greig D (2009) Reproductive isolation in Saccharomyces. Heredity 102:39–44

Greig D, Leu JY (2009) Natural history of budding yeast. Curr Biol 19:R886–R890

Greig D, Borts RH, Louis EJ, Travisano M (2002) Epistasis and hybrid sterility in Saccharomyces.

Proc R Soc Lond B Biol Sci 269:1167–1171

62 E. Kroll et al.

Guerin E, Cambray G, Sanchez-Alberola N, Campoy S, Erill I, Da Re S, Gonzalez-Zorn B, Barbe J,

PloyMC,Mazel D (2009) The SOS response controls integron recombination. Science 324:1034

Hartl DL, Lohe AR, Lozovskaya ER (1997) Regulation of the transposable element mariner.

Genetica 100:177–184

Hastings PJ, Bull HJ, Klump JR, Rosenberg SM (2000) Adaptive amplification. An inducible

chromosomal instability mechanism. Cell 103:723–731

Hauffe HC, Searle JB (1998) Chromosomal heterozygosity and fertility in house mice (Mus

musculus domesticus) from Northern Italy. Genetics 150:1143–1154

Hawley RS (2002) Meiosis: how male flies do meiosis. Curr Biol 12:R660–R662

He AS, Rohatgi PR, Hersh MN, Rosenberg SM (2006) Roles of E. coli double-strand-break-repair

proteins in stress-induced mutation. DNA Repair 5:258–273

Henikoff S, Ahmad K, Malik HS (2001) The centromere paradox: stable inheritance with rapidly

evolving DNA. Science 293:1098–1102

Herrmann RG, Maier RM, Schmitz-Linneweber C (2003) Eukaryotic genome evolution: rear-

rangement and coevolution of compartmentalized genetic information. Philos Trans R Soc

Lond B Biol Sci 358:87–97, discussion 97

Hughes TR, Roberts CJ, Dai H, Jones AR, Meyer MR, Slade D, Burchard J, Dow S, Ward TR,

Kidd MJ, Friend SH, Marton MJ (2000) Widespread aneuploidy revealed by DNA microarray

expression profiling. Nat Genet 25:333–337

Hunter N, Chambers SR, Louis EJ, Borts RH (1996) The mismatch repair system contributes to

meiotic sterility in an interspecific yeast hybrid. EMBO J 15:1726–1733

Hutter H, Vogel BE, Plenefisch JD, Norris CR, Proenca RB, Spieth J, Guo C, Mastwal S, Zhu X,

Scheel J, Hedgecock EM (2000) Conservation and novelty in the evolution of cell adhesion and

extracellular matrix genes. Science 287:989–994

Infante JJ, Dombek KM, Rebordinos L, Cantoral JM, Young ET (2003) Genome-wide amplifica-

tions caused by chromosomal rearrangements play a major role in the adaptive evolution of

natural yeast. Genetics 165:1745–1759

Jang JK, Sherizen DE, Bhagat R, Manheim EA, McKim KS (2003) Relationship of DNA double-

strand breaks to synapsis in Drosophila. J Cell Sci 116:3069–3077

Jinks-Robertson S, Sayeed S, Murphy T (1997) Meiotic crossing over between nonhomologous

chromosomes affects chromosome segregation in yeast. Genetics 146:69–78

Kai M, Wang TS (2003) Checkpoint activation regulates mutagenic translesion synthesis. Genes

Dev 17:64–76

Kellis M, Patterson N, Endrizzi M, Birren B, Lander ES (2003) Sequencing and comparison of

yeast species to identify genes and regulatory elements. Nature 423:241–254

King M (1993) Species evolution: the role of chromosome change. Cambridge University Press,

Cambridge

Koch AL (1971) The adaptive responses of Escherichia coli to a feast and famine existence. Adv

Microb Physiol 6:147–217

Lande R (1989) Fisherian and Wrightian theories of speciation. Genome 31:221–227

Lee HY, Chou JY, Cheong L, Chang NH, Yang SY, Leu JY (2008) Incompatibility of nuclear and

mitochondrial genomes causes hybrid sterility between two yeast species. Cell 135:1065–1073

Lichten M (2001) Meiotic recombination: breaking the genome to save it. Curr Biol 11:

R253–R256

Livingstone K, Rieseberg L (2004) Chromosomal evolution and speciation: a recombination-

based approach. New Phytol 161:107–112

Lombardo MJ, Aponyi I, Rosenberg SM (2004) General stress response regulator RpoS in

adaptive mutation and amplification in Escherichia coli. Genetics 166:669–680

Mayr E (1942) Systematics and the origins of species. Columbia University Press, New York

Mayr E (1966) Animal species and evolution. Harvard University Press, Cambridge

Mayr E (1996) What is a species and what is not? Philos Sci 63:262–277

McKenzie GJ, Harris RS, Lee PL, Rosenberg SM (2000) The SOS response regulates adaptive

mutation. Proc Natl Acad Sci USA 97:6646–6651


Muller HJ (1940) Bearing of the Drosophila work on systematics. In: Huxley J (ed) The new

systematics. Clarendon, Oxford, pp 185–268

Natsoulis G, Thomas W, Roghmann MC, Winston F, Boeke JD (1989) Ty1 transposition in

Saccharomyces cerevisiae is nonrandom. Genetics 123:269–279

Naumov GI, James SA, Naumova ES, Louis EJ, Roberts IN (2000) Three new species in the

Saccharomyces sensu stricto complex: Saccharomyces cariocanus,Saccharomyces kudriavze-

vii and Saccharomyces mikatae. Int J Syst Evol Microbiol 50(Pt 5):1931–1942

Navarro A, Barton NH (2003) Chromosomal speciation and molecular divergence–accelerated

evolution in rearranged chromosomes. Science 300:321–324

Noor MA, Grams KL, Bertucci LA, Reiland J (2001) Chromosomal inversions and the reproduc-

tive isolation of species. Proc Natl Acad Sci USA 98:12084–12088

Orr HA, Presgraves DC (2000) Speciation by postzygotic isolation: forces, genes and molecules.

Bioessays 22:1085–1094

Page SL, Hawley RS (2003) Chromosome choreography: the meiotic ballet. Science 301:785–789

Peoples TL, Dean E, Gonzalez O, Lambourne L, Burgess SM (2002) Close, stable homolog

juxtaposition during meiosis in budding yeast is dependent on meiotic recombination, occurs

independently of synapsis, and is distinct from DSB-independent pairing contacts. Genes Dev

16:1682–1695

Petit MA, Dimpfl J, Radman M, Echols H (1991) Control of large chromosomal duplications in

Escherichia coli by the mismatch repair system. Genetics 129:327–332

Phadnis N, Orr HA (2009) A single gene causes both male sterility and segregation distortion in

Drosophila hybrids. Science 323:376–379

Pialek J, Hauffe HC, Rodriguez-Clark KM, Searle JB (2001) Raciation and speciation in house

mice from the Alps: the role of chromosomes. Mol Ecol 10:613–625

Radman M (1975) SOS repair hypothesis: phenomenology of an inducible DNA repair which is

accompanied by mutagenesis. Basic Life Sci 5A:355–367

Radman M, Wagner R (1993) Mismatch recognition in chromosomal interactions and speciation.

Chromosoma 102:369–373

Rice W, Hostert E (1993) Laboratory experiments on speciation: what have we learned in 40

years? Evolution 47:1637–1653

Rieseberg LH (2001) Chromosomal rearrangements and speciation. Trends Ecol Evol 16:351–358

Roeder GS, Bailis JM (2000) The pachytene checkpoint. Trends Genet 16:395–403

San Filippo J, Sung P, Klein H (2008) Mechanism of eukaryotic homologous recombination. Annu

Rev Biochem 77:229–257

Schilthuizen M (2000) Dualism and conflicts in understanding speciation. Bioessays 22:1134–1141

Schliewen UK, Tautz D, Paabo S (1994) Sympatric speciation suggested by monophyly of crater

lake cichlids. Nature 368:629–632

Schmidt KH, Pennaneach V, Putnam CD, Kolodner RD (2006) Analysis of gross-chromosomal

rearrangements in Saccharomyces cerevisiae. Methods Enzymol 409:462–476

Searle JB (1998) Speciation, chromosomes, and genomes. Genome Res 8:1–3

Seoighe C, Federspiel N, Jones T, Hansen N, Bivolarovic V, Surzycki R, Tamse R, Komp C,

Huizar L, Davis RW, Scherer S, Tait E, Shaw DJ, Harris D, Murphy L, Oliver K, Taylor K,

Rajandream MA, Barrell BG, Wolfe KH (2000) Prevalence of small inversions in yeast gene

order evolution. Proc Natl Acad Sci USA 97:14433–14437

Sinervo B, Svensson E (2002) Correlational selection and the evolution of genomic architecture.

Heredity 89:329–338

Smets B, Ghillebert R, De Snijder P, Binda M, Swinnen E, De Virgilio C, Winderickx J (2010)

Life in the midst of scarcity: adaptations to nutrient availability in Saccharomyces cerevisiae.

Curr Genet 56:1–32

Taddei F, Matic I, Radman M (1995) cAMP-dependent SOS induction and mutagenesis in resting

bacterial populations. Proc Natl Acad Sci USA 92:11736–11740

Taddei F, Vulic M, Radman M, Matic I (1997) Genetic variability and adaptation to stress. EXS

83:271–290

64 E. Kroll et al.

Ting CT, Tsaur SC, Wu ML, Wu CI (1998) A rapidly evolving homeobox at the site of a hybrid

sterility gene. Science 282:1501–1504

Toczyski DP, Galgoczy DJ, Hartwell LH (1997) CDC5 and CKII control adaptation to the yeast

DNA damage checkpoint. Cell 90:1097–1106

Turelli M, Barton NH, Coyne JA (2001) Theory and speciation. Trends Ecol Evol 16:330–343

Via S (2001) Sympatric speciation in animals: the ugly duckling grows up. Trends Ecol Evol

16:381–390

Vulic M, Lenski RE, Radman M (1999) Mutation, recombination, and incipient speciation of

bacteria in the laboratory. Proc Natl Acad Sci USA 96:7348–7351

White MJD (1978) Modes of speciation. W.H. Freeman& Co, SanFrancisco

Witkin EM (1976) Ultraviolet mutagenesis and inducible DNA repair in Escherichia coli. Bacte-

riol Rev 40:869–907

Witkin EM, Wermundsen IE (1979) Targeted and untargeted mutagenesis by various inducers of

SOS functions in Escherichia coli. Cold Spring Harb Symp Quant Biol 43(Pt 2):881–886

Zinser ER, Kolter R (1999) Mutations enhancing amino acid catabolism confer a growth advan-

tage in stationary phase. J Bacteriol 181:5800–5807


Chapter 4

Populations of RNAMolecules as Computational

Model for Evolution

Michael Stich, Carlos Briones, Ester Lazaro, and Susanna C. Manrubia

Abstract We consider populations of RNA molecules as computational model for

molecular evolution. Based on a large body of previous work, we review some

recent results. In the first place, we study the sequence–structure map, its implica-

tions on the structural repertoire of a pool of random RNA sequences and its

relevance for the RNA world hypothesis of the origin of life. In a scenario where

template replication is possible, we discuss the internal organization of evolving

populations and its relationship with robustness and adaptability. Finally, we

explore how the effect of the mutation rate on fitness changes depends on the

degree of adaptation of an RNA population.

4.1 Introduction

Molecular evolution covers a huge area of research, ranging from prebiotic chem-

istry and questions on the origin of life, through many aspects related to the origin

of and the relationships among species, the study of viral and bacterial evolution

and their medical implications up to the artificial design and in vitro selection of

molecules, with all their applications in nano- and biotechnology. In this chapter,

we do not aim to give a complete overview of that wide research field, but focus on

the use of populations of RNA molecules as a model to understand evolution of

prebiotic replicators in the RNA world. As RNA viruses share many characteristics

with primitive RNA molecules with replicative ability, these studies can also be

used to tackle many aspects of viral evolution. Although a large body of our work is

inspired by experiments, in this chapter we focus on theoretical approaches for

understanding evolutionary processes.

M. Stich, C. Briones, E. Lazaro, and S.C. Manrubia

Dpto de Evolucion Molecular, Centro de Astrobiologıa (CSIC-INTA), Ctra de Ajalvir, km 4,

28850 Torrejon de Ardoz (Madrid), Spain



67

RNAmolecules are a very well suited model for studying evolution because they

incorporate, in a single molecular entity, both genotype and phenotype. While

errors in the replication process introduce mutations in the RNA sequence (geno-

type), selection acts upon the function (phenotype) of the molecule. Since in many

cases the spatial structure of the molecule is crucial for its biochemical function, the

structure of an RNA molecule can be considered as a minimal representation of the

phenotype.

In current biology, RNA viruses are the paradigmatic example for evolving

populations: replication is fast, it takes place with a relatively high error rate, and

population sizes are large. This has made RNA viruses an often used example for

quasispecies, a concept originally proposed by Eigen (1971) and developed over

the last decades in the context of virology (Domingo 2006). It states that a popula-

tion of replicators, e.g., an RNA virus evolving within an infected host, cannot be

represented by only one, fittest, genome, but by the spectrum of related mutants that

are present in the population. The quasispecies evolves under a certain error

(mutation) rate and the cloud of mutants enables the population to adapt quickly

to new environmental situations, such as population bottlenecks and changed

selective pressures. Under constant external conditions, a quasispecies approaches

a dynamic equilibrium between selection of favorable sequences (what we mean by

favorable, will be specified below) and the diversity constantly introduced by muta-

tion. Therefore, the mutation rate is of crucial importance in the study of such

heterogeneous populations in molecular evolution (Huynen et al. 1996; Biebricher

and Eigen 2005): if the mutation rate becomes too large, selection becomes ineffi-

cient, the correlations between the genomes within the population decay, and the

whole population may even become extinct. There are many reported examples of

the extinction of RNA virus populations when replication takes place at increased

error rates due to the presence of mutagenic agents (Sierra et al. 2000; Domingo

2005; Cases-Gonzalez et al. 2008). These results have inspired a new promising

antiviral strategy named lethal mutagenesis (Loeb et al. 1999).

Another field of research within molecular evolution is the quest for understand-

ing the origin and early evolution of life. One of the most appealing theories in this

context is the so-called RNA world hypothesis. It is based on the facts that RNA

cannot only represent a genetic code, like DNA in present-day cells, but also can act

as catalyst of biochemical reactions, like present-day enzymes. Therefore, a single

RNA molecule could have been endowed with the two main features of living

matter, providing the genome (i.e., the blueprint for replication) and the primordial

machinery for replication and metabolism. One of the open questions in this context

is how the first template-dependent RNA polymerase ribozyme could have

emerged. Experimentally, a minimum size of approximately 165 nucleotides has

been established for such a molecule (Johnston et al. 1999; Joyce 2004), a length

three to four times that of the longest RNA oligomers obtained by random poly-

merization (Huang and Ferris 2003, 2006). Hence, one of the main challenges

within the RNA world scenario is to convincingly bridge this gap.

In this chapter, we will review some recent results obtained in our lab (Manrubia

and Briones 2007; Stich et al. 2007, 2008, 2010; Briones et al. 2009) and put them

68 M. Stich et al.

into the context of the aforementioned issues. The first part of this chapter tries to

deepen our understanding of the sequence–structure map, relevant for the RNAworld

model. Then, we discuss the internal organization of evolving populations and its

relevance for robustness and adaptability. Subsequently, we explore the relationship

between microscopic mutation rate and the fractions of beneficial and deleterious

mutations, as observed in experiments or used in phenomenological models.

4.2 Structural Repertoire of RNA Pools

RNA structure is crucial for biochemical function of an RNA molecule. A lot of

research efforts are dedicated to the folding process that relates RNA sequences

with RNA structures. For our purpose, it is sufficient to consider two-dimensional

secondary structures as good approximation of real three-dimensional structures.

Two fundamental properties of the sequence–structure map are that (1) the number

of different sequences is much higher than the number of structures and (2) not all

possible structures are equally probable (Fontana et al. 1993; Schuster et al. 1994).

In this context, common structures are those which have many different sequences

folding into them and rare structures are those which have only few sequences

folding into them. In this section, we explore the structural repertoire of a pool of

random sequences.

We first describe the results of the folding of 108 RNA molecules of length 35 nt

consisting of random sequences composed of the four types of nucleotides A, C, G,and U (Stich et al. 2008). As secondary structure of each molecule, we take the

minimum free energy structure as given by the fold () routine from the Vienna

RNA Package (Hofacker et al. 1994).

RNA secondary structures consist of stems, where base pairing (A–U, G–C,G–U) between nucleotides occurs, and unpaired regions. In standard bracket nota-

tion, nucleotides paired with each other are denoted by “(” and “)”, while unpaired

nucleotides are represented by “.”. Among unpaired regions, we can distinguish

dangling ends and different kinds of loops: hairpin loops, bulges, interior loops, and

multiloops. The simplest structure is called a stem–loop, it consists of one hairpin

loop and one stem, and possibly one or two dangling ends. While there are 4n

sequences of length n (the so-called sequence space), the number Sn of different

structures (the structure space) is much smaller. Based on theoretical studies

(Waterman 1978), the expression Sn � 0.7131 � n�3/2 (2.2888)n has been given

(Gr€uner et al. 1996). Therefore, different sequences will actually fold into the same

secondary structure, grouping into neutral networks of genomes (Gr€uner et al. 1996;Huynen et al. 1996). Neutral networks are formed by genomes sharing the same

phenotype, here secondary structure, and which are connected by (single) muta-

tional events. The sequence–structure map turns out to be very complex. Two

sequences that are just one mutation apart may fold into structures very different

from each other. At the same time, in a relatively small neighborhood of any

sequence, almost all common structures can be found (Fontana et al. 1993).

4 Populations of RNA Molecules as Computational Model for Evolution 69

In our case, 108 sequences folded into 5,163,324 structures (Stich et al. 2008). A

way to visualize the uneven distribution of sequences into structures is the frequen-

cy–rank diagram. In Fig. 4.1a, we have ranked the structures according to the number

of sequences folding into them. One can see that there are around thousand common

structures, each of them obtained from about 104 different sequences. On the other

hand, we also find a few million rare structures yielded by only one or two sequences.

Although for a much smaller pool, this has already been reported before (Schuster

et al. 1994; Gr€uner et al. 1996; Schuster and Stadler 1994; Tacker et al. 1996).

In order to study the distribution of common vs. rare structures in more detail, we

have proposed a classification where we characterize a structure in terms of three

numbers (Stich et al. 2008): (a) the number of hairpin loops, H, (b) the sum of

bulges and interior loops, I, and (c) the number of multiloops, M. For example, a

simple stem–loop structure, denoted as SL, is characterized by (H,I,M) ¼ (1,0,0),

and all stem–loop structures found in the pool are grouped into that structure family.Other important families are the hairpin structure family, HP, with one interior loop

or bulge (1,1,0), the double stem–loop, DSL, represented by (2,0,0), and the simple

hammerhead structure, HH, by (2,0,1). Of course, there exist more complicated

structure families, as detailed in Stich et al. (2008). For the pool that we have

folded, we find that only 21 structure families are enough to cover all the 5.2 million

structures identified.

Our analysis, displayed in Fig. 4.1b, shows that the vast majority of sequences

fold into simple structure families. For example, 79.0% of all sequences belong to

only three structure families (HP, HP2, SL, in decreasing abundance), and 92.1% of

all sequences fold into simple structures with at most 3 stems (HP, HP2, SL, DSL,

DSL2, HH). Note that 2.1% of all sequences remain open and do not fold. Our data

is in agreement with other findings on the structural repertoire of RNA sequence

open

SL

HP

HP2

HP3HH

DSL

DSL2

rest

100 101 102 103 104 105 106 107

Rank

100 101 102 103 104 105 106 107

Rank

100

101

102

103

104

105

Freq

uenc

y

10–4

10–3

10–2

10–1

100

101

102

103

104

105

Bin

ned

abso

lute

fre

quen

cy

HPHP2SLDSLHP3DSL2

a b c

Fig. 4.1 (a) Frequency–rank diagram of the 5,163,324 different secondary structures, obtained by

folding 108 RNA sequences of length 35 nt. (b) Distribution of the sequences in structure families

according to their frequency. Higher-order hairpins, HPx, are defined as (H,I,M) ¼ (1,x,0), beingx � 2, higher-order double stem–loops, DSLx, as (H,I,M) ¼ (2, x�1,0), and higher-order ham-

merheads, HHx, as (2, x�1,1). (c) Frequency–rank diagram according to the structural family. The

upper thick solid curve denotes the same curve as in (a). Parts (a) and (c) after Stich et al. (2008)

70 M. Stich et al.

pools where the influence of the sequence length (Sabeti et al. 1997; Gevertz et al.

2005), the nucleotide composition (Knight et al. 2005; Kim et al. 2007), and pool

size (Gevertz et al. 2005) has been studied.

Now, we can reconsider the frequency–rank diagram. We sum up all structures

of a given structure family within a rank interval. Through this binning procedure,

we obtain for each structure family a curve which describes its relative frequency

compared with that of the other families. The curves for the most frequent families

are shown in Fig. 4.1c. We immediately see that the most frequent structures belong

to the stem–loop family, followed by the hairpin family, double stem loops, higher-

order hairpin families, and hammerheads. For low ranks, the SL curve is identical

with the curve describing all structures. For ranks between 4 � 103 and 104, it is the

HP curve which practically coincides with the total curve. Interestingly, the posi-

tion of the bump around rank 103 falls together with the locations where the SL

and HP families are equally present. Hence, we conclude that the bumps in the

frequency–rank diagram correspond to the succession of different structural

families and are not smoothed by better sampling of the sequence space.

What implications have these findings for the RNA world scenario? The stan-

dard view of the RNA world hypothesis states that the first chains of polymerized

polynucleotides consisted of random sequences. Therefore, it is important to study

the structural and subsequently the functional repertoire of such short sequences.

We have seen that a random pool is very rich in simple structures. However, as

already mentioned above, short molecules cannot perform template-dependent

replication. Therefore, we devised a four-step model of modular evolution as a

possible pathway for the emergence of functional and progressively longer mole-

cules starting with a random pool of RNA oligomers (Briones et al. 2009). The first

step is the random polymerization of RNA molecules up to 40-mers. The second

step is the folding of these sequences, leading to high fractions of simple structures

like hairpins, as just shown. The third step is based on the observation that simple

hairpin structures, similar to those formed by short random sequences in huge

amounts, are actually known to show catalytic activity, leading to RNA–RNA

ligation (Puerta-Fernandez et al. 2003). If a certain fraction of the hairpin molecules

originated is capable of displaying ligase activity, longer molecules may be formed.

Even though the majority of the long molecules may not perform ligase activity,

some of them will keep the modular structure of their building blocks and remain

active to catalyze further RNA–RNA ligations (Manrubia and Briones 2007). This

suggests that hairpin ribozymes, both in individual modules and in combined

structures, could have catalyzed the synthesis of progressively longer RNA mole-

cules from short and structurally simpler modules (Briones et al. 2009). Finally, the

fourth step of the model consists of a maturation of these ligating RNAmolecules of

intermediate length into self-replicating RNA ligase networks, which could coexist

and even compete with each other, leading eventually to a molecule long and

complex enough to perform template-dependent RNA replication [further details

in Briones et al. (2009)]. It is important to emphasize that the whole model relies

strongly on the observation that simple structures like hairpins – with potential

ligase activity – are ubiquitous in pools of random RNA sequences.


4.3 Internal Organization of Evolving Populations

Above, we have discussed the static picture of the sequence–structure map. Once

replication within a population is possible, evolution through Darwinian selection is

triggered. Here, RNA serves as a model to study the interplay between mutation,

selection, and the diversity sustained in populations of fast mutating replicators

(Stich et al. 2007).

First, we briefly describe the evolutionary algorithm. Our system consists of a

population of N replicating RNA sequences, each of length n nucleotides. At the

beginning of the simulation, every molecule is initialized with a random sequence.

Every time that a sequence replicates, each of its nucleotides has a probability m(mutation rate) to be replaced by another nucleotide, randomly chosen among the

four possibilities A, C, G, U.At each generation, the sequences are folded into secondary structures as

described above. We define a target structure that represents in a simple way

optimal performance in a given environment. It can be a hairpin, hammerhead, or

any other structure: the qualitative behavior of the system does not depend on this

choice. We compare every folded structure with the target structure by means of the

base pair distance di, defined as the number of base pairs that have to be opened and

closed to transform a given structure into the target structure (Hofacker et al. 1994).

The closer a secondary structure is to the target structure, the higher the probability

p(di) that the corresponding sequence i replicates:

p dið Þ ¼ exp �bdið ÞPNi¼1 exp �bdið Þ : (4.1)

The parameter b denotes the selective pressure and is here chosen as b ¼ 2/n.Generations in our simulations are nonoverlapping and the offspring generation is

calculated according to Wright–Fisher sampling.

Two relevant quantities to characterize the state of the population are the

average distance d ¼PNi¼1 di=N to the target structure and the fraction r of struc-

tures in the population folding exactly into the target structure. Because of the

persisting action of mutation, both quantities fluctuate in time even after reaching

the asymptotic regime. Therefore, we perform averages over long time intervals

(and different realizations, starting from distinct initial RNA populations), obtain-

ing mean values denoted by �d and �r, respectively.In order to quantify collective properties of the molecular ensemble, we first

determine the consensus sequence of the population, given by, for each position

along the sequence, the most frequent type of nucleotide found within the popu-

lation. In real RNA molecular and viral quasispecies, the consensus sequence is

obtained by means of population sequencing (Thurner et al. 2004; Simmonds et al.

2004; Domingo 2006), and it does not necessarily correspond to any of the indi-

vidual sequences present in the population. It is straightforward to fold the con-

sensus sequence and obtain the structure of the consensus sequence, for which its

72 M. Stich et al.

coincidence with the target structure can be determined. At each time step we count

either one, corresponding to coincidence, or zero, otherwise. Averages over time

(and realizations) of this binary variable yield �rC, which corresponds to the

probability that, at a randomly chosen time step, the structure of the consensus

sequence coincides with the target structure.

We further define a consensus structure. It is calculated by determining, for each

position along the molecule, the most frequent structural state found within the

population, i.e., unpaired “.”, paired upstream “(”, or paired down-stream “)”. Due

to this definition, the consensus structure does not necessarily represent a valid

secondary structure of an RNA molecule. This procedure is hence fundamentally

different from assigning a consensus structure to an alignment of sequences

(Hofacker et al. 2002). Averages over time (and realizations) of the coincidence

between the consensus structure and the target structure yield the probability �r S.

Within this model, evolution takes place in the following way: sequences which

fold into structures similar to the target structure will replicate more likely and their

fraction in the population increases. Mutation introduces diversity and enables the

system to find structures that are closer to the target, and finally find and fix the

target structure. Starting from a random set of sequences, we can distinguish several

phases of evolution: the search phase, where d decreases while r ¼ 0. This phase

finishes at generation gA when a molecule folds into the target structure for the first

time. Then, the phase of fixation begins, where – on average – d still decreases and

r increases. However, due to the stochastic nature of mutation – and hence in

particular for large mutation rates as will be explored further below – the population

may lose again the target structure (and r drops down to zero). If r does not drop to

zero for 500 consecutive generations, we say that the target structure has been fixed

at generation gF. Then, the asymptotic regime is reached, where d and r fluctuate

around constant values and which corresponds to a mutation–selection equilibrium.

If the mutation rate m is too large, the population is unable to maintain the target

structure within the population. In absence of an analytic theory for the system we

are studying, we determine the fixation threshold as the value mF at which the curvegF(m) diverges.

Since we now have defined the main quantities to describe the population, we

show the results in Fig. 4.2. They were obtained from simulations for a system of

N ¼ 1,000 RNAmolecules of length n ¼ 30 nt evolving toward a hairpin structure.

In (a) we show the curves for �r; �rC, and �rS. The quantity �r describes the funda-

mental property of a quasispecies at mutation–selection equilibrium. For small m, �rtakes maximal values. This means that a population contains the largest fraction of

correctly folded molecules if it evolves at small mutation rates. As m increases, �rdecreases monotonously until it approaches zero. To determine the fixation thresh-

old, we look at Fig. 4.2b where we show the curves of the search time and search

plus fixation time. The solid curve represents the search time. We observe that for

small m finding the target structure is difficult because only little diversity is

introduced and the search process is slow. Therefore, fixation takes a long time.

As m increases, the introduced diversity in the population becomes larger and both

search and search plus fixation times decrease. However, fixation turns out to be a


difficult task if m is too large, and the curves for search and search plus fixation start

to deviate. The search plus fixation time gF (dotted curve) diverges around

m � 0.045, where we approximately locate the fixation threshold for this n and

target structure. This means that while the population shows largest �r for small mand highest degree of diversity close to the fixation threshold, the search and

fixation times are optimized for intermediate mutation rates around m � 0.025

well below the fixation threshold.

Coming back to Fig. 4.2a, we now have a look at the curves for �rC and �rS. Thecurve of �rC lies for all considered mutation rates above the curve of �r. This means

that based upon the information of the consensus sequence only, one may overesti-

mate the evolutionary success. This effect is observed both below and above the

fixation threshold. For example, for m ¼ 0.05, where only 0.5% sequences fold into

the target structure, and only into an intermittent way, the probability that the

consensus sequence folds into the target structure is still 18%. Consequently, the

population remains close to sequences that actually fold into the target structure

although it is unable to fix it. Obviously, this is related to the fact that at least part of

the population are descendents from the same sequence and hence are closely

related to each other. Note that the probability that a sequence of the population

folds into the target structure is different from the probability that the consensus

sequence does. Since consensus sequences are readily obtained from molecular or

viral quasispecies, one should take into account this difference.

Considering now the curve for �rS, we observe a qualitatively different behavior:

for m < 0.025, the probability that the consensus structure coincides with the target

structure is practically one, while for m > 0.025, it approaches zero. For small m,this effect can be easily explained: the weight of all the correctly folded molecules

is strong enough to keep �rS high. But in Stich et al. (2007), we showed that even

0 0.01 0.02 0.03 0.04 0.05 0.06

μ

0

0.2

0.4

0.6

0.8

1a b

ρρC

ρS

0.07 0.08 0 0.01 0.02 0.03 0.04 0.05 0.06

μ0.07 0.08

0

50

100

150

200

250

300

gAgF

Fig. 4.2 (a) Asymptotic properties of a population of size N ¼ 1,000 and molecules of length

n ¼ 30 nt as function of the mutation rate m. Displayed are the average fraction of correctly foldedstructures �r, and the quantities �rC and �rS. Averaging has been performed over 4,000 generations

and 20 realizations, disregarding the first 2,000 generations. (b) Search time gA and search plus

fixation time gF. We locate the fixation threshold where gF diverges. Averaging has been

performed over 200 realizations. The population evolves toward a hairpin target structure given

by ..((((((. . .(((. . .))). . ..)))))) in bracket notation

74 M. Stich et al.

neglecting the correctly folded molecules and for large mutation rates, among the

remaining sequences there is a sufficiently large fraction of those molecules which

have a similar structure to the target structure. An analogous effect is known for

random sequences: in a small neighborhood of a given sequence, the most probable

structures are identical or very similar to the structure of the reference sequence

(Fontana et al. 1993). Even where rS ¼ 0, the distribution of the structure states

along the chain may still resemble the target structure and the positions where the

concordance is broken correspond to positions that are actually less stable.

While �rC senses the similarity among the sequences and �rS the similarity among

the structures, both quantities take superior values than �r for most of the mutation

rates in spite of the fact that selection is actually acting upon structure (not

sequence) and that the corresponding fitness landscape is rough. This means that

the population retains relevant structural information in a distributed fashion even

above the fixation threshold. This represents a strong structural robustness and

suggests that certain functional RNA secondary structures may effectively with-

stand high mutation rates (Stich et al. 2007).

4.4 Phenotypic Effect of Mutations

In the last section, we have already discussed the optimal mutation rate to promote

adaptation in an evolving system. Here, we calculate the distribution of the effects

of mutations on fitness and the relative fractions of beneficial and deleterious

mutations (Stich et al. 2010). It is important to recall that the effect of mutations

on the phenotype depends on the genomic and populational context. We explore

two different situations: the mutation–selection equilibrium (equilibrated popula-

tion) and the first stages of the adaptation process (adapting population).

Here, we consider a population of N ¼ 1,000 molecules of length n ¼ 50 nt

evolving toward a hairpin target structure. The change in fitness of an RNA

sequence under replication is quantified by the change of distance to the target

structure, i.e., by Dij ¼ di – dj, where i denotes the mother and j the daughter

sequence. Hence, for Dij > 0 (Dij < 0), the mutations lead to an increase (decrease)

of fitness and hence are beneficial (deleterious). If Dij ¼ 0, either no mutation

occurred or the mutations had no effect on fitness (were neutral). As we sum up

over N values of Dij at each generation (and over generations and realizations as

specified below), we obtain a probability distributionP(D) of the changes in fitness.In Fig. 4.3a, we show for three different mutation rates the distributions P(D),

obtained for populations at mutation–selection equilibrium. The part of the distri-

bution with the largest weight represents replication events with no or neutral

mutations (D ¼ 0). For a very low mutation rate, negative fitness events strongly

dominate over the positive ones and hence beneficial mutations are rare. As the

mutation rate increases, the curves move up for positive and negative D since there

are more mutation events. Although in particular beneficial mutations occur more

often, negative fitness effects still dominate in absolute numbers.


From the distribution P we can calculate the fraction of deleterious changes

p and beneficial changes q in the following way:

q ¼Z 1

0þP Dð ÞdD; (4.2)

p ¼Z 0�

�1P Dð ÞdD: (4.3)

These quantities represent the beneficial and deleterious phenotypic mutation

rates which shall not be confounded with the microscopic mutation rate m. Bydefinition, p þ q þ P(0) ¼ 1.

Π(Δ

)

–10 0 10–30 –20 20 30

Δ

10–6

10–4

10–2

100a

c

b

d

10–6

10–4

10–2

100

Π(Δ

)

μ = 5x10–4

μ = 1x10–2

μ = 4x10–2

μ = 5x10–4

μ = 1x10–2

μ = 4x10–2

10–4 10–3 10–2 10–1 100

μ

10–5

–10 0 10–30 –20 20 30

Δ10–4 10–3 10–2 10–1 100

μ

10–5

10–4

10–3

10–2

10–1

100

10–4

10–3

10–2

10–1

100

pqΠ(0)

pqΠ(0)

μF

μF

Fig. 4.3 Phenotypic changes of mutations for optimized (a, b) and adapting (c, d) populations. (a)

Probability distribution P(D) obtained from 300 generations in the asymptotic regime and for

three different values of m. (b) Beneficial (q) and deleterious (p) phenotypic mutation rates as

function of the microscopic mutation rate m for optimized populations. Replication events without

fitness change are given byP(0). (c) As (a), but for adapting populations (probability distributions

obtained from the first 50 generations and 6 different realizations). (d) As (b), but for adapting

populations. The thin curves denote the curves from (b). The target structure is ((((. . .. . ..(((((.(((((. . .. . .))))).))))). . .. . ..)))) in bracket notation. After Stich et al. (2010)

76 M. Stich et al.

How q and p depend on m is depicted in Fig. 4.3b. For low mutation rates, we see

that p is more than two orders of magnitude larger than q. As m increases, both p andq increase, although p > q for all m, in particular for mutation rates below the

fixation threshold, for this n and target structure approximately located at

mF ¼ 0.02. As m increases, the fraction of replication events with no change in

fitness, given by P(0), decreases. The ratio p/q decreases from more than two

orders of magnitude to less than one close to mF. This reflects the fact that the higherthe mutation rate at which a population has reached mutation–selection equilibrium

the lower the fraction of correctly folded molecules, and hence beneficial mutations

are more probable. However, these beneficial mutations do not increase the degree

of adaptation of the population due to the difficulties to get fixed at high error rate.

In Fig. 4.3c,d, we show the distribution P(D) and the functional behavior of

(p, q) ¼ f(m) for adapting populations. In this case, fitness changes are measured

before the target structure has been found. The distributionsP(D) behave in a qualita-tively similar way, although quantitative differences to Fig 4.3a can be seen, e.g., for

m ¼ 0.0005: The range of negativeD is smaller than for an equilibrated population, so

very deleterious mutations are not present, and also the overall level of deleterious

mutations is lower. At the same time, beneficial mutations are more common. This

observation can be explained by the fact that since the population is still relatively far

from target, mutations that drive a sequence even further are less likely. For the same

reason, mutations that have a positive effect on fitness are more probable.

Figure 4.3d summarizes the results: In an adapting population, p is smaller than

at equilibrium, and q is larger, although these differences get much lower as the

error rate increases. However, in all cases there are still more deleterious mutations

than beneficial ones. Again, both phenotypic mutation rates increase as m increases,

while replication events without phenotypic change decrease.

4.5 Summary

Here, we have presented recent results with RNA populations as computational

model to explore and understand evolutionary processes, using the complex under-

lying sequence–structure–function relationship of RNA molecules.

In the first section, we showed some observations on the structural repertoire of

random RNA sequences (Stich et al. 2008). One important result is that simple

structures like stem–loops and hairpins are dominant in pools of short sequences.

This finding, together with other results and arguments, allowed us to devise a

stepwise model of modular evolution for the origin of the RNA world (Briones et al.

2009).

In the second section, we introduced an algorithm of RNA evolution in silico

(Stich et al. 2007). After characterizing the asymptotic state of the population (at

mutation–selection equilibrium), we showed that search and fixation times are

optimized for intermediate mutation rates, far from the fixation threshold where

the creation of diversity is maximal and far from the regime of low mutation rates


where evolutionary success is optimized (in terms of correctly folded molecules).

These results have important implications for the adaptability of virus and repli-

cator populations that, due to the changes in the selective pressures that they

continuously experience, need to have the capability to adapt rapidly, which can

be obtained by the selection of high mutation rates. However, the difficulties for the

fixation of beneficial mutations, together with the low fitness values attained when

replication takes place at mutation rates close to the error threshold, suggest that

viral quasispecies operate at mutation rates considerably smaller.

Furthermore, close to and even beyond the fixation threshold, RNA populations

show clear signatures of the target structure they try to approach (Stich et al. 2007).

For example, even a population that contains practically no molecule that folds into

the correct structure, as a whole may actually harbor the target structure as the

structure of its consensus sequence. This demonstrates that the evolutionary success

of the population is more robust than suggested by the spectrum of its mutants alone.

Finally, we have established a connection between the microscopic mutation rate

m and the phenotypic mutation rates p and q (Stich et al. 2010). These mutation rates

are used in phenomenological models of population dynamics and also in fitting

models of data obtained from experiments (Eyre-Walker and Keightley 2007). We

find that adapting populations have a much larger fraction of beneficial mutations

than equilibrated ones, especially for small mutation rates. Furthermore, we have

shown that increases in m do not cause linearly proportional increases in p and q, asoften assumed in simple models of population evolution.

In summary, our results encourage the combined approach of experimental

research and computational modeling for studying molecular evolution.

Acknowledgments The authors acknowledge support from Spanish MICIIN through projects

FIS2008-05273 and BIO2007-67523, from INTA, and from Comunidad Autonoma de Madrid,

project MODELICO (S2009/ESP-1691).

References

Biebricher CK, Eigen M (2005) The error threshold. Virus Res 107:117–127

Briones C, Stich M, Manrubia SC (2009) The dawn of the RNA world: Toward functional

complexity through ligation of random RNA oligomers. RNA 15:743–749

Cases-Gonzalez C, Arribas M, Domingo E, Lazaro E (2008) Beneficial effects of population

bottlenecks in an RNA virus evolving at increases error rate. J Mol Biol 384:1120–1129

Domingo E (ed) (2005) Virus entry into error catastrophe as a new antiviral strategy. Virus Res

107:115–228

Domingo E (ed) (2006) Quasispecies: concept and implications for virology. Springer, Berlin

Eigen M (1971) Self-organization of matter and the evolution of biological macromolecules.

Naturwissenschaften 58:465–523

Eyre-Walker A, Keightley PD (2007) The distribution of fitness effects of new mutations. Nat Rev

Genet 8:610–618

Fontana W, Konings DAM, Stadler PF, Schuster P (1993) Statistics of RNA secondary structures.

Biopolymers 33:1389–1404

78 M. Stich et al.

Gevertz J, Gan HH, Schlick T (2005) In vitro RNA random pools are not structurally diverse: a

computational analysis. RNA 11:853–863

Gr€uner W, Giegerich R, Strothmann D, Reidys C, Weber J, Hofacker IL, Stadler PF, Schuster P

(1996) Analysis of RNA sequence structure maps by exhaustive enumeration. I. Neutral

networks. Monatsh Chem 127:355–374

Hofacker IL, FontanaW, Stadler PF, Bonhoeffer LS, Tacker M, Schuster P (1994) Fast folding and

comparison of RNA secondary structures. Monatsh Chem 125:167–188

Hofacker IL, Fekete M, Stadler PF (2002) Secondary structure prediction for aligned RNA

sequences. J Mol Biol 319:1059–1066

Huang W, Ferris JP (2003) Synthesis of 35–40 mers of RNA oligomers from unblocked mono-

mers. A simple approach to the RNA world. Chem Commun 12:1458–1459

Huang W, Ferris JP (2006) One-step, regioselective synthesis of up to 50-mers of RNA oligomers

by montmorillonite catalysis. J Am Chem Soc 128:8914–8919

Huynen MA, Stadler PF, Fontana W (1996) Smoothness within ruggedness: the role of neutrality

in adaptation. Proc Natl Acad Sci USA 93:397–401

Johnston WK, Unrau PJ, Lawrence MS, Glasner ME, Bartel DP (1999) RNA-catalyzed RNA

polymerization: accurate and general RNA-templated primer extension. Science 292:1319–1325

Joyce GF (2004) Directed evolution of nucleic acid enzymes. Annu Rev Biochem 73:791–836

Kim N, Gan HH, Schlick T (2007) A computational proposal for designing structured RNA pools

for in vitro selection of RNAs. RNA 13:478–492

Knight R, De Sterck H, Markel R, Smit S, Oshmyansky A, Yarus M (2005) Abundance of

correctly folded RNA motifs in sequence space, calculated on computational grids. Nucleic

Acids Res 33:5924–5935

Loeb LA, Essigmann JM, Kazazi F, Zhang J, Rose KD, Mullins JI (1999) Lethal mutagenesis of

HIV with mutagenic nucleoside analogs. Proc Natl Acad Sci USA 96:1492–1497

Manrubia SC, Briones C (2007) Modular evolution and increase of functional complexity in

replicating RNA molecules. RNA 13:97–107

Puerta-Fernandez E, Romero-Lopez C, Barroso-delJesus A, Berzal-Herranz A (2003) Ribozymes:

recent advances in the development of RNA tools. FEMS Microbiol Rev 27:75–97

Sabeti PC, Unrau PJ, Bartel DP (1997) Accessing rare activities from random RNA sequences: the

importance of the length of molecules in the starting pool. Chem Biol 4:767–774

Schuster P, Stadler PF (1994) Landscapes: complex optimization problems and biopolymer

structures. Comput Chem 18:295–324

Schuster P, Fontana W, Stadler PF, Hofacker IL (1994) From sequences to shapes and back: a case

study in RNA secondary structures. Proc R Soc Lond B Biol Sci 255:279–284

Sierra S, Davila M, Lowenstein PR, Domingo E (2000) Response of foot-and-mouth disease virus

to increased mutagenesis. J Virol 74:8316–8323

Simmonds P, Tuplin A, Evans DJ (2004) Detection of genome-scale ordered RNA structure

(GORS) in genomes of positive-stranded RNA viruses: implication for virus evolution and

host persistence. RNA 10:1337–1351

Stich M, Briones C, Manrubia SC (2007) Collective properties of evolving molecular quasispe-

cies. BMC Evol Biol 7:110

Stich M, Briones C, Manrubia SC (2008) On the structural repertoire of pools of short, random

RNA sequences. J Theor Biol 252:750–763

Stich M, Lazaro E, Manrubia SC (2010) Phenotypic effect of mutations in evolving populations of

RNA molecules. BMC Evol Biol 10:46

Tacker M, Stadler PF, Bornberg-Bauer EG, Hofacker IL, Schuster P (1996) Algorithm indepen-

dent properties of RNA secondary structure predictions. Eur Biophys J 25:115–130

Thurner C, Witwer C, Hofacker IL, Stadler PF (2004) Conserved RNA secondary structures in

flaviviridae genomes. J Gen Virol 85:1113–1124

Waterman MS (1978) Secondary Structure of Single-stranded Nucleic Acids. In: Rota G-C (ed)

Studies in Foundation and Combinatorics, vol 1 of: Advances in Mathematics Supplementary

Studies. Academic Press, New York, pp 167–212


Chapter 5

Pseudaptations and the Emergence

of Beneficial Traits

Steven E. Massey

Abstract There is increasing evidence for the emergence of some beneficial

traits in biological systems in the absence of direct selection. Many of these

encompass mutational robustness, which increasingly appears to arise as a by-

product of natural selection, as a consequence of the biased incremental change of

complex biological systems. Understanding the emergence of robustness in dis-

parate biological systems is facilitated by the use of graph theory and the concept

of connectivity. A particular case that is explored here is that of the standard

genetic code (SGC). The SGC is arranged so that mutations tend to result in

conservative as opposed to radical amino acid changes, a property termed “error

minimization”. A commonly cited explanation for this property is the “Adaptive

Code” hypothesis, which proposes that error minimization has been directly

selected for. However, it is shown that direct selection of the error minimization

property is mechanistically difficult. In addition, it is apparent that error minimi-

zation may arise simply as a result of code expansion, this is termed the

“emergence” hypothesis. The emergence of error minimization in the genetic

code is likened to other biological examples, where mutational robustness arises

from the innate dynamics of complex systems; these include neutral networks and

a variety of subcellular networks. The concept of “biased incrementalism” is

introduced to account for the emergence of robustness in these diverse systems,

while the term “pseudaptation” is used for such traits that are beneficial to fitness,

but are not directly selected for.

S.E. Massey

Biology Department, University of Puerto Rico – Rio Piedras, P.O. Box 23360, San Juan, Puerto

Rico 00931, USA



81

5.1 Adaptive Evolution and Natural Selection

The modern definition of an adaptation is tautological in relation to natural selection;

fromMayr’s book “What Evolution Is” (Mayr 2001), adaptations are beneficial traits

that arise by natural selection, of if they occur by chance are maintained by natural

selection. From a panselectionist perspective, all beneficial phenotypes are to be

regarded as adaptations, arising from natural selection. However, it may be argued

that the definition of “adaptation” is not inviolate; indeed, it is worth remembering

that until the modern synthesis natural selection was not widely accepted as the

predominant force behind adaptive evolution; the so-called “eclipse of Darwinism”

(Huxley 1942). The theme of this work is to clarify the definition of adaptation, in the

context of natural selection, and to examine examples of beneficial traits that have

arisen in the absence of direct selection, and how they should be defined.

5.2 Emergence as a By-Product of Natural Selection

Emergence is a term used in studies of complexity, to describe properties that arise

from the summation of numerous individual interactions. Diverse examples include

the emergence of nonrandom network topologies (ranging from biological net-

works such as metabolic networks, social interaction networks such as sexual

contact and scientific collaboration networks, infrastructure networks such as the

Internet and power networks and chemical networks; Gleiss et al. 2001; Albert and

Barabasi 2002), weather features such as hurricanes, and Adam Smith’s “invisible

hand” that self-regulates the market. In biological systems emergent properties may

be directly selected for; examples of this include termite mounds, the shoaling

behavior of fish or the ability of ant colonies to solve geometric problems, such as

the shortest route to a food source. In contrast, this chapter is devoted to addressing

cases where beneficial traits emerge in the absence of direct selection.

5.3 “Pseudaptation” as a Descriptor of Beneficial Traits

That Arise in the Absence of Direct Selection

The term “spandrel” was coined to describe phenotypes that arise without the

direct agency of natural selection (Gould and Lewontin 1979; Gould 1997). How-

ever, it is unclear in the definition whether these traits are beneficial to fitness.

Therefore, it is proposed that the term “spandrel” should be used to refer to

phenotypes that arise nonadaptively, as a side-product of natural selection, but

are not clearly beneficial to fitness. This work is devoted to discussing beneficial

traits that are not directly selected for, hence a term is required for such phenomena;

it is suggested that the term “pseudaptation” is used for such traits (Massey 2010).

82 S.E. Massey

The prefix “pseud” is used to indicate the potential tendency to misinterpret such

traits as true adaptations resulting from natural selection. In contrast, therefore,

“adaptations” are beneficial traits that result from the agency of natural selection.

The vast majority of beneficial traits are expected to be true adaptations.

5.4 The Genetic Code as a Case Study

The standard genetic code (SGC) will be used as a case study for illustrating how a

pseudaptation may emerge in a complex system. The arrangement of amino acids to

codons in the SGC is such that proteins are remarkably robust to the deleterious

effects of mutations and transcriptional/translational errors, in comparison to ran-

domly generated genetic codes (Alff-Steinberger 1969; Di Giulio 1989; Haig and

Hurst 1992; Ardell 1998; Freeland et al. 2000; Gilis et al. 2001; Goodarzi et al.

2004, etc.). This property of the SGC is termed “error minimization” (EM) and

results in a tendency for conservative as opposed to radical amino acid substitutions

(Fig. 5.1). EM can be expressed mathematically by the “EM value”. This is a

Fig. 5.1 The influence of the structure of the standard genetic code on the proportions of

conservative or radical amino acid substitutions. There are 75 different amino acid substitutions

that can result from a single point mutation, due to the structure of the SGC. The similarity of the

amino acids separated by a single point mutation was defined according to the Grantham matrix.

The proportion of substitutions that corresponded to different Grantham values was binned

accordingly. The chart shows a strong skew toward conservative substitutions

5 Pseudaptations and the Emergence of Beneficial Traits 83

parameter that calculates the average difference between two amino acids arising

from a nonsynonymous mutation and is defined as follows:

EM ¼X61i¼1

XNtN¼1

dNi=Nt

!,61 Massey 2008ð Þ;

where there are i sense codons, Nt is the total number of sense codons separated by a

single point mutation from the ith codon under consideration, dNi is the physico-

chemical distance between the amino acids coded for by the ith sense codon and theNth sense point mutation, according to the 20 � 20 Grantham physicochemical

similarity matrix (Grantham 1974).

The smaller the value between two amino acids in the Grantham matrix, the

more similar they are, thus the smaller the EM value the larger the extent of EM in a

genetic code. The EM value of the SGC is 60.7, while the EM value of a computa-

tionally randomly generated code is 74.5. Only 0.03% of computationally randomly

generated genetic codes possess EM values equal or better than that of the SGC,

which is indication of the remarkable optimization of the SGC (Massey 2008a).

Thus, the code is near optimal for the property of EM.

The EM value of the SGC can be understood as representing the average con-

nectivity of all the codons. Figure 5.2 shows how a typical codon may be repre-

sented as the node of a graph, with edges representing point mutations to different

codons. Each codon may be represented this way, thus the SGC may be envisaged

as a graph, composed of 64 nodes. The EM value represents the average connectivity

of the SGC, thus robustness arises from a maximization of the average connectivity

of the code in terms of neutrality. Another way of putting this is that the amino acids

are assigned to codon blocks so that the likelihood of an amino acid substitution

being selectively neutral is high. The property of EM is beneficial in that it limits

the deleterious effects of mutations and transcriptional/translational errors. Thus,

the “Adaptive Code” hypothesis proposes that the EM property is a beneficial trait

that has been selected via natural selection (Freeland et al. 2000). The Adaptive

Code hypothesis implies that “code space”, the space of alternative genetic codes,

was “searched” by natural selection until a near-optimal code was reached (the

SGC). However, there are problems with this scenario, discussed next.

5.4.1 Challenges for the Adaptive Code Hypothesis to Explainthe Origin of the EM Property of the SGC

There are several challenges for explaining how the EM property was directly

selected for by natural selection. First, in order to “find” an optimal or near-optimal

genetic code for the property of EM, the code space of alternative codes needs to be

searched. This necessitates the occurrence of “codon reassignments”, which are

where the amino acid identity of a codon(s) is reassigned from one of the 20 amino

acids to another. While these have occurred in nature, mainly in mitochondrial

84 S.E. Massey

Table 5.1 The emergence of EM in simulations of genetic code evolution

V A D G1 2 3 4

V A D G1 2 3 4

5 6 7 89 10 11 12

5 6 7 8

9 10 10

865

1314

A2V1 G4

11 1211 12

13 14 15 10

7

17

15

1918

9

16

D3

20

10

88

5 6 7 8

V1 A2 D3 G4

1 2 3 4V A D G

i) ii)

iv) v)

iii)

Selective criteria

(amino acid

difference according

to the Grantham

matrix)

Average EM value

of alternate codes

Average percentage

optimization of alternate

codes compared with the

standard genetic code

Percentage of alternate

codes that have equal or

superior error

minimization than the

standard genetic code

<150 70.2 31.2 0.3

<140 69.6 35.5 0.3

<130 68.8 41.3 0.7

<120 67.8 48.6 1.2

<110 67.4 51.4 1.1

<100 66.9 55.1 2

<90 65.3 66.7 6.8

<80 64.2 74.6 13.6

<70 64.5 72.5 12.2

<60 62.7 85.5 21.9

<50 64 76.1 17.7

<40 65.2 67.4 10.1

Genetic code evolution was simulated according to the 213 model (Massey 2006), utilizing a

model of code expansion facilitated by gene duplication of charging enzymes and adaptor

molecules (Massey 2008a). The 213 model proposes that the middle (2nd) position of the codon

became informational (coding) first, followed by the 1st position of the codon and finally the 3rd

position of the codon; hence “213”. The simulation was conducted as follows. Starting codons

were assigned to valine, alanine, aspartate, and glycine. These were chosen for their likely ancient

ancestry (Massey 2008a). Amino acids were randomly added to the expanding code according to

their similarity with a parent amino acid and its requisite codon. The similarity criteria were based

on the Grantham matrix and were set for each simulation. The figure shows the scheme of code

expansion according to the 213 model. There are four initial amino acids and codons; new amino

acids are added to the expanding code via duplication events, randomly but according to the

similarity criterion, until the structure of the SGC is achieved. The similarity criteria under which

code evolution was simulated are shown in the left hand column of the table. 10000 codes were

simulated for each selective criterion. The average EM values of the alternative codes are

displayed, along with the optimality of the alternative codes compared with the SGC. Table and

figure reproduced from Massey (2008a)


genomes, they are extremely rare in free-living genomes. There are two main

mechanisms for codon reassignment; the Codon Capture mechanism (Osawa and

Jukes 1989) and the Ambiguous Intermediate mechanism (Schultz and Yarus

1994). The Codon Capture mechanism proposes that an AT or GC rich codon(s)

was initially lost under extreme GC/AT pressure, and on its reappearance in the

genome, after reversal of the GC/AT pressure, was reassigned to code for a different

amino acid. The Ambiguous Intermediate mechanism proposes that a codon(s)

undergoes a period of coding ambiguity, while it is reassigned to code from one

amino acid to another.

The searching of code space was simulated computationally to test the plausibi-

lity of selecting for an error minimized genetic code (Massey 2010). An initial

alternative genetic code was generated by randomly assigning the 20 amino acids to

the 20 codon groupings of the SGC. The EM value of this code was determined.

Then a random codon reassignment was conducted, according either to the Codon

Capture or Ambiguous Intermediate mechanism. If the EM value of the new code

was better than the old, then the new code was accepted and the process repeated. If

not, then the new code was rejected, and a new random codon reassignment was

conducted on the previous code. The process was continued until the code attained

an EM value better or equal to that of the SGC. In this way, the average number of

codon reassignments required to achieve EM on a par with the SGCwas determined.

It was found that it is very difficult to select an optimal code via the Codon

Capture mechanism, with only 1.2–3.2% searches resulting in success (Table 5.2).

This effectively rules out searching of code space via the Codon Capture

Fig. 5.2 Connectivity of a

codon in the genetic code.

The connectivity of the

tyrosine TAT codon is shown,

the arrows represent point

mutations that lead to a

different sense codon. The

average value of a point

mutation of the TAT codon

is 93.25 (resulting from

(22 þ 144 þ 0 þ 0 þ 0 þ83 þ 143 þ 160 þ 194)/8,

where mutations to stop

codons and synonymous

mutations are equivalent to

zero. These values are

obtained from the Grantham

matrix). All codons in the

SGC are likewise connected,

thus the SGC may be viewed

as a regular network or graph

of 64 nodes

86 S.E. Massey

mechanism. In addition, when searching of code space is simulated using the

Ambiguous Intermediate mechanism, it is shown that 20–31 codon reassignments

on average are required to achieve an alternative code with equivalent or better EM

than the SGC (Table 5.2). This implies that there was a “burst” of codon reassign-

ments up to the last universal ancestor, which possessed the SGC, and stasis since

then. As it stands, the Adaptive Code does not provide an explanation for why this

should be. In addition, if the code were optimized via a search through code space

then it seems that the search ceased before full optimality had been achieved

Table 5.2 The number of codon reasignments required to produce a robust genetic code

Model for searching

of code space

(a) Average number of

codon reassignments

required to produce an

optimal genetic code (a

single codon reassigned

at a time)

(b) Average number of

codon reassignments


optimal genetic code

(two codons reassigned

at a time)

(c) Average number of

codon reassignments


optimal genetic code

(one or two codons

reassigned)

1. Selection for

superior EM

31 (SD ¼ 10)

0 failures

23 (SD ¼ 7)

292 failures

23 (SD ¼ 7)

0 failures

2. Selection for

superior EM,

with codon

adjacency

constraint

31 (SD ¼ 10)

3 failures

20 (SD ¼ 7)

668 failures

23 (SD ¼ 7)

0 failures

3. Selection for

superior EM,

with GC/AT

content

constraint

16 (SD ¼ 9)

968 failures

Two purine (A and G)

ending codons or

two pyrimidine

(T and C) ending

codons cannot be

reassigned under

the GC/AT

constraint


ending codons or

two pyrimidine

(T and C) ending

codons cannot be

reassigned under

the GC/AT

constraint

4. Selection for

superior EM,

with adjacency

constraint and

GC/AT content

contraint

16 (SD ¼ 6)

988 failures


ending codons or

two pyrimidine

(T and C) ending

codons cannot be

reassigned under

the GC/AT

constraint


ending codons or

two pyrimidine

(T and C) ending

codons cannot be

reassigned under

the GC/AT

constraint

1000 random codes were generated, and then the average number of codon reassignments required

to produce a code that was equal or more optimized than the SGC for EMwas determined for each.

Averages were calculated from 1,000 initial codes, except for model (3) where the average value

was calculated from the simulations out of 1,000 that produced optimal codes. Simulations that

failed to achieve an optimized genetic code are described as resulting in “failures”. Two con-

straints were applied to the simulations. The GC/AT constraint is that the codon reassigned should

be either composed of GC only or AT only. A GC rich codon is likely to be lost in an extremely

AT-biased genome, while AT rich codon is likely to be lost in an extremely GC-biased genome.

This is part of the Codon Capture mechanism. The adjacency constraint is that the amino acid

should be reassigned to a codon block adjacent to the original codon block, this follows a pattern

observed in extant codon reassignments. Reproduced from Massey (2010)


(Fig. 5.3). Again, the Adaptive Code hypothesis does not provide an explanation for

why this should be.

Additional difficulties with the Adaptive Code hypothesis are as follows. First,

when searching of code space is simulated via the Ambiguous Intermediate mecha-

nism, the structures of the codes generated differ from that of the SGC; the code

resulting from codon reassignments of single codons is more fragmented than the

SGC, the code resulting from the codon reassignment of double codons has three

amino acids (M, Q, T) that have large numbers of codons distributed throughout the

code, and the code resulting from the codon reassignment of single and double

codons displays both these features (Fig. 5.4). Second, no present day codon reas-

signment displays improved EM (Freeland et al. 2000). It is hard to envisage how a

Fig. 5.3 Increase in error minimization of random alternative genetic codes, with increasing

numbers of codon reassignments. The increase in error minimization of a randomly generated

genetic code with increasing numbers of codon reassignments was followed according to the

selective models described in Table 5.2. Each codon reassignment resulted in an increase in the

EM of the code. (a) One codon reassigned; (b) two codons reassigned; (c) one or two codons

reassigned. Data taken from Table 1. Reproduced from Massey (2010)

88 S.E. Massey

Fig.5.4

Typicalalternativegeneticcodes

thathaveundergoneoptimization.A

lternativegeneticcodes

wereproducedbycomputationalsimulation,following

theAmbiguousInterm

ediate

mechanism.Typical

codestructuresaredisplayed

asfollows:(a)resultingfrom

reassignmentsofsingle

codonsonly,having

undergone31codonreassignments;(b)resultingfrom

reassignments

oftwocodonsonly,havingundergone23codonreassignments;(c)resultingfrom

reassignmentsofacombinationof1or2codons,havingundergone23codonreassignments.Datataken

from

Table

1.Reproducedfrom

Massey(2010)


codon reassignment that improves EM can be selected for, given that every codon in

every protein would be affected by the codon reassignment andmost fitness-affecting

changes would be expected to be deleterious, whereas improved EM would only

affect a fraction of the total number of codons, for improvedmutational robustness or

robustness to phenotypic mutations. This can be expressed as follows:

Maximum number of codons j in the genome affected by improved EM

(for genotypic mutations) ¼ 3Nj� mg

where Nj is the total number of codon j in the genome and mg is the genomic

mutation rate per bp. The factor of 3 is used to convert the triplet codon. As mg isvery small, then the maximum number of codons j affected by improved EM is also

very small. In contrast, the maximum number of codons that may be adversely

affected by the reassignment, if occurring via the Ambiguous Intermediate mecha-

nism, is Nj. Indeed, there is evidence that mitochondrial codon reassignments are

deleterious (Massey and Garey 2007; Massey 2008b). This means that the potential

benefit is far outweighed by the likely deleterious effect of a codon reassignment.

This reasoning is also applicable to any improvements that a codon reassignment

may confer against transcriptional/translational errors.

Third, multiple extinctions are implied by the selective mechanism, with all

species with previous suboptimal codes being subject to mass extinction, given that

there are no extant organisms with codes that represent ancestors of the SGC. This

problem is also applicable to the alternative “Emergence” hypothesis, though to a

lesser degree.

5.4.2 The Emergence Hypothesis as an Explanationfor the Origin of EM in the SGC

If selecting for an error-minimized genetic code is problematic, what is an alterna-

tive explanation for the EM property? A mechanism for the origin of the EM

property has been proposed based on code expansion (growth) via gene duplication

of charging enzymes (aminoacyl-tRNA synthetases) and adaptor molecules

(tRNAs; Massey 2008a). An allusion to this mechanism was made by Crick

(1968), who proposed that the process of genetic code expansion occurred via

gene duplication of charging enzymes and adaptor molecules, resulting in “similar

codons coding for similar amino acids”.

In the 2008 study, it was demonstrated that a substantial amount of EM might

arise neutrally (i.e., emerge), simply as a result of the addition of similar amino

acids to similar codons, by mimicking the process of charging enzyme duplication.

Simulations were conducted on three different mechanisms of code evolution.

A substantial amount of EM was shown to arise in all three of these different

models of genetic code evolution. This result implies that no matter what the actual

90 S.E. Massey

details of code expansion (which remain to be determined), given the requirement

for charging enzymes and adaptor molecules, and the observation that gene dupli-

cates are likely to possess physicochemically related substrates, then EM is likely

an emergent property of code evolution. Thus, the conclusion is that at least a

proportion of the EM property has arisen neutrally, and was not directly selected

for, hence constituting a pseudaptation. The “emergence” of the EM property in the

simulations results from the incremental growth of the code, from the bias with

which similar amino acids are added to similar codons, and from a parent codon.

Thus, a process of “biased incrementalism” is responsible for the emergence of

mutational robustness. This process of biased incrementalism also appears to be

responsible for the emergence of mutational robustness in scale-free networks and

neutral networks as discussed below. Hence this may be a universal process in

complex systems, leading to the emergence of robustness. It is also potentially

significant that all three systems, such as SGC, scale-free networks, and neutral

networks, may be represented as graphs or networks.

5.5 The Emergence of Robustness in Other Biological Systems

5.5.1 Scale-Free Networks

There are many different types of networks in nature, from protein–protein interac-

tion networks, neuronal networks, up to ecosystem food webs and social interaction

networks. A common property in these networks is that they are usually scale-free.

The term scale-free refers to the distribution of connections of the nodes in the

network, whereby the distribution follows a power law; P(k) � k–y, where P(k) isthe fraction of nodes in the network having k connections to other nodes, and y is theexponent; y is usually between 2 and 3 in empirically observed networks. This type

of distribution means that a few nodes have many connections, while many of the

nodes only have a few connections. As most nodes only have a few connections,

scale-free networks are robust to the removal of nodes (Albert et al. 2000). In

subcellular biological networks such as metabolic and gene networks, this results in

mutational robustness, discussed below.

The scale-free property is also widespread in nonbiological systems such as the

Internet, electricity distribution networks, and transportation networks. This implies

that when observed in biological networks the property is not a product of natural

selection, but is an emergent property and hence represents a pseudaptation.

A widely accepted mechanism for the origin of the scale-free property is that of

preferential attachment during the growth of a network (Barabasi and Albert 1999).

According to the model, a well-connected node is more likely to gain more

connections, as nodes and connections are added to a growing network. This

“rich gets richer” model results in the scale-free property of the network, and it is

a passive process in that the structure of the network is not designed or selected for.


The concept falls into the larger concept of “biased incrementalism”, introduced

above to account for the emergent property of robustness in the SGC, as new

connections are added incrementally in a biased fashion (preferentially to highly

connected nodes). Three biological networks that possess the scale-free property

are discussed next.

Metabolic networks describe the metabolism of an organism. The nodes of the

network represent metabolites, while connections between the nodes represent

chemical reactions that convert one metabolite into another. These reactions are

catalyzed by enzymes. The overall flux of metabolic networks is robust to gene

deletion (Edwards and Palsson 1999, 2000). Metabolic networks are typically

scale free (Jeong et al. 2000; Ravasz et al. 2002), which means that most metabo-

lites are connected to only a few other metabolites, while a few serve as “hubs” and

are involved in many reactions. The robustness of metabolic networks to gene

deletion may be accounted for by the scale-free property (Jeong et al. 2000). The

origin of the scale-free property is unclear. There is some evidence that preferential

attachment has given rise to the scale-free property (Light et al. 2005). This

suggests that new enzymes are added to metabolism by gene duplication retain

some of the metabolites in the original reaction. This would mean that the most

connected metabolites are the most ancient, for which there is some evidence (Light

et al. 2005; Wagner 2006). Thus, the scale-free property of metabolic networks

appears to be an emergent property, and one that is beneficial to the organism in

terms of increased robustness to genetic perturbation.

Protein interaction networks attempt to characterize all of the protein�protein

interactions present within a cell. High-throughput technologies based on the yeast

two-hybrid technique or mass spectroscopy allow such networks to be constructed

on a large scale. Protein interaction networks are also typically scale-free (e.g.,

S. cerevisiae, Wagner 2001; human, Stelzl et al. 2005). Thus, these networks appear

to be tolerant of gene deletions, i.e., they are mutationally robust (Li et al. 2006),

although there is some debate as to whether the disruption of highly connected

nodes is truly more deleterious than of less connected nodes (Hahn et al. 2004). The

scale-free property appears to have arisen by preferential attachment of new edges

to highly connected nodes, without the direct action of natural selection (Wagner

2003; Berg et al. 2004).

Transcription factor networks represent the known regulatory interactions at the

transcriptional level within a cell and may be derived from various sources such as

molecular genetics and high-throughput technologies. A transcription factor net-

work consists of two types of nodes representing transcription factor genes and

genes that are the target of regulation; the transcription factor nodes are direction-

ally connected to regulated gene nodes. These networks are scale-free in humans

(Rodriguez-Caso et al. 2005), E.coli, and yeast (Maslov and Sneppen 2006), and as

a result they are robust to disruption (Balaji et al. 2006; Krishnan et al. 2007;

Guzman-Vargas and Santillan 2008). The precise mechanics leading to the scale-

free property remain to be determined, but appear to be related to the observation

that transcription factors with transcripts that have a short half life are more highly

connected (Wang and Purisima 2005).

92 S.E. Massey

5.5.2 Neutral Networks

Neutral networks are hyperdimensional regions of sequence space theoretically

applicable to both RNA and protein sequences and are represented using graph

theory. Nodes in the networks represent individual sequences, while connections

between the nodes represent neutral mutations leading from one sequence to

another. Migration of sequences to highly connected regions of the neutral network

is likely, simply by chance (van Nimwegen, Wilke 2001; Fig. 5.5). By definition,

sequences residing in such regions would have a greater number of connections,

and thus a greater proportion of potential neutral mutations, increasing their muta-

tional robustness. Hence, the stochastic migration of sequences to highly connected

regions of a neutral network is expected to result in the passive emergence of

mutational robustness in the absence of direct selection, this constitutes a pseudap-

tation. Natural selection still has an important role, in the formation of the neutral

network in the first place. The migration of sequences through sequence space to

highly connected regions is consistent with the concept of biased incrementalism, in

that a sequence will change incrementally a point mutation at a time (usually), and

the type of mutation that gets fixed is biased toward neutral mutations.

5.5.2.1 Neutral Networks in RNA

Efficient algorithms exist to predict RNA secondary structures (Tacker et al. 1996).

Simulation studies using simple RNA structures have demonstrated the existence of

neutral networks in RNA molecules (Schuster et al. 1994). The acquisition of

robustness by stochastic migration to highly connected regions of the neutral

network has also been demonstrated using computer simulations of simple RNA

structures, consistent with one of the predictions of neutral network theory (van

Nimwegen et al. 1999; Szollosi and Derenyi 2008). There is some evidence that

viral RNAs (Huynen et al. 1993; Wagner and Stadler 1999; Sanjuan et al. 2006) and

Fig. 5.5 Migration to a highly connected region of a neutral network. On the left is a simplified

representation of a neutral network structure. Nodes represent sequences, while edges represent

neutral point mutations between the sequences. The structure on the right represents the most

likely distribution of sequences that will be achieved by simple stochastic change, with larger

nodes representing more likely sequences


micro RNAs (Borenstein and Ruppin 2006; Shu et al. 2008) are mutationally

robust. The extent to which the robustness of wild type RNA molecules has arisen

by migration through neutral networks remains to be determined.

5.5.2.2 Neutral Networks in Proteins

The concept of a neutral network was initially proposed to account for the obser-

vation that there was likely extensive sequence redundancy amongst proteins

(Maynard Smith 1970. The term “neutral network” was coined by Schuster et al.

1994). There is indirect evidence for the presence of neutral networks in proteins.

This includes the observation that proteins may retain the same structure and

function, but vary extensively in sequence. The existence of neutral networks has

been shown by simulation studies in 2D lattice models (Govindarajan and Goldstein

1997; Bornberg-Bauer and Chan 1999; Xia and Levitt 2002). Babjide et al. (1997)

also present evidence for the existence of neutral networks in wild type proteins

using knowledge-based interaction potentials to calculate the stability of a sequence

in a given 3D structure. Simulation studies on 2D lattice proteins suggest that the

acquisition of mutational robustness may occur in these simplified models (Taverna

and Goldstein 2002). There is a substantial literature demonstrating the robustness

of proteins to point mutations (reviewed in Wagner 2005, Chap. 5). Whether extant

proteins have acquired robustness from migration through neutral networks is

unclear, but should be a fruitful area of future research.

5.6 The Difficulty of Selecting for Mutational Robustness

Mutational robustness may take two forms; intrinsic and extrinsic (Elena et al.

2006). Intrinisic robustness is an innate property of a sequence or network, while

extrinsic robustness is robustness that arises by the application of an external factor,

such as a heat shock protein, DNA repair or any homeostatic mechanism. In this

chapter, we have been concerned only with intrinsic robustness. The ability of

natural selection to select for intrinsic mutational robustness is unclear. Theoreti-

cally, intrinsic mutational robustness could be directly selected for via group

selection, whereby mutationally robust populations would be more successful in

competing with less robust populations. However, evidence for such a group

selection effect is elusive (Okasha 2001). The occurrence of high mutation rates

has been proposed to lead to selection for intrinsic mutational robustness (Schuster

and Swetina 1988); this has been demonstrated in digital organisms (Wilke et al.

2001), but examples from nature are scarce. The mechanistic difficulties of directly

selecting for intrinsic mutational robustness are consistent with the observations of

the emergence of mutational robustness in various biological systems in the

absence of direct selection.

94 S.E. Massey

A mechanism that may indirectly produce intrinsic mutational robustness is to

select for robustness to phenotypic mutations (defined as errors in transcription and/

or translation). Sequences that have been selected to be robust to phenotypic

mutations will by default be robust to genotypic mutations. This process has been

demonstrated experimentally (Goldsmith and Tawfik 2009). In this case, although

beneficial, the property of mutational robustness is not directly selected for, and so

constitutes a pseudaptation.

5.7 Is Mutational Robustness the Only Pseudaptation?

Here, several instances of mutational robustness have been described as pseudapta-

tions, that is, beneficial traits that have not been directly selected for. However,

while these may be beneficial in the short term for ameliorating the effects of

deleterious mutations, mutational robustness may not be beneficial in the long term

because it reduces phenotypic variability (Lenski et al. 2006). However, increased

robustness is not only detrimental in the long term; robustness also results in an

increase in the amount of neutrality, which can act to improve adaptability by

increasing the accessibility of sequence space. For instance, an interesting example

of how the robustness of the SGC may improve adaptability is explored by Zhu and

Freeland (2006). This apparent contradiction may be resolved by distinguishing

between genotypic and phenotypic robustness; genotypic (sequence) robustness

tends to decrease adaptability by reducing variation, while phenotypic robustness

tends to increase adaptability by allowing sequences to vary, thus accessing novel

areas of sequence space (Wagner 2008), for which there is empirical evidence

(Bloom et al. 2006; Amitai et al. 2007). As described here, emergent robustness

appears to arise by a process of biased incrementalism in a series of biological

systems; the genetic code, neutral networks, and cellular networks. It may be worth

noting that natural selection itself is a form of biased incrementalism; biased in that

over time generations see an incremental increase in fitness due to the biased

fixation of adaptive mutations. Fitness is analogous to survivability, which could

be viewed as a form of robustness.

Evolvability, or adaptability, is another beneficial trait that is sometimes

described as an adaptation. Evolvability refers to the ability to adapt to new

environmental challenges, thus the long-term survival of a lineage with enhanced

evolvability will be more likely. Whether evolvability can be selected for is unclear,

the anticipatory nature of selecting for evolvability being problematic. An example

is the evolution of sex as a mechanism to increase adaptability (Otto and

Lenormand 2002, a review). Acknowledging that enhanced evolvability will lead

to improved success of a population over the long term, the process of selecting for

evolvability remains to be demonstrated; thus, some cases may turn out to be

pseudaptations that are not selected for at the individual level. For example, lateral

gene transfer may lead to increased adaptability, notably in bacteria, but whether it

is a trait that is selected for that purpose is unclear. Likewise, human social


behaviors such as worship, music, and morality are presently at the center of a

debate as to whether they may be regarded as adaptations, the product of natural

selection, or are nonadaptive in origin. The field of evolutionary psychology

promises to reveal which of these behaviors were directly selected for. Those

behaviors that are beneficial to an individual, but turn out not to have been selected

for may be better described as pseudaptations.

While adaptations often typify an individual species, the instances of mutational

robustness characterized here as pseudaptations do not appear to. For instance, the

mutational robustness of the genetic code is universal to all organisms. In addition,

emergent properties, as they are often invisible to selection, are not necessarily

restricted to beneficial traits, but are also likely to encompass deleterious traits also.

Examples of such deleterious emergent traits await characterization.

Acknowledgments This work was supported by the Biology Department, University of Puerto

Rico – Rio Piedras, Puerto Rico.

References

Albert R, Barabasi AL (2002) Statistical mechanics of complex networks. Rev Modern Phy

74:47–94

Albert R, Jeong H, Barabasi AL (2000) Error and attack tolerance of complex networks. Nature

406:378–382

Alff-Steinberger C (1969) The genetic code and error transmission. Proc Natl Acad Sci USA

64:584–591

Amitai G, Devi Gupta R, Tawfik DS (2007) Latent evolutionary potentials under the neutral

mutational drift of an enzyme. HFSP J 1:67–78

Ardell DH (1998) On error minimization in a sequential origin of the standard genetic code. J Mol

Evol 47:1–13

Babjide A, Hofacker IL, Sippl MJ, Stadler PF (1997) Neutral networks in protein space:

a computational study based on knowledge-based potentials of mean force. Fold Des

2:261–269

Balaji S, Iyer LM, Aravind L, Babu MM (2006) Uncovering a hidden distributed architecture

behind scale-free transcriptional regulatory networks. J Mol Biol 360:204–212

Barabasi AL, Albert R (1999) Emergence of scaling in random networks. Science 286:509–512

Berg J, Lassig M, Wagner A (2004) Structure and evolution of protein in interaction networks:

a statistical model for link dynamics. BMC Evol Biol 4:51

Bloom JD, Labthavikul ST, Otey CR, Arnold FH (2006) Protein stability promotes evolvability.


Bornberg-Bauer E, Chan HS (1999) Modeling evolutionary landscapes: mutational stability,

topology and superfunnels in sequence space. Proc Natl Acad Sci USA 96:10689–10694

Borenstein E, Ruppin E (2006) Direct evolution of genetic robustness in microRNA. Proc Natl

Acad Sci USA 103:6593–6598

Crick FHC (1968) The origin of the genetic code. J Mol Biol 38:367–379

Di Giulio M (1989) The extension reached by the minimization of polarity distances during the

evolution of the genetic code. J Mol Evol 29:288–293

Edwards JS, Palsson BO (1999) Systems properties of the Haemophilus influenzae Rd metabolic

genotype. J Biol Chem 274:17410–17416

Edwards JS, Palsson BO (2000) Robustness analysis of the Escherichia coli metabolic network.

Biotech Prog 16:927–939

96 S.E. Massey

Elena SF, Carrasco P, Daros J-A, Sanjuan R (2006) Mechanisms of genetic robustness in RNA

viruses. EMBO Rep 7:168–173

Freeland SJ, Knight RD, Landweber LF, Hurst LD (2000) Early fixation of an optimal genetic

code. Mol Biol Evol 17:511–518

Gilis D, Massar S, Cerf NJ, Rooman M (2001) Optimality of the genetic code with respect to

protein stability and amino-acid frequencies. Genome Biol 2:11

Gleiss PM, Stadler PF, Wagner A (2001) Relevant cycles in chemical reaction networks. Adv

Complex Sys 1:1–18

Goldsmith M, Tawfik DS (2009) Potential role of phenotypic mutations in the evolution of protein

expression and stability. Proc Natl Acad Sci USA 106:6197–6202

Goodarzi H, Nejad HA, Torabi N (2004) On the optimality of the genetic code, with consideration

of termination codons. BioSystems 77:163–173

Gould SG (1997) The exaptive excellence of spandrels as a term and prototype. Proc Natl Acad Sci

USA 94:10750–10755

Gould SG, Lewontin RC (1979) The spandrels of San Marco and the panglossian paradigm: a

critique of the adaptionist program. Proc R Soc Lond B 205:581–598

Govindarajan S, Goldstein RA (1997) Evolution of model proteins on a foldability landscape.

Proteins 29:461–464

Grantham R (1974) Amino acid difference formula to help explain protein evolution. Science

185:862–864

Guzman-Vargas L, Santillan M (2008) Comparative analysis of the transcription-factor gene

regulatory networks of E.coli and S.cerevisiae. BMC Syst Biol 2:13

Hahn MW, Conant GC, Wagner A (2004) Molecular evolution in large genetic networks: does

connectivity equal constraint. J Mol Evol 58:203–211

Haig D, Hurst LD (1992) A quantitative measure of error minimization in the genetic code. J Mol

Evol 33:412–417

Huxley J (1942) Evolution: the modern synthesis. MIT Press, Cambridge Massachusetts

Huynen MA, Konings DAM, Hogeweg P (1993) Multiple coding and the evolutionary properties

of RNA secondary structure. J Theor Biol 185:251–267

Jeong H, Tombor B, Albert R, Oltvai ZN, Barabasi AL (2000) The large-scale organization of

metabolic networks. Nature 407:651–654

Krishnan A, Tomita M, Giuliani A (2007) Evolution of gene regulatory networks: robustness as an

emergent property of evolution. Phys Stat Mech Appl 387:2170–2186

Lenski RE, Barrick JE, Ofria C (2006) Balancing robustness and evolvability. PLOS Biol 4:e428

Li D, Li J, Ouyang S, Wang J, Wu S, Wan P, Zhu Y, Xu X, He F (2006) Protein interaction

networks of Saccharomyces cerevisiae, Caenorhabditis elegans andDrosophila melanogaster:Large-scale organization and robustness. Proteomics 6:456–461

Light S, Kraulis P, Elofsson A (2005) Preferential attachment in the evolution of metabolic

networks. BMC Genom 6:159

Maslov S, Sneppen K (2006) Large-scale topological properties of molecular networks. In:

Koonin E, Wolf Y, Karev G (eds) Power laws, scale-free networks and genome biology.

Springer, New York

Massey SE (2006) A sequential “2-1-3” model of genetic code evolution that explains codon

constraints. J Mol Evol 62:809–810

Massey SE, Garey JR (2007) A comparative genomics analysis of codon reassignments reveals a

link with mitochondrial proteome size and a mechanism of genetic code change via suppressor

tRNAs. J Mol Evol 64:399–410

Massey SE (2008a) A neutral origin of error minimization in the genetic code. J Mol Evol

67:510–516

Massey SE (2008b) The proteomic constraint and its role in molecular evolution. Mol Biol Evol

25:2557–2565

Massey SE (2010) Searching of code space for an error minimized genetic code via Codon Capture

leads to failure, or requires at least 20 improving codon reassignments via the Ambiguous

Intermediate mechanism. J Mol Evol 70:106–115


Maynard Smith J (1970) Natural selection and the concept of a protein space. Nature 225:563–564

Mayr E (2001) What evolution is. Basic Books, New York

Okasha S (2001) Why won’t the group selection controversy go away? Brit J Phil Sci 52:25–50

Osawa S, Jukes TH (1989) Codon reassignment (codon capture) in evolution. J Mol Evol

28:271–278

Otto SP, Lenormand T (2002) Resolving the paradox of sex and recombination. Nat Rev Gen

3:252–261

Ravasz E, Somera AL, Mongru DA, Oltvai ZN, Barabasi AL (2002) Hierarchical organization of

modularity in metabolic networks. Science 297:1551–1555

Rodriguez-Caso C, Medina MA, Sole RV (2005) Topology, tinkering and evolution of the human

transcription factor network. FEBS J 272:6423–6434

Sanjuan R, Forment J, Elena SF (2006) In silico predicted robustness of viroids RNA secondary

structures. I. The effect of single mutations. Mol Biol Evol 23:1427–1436

Schultz DW, Yarus M (1994) Transfer RNA mutation and the malleability of the genetic code.

J Mol Biol 235:1377–1380

Schuster P, Swetina J (1988) Stationary mutant distributions and evolutionary optimization. Bull

Math Biol 50:635–660

Schuster P, Fontana W, Stadler PF, Hofacker IL (1994) From sequences to shapes and back: a case

study in RNA secondary structures. Proc Roy Soc Lond B 255:279–284

Shu W, Ni M, Bo X, Zheng Z, Wang S (2008) In silico genetic robustness analysis of secondary

structural elements in the miRNA gene. J Mol Evol 67:560–569

Stelzl U et al (2005) A human protein–protein interaction network: a resource for annotating the

proteome. Cell 122:957–968

Szollosi GJ, Derenyi I (2008) The effect of recombination on the neutral evolution of genetic

robustness. Math Biosci 214:58–62

Tacker M, Stadler P, Bornberg-Bauer E, Hofacker I, Schuster P (1996) Algorithm independent

properties of RNA secondary structure predictions. Eur Biophys J Biophys Lett 25:115–130

Taverna DM, Goldstein RA (2002) Why are proteins so robust to site mutations? J Mol Biol

315:479–484

Van Nimwegen E, Crutchfield JP, Huynen M (1999) Neutral evolution of mutational robustness.


Wagner A (2001) The yeast protein interaction network evolves rapidly and contains few redun-

dant duplicate genes. Mol Biol Evol 18:1283–1292

Wagner A (2003) How the global structure of protein interaction networks evolves. Proc R Soc

Lond B 270:457–466

Wagner A (2005) Robustness and evolvability in living systems. Princeton University Press,

Princeton

Wagner A (2006) The connectivity of large genetic networks. Design, history or mere chemistry?

In: Koonin E, Wolf Y, Karev G (eds) Power laws, scale-free networks and genome biology.

Springer, New York

Wagner A (2008) Robustness and evolvability: a paradox resolved. Proc R Soc B 275:91–100

Wagner A, Stadler PF (1999) Viral RNA and evolved mutational robustness. J Exp Zool

285:119–127

Wang E, Purisima E (2005) Network motifs are enriched with transcription factors whose

transcripts have short half-lifes. Trends Gen 21:492–495

Wilke CO (2001) Adaptive evolution on neutral network. Bull Math Sci 63:715–730

Wilke CO, Wang JL, Ofria C, Lenski RE, Adami C (2001) Evolution of digital organisms at high

mutation rates leads to survival of the flattest. Nature 412:331–333

Xia Y, Levitt M (2002) Roles of mutation and recombination in the evolution of protein thermo-

dynamics. Proc Natl Acad Sci USA 99:10382–10387

Zhu W, Freeland S (2006) The standard genetic code enhances adaptive evolution of proteins.

J Theor Biol 239:63–70

98 S.E. Massey

Part IIGenome / Molecular Evolution

Chapter 6

Transferomics: Seeing the Evolutionary Forest

Using Phylogenetic Trees

John W. Whitaker and David R. Westhead

Abstract Horizontal gene transfer (HGT) is the movement of genetic material

between species that would otherwise have isolated heritages. The immediate gain

of a gene, or sets of genes, allows traits to be acquired far more rapidly than through

Darwinian evolution. The entire set of genes within a species that were acquired

through HGT is known as its transferome. Studies of prokaryotes transferomes have

revealed that the propensity of a gene to be transferred is related to biological

network structure. Recent increases in the number of sequenced eukaryotic gen-

omes have made it possible to carry out analysis of their transferomes, and this has

revealed novel insight into eukaryotic evolution. In this chapter, we present a

review of some studies that have increased our understanding of transferomes.

6.1 Introduction

Inheritance is the movement of genetic information from one generation to the next.

Traditionally information only flows vertically, from parent to offspring, within the

same species. This rule of inheritance is so important that it forms the commonly

used law upon which a species is defined. That is, for two groups of organisms to be

considered the same species, they must be able to reproduce and the resulting

offspring must be fertile. During traditional inheritance, new traits can be acquired

within a species through the processes of mutation and selection. This has led to the

idea of evolution forming a tree, with the last universal common ancestor at the base

and modern day species at the leaves. Horizontal gene transfer (HGT) breaks the

vertical law of inheritance and allows genes to move between species. The gain of a

J.W. Whitaker and D.R. Westhead

Institute of Molecular and Cellular Biology, Garstang Building, University of Leeds, Leeds

LS2 9J, UK

e-mail: [email protected]; [email protected]


101

gene through horizontal transfer can be of great advantage as it allows traits to be

acquired far more rapidly than through mutation.

The most powerful method of predicting the genes that have been acquired through

horizontal transfer is through the construction of phylogenetic trees (Whitaker

et al. 2009c). A phylogenetic tree allows the horizontally transferred genes to be

identified, because the grouping of species within the tree will differ from that of the

accepted taxonomy. The publication of whole genome sequences brings with it

the opportunity to carry HGT predictions on a genomic scale. The study of HGT on

a genomic level is known as transferomics (Whitaker et al. 2009a) and allows for

the comparison of the levels of gene transfer between species. Moreover, it allows

gene transfer to be considered in the context of biological systems, such as

metabolic networks, thus revealing the underlying evolutionary pressures that

influence the process of gene transfer. In this chapter, we review several studies

which have carried out analysis of transferomes in the context of biological

systems. Initially, we start with some seminal work which analysed the transfer-

omes of prokaryotes and eukaryotes. Then we discuss our recent work which has

looked at the transferomes of unicellular eukaryotes.

6.2 Horizontal Gene Transfer in Prokaryotes

The process of HGT has been most extensively studied in prokaryotes (Beiko et al.

2005; Lerat et al. 2005; Zhaxybayeva et al. 2006). Prokaryotes have a number of

mobile genetic elements, including plasmids, transposons and intergrons, which

can carry genes from one species to another. To encourage the movement of genetic

elements, bacteria are able to swap DNA between cells, through conjugation, or

take it up from the environment, through transformation. Furthermore, prokaryotic

viruses, phages, can carry genes between prokaryotes through a process known as

transduction. As prokaryotes are so well adapted to HGT it is not surprising that it

occurs extensively. It has been estimated that since E. coli diverged from the

Salmonella lineage, 100 million years ago, 18% of its 4,288 genes have been

acquired through HGT (Lawrence and Ochman 1998). The observation of such

extensive HGT has led to the suggestion that the ancestry of prokaryotes would be

better represented by a network than a tree (William 1999).

The genomes of prokaryotes can be split into two parts, a core genome and a

dispensable genome, which together are termed the pan-genome (Tettelin et al.

2005). The core genome represents the genes that are common to all members of a

species and carry out the core functions. The dispensable genome represents genes

that are only partially shared between strains of a species. In Streptococcus aga-lactiae, the dispensable genome was estimated to make up 20% of the genes within

the pan-genome (Tettelin et al. 2005). Analysis of the types of genes that are

commonly transferred has brought the “complexity hypothesis”. According to the

complexity hypothesis, genes which are involved in complex interactions are

unlikely to be transferred, while genes that are in few interactions are more likely

102 J.W. Whitaker and D.R. Westhead

to be transferred. More particularly, “informational genes” (e.g. genes involved in

transcription, translation, and related process) are less likely to be transferred than

“operational genes” (e.g. genes encoding metabolic enzymes). Furthermore, analy-

sis of HGT within prokaryote metabolic networks found that genes on the periphery

of the network were more likely to be transferred than those at the core (Pal et al.

2005). Taken together, these findings have led to the suggestion that prokaryotic

genomes are in a constant state of flux, with new environment specific genes being

constantly acquired, allowing rapid adaptation (Thomason and Read 2006). In this

process, environmentally specific genes join at the networks edge allowing adapta-

tion of the existing network to the new environment. For example, the gain of an

enzyme might enable the breakdown of a rare suger, allowing it to enter glycolysis.

6.3 Horizontal Gene Transfer in Eukaryotes

HGT has not been studied as extensively in eukaryotes. In multicellular eukaryotes,

where DNA would have to be transferred into the germ line, it seems unlikely that

HGT will occur (Salzberg et al. 2001); although, it cannot be ruled out altogether. In

unicellular eukaryotes, this barrier does not exist; however, they do not possess the

same HGT machinery as prokaryotes, and therefore, it is unlikely to be as prevalent.

Sources of HGT in eukaryotes include viruses, absorption from the environment,

phagocytosis and endosymbiosis. Additionally, there are well characterised exam-

ples of gene transfer from prokaryotes to eukaryotes, e.g. the tumour inducing

plasmid of Agrobacterium tumefaciens being transferred into a plant (Chilton et al.

1977). With the sequencing of many unicellular eukaryotic genomes, it has recently

become feasible to study the extent to which HGT has occurred.

HGT that accompanies endosymbiosis is termed endosymbiotic gene transfer

(EGT) and was important in establishing the eukaryotic organelles: the mitochon-

dria and plastids. In addition to the primary endosymbiosis events that established

plastids as eukaryotic organelles, multiple endosymbioses have occurred in unicel-

lular eukaryotes (Reyes-Prieto et al. 2007; Yoon et al. 2005). An important example

is the event, or events, which gave rise to the chromalveolates, in which a hetero-

trophic eukaryote gained a plastid through endocytosis of a plastid-containing red

alga (Cavalier-Smith 1999). This brought together five genomes in one cell; two

nuclear, two mitochondrial and one plastid; and with them came the opportunity for

large scale EGT (Huang et al. 2004a; Tyler et al. 2006) (see Fig. 6.1). A further

potential source of EGT in eukaryotes is from chlamydia and may have occurred

during the establishment of the primary plastid (Becker et al. 2008; Huang and

Gogarten 2007).

Over the past few years there have been many studies that have looked at the

levels of HGT which occurred in unicellular eukaryotes (Alsmark et al. 2009;

Andersson et al. 2007; Carlton et al. 2007; Huang et al. 2004a, b; Nosenko and

Bhattacharya 2007; Richards et al. 2006; Striepen et al. 2004). Within these studies,

the genes that are most commonly found to have been gained through transfer are

6 Transferomics: Seeing the Evolutionary Forest Using Phylogenetic Trees 103

those which encode metabolic enzymes. This finding is in keeping with the com-

plexity hypothesis (Alsmark et al. 2009). Eukaryotes that have phagotrophic life-

styles have been shown to have many HGTs. The predominant levels of HGT

associated with phagocytosis and endosymbiosis have led to the suggestion that

“you are what you eat” (Doolittle 1998). Here, the host nuclei are being repeatedly

exposed to genes from endosymbionts or food bacteria. The repeated exposure

means that a gene already present in the host nucleus could be replaced by a foreign

gene. When a gene is replaced by a gene transfer of the same function, it is termed

orthologous replacement.

A recent study has identified a case of large scale HGT from bacteria, fungi and

plants into the Bdelloid rotifers (Gladyshev et al. 2008). Bdelloid rotifers are

asexual multicellular animals that are highly desiccation tolerant. It is believed

that the gain of so many genes may have been facilitated by repeated desiccation

and recovery. Furthermore, it is possible that genetic exchange between Bdelloid

rotifers could be occurring by the same mechanism. This could explain how they

have survived asexually for millions of years.

6.4 metaTIGER and Its Application to Transferomics

In this section, we shall discuss our recent studies of the transferomes of unicellular

eukaryotes. We shall begin by describing the construction and functionality of the

metabolic evolution resource, metaTIGER (Whitaker et al. 2009b). Then we shall

describe how metaTIGER was used to investigate the transferomes of ten groups of

unicellular eukaryotes (Whitaker et al. 2009a). Detailed descriptions of these works

can be found in the corresponding publications; herein, we provide a brief summary

of the studies.

Fig. 6.1 Secondary endosymbiosis and gene transfer. The large cell represents a primordial

eukaryote that initially does not possess a plastid. The smaller cell represents an alga that does

possess a plastid. The two nuclei are shown by black dotted lines, the two mitochondria are shown

by grey ovals with black boarders and the plastid is shown by a black oval with a white and greyboarder. In the image on the left, the two cells are living autonomously but in a symbiotic

relationship. In the middle image, the alga cell has been engulfed by the other cell to maximise

the surface area between the two cells allowing more efficient exchange of nutrients. Owing to the

close proximity of the cells DNA from the alga genome may be transferred into the nuclear

genome of the other cell. Over time this leads to a reduction in the size of the algal genomes. In the

image on the right, the alga nucleuses and mitochondria have been lost altogether leaving only the

plastid


6.4.1 metaTIGER

The reconstruction of metabolic networks is an essential aspect of genome analysis.

metaTIGER is the first resource to bring together the reconstructed metabolic

networks of 121 eukaryotes with detailed evolutionary information. The construc-

tion and functionality of metaTIGER are summarised in Fig. 6.2.

6.4.1.1 Enzyme Prediction

To construct metaTIGER, the websites of online sequence repositories and

sequencing centres were searched for genomic sequence data. The quality of

sequence data that was used varied from assembled genomes to expressed sequence

tags (ESTs). The enzymes that are present within the genomes were predicted

through homology to PRIAM enzymes sequence profiles (Claudel-Renard et al.

2003) using the program SHARKhunt (Pinney et al. 2005). The PRIAM enzyme

sequences profiles correspond to the conserved domains of proteins that all share

the same enzymatic function. For each of the profiles, enzymatic function is

denoted by an E.C. number. When used by SHARKhunt the conserved domains

are used to make position specific scoring matrices (PSSMs) and hidden Markov

models (HMMs), which are respectively used by PSI-BLAST (Altschul et al. 1997)

Fig. 6.2 An overview of metaTIGER website. On the left are the sources of the information that

were used in the construction of metaTIGER. In the centre grey box are the three main elements of

the metaTIGER site. On the right are the ways that a user can interact with the metaTIGER site.

Arrows show the flow of information


and GeneWise2 (Birney and Durbin 2000) to search the genomic sequence data.

SHARKhunt works by running an initial PSI-BLAST search that quickly identifies

regions of the genome which are similar to the PRIAM PSI-BLAST profile. Then

these regions are extracted matched against the corresponding HMM using Gene-

Wise2. The region of the genome that matches the HMM is extracted and used to

create a polypeptide sequence. The polypeptide is then tested using PSI-BLAST

and the original PRIAM PSI-BLAST profile. The E-values and sequences are then

given as output.

6.4.1.2 Website Construction

The enzyme predictions were uploaded into the metaTIGER relational database. To

allow the enzyme predictions to be searched and interpreted in relation to the

metabolic network, parts of the KEGG Ligand database (Kanehisa et al. 2006;

Ogata et al. 1999) were also loaded into the metaTIGER database. Furthermore,

custom KEGG pathway images, for each organism, were produced to allow the

predicted enzymes to be viewed in the context of metabolic pathways. To allow

comparative analysis of pathways two tools are provided on the metaTIGER

website. These tools allow the enzymes that are present within a particular pathway

to be compared between multiple organisms.

6.4.1.3 Phylogenetic Trees

Integrated into the metaTIGER site is evolutionary information in the form of a

phylogenetic tree for each of the predicted enzymes. When producing the phylogenetic

trees care was taken to make them in a way that reduced the chance of artefacts

and makes them suitable for the prediction of HGTs. The trees can be viewed in the

site using the interactive tree viewer iTOL (Letunic and Bork 2007) (see Fig. 6.5 for

an example of phylogenetic tree viewed via the metaTIGER site) or they can be

searched for clades of interest using PhyloGenie (Frickey and Lupas 2004). The tree

searching functionality is of particular importance, when metaTIGER is applied

to transferomic analysis, as it allows phylogenetic trees that depict HGT events to

be rapidly identified.

6.4.2 Transferome Analysis

The tools and data that are intergraded into the metaTIGER website were used to

investigate the process of HGT in unicellular eukaryotes (Whitaker et al. 2009a).

The transferome analysis was made up of four sections: identification of a high-

confidence HGT dataset; comparison of the gene transfer levels and identification

of drug targets; connectivity analysis; and enrichment analysis.


6.4.2.1 The Identification of a High-Confidence HGT Dataset

To establish a dataset of putative EGTs and HGTs (E/HGTs), the metaTIGER tree

search facility was used. Four different types of gene transfer events were identi-

fied: EGT from Cyanobacteria, EGT from Chlamydia, EGT from archaeplastida

(land plants, green alga, red alga and glaucophytes) and HGT from bacteria.

The gene transfer events were identified in ten groups of unicellular eukaryotes:

Plasmodium, Theileria, Toxoplasma, Cryptosporidium, Leishmania, Trypanosoma,Phytophthora, diatoms, Ostreococcus and Saccharomyces. These ten groups were

used as each of them had more one completed genome sequence. By using only

groups with more than one completed genome sequences, it meant that contamina-

tion in a single genome sequence would not influence the results of the analysis.

PhyloGenie tree queries were designed to identify all phylogenetic trees that

depicted the corresponding E/HGT event. The tree queries constitute a way of

screening a large database of trees for trees of potential interest; subsequent manual

checking of the identified trees is necessary and was carried out, and unconvincing

examples were rejected. When searching for trees depicting high-confidence HGT

events, only clades with bootstrap support of 70% or above were considered. A cut-

off of 70% was used because it corresponds to at least a 95% chance that the clade is

correct (Hillis and Bull 1993). The E/HGT predictions that were made can be

viewed as high-confidence predictions as three steps have been taken to ensure

their quality: (1) the use of organism groups with more than one genome sequence;

(2) the manual inspection of E/HGT depicting trees; and (3) a bootstrap cut-off of

greater than 70%.

6.4.2.2 Comparison of the Gene Transfer Levels and Identification

of Drug Targets

The level of predict E/HGTs was compared between the ten groups of unicellular

eukaryotes and is shown in Fig. 6.3. The largest number of EGTs was found in the

two photosynthetic groups (Ostreococcus and the diatoms), which confirmed that

the high-confidence dataset of E/HGT predictions were suitable for revealing gene

transfer trends. Organisms that posses a plastid like organelle but are not photosyn-

thetic (Toxoplasma, Theileria and Plasmodium) were found to retain EGTs, indi-

cating non-photosynthetic metabolic activities have been gained through EGT.

Furthermore, organism groups that have believed to have once possessed a plastid,

which is now lost (Cryptosporidium and Phytophthora), have also retained EGTs,

indicating that enzymes which function outside of the plastid have been gained. No

EGTs were found in Saccharomyces, which are not thought to have ancestrally

possessed a plastid.

There are two reasons why HGT may be good drug targets. First, the acquired

genes could have previously been specific to prokaryotes and therefore be absent

from the parasites host. Second, if the acquired gene is present within the pathogens

host but acquired version is highly divergent from the hosts copy (e.g. the acquired


gene is of prokaryotic origin) then parasite specific inhibitors can be produced. The

trypanosomatids (Trypanosoma and Leishmania) were found to possess a large

number of genes that have been gained through HGT from bacteria. As there is a

great need for new therapeutic strategies to combat these parasites, the predicted

HGTs were investigated further. This revealed that one of the HGTs, Pyruvate

decarboxylase, was already a target for the drug omeprazole, which is currently

used in the treatment of Leishmania tropica. Moreover, three HGTs were suggested

a possible new drug targets: isopentenyl pyrophosphate isomerase (see Fig. 6.4),

isocitrate dehydrogenase and pyrroline-5-carboxylate reductase.

To investigate the idea that Chlamydia assisted in the establishment of the primary

plastid (Becker et al. 2008; Huang and Gogarten 2007) the predicted EGTs were

considered. The EGT predictions were compared to identify gene transfers, where

a gene had been transferred from Chlamydia into the archaeplastida, then into a thirdlineage during secondary endosymbiosis. The following examples were identified:

four genes in the diatoms; three genes within Plasmodium and Toxoplasma; and one

gene in Theileria. These results support the idea that Chlamydia assisted in

the establishment primary plastid and show that Chlamydial genes have been trans-

ferred during secondary endosymbiosis. The metaTIGER phylogenetic tree of enoyl-

[acyl-carrier-protein] reductase is shown as an example in Fig. 6.5.

6.4.2.3 Connectivity Analysis

The enzymes whose genes have been identified as being gained through E/HGT

have successfully integrated into their new hosts metabolic networks. For the genes

to have become fixed within the lineages they must have provided an evolutionary

advantage through enhancement of the metabolic network. If two or more enzyme

encoding genes, which are connected within a metabolic pathway, are co-trans-

ferred they could provide a greater enhancement to the metabolic network than two

Fig. 6.3 The metabolic transferomes. The total number of enzymes found in metaTIGER with an

E-value beneath 1.0 � 10�30 are shown for each of the groups of unicellular eukaryotes. The

counts of E/HGTs are indicated by the differently coloured bars


enzymes that are not connected. This greater potential for enhancement could

provide greater evolutionary pressure for the fixation of co-transferred genes,

which encode enzymes that are connected within metabolic pathways. To investi-

gate this, the number of connexions between enzymes, whose genes were acquired

via horizontal transfer, was considered.

The connectivity analysis worked by comparing the number of connexions

between E/HGTs to the number of connexions between enzymes picked at random

from the species metabolic network. Enzymes were taken as being connected if

they carried out consecutive reactions within the organisms metabolic network.

This analysis was carried out separately for EGTs and HGTs. For EGTs, it was

found that the number of connections was significantly greater than random. In

particular, the number of connections between EGTs from the organisms that have

now lost their plastids (Cryptosporidium and Phytophthora) was found to be

significantly greater than random. This demonstrates the co-transfer of enzyme

encoding genes that are connected within the metabolic network but do not function

within the plastid.

When the connectivity analysis was applied to the HGTs from bacteria, no

organism groups were found to be significantly more connected than random.

Fig. 6.4 The mevalonate

isoprenoid biosynthesis

pathways in T. cruzi.Enzymes are shown by

E.C. number. The enzyme

isopentenyl pyrophosphate

isomerase (5.3.3.2) is a

predicted HGT in the

trypanosomatids. The enzyme

farnesyl diphosphate synthase

(2.5.1.10) has been validated

as a drug target in T. cruzi,suggesting that isopentenyl

pyrophosphate isomerase

would also be an effective

drug target


Fig.6.5

ThemetaT

IGERphylogenetic

tree

ofenoyl-[acyl-carrier-protein]reductase.Thephylogenetic

tree

ontheleftshowstheentire

phylogenetic

tree

of

enoyl-[acyl-carrier-protein]reductase(1.3.1.9)as

viewed

onthemetaT

IGERwebsite.Thephylogenetictree

ontherigh

tshowsasinglecladewhichhas

been

enlarged

andhad

certaintaxahighlighted.Bacterialtaxaarehighlightedwithagrey

backgroundandeukaryotictaxaarehighlightedwithablackbackground.

Certain

taxahavebeenenlarged

tomakethetree

easier

tointerpret(N

B:ThediatomsarePha

eoda

ctylum

tricornu

tum

andTha

lassiosira

pseudo

nana

)


However, the two groups with the largest numbers of HGTs (Leishmania and

Ostreococcus) approached statistical significance. Thus, suggesting that if a less

stringent criterion had been used during HGT prediction, significances might have

also been found for HGTs from bacteria.

6.4.2.4 Enrichment Analysis

Genes are commonly characterised according to three ontology categories: molec-

ular function, biological process and cellular location. In the case of metabolic

enzymes, these categories relate to the chemical reactions they catalyse, the path-

ways and sub-networks within which they function and the location within the cell

where they function. Enrichment analysis was carried out to investigate if the genes

encoding enzymes in particular pathways or of particular molecular function are

more likely to have been acquired through HGT. Enrichment analysis of cellular

location was not carried out as it is likely to be less conserved between organism

groups than other aspects of ontology.

Of the different enrichment analyses that were used it was only enrichment of

KEGG metabolic pathways that found significant results. The pathways enrichment

was carried out on three levels: KEGG map group (large groups of related path-

ways); KEGG maps (a set of closely related pathways); and KEGG modules

(specific metabolic pathways). Aspects of plastid metabolism that are known to

occur within specific organism groups were found to be enriched with EGTs. Thus,

demonstrating that the high-confidence E/HGTs predictions can uncover the under-

lying trends of enrichment. The most significant and unexpected trend at the level

of KEGG map groups level was an enrichment of EGTs within carbohydrate

metabolism. Several examples of metabolic pathways that could be important to

the pathogenicity of some of the parasites being studied were identified: xylose

metabolism in Leishmania; 1-3-beta-glucan metabolic in Phytophthora; trehalosemetabolism in Phytophthora; and lipopolysaccharide biosynthesis in Phytophthora.

Of the pathway enrichments that may be important to pathogenicity, lipopoly-

saccharide biosynthesis pathway is the most exciting because it has not been seen

before in eukaryotes, outside than members to the archaeplastida. Moreover,

lipopolysaccharides are important pathogenicity factor in Gram-negative bacteria.

Thus, it is possible that lipopolysaccharides may be important to the pathogenicity

of Phytophthora and its discovery may aid the development of new control agents.

6.5 Conclusions

Transferomics is the study of HGT on a genomic scale and can be used to reveal the

underlying trends that influence gene transfer. Large scale transferomic studies are

no longer exclusive to prokaryotes. Transferomic analysis of eukaryotes can be


used to reveal insight into their evolution which may be useful in the development

of new therapeutic strategies.

References

Alsmark UC, Sicheritz-Ponten T, Foster PG, Hirt RP, Embley TM (2009) Horizontal gene transfer

in eukaryotic parasites: a case study of Entamoeba histolytica and Trichomonas vaginalis. In:Gogarten MB, Gogarten JP, Olendzenski L (eds) Horizontal gene transfer genomes in flux, vol

532, Methods in molecular biology. Springer, Heidelberg, pp 489–500

Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ (1997) Gapped

BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic

Acids Res 25:3389–3402

Andersson JO, Sjogren AM, Horner DS, Murphy CA, Dyal PL, Svard SG, Logsdon JM Jr,

Ragan MA, Hirt RP, Roger AJ (2007) A genomic survey of the fish parasite Spironucleussalmonicida indicates genomic plasticity among diplomonads and significant lateral gene

transfer in eukaryote genome evolution. BMC Genomics 8:51

Becker B, Hoef-Emden K, Melkonian M (2008) Chlamydial genes shed light on the evolution of

photoautotrophic eukaryotes. BMC Evol Biol 8:203

Beiko RG, Harlow TJ, Ragan MA (2005) Highways of gene sharing in prokaryotes. Proc Natl

Acad Sci USA 102:14332–14337

Birney E, Durbin R (2000) Using genewise in the Drosophila annotation experiment. Genome Res

10:547–548

Carlton JM, Hirt RP, Silva JC, Delcher AL, Schatz M, Zhao Q, Wortman JR, Bidwell SL,

Alsmark UCM, Besteiro S, Sicheritz-Ponten T, Noel CJ, Dacks JB, Foster PG, Simillion C,

Van de Peer Y, Miranda-Saavedra D, Barton GJ, Westrop GD, Muller S, Dessi D, Fiori PL,

Ren Q, Paulsen I, Zhang H, Bastida-Corcuera FD, Simoes-Barbosa A, Brown MT, Hayes RD,

Mukherjee M, Okumura CY, Schneider R, Smith AJ, Vanacova S, Villalvazo M, Haas BJ,

Pertea M, Feldblyum TV, Utterback TR, Shu C-L, Osoegawa K, de Jong PJ, Hrdy I,

Horvathova L, Zubacova Z, Dolezal P, Malik S-B, Logsdon JM Jr, Henze K, Gupta A,

Wang CC, Dunne RL, Upcroft JA, Upcroft P, White O, Salzberg SL, Tang P, Chiu C-H,

Lee Y-S, Embley TM, Coombs GH, Mottram JC, Tachezy J, Fraser-Liggett CM, Johnson PJ

(2007) Draft genome sequence of the sexually transmitted pathogen Trichomonas vaginalis.Science 315:207–212

Cavalier-Smith T (1999) Principles of protein and lipid targeting in secondary symbiogenesis:

euglenoid, dinoflagellate, and sporozoan plastid origins and the eukaryote family tree.

J Eukaryot Microbiol 46:347–366

Chilton M-D, Drummond MH, Merlo DJ, Sciaky D, Montoya AL, Gordon MP, Nester EW (1977)

Stable incorporation of plasmid DNA into higher plant cells: the molecular basis of crown gall

tumorigenesis. Cell 11:263

Claudel-Renard C, Chevalet C, Faraut T, Kahn D (2003) Enzyme-specific profiles for genome

annotation: PRIAM. Nucleic Acids Res 31:6633–6639

Doolittle WF (1998) You are what you eat: a gene transfer ratchet could account for bacterial

genes in eukaryotic nuclear genomes. Trends Genet 14:307–311

Frickey T, Lupas AN (2004) PhyloGenie: automated phylome generation and analysis. Nucleic

Acids Res 32:5231–5238

Gladyshev EA, Meselson M, Arkhipova IR (2008) Massive horizontal gene transfer in bdelloid

rotifers. Science 320:1210–1213

Hillis DM, Bull JJ (1993) An empirical test of bootstrapping as a method for assessing confidence

in phylogenetic analysis. Syst Biol 42:182


Huang J, Gogarten JP (2007) Did an ancient chlamydial endosymbiosis facilitate the establishment

of primary plastids? Genome Biol 8:R99

Huang J, Mullapudi N, Lancto CA, Scott M, AbrahamsenMS, Kissinger JC (2004a) Phylogenomic

evidence supports past endosymbiosis, intracellular and horizontal gene transfer in Cryptospo-ridium parvum. Genome Biol 5:R88

Huang J, Mullapudi N, Sicheritz-Ponten T, Kissinger JC (2004b) A first glimpse into the pattern

and scale of gene transfer in Apicomplexa. Int J Parasitol 34:265–274

Kanehisa M, Goto S, Hattori M, Aoki-Kinoshita KF, Itoh M, Kawashima S, Katayama T, Araki M,

Hirakawa M (2006) From genomics to chemical genomics: new developments in KEGG.

Nucleic Acids Res 34:D354–D357

Lawrence JG, Ochman H (1998) Molecular archaeology of the Escherichia coli genome. Proc Natl

Acad Sci USA 95:9413–9417

Lerat E, Daubin V, Ochman H, Moran NA (2005) Evolutionary origins of genomic repertoires in

bacteria. PLoS Biol 3:e130

Letunic I, Bork P (2007) Interactive Tree Of Life (iTOL): an online tool for phylogenetic tree

display and annotation. Bioinformatics 23:127–128

Nosenko T, Bhattacharya D (2007) Horizontal gene transfer in chromalveolates. BMC Evol Biol

7:173

Ogata H, Goto S, Sato K, Fujibuchi W, Bono H, Kanehisa M (1999) KEGG: Kyoto Encyclopedia

of Genes and Genomes. Nucleic Acids Res 27:29–34

Pal C, Papp B, Lercher MJ (2005) Adaptive evolution of bacterial metabolic networks by

horizontal gene transfer. Nat Genet 37:1372–1375

Pinney JW, Shirley MW, McConkey GA, Westhead DR (2005) metaSHARK: software for

automated metabolic network prediction from DNA sequence and its application to the

genomes of Plasmodium falciparum and Eimeria tenella. Nucleic Acids Res 33:1399–1409Reyes-Prieto A, Weber APM, Bhattacharya D (2007) The origin and establishment of the plastid

in algae and plants. Annu Rev Genet 41:147–168

Richards TA, Dacks JB, Jenkinson JM, Thornton CR, Talbot NJ (2006) Evolution of filamentous

plant pathogens: gene exchange across eukaryotic kingdoms. Curr Biol 16:1857–1864

Salzberg SL, White O, Peterson J, Eisen JA (2001) Microbial genes in the human genome: lateral

transfer or gene loss? Science 292:1903–1906

Striepen B, Pruijssers AJ, Huang J, Li C, Gubbels MJ, Umejiego NN, Hedstrom L, Kissinger JC

(2004) Gene transfer in the evolution of parasite nucleotide biosynthesis. Proc Natl Acad Sci

USA 101:3154–3159

Tettelin H, Masignani V, Cieslewicz MJ, Donati C, Medini D, Ward NL, Angiuoli SV, Crabtree J,

Jones AL, Durkin AS, DeBoy RT, Davidsen TM, Mora M, Scarselli M, Margarit y Ros I,

Peterson JD, Hauser CR, Sundaram JP, Nelson WC, Madupu R, Brinkac LM, Dodson RJ,

Rosovitz MJ, Sullivan SA, Daugherty SC, Haft DH, Selengut J, Gwinn ML, Zhou L, Zafar N,

Khouri H, Radune D, Dimitrov G, Watkins K, O’Connor KJB, Smith S, Utterback TR, White O,

Rubens CE, Grandi G, Madoff LC, Kasper DL, Telford JL, Wessels MR, Rappuoli R, Fraser CM

(2005) Genome analysis of multiple pathogenic isolates of Streptococcus agalactiae: implications

for the microbial “pan-genome”. Proc Natl Acad Sci USA 102:13950–13955

Thomason B, Read TD (2006) Shuffling bacterial metabolomes. Genome Biol 7:204

Tyler BM, Tripathy S, Zhang X, Dehal P, Jiang RH, Aerts A, Arredondo FD, Baxter L, Bensasson D,

Beynon JL, Chapman J, Damasceno CM, Dorrance AE, Dou D, Dickerman AW, Dubchak IL,

Garbelotto M, Gijzen M, Gordon SG, Govers F, Grunwald NJ, Huang W, Ivors KL, Jones RW,

Kamoun S, Krampis K, Lamour KH, Lee MK, McDonald WH, Medina M, Meijer HJ,

Nordberg EK, Maclean DJ, Ospina-Giraldo MD, Morris PF, Phuntumart V, Putnam NH,

Rash S, Rose JK, Sakihama Y, Salamov AA, Savidor A, Scheuring CF, Smith BM, Sobral BW,

Terry A, Torto-Alalibo TA, Win J, Xu Z, Zhang H, Grigoriev IV, Rokhsar DS, Boore JL (2006)

Phytophthora genome sequences uncover evolutionary origins and mechanisms of pathogenesis.

Science 313:1261–1266


Whitaker J, McConkey G, Westhead D (2009a) The transferome of metabolic genes explored:

analysis of the horizontal transfer of enzyme encoding genes in unicellular eukaryotes.

Genome Biol 10:R36

Whitaker JW, Letunic I, McConkey GA, Westhead DR (2009b) metaTIGER: a metabolic evolu-

tion resource. Nucleic Acids Res 37:D531–D538

Whitaker JW, McConkey GA, Westhead DR (2009c) Prediction of horizontal gene transfers in

eukaryotes: approaches and challenges. Biochem Soc Trans 37:792–795

William M (1999) Mosaic bacterial chromosomes: a challenge en route to a tree of genomes.

Bioessays 21:99–104

Yoon HS, Hackett JD, Van Dolah FM, Nosenko T, Lidie KL, Bhattacharya D (2005) Tertiary

endosymbiosis driven genome evolution in dinoflagellate algae. Mol Biol Evol 22:1299–1308

Zhaxybayeva O, Gogarten JP, Charlebois RL, Doolittle WF, Papke RT (2006) Phylogenetic

analyses of cyanobacterial genomes: quantification of horizontal gene transfer events. Genome

Res 16:1099–1108


Chapter 7

Comparative Genomics and Transcriptomics

of Lactation

Christophe M. Lefevre, Karensa Menzies, Julie A. Sharp,

and Kevin R. Nicholas

Abstract Lactation is an important characteristic of mammalian reproduction

sometimes referred to as the quintessence of mammals. Comparative genomics

and transcriptomics experiments are allowing a more in-depth molecular analysis of

the evolution of lactation throughout the mammalian kingdom and these recent

results are reviewed here. Milk cell and mammary gland gene expression analysis

with sequencing methodology have started to reveal conserved or specific milk

protein and components of the lactation system of monotreme, marsupial and

eutherian lineages. These experiments have confirmed the ancient origin of the

complex lactation system and provided useful insight into the function of specific

milk proteins in the control of the lactation programme or the role of milk in the

regulation of growth and development of the young beyond simple nutritive

aspects.

C.M. Lefevre

Institute for Technology Research and Innovation, Deakin University, Waurn Ponds, Geelong,

VIC 3217, Australia

CRC for Innovative Dairy Products, Department of Zoology, University of Melbourne,

Melbourne, VIC 3010, Australia

Victorian Bioinformatics Consortium, Monash University, Clayton, Melbourne, VIC 3080,

Australia


K. Menzies, J.A. Sharp, and K.R. Nicholas

Institute for Technology Research and Innovation, Deakin University, Waurn Ponds, Geelong,

VIC 3217, Australia

CRC for Innovative Dairy Products, Department of Zoology, University of Melbourne,

Melbourne, VIC 3010, Australia


115

7.1 Introduction: Lactation Evolution and Diversity

Lactation consists in the nourishment of the young with copious milk secretion by

the mammary gland. This aspect of mammalian reproduction is a defining charac-

teristic of mammals and it is often referred to as the quintessence of mammals

despite the existence of other differentiating characters (jaw bone structure, fur. . .).It has also been suggested that a key role of lactation during mammalian evolution

has been to allow the development of affectivity with an opportunity for learning,

therefore providing a substrate for the development of intelligence in man (Peaker

2002). Thus, due to its essential role in reproduction, lactation is in part responsible

for the evolutionary success of mammals.

Milk provision is a complex process, with changes in milk composition and

interactions between parent and young beyond the straightforward nutritional

function. The precise mechanism of how lactation evolved and its ancestral role

are still unclear, but a diversity of lactation strategies has been adopted by mam-

mals. Fossil and molecular evidence point to the appearance of early mammals

toward the end of the Triassic on the synaptid branch of the tree of life separating

mammalian ancestors from other living creatures during the Permian (about 320

million years ago, Fig. 7.1). Comparative genome analysis has recently emphasised

at the molecular level the ancient origin of the essential components of the lactation

system. This complex lactation system has gradually evolved during therapsid

evolution in the Triassic period and was already well established in the crown

mammals and probably in the preceding mammaliaforms of the late triassic.

Today, after more than 200 million years of evolution, the diversity of mamma-

lian species and the extreme variations in their reproductive strategies affecting in

particular the lactation cycle provide numerous examples of lineage or species-

specific adaptations of the lactation system during evolution. The earliest split in the

mammalian phylogeny established the Prototheria (monotreme or Monotrema)

which separated from the Theria about 166 (Bininda-Emonds et al. 2007) to 220

(Madsen 2009) million years ago. Theria latter split into Metatheria (marsupials or

Marsupialia) and Eutheria (eutherian or placentalia) lineages as illustrated in

Fig. 7.1. Only two genera of Monotremes have survived in Australiasia; the

platypus (Ornithorhynchus anatinus) and echidnas (Tachyglossus and Zaglosusgenera). These egg-laying Monotremes are often regarded as representative of

early mammals with a more primitive prototherian lactation system. Genomics

and transcriptomics approaches have recently enable the molecular analysis of

monotreme (Lefevre et al. 2009), marsupial (Lefevre et al. 2007) and eutherian

(Lemay et al. 2009) lactation. Comparative approaches have started to allow a

detailed analysis of the functional evolution of specific molecular components of

lactation (Menzies et al. 2009c; Topcic et al. 2009).

The recent advances in genome sequencing of a number of mammalian species

have provided invaluable resources for the comparative evolutionary analysis of

milk proteins and other genes involved in lactation. The recent release of the bovine

(Bos Taurus) genome draft has stimulated intense activity in lactation genomics

116 C.M. Lefevre et al.

(Elsik et al. 2009). Lactation gene sets have been compiled from mammary gland

cDNA libraries at multiple stages of mammary development or lactation status to

identify unique milk proteins or important mammary genes in the cow (Lemay et al.

2009) and other species including monotremes (Lefevre et al. 2009) and Marsupials

(Lefevre et al. 2007). Some of these results are reviewed here.

7.2 Milk Cell Sequencing and Monotreme Lactation

Egg-laying monotremes are regarded as close representatives of ancient mam-

mals. Tiny hatchlings are highly altricial and depend completely on milk as a

source of nutrition during the period of suckling, which is prolonged relative to

gestation and incubation (Griffiths 1978). Monotremes have no teat and milk is

excreted from a series of ducts opening directly on the surface of the ventral skin

patch of the areola. However, monotreme young exhibit a real suckling behaviour

Car

boni

fero

us

0

65

146

208

250

290

325

360

Pal

eozo

icM

esoz

oic

Cen

ozoi

c

Per

mia

nT

riass

icJu

rass

icT

ertia

ryC

reta

ceou

s

Mon

otre

mes

Mar

supi

als

Eut

heria

ns

Prototherian Therian

Metatherian

Amniotes

Synapsids Sauropsids

Therapsids

Cynodonts

Mammaliaformes

Mammals turtles,crocodiles,

dinosaurs & birds,

Placentation, viviparity

GradualAccrual of

Milk secretionby cutaneous

glands

Complex lactation system established

oviparity

Constantsecretion

ofcomplex

milk

changesin

MilkThroughout

lactation

secretionof

complexmilk

Fig. 7.1 Evolution of mammals and lactation

7 Comparative Genomics and Transcriptomics of Lactation 117

and do not simply leak the milk secretion. The role of milk in monotreme young

development remains to be established and, apart from the initial lactation period,

changes in milk composition similar to those reported in marsupials with lactation

phase-specific changes in milk protein gene expression are still controversial.

Changes in milk fat composition have been described in echidna but an effect of

diet on milk fat content has been demonstrated and milk taken from the same

platypus over a 3-month period in the wild showed no significant change in milk

fat content (Griffiths 1978). However, in Echidna, changes in milk protein

expression profiles have been reported ( Joseph and Griffiths 1992). The protein

composition of monotreme milk has been investigated. Whey proteins including

alpha-lactalbumin, lyzosyme (Guss et al. 1997; Messer et al. 1997, Shaw et al.

1993) and, more recently, whey acidic proteins WAP and WFDC2 (Sharp et al.

2007) as well as a complete set of caseins and other proteins (Lefevre et al. 2009)

have been characterised.

7.2.1 Milk Cell cDNA Sequencing

In order to collect molecular probes and develop a non-invasive sequencing

approach for the analysis of lactation in protected species such as the platypus

(Ornithorhynchus anatinus) and the short-beak echidna (Tachyglossus aculeatus), amilk cell cDNA sequencing approach has been developed (Lefevre et al. 2009).

Similar approaches may be more generally useful for the comparative analysis of

lactation in mammalian species, which may be protected or not easily bred in the

laboratory. Non-destructive approach may also be used in future experiments for

the controlled study of variation of gene expression in the same mammal during the

full course of lactation. Better knowledge of milk composition in endangered

species may be useful to conservation programmes to determine best substitution

practice for artificial feeding or cross species fostering. Milk cells preparation may

contain cells not only from mammary epithelia origin but also immune cells or cells

from skin or sebaceous glands. For example, at the end of lactation when milk

production stops, massive infiltration of immune cells into the mammary gland of

monotremes has been described (Griffiths 1978). The areola also contains seba-

ceous glands. Thus, milk cells may include skin cells, immune cells, exfoliated

epithelial cells from ducts and mammary or sebaceous glands. The purification of

milk fat globules mRNA from milk has been proposed as one possible approach for

the enrichment of mammary epithelial transcripts from human milk (Maningat et al.

2007). However, shallow milk cell cDNA sequencing during peak lactation in

monotremes has provided information about a number of caseins and whey protein

genes. Milk proteins transcripts were detected at high level indicating that mono-

treme milk is enriched in exfoliated mammary epithelial cells (Lefevre et al. 2009).

Potentially, all-milk cell fraction analysis may reveal changes in mammary gene

expression signatures from non-epithelial compartments as well. In the future, deep


sequencing will be useful to analyse the variation of gene expression and the cell

composition of milk during the course of lactation in monotremes and other species.

In exploratory experiments, milk protein sequences from platypus and echidna

were characterised including a full set of caseins. Some of these genes could not be

accurately predicted from the current platypus genome sequence annotation alone.

Sequence divergence between monotremes and other mammals represents an

average of one change per nucleotide so that neutral or rapidly evolving sequences

of monotremes and eutherians cannot be easily aligned for efficient annotation

(Warren et al. 2008). The problem is also compounded with the rapid evolution of

milk proteins such as casein (Mercier et al. 1976) typically genetically encoded by

diverse combinations of short exons and the presence of unresolved gaps in the draft

genome sequence of the platypus.

7.2.2 Monotreme Casein and the Ancient Originof the Casein Gene Cluster

Caseins are major milk proteins and their dual functionality is to serve as a source

of amino acids as well as to transport phosphate and calcium to support bone

growth of the young. Alpha and beta caseins (CSN1 and CSN2) and their variants

are also called “calcium sensitive caseins” because they precipitate easily in low

to moderate calcium concentrations. They are secreted as large calcium-depen-

dent aggregates or casein micelles sequestering calcium under the stabilising

action of kappa casein (CSN3). It was previously believed that CSN3 was evolu-

tionary unrelated to other caseins (Jones et al. 1985). However, this view has been

challenged and gene structure analysis has revealed the similar and peculiar

organisation of all casein genes, with short all in-frame exons, placing them

together with other calcium-binding phosphoproteins in a new protein family

(Kawasaki and Weiss 2003). Monotreme milk cells express all types of caseins

and casein variants (Lefevre et al. 2009) similar to those reported in other

mammals (Rijnkels 2002). In Fig. 7.2, the organisation of the monotreme casein

cluster locus is compared with other mammals, including a marsupial (the opos-

sum Monodelphis domestica) and eutherians. A physical linkage of casein genes

is seen in the casein locus of all mammalian genomes examined and the locus has

expended during mammalian radiation. A recent duplication of beta casein

occurred in the monotreme lineage. Similar duplications have also occurred

recursively along eutherian lineages (Rijnkels 2002). Casein sequences exhibit a

rapid evolution. This is in part due to extensive exon usage variation. As in other

mammals, a number of casein splice variants have been identified in monotremes

and the platypus or the echidna may use different exons. Despite this variability,

the close genomic proximity of the main alpha and beta casein genes in an

inverted tail–tail orientation and the relative orientation of additional casein-

like genes and the more distant kappa casein gene are similar in all mammalian


genome sequences so far available. This configuration is likely to be important for

the concerted expression of casein genes. During mammalian evolution, the

casein cluster has expended by gene duplication within the cluster. Eutherian

have expended the most, acquiring new genes including caseins or additional

calcium-binding phosphoproteins from salivary secretion or enamel matrix

(Kawasaki and Weiss 2003). Marsupials seem to possess only one copy of each

CSN1 and CSN2 (Lefevre et al. 2007). Interestingly, marsupial beta caseins are

longer than in other mammals (Lefevre et al. 2007) suggesting that the absence of

a third casein homolog may be compensated by an apparent elongation of the

CSN2 sequence. Two models are presented for the ancient organisation of the

casein cluster in the crown mammals with either two or three calcium-binding

casein in addition to kappa-casein (Fig. 7.2). Importantly, the most complex

model is supported by similar gene organisation of eutherian CSN1S2 and full-

length monotreme CSN2b and the presence of a canonical phosphorylation site in

the most ancient monotreme CSN2b coding region. This model also implies the

deletion of the ancestral CSN1S2 in the marsupial lineage, supported by the

presence of several retrotransposon type repeats in the corresponding region of

the opossum casein locus. The simpler model is more difficult to explain as it

implies the opportunistic construction of a strong casein-like phosphorylation site

from the more ancient, non-duplicated, genome sequence and independent dupli-

cations in the eutherian lineage. Thus, it is certain that the ancestral casein locus

was already highly organised before the common ancestor of extant mammals,

and it is likely that three calcium sensing casein had already arisen from duplica-

tion in a more ancient ancestor (Lefevre et al. 2009).

0 0.05 0.1 0.15 0.2 0.25 0.3 Mb

CSN2

CSN2

CSN1S1 CSN3

CSN2

CSN1S2STAT HIS3 HIS1

NP_999876.1

ODAM

FDSCPCSN1S2b

CSN1 CSN2b CSN3

CSN1 CSN3Odam

CSN1S1

CSN2

CSN3CSN1S2 OdamHIS STAT

csnkcsna Csn1s2a Csn1s2b OdamSTAT HIS

csnb

AK05291

Platypus ultra362

Opossum chr.5

Cattle chr.6

Mouse chr.5

Human chr.4

Fig. 7.2 Comparative analysis of the casein locus in mammals


7.2.3 Monotreme Milk Transcriptome

Other genes have been identified from monotreme milk cells. The milk cell

transcriptome of platypus and echidna estimated by cDNA sequencing is presented

in Fig. 7.3. A global discrepancy was seen between the transcript frequencies in the

two species, with the platypus transcriptome largely dominated by beta-lactoglo-

bin and casein transcripts while echidna milk cell RNA includes a higher propor-

tion of whey proteins. This discrepancy is consistent with the observation that

platypus milk contains fewer whey proteins than echidna milk (Hopper and

McKenzie 1974). A number of whey protein such as alpha-lactalbumin, lacto-

transferin and WAP and WDC2 have been identified. WAP has shown extensive

rearrangements in mammalian lineages leading to a reorganisation of the number

of exons from monotremes to marsupials and eutherian while a functional gene

has been lost in human, cow and goats (Sharp et al. 2007). Interestingly, WAPdomains have been shown to carry specific functional activities in different

lineages (Topcic et al. 2009). However, the function of WAP is not fully under-

stood. The monotreme ortholog of human C6orf58, a protein of unknown function

expressed in epithelial cells of the digestive track of other mammals but not

previously identified in milk, was found to be expressed at high level in mono-

treme milk cells. Putative proteins or proteins of unknown function have been

identified including a gene with high similarity to chondromodulin II which is a

positive regulator of chondrocyte proliferation (Mori et al. 1997; Yamagoe et al.

1998), a gene with similarity to prolactin inducible protein PIP (Murphy et al.

1987), and ovostatin.

Fig. 7.3 Quantitative estimates of gene expression in milk cells from monotremes. (a) Platypus

milk cell transcriptome. (b) Echidna milk cell transcriptome


7.2.4 Ancient Origins and Variability of the Lactation System

Overall the conservation of the key milk caseins, in particular their consistent

genomic organisation, indicates the early, pre-monotreme development of the

fundamental lactation mechanism across all mammals. In contrast, either the

lineage-specific gene duplications that have occurred specifically within the casein

locus of monotremes and eutherians but not marsupials or the more complex

rearrangements and losses that have occurred in WAP genes (Sharp et al. 2007),

as well as the presence of putative lineage-specific milk proteins, emphases the

independent selection on milk provision strategies to the young, likely to be linked

to different developmental strategies. The monotremes therefore provide insight

into the ancestral drivers for lactation and how these have adapted in different

lineages, including our own.

7.3 Marsupial Lactation: The Marsupial Lactation Cycle

and Mammary Gland Sequencing

Amongst mammals, marsupials exhibit one of the most interesting lactation system

with complex changes during the lactation cycle.

7.3.1 Marsupial Lactation

After a short gestation period, marsupials give birth to a relatively immature

newborn that is totally dependant on milk for growth and development during a

relatively long lactation period. Important changes occur during the lactation cycle

of marsupials in terms of mammary gland development, milk production, milk

composition as well as development or behaviour of the young (Green et al. 1983).

This is in sharp contrast with eutherians with a larger investment during gestation

(Tyndale-Biscoe et al. 1988) and milk of a relatively constant composition, apart

from the initial colostrum during the immediate postpartum period (Jenness 1986).

Marsupial milk provides essential nutrients and putative growth factors for the

development of the young and cross-fostering experiments have shown that milk

controls post-natal development (Ballard et al. 1995; Trott et al. 2003; Waite et al.

2005). Endocrine and others factors, potentially intrinsic to the mammary gland,

are likely to control milk secretion (Hendry et al. 1998) and marsupial milk con-

tains autocrine/paracrine regulators of the mammary gland (Brennan et al. 2007;

Nicholas et al. 1997). In special circumstances, macropod marsupials such as the

tammar wallaby (Macropus eugenii) and red kangaroo (Macropus Rufus) may

present asynchronous concurrent lactation, feeding concurrently two young of

different ages with milk of different compositions from adjacent mammary glands;


a new born pouch young and a few month older animal (Lemon and Bailey 1966;

Nicholas 1988). Although teat-sealing experiments have also shown gland-specific

involution in mice, the case of marsupials goes farther to demonstrate the impor-

tance of local control in the complex lactation programme of marsupial. However,

the molecular control mechanisms of marsupial milk composition are not fully

known.

7.3.2 The Tammar Wallaby: An Animal Model of MarsupialLactation

The tammar wallaby (Macropus Eugenii) is one of the most studied marsupial

models. It is an annual breeder characterised by a short pregnancy lasting 26 days

followed by an extended lactation period of about a 300 days. The lactation cycle is

divided into three phases of approximately 100 days each based on the sucking

pattern of the young (permanently attached to the teat, permanently in the pouch

and intermittently sucking, in and out of the pouch) and milk composition. Shortly

after birth, the single young weighing only 400 mg crawls into the pouch and

attaches to one of four teats, each associated with a separate mammary gland. The

chosen teat will provide all the milk during the entire period of lactation with

massive growth of the associated glandular tissue while the other three glands do

not generally participate in any milk production.

Changes in expression levels of milk protein genes have been described for a

number of milk proteins in several marsupial species. In particular, lactation stage-

specific genes, such as early lactation protein (ELP), mid-late whey acidic protein

(WAP) and late lactation proteins (LLP-A and LLP-B), have been characterised in

the tammar and other marsupial species (Bird et al. 1994; Demmer et al. 2001;

Green et al. 1980, 1991; Nicholas et al. 1987, 2001; Simpson et al. 1998; Trott et al.

2002). With the exception of WAP which is also found in milk of many eutherians

(Hennighausen and Sippel 1982) but not in humans, goat and ewe (Hajjoubi et al.

2006), all of these phase-specific milk proteins are marsupial-specific and have not

been found in eutherian or monotreme milk. Other marsupial-specific milk proteins

include trichosurin (Piotte et al. 1998) and the newly identified putative proteins

include PTMP-1 and PTMP-2 (Lefevre et al. 2007). PTMP-1 does not occur in

the genome sequence of the American marsupial opossum and may be Macropod

lineage-specific.

7.3.3 Tammar Mammary Transcriptome Sequencing

We have also reported expression of marsupial genes quantified by sampling the

mammary transcriptome at specific stages of the tammar lactation cycle (Lefevre


et al. 2007) by shallow and deep cDNA sequencing methods (Fig. 7.4). Ten percent

of the mammary transcriptome was estimated to represent marsupial-specific genes

and 15% mammal-specific genes. These results have also identified non-coding

RNA expressed during lactation. PTNC-1 is a novel non-coding RNA derived from

a region of the genome that is ultra-conserved in mammals suggesting an important

functional role. Other non-coding RNAs candidates have also been identified.

Further work will be required to characterise the function of these molecules.

During the course of lactation, the tammar mammary gland expresses a limited

number of common or phase-specific milk protein genes at high and increasing

levels. This accounts for over 60% of all transcripts during copious late lactation.

The remaining transcripts predominantly represent translational machinery compo-

nents, immune-related product or genes involved in energy production. These

results depict the lactating mammary gland as an organ highly specialised in the

synthesis of milk. Observations from the mammary tissue of late pregnant animals

have shown how the late pregnant mammary gland is primed for the rapid com-

mencement of milk production after parturition. Similarly, the large increase in

protein content of tammar milk during mid to late lactation is accompanied by an

increase of secreted milk protein gene expression in the mammary gland. Secreted

protein gene expression correlates with growth of the mammary gland, growth of

the young, milk production and milk protein synthesis, which all steadily increase

during the lactation cycle (Findlay 1982; Green et al. 1980). This global change of

gene expression in the mammary gland may reflect a combination of changes in

cellular gene expression and cell type populations within the tissue. As the mam-

mary gland size steadily increases during the lactation cycle, progressive replace-

ment of the stroma by alveolar tissue during the course of pregnancy and lactation,

and a marked increase of alveolar size during late lactation have been described

(Findlay 1982). The increase in relative abundance of milk protein transcripts may

correspond to an increase of milk protein gene expression in mammary epithelial

cells (lactogenesis) only or an increase in the number and proportion of secretory

epithelial cells in the mammary gland during lactation (mammogenesis).

Fig. 7.4 Mammary gland transcriptome from the Marsupial tammar wallaby at different stages of

the lactation cycle


The mammary transcriptome most likely represent a combination of these pro-

cesses. Transcriptomics of milk cells in this species as described above for mono-

tremes would provide interesting complementary data.

The combination of cDNA and signature digital sequencing methodologies has

highlighted some of the caveat and limitations of sequencing approaches for the

study of gene expression in the highly specialised mammary gland. In lactating

tissue with a large dominance of milk protein transcripts, sequencing is less

effective method for gene discovery. Next generation sequencing might over-

come these limitations in the future. One advantage of digital sequencing for the

estimation of gene expression over differential gene expression estimation by

microarray is that it provides an estimation of relative mRNA levels. However,

the ongoing development of marsupial microarrays will allow the detail analysis of

differential expression of a larger gene catalogue to investigate the molecular,

hormonal and cellular mechanisms involved in the regulation of lactation in

marsupials.

7.4 Eutherian Lactation: Fur Seal Adaptation and

Mammary Gland Involution

Within eutherian animal diversity the Pinniped family includes a variety of extreme

adaptations of the lactation system, containing species with some of the shortest

lactation periods or, most interestingly, species with the most elongated periods

between successive nursing periods.

7.4.1 Adaptations of Lactation in Pinnipeds

The three families of Pinnipeds, comprising Phocids (true seals), Odobenids

(walrus), and Otariids (sea lions, fur seals), evolved from a carnivorous ancestor

around 25 million years ago and diverged during the middle Miocene (10 million

years ago) (Fordyce 2002). Each family adopted different approaches to lactation.

The walrus has the lowest reproductive rate of any pinniped species. Calves

accompany their mother from birth, nursing on demand during these trips and are

not weaned for 2 years or more. Phocid seals evolved large sizes to reduce heat loss,

risk of predation and increased body reserves. This enabled them to adopt a “fasting

strategy” of lactation (Oftedal et al. 1987) whereby amassed body reserves of stored

nutrients facilitate fasting on land during continuous milk production over rela-

tively short periods (4–42 days, depending on the species).

In contrast, otariid seals retained smaller body sizes and insulating fur adopting a

“foraging lactation” strategy, breeding at rockeries to gain proximity to local prey

resources (Bonner 1984). Reduced prey availability and the need to exploit


resources farther off shore led to extended lactation (4–12 month) with a reduction

of foraging trip frequency and an extension of the foraging period. Otariid seals

produce milk with no detectable lactose and have adopted a lactation strategy which

is characterised by alternation between periods of several days of copious milk

production on shore and extended periods of maternal foraging at sea (Bonner

1984). Inter-suckling intervals of up to 23 days are among the longest ever recorded

for a mammal (Bonner 1984). For other mammals in general, accumulation of milk

in the mammary gland when suckling is interrupted causes rapid down regulation of

milk protein gene expression, followed by involution via apoptotic cell loss after

a few days (Li et al. 1997). However, in otariid the mammary gland remains

functional despite sustained interruptions in suckling activity.

7.4.2 The Mammary Transcriptome of an Otariid:The Lactating Fur Seal During the Foraging Period

The mammary transcriptome from the mammary gland of a foraging Cape fur seal

(Arctocephalus pusillus) is represented in Fig. 7.5. In contrast to the tammar

transcriptome in Fig. 7.4, milk proteins are less predominant. During foraging

periods at sea in the absence of sucking, fur seal mammary glands have been

recorded to produce 80% less milk than when lactating on land (Arnould and

Lysozyme

CSN1S2

CSN1S1

CSN2

CSN3

Serum amyloid A-3

IGJ

?

Fig. 7.5 Mammary gland transcriptome of lactating fur seal during the foraging period


Boyd 1995), and milk protein gene expression decreases (Sharp et al. 2006), aspects

which are common with cessation of sucking in other mammalian species and

characteristic of the reversible initiation phase of involution. However, in other

mammals these events are rapidly followed by involution with marked apoptotic

mammary gland cell death. The fur seal mammary gland does not pursue involution

at this time (Sharp et al. 2006) and remains active in readiness for return to land to

continue nursing the young.

7.4.3 Adaptation of Otariid Lactation Suggest a Key Rolefor Alpha-Lactalbumin in Mammary Gland Involution

Transcriptomics and genome analysis of three otariids: Cape fur seal (Arctocepha-lus pusillus), California sea lion (Zalophus californianus) and Antarctic fur seal

(Arctocephalus gazella) and three phocids: grey seal (Halichoerus grypus), ringedseal (Pusa hispida) and harbour seal (Phoca vitulina) have shown that the expres-

sion of LALBA has been knocked down during otariid evolution due to a cis-actingmutations in the promoter region (Sharp et al. 2008). LALBA encodes alpha-

lactalbumin, a milk protein involved in lactose synthesis. There are other examples

in nature where lactose is not required for milk production. In tammar wallaby

(Macropus eugenii) milk, carbohydrate is low and lactose is absent throughout peak

lactation (Messer and Elliott 1987) during which time other unknown factors act as

the major osmole, demonstrating that lactose is not necessary for milk production.

LALBA was reported to cause apoptosis of mouse and human mammary epithelial

cell lines and fur seal primary mammary cells (Sharp et al. 2008). Modified LALBA

(combined with oleic acid) also causes apoptosis of tumour cell lines (Tolin et al.

2009). Whether absence of LALBA alone or in combination with other changes is

responsible for the delay of involution in otariid remains to be established, but the

extinction of LALBA expression apparently represents a key event in the evolution

of lactation in this family.

7.5 Discussion

Genome analysis has shown that, in general, milk protein genes are not co-clustered

together in the genome except for the caseins. The conserved genomic organisation

of the caseins genes (Lefevre et al. 2009; Warren et al. 2008) or the co-clustering of

other milk proteins with mammary genes in the bovine genome (Lemay et al. 2009)

suggests that the need for coordinate expression during lactation may be an

influential factor in shaping the genome of mammals. Compared with other genes

of the bovine genomes, mammary and milk genes are more conserved in mammals

and evolve slowly in the bovine lineage. The most conserved proteins are


associated with secretory processes, especially component of the milk fat globule,

while the most divergent are associated with nutritional and immunological com-

ponents of milk (Lemay et al. 2009). In all, the high conservation of mammary

genes suggests that lactation evolved by co-opting existing structures and pathways

for the synthesis and secretion of copious milk (Lemay et al. 2009; Menzies et al.

2009c), and that a complex lactation system was already fully implemented in early

mammals. The apparently strong negative selection and the absence of positive

selection in milk and mammary genes support the hypothesis that milk evolution

has been constrained to optimise survival of both mother and offspring. Further

analysis of the mammalian diversity will be needed to confirm this or identify

differential constraints on the molecular components and biological pathways of

lactation.

Significantly more mammary gene duplications have occurred since the diver-

gence of the monotremes and therians than for other bovine genes. This variability

in copy number may be in part responsible for the variability in milk composition.

The regulation of transcription and other physiological energy partitioning pro-

cesses may also play a role and studies on the transcriptional regulation of genes in

epithelial cell culture, mammary explants or mammary gland tissue from a number

of animal models are starting to address this aspect (Brennan et al. 2008; Lemay

et al. 2007; Menzies et al. 2009a, b, c; Rudolph et al. 2003). A number of genera-

specific milk proteins have also been identified, especially in marsupials. For

ubiquitous milk proteins such as caseins and WAP, lineage-specific recombination

of protein domains has been described. A detailed analysis of the structure of WAP

has shown extensive rearrangements of the genes in mammalian lineages leading

to a reorganisation of the number of exons from monotremes to marsupials and

eutherians while a functional gene has been lost in human, cow and goats (Sharp

et al. 2007). Preliminary experiments suggest specific WAP domains carry unique

functional activities in different lineages (Topcic et al. 2009).

These studies provide a broad picture of the evolutionary landscape of lactation

revealing the importance of conserved metabolic and secretory pathways concur-

rently with the modular reorganisation of existing milk components or the appear-

ance of specific milk proteins. This mix of robustness and flexibility allows the

adoption of a diversity of lactation strategies under physiologic, behavioural or

environmental conditions. Thus, both the ancient and highly conserved or the more

variable and specific molecular components of lactation are, in part, responsible for

the success of the mammals to survive, adapt and evolve.

7.6 Conclusion

During mammalian radiation, species have diversified lactation strategies to accom-

modate reproductive success and adapt to the environment. There is much to learn

from the natural resource of animal diversity about the genetics of lactation. This

has been illustrated by the comparative analysis of gene expression in a variety of


lactating mammalian lineages. Sequencing approaches will enable a broader explo-

ration of lactation diversity. We have shown that milk cells provide easy access to

functional data. Comparative genome analysis of the lactation system is also a new

and complementary methodology. It will then become possible to study in detail

how the evolutionary constraints on lactation vary between lineages depending on

lactation strategies or environmental adaptations.

In all mammals, milk provision is a complex process with changes in milk

composition and interactions between parent and young beyond the straightforward

nutritional function. The role of milk on the mammary gland or the development of

the young is starting to emerge through studies of lactation in mammals with

extreme adaptations of the lactation systems such as fur seals or marsupials. Such

adaptations provide valuable models to enhance our understanding of the biology of

lactation. The central role of milk is best studied in animal models with extreme

adaptation to lactation that allow researchers to more easily identify regula-

tory mechanisms that are present, but not as readily apparent in eutherian species

(Nicholas et al. reviews). Early development of the eutherian young is programmed

and regulated in utero. Inappropriate signalling results in abnormal development

and mature onset disease. The marsupial gives birth to an altricial young and much

of the early development is regulated by milk. It is now apparent that new roles for

milk are emerging and future studies using the marsupial and other models will

allow researchers to more fully understand the central role of milk to deliver time-

dependent signals for both growth and development of the young, protect the young

and mammary gland from infection and regulate the development and function of

the mammary gland. A better understanding of the temporal delivery of these

signals will provide new opportunities for treatment and prevention of disease.

The results presented here have illustrated how comparative analysis of lactation

by genomics and transcriptomics enables a better understanding of the role of milk

in the programming of mammalian development.

References

Arnould JPY, Boyd IL (1995) Temporal patterns of milk production in Antarctic fur seals

(Arctocephalus gazella). J Zool 237:1–12Ballard FJ, Grbovac S, Nicholas KR, Owens PC, Read LC (1995) Differential changes in the milk

concentrations of epidermal growth factor and insulin-like growth factor-I during lactation in

the tammar wallaby, Macropus eugenii. Gen Comp Endocrinol 98:262–268

Bininda-Emonds OR, Cardillo M, Jones KE, MacPhee RD, Beck RM et al (2007) The delayed rise

of present-day mammals. Nature 446:507–512

Bird PH, Hendry KA, Shaw DC, Wilde CJ, Nicholas KR (1994) Progressive changes in milk

protein gene expression and prolactin binding during lactation in the tammar wallaby (Macro-

pus eugenii). J Mol Endocrinol 13:117–125

Bonner WN (1984) Lactation strategies in pinnipeds: problems for a marine mammalian group.

Symp Zool Soc Lond 51:253–272


Brennan AJ, Sharp JA, Digby MR, Nicholas KR (2007) The tammar wallaby: a model to examine

endocrine and local control of lactation. IUBMB Life 59:146–150

Brennan AJ, Sharp JA, Lefevre CM, Nicholas KR (2008) Uncoupling the mechanisms that

facilitate cell survival in hormone-deprived bovine mammary explants. J Mol Endocrinol

41:103–116

Demmer J, Stasiuk SJ, Grigor MR, Simpson KJ, Nicholas KR (2001) Differential expression of the

whey acidic protein gene during lactation in the brushtail possum (Trichosurus vulpecula).

Biochim Biophys Acta 1522:187–194

Elsik CG, Tellam RL, Worley KC, Gibbs RA, Muzny DM et al (2009) The genome sequence of

taurine cattle: a window to ruminant biology and evolution. Science 324:522–528

Findlay L (1982) The mammary glands of the tammar wallaby (Macropus eugenii) during

pregnancy and lactation. J Reprod Fertil 65:59–66

Fordyce RE (ed) (2002) Fossil record. Academic Press, San Diego, California, USA, pp 453–471

Green B, Newgrain K, Merchant J (1980) Changes in milk composition during lactation in the

tammar wallaby (Macropus eugenii). Aust J Biol Sci 33:35–42

Green B, Griffiths M, Leckie RM (1983) Qualitative and quantitative changes in milk fat during

lactation in the tammar wallaby (Macropus eugenii). Aust J Biol Sci 36:455–461

Green B, VandeBerg JL, Newgrain K (1991) Milk composition in an American marsupial

(Monodelphis domestica). Comp Biochem Physiol B 99:663–665

Griffiths M (1978) The biology of monotremes. Academic Press, New York, NY

Guss JM, Messer M, Costello M, Hardy K, Kumar V (1997) Structure of the calcium-binding

echidna milk lysozyme at 1.9 A resolution. Acta Crystallogr D Biol Crystallogr 53:355–363

Hajjoubi S, Rival-Gervier S, Hayes H, Floriot S, Eggen A et al (2006) Ruminants genome no

longer contains whey acidic protein gene but only a pseudogene. Gene 370:104–112

Hendry KA, Simpson KJ, Nicholas KR, Wilde CJ (1998) Autocrine inhibition of milk secretion in

the lactating tammar wallaby (Macropus eugenii). J Mol Endocrinol 21:169–177

Hennighausen LG, Sippel AE (1982) Characterization and cloning of the mRNAs specific for the

lactating mouse mammary gland. Eur J Biochem 125:131–141

Hopper KE, McKenzie HA (1974) Comparative studies of alpha-lactalbumin and lysozyme:

echidna lysozyme. Mol Cell Biochem 3:93–108

Jenness R (1986) Lactational performance of various mammalian species. J Dairy Sci 69:869–885

Jones WK, Yu-Lee LY, Clift SM, Brown TL, Rosen JM (1985) The rat casein multigene family.

Fine structure and evolution of the beta-casein gene. J Biol Chem 260:7042–7050

Joseph M, Griffiths M (1992) Whey proteins in milks of monotremes and wallabies. Australian

Mammology 14:125–127

Kawasaki K, Weiss KM (2003) Mineralized tissue and vertebrate evolution: the secretory calcium-

binding phosphoprotein gene cluster. Proc Natl Acad Sci USA 100:4060–4065

Lefevre CM, Digby MR, Whitley JC, Strahm Y, Nicholas KR (2007) Lactation transcriptomics

in the australian marsupial, Macropus eugenii: transcript sequencing and quantification. BMC

Genomics 8:417

Lefevre CM, Sharp JA, Nicholas KR (2009) Characterisation of monotreme caseins reveals lineage-

specific expansion of an ancestral casein locus in mammals. Reprod Fertil Dev 21:1015–1027

Lemay DG, Neville MC, Rudolph MC, Pollard KS, German JB (2007) Gene regulatory networks

in lactation: identification of global principles using bioinformatics. BMC Syst Biol 1:56

Lemay DG, Lynn DJ, Martin WF, Neville MC, Casey TM et al (2009) The bovine lactation

genome: insights into the evolution of mammalian milk. Genome Biol 10:R43

Lemon M, Bailey LF (1966) A specific protein difference in the milk from two mammary glands

of a red kangaroo. Aust J Exp Biol Med Sci 44:705–707

Li M, Liu X, Robinson G, Bar-Peled U, Wagner KU et al (1997) Mammary-derived signals

activate programmed cell death during the first stage of mammary gland involution. Proc Natl


Madsen O (2009) Mammals (mammalia). In: Hedges SB, Kumar SB (eds) The timetree of life.

Oxford Univeristy Press, Oxford, pp 459–461


Maningat PD, Sen P, Sunehag AL, Hadsell DL, Haymond MW (2007) Regulation of gene expres-

sion in human mammary epithelium: effect of breast pumping. J Endocrinol 195:503–511

Menzies KK, Lee HJ, Lefevre C, Ormandy CJ, Macmillan KL, Nicholas KR (2009a) Insulin, a key

regulator of hormone responsive milk protein synthesis during lactogenesis in murine mam-

mary explants. Funct Integr Genomics 10(1):87–95

Menzies KK, Lefevre C, Macmillan KL, Nicholas KR (2009b) Insulin regulates milk protein

synthesis at multiple levels in the bovine mammary gland. Funct Integr Genomics 9:197–217

Menzies KK, Lefevre C, Sharp JA, Macmillan KL, Sheehy PA, Nicholas KR (2009c) A novel

approach identified the FOLR1 gene, a putative regulator of milk protein synthesis. Mamm

Genome 20:498–503

Mercier JC, Chobert JM, Addeo F (1976) Comparative study of the amino acid sequences of the

caseinomacropeptides from seven species. FEBS Lett 72:208–214

Messer M, Elliott C (1987) Changes in alpha-lactalbumin, total lactose, UDP-galactose hydrolase

and other factors in tammar wallaby (Macropus eugenii) milk during lactation. Aust J Biol Sci

40:37–46

Messer M, Griffiths M, Rismiller PD, Shaw DC (1997) Lactose synthesis in a monotreme, the

echidna (Tachyglossus aculeatus): isolation and amino acid sequence of echidna alpha-lactal-

bumin. Comp Biochem Physiol B Biochem Mol Biol 118:403–410

Mori Y, Hiraki Y, Shukunami C, Kakudo S, Shiokawa M et al (1997) Stimulation of osteoblast

proliferation by the cartilage-derived growth promoting factors chondromodulin-I and -II.

FEBS Lett 406:310–314

Murphy LC, Tsuyuki D, Myal Y, Shiu RP (1987) Isolation and sequencing of a cDNA clone for a

prolactin-inducible protein (PIP). Regulation of PIP gene expression in the human breast

cancer cell line, T-47D. J Biol Chem 262:15236–15241

Nicholas KR (1988) Asynchronous dual lactation in a marsupial, the tammar wallaby (Macropus

eugenii). Biochem Biophys Res Commun 154:529–536

Nicholas KR, Messer M, Elliott C, Maher F, Shaw DC (1987) A novel whey protein synthesized

only in late lactation by the mammary gland from the tammar (Macropus eugenii). Biochem

J 241:899–904

Nicholas K, Simpson K, Wilson M, Trott J, Shaw D (1997) The tammar wallaby: a model to study

putative autocrine-induced changes in milk composition. J Mammary Gland Biol Neoplasia

2:299–310

Nicholas KR, Fisher JA, Muths E, Trott J, Janssens PA et al (2001) Secretion of whey acidic

protein and cystatin is down regulated at mid-lactation in the red kangaroo (Macropus rufus).

Comp Biochem Physiol A Mol Integr Physiol 129:851–858

Oftedal OT, Boness DJ, Tedmam RA (1987) The behaviour, physiology, and anatomy of lactation

in the Pinnipedia. Curr Mammal 1:175–245

Peaker M (2002) The mammary gland in mammalian evolution: a brief commentary on some of

the concepts. J Mammary Gland Biol Neoplasia 7:347–353

Piotte CP, Hunter AK, Marshall CJ, Grigor MR (1998) Phylogenetic analysis of three lipocalin-

like proteins present in the milk of Trichosurus vulpecula (Phalangeridae, Marsupialia). J Mol

Evol 46:361–369

Rijnkels M (2002) Multispecies comparison of the casein gene loci and evolution of casein gene

family. J Mammary Gland Biol Neoplasia 7:327–345

Rudolph MC, McManaman JL, Hunter L, Phang T, Neville MC (2003) Functional development of

the mammary gland: use of expression profiling and trajectory clustering to reveal changes in

gene expression during pregnancy, lactation, and involution. J Mammary Gland Biol Neoplasia

8:287–307

Sharp JA, Cane KN, Lefevre C, Arnould JP, Nicholas KR (2006) Fur seal adaptations to lactation:

insights into mammary gland function. Curr Top Dev Biol 72:275–308

Sharp JA, Lefevre C, Nicholas KR (2007) Molecular evolution of monotreme and marsupial whey

acidic protein genes. Evol Dev 9:378–392


Sharp JA, Lefevre C, Nicholas KR (2008) Lack of functional alpha-lactalbumin prevents involu-

tion in cape fur seals and identifies the protein as an apoptotic milk factor in mammary gland

involution. BMC Biol 6:48

Shaw DC, Messer M, Scrivener AM, Nicholas KR, Griffiths M (1993) Isolation, partial character-

isation, and amino acid sequence of alpha-lactalbumin from platypus (Ornithorhynchus anati-

nus) milk. Biochim Biophys Acta 1161:177–186

Simpson K, Shaw D, Nicholas K (1998) Developmentally-regulated expression of a putative

protease inhibitor gene in the lactating mammary gland of the tammar wallaby, Macropus

eugenii. Comp Biochem Physiol B Biochem Mol Biol 120:535–541

Tolin S, De Franceschi G, Spolaore B, Frare E, Canton M et al (2009) The oleic acid com-

plexes of proteolytic fragments of alpha-lactalbumin display apoptotic activity. FEBS J

277(1):163–173

Topcic D, Auguste A, De Leo AA, Lefevre C, Digby MR, Nicholas KR (2009) Characterization of

the tammar wallaby (Macropus eugenii) whey acidic protein gene: new insights into the

function of the protein. Evol Dev 11:363–375

Trott JF, Wilson MJ, Hovey RC, Shaw DC, Nicholas KR (2002) Expression of novel lipocalin-like

milk protein gene is developmentally-regulated during lactation in the tammar wallaby,

Macropus eugenii. Gene 283:287–297

Trott JF, Simpson KJ, Moyle RL, Hearn CM, Shaw G et al (2003) Maternal regulation of milk

composition, milk production, and pouch young development during lactation in the tammar

wallaby (Macropus eugenii). Biol Reprod 68:929–936

Tyndale-Biscoe H, Janssens PA, Australian Academy of Science, Australian Society for Repro-

ductive Biology, Australian Mammal Society (1988) The developing marsupial: models for

biomedical research, vol viii. Springer-Verlag, Berlin, p 245

Waite R, Giraud A, Old J, Howlett M, Shaw G et al (2005) Cross-fostering in Macropus eugenii

leads to increased weight but not accelerated gastrointestinal maturation. J Exp Zool

303:331–344

Warren WC, Hillier LW, Marshall Graves JA, Birney E, Ponting CP et al (2008) Genome analysis

of the platypus reveals unique signatures of evolution. Nature 453:175–183

Yamagoe S, Mizuno S, Suzuki K (1998) Molecular cloning of human and bovine LECT2 having a

neutrophil chemotactic activity and its specific expression in the liver. Biochim Biophys Acta

1396:105–113


Chapter 8

Evolutionary Dynamics in the Aphid Genome:

Search for Genes Under Positive Selection

and Detection of Gene Family Expansions

Morgane Ollivier and Claude Rispe

Abstract Aphids have a high adaptative potential and their capacity to adapt to

various environments could be linked with specific expansions in gene repertoires.

A large scale acquisition of genomic data has been recently undertaken with the

genome of Acyrthosiphon pisum (reference gene set) and EST data from three other

species: Myzus persicae, Aphis gossypii and Toxoptera citricida. We identified

paralogs through an intra-genomic Reciprocical Best Hit search in A. pisum and

highlighted a high and steady level of duplications in A. pisum. We assembled,

ESTs, predicted coding sequences and identified pairs of orthologs with A. pisum.We identified a fraction of fast-evolving sequences (high ratio of non-synonymous

to synonymous rates) including genes shared by aphids but not identified in non-

aphid species. Phylogenetic study of fast-evolving genes (Apo, C002, Spaetzel)

shows that rate accelerations and duplication events are linked and could favour the

emergence of specific biological functions.

8.1 Introduction

Studies of the adaptation of species to their environment have historically been

focused on analyses of phenotypic variation. The enormous increase in sequence

data now allows to directly detect at the gene level processes which contribute

to adaptation. In a given population, genes are under drift and selection effects.

Selection can act against deleterious mutation or in favour of advantageous

mutations. It is possible to detect traces of selection on genomes by comparing

M. Ollivier and C. Rispe

INRA, UMR 1099 BiO3P, Domaine de la Motte, F-35653, Le Rheu, France



133

homologous sequences from different organisms and by computing maximum

likelihood (ML) synonymous (dS) and non-synonymous (dN) evolutionary rates.

The ratio omega (dN/dS) is indeed used as an indicator of variable evolutio-

nary pressures among protein-coding genes: low ratios are typical of highly

constrained sequences (under purifying selection), while values close to unity

would reflect relaxed selection and values above unity would result from positive

selection.

Another interesting point to consider are gene duplications. Whereas the major-

ity of duplicated sequences are removed from a genome, duplications can provide

new evolutionary opportunities as duplicated genes are often under particular

selective pressures (either relaxation or positive selection).

Aphids (Insecta: Hemiptera) are small insects that feed on plant sap. Some

species are crop-feeding and considered as pests in agriculture. Their effects

on crops are enhanced by host–plant specialisation (Hawthorne and Via 2001;

Hufbauer and Via 1999) and their rapid demographic increases due to viviparous

clonal reproduction (parthenogenesis). Phenotypic plasticity (of reproductive

mode, of dispersal) enhances their high adaptive potential. Their life cycle is

remarkable as it shows alternation of asexual and sexual reproduction (Fig. 8.1).

The capacity of aphids to adapt to various environments could be linked with

shifts in gene repertories (more genes/specific gene regulation) and expansions of

specific gene families. Recently, the genome of the pea aphid Acyrthosiphon pisum(Aphidinae, Macrosiphini) has been completely sequenced as a joint effort of the

International Aphid Genomics Consortium (2010); it comprises close to 34,000

predicted genes. Collections of ESTs are available for three other aphid species,

Myzus persicae, Toxoptera citricida and Aphis gossypii. This body of data providesa significant material to analyse fine-scale evolutions (selective pressures on the

different genes, amplification of some gene families) and to relate specific evolu-

tions in the aphid genome and biological adaptations in the pea aphid genome and

between aphid species.

To better evaluate the adaptive aspects of gene repertoires and gene sequence

evolution in aphids as a group, we developed a two step approach. The first

consisted in an evaluation of the importance of duplication in the pea aphid

genome, by comparison with two others insect genomes. The second study was

a comparison of the coding genomes of different aphid species comprising

two different tribes (Macrosiphini and Aphidini) and characterised by different

life-cycle and host–plant preferences: A. pisum, M. persicae, A. gossypii andT. citricida. For these comparisons, we used the pea aphid genome and all ESTs

available for the three other species. We quantified the fraction of genes shared by

different aphid species but unknown from other insects, which thus could play a

special role in the biology of aphids. We also analysed the patterns of divergence

among putative orthologs, and especially focus on fast-evolving sequences, which

could be so as a result of positive selection and rapid adaptation to environmental

changes, or strong co-evolutionary interactions as those between insects and their

host–plant.

134 M. Ollivier and C. Rispe

8.2 Dynamic of Duplication During the Evolutionary

Time (IAGC, Plos Biology, 2010)

Genome comparisons are very efficient to detect specificity of gene repertoires

among species, like relative duplication phenomenon (e.g. as it has been done in the

Drosophila genus Zdobnov et al. 2002; Heger and Ponting 2007). In a group, it is

possible to measure the relative importance and the dynamic of duplications

between genomes. A “self-blast” of the coding sequences (CDS) from a genome

can indeed allow identifying paralogous genes. We then can measure the diver-

gence time between copies with the dS rate which is a rough measure of evolution-

ary time since duplication. This method has been efficient to detect global

duplication events in Arabidopsis thaliana (Blanc and Wolfe 2004) or Paramecium

Viviparous female

Parthenogenetic femalesn clonal generations

Eggs

“Sexual” lineage1 Sexual generation

Sexuals

Winter

Fall

Spring

Summer

Fig. 8.1 Life cycle of the pea aphid Acyrthosiphon pisum. A parthenogenetic female generates

several clonal generations. In fall, the photoperiod decreases and parthenogenetic female produces

males and oviparous females that can mate. Oviparous females produce eggs that can stay in

diapause during the winter. In spring, Viviparous females emerge from the eggs

8 Evolutionary Dynamics in the Aphid Genome 135

tetraurelium (Aury et al. 2006), which appeared as very clear peaks in the distribu-

tion of dS distances among all paralogs.

With this method, we studied the A. pisum genome and for comparison two other

insect genomes, D. melanogaster (Adams et al. 2000), which comprises more than

14,000 predicted genes, and Apis mellifera (Weinstock et al. 2006), which com-

prises about 9,000 predicted genes. Each coding genome was blasted on itself

(blastP, Evalue ¼ 1.0 e�10). Reciprocical Best Hit (RBH) (Hirsh and Fraser

2001; Jordan et al. 2002) in each genome were considered as potential gene copies

dating back to the nearest duplication event. We then aligned and computed the

synonymous mutation rate between all RBH pairs of sequences in the three

genomes using a codon-based model (Codeml from PAML; Yang 1997).

Comparison of dS gene value distributions across the pea aphid, fruitfly and

honeybee genomes (Fig. 8.2) shows a particularly high and steady level of duplica-

tions in the pea aphid genome, well above that observed in the bee and fruitly

genomes.

4000

3500

3000

2500

Pai

rs o

f par

alog

s

Classes of dS (synonymous changes per sites)

2000

1500

1000

500

00.25 – 0.500 – 0.25 0.50 – 0.75 0.75 – 1.00 1.00 – 1.25 1.25 – 1.50

Fig. 8.2 Widespread gene duplication in an ancestor of the pea aphid as suggested by the

distributions of synonymous divergences among pairs of recent paralogs (Reciprocal Best Hits)

within pea aphid, honey bee and drosophila. Black: Acyrthosiphon pisum, grey: Drosophilamelanogaster, white: Apis mellifera


8.3 Comparative Analysis of the A. pisum Genome and

EST-Based Genes Sets from Other Aphid Species

(Ollivier et al. IMB, Accepted)

Comparisons of the gene repertoires of related organisms and of the evolutionary

rates of genes may bring insights about the genes and functions that are particularly

significant at the biological level for that group of organisms.

8.3.1 Search for Orthologous Genes

We assembled ESTs for three aphid species: Myzus persicae, Aphis gossypii andToxoptera citricida. From these collections of unigenes, we predicted CDS in each

species. They are available in Aphidbase (http://www.aphidbase.com; Legeai et al.

2010). We identified putative orthologs thanks to RBH method and found 259 RBH

between the four species (restricted set biased towards genes with a high level of

expression), 4649 RBH between A. pisum and M. persicae, 1789 RBH between

A. pisum and A. gossypii and 982 RBH between A. pisum and T. citricida. Evolu-tionary rates (non-synonymous mutation rates, dN; Synonymous mutation rate, dSand dN/dS ratio) were computed between all orthologous pairs of sequences, using

codeml from PAML.

8.3.2 Pairwise Comparisons and Estimation ofEvolutionary Rates

Distributions of dN/dS ratios (Fig. 8.3) were similar for the three pairwise compa-

risons, A. pisum/M. persicae, A. pisum/A. gossypii and A. pisum/T. citricida. We

observed three L-shaped distributions with a low mode and a long right tail

corresponding to RBH with the highest ratios in all comparisons. We focused on

those sequences, as they might be fast-evolving genes of particular interest, and

found 248, 60 and 32 genes for which dN/dS > 0.4, for the three comparisons

respectively.

We also recorded all sequences that were RBH in the three pairwise comparisons

and which had no hit in Uniprot (tentative aphid-specific genes). This category

comprised 10% of all pairwise RBH, so 445, 159 and 66 genes respectively in the

three comparisons. In this sets, dN and dN/dS ratios were three times higher than

in the reference set (P value >10 � 10�3, Z-test), and 50% of those genes had a

dN/dS > 0.40.

This suggests that those genes are evolving particularly fast at the proteic level

and are under positive or relaxed selection. It can also explain why those genes are


http://www.aphidbase.com

only recognised within aphids: they may have diverged too much from other related

sequences in other animals group.

8.3.3 Phylogenetic Analyses of Two Fast-Evolving Sequences

8.3.3.1 Gene “Apo”: Example of Specific Lineage Duplications

This gene, with no similarity in Uniprot database, presented high dN/dS ratios in

pairwise comparisons. We found this gene in all aphid species and in four copies in

A. pisum. AnML phylogenetic tree (Fig. 8.4) strongly supported the grouping of the

four A. pisum copies, suggesting a lineage-specific duplication. A free–ratio model

(PAML) was significant and showed an increase of the dN/dS ratio for Apo2 (1.69),the Apo3/Apo4 (1.66) group and the ancestral branch to A. pisum (2.02); whereas

the dN/dS ratio for M. persicae, T. citricida and A. gossypii branches are under

0.40. The ratio increases were related with duplication events. Similar pattern was

found for other sequences like Juvenil Hormone Acid Methyl transferase and

Glycosyl-hydrolase (see Ollivier et al. 2010, IMB). In each case, we found strong

increases of the dN/dS ratios consistent with specific lineage duplications. This

shows that duplication strongly influenced evolutionary rates, possibly as the result

of an adaptative process.

2500

2000

1500

1000

500

00 – 0.1 0.1 – 0.2 0.2 – 0.3

dN/dS ratio

Num

ber

of R

BH

0.3 – 0.4 > 0.4

Fig. 8.3 Distribution of the estimated pairwise ratio of non-synonymous to synonymous diver-

gence, for RBH genes among the pea aphid (complete genome) and EST-based gene sets from

each of three other aphid species. White: A. pisum/M. persicae, black: A. pisum/T. citricida, grey:A. pisum/A. gossypii


8.3.3.2 Protein C002: A Specific Protein of Aphid Lineage

This protein, as an example, presents a high dN/dS ratio between A. pisum and

M. persicae (0.57). This gene has no hit in uniprot. We found these genes in a single

copy in the four species considered. The global dN/dS ratio (one–ratio model from

PAML) computed on the species tree was exceptionally high at 0.73. This gene has

recently been identified as specific to salivary glands and essential in feeding (Mutti

et al. 2008): this protein is transferred from aphid to plant during feeding; C002

knock-down insects die prematurely. We may thus interpret this very high rate as

the result of an adaptative response of strong plant interaction. The fact that these

gene has no homologs in other insects group suggests too a specific adaptation.

8.3.4 Functional Annotation of Fast-Evolving Genes

We compiled the 5139 A. pisum sequences found in RBH pairs: 3141 could be

annotated through Blast2GO (Conesa et al. 2005; http://www.blast2go.org/) with

26.138 GO terms. We found an annotation for 60% of A. pisum sequences, but,

A. pisum - Apo1

A. pisum - Apo2

A. pisum - Apo3

A. pisum - Apo4

M. persicae

A. gossypii

T. citricida

95

100

100

0.05

Fig. 8.4 Maximum likelihood tree of “Apo” gene in four aphid species (�Lnl ¼ 1683.99,

Gamma ¼ 2.21; Likelihood settings from best-fit model (TrNþG) selected by AIC in Modeltest).

Bootstraps values indicated under nodes


http://www.blast2go.org/

analysing separately the “Fast-Evolving” (dN/dS > 0.40) genes, only 30% were

annotated. The sets of annotated A. pisum sequences were too small to make

statistical comparisons in A. pisum/A. gossypii and A. pisum/T. citricida compar-

isons. However, in the A. pisum/M. persicae comparison, we found significant

differences among frequencies of GO categories between the “fast-evolving” subset

of sequences and the rest of genes. 22 GO were over represented in the subset

(P value < 0.01, exact Fisher’s test). One category that appears significantly

enriched under Fisher’s test is of particular interest: genes annotated as “defence

response to fungus”, genes “cactus” and “Spaetzel”. They are involved in develop-

ment and innate immunity in the Toll signalling pathway. Genes involved in defence

and immunity are relatively few in A. pisum overall (Gerardo et al. 2010). dN/dSratios are respectively 0.50 and 0.44 for the Cactus and Spaetzel gene, and while

Cactus is single copy, we found five copies in Spaetzel gene resulting from a serial

lineage duplication. These duplications may have enhanced increases of non-synon-

ymous substitution rates in Spaetzel lineage. Aphids present a particular immune

system pattern and genes involved in this function seem evolving in a particular

pathway. These genes are thus probably under strong selective pressure.

8.4 Conclusion and Prospects

We highlighted an unusually high rate of duplication in A. pisum genome. This

finding can give us new insights to test theoretical predictions on the relation between

duplications and evolutionary rates (Ohno 1970). Because cases of positive selection

(Hugues 1994) often occur among gene families, we expected that a large fraction of

the pea aphid genome is thus concerned by patterns of accelerated evolution, which

could favour the emergence of new biological functions and of adaptations.

The comparisons of A. pisum genome and EST-based gene sets from three

other species, even though they constitute partial genomes helped highlighting

two particular gene sets: fast-evolving genes and/or genes that are aphid specific.

The fact that some genes have no hit in non-Aphid databases can reflect a deep

divergence of those genes with their ortholog in other non-aphid species. These

genes could have evolved for specific functions in link to aphid biology.

We have shown that duplications can strongly influence evolution rates of at

least some of the gene copies. We have developed some examples of fast-evolving

genes, some of them being “aphid-specific”. These genes may be under positive or

relaxed selection and could be the result of an adaptative process.

However, our study has been limited by relatively small number of homologous

genes and the exact role of duplication in aphid adaptation remains to be demonstrated

in a larger scale. We will consider, in our future prospects, two main objectives:

1. A fine-scale study of the high level of duplication and of influence of duplica-

tions on evolutionary rates.

2. We will focus on a particular biological feature in aphids: reproduction

polyphenism. Some aphid species are considered as sexual and present, in


their biological cycle, an asexual and a sexual phase, as previously described

(Fig. 8.1). But some species have lost the sexual phase and have become entirely

clonal. Loss of sexuality and of recombination is expected to result in an

accumulation of deleterious mutations and then in the doom of asexual lineages

(Kondrashov 1988). We aim to evaluate the extent by which clonal aphid species

are affected by mutation accumulation and to determine their evolutionary time

of persistence. For this particular project, we have obtained the sequencing of

20,000 ESTs sequences for six new Aphid species, including both taxa that

maintain a sexual reproduction and taxa that are entirely clonal.

Genomic data will then soon be available for more aphid species, including one

complete genome (A. pisum) and partial genomes (ESTs-based data or genomic

data from low-coverage sequencing projects). In such situation, as we start to refine

our knowledge of genomes in the whole aphid group, a relevant strategy is to

determine all possible phylomes. The group of Tonı Gabaldon (“Comparative

Genomics”, CRG Barcelone) has for example developed a pipeline to generate

phylomes from partial or entire genomes of several species (Huerta-Cepas et al.

2008; http://phylomedb.org/). Thanks to collaboration with this group, in Autumn

2009, we have started to generate such phylomes with the extant genomic data for

aphids. This will allow us to retrieve all orthologs available between all species.

Between pair of asexual and sexual species, we will thus be able to compare the

accumulation of non-synonymous mutations in sexual and asexual taxa. We will

also be able to quantify duplication patterns along the different branches of the

aphid species tree. Finally, we will test the correlation between duplication, accel-

eration of evolution and specific aphid biological features.

References

Adams MD, Celniker SE et al (2000) The genome sequence of Drosophila melanogaster. Science287(5461):2185–2195

Aury JM, Jaillon O et al (2006) Global trends of whole-genome duplications revealed by the ciliate

Paramecium tetraurelia. Nature 444(7116):171–178Blanc G, Wolfe KH (2004) Widespread paleopolyploidy in model plant species inferred from age

distributions of duplicate genes. Plant Cell 16(7):1667–1678

Conesa A, Gotz S et al (2005) Blast2GO: a universal tool for annotation, visualization and analysis

in functional genomics research. Bioinformatics 21(18):3674–3676

Gerardo NM, Altincicek B et al (2010) Immunity and defense in pea aphids Acyrthosiphon pisumGenome Biol 11:R21

Hawthorne DJ, Via S (2001) Genetic linkage of ecological specialization and reproductive

isolation in pea aphids. Nature 412(6850):904–907

Heger A, Ponting CP (2007) Evolutionary rate analyses of orthologs and paralogs from 12

Drosophila genomes. Genome Res 17(12):1837–1849

Hirsh AE, Fraser HB (2001) Protein dispensability and rate of evolution. Nature 411(6841):

1046–1049


http://phylomedb.org/

Huerta-Cepas J, Bueno A et al (2008) PhylomeDB: a database for genome-wide collections of

gene phylogenies. Nucleic Acids Res 36:D491–D496

Hufbauer RA, Via S (1999) Evolution of an aphid–parasitoid interaction: variation in resistance

to parasitism among aphid populations specialized on different plants. Evolution 53(5):

1435–1445

Hugues A (1994) The evolution of functionally novel proteins after gene duplication. Proc Biol Sci

256:119–124

IAGC (2010) Genome sequence of the Pea Aphid Acyrthosiphon pisum. Plos Biol doi 10.1311/journal.phio.1000313

Jordan IK, Rogozin IB et al (2002) Essential genes are more evolutionarily conserved than are

nonessential genes in bacteria. Genome Res 12(6):962–968

Kondrashov AS (1988) Deleterious mutations and the evolution of sexual reproduction. Nature

336:435–441

Legeai F, Shigenobu S et al (2010) AphidBase: a centralized bioinformatic resource for annotation

of the pea aphid genome. Insect Mol Biol 19(2):5–12

Mutti NS, Louis J et al (2008) A protein from the salivary glands of the pea aphid, Acyrthosiphonpisum, is essential in feeding on a host plant. Proc Natl Acad Sci USA 105(29):9965–9969

Ohno S (ed) (1970) Evolution by gene duplication. New York, Springer

Ollivier M, Legeai F, Rispe C (2010) Comparative analysis of the Acyrthosiphon pisum genome

and EST-based gene sets from other aphid species. Insect Mol Biol 19(2):33–45

Weinstock GM, Robinson GE et al (2006) Insights into social insects from the genome of the

honeybee Apis mellifera. Nature 443(7114):931–949Yang ZH (1997) PAML: a program package for phylogenetic analysis by maximum likelihood.

Comput Appl Biosci 13(5):555–556

Zdobnov EM, von Mering C et al (2002) Comparative genome and proteome analysis of Anophe-les gambiae and Drosophila melanogaster. Science 298(5591):149–159


Chapter 9

Mammalian Chromosomal Evolution: From

Ancestral States to Evolutionary Regions

Terence J. Robinson and Aurora Ruiz-Herrera

Abstract Chromosome painting by fluorescence in situ hybridization (FISH) has

allowed the detection of regions of orthology in most orders of mammals permitting

the formulation of ancestral mammalian karyotypes at higher taxonomic levels. We

show (1) how the availability of genome sequence data from outgroup species has

facilitated the identification of chromosomes and chromosomal segments that

define eutherian monophyly, and (2) that FISH together with in silico analysis of

genomic sequences point to a nonrandom distribution of evolutionary breakpoints

that are rich in repeat elements and segmental duplications. These regions may

mediate rearrangement by nonallelic homologous recombination between mis-

aligned copies of duplicated regions and lead to breakpoint reuse. Characters that

have arisen convergently (i.e., homoplasy), pose a significant challenge in system-

atics, as does lineage sorting of genetic polymorphisms across successive speciation

nodes (hemiplasy). We show how hemiplasy, a theoretically plausible evolutionary

phenomenon, can materially affect data sets and explore the distinction between

homoplasy and hemiplasy based on persistence times of phylogenetic markers.

T.J. Robinson

Evolutionary Genomics Group, Department of Botany & Zoology, University of Stellenbosch,

Private Bag X1, Matieland 7602, South Africa


A. Ruiz-Herrera

Unitat de Citologia i Histologia, Departament de Biologia Cellular, Fisiologia i Inmunologia,

Universitat Autonoma de Barcelona, Campus Bellaterra, 08193 Barcelona, Spain

Institut de Biotecnologia i Biomedicina, Universitat Autonoma de Barcelona, Campus Bellaterra,

08193 Barcelona, Spain


This manuscript is a synthesis of spoken presentations by: Robinson TJ: Molecular discoveries at

the root of the eutherian tree: Homoplasy, hemiplasy and ancestral states in the phylogenetic

reconstructions of mammalian karyotypes. Ruiz-Herrera A: The genomic puzzle of mammalian

evolutionary breakpoints: can we track any trend?


143

9.1 Introduction

Chromosome reorganization resulting from inversions, translocations, fusions, and

fissions, among other structural changes, contributes to the shuffling of the mamma-

lian genome and thus to the generation of new chromosomal forms on which natural

selection may work. These rearrangements can be caused by the improper repair

of double strand breaks (DSBs) and if the DNA damage occurs in the germ line

and the structural rearrangements are transmissible, the modified chromosome(s)

have the potential to establish in a population through selection and/or stochastic

processes. It is this context that mammalian phylogenomics (the combination of

genomics and phylogenetics that elucidates the phylogenetic relationships among

species by analysis of their entire genomes) has become one of the most integrative

fields in evolutionary biology. A component of this, specifically how chromosomal

rearrangements are involved in speciation and macroevolution, is fundamental for

understanding the dynamics of mammalian chromosomal evolution.

In this overview, we focus on recent developments related to three topical issues

in chromosomal phylogenomics. We report on recent attempts to cladistically

define chromosomal characters that are consistent with eutherian monophyly by

examining the composition of the putative eutherian ancestral karyotype (defined

by cross-species chromosome painting, Ferguson-Smith and Trifonov 2007) and

the genome assemblies of two outgroup species, the opossum (Monodelphis domes-tica) and chicken (Gallus gallus). Second, we summarize evidence supporting a

causal relationship between segmental duplication, repetitive elements, and evolu-

tionary breakpoints at the junction of conserved syntenies, and the propensity for

breakpoint reuse among eutherian species. Finally, we examine the complications

attendant in inferring evolutionary relationships from the cladistic analysis of

chromosomal characters (so called rare genomic changes, Rokas and Holland

2000). We suggest that the critical distinction between homoplasy (convergence

or reversal) and hemiplasy (persistence of the rearrangement across speciation

nodes, Avise and Robinson 2008) may be resolved in instances where divergence

times for nodes are well defined, and the persistence time is less than the divergence

time from a common ancestor.

9.2 Chromosomal Evolution

Nadeau and Taylor (1984) proposed the random-breakage model of chromosomal

evolution. Their thesis, which extended earlier work by Ohno (1973), emphasized

three important points: (1) chromosomal segments are expected to be conserved

among species, (2) that a diploid number of 48 was likely for the common ances-

tor of all mammals, and (3) chromosomal rearrangements are randomly distri-

buted within genomes. Almost 40 years later, and given advances resulting from

144 T.J. Robinson and A. Ruiz-Herrera

molecular cytogenetics, large-scale genome sequencing projects, and new mathe-

matical algorithms, it is interesting to assess how prescient these early observations

were.

9.2.1 Ancestral Karyotypes

Ancestral reconstructions are of interest for different reasons: (1) conserved synte-

nies among species allow the prediction of gene locations based on chromosomal

orthologs (with clear application to species for which genomic data are not avail-

able), (2) ancestral reconstructions provide a framework for estimating rates and

directions of chromosomal change, and (3) mapping karyotypic characters on

evolutionary trees can highlight the importance of chromosomal change in phylo-

genetic reconstructions.

Data derived from cross-species fluorescence in situ hybridization (Zoo–FISH)

are useful for inferring the composition of ancestral karyotypes at various taxo-

nomic and hierarchical levels, i.e., Eutheria (Chowdhary et al. 1998; Richard et al.2003; Yang et al. 2003; Svartman et al. 2004; Ferguson-Smith and Trifonov 2007;

Robinson and Ruiz-Herrera 2008), Boreoeutheria (Froenicke et al. 2006; Robinson

et al. 2006), Rodentia (Graphodatsky et al. 2008), Primates (Stanyon et al. 2008),

Carnivora (Graphodatsky et al. 2002), Cetartiodactyla (Balmus et al. 2007), and

Perissodactyla (Trifonov et al. 2008).

Of the 46 chromosomes in the putative ancestral eutherian karyotype (Fig. 9.1a)

Robinson and Ruiz-Herrera (2008) show that two intact chromosome pairs

(corresponding to human chromosomes 13 and 18) and three conserved chromo-

some segments (10q, 8q, and 19p in the human karyotype) are probably symple-

siomorphic for Eutheria because they are also present as unaltered orthologs in one

or both of the outgroup species (opossum and chicken). Seven additional syntenies

(4q/8p/4pq, 3p/21, 14/15, 10p/12pq/22qt, 19q/16q, 16p/7a, and 12qt/22q), each

involving human chromosomal segments that in combination correspond to intact

chromosomes in the ancestral eutherian karyotype, are also present in one or both

outgroup taxa and thus are probable symplesiomorphies for Eutheria. However,

eight chromosome pairs (corresponding in toto to human chromosomes 1, 5, 6, 9,

11, 17, 20, and the X) and three chromosome segments (2p-q13, 7b, and 2q13-qter)

are derived characters that support the monophyly of eutherian mammals.

There is also considerable recent support for a 2n ¼ 46 chromosome number

in the boreoeutherian ancestor that is not dissimilar to Ohno’s 2n ¼ 48. The

boreoeutherian ancestor originally proposed by Froenicke (2005) is virtually

identical to the eutherian karyotype presented by Ferguson-Smith and Trifonov

(2007) with both benefiting from refinements by Robinson and Ruiz-Herrera

(2008), i.e., the HSA 4q/8p/4pq, HSA2p-q13, HSA10p/12pq/22qt, HSA 19q/

16q, HSA 16p/7a, and HSA 12qt/22q syntenies (see Table 1 in Robinson and

Ruiz-Herrera 2008).

9 Chromosomal Evolution 145

Fig. 9.1 (a) The ancestral eutherian autosomal karyotype based on Ferguson-Smith and Trifonov

(2007) with refinements by Robinson and Ruiz-Herrera (2008). The X chromosome is conserved

across all eutherianmammals and is not included here. (Asterisk) Analysis of reciprocal chromosome

painting data together with genome sequence information indicates that the breakpoint is located in

HSA 3p (see Ruiz-Herrera and Robinson 2007). (b) Schematic representation of orthologous blocks

detected in different mammals that correspond to human chromosome 3. Species included in the

comparison have been studied by reciprocal chromosome painting providing for a rigorous delimi-

tion of the boundaries of synteny. Adapted from Ruiz-Herrera and Robinson (2008)


9.2.2 Evolutionary Breakpoints

In silico analysis led to the formulation of the fragile-breakage model (Bourque

and Pevzner 2002; Pevzner and Tesler 2003; Bourque et al. 2004). Contrary to the

random-breakage model (Sect. 9.2 above), Pevzner and Tesler (2003) showed that

transformation of the mouse gene order to that in human would require consider-

able breakpoint reuse due to the large number of syntenic blocks less than 1 Mb in

size. This suggests that chromosomal rearrangements are not randomly distributed

in the genome, but are concentrated rather in certain regions that can be consid-

ered “hot spots” for recombination – an observation substantiated by chromosome

painting studies (Fig. 9.1b; see also Froenicke 2005; Ruiz-Herrera et al. 2005) that

indicated some genomic regions areas are more prone to breakage and reorgani-

zation than others (Bourque et al. 2004; Murphy et al. 2005; Ruiz-Herrera et al.

2005, 2006; Ma et al. 2006, Kemkemer et al. 2009; Larkin et al. 2009). In a

phylogenetic context, the term “breakpoint reuse” accounts for the recurrence of

the same breakpoint in two different species but, based on comparison with an

outgroup lineage, not in the common ancestor (Murphy et al. 2005; Larkin et al.

2009; Sankoff 2009).

The assumption that some chromosome regions have been reused during the

mammalian chromosomal evolution raises several intriguing questions. (a) Is any

particular DNA configuration or sequence composition driving chromosome evo-

lution?, (b) how are these regions organized in the three-dimensional cell nucleus?,

and (c) by which mechanisms are they regulated in the germ line?

The chromosomal rearrangements that shape mammalian genomes originate as

DSBs. This type of lesion can result from exogenous factors (ionizing radiation

and chemical agents), endogenous agents (free radicals or a stalled replication

fork), or through highly specialized cellular processes that include meiosis and the

recombination of immunoglobulins in the immune system. In all instances,

however, mammalian cells repair DSBs by homologous recombination (HR) or

nonhomologous end joining (NHEJ) (Karran 2000). NHEJ dominates during G1

to the early S phase of the cell cycle, and HR occurs mainly in late S and the G2

phases. Should either mechanisms (HR or NHEJ) fail, DSBs are ineffectively

repaired leading to cell death, or enhanced genomic instability as reflected by

large-scale chromosomal alterations (i.e., deletions, duplications, translocations).

In somatic cells these rearrangements often distinguish neoplasms (see Ruiz-

Herrera and Robinson 2008 and references therein). If these new chromosomal

forms are produced in the germ line, however, they may be coincidental with the

formation of new species.

An interesting aspect to emerge from comparative genomic studies is the finding

that breakpoint regions are rich in repetitive elements. These include tandem

repeats (Puttagunta et al. 2000; Kehrer-Sawatzki et al. 2005), segmental duplica-

tions (SD) (Goidts et al. 2004; Carbone et al. 2006; Bailey and Eichler 2006;

Kehrer-Sawatzki and Cooper 2008), and transposable elements (TEs) (Caceres

et al. 1999; Carbone et al. 2009; Longo et al. 2009), each of which is dealt with


serially below. Additionally, new data suggest that the permissiveness of some

regions of the genome to undergo chromosomal breakage could be determined by

changes in chromatin conformation (Carbone et al. 2009; Lemaitre et al. 2009).

9.2.3 Tandem Repeats

Tandem repeats have been regarded as an important source of DNA variation and

mutation (Armour 2006) having the capacity to form a variety of secondary

structures such as hairpins and bipartite triplexes (Catasti et al. 1999). The instabil-

ity that characterizes tandem repeats is thought to result from slippage during DNA

replication and recombination during meiosis (Usdin and Grabczyk 2000). Expan-

sions of the repeat array occur when an unusual secondary structure is formed in the

lagging daughter strand during DNA replication. Deletions, on the other hand,

occur when an unusual configuration develops in the template for lagging-strand

DNA synthesis (Usdin and Grabczyk 2000). It seems probable that just as tandem

repeats are affected by deletions and expansions in some well-known human

diseases, so too are they implicated in the formation of evolutionary breakpoints.

Some simple tandem repeats have been detected in breakpoint regions, for

instance, the dinucleotide [TA]n (Kehrer-Sawatzki et al. 2005) and [TCTG]n,

[CT]n and [GTCTCT]n (Puttagunta et al. 2000). These early observations led to

further investigations of the possible role of tandem repeat in shaping mammalian

genome architecture (Ruiz-Herrera et al. 2006). The analysis of the distribution

of tandem repeats in human chromosomes by Ruiz-Herrera and colleagues (Ruiz-

Herrera et al. 2006), and their spatial relationship to evolutionary breakpoints

highlights two important points. First, it emphasizes the high concentration of

tandem repeats found at the telomeres and the pericentromeric areas (in agreement

with recent reports on the distribution of duplicated regions by Schueler and

Sullivan 2006 and Riethman 2008). The second is the concentration of tandem

repeats at evolutionary chromosomal bands. Although this is by no means ubiqui-

tous, the correspondence is typified by human chromosomes 3 and 7 (Robinson

et al. 2006; Ruiz-Herrera and Robinson 2008). For example, bands with the greatest

number of tandem repeats in human chromosome 3 (3p25, 3p21.3, 3p12, 3q13.1,

3q21, and 3q29) are also the chromosomal regions that have been implicated in

evolutionary rearrangements (Ruiz-Herrera and Robinson 2008).

9.2.4 Segmental Duplications

SD, or large blocks of genomic sequence (from 1 kb to hundreds of kb) that share

>90% of sequence identity, constitute at least 5% of the human genome (Eichler

2001). They are unevenly distributed along different human chromosomes but


concentrate mainly in the pericentromeric and subtelomeric regions of chromo-

somes (Vallente-Samonte and Eichler 2002).

From an evolutionary perspective, sequence data have identified SD as an

important element in large-scale genome reorganization that underpins evolution-

ary lineages (reviewed in Bailey and Eichler 2006). Nonallelic homologous recom-

bination (NAHR, homologous recombination among paralogous sequences)

mediated by duplicated sequences can, depending on their orientation, result in

deletions, duplications, inversions, and translocations (Bailey and Eichler 2006;

Turner et al. 2008; Marques-Bonet et al. 2009). For example, Armengol et al.

(2003) found an accumulation of SD in rearrangement breakpoints when comparing

the human and mouse whole-genome assemblies; these findings were subsequently

extended to the rat genome (Armengol et al. 2005).

The presence of SDs in evolutionary breakpoint regions has similarly been

shown in primates (Antonell et al. 2005; Nickerson and Nelson 1998; Carbone

et al. 2006; Kehrer-Sawatzki and Cooper 2008). Nine pericentric inversions distin-

guish the human and the chimpanzee karyotypes in addition to the ancestral fusion

of human chromosome 2 (Yunis and Prakash 1982). SDs are located at the break-

points of six of these pericentric inversions, affecting human chromosomes 1, 9, 12,

15, 16, and 18 (Kehrer-Sawatzki and Cooper 2008). The gorilla specific transloca-

tion t(4;19) also appears to be rich in SDs (Stankiewicz et al. 2004).

An analysis of human chromosome 3 typifies how SDs have shaped the evolu-

tionary architecture of mammalian genomes (Ruiz-Herrera and Robinson 2008).

This chromosome contains 2,062 duplicated regions (90% homology and 1 kb

length) accounting for 1.7% (3.3 Mbp) of its length (see http://genome.icsc.edu).

Of these duplicated regions 480 (23.28%) represent 10 kb of continuous sequence,

36 of which occur in 3p25 (7.5%), 160 in 3p12 (33.33%), 173 in 3q21 (36.04%),

and 89 in 3q29 (18.5%) (Ruiz-Herrera and Robinson 2008). The accumulation of

SD in 3q29 is not surprising given that transchromosomal duplications tend to

concentrate in the subtelomeric and pericentromeric areas (Eichler 2001). Of

interest is the fact that three of the four chromosomal bands implicated as evolu-

tionary breakpoints during the eutherian evolution (3p25, 3p12 and 3q21; Fig. 9.1b)

also have the highest concentration of SDs in HSA3. These values contrast sharply

with bands not implicated in evolutionary breakpoints such as 3p14, 3q13.3, and

3q26 (Ruiz-Herrera and Robinson 2008).

9.2.5 Transposable Elements

Other repetitive elements, such as TEs, have been implicated in genomic reorgani-

zation and structural variation by mechanisms that include HR and transposition

(Gray 2000; Ostertag and Kazazian 2001; Feschotte and Prithman 2007; Cordaux

and Batzer 2009). TEs are DNA sequences that are able to move from one locus

to another, often duplicating themselves in the process. They are classified into

two classes according to their sequence structure and mechanism of transposition


http://genome.icsc.edu

(Wicker et al. 2007): Class 1 includes those that transpose through reverse tran-

scription of an RNA intermediate (retrotransposons), and Class 2 refers to DNA

transposons that move through transposition of a DNA intermediate. Retrotranspo-

sons have been the most successful TEs to colonize mammalian genomes – they

make up approximately 40 and 50% of the human and opossum genomes, respec-

tively (Lander et al. 2001; Gentles et al. 2007).

TEs, as with SDs, have the capacity to influence genome plasticity. This can be

done, for example, by (1) the alteration of gene function and regulation, (2) con-

tributing to the creation of new genes, and (3) inducing chromosomal rearrange-

ments (see Feschotte and Prithman 2007 and Cordaux and Batzer 2009 for reviews).

TE-triggered chromosomal rearrangements have been extensively recorded in

plants and animals such as maize and Drosophila (Walker et al. 1995; Caceres

et al. 1999). In the case of primates not all the inversion breakpoints between human

and chimpanzee map to regions of SDs. The breakpoints of the inversions affecting

human chromosome 4, 5, and 17 are rich in Alu elements (Kehrer-Sawatzki and

Cooper 2008). Moreover, a high proportion of Alu elements at the ends of SDs

suggest that they were generated by Alu mispairing, followed by HR in the human

genome (Bailey et al. 2003). There is also evidence in the recent literature for an

accumulation of L1 elements in evolutionary breakpoint regions (Zhao

and Bourque 2009). Longo and collaborators (Longo et al. 2009), for example,

described an accumulation of L1 elements and ERVs (endogenous retroviruses) in

an evolutionary breakpoint in the tammar wallaby genome, a marsupial species.

Gibbons (Family Hylobatidae) represent an interesting case among Hominoidea

(which also include humans and the other great apes, i.e., chimpanzee, gorilla, and

orang-utan), as they are characterized by a strikingly unstable karyotype – this in

sharp contrast to the stability observed for the great apes and most of the more

distantly related primate species (Muller et al. 2003). In a series of elegant studies,

Carbone and co-workers have established a physical map containing most of the

synteny disruptions existing in the white-cheeked gibbon (Nomascus leucogenys)(Carbone et al. 2006, 2009). They isolated most of the synteny-breakpoints in gibbon

BAC clones and subsequently identified them at highest resolution. Their results

revealed an enrichment of active Alu in the gibbon breakpoints, these being less

methylated (CpG-rich) than their orthologous counterparts in the human genome.

The authors hypothesized that this epigenetic state could promote changes into an

open chromatin configuration that, in turn, may be responsible for the higher rate

of chromosomal breakage characterizing the Hylobatidae (Carbone et al. 2009).

9.3 Hemiplasy

During the course of comparing the syntenic blocks in eutherian mammals (see

Sect. 9.2.1 above), we noticed several candidate examples of hemiplasy (two ofwhich

involved chiropterans and afrotherians; Robinson et al. 2008). It was apparent from

these comparisons that a complication with using chromosomal characters to infer


phylogenetic relationships concerns the distinction between characters that have

arisen convergently (i.e., are homoplasic), and those that are due to common ancestry

but which result in homoplasy-like outcomes even though the character states

themselves are genuinely homologous – i.e., are hemiplasic (Avise and Robinson

2008). A likely outcome of the failure to identify hemiplasy (as with homoplasy) is a

misleading phylogenetic interpretation of chromosomal characters, and hence

attempts to disentangle the effects of homoplasy and hemiplasy in a specific phylog-

eny are both useful and conceptually interesting.

9.3.1 Defining Hemiplasy

In brief, hemiplasy can arise where character states a, b, and c represent any type ofgenetic polymorphism (including alternative states of karyotypic features – see

Fig. 9.2a). The more persistent the polymorphic state, the greater the probability of

an eventual discordance between a species tree and a gene tree.

Figure 9.2a illustrates how idiosyncratic lineage sorting can eventuate in gene–

tree/species–tree discordance, and how alternative explanations are possible where

conflicting hypotheses are suggested by different data sets. For example, sequence-

based phylogenies have suggested an association of elephant shrew, tenrec, and

golden mole to the exclusion of aardvark (Amrine-Madsen et al. 2003; Murphy

et al. 2007a) or, alternatively, aardvark, tenrec, and golden mole to the exclusion of

elephant shrews (Waddell and Shelley 2003). In contrast, molecular cytogenetic data

point to a sister relationship between elephant shrew and aardvark to the exclusion of

golden mole (Robinson et al. 2004). This latter association would contradict much

other phylogenetic evidence and we have argued (Robinson et al. 2008) that this

conflict may be explained by the polymorphic state of the 10q/17 and 3/20 syntenies

in an afroinsectiphillian common ancestor that subsequently sorted idiosyncratically

to produce a gene tree/species tree discordance (Fig. 9.2b). Both the 10q/17 and 3/20

syntenies in the aardvark and elephant shrew are caused by centric fusions that must

have arisen in the common ancestor to Afroinsectiphillia prior to the basal divergence

of aardvark �75 mya. They then became independently fixed in the lineage leading

to the elephant shrew (thought to have diverged at 73 mya), but were lost in the

lineage to Afroinsectivora (represented in our analysis only by the golden mole)

subsequent to the divergence of this clade �65 mya meaning also that the character

states themselves would be genuinely homologous and have persisted minimally as

polymorphic states for �2 million years.

9.3.2 Distinguishing Hemiplasy

In an attempt to emphasize the distinction between hemiplasy and homoplasy of

chromosomal characters, consider the tree presented in Fig. 9.2c. This scheme

shows two Robertsonian fusions (A/B and C/E) associated with divergence dates


Fig. 9.2 (a) A schematic representation of how a chromosomal polymorphism that traversed

successive speciation nodes can become fixed in the descendant species in a pattern that appears

discordant with the species phylogeny. Idiosyncratic sorting of a Robertsonian (Rb) fusion

polymorphism (a, b, c) into the descendant taxa would result in lineages that are fixed for the

karyotypic state prior to fusion (i.e., the 2n is unaltered), and those that are homozygous for the

rearrangement (i.e., 2n-2). Note that allele “c” in is a derived character state that it is shared by two

descendant taxa (II and III) that do not constitute a clade at the organismal level. (b) A diagramme

showing how the Robertsonian fusions 10q/17 and 3/20 arose �75 mya in a common ancestor to

Afroinsectiphillia and sorted idiosyncratically (oval) suggesting that these derived chromosomal

syntenies must have persisted for at least �2 million years in order to temporally encompass the

relevant speciation nodes. (c) A hypothetical phylogeny showing the presence of chromosomal

characters A/B, C/D, and C/E in five species (I–V). Two alternative hypotheses can be proposed to

accommodate the distribution of the characters among species (see text)


that vary from 15 to 2 mya for pertinent nodes. It is instructive first to examine the

A/B adjacent synteny. Hemiplasy would require an unlikely persistence time of

13 mya to account for the presence of A/B in distant parts of the species tree (i.e.,

species I and II). The alternative – and more likely explanation – is convergence,

with A/B arising independently in both lineages (homoplasy).

In contrast to this pattern, we argue that chromosomal character C/E most

likely reflects an instance of hemiplasy. As with A/B, two mutually exclusive

hypotheses can be advanced to explain the pattern shown in Fig. 9.2c. First, it

could be argued that the rearrangement (C/E) was present in the common ancestor

of II–V (dated at 4 mya), and its absence in II is due to reversal. Alternatively, it

was fixed in the common ancestor to IV þ V (2 mya), and convergently so in III.

Two “rare genomic changes” would be required in either scenario. Second, and in

contrast to the first hypothesis, hemiplasy would suggest the origin of a single

rearrangement (¼ a single “rare genomic change”) at the common node (4 mya),

followed by incomplete lineage sorting when the ancestral polymorphism is

retained through speciation events, i.e., C/E becomes fixed in the lineages leading

to species III–V and is lost in the lineage to II. The maximum persistence time

required for retention of the chromosomal polymorphism under this scenario is

2 million years. Moreover, this latter explanation most parsimoniously accounts

for the presence of the C/D synteny in species II. This is that the C/E rearrange-

ment was present in a polymorphic state in the common ancestor of II–V (i.e., a

fused C/E and the unfused homologues C and E), a combination that would permit

the independent fusion of C with D. The alternative explanation (the de novo

fission of C/E on the branch leading to species II followed by a fusion of C with D)

being considered less likely.

This scheme emphasizes a critical distinction between hemiplasy and homo-

plasy. This is that hemiplasy is generally more likely for near neutral polymorph-

isms or those that are overdominant. It is also more likely when the internodal

distances in a phylogenetic tree are short (relative to effective population sizes, see

Robinson et al. 2008). On the other hand, homoplasy is less likely to be constrained

by narrow divergence times – the greater the temporal distance, the more likely the

possibility of convergence and reversals of chromosomal rearrangements.

9.4 Conclusions

Contemporary studies of mammalian chromosome evolution are informed by

factors that include data from various sources. First, ancestral karyotypes (and the

critical distinction between symplesiomorphic and synapomorphic characters that

can only be inferred using appropriate outgroups) usually form the comparative

basis for determining the mode and often, the tempo of karyotypic change. This in

turn is reliant on the correct identification of orthologous blocks (either by FISH or

chromosome banding), and is further shaped by knowledge of segmental duplica-

tion, repetitive elements, and breakpoint reuse. In turn these data can have bearing


on the phylogenetic distinction between characters that are convergent/reversals

(i.e., homoplasious), and those that potentially reflect persistence of characters

across species nodes (hemiplasy).

Considerable progress has been made in determining the major features of

mammalian chromosomal evolution. However, recent developments in sequencing

efficiency and expectations of an improvement in annotation technology make it

likely that initiatives such as the recent proposal to target 10,000 vertebrate species

for whole-genome sequencing (Genome 10K Community of Scientists 2009) will

provide a level of resolution and taxonomic scope that is unprecedented for

studying vertebrate and, in particular, mammalian evolutionary relationships. It

can be anticipated that data generated by the G10KCOS initiative will provide

detailed answers on the mechanisms of genomic change, including rearrangements,

duplications, and losses, and definitive insights into the origin of mammalian

karyotypic diversity.

Acknowledgments Financial support to TJR (National Research Foundation, South Africa) and

ARH (Parque Zoologico de Barcelona, Spain) is gratefully acknowledged. Anne Ropiquet is

thanked for discussion on chromosomal phylogenies and Clement Gilbert for comments on an

earlier version of this manuscript.

References

Amrine-Madsen H, Koepfli K-P, Wayne RK, Springer MS (2003) A new phylogenetic marker,

apolipoprotein B, provides compelling evidence for eutherian relationships. Mol Phylogenet

Evol 28:225–240

Antonell A, de Luis O, Domingo-Roura X, Perez-Jurado LA (2005) Evolutionary mechanisms

shaping the genomic structure of the Williams-Beuren syndrome chromosomal region at

human 7q11.23. Genome Res 15:1179–1188

Armengol L, Pujana MA, Cheung J, Scherer SW, Estivill X (2003) Enrichment of segmental

duplications in regions of breaks of synteny between the human and mouse genomes suggest

their involvement in evolutionary rearrangements. Hum Mol Genet 12:2201–2208

Armengol L, Marques-Bonet T, Cheung J, Khaja R, Gonzalez JR, Scherer SW, Navarro A,

Estivill X (2005) Murine segmental duplications are hot spots for chromosome and gene

evolution. Genomics 86:692–700

Armour JA (2006) Tandemly repeated DNA: why should anyone care? Mutat Res 598:6–14

Avise JC, Robinson TJ (2008) Hemiplasy: a new term in the lexicon of phylogenetics. Syst Biol

57:503–507

Bailey JA, Liu G, Eichler EE (2003) An Alu transposition model for the origin and expansion of

human segmental duplications. Am J Hum Genet 73(4):823–834

Bailey JA, Eichler EE (2006) Primate segmental duplications: crucibles of evolution, diversity and

disease. Nat Rev Genet 7:552–564

Balmus G, Trifonov VA, Biltueva LS, O’Brien PC, Alkalaeva ES, Fu B, Skidmore JA, Allen T,

Graphodatsky AS, Yang F, Ferguson-Smith MA (2007) Cross-species chromosome painting

among camel, cattle, pig and human: further insights into the putative Cetartiodactyla ancestral

karyotype. Chromosome Res 15(4):499–515

Bourque G, Pevzner PA (2002) Reconstructing gene orders in the ancestral genomes. Genome Res

12:26–36


Bourque G, Pevzner PA, Tesler G (2004) Reconstructing the genomic architecture of ancestral

mammals: lessons from human, mouse, and rat genomes. Genome Res 14:507–516

Caceres M, Ranz JM, Barbadilla A, Long M, Ruiz A (1999) Generation of a widespread

Drosophila inversion by a transposable element. Science 285:415–418

Carbone L, Vessere GM, ten Hallers BF, Zhu B, Osoegawa K, Mootnick AR, Kofler A,

Wienberg J, Rogers J, Humphray S, Scott C, Harris RA, Milosavljevic A, de Jong P (2006)

A high-resolution map of synteny disruptions in gibbon and human genomes. PLoS Genet

2:223

Carbone L, Harris RA, Vessere GM, Mootnick AR, Humphray S, Rogers J, Kim SK, Wall JD,

Martin D, Jurka J, Milosavljevic A, de Jong PJ (2009) Evolutionary breakpoints in the gibbon

suggest association between cytosine methylation and karyotype evolution. PLoS Genet

5:e1000538

Catasti P, Chen X, Mariappan SVS, Bradbury EM, Gupta G (1999) DNA repeats in the human

genome. Genetica 106:15–36

Chowdhary BP, Raudsepp T, Froenicke L, Scherthan H (1998) Emerging patterns of comparative

genome organization in some mammalian species as revealed by Zoo-FISH. Genome Res

8:577–589

Cordaux R, Batzer MA (2009) The impact of retrotransposons on human genome evolution. Nat

Rev Genet 10:691–703

Eichler EE (2001) Recent duplication, domain accretion and the dynamic mutation of the human

genome. Trends Genet 17:661–669

Ferguson-Smith MA, Trifonov V (2007) Mammalian karyotype evolution. Nat Rev Genet

8:950–962

Feschotte C, Prithman EJ (2007) DNA transposons and the evolution of the eukaryotic genomes.

Annu Rev Genet 41:331–368

Froenicke L (2005) Origins of primate chromosomes – as delineated by Zoo-FISH and alignments

of human and mouse draft genome sequences. Cytogenet Genome Res 108:122–138

Froenicke F, Wienberg J, Stone G, Adams L, Stanyon R (2003) Towards the delineation of the

ancestral eutherian genome organization: comparative genome maps of human and the African

elephant (Loxodonta africana) generated by chromosome painting. Proc R Soc Lond B Biol

Sci 270:1331–1340

Froenicke L, Caldes MG, Graphodatsky A, M€uller S, Lyons LA, Robinson TJ, Volleth M, Yang F,

Wienberg J (2006) Are molecular cytogenetics and bioinformatics suggesting contradictory

models of ancestral mammalian genomes? Genome Res 16:306–310

Genome 10K Community of Scientists (2009) Genome 10K: a proposal to obtain whole-genome

sequence for 10,000 vertebrate species. J Hered 100:659–674

Gentles AJ, Wakefield MJ, Kohany O, Gu W, Batzer MA, Pollock DD, Jurka J (2007) Evolution-

ary dynamics of transposable elements in the short-tailed opossum Monodelphis domestica.

Genome Res 17:992–1004

Goidts V, Szamalek JM, Hameister H, Kehrer-Sawatzki H (2004) Segmental duplication asso-

ciated with the human-specific inversion of chromosome 18: a further example of the impact of

segmental duplications on karyotype and genome evolution in primates. Hum Genet

117:168–176

Graphodatsky AS, Yang F, Perelman PL, O’Brien PC, Serdukova NA, Milne BS, Biltueva LS,

Fu B, Vorobieva NV, Kawada SI, Robinson TJ, Ferguson-Smith MA (2002) Comparative

molecular cytogenetic studies in the order Carnivora: mapping chromosomal rearrangements

onto the phylogenetic tree. Cytogenet Genome Res 96:137–145

Graphodatsky AS, Yang F, Dobigny G, Romanenko SA, Biltueva LS, Perelman PL, Beklemisheva

VR, Alkalaeva EZ, Serdukova NA, Ferguson-Smith MA, Murphy WJ, Robinson TJ (2008)

Tracking the evolution of genome organization in rodents by ZOO-FISH. Chromosome Res

16:261–274

Gray YH (2000) It takes two transposons to tango: transposable-element-mediated chromosomal

rearrangements. Trends Genet 16:461–468


Karran P (2000) DNA double strand break repair in mammalian cells. Curr Opin Genet Dev

10:144–150

Kehrer-Sawatzki H, Cooper DN (2008) Molecular mechanisms of chromosomal rearrangement

during primate evolution. Chromosome Res 16:41–56

Kehrer-Sawatzki H, Szamalek JM, Tanzer S, Platzer M, Hameister H (2005) Molecular character-

ization of the pericentric inversion of chimpanzee chromosome 11 homologous to human

chromosome 9. Genomics 85:542–550

Kemkemer C, Kohn M, Cooper DN, Froenicke L, Hogel J, Hameister H, Kehrer-Sawatzki H

(2009) Gene synteny comparisons between different vertebrates provide new insights

into breakage and fusion events during mammalian karyotype evolution. BMC Evol Biol

9:84

Korstanje R, O’Brien PCM, Yang F, Rens W, Bosma AA, van Lith HA, van Zutphen LF,

Ferguson-Smith MA (1999) Complete homology maps of the rabbit (Oryctolagus cuniculus)and human by reciprocal chromosome painting. Cytogenet Cell Genet 86:317–322

Lander ES and the Int Human Genome Sequencing Consortium (2001) Initial sequencing and

analysis of the human genome. Nature 409:860–921

Larkin DM, Pape G, Donthu R, Auvil L, Welge M, Lewin HA (2009) Breakpoint regions and

homologous synteny blocks in chromosomes have different evolutionary histories. Genome

Res 19:770–777

Lemaitre C, Zaghloul L, Sagot MF, Gautier C, Arneodo A, Tannier E, Audit B (2009) Analysis of

fine-scale mammalian evolutionary breakpoints provides new insight into their relation to

genome organisation. BMC Genomics 10:335

Li T, O’Brien PCM, Biltueva L, Fu B, Wang J, Nie W, Ferguson-Smith MA, Graphodatsky AS,

Yang F (2004) Evolution of genome organizations of squirrels (Sciuridae) revealed by cross-

species chromosome painting. Chromosome Res 12:317–335

Longo MS, Carone DM, NISC Comparative Sequencing Program, Green ED, O’Neill MJ, O’Neill

RJ (2009) Distinct retroelement classes define evolutionary breakpoints demarcating sites of

evolutionary novelty. BMC Genomics 10:334

Ma J, Zhang L, Suh BB, Raney BJ, Burhans RC, Kent WJ, Blanchette M, Haussler D, Miller W

(2006) Reconstructing contiguous regions of an ancestral genome. Genome Res 16:1557–1565

Marques-Bonet T, Girirajan S, Eichler EE (2009) The origins and impact of primate segmental

duplications. Trends Genet 25:443–545

Muller S, Stanyon R, O’Brien PCM, Ferguson-Smith MA, Plesker R, Wienberg J (1999) Defining

the ancestral karyotype of all primates by multidirectional chromosome painting between tree

shrews, lemurs and humans. Chromosoma 108:393–400

Muller S, Hollatz M, Wienberg J (2003) Chromosomal phylogeny and evolution of gibbons

(Hylobatidae). Hum Genet 113:493–501

Murphy WJ, Larkin DM, Everts-van-der Wind A, Bourque G, Tesler G, Auvil L, Beever JE,

Chowdhary BP, Galibert F, Gatzke L, Hitte C, Meyers SN, Milan D, Ostrander EA, Pape G,

Parker HG, Raudsepp T, Rogatcheva MB, Schook LB, Skow LC, Welge M, Womack JE,

O’brien SJ, Pevzner PA, Lewin HA (2005) Dynamics of mammalian chromosome evolution

inferred from multispecies comparative maps. Science 309:613–617

Murphy WJ, Pringle TH, Crider TA, Springer MS, Miller W (2007a) Using genomic data to

unravel the root of the placental mammal phylogeny. Genome Res 17:413–421

Murphy WJ, Davis B, David VA, Agarwala R, Schaffer AA, Pearks Wilkerson AJ, Neelam B,

O’Brien SJ, Menotti-Raymond M (2007b) A 1.5-Mb-resolution radiation hybrid map of the

cat genome and comparative analysis with the canine and human genomes. Genomics

89:189–196

Nadeau JH, Taylor BA (1984) Lengths of chromosomal segments conserved since divergence of

man and mouse. Proc Natl Acad Sci USA 81:814–818

Nickerson E, Nelson DL (1998) Molecular definition of pericentric inversion breakpoints occur-

ring during the evolution of humans and chimpanzees. Genomics 50:368–372

Ohno S (1973) Ancient linkage groups and frozen accidents. Nature 244:259–262


Ostertag EM, Kazazian HH (2001) Twin priming: a proposed mechanism for the creation of

inversions in L1 retrotransposition. Genome Res 11:2059–2065

Perelman PL, Graphodatsky AS, Serdukova NA, Nie W, Alkalaeva EZ, Fu B, Robinson TJ,

Yang F (2005) Karyotypic conservatism in the suborder Feliformia (Order Carnivora). Cyto-

genet Genome Res 108:348–354

Pevzner P, Tesler G (2003) Human and mouse genomic sequences reveal extensive breakpoint

reuse in mammalian evolution. Proc Natl Acad Sci USA 100:7672–7677

Puttagunta R, Gordon LA, Meyer GE, Kapfhamer D, Lamerdin JE, Kantheti P, Portman KM,

Chung WK, Jenne DE, Olsen AS, Burmeister M (2000) Comparative maps of human 19p13.3

and mouse chromosome 10 allow identification of sequences at evolutionary breakpoints.

Genome Res 10:1369–1380

Richard F, LombardM, Dutrillaux B (2003) Reconstruction of the ancestral karyotype of eutherian

mammals. Chromosome Res 11:605–618

Riethman H (2008) Human telomere structure and biology. Annu Rev Genomics Hum Genet

9:1–19

Robinson TJ, Ruiz-Herrera A (2008) Defining the ancestral eutherian karyotype: a cladistic

interpretation of chromosome painting and genome sequence assembly data. Chromosome

Res 16:1133–1141

Robinson TJ, Fu B, Ferguson-Smith MA, Yang F (2004) Cross-species chromosome painting in

the golden mole and elephant shrew: support for the mammalian clades Afrotheria and

Afroinsectiphillia but not Afroinsectivora. Proc Biol Sci 271:1477–1484

Robinson TJ, Ruiz-Herrera A, Froenicke L (2006) Dissecting the mammalian genome – new

insights into chromosomal evolution. Trends Genet 22:297–301

Robinson TJ, Ruiz-Herrera A, Avise JC (2008) Hemiplasy and homoplasy in the karyotypic

phylogenies of mammals. Proc Natl Acad Sci USA 105:14477–14481

Rokas A, Holland PW (2000) Rare genomic changes as a tool for phylogenetics. Trends Ecol Evol

15:454–459

Ruiz-Herrera A, Robinson TJ (2008) Evolutionary plasticity breakpoints in human chromosome 3.

BioEssays 30:1126–1137

Ruiz-Herrera A, Garcia F, Mora L, Egozcue J, Ponsa M, Garcia M (2005) Evolutionary conserved

chromosomal segments in the human karyotype are bounded by unstable chromosome bands.

Cytogenet Genome Res 108:161–174

Ruiz-Herrera A, Castresana J, Robinson TJ (2006) Is mammalian chromosomal evolution driven

by regions of genome fragility? Genome Biol 7:R115

Ruiz-Herrera A, Robinson TJ (2007) Chromosomal instability in Afrotheria: fragile sites, evolu-

tionary breakpoints and phylogenetic inference from genome sequence assemblies. BMC Evol

Biol 7:199

Sankoff D (2009) The where and wherefore of evolutionary breakpoints. J Biol 8:66

Schueler MG, Sullivan BA (2006) Structural and functional dynamics of human centromeric

chromatin. Annu Rev Genomics Hum Genet 7:301–313

Stankiewicz P, Shaw CJ, Withers M, Inoue K, Lupski JR (2004) Serial segmental duplications

during primate evolution result in complex human genome architecture. Genome Res

14:2209–2220

Stanyon R, Rocchi M, Capozzi R, Roberto R, Misceo D, Ventura M, Cardone MF, Bigoni F,

Archidiacono N (2008) Primate chromosome evolution: ancestral karyotypes, marker order

and neocentromeres. Chromosome Res 16:17–39

Svartman M, Stone G, Page JE, Stanyon R (2004) A chromosome painting test of the basal

eutherian karyotype. Chromosome Res 12:45–53

Trifonov VA, Stanyon R, Nesterenko AI, Fu B, Perelman PL, O’Brien PC, Stone G, Rubtsova NV,

Houck ML, Robinson TJ, Ferguson-Smith MA, Dobigny G, Graphodatsky AS, Yang F (2008)

Multidirectional cross-species painting illuminates the history of karyotypic evolution in

Perissodactyla. Chromosome Res 16:89–107


Turner DJ, Miretti M, Rajan D, Fiegier H, Carter NP, Blayney ML, Beck S, Hurles ME (2008)

Germline rates of the novo meiotic deletions and duplications causing several genomic

disorders. Nat Genet 40:90–95

Usdin K, Grabczyk E (2000) DNA repeat expansions and human disease. Cell Mol Life Sci

57:914–931

Vallente-Samonte R, Eichler EE (2002) Segmental duplications and the evolution of the primate

genome. Nat Rev Genet 3:65–72

Waddell PJ, Shelley S (2003) Evaluating placental inter-ordinal phylogenies with novel sequences

including RAG1, gamma-fibrinogen, ND6, and mt-tRNA, plus MCMC-driven nucleotide,

amino acid, and codon models. Mol Phylogenet Evol 28:197–224

Walker EL, Robbins TP, Bureau TE, Kermicle J, Dellaporta SL (1995) Transposon-mediated

chromosomal rearrangements and gene duplications in the formation of the maize R-r

complex. EMBO J 14:2350–2363

Wicker T, Sabot F, Hua-Van A, Bennetzen JL, Capy P, Chalhoub B, Flavell A, Leroy P, Morgante

M, Panaud O, Paux E, SanMiguel P, Schulman AH (2007) A unified classification system for

eukaryotic transposable elements. Nature Rev Genet 8:973–982

Yang F, Alkalaeva EZ, Perelman PL, Pardini AT, Harrison WR, O’Brien PC, Fu B, Graphodatsky

AS, Ferguson-Smith MA, Robinson TJ (2003) Reciprocal chromosome painting among

human, aardvark, and elephant (superorder Afrotheria) reveals the likely eutherian ancestral

karyotype. Proc Natl Acad Sci USA 100:1062–1066

Yang F, Fu B, O’Brien PCM, Nie W, Ryder OA, Ferguson-Smith MA (2004) Refined genome-

wide comparative map of the domestic horse, donkey and human based on cross-species

chromosome painting: insight into the occasional fertility of mules. Chromosome Res

12:65–76

Yunis JJ, Prakash O (1982) The origin of man: a chromosomal pictorial legacy. Science

215:1525–1530

Zhao H, Bourque G (2009) Recovering genome rearrangements in the mammalian phylogeny.



Chapter 10

Mechanisms and Evolution of Dorsal–Ventral

Patterning

Claudia Mieko Mizutani and Rui Sousa-Neves

Abstract In the last two decades, a great progress has been made with the dis-

covery and understanding of conserved signaling pathways, in particular those

involved in embryonic dorsal–ventral patterning and the organization of the ner-

vous system. Remarkably, the spatial distribution of these signal molecules appears

conserved across a large group of animals that have centralized nervous systems.

Despite these achievements, there are still many unanswered questions on how the

nervous system organization evolves and responds to variations in organism size.

In this review, we discuss the progression of the field from early observations made

more than a century ago and introduce future challenges regarding the problem of

scaling of the nervous system during evolution.

10.1 Introduction

Animal development can lead to diverse life forms from a relatively limited number

of genes. A great progress to our understanding of the mechanisms of development

has been made using model organisms suitable to genetic and molecular analyses.

These model organisms are likely to continue uncovering mechanisms relevant to a

wide variety of species and of significance for human health. One example is the

conservation of the molecular components employed to differentiate neural tissues

C.M. Mizutani

Department of Biology, Case Western Reserve University, 10900 Euclid Ave, Cleveland, OH

447080, USA

Department of Genetics, Case Western Reserve University, 10900 Euclid Ave, Cleveland, OH

447080, USA


R. Sousa-Neves

Department of Biology, Case Western Reserve University, 10900 Euclid Ave, Cleveland, OH

447080, USA



159

from epidermis and the subsequent subdivision of the nervous system into discrete

regions of gene expression. Most recently, the sequencing of several genomes and

technological advances made in the past decade brought previously intractable

organisms to scrutiny. These advances also opened the possibility to tackle ques-

tions that could not have been answered before and deserve attention. One of them

is how do organisms change over time? Another is how the body plan and organs

can be rescaled across species? Answers to these questions are essential to our

understanding of the evolution of novel body plans.

Broadly, two general mechanisms have been proposed to explain the generation

of different body plans and tissues, which in principle should apply to the dorsal–

ventral (D/V) axis formation. The first proposes that the evolution of cis-regulatory

sequences that control gene expression plays a significant role in body plan

diversity, while the second implicates the evolution of coding sequences in key

patterning genes. The first possibility has been tested by transferring previously

isolated cis-regulatory sequences from one organism to another, and assaying the

expression patterns generated by means of a reporter gene. In many cases, the

patterns of expression observed largely resemble that of the host (Kassis 1990;

Ludwig et al. 1998; Crocker et al. 2008; Liberman and Stathopoulos 2009). That is,

despite extensive modifications in regulatory sequences, the final expression pattern

resembles that of the species that implements the information rather than the donor

of these regulatory sequences. In other cases documented so far, we observe the

inverse: the patterns generated resemble those of the donor (Wittkopp et al. 2002;

Gompel et al. 2005; Crocker et al. 2008).

In addition to mutations in cis-regulatory sequences, there is also evidence that

changes in coding sequences lead to different developmental programs. One exam-

ple is the case of hybrid lethal systems, which provides an effective way of making

the development of two similar species incompatible. Such complementary lethal

genes are innocuous when present in individuals of a single species, but cause

lethality and/or sterility when combined in a hybrid between different species

(Sturtevant 1929; Yamamoto et al. 1997; Brideau et al. 2006). The molecular

identification of hybrid lethal genes isolated so far reveals that differences in the

coding sequences are responsible for the developmental incompatibilities observed.

Thus, both changes in regulatory sequences, as well as changes in coding sequences,

can lead to the generation of developmentally distinct processes and consequently

novel life forms. In addition, these results also highlight that gene networks, rather

than individual genes, are coevolving to adapt to mutations in both coding and

noncoding sequences.

In this review, we discuss the early molecular events that contribute to germ

layer specification, with an emphasis on the establishment of D/V morphogenetic

gradients that regulate patterns of neural gene expression in Drosophila. Thisproblem traces back to the nineteenth century, and recent investigation led to the

identification of key molecular players and a unifying view of neural development.

We also discuss the problem of morphogenetic scaling across species and possible

mechanisms that could explain how patterns of gene expression are reshaped in

response to size changes.

160 C.M. Mizutani and R. Sousa-Neves

10.1.1 The Unity of Plan Hypothesis and Body Axis Inversion

From humans to small bees and worms, animals exhibit complex behaviors and

social organizations generated by nervous systems of great complexity. Three

questions stand out when we observe these complex structures: (1) to what extent

different nervous systems share a similar and conserved molecular architecture; (2)

how and when did this organization arise; and (3) how do these structures evolve

and become more complex? Over the past 20 years, key findings from the field

of developmental biology have provided answers to some of these questions,

unlocking clues on the origins and evolution of the nervous system.

The advent of developmental biology as a field combining anatomy, embryology

genetics and molecular biology brought together two important discoveries sepa-

rated by a large number of years. The first one was an observation made by the

French anatomist Ethienne Goeffroy Saint-Hilaire in 1822, a proponent of the

“unity of plan” hypothesis (Geoffroy St. Hilaire 1822). Based on the anatomy of

a lobster to that of vertebrates, he suggested that invertebrates and vertebrates

shared the same elements of body construction, which could be explained by an

inversion of the embryonic D/V axis that caused the ventral position of the

invertebrate nervous system vs. a dorsal position in vertebrates. The second discov-

ery was the classical neural induction transplantation experiment carried out by

Spemann and Mangold, a century later in 1924 (Spemann and Mangold 1924).

Their experiment led to the identification of the Spemann organizer, a region of the

embryo capable of inducing surrounding cells to differentiate as neural tissue. What

are the signals released by the Organizer that result in neural induction and could

the D/V inversion be confirmed at the molecular level? Several decades had to

elapse before the answers to these questions were obtained and the final outcome of

those efforts was quite remarkable.

At the center of the mechanism of neural induction was the discovery of a gene

cassette that function antagonistically: the invertebrate genes short gastrulation(sog) and decapentaplegic (dpp) and their vertebrate counterparts BMP-4 and

Chordin (Chd). Genetic manipulations of these genes revealed that dpp/BMP-4

encodes a secreted protein belonging to the TGF-b family of transforming growth

factors (Padgett et al. 1987), which has a dual function; it signals to cells to promote

epidermal specification (Irish and Gelbart 1987; Wharton et al. 1993) and at the

same time, it blocks neural development. In vertebrates, Chd is secreted by the

Spemann organizer and it promotes neural development by blocking the BMP-4

anti-neural signal (Sasai et al. 1994). Similarly in flies, Sog is an antagonist of Dpp

and also protects the future site of neuroectoderm by binding to Dpp and preventing

it to activate its receptors (Francois et al. 1994; Biehs et al. 1996). Thus, neural

induction is achieved by a double-negative mechanism whereby neural develop-

ment is a result of repression of a repressive signal. The exciting side of this

research was that not only these long sought morphogens were finally isolated

and provided a mechanistic basis for neural induction, but also they were shown to

be completely interchangeable between vertebrates and invertebrates, and finally,

10 Mechanisms and Evolution of Dorsal–Ventral Patterning 161

their opposite expression patterns along the D/V axis were also shown to be upside-

down in these organisms (Padgett et al. 1993; Francois et al. 1994; Schmidt et al.

1995; Holley et al. 1995). That is, sog is expressed ventrally in invertebrates, whileChd is expressed dorsally, and in both cases, their expression domains demarcate

the future site of nervous system development. Together, these facts highlighted the

preceding ideas of axis inversion set forth by Saint-Hilaire and were suggestive of a

common ancestry among vertebrates and invertebrates (Arendt and Nubler-Jung

1994; De Robertis and Sasai 1996; Ferguson 1996; Bier 1997).

10.1.2 Dorsal, a Gene at Odds with the EvolutionaryConservation of D/V Patterning

Before the discovery of neural inducers, a series of studies in Drosophila demon-

strated that the early embryo is initially patterned by a ventral-to-dorsal gradient of

another protein called Dorsal, an NFk-B-related transcription factor. The Dorsal

nuclear gradient is established via a complex proteolytic cascade of exclusively

maternal information that culminates with a regulated transport of Dorsal into the

nucleus, resulting in a nuclear concentration gradient with high levels of Dorsal in

ventral most nuclei, moderate levels in lateral nuclei, and very low or absent levels

in dorsal nuclei (Roth et al. 1989; Rushlow et al. 1989; Steward 1989). Once inside

the nucleus, Dorsal can activate or repress the expression of several zygotic target

genes that implement the differentiation of the three primary germ layers of the

embryo (Ray et al. 1991; Stathopoulos et al. 2002). High levels of Dorsal activate

mesodermal genes (e.g., snail and twist) in the ventral side of the embryo, while

moderate levels activate neureoctodermal genes (e.g., sog). In contrast, Dorsal

represses ectodermal genes (e.g., dpp), and as a consequence these genes have

their expression restricted to the dorsal region of the embryo, where there are low or

undetectable levels of Dorsal (Fig. 10.1).

Even though the Dorsal gradient is crucial for defining the three primary germ

layers in Drosophila, this does not seem to represent the ancestral role of the

Dorsal/NFkB signaling pathway. Rather, Dorsal/NFkB pathway is involved in

immune response in both vertebrates and invertebrates (reviewed by Ferrandon

et al. 2007), whereas the recruitment of this signaling pathway in D/V patterning is

likely to be an innovation found in some invertebrates. One can also speculate that

the innovative role of the Dorsal gradient in D/V patterning is under a rapid process

of evolution. Recent work in divergent insect groups indicates that the mechanisms

controlling the formation of the Dorsal gradient is highly variable within insects,

possibly reflecting adaptations of this gradient to short germ band (e.g., tribolium)

vs. long germ band (e.g., flies) modes of development (Chen et al. 2000; Nunes da

Fonseca et al. 2008).

A number of studies indicate that the Dorsal gradient influences the further

subdivision of the Drosophila neuroectoderm into restricted domains of


neuroectodermal gene expression, as discussed in the next section. This observation

stands in contrast to the subdivision of the neural tube in vertebrates, which

employs the morphogens BMP and Sonic Hedgehog (Shh) (Liem et al. 1995,

2000; Briscoe et al. 1999; Litingtung and Chiang 2000). Those differences led to

view that the D/V patterning of the nervous system of Drosophila and vertebrates

have arisen from completely different molecular mechanisms and may have

evolved by convergent evolution.

10.1.3 From Saint Hilaire and Spemann Toward a UnifyingMechanism for Neural Organization

Recently, research on nervous system origins has sparked another round of interest.

First, further analyses of the patterning of nervous system into organized D/V

domains of gene expression became available, along with studies of upstream

Fig. 10.1 Formation of dorsal–ventral gradients in the Drosophila embryo. (a) Scheme of an early

Drosophila embryo, in lateral view (anterior to the right). The embryo develops as a syncitium

blastoderm. Nuclei divide and migrate to the periphery of the embryo, where cellularization takes

place. (b) Dorsal–ventral gradients emanating from ventral and dorsal regions subdivide the

embryo into three primary domains that give rise to the mesoderm (MES), neuroectoderm (NE),

and ectoderm (ECT). (c and d) Cross-section view of embryo. (c) Representation of the Dpp and

Dorsal gradients. Small blue dots represent Dpp molecules that form a dorsal-to-ventral gradient in

the extracellular domain. The nuclear Dorsal gradient is represented by red colored nuclei. (d)

Expression domains of dorsal–ventral genes that elicit the differentiation of mesoderm, neuroec-

toderm, and ectoderm


signaling events that generate this patterning (Mizutani et al. 2006; Mizutani and

Bier 2008). Second, experiments carried out in organisms that belong to other

phylogenetic branches, such as hemichordates, annelids, cnidarians, and sea

anemone, have served as outgroups for valuable comparative studies of nervous

system and axis formation evolution (Samuel et al. 2001; Rentzsch et al. 2006;

Lowe et al. 2006; Denes et al. 2007; Lapraz et al. 2009; Nomaksteinsky et al. 2009;

Saina et al. 2009).

The idea that the nervous system patterning predated the split between verte-

brates and invertebrates implies that a centralization and organization of the

nervous system must have originated a long time ago, an estimated time of

500–600 million years. Supporting this view, the BMP/dpp signaling pathway has

clearly emerged as a conserved pathway in all bilaterian organisms studied so far,

and in most cases it has been shown to be involved in not only nervous system

centralization, but also in its patterning (De Robertis 2008; Mizutani and Bier

2008). In the next section, we focus on the D/V subdivision of the nervous system

of Drosophila, and subsequently we discuss how morphogenetic gradients involved

in the overall D/V patterning of the primary germ layers may evolve in closely

Drosophila species. More detailed discussions on the evolution of the nervous

system have been reviewed elsewhere (Lowe et al. 2006; Mizutani and Bier

2008; Arendt et al. 2008; Holland 2009).

10.2 Neural Patterning and Specification of Neuroblasts

in Drosophila

Due to its simplicity, the ventral nervous cord of insects has served as a paradigm to

the study of differentiation of neuroblasts or neural stem cells. At early embryonic

stages, once the neural and nonneural ectodermal domains are established by the

activity of BMP/dpp and Chd/sog, the neural domain is further subdivided into

expression domains of key transcription factors that confer a unique identity to each

of the 30 neuroblasts per hemisegment that delaminate from the neuroectoderm

(reviewed in Bhat 1999; Technau et al. 2006). Each Neuroblast is committed to

generate a stereotyped neural cell lineage (Doe and Skeath 1996; Doe 1992, 2008)

after receiving “positional information” from both D/V and anterior–posterior

(A/P) expressing genes (Fig. 10.2). Information provided from the D/V axis dictates

the formation of main neural cell types, such as motorneurons, serotonergic, and

sensory neurons (Schmid et al. 1999). In the Drosophila embryo, the neural identity

genes responsible for D/V patterning are ventral nervous system defective (vnd),intermediate neuroblasts defective (ind), and muscle segment homeobox/Drop(msh/Dr), which are expressed in nonoverlapping domains. vnd is expressed in

the ventral most layer of the neuroectoderm, while ind is expressed in the interme-

diate region, and finally msh is expressed in the dorsal most region (Jimenez et al.

1995; Isshiki et al. 1997; McDonald et al. 1998; Weiss et al. 1998; Mellerick and

Modica 2002) (Fig. 10.1d).


The study of nervous system patterning along the D/V axis provided additional

evidence for the common ancestry of the nervous system, since the vertebrate

homologues for vnd, ind, and msh (Nkx2.2., Gsh, and Msx) are also expressed in

the same arrangement along the D/V axis of the neural tube after it is inverted

(Valerius et al. 1995; Wang et al. 1996; Suzuki et al. 1997; Weiss et al. 1998;

Briscoe et al. 1999; Liu et al. 2004; Kriks et al. 2005). It has been shown that the

BMP signaling pathway is responsible for repressing the expression of neural

identity genes in a dosage-dependent fashion by reaching the adjacent neural

domain, such that ventrally expressing genes are more sensitive to its repression

than dorsally expressing genes are. As a result, the domains of vnd/NKx2.2. andind/Gsh are pushed away from the dorsal source of BMP secretion, while msh/Msx

domain is placed more dorsally since this gene can tolerate high levels of BMPs

before being repressed. In addition to this differential sensitivity to BMP levels,

there is also a cross-regulatory interaction among those neural identity genes that

cooperate in this patterning. Namely, they can repress each other in the ventral-

to-dorsal direction, an interaction referred to as “ventral dominance” (Cowden and

Levine 2003). Thus, vnd represses ind, while both vnd and ind repress msh expres-

sion. This same relationship also appears to be at least partially conserved in

vertebrates (Mizutani et al. 2006; Illes et al. 2009). It is noteworthy that even

though the specification of D/V neural cell types in the nervous system is also

dependent on other morphogens in vertebrates and invertebrates (i.e., Shh and

Dorsal, respectively), the BMP signaling can provide most of the information for

neural patterning in the absence of these additional cues (Jacob and Briscoe 2003;

Mizutani et al. 2006).

The findings above reconcile discrepancies found in some vertebrate and inver-

tebrate lineages of noncentralized nervous systems, which more likely represent

highly derived forms, and establish a common unifying mechanism that patterns the

Fig. 10.2 Neuroblast formation and neural determination in Drosophila. (a) Blastoderm stage

embryo. Germ layers are indicated, as well as the D/V neuroectodermal domains. (b) Mesodermal

cells invaginate, bringing the two halves of the neuroectoderm together at the ventral midline.

(c) Delamination of neuroblasts from respective neuroectodermal domains. (d) Ventral view of

neuroblast map, roughly representing the 30 neuroblasts per hemisegment. (e) Neuron types

formed along the D/V axis. Serotonergic neurons are formed in ventral region, sensory neurons

in lateral regions, and motoneurons in all three domains. Colors of neuroblasts in (d) and neurons

in (e) indicate their ventral (blue), lateral (green), and dorsal (red) identities


nervous system. Further evidence of the ancestral role of BMP signaling in neural

patterning was substantiated by studies carried out in an outgroup organism, the

marine annelid Platynereis dumerilii, which belongs to the second major inverte-

brate branch of lophotrochozoa (Denes et al. 2007).

Thus, the nervous system evolution seems highly conservative and is likely to

have relied on the ancestral BMP signaling pathway to generate a similar architec-

ture of neural cell types arranged along the D/V axis for millions of years. The

picture that emerges from these studies also suggest that this ancestral signaling

cassette can be superimposed to other graded morphogenetic signals, such as Dorsal

in the case of Drosophila, and Shh in vertebrates. Remarkably, in the case of

insects, the whole system must still be able to maintain the layers of gene expres-

sion of vnd, ind, andmshwith similar number of cells. This view is supported by the

highly stereotyped neuroblast maps between divergent insects such as grasshopper,

Drosophila, and silverfish, and even more distant arthropods such as crustaceans

(Thomas et al. 1984; Doe 1992; Whitington 1996; Ungerer and Scholtz 2008).

Genetic experimentation in Drosophila has shown that alterations in the width of

expression domains of vnd, ind, and msh can lead to profound alterations of loss or

duplication of specific neuron cell types (Fig. 10.3). Even though some partial

modifications in the patterns of expression of those neural identity gene may exist,

Loss of ventral neuronsDuplication of RP2

Loss of RP2

Early NE domains

b c d

awt vnd- ind-

Late stage neurons (ventral and intermediate)

v iRP2

Fig. 10.3 Alterations in width of neuroectodermal domains lead to loss or duplications of specific

neurons. (a) Early neuroectodermal expression domains in wt and in vnd and indmutants. The vnddomain is represented in green, ind in blue, and msh in red. Position of ventral midline is indicated

by arrowhead. In vnd mutant, ind expression domain is expanded, while in ind mutant, both vndand ind are expanded. (b–d) Late stage embryos stained for even-skipped, which recognizes

neurons of ventral and intermediate fate (v and i). (b) Wild type. (c) vnd mutant displaying loss

of ventral neurons and duplication of RP2 motorneuron, an intermediate neuron. (d) ind mutant

with loss of RP2 neurons (red arrows). [(b and c) pictures were reproduced from McDonald et al.

1998. Picture in (d) was reproduced from Weiss et al 1998]


as it has been reported in the bettle tribolium (Wheeler et al. 2005), in general there

seems to be a strong pressure to maintain a conserved organization of neuroblast

number and types. Therefore, it would not be surprising if there were a robust

mechanism that assures that the same number of cells is maintained in the nervous

system of insects despite their differences in embryo size.

10.3 Scaling of Germ Layers During Evolution

It seems intuitive that to understand the evolution of the nervous system, the

mechanism of scaling of animals and tissues will have to be considered. One way

to address this problem could be through the investigation of related organisms that

differ in size. If the patterning of the nervous system requires polarized morphoge-

netic signals that emanate from opposing sides of the embryo, then we might expect

that species of different embryonic sizes in which the source of morphogenetic

signals are located further apart should have variations in the organization of the

nervous system or other cell fates along the D/V axis.

Recent progress on the mechanisms of morphogenetic gradient scaling and

evolution has been made for A/P patterning in different fly species (McGregor

et al. 2001; Gregor et al. 2005, 2008; Lott et al. 2007). However, there are still a

number of gaps regarding scaling in the case of D/V patterning. For instance,

comparisons across species that differ in size using molecular markers for germ

layer domains are necessary to assess changes occurred during evolution of D/V

patterning. Also, quantitative expression profiles might resolve the question of

whether the levels of morphogens across related organisms that differ in size are

similar or significantly different. On a first estimate, divergent Drosophilids appear

to display variations in the width of peak levels of the Dorsal gradient (Crocker

et al. 2008), although a more precise quantitative measurement for those differences

is still lacking. Such comparisons are important to test the generality of predictions

made by current mathematical models based on D/V morphogenetic activity in one

species (Eldar et al. 2002; Mizutani et al. 2005; Zinzen et al. 2006; Kanodia et al.

2009) and begin elucidating the general principles that control the number of cells

allocated to particular tissue types. A better understanding of mechanisms that

govern tissue size and pattern is essential to manipulate tissue regeneration,

which is of relevance to the field of stem cell biology.

10.3.1 Investigation of Drosophila Sibling Specieswith Embryos That Vary in Size

In addition to those comparative studies, the investigation of closely related species

that can hybridize has the potential to clarify mechanisms of scaling by direct


experimentation. For instance, the Drosophila D/V patterning relies on both mater-

nal (e.g., Dorsal) and zygotic (e.g., Dpp/Sog) cues that can be completely separated

by generating hybrid embryos from the cross of species that produce embryos of

different sizes (i.e., which receive maternal information exclusively from one

species and zygotic information from both parents). In this regard, the D. melano-gaster subgroup of sibling species offers unique advantages to such studies.

D. simulans and D. sechellia became separated from the ancestor of D. melanoga-ster approximately 5 million years ago. D. sechellia is believed to have differen-

tiated from D. simulans more recently in Seychelles Islands some 0.5 million years

ago (Lachaise et al. 1986). The external anatomy of those three sibling species is

very similar and the only way to reliably distinguish them is by differences in the

male genitalia and to a lesser extent by the pigmentation of the sixth abdominal

tergite in females. Both at the genomic and chromosomal levels, these species are

almost identical (Horton 1939; Lemeunier and Ashburner 1984; Clark et al. 2007).

However, D. melanogaster and D. sechellia produce eggs of considerably differentsizes (Fig. 10.4) (Lott et al. 2007), which has been shown to be genetically

determined and under little influence by environmental factors (Warren 1924).

The similarity and discrete differences among these sibling species, coupled to

the ability of hybridizing them, offer the opportunity to begin addressing the

questions raised above. When the D/V partition of these species was analyzed

using D/V markers such as snail, the larger sized embryo of D. sechellia has a

significantly wider mesoderm than D. melanogaster (Mizutani, C.M., unpublished

data). This difference in mesoderm size is remarkable, given the recent divergence

of these two species. However, it is not yet clear how the maternal gradient of

Dorsal and the zygotic Dpp gradient behave when the scale is modified to sustain

similar neuroectodermal domains. It remains to be determined if larger animals need

to produce higher levels of these morphogens or whether compensatory mechanisms

Fig. 10.4 Scaling of

neuroectodermal domains and

germ layers in Drosophilaspecies with embryos of

different sizes. (a) D. busckii.(b) D. melanogaster.(c) D. sechellia. Theneuroectodermal domains

(NE) are maintained in all

three species, but the

mesodermal domains (MES)

vary in size


to circumvent variations in distance are at play. Other questions that these observa-

tions raise are whether a maternal gradient of one species can allocate the correct

number of cells per germ layer of another species, or if this process relies on zygotic

activity. As mentioned above, hybridization experiments should resolve these and

other issues regarding scaling. For instance, if only maternal cues define species-

specific D/V patterning, then hybrid embryos between those two species should

have a D/V subdivision similar to that provided by the mother of one of the species,

since information for the embryo size plus the entire machinery dedicated to

establish the Dorsal gradient are provided solely by the mother. Conversely, if

hybrid embryos displayed an intermediate D/V subdivision between two species,

then this would be indicative that zygotic determinants participate in the species-

specific partition of the germ layers. Ultimately, such experimental tests might help

define more precisely how D/V patterning evolves.

If the large embryo of D. sechellia has an increased mesoderm than D. melano-gaster, then we should expect a smaller embryo to have a narrower mesodermal

domain to compensate for the conserved size of the neuroectoderm observed in

several other insects (Whitington 1996). D. buskii lays embryos of about one third

of the size of D. melanogaster (Fig. 10.4) (Gregor et al. 2005), and indeed the

miniature embryos of this species have proportionally less mesodermal cells than

D. melanogaster and D. sechellia (Mizutani, C.M., unpublished data). Thus, this

is again in agreement that the D/V partition of the embryo should be sensitive to

the sources of morphogenetic information, distances, and consequently embryo

size. However, in contrast to the mesodermal variation, the number of cells con-

fined to the neuroectoderm is the same in all three species. Although this is

consistent with the fact that the nervous system patterning is under a strong pressure

to maintain an organization that preserves its function, the mechanisms that limit

the number of cells in the neuroectoderm cannot be explained by a D/V partition

based merely on either zygotic or maternal morphogenetic gradients. What are the

alternatives to explain this paradox? At this juncture, these observations are difficult

to reconcile and suggest we might be entering in new avenues of investigation of

evolutionary mechanisms of body plan formation. The ideas we would like to

discuss below are still highly speculative and based on a recent discovery of nuclei

movements in the Drosophila embryo.

10.3.2 Do Embryos Employ a Cell Counting Mechanism ThatCouples Nuclear Density and Morphogenetic Activity?

In general, the action of morphogenetic gradients is depicted as involving two static

cell populations: one committed to the role of sending a signal and another naıve

that receives and implements these signals. This view is convenient to establish the

differences in these two cell populations, but in living organisms, cell populations

are spatially displaced during cell divisions and morphogenetic movements. In the


Drosophila syncitium embryo, cell nuclei have a dynamic movement toward the

periphery of the embryo and undergo a few rounds of division while the Dorsal

gradient is being established (Roth et al. 1989; DeLotto et al. 2007; Kanodia et al.

2009). Once the embryo enters the 14th cycle, a long pause takes place without

any further cell divisions, and cellularization occurs with the invagination of

cell membranes during blastoderm stage E5 (Fig. 10.1a). During this stage, which

lasts about 40 min, most zygotic genes are regulated in response to D/V and A/P

gradients.

Contrary to the classical view of the blastoderm as being a stationary stage when

the nuclei stop dividing and no major cell movements of invaginations or germ

band extension occur until later in gastrulation, Keranen and colleagues recently

demonstrated that complex and ordered nuclei movements do occur during stage

E5, ultimately contributing to a highly stereotyped nuclear density packing in the

embryo (Keranen et al. 2006). Those authors show that some nuclei can move as far

as 20 mm (or three cell diameters) in a stereotyped fashion. In normal embryos,

lateral nuclei move toward the dorsal region, increasing their density along the

dorsal midline. Most ventral nuclei have a limited movement and reach a lower

density than other regions of the embryo by the end of stage E5. Interestingly, these

movement patterns are affected in mutants that disrupt the Dorsal gradient forma-

tion. It is well known that mutations in gd7 and Toll3 create apolar embryos without

any Dorsal (gd7) or ubiquitous Dorsal (Toll3) (Konrad et al. 1988; Schneider et al.

1991; Stathopoulos et al. 2002; Mizutani et al. 2006). What has escaped previous

analyses is the fact that those apolar embryos also exhibit a different distribution of

nuclei densities (Keranen et al. 2006), suggesting that the Dorsal signaling might be

required for the orderly control of cell movements observed in wild type embryos.

The control of nuclear movements can be directly or indirectly controlled by

Dorsal, and it is also possible to involve the zygotic expression of Dpp, since in the

apolar embryos, the expression of Dpp is either ubiquitous or absent. In either case,

it seems that the D/V morphogenetic gradients can modulate the final number of

cell nuclei that occupy different regions across the D/V axis, including the lateral

region that gives rise to the neuroectoderm. If morphogenetic gradients indeed

influence nuclear density as the data suggest, then this might be an important piece

of information to resolve the scaling paradox of the nervous system. The high levels

of Dorsal observed in ventral the nuclei could be the result of not only an increased

translocation of Dorsal to fixed positioned nuclei, but might also involve a prior (or

concomitant) control of the nuclear density in the ventral region about to achieve

the highest accumulation of nuclear Dorsal. This mechanism could in principle

limit the number of prospective neuroectodermal cells that acquire intermediate

levels of Dorsal, and potentially explain the constant width of the neuroectoderm of

closely related species that vary in the width of the mesoderm (Fig. 10.5). In this

case, the delimitation of cells within the ventral, intermediate, and dorsal domains

of the neuroectoderm would be under the control of the nuclear clustering activities

of D/V morphogens. However, in the mesoderm, Dorsal would be functioning in its

well-characterized role of threshold regulation of target genes, and thus susceptible

to variations in embryonic size.


10.4 Emergence of Novel Nervous System Properties

Despite Conservation in Cellular Architecture

In this review, we discussed that the organization of the nervous system is robust

and highly conservative. However, it is clear that this system finds breaches in this

robustness to create novelty. This observation is particularly pertinent in the light of

the sharp behavioral mating preferences and ecological differences that exist among

Fig. 10.5 Distribution of dorsal–ventral gradients, cell fate positions and nuclei density packing in

different Drosophila species. Graph representing Dorsal (red line) and Dpp (blue line) gradientlevels. Abscise indicates position of cell fates and gene expression domains along the D/V axis:

sna (red), vnd (blue), ind (green), msh (magenta), and dpp (yellow). Colored bar indicates nucleardensity packing (orange, low density; black, intermediate density; blue, high density). All three

species, D. melanogaster (a), D. sechellia (b), and D. busckii (c) have equally sized neuroecto-

dermal domains (vnd, ind and msh), but variable mesodermal domains (sna). Peak levels of Dorsalgradient inD. melanogaster (a) are higher than inD. sechellia (b), while the gradient is wider inD.sechellia than in D. melanogaster (Mizutani, unpublished data). One can speculate that D. busckiihas a Dorsal gradient with higher peak and narrower width than D. melanogaster (c). Nuclei

distribution for D. sechellia and D. busckii is hypothetical and takes into consideration a higher

concentration of nuclei in the ventral region in the case of D. sechellia and lower concentration in

D. busckii, in comparison to D. melanogaster. Such distribution could in principle change the finalDorsal gradient shape. Finally, another hypothetical representation is the Dpp gradient, which

would scale with size in all three species, based on models made in Xenopus (Ben-Zvi et al. 2008)


theD. melanogaster sibling species (Watanabe and Kawanishi 1979; Lachaise et al.

1986). There are many alternative ways to create flexibility in nervous system

function without necessarily changing neuronal cell identities in closely related

species (reviewed in Katz and Harris-Warrick 1999). For instance, one can specu-

late that the increase in mesoderm verified in D. sechellia could lead to changes in

peripheral muscle tissue, with the consequence of modifying neural muscular

junction connections. Such modifications could in turn alter output responses

distinct from D. melanogaster in a variety of behaviors, including locomotion, or

mating courtship rituals produced by males. Indeed, the muscle of Lawrence,

responsible for wing vibration in males during courtship, has an increased number

of fibers inD. sechellia thanD. melanogaster, which could explain the difference inlove song frequencies in the two species (Orgogozo et al. 2007). Another way to

create novel behavioral functions would be through mutations in single genes that

regulate some aspect of neural physiology or response. One example is the loss of

olfactory and gustatory receptors in D. sechellia, which allowed this species to

specialize in feeding on Morinda fruit (Matsuo et al. 2007; McBride 2007). This

fruit contains toxic levels of octanol and is avoided by all other sibling species,

D. melanogaster, D. simulans, and D. mauritiana that retained these receptors.

With the sequencing of the genome of twelve Drosophila species (Clark et al.

2007), pair wise genome comparison among the D. melanogaster subgroup has

allowed the discovery of ancestral and fast-evolving alleles with predicted neural

functions, including potassium channels and additional gustatory and odorant

receptors (Sousa-Neves and Rosas, 2010).

10.5 Conclusion

The nervous system organization along the D/V axis into distinct domains of gene

expression is conserved in most bilaterian organisms and appears to rely on the

ancient BMP/dpp and Chd/sog signaling cassette. Previous work in insect embryos

has shown that this conservation can be resolved at the cellular level, since

neuroblast maps among divergent insects are very similar in terms of number and

types of cells. Evolutionary changes in embryo size must pose a tremendous

challenge to the scaling properties of morphogenetic gradients to constrain the

number of cells within these neural domains, and at the same time, novel body plans

can be created by altering the determination of other germ layers under a low

evolutionary pressure. We speculate that in addition to its traditional role in

defining cell fate and proliferation, morphogenetic gradients may also coordinate

nuclear clustering and distribution, which may function as a cell counting mecha-

nism that allocates the correct number of cells within specific dorsal–ventral

domains of embryos in Drosophilids. Future experimental and computational

modeling studies in closely related Drosophila species might reveal emerging

properties of evolutionary mechanisms of germ layer formation.


References

Arendt D, Nubler-Jung K (1994) Inversion of dorsoventral axis? Nature 371:26

Arendt D, Denes AS, Jekely G, Tessmar-Raible K (2008) The evolution of nervous system

centralization. Philos Trans R Soc Lond B Biol Sci 363:1523–1528

Ben-Zvi D, Shilo BZ, Fainsod A, Barkai N (2008) Scaling of the BMP activation gradient in

Xenopus embryos. Nature 26:1205–1211

Bhat KM (1999) Segment polarity genes in neuroblast formation and identity specification during

Drosophila neurogenesis. Bioessays 21:472–485

Biehs B, Francois V, Bier E (1996) The Drosophila short gastrulation gene prevents Dpp

from autoactivating and suppressing neurogenesis in the neuroectoderm. Genes Dev 10:

2922–2934

Bier E (1997) Anti-neural-inhibition: a conserved mechanism for neural induction. Cell

89:681–684

Brideau NJ, Flores HA, Wang J, Maheshwari S, Wang X, Barbash DA (2006) Two Dobzhansky–

Muller genes interact to cause hybrid lethality in Drosophila. Science 314:1292–1295

Briscoe J, Sussel L, Serup P, Hartigan-O’Connor D, Jessell TM, Rubenstein JL, Ericson J (1999)

Homeobox gene Nkx2.2 and specification of neuronal identity by graded Sonic hedgehog

signalling. Nature 398:622–627

Chen G, Handel K, Roth S (2000) The maternal NF-kappaB/dorsal gradient of Tribolium casta-

neum: dynamics of early dorsoventral patterning in a short-germ beetle. Development

127:5145–5156

Clark AG, Eisen MB, Smith DR, Bergman CM, Oliver B, Markow TA, Kaufman TC, Kellis M,

Gelbart W, Iyer VN et al (2007) Evolution of genes and genomes on the Drosophila phylogeny.

Nature 450:203–218

Cowden J, Levine M (2003) Ventral dominance governs sequential patterns of gene expression

across the dorsal-ventral axis of the neuroectoderm in the Drosophila embryo. Dev Biol

262:335–349

Crocker J, Tamori Y, Erives A (2008) Evolution acts on enhancer organization to fine-tune

gradient threshold readouts. PLoS Biol 6:e263

De Robertis EM (2008) Evo-devo: variations on ancestral themes. Cell 132:185–195

De Robertis EM, Sasai Y (1996) A common plan for dorsoventral patterning in Bilateria. Nature

380:37–40

DeLotto R, DeLotto Y, Steward R, Lippincott-Schwartz J (2007) Nucleocytoplasmic shuttling

mediates the dynamic maintenance of nuclear dorsal levels during Drosophila embryogenesis.

Development 134:4233–4241

Denes AS, Jekely G, Steinmetz PR, Raible F, Snyman H, Prud’homme B, Ferrier DE, Balavoine G,

Arendt D (2007) Molecular architecture of annelid nerve cord supports common origin of

nervous system centralization in bilateria. Cell 129:277–288

Doe CQ (1992) Molecular markers for identified neuroblasts and ganglion mother cells in the

Drosophila central nervous system. Development 116:855–863

Doe CQ (2008) Neural stem cells: balancing self-renewal with differentiation. Development

135:1575–1587

Doe CQ, Skeath JB (1996) Neurogenesis in the insect central nervous system. Curr Opin

Neurobiol 6:18–24

Eldar A, Dorfman R, Weiss D, Ashe H, Shilo BZ, Barkai N (2002) Robustness of the BMP

morphogen gradient in Drosophila embryonic patterning. Nature 419:304–308

Ferguson EL (1996) Conservation of dorsal–ventral patterning in arthropods and chordates. Curr

Opin Genet Dev 6:424–431

Ferrandon D, Imler JL, Hetru C, Hoffmann JA (2007) The Drosophila systemic immune

response: sensing and signalling during bacterial and fungal infections. Nat Rev Immunol

7:862–874


Francois V, Solloway M, O’Neill JW, Emery J, Bier E (1994) Dorsal–ventral patterning of the

Drosophila embryo depends on a putative negative growth factor encoded by the shortgastrulation gene. Genes Dev 8:2602–2616

Geoffroy St.-Hilaire E (1822) Considerations generales sur la vertebre. Mem Mus Hist Nat

9:89–119

Gompel N, Prud’homme B, Wittkopp PJ, Kassner VA, Carroll SB (2005) Chance caught on the

wing: cis-regulatory evolution and the origin of pigment patterns in Drosophila. Nature

433:481–487

Gregor T, Bialek W, de Ruyter van Steveninck RR, Tank DW, Wieschaus EF (2005) Diffusion

and scaling during early embryonic pattern formation. Proc Natl Acad Sci USA 102:

18403–18407

Gregor T, McGregor AP, Wieschaus EF (2008) Shape and function of the bicoid morphogen

gradient in dipteran species with different sized embryos. Dev Biol 316:350–358

Holland LZ (2009) Chordate roots of the vertebrate nervous system: expanding the molecular

toolkit. Nat Rev Neurosci 10:736–746

Holley SA, Jackson PD, Sasai Y, Lu B, De Robertis EM, Hoffmann FM, Ferguson EL (1995) A

conserved system for dorsal-ventral patterning in insects and vertebrates involving sog and

chordin. Nature 376:249–253Horton IH (1939) A comparison of the salivary gland chromosomes of Drosophila melanogaster

and D. simulans. Genetics 24:234–243

Illes JC, Winterbottom E, Isaacs HV (2009) Cloning and expression analysis of the anterior

parahox genes, Gsh1 and Gsh2 from Xenopus tropicalis. Dev Dyn 238:194–203

Irish VF, Gelbart WM (1987) The decapentaplegic gene is required for dorsal–ventral patterning

of the Drosophila embryo. Genes Dev 1:868–879

Isshiki T, Takeichi M, Nose A (1997) The role of the msh homeobox gene during Drosophilaneurogenesis: implication for the dorsoventral specification of the neuroectoderm. Develop-

ment 124:3099–3109

Jacob J, Briscoe J (2003) Gli proteins and the control of spinal-cord patterning. EMBO Rep

4:761–765

Jimenez F, Martin-Morris LE, Velasco L, Chu H, Sierra J, Rosen DR, White K (1995) vnd, a generequired for early neurogenesis of Drosophila, encodes a homeodomain protein. EMBO

J 14:3487–3495

Kanodia JS, Rikhy R, Kim Y, Lund VK, DelottoR Lippincott-Schwartz J, Shvartsman SY (2009)

Dynamics of the dorsal morphogen gradient. Proc Natl Acad Sci USA 106:21707–21712

Kassis JA (1990) Spatial and temporal control elements of the Drosophila engrailed gene. Genes

Dev 4:433–443

Katz PS, Harris-Warrick RM (1999) The evolution of neuronal circuits underlying species-specific

behavior. Curr Opin Neurobiol 9:628–633

Keranen SV, Fowlkes CC, Luengo Hendriks CL, Sudar D, Knowles DW, Malik J, Biggin MD

(2006) Three-dimensional morphology and gene expression in the Drosophila blastoderm at

cellular resolution II: dynamics. Genome Biol 7:R124

Konrad KD, Goralski TJ, Mahowald AP (1988) Developmental genetics of the gastrulation

defective locus in Drosophila melanogaster. Dev Biol 127:133–142

Kriks S, Lanuza GM, Mizuguchi R, Nakafuku M, Goulding M (2005) Gsh2 is required for the

repression of Ngn1 and specification of dorsal interneuron fate in the spinal cord. Development

132:2991–3002

Lachaise D, David JR, Lemeunier F, Tsacas L, Ashburner M (1986) The reproductive relationship

of Drosophila sechellia with Drosophila mauritiana, Drosophila simulans and Drosophila

melanogaster from the afro-tropical region. Evolution 40:262–271

Lapraz F, Besnardeau L, Lepage T (2009) Patterning of the dorsal–ventral axis in echinoderms:

insights into the evolution of the BMP-chordin signaling network. PLoS Biol 7:e1000248

Lemeunier F, Ashburner M (1984) Relationships within the melanogaster species subgroup of the

genus Drosophila (Sophophora). Chromosoma 89:343–351


Liberman LM, Stathopoulos A (2009) Design flexibility in cis-regulatory control of gene expres-

sion: synthetic and comparative evidence. Dev Biol 327:578–589

Liem KF Jr, Tremml G, Roelink H, Jessell TM (1995) Dorsal differentiation of neural plate cells

induced by BMP-mediated signals from epidermal ectoderm. Cell 82:969–979

Liem KF Jr, Jessell TM, Briscoe J (2000) Regulation of the neural patterning activity of sonic

hedgehog by secreted BMP inhibitors expressed by notochord and somites. Development

127:4855–4866

Litingtung Y, Chiang C (2000) Specification of ventral neuron types is mediated by an antagonistic

interaction between Shh and Gli3. Nat Neurosci 3:979–985

Liu Y, Helms AW, Johnson JE (2004) Distinct activities of Msx1 and Msx3 in dorsal neural tube

development. Development 131:1017–1028

Lott SE, Kreitman M, Palsson A, Alekseeva E, Ludwig MZ (2007) Canalization of segmentation

and its evolution in Drosophila. Proc Natl Acad Sci USA 104:10926–10931

Lowe CJ, Terasaki M, WuM, Freeman RM Jr, Runft L, Kwan K, Haigo S, Aronowicz J, Lander E,

Gruber C et al (2006) Dorsoventral patterning in hemichordates: insights into early chordate

evolution. PLoS Biol 4:e291

Ludwig MZ, Patel NH, Kreitman M (1998) Functional analysis of eve stripe 2 enhancer evolution

in Drosophila: rules governing conservation and change. Development 125:949–958

Matsuo T, Sugaya S, Yasukawa J, Aigaki T, Fuyama Y (2007) Odorant-binding proteins OBP57d

and OBP57e affect taste perception and host-plant preference in Drosophila sechellia. PLoS

Biol 5:e118

McBride CS (2007) Rapid evolution of smell and taste receptor genes during host specialization in

Drosophila sechellia. Proc Natl Acad Sci USA 104:4996–5001

McDonald JA, Holbrook S, Isshiki T, Weiss J, Doe CQ, Mellerick DM (1998) Dorsoventral

patterning in the Drosophila central nervous system: the vnd homeobox gene specifies ventral

column identity. Genes Dev 12:3603–3612

McGregor AP, Shaw PJ, Hancock JM, Bopp D, Hediger M, Wratten NS, Dover GA (2001) Rapid

restructuring of bicoid-dependent hunchback promoters within and between Dipteran species:

implications for molecular coevolution. Evol Dev 3:397–407

Mellerick DM, Modica V (2002) Regulated vnd expression is required for both neural and glial

specification in Drosophila. J Neurobiol 50:118–136

Mizutani C, Bier E (2008) EvoD/Vo: the origins of BMP signalling in the neuroectoderm. Nat Rev

Genet 9:663–677

Mizutani CM, Nie Q, Wan FY, Zhang YT, Vilmos P, Sousa-Neves R, Bier E, Marsh JL,

Lander AD (2005) Formation of the BMP activity gradient in the Drosophila embryo. Dev

Cell 8:915–924

Mizutani CM, Meyer N, Roelink H, Bier E (2006) Threshold-dependent BMP-mediated repres-

sion: a model for a conserved mechanism that patterns the neuroectoderm. PLoS Biol 4:e313

Nomaksteinsky M, Rottinger E, Dufour HD, Chettouh Z, Lowe CJ, Martindale MQ, Brunet JF

(2009) Centralization of the deuterostome nervous system predates chordates. Curr Biol

19:1264–1269

Nunes da Fonseca R, von Levetzow C, Kalscheuer P, Basal A, van der Zee M, Roth S (2008) Self-

regulatory circuits in dorsoventral axis formation of the short-germ beetle Tribolium casta-

neum. Dev Cell 14:605–615

Orgogozo V, Muro NM, Stern DL (2007) Variation in fiber number of a male-specific muscle

between Drosophila species: a genetic and developmental analysis. Evol Dev 9:368–377

Padgett RW, St Johnston RD, Gelbart WM (1987) A transcript from a Drosophila pattern gene

predicts a protein homologous to the transforming growth factor-beta family. Nature

325:81–84

Padgett RW, Wozney JM, Gelbart WM (1993) Human BMP sequences can confer normal

dorsal–ventral patterning in the Drosophila embryo. Proc Natl Acad Sci USA 90:2905–2909

Ray RP, Arora K, Nusslein-Volhard C, Gelbart WM (1991) The control of cell fate along the

dorsal–ventral axis of the Drosophila embryo. Development 113:35–54


Rentzsch F, Anton R, Saina M, Hammerschmidt M, Holstein TW, Technau U (2006) Asymmetric

expression of the BMP antagonists chordin and gremlin in the sea anemone Nematostella

vectensis: implications for the evolution of axial patterning. Dev Biol 296:375–387

Roth S, Stein D, Nusslein-Volhard C (1989) A gradient of nuclear localization of the dorsal protein

determines dorsoventral pattern in the Drosophila embryo. Cell 59:1189–1202

Rushlow CA, Han K, Manley JL, Levine M (1989) The graded distribution of the dorsal

morphogen is initiated by selective nuclear transport in Drosophila. Cell 59:1165–1177

Saina M, Genikhovich G, Renfer E, Technau U (2009) BMPs and chordin regulate patterning of

the directive axis in a sea anemone. Proc Natl Acad Sci USA 106:18592–18597

Samuel G, Miller D, Saint R (2001) Conservation of a DPP/BMP signaling pathway in the

nonbilateral cnidarian Acropora millepora. Evol Dev 3:241–250

Sasai Y, Lu B, Steinbeisser H, Geissert D, Gont LK, De Robertis EM (1994) Xenopus

chordin: a novel dorsalizing factor activated by organizer-specific homeobox genes. Cell

79:779–790

Schmid A, Chiba A, Doe CQ (1999) Clonal analysis of Drosophila embryonic neuroblasts: neural

cell types, axon projections and muscle targets. Development 126:4653–4689

Schmidt J, Francois V, Bier E, Kimelman D (1995) Drosophila short gastrulation induces an

ectopic axis in Xenopus: evidence for conserved mechanisms of dorsal–ventral patterning.

Development 121:4319–4328

Schneider DS, Hudson KL, Lin TY, Anderson KV (1991) Dominant and recessive mutations

define functional domains of Toll, a transmembrane protein required for dorsal–ventral polarity

in the Drosophila embryo. Genes Dev 5:797–807

Sousa-Neves R, Rosas A (2010) An Analysis of Genetic Changes during the Divergence of

Drosophila species. PloS One 5(5): e10485. doi:10.1371/journal.pone.0010485

Spemann H, Mangold H (1924) Uber induction von embryonanlagen durch implantation artfrem-

der organis atoren. W Roux’ Arch Ent Org 100:599–638

Stathopoulos A, Van Drenth M, Erives A, Markstein M, Levine M (2002) Whole-genome analysis

of dorsal–ventral patterning in the Drosophila embryo. Cell 111:687–701

Steward R (1989) Relocalization of the dorsal protein from the cytoplasm to the nucleus correlates

with its function. Cell 59:1179–1188

Sturtevant AH (1929) Contributions to the genetics of Drosophila simulans and Drosophila

melanogaster. I. The genetics of Drosophila simulans. Publs Carnegie Instn 399:1–62

Suzuki A, Ueno N, Hemmati-Brivanlou A (1997) Xenopus msx1 mediates epidermal induction

and neural inhibition by BMP4. Development 124:3037–3044

Technau GM, Berger C, Urbach R (2006) Generation of cell diversity and segmental pattern in the

embryonic central nervous system of Drosophila. Dev Dyn 235:861–869

Thomas JB, Bastiani MJ, Bate M, Goodman CS (1984) From grasshopper to Drosophila: a

common plan for neuronal development. Nature 310:203–207

Ungerer P, Scholtz G (2008) Filling the gap between identified neuroblasts and neurons in

crustaceans adds new support for Tetraconata. Proc Biol Sci 275:369–376

Valerius MT, Li H, Stock JL, Weinstein M, Kaur S, Singh G, Potter SS (1995) Gsh-1: a novel

murine homeobox gene expressed in the central nervous system. Dev Dyn 203:337–351

WangW, Chen X, Xu H, Lufkin T (1996) Msx3: a novel murine homologue of the Drosophila msh

homeobox gene restricted to the dorsal embryonic central nervous system. Mech Dev

58:203–215

Warren DC (1924) Inheritance of Egg Size in Drosophila melanogaster. Genetics 9:41–69

Watanabe TK, Kawanishi M (1979) Mating preference and the direction of evolution in drosoph-

ila. Science 205:906–907

Weiss JB, Von Ohlen T, Mellerick DM, Dressler G, Doe CQ, Scott MP (1998) Dorsoventral

patterning in the Drosophila central nervous system: the intermediate neuroblasts defectivehomeobox gene specifies intermediate column identity. Genes Dev 12:3591–3602


Wharton KA, Ray RP, Gelbart WM (1993) An activity gradient of Decapentaplegic is necessary

for the specification of dorsal pattern elements in the Drosophila embryo. Development

117:807–822

Wheeler SR, Carrico ML, Wilson BA, Skeath JB (2005) The Tribolium columnar genes reveal

conservation and plasticity in neural precursor patterning along the embryonic dorsal–ventral

axis. Dev Biol 279:491–500

Whitington PM (1996) Evolution of neural development in the arthropods. Semin Cell Dev Biol

7:605–614

Wittkopp PJ, Vaccaro K, Carroll SB (2002) Evolution of yellow gene regulation and pigmentation

in Drosophila. Curr Biol 12:1547–1556

Yamamoto MT, Kamo M, Yamamoto S, Watanable TK (1997) Cytogenetic mapping of lethal

hybrid rescue gene of Drosophila simulans. Genes Genet Syst 72:297–301Zinzen RP, Senger K, Levine M, Papatsenko D (2006) Computational models for neurogenic gene

expression in the Drosophila embryo. Curr Biol 16:1358–1365


Chapter 11

Evolutionary Genomics for Eye Diversification

Atsushi Ogura

Abstract There are several types of eyes in morphology such as camera eye,

compound eye, mirror eye, and single lens eye, and all the eye types have been

evolved from the same origin, the prototype eye. Even though there are conserved

genes and networks in the eye evolution, little is known about what kinds of genetic

basis have been contributed to the eye diversification. It is essential for discovering

genes for the morphological diversification to develop a platform of genomic and

transcriptomic comparison among species. We, therefore, developed microarray

that cover the genes related to development, function, and structure of molluscan

eye, as an example, for the evolutionary genomic studies.

11.1 Evolutionary Genomics for Eye Diversification

11.1.1 Evolution of the Eye

The eye is one of the most elaborate organs in animals and the study of its evolution

is of particular interest. The evolution of animal eyes has been one of the most

fundamental and classical subjects in the field of biology dating back to the time of

Darwin. However, it has been difficult to understand how this complex organ arose

simply from mutations and selections. Darwin discussed this matter in his “On the

Origin of Species by Means of Natural Selection” in a chapter titled, “Difficulties of

the Theory”, in which he wrote that “organs of extreme perfection and complica-

tion” such as the eye remained inexplicable by his theory (Darwin 1859).

A. Ogura

Division of Advanced Sciences, Ochadai Academic Production, Ochanomizu University, Ohtsuka

2-1-1, Bunkyo, Tokyo 112-8610, Japan



179

The evolutionary study of animal eyes was also difficult for a long time from the

viewpoint of molecular evolution and biology. There were only a few molecular

theories to link primitive eyes to the elaborate and varied eye organs commonly

seen today. It seemed that natural selection could not adequately explain the

evolutionary mechanism underlying the development of complex animal eyes.

However, studies based on basic control genes in the developmental processes in

animal eyes have revealed that there is a conserved key regulatory network repre-

sented by the Pax6 genes among almost all animals (Gehring 1996; Fernald 2000).

Even though there are no clear evolutionary tracks between the various types of

animal eyes, the evolutionary history of the eye can be explained from the con-

served molecular mechanisms. Recent studies have also reported that not only the

core gene regulatory network for eye development but also genes downstream of

the network and other peripheral genes related to the function and structure of eyes

have been conserved among animals at least since the split of bilateral animals

(Box 11.1). The origin and ancestral prototype of the eye as well as the molecular

mechanism underlying the diversification of the various eye types, however, remain

unclear.

Box 11.1

akhirin, apkc, apterous, arf4, arm, arr3, Arrestin, Ascll, ash1, ath5, Atoh4,

Atoh7, atonal, bad, BarH1, BarH2, barhl2, baz, bbs1, bbs2, bHLH, Big

brother, blimp1, blue-opsin, Bmp4, Bmp7, BRD-U, brn1, Brn3a, Brn3b,

Brother, bunched, c-kit, c-myc, calb2a, calb2b, Calphotin, CG13030, chaop-

tic, Chx10, cko, Cpsf1, crb, crb1, cre, crx, Cryba4, Cryz, cut, dachshund,

daughterless, dbx1/2, delta1, dkk1, Dlx1, Dlx2, drosocrystallin, dynein,

ectopic, eli, elk3, En-1, equarin, err2beta, ERRbeta, esrrb, Etv6, extramacro-

chaete, ey, Eya, flox, flr, Foxa2, FoxC1, FoxC2, FoxD1, FoxE3, FoxF1,

FoxG1, FoxK1, FoxL1, FoxM1, FoxN2, FoxN3, FoxN4, Foxn5, FoxO,

FoxO3, FoxP1, FoxP2, FoxP4, FoxS1, fzd4/5, gam1, gapeh, glass, gli2,

gli3, glu, hairy, Hes1, hes1, hes1, hes2, hes5, hmgb3, homothorax, Hoxb1,

Ihx, inaF, Islet-1, Jagged1, jmjd2c, jmjN/C, kif3, klingon, Krt1-12, L1cam,

lazaro, lfe1, Lhx2, Lhx9, Lmo4, lok, lozenge, lrp5, lrx3, m-opsin, mab21l1,

mab21l2, Maf, Math3, Math5, mdka, mdkb, meis2, melanopsin, mirror, Mitf,

Mmp9, Mocs3, mts, Munster, musashi, myc, nanog, ncad, ncam1, Necab2,

nestin, NeuroD, NeuroD1, ngn3, nkx2, nkx6, Nlz1, Nlz2, noggin, nohp1,

notch, nphp, nr2e1, Nr2e3, nr2f1, nr2f2, nr3b2, nrl, nrx, ocelliless, oct4, of,

onecut, opsin1, Optix, OTX1, OTX2, otx2, ovl, p27Xic1, p57kip2, par3, par6,

patj, Pax2, Pax6, pax6cre, pax7, PaxB, pebbled, peripherin, PhospholipaseD,

phyllopod, pi3p, Pias3, pikachurin, Pitx3, PNR, pp2a, pralemmin, Prep1,

prospero, Prox1, ptc, pten, Rab, RARa, RARb, RARg, rax, Rb1, recoverin,retp1, Rex1, rhodopsin, ror, rough, rpgrip1, rs1, runt, rx1, rx2, rx3, rxr, sara,

scabrous, shaven, Shh, sit, six1, six3, six6, smo, snare, so, sox1, sox11, sox2,

sox3, sox4a, sox8, sox9, SoxN, spineless, ssea-1, stardust, sufu, sufuko,

(continued)

180 A. Ogura

sumo1, syn3, tangerinA, target, tbx3, tbx5, teashirt, TFIID, TGIF, TGIF2,

timeless, tiptop, to, tramtrack, TRbeta1, TRbeta2, trp-like, trpgamma, trpm1,

tsk, tws, ubc9, vax2, vsx1, vsx2, warts, wnt, wnt2b, xhmgb3, xwnt8, zic1,

zic2, zic3

These genes were collected from NCBI and Pubmed and considered to be related to

development, structure, and function of animal eyes.

11.1.2 Origin and Prototype of the Eye

The evolution of different eye types might have occurred many times as indepen-

dent events in the lineages of different animals. However, few eyes have ever been

found as fossils because they are soft organs, thereby making it difficult to examine

the origin of animal eyes. Only in some animals, such as trilobites, the eyes consist

of calcite lenses can be fossilized. Trilobite fossils with compound eyes have been

found that date back to the early Cambrian period some 540 million years ago. This

suggests that the origin of eyes occurred before the preCambrian period. Cnidarian

is one of the most primitive animals to possess eyes, and intensive studies of the

cnidarian eye have revealed the fundamental mechanisms for eye formation and

development (Kozmik et al. 2003). Once fundamental genes for animal eyes and a

common type of photoreceptor cell were discovered to be conserved among

animals from the common origin, the evolution and diversification of different

eye types could be considered, not as independent events, but as divergent events

that originated from a prototype eye present in the ancestral species. This raises the

question as to the exact form and structure of the prototype eye. Gehring and Ikeo

have inferred a two celled prototype eye consisting of one photoreceptor cell and

one pigment cell (Gehring and Ikeo 1999). Recently, Gehring has suggested that the

eye organelle in the Protist, dinoflagellate, might be the prototype eye, and the

origin of all animals eyes (Gehring 2005). The characteristics of the prototype eye

can be estimated by a comparison of the structure and molecular basis of extant

animal eyes. The next question for researches is how various types of eyes came to

be diverged from the prototype eye.

11.1.3 Diversification of the Eye

Photoreceptors, as suggested by Salvini-Plawen and Mayr on the basis of morpho-

logical and embryological studies, have evolved independently in 40–65 different

lineages (Salvivi-Plawen and Mayr 1977). However, studies based on molecular

biology and evolution have revealed that, even though the evolutionary processes of

11 Evolutionary Genomics for Eye Diversification 181

different types of eyes seem different, the molecular basis is shared among the

various eye types and they arose by divergent evolution (Nilsson 2004; Serb and

Eernisse 2008). These phenomena have often been explained by the concepts of

convergent and divergent evolution. Convergent evolution is defined as the mecha-

nism by which similar tissue or organ structures can be evolved from different

origins or via different processes. Divergent evolution, on the other hand, is defined

as the evolutionary process in which different types of tissues and organs can be

evolved from the same origin. The camera eye of vertebrates and cephalopods, in

spite of the outward similarities, can be considered to be the result of divergent

evolution using the same gene source and genetic mechanisms (Ogura et al. 2004).

Jumping spiders also possess highly evolved camera eye like vertebrates but they

have acquired their eyes independently, which was validated by the phylogenetic

analysis (Su et al. 2007). Camera eye can be also found in more primitive species of

Cnidaria, cubozoan jellyfish (Nilsson 2004). These divergent mechanisms are the

key to explaining the diversification of not only the evolution of camera eye but also

that of various eye types.Molluscs provide one of the best targets for the study of this

topic because, even within one lineage of molluscs, all eye types can be found

(Kozmik et al. 2008). Squid and octopuses have a camera eye, the nautilus has a

pinhole eye, the scallop has a mirror eye, and the ark shell has a compound eye

(Fig. 11.1).

Fig. 11.1 Eyes of molluscs. Pictures show various types of eyes in molluscs; (a) Loligo vulgaris,a squid belonging to the family Loliginidae. (b) An eyeball extracted from Loligo. (c) Embryo of

idiosepius, pygmy squid. (d) Nautilus pompilius that has a pinhole eye. (e) Pecten yessoensis,a Japanese sea scallop that has a hundred of tiny mirror eyes

182 A. Ogura

The vertebrate camera eye was developed from the neural plate and formed an

optic vesicle, which was subsequently invaginated to form an optic cup. On the

contrary, the cephalopod camera eye developed as an evagination of the brain

leading to an invagination of the ectoderm. These differences in origin have

resulted in distinct differences in the orientation of the photoreceptor cells between

vertebrates and cephalopods, in which they face the light source in cephalopods but

face in the opposite direction in vertebrates.

The compound eye has a complex structure and is found inmany species including

Arthropoda and Mollusca. They are very different from the camera eye and consist

of hundreds of individual eyes with lenses and photoreceptors. In Drosophila, for

example, the eye primordia formed as an invagination of the embryonic ectoderm

that forms the eye imaginal disk in the larvae. During metamorphosis, the eye disk

organizes itself to form the compound eye, the photoreceptor cells of which extend

their axons backwards from the periphery to establish contact with the brain.

11.1.4 Genomic and Transcriptomic Approachesto Eye Evolution

Recent work on evolutionary genomics in various types of eyes, together with

comparative analyses of gene expression comparison among closely related spe-

cies, has led to the hypothesis of a dynamic mechanism for the diversification of

eyes (Wistow 2006; Choy et al. 2006; Bao and Friedrich 2009; Baker et al. 2009).

The advantages of these large-scale genomic and transcriptomic studies of animal

eyes are that they can trace the evolutionary process of not only the key regulatory

genes, such as Pax6, but also genes related to eye function and maintenance through

the analyses of orthologous gene sets involved in eye evolution. These advance-

ments were achieved by large-scale analyses using microarray technologies and

next generation sequencers.

Molluscs provide a good example of the application of evolutionary genomics

studies, as all eye types have evolved in one lineage. It is essential to identify the

genes responsible for the morphological diversification so as to allow the develop-

ment of a platform for the comparison of gene expression among species. In this

example, a microarray that covers the genes related to the development, function,

and structure of the molluscan eye was developed by constructing full-length cDNA

libraries for the octopus, nautilus, scallop, and two squid species (Fig. 11.2). This

strategy provides comparative genomic and transcriptomic approaches to the

molecular mechanism for diversification in the molluscan eye. The Molluscan

Eye Array, based on the above microarray, is designed for the comparative gene

expression analysis of the molluscan eye with genes expressed in eye of loligo,

octopus, nautilus, and pecten, as well as genes known to be expressed in vertebrate

eyes, and genes expressed in the idiosepius, the pygmy squid, and brain. We have

designed conserved regions of the genes for the microarray probes to detect the

gene expression of orthologous genes.


As a result of the Molluscan Eye Array experiments using RNA samples from

the adult eye of the idiosepius, nautilus, and pecten, we could estimate the genes

expressed differentially among species that played an important role in the diversi-

fication of eye structure. More than 88% of the probe designed from the same

species tested in the experiment could be identified as expressed genes, and

10–30% of the probes could be detected by the RNA samples of different species

that were unknown transcripts ever (Fig. 11.3a). To validate the reliability of

interspecies array, we have tested gene expression of Pax6 in idiosepius with the

probe designed from zebrafish Pax6 gene and confirmed its expression by in situ

hybridization.

Next, to distinguish the stage-specific and camera eye-specific expression of eye

genes in cephalopods, we used RNAs from three different embryonic stages of the

pygmy squid eye for the array. We found that 2,893 genes are expressed in the squid

embryonic eye but not in the eyes of nautilus or pecten. Only 269/2,893 (9.3%)

genes were adult-specific expression in the idiosepius. In addition, 634/2,893 genes

are commonly observed in the gene expression databases of vertebrate eye and

retina (Fig. 11.3b). These results show that this approach provides an efficient

platform and database for searching candidate genes involved in camera eye

acquisition.

Furthermore, expression diversities of eye-related genes in molluscs were exam-

ined by calculating how much genes were shared to be expressed among species.

Pecten shows lower gene expression diversity comparing with squid and nautilus

statistically (Fig. 11.3c). This result indicates that pecten tended to conserve

commonly used genes since the last common ancestor of mollusca and not to

acquire novel gene much more than other molluscs. Cephalopods, on the other

Fig. 11.2 Scheme of Molluscan Eye Array design was illustrated. Eyeballs of five different

species, loligo, idiosepius, octopus, nautilus, and pecten, were extracted and used for the construc-

tion of cDNA libraries

184 A. Ogura

hand, tended to acquire lineage-specific genes in relation to the evolution of camera

eye structure, which makes their expression diversities higher than pecten.

Thus, evolutionary genomic and transcriptomic approaches might contribute to

the elucidation of the diversification mechanisms of animal eyes by searching

common and unique genes in the developmental processes and eye structures.

References

Baker RH et al (2009) Genomic analysis of a sexually-selected character: EST sequencing and

microarray analysis of eye-antennal imaginal discs in the stalk-eyed fly Teleopsis dalmanni

(Diopsidae). BMC Genomics 10(1):361

Bao R, Friedrich M (2009) Molecular evolution of the Drosophila retinome: exceptional gene gain

in the higher Diptera. Mol Biol Evol 26(6):1273–1287

Fig. 11.3 Characteristics of eye gene expressions in molluscs were shown in the figure. (a)

Proportions of probes designed from idiosepius, nautilus, and pecten hybridized that were detected

as expressed genes in the three Molluscan Eye Array experiments, idiosepius mRNA, nautilus

mRNA, and pecten mRNA were indicated. (b) Squid camera eye-specific genes were estimated by

comparing mRNA expression in the Molluscan Eye Array. (c) Species-specific expressions

represent exclusive gene expression in a particular species, and conserved expressions represent

gene expression that were observed in more than one species


Choy KW et al (2006) Genomic annotation of 15, 809 ESTs identified from pooled early gestation

human eyes. Physiol Genomics 25(1):9–15

Darwin (1859) On the Origin of Species by Means of Species

Fernald RD (2000) Evolution of eyes. Curr Opin Neurobiol 10(4):444 – 450

Gehring WJ (1996) The master control gene for morphogenesis and evolution of the eye. Genes

Cells 1:11–15

Gehring WJ (2005) New perspectives on eye development and the evolution of eyes and photo-

receptors. J Hered 96(3):171–184

Gehring WJ, Ikeo K (1999) Pax 6: mastering eye morphogenesis and eye evolution. Trends Genet

15(9):371–377

Kozmik Z et al (2003) Role of Pax genes in eye evolution: a cnidarian PaxB gene uniting Pax2 and

Pax6 functions. Dev Cell 5(5):773–785

Kozmik Z et al (2008) Assembly of the cnidarian camera-type eye from vertebrate-like compo-

nents. Proc Natl Acad Sci USA 105(26):8989–8993

Nilsson DE (2004) Eye evolution: a question of genetic promiscuity. Curr Opin Neurobiol 14(4):

407–414

Ogura A et al (2004) Comparative analysis of gene expression for convergent evolution of camera

eye between octopus and human. Genome Res 14(8):1555–1561

Salvivi-Plawen LV, Mayr E (1977) On the evolution of photoreceptors and eyes. Evol Biol

10:207–263

Serb JM, Eernisse DJ (2008) Charting evolution’s trajectory: using molluscan eye diversity to

understand parallel and convergent evolution. Evol Educ Outreach 1(4):439–447

Su KF et al (2007) Convergent evolution of eye ultrastructure and divergent evolution of vision-

mediated predatory behaviour in jumping spiders. J Evol Biol 20(4):1478–1489

Wistow G (2006) The NEIBank project for ocular genomics: data-mining gene expression in

human and rodent eye tissues. Prog Retin Eye Res 25(1):43–77

186 A. Ogura

Chapter 12

Do Long and Highly Conserved Noncoding

Sequences in Vertebrates Have Biological

Functions?

Yoichi Gondo

Abstract Vertebrate genomes consist of only a small fraction of protein-coding

sequences with vast majority of repetitive and nonrepetitive noncoding sequences.

Based on the completion of whole genome sequencing including human, it has

become possible to characterize the genomic structure directly at the DNA sequence

level.With the first approximation of the functional portion of the genome to be highly

evolutionary conserved, comparative genomics with bioinformatics and experimental

tools are now revealing the details of each element in the genome. In this chapter,

recent efforts to extract highly conserved sequences are reviewed with particularly

focusing on noncoding and nonrepetitive human and rodent genomes. Strikingly,

extracted highly conserved sequences in noncoding sequences exhibit much higher

conservation in many vertebrate genomes but not in other invertebrate species than

actually functional protein-coding sequences do. Some testable working hypotheses

to maintain such highly conserved sequences are also reviewed and discussed.

Abbreviations

LINE Long interspersed elements

SINE Short interspersed elements

UTR Untranslated region

SNP Single nucleotide polymorphism

CNG Conserved non-genic sequence

UCE Ultraconserved element

POLA DNA polymerase alpha catalytic subunit gene

LCNS Long conserved noncoding sequence

Y. Gondo

Mutagenesis and Genomics Team, RIKEN BioResource Center, 3-1-1 Koyadai, Tsukuba

305-0074, Japan



187

HNRNPD Heterogeneous nuclear ribonucleoprotein D

HNRPDL Heterogeneous nuclear ribonucleoprotein D-like

KO Knockout

12.1 Introduction

Most of higher eukaryotes contain noncoding sequences in the genome. Classically,

the DNA reassociation kinetics analyses by using the self-hybridization of frag-

mented genomic DNA, called Cot curve analysis, experimentally revealed that

significant portions of higher eukaryotes encompassed various types of repetitive

sequences (e.g., Britten and Kohne 1968; Wetmur and Davidson 1968).

The gene-coding sequences were also estimated by various methods including

RNA–DNA reassociation kinetics or Rot curve analysis. For instance, the complex-

ity of RNA expression was studied by RNA–DNA association kinetics (Chikaraishi

et al. 1978). They found that a unique fraction (31.2%) of rat genomic DNA was

found in nuclear RNA of the rat brain and exhibited the highest RNA complexity

among various tested rat tissues. Based on the average length of the rat nuclear

RNA (4,500 nucleotides) (Bantle and Hahn 1976) and finding that two-thirds (4,500

nucleotides) (1.9 Gb) of the rat genome are unique sequences, Chikaraishi et al.

(1978) estimated that the total number of rat gene was 130,000.

Based on the spontaneous mutagenesis studies of viability polygenes inDrosoph-ila melanogaster, Mukai (1978) suggested that most of the functional mutations

affecting viability polygenes occurred in noncoding sequences. He and others esti-

mated the spontaneous mutation rate of viability polygenes on the second chromo-

some ofD. melanogaster to be at least 0.14 per generation (Mukai 1964; Mukai et al.

1972; Ohnishi 1977). Estimating the number of the protein-coding genes on the

second chromosome to be 2,200 based on the “one-band one-gene hypothesis” (Judd

et al. 1972) and the average spontaneous mutation rate per protein-coding gene per

generation to be�10�5 or less, Mukai (1978) calculated the total mutation rate of the

protein-coding genes on the second chromosome to be 0.022 per generation, which

could explain at most 16% (¼022/0.14) of the mutations of viability polygenes.

Based on these considerations,Mukai (1978) concluded that most (>84%)mutations

affecting viability polygenes occurred in noncoding sequences.

Since the completion of the human genome project (International Human

Genome Sequencing Consortium 2001), it has become possible to identify the

genomic structure directly at the level of the DNA sequence. Bioinformatics and

various software programs have been developed not only to detect repetitive

sequences but also to identify and predict known and unknown protein-coding

sequences in whole genomic DNA sequences. Functional genomicists are now

working to reveal the biological functions of protein-coding as well as noncoding

sequences.

188 Y. Gondo

12.2 Genomic Structure of Human and Mammalian Genome

The initial analysis by the International Human Genome Sequencing Consortium

(2001) revealed that the interspersed repetitive sequences of long interspersed

elements (LINEs), short interspersed elements (SINEs), retrovirus-like elements,

and DNA transposon fossils occupied 21, 13, 8, and 3% of the human genomic DNA

sequences, respectively. In short, interspersed repetitive sequences comprise

45% of the human genome. In consideration of the amount of chromosomal dupli-

cations (3.6%) and other repetitive elements, repetitive sequences were expected to

occupy at least 50% of the human genome. Additionally, the total number of

protein-coding gene in the draft human genome sequence was estimated to be

30,000–35,000, with an average coding length of �1,400 bp (International Human

Genome Sequencing Consortium 2001).

After completing the euchromatic sequencing of the human genome, the Interna-

tional Human Genome Sequencing Consortium (2004) reported the size of the human

genome to be 3.08 Gb, with the total finished sequences of 2.85 Gb and estimated

total gaps of 0.23 Gb. They estimated�28Mb of the gaps as euchromatic, concluding

that the total euchromatic human genome was thus 2.88 Gb. Coding sequences were

estimated to be �1.2% of the euchromatic genome or �34 Mb in total. Based on the

number of protein-coding genes in the human genome (�25,000), the average length

of the protein-coding sequence was expected to be �1,400 bp.

Based on the subsequent DNA sequencing of whole genomes of mouse and other

mammals, mammalian genomes were found to have a similar structure to the

human genome (Fig. 12.1) (e.g., International Mouse Sequencing Consortium

2002). As a rough approximation, 1.2–1.5% coding sequences in which

C

R1

R2

R3

R4R5

N

Fig. 12.1 Composition of human genome deduced from the analysis of whole genomic DNA

sequence. International Human Genome Sequencing Consortium (2001) depicted that the portion

of protein-coding sequences (C) is only 1.2–1.5% of the genome. Roughly a half of the genomic

sequences are various classes of repetitive sequences: R1, LINEs; R2, SINEs; R3, retrovirus-like

elements; R4, DNA transposon fossils and R5, other repetitive elements. Another �half of the

genome consists of noncoding and nonrepetitive sequences (N)

12 Do Long and Highly Conserved Noncoding Sequences in Vertebrates 189

approximately 25,000–30,000 genes are coded in the 3 Gb of the mammalian

genome, although the size of the euchromatic mouse genome was estimated

to be �2.5 Gb and significantly smaller (International Mouse Sequencing Consor-

tium 2002).

12.3 Noncoding Sequences in Mammalian Genome

Most (98–99%) of the mammalian genomic sequence is, therefore, noncoding.

Repetitive and nonrepetitive sequences each occupy roughly half of the mammalian

genome. The biological functions of repetitive as well as non-repetitive noncoding

sequences in the genome remain to be elucidated. As an alternative, the “junk DNA”

hypothesis has been raised that most of the noncoding DNAs may not have any

biological functions (e.g., Nowak 1994). Nonfunctional genomic DNA sequences are

expected to have a fast base-substitution rate in the course of evolution due to a lack

of evolutionary constraints that would eliminate detrimental base substitutions. For

instance, genomic noncoding sequences such as pseudogenes, introns, untranslated

regions (UTR) of mRNA, and intergenic sequences, assumed to be less functional

sequences than protein-coding sequences, usually have more single nucleotide poly-

morphisms (SNPs) within species and exhibit higher divergence between the homo-

logs among various species than the protein-coding sequences do. In turn, the degree

of homology detected by aligning the syntenic sequences between different species

has been used to find protein-coding sequences and functional regulatory elements

(reviewed by O’Brien et al. 1999). For instance, sequences that are more than 80%

similar between human and rodents are empirically recognized as good candidates for

protein-coding sequences and/or functionally constrained parts of the genome. The

overall similarity between human and mouse genome was estimated to be 66.7%

(International Mouse Sequencing Consortium 2002).

12.4 Conserved Noncoding Sequences

The human genome project and following whole genome sequencing of various

species have also allowed us to conduct more precise alignment and comparison of

genomic DNA sequences. Also, various conserved sequences have been identified

not only in protein-coding sequences but also in noncoding fractions of the

genome. For the first time, the capacity to search such highly conserved noncoding

sequences in a large scale between vertebrate species became available when the

whole mouse genomic sequence was released. The initial publication of the whole

mouse genome sequence (Mouse Genome Sequencing Consortium 2002)

described the comparison of whole genomic sequences between human and

mouse. Approximately, 5% of the 50–100 bp windows in the human genome

was conserved in the mouse genome. Since protein-coding sequences comprise

1.5% of the genome at most, the noncoding portion of the identified conserved

190 Y. Gondo

sequences is 3.5% or more in the human genome. Dermitzakis et al. (2002) in the

same issue of Nature reported more extensive searches focusing on the human

chromosome 21. They searched �100 bp with �70% identical sequences between

human chromosome 21 and syntenic mouse sequences after masking the repetitive

sequences. By further eliminating known coding sequences, Dermitzakis et al.

(2002) finally obtained 2,262 of conserved nongenic sequences (CNGs). They

further analyzed 220 CNG in 20 mammalian species and found that CNGs are

more conserved than protein-coding sequences and noncoding RNAs (Dermitzakis

et al. 2003). Indeed, approximately 80% similarity in the protein-coding genomic

sequences is high enough to keep the 100% identity of the amino acid sequence in

the protein. Because of the degeneracy of the genetic code, base-substitutions

between purines (A<>G) and pyrimidines (T<>C) do not change the coded

amino acid residue except three cases out of the 64 codons. Thus, to have more

than 90% homologies in any two sequences in a significant stretch, there must be

some mechanism(s) to maintain or create such long conserved sequences other

than to maintain the protein function.

12.4.1 Ultraconserved Elements

Bejerano et al. (2004) expanded the extraction of such highly conserved sequences

to the whole genomes of human, mouse and rat with more stringent condition of

�200 bp in length with 100% identity. They found 481 ultraconserved elements

(UCEs). Most of the UCEs are also conserved in many vertebrate species. For

instance, 477, 467, and 324 UCEs exhibited averages of 99.2, 95.7, and 76.8%

identities in dog, chicken, and fugu fish genomes, respectively (Bejerano et al.

2004). The distribution frequencies of SNPs were also analyzed within human and

chimp populations. Comparing to the distribution frequencies of SNPs in entire

genomes, Bejerano et al. (2004) found 20-fold fewer SNPs in the UCEs from both

genomes.

Bejerano et al. (2004) also analyzed the 481 UCEs with respect to their genomic

neighborhood. UCEs were found in both exons (111) and non-exons (256). The

remaining 114 UCEs were unknown in terms of this definition. Nonexonic UCEs

are further classified to 100 intronic UCEs and 156 intergenic UCEs. Thus, at least,

211 UCEs, which were exonic and intronic, were clearly transcribed. Nonexonic

UCEs tended to cluster in gene deserts. Further analysis of UCE locations suggested

that exonic UCEs were likely to exist close to known RNA regulating genes

whereas intergenic nonexonic UCEs had a tendency to flank genes for regulation

of transcription, DNA binding, and development. Intronic UCEs were also often

found in development-related genes. The longest UCEs (779, 770, and 731 bp) were

clustered in introns of the DNA polymerase alpha catalytic subunit gene (POLA)on the X chromosome. Particularly, the 779 bp UCE was adjacent to another

275 bp UCE, comprising a total of 1,046 bp highly conserved sequence (Bejerano

et al. 2004).


12.4.2 Long Conserved Noncoding Sequences

12.4.2.1 Discovery of LCNS

Just after the completion of whole mouse genome sequencing (Mouse Genome

Sequencing Consortium 2002), we have independently started genomewide extrac-

tions of highly conserved noncoding sequences between human and mouse (Sakuraba

et al. 2008). We firstly masked not only repetitive sequences but also all the protein-

coding sequences from human and mouse genomes, thereby extracting only, non-

coding and nonrepetitive portion of each genome. Then highly homologous

sequences of�500 bp in length and�95% identity were extracted by BLAST search

between the human and mouse nonrepetitive noncoding sequences. The human and

mouse genomic sequence databases have been updated many times. In response, we

conducted the extraction of highly conserved noncoding/nonrepetitive sequences

three times during 2002–2007 (Sakuraba et al. 2008). A total of 611 long conserved

noncoding sequences (LNCS) were found. We did not consider synteny when we

conducted the extraction of LNCS. Nevertheless, the LCNS pairs were syntenic

(Sakuraba et al. 2008).

12.4.2.2 Similarity among 611 LCNS

In spite of the repeat masking, minor duplications may exist in the extracted LCNS.

We conducted BLAST search for each LCNS to the other and found that six LCNS

pairs exhibited some similarity (Sakuraba et al. 2008). The result is summarized in

Table 12.1. Four pairs (LCNS 504 and 719, LCNS744 and 767, LCNS 596 and 835

and LCNS 541 and 788) had 85–92.2% similarity but were rather short stretch of

84–294 bp. Each LCNS of the 4 pairs was located on separate chromosomes.

The LCNS501 and 503 pair were very similar (90.5%) in short length (63 bp) but

they were located very close to each other on the same chromosome in human

(chromosome 4) as well as in mouse (chromosome 5). LCNS501 and 503 were

found in an intron of the heterogeneous nuclear ribonucleoprotein D (HNRNPD)and heterogeneous nuclear ribonucleoprotein D-like (HNRPDL) genes, respec-

tively. Thus, LCNS501 and 503 seemed to be a part of the intrachromosomal

duplication around the HNRNPD and HNRPDL genes. Since this structure is the

same in the mouse (Hnrnpd gene and Hnrpdl genes on mouse chromosome 5), the

duplication for these paralogs seems to have occurred before the divergence

between human and mouse lineages. Then, mutations may have accumulated to

reduce the similarity to 90.5% in 65 bp. However, the orthologs are extremely

similar (95.7% and 97.1%) in very long stretch of 580 and 686 bp in LCNS501 and

503, respectively (Table 12.1) so they were extracted as one LCNS. It may be

possible to explain the huge difference in similarity between orthologs and paralogs

as follows. The intrachromosomal duplication around the ancestral HNRNPD gene

occurred long time before the common ancestor of human and mouse appeared and

the paralogous sequences accumulated many mutations. On the other hand,

192 Y. Gondo

Table

12.1

Six

pairs

ofLCNSwithsequence

similarity

Conservation

Length

aSim

ilarity

Human

(hg18)

Mouse

(mm9)

LCNSI.D.

Length

a%

Identity

Chr.

Start

End

Chr.

Start

End

242

504

98.6

133

85.0%

chr8

37,357,379

37,357,882

chr8

27,787,978

27,788,481

352

719

95.3

chr10

77,076,129

77,076,843

chr14

23,017,453

23,018,171

259

744

97.8

294

92.2%

chr16

53,780,778

53,781,520

chr8

95,069,373

95,070,116

580

767

98.0

chr5

3,565,400

3,566,166

chr13

72,166,145

72,166,910

269

596

99.7

84

90.5%

chr3

138,466,129

138,466,724

chr9

100,171,994

100,172,686

610

835

95.4

chr13

94,416,647

94,417,478

chr14

118,834,842

118,835,707

289

541

95.4

104

90.5%

chr2

58,711,520

58,712,060

chr11

25,908,519

25,909,051

340

788

95.3

chr14

96,500,728

96,501,514

chr12

107,367,078

107,367,862

501

580

95.7

63

90.5%

chr4

83,494,858

83,495,432

chr5

100,390,903

100,391,482

503

686

97.1

chr4

83,565,412

83,566,096

chr5

100,464,327

100,465,010

522

1,023

95.9

1,019

90.5%

chrX

41,093,072

41,094,094

chr6

102,886,532

102,887,550

654

1,061

97.2

chrX

41,093,034

41,094,094

chrX

12,869,761

12,870,815

Six

pairs

outof611LCNShad

somesequence

similarityandareshownin

pairw

ise

aLength

isin

bp


mutations were hardly accumulated in these syntenic regions after the human and

mouse speciated so that the LCNS501 and 503 were highly conserved. It, however,

does not explain why only the LCNS501 and 503 were conserved but the flanking

syntenic regions had diverged between human and mouse. It may be because of the

evolutionary constraint against very functional sequences of LCNS501 and 503 and

will be discussed furthermore in Sect. 12.5.

The LCNS522 and 654 were another peculiar pair. They showed much longer

homology of 1,019 bp with 90.5% similarity compared with the other five LCNS

pairs (Table 12.1). LCNS 522 and 654 located on the different chromosomes in the

mouse but they were almost the same stretch on human X chromosome. In other

words, LCNS522 and 654 were one identical sequence in human but duplicated

interchromosomally in the mouse genome. Thus, strictly speaking, human and

mouse have 610 and 611 LCNS, respectively.

12.4.2.3 Comparison of LCNS with UCE

We compared the contents of 611 LCNS with the 481 UCE (Bejerano et al. 2004),

since the total extracted numbers of the conserved elements in the two independent

studies were quite similar in spite of different extraction criteria. The result is

summarized in Fig. 12.2. As depicted, 138 (23%) LCNS and 150 (31%) of UCE

overlap. LCNS are longer than UCEs by definition and 12 LCNS indeed encom-

passed two different UCEs. Another new set of 473 LCNS, independent from the

UCEs was, therefore, found (Sakuraba et al. 2008); Bejerano et al. (2004) extracted

the 481 UCEs with whole genome comparison including protein-coding as well as

repetitive sequences. Thus, 69 and 9 UCE overlap protein-coding and repetitive

sequences, respectively (Fig. 12.2), which were naturally different from the 611

LCNS. In addition, the 138 LCNS that contained one or two UCEs had extra

sequences that did not overlapped any UCEs. Such nonoverlapping portions of

the 138 LCNS to UCE were also newly identified highly conserved noncoding/

nonrepetitive sequences. It may be noteworthy that no UCEs were identified on

human chromosome 21 (Bejerano et al. 2004); on the other hand, we found three

473138

253

611 LCNS481 UCE

150

699

Fig. 12.2 Sequence

comparison between 611

LCNS and 481 UCE. In

addition to the 481 UCE, 473

additional LCNS were

discovered as highly

conserved elements in

vertebrates. NonUCE-

overlapping sequences in 138

LCNS that contained a part of

a UCE are also newly added

highly conserved sequences

194 Y. Gondo

LCNS on human chromosome 21 in the syntenic region to mouse chromosome 16

(see Supplementary Table 1 of Sakuraba et al. 2008).

12.4.2.4 Length, Identity, and Location of LCNS

Some characteristics of LCNS length and identity are summarized in Tables 12.2

and 12.3. Naturally, the length of LCNS was much longer than UCE. Table 12.2

depicted top 20 largest and 19 shortest LCNS whose mean identities were 96.1% in

both. The longest LCNS146 was 1,865 bp with 95.1% identity, barely satisfying the

similarity criterion. The second longest LCNS572, however, exhibited 98.0%

identity in the stretch of 1,768 bp. Forty-five LCNS were longer than 1,000 bp.

The mean and median lengths of 611 LCNS were 685 and 636 bp, respectively. The

20 most and least similar LCNS are also described in Table 12.3, the mean lengths

of which were 617 and 627 bp, respectively. The mean and median identities of 611

LCNS are 95.6 and 96.2%, respectively.

The locations of LCNS were classified as UTR, intronic, or intergenic

(Table 12.4 and Sakuraba et al. 2008) as done by Bejerano et al. (2004). We

eliminated protein-coding exons but kept UTR and still 22 (3.6%) of LCNS were

found in 50 or 30 UTRs. A large fraction of LCNS located in intron and more than

half (55.3%) were found in intergenic region often distant from nearby genes. Two

hundred and seventy nine LCNS were more than 10 kb distant from the closest

gene. In spite of the intrinsic differences of length and homology, overall location

was quite similar between LCNS and UCE. The LCNS were also clustered as the

case for UCE (see Fig. 1 of Sakuraba et al. 2008). As shown above, LCNS located

all chromosomes including human chromosome 21 except Y chromosome. The

distribution of LCNS varies. For instance, human chromosome 2 carries more than

2-folds of LCNS than average. Mouse chromosome 2 that is mostly syntenic to

human chromosome 2 also had 2 - 3 folds enriched with LCNS than average.

Contrarily, the number of LCNS on human chromosome 12, 21, 22, and Y and

mouse chromosome 10, 15, 16, and Y were extremely underrepresented (Sakuraba

et al. 2008).

Another key similarity of LCNS to UCE was the degree of conservation in the

other species. The 611 LCNS were surveyed in genomic DNA database of dog,

chicken, frog, fugu, tetraodon, and zebrafish in which 606, 493, 397, 82, 58, and 83

LCNS were identified, respectively. Three invertebrate species (Ciona intestinalis,Ciona savignyi, and Drosophila melanogaster) were also surveyed but no LCNS

homologs were found. Interestingly, the degree of conservation of LCNS in verte-

brate species was more or less inversely proportional to the evolutionary distance.

Not only the number of LCNS identified in the six vertebrate species but also the

average identities were negatively correlated to the evolutionary distance. The

average identities of LCNS located in dog, chicken, frog, fugu, tetraodon, and

zebrafish genomes were 95.6%, 94.1%, 91.6%, 90.8%, 90.9%, and 90.8%, respec-

tively (Sakuraba et al. 2008).


Table

12.2

Twenty

longestandnineteenshortestLCNS

I.D.

Conservation

Human

(hg18)

Mouse

(mm9)

Length

a%

Identity

Chr.

Start

End

Chr.

Start

End

Location

Distance

b

LCNS146

1,865

95.1

chr6

99,531,674

99,533,530

chr4

22,228,018

22,229,881

Intergene

>10kb

LCNS572

1,768

98.0

chr14

25,985,222

25,986,988

chr12

47,799,178

47,800,939

30 U

TR

LCNS230

1,722

96.9

chr19

35,532,814

35,534,534

chr7

38,435,408

38,437,128

Intron

LCNS076

1,548

95.5

chr2

156,292,959

156,294,493

chr2

56,308,034

56,309,578

Intergene

>100kb

LCNS033

1,473

95.1

chr10

23,526,498

23,527,964

chr2

19,372,376

19,373,845

Intergene

�10kb

LCNS200

1,436

95.1

chr1

91,071,304

91,072,739

chr5

106,987,921

106,989,346

Intergene

>10kb

LCNS557

1,359

96.2

chr6

86,377,974

86,379,319

chr9

88,347,742

88,349,092

30 U

TR

LCNS440

1,291

97.4

chrX

24,825,753

24,827,039

chrX

90,654,218

90,655,508

Intron

>10kb

LCNS577

1,282

95.2

chr14

25,983,926

25,985,203

chr12

47,797,882

47,799,156

30 U

TR

LCNS334

1,257

95.1

chr14

35,884,301

35,885,557

chr12

57,488,091

57,489,341

Intergene

>10kb

LCNS478

1,253

97.2

chr3

159,508,610

159,509,860

chr3

66,995,021

66,996,272

Intron

>10kb

LCNS050

1,250

96.2

chr2

143,820,605

143,821,853

chr2

43,833,513

43,834,757

Intron

>10kb

LCNS583

1,232

96.4

chr5

91,478,639

91,479,869

chr13

80,139,355

80,140,585

Intergene

>100kb

LCNS639

1,219

95.1

chr10

103,201,169

103,202,380

chr19

45,520,854

45,522,070

Intron

>10kb

LCNS111

1,213

95.4

chr2

176,462,362

176,463,573

chr2

74,328,073

74,329,285

Intergene

>10kb

LCNS482

1,202

97.6

chr3

159,258,850

159,260,050

chr3

66,739,049

66,740,249

Intergene

>10kb

LCNS632

1,195

96.4

chr18

70,637,515

70,638,707

chr18

84,583,528

84,584,719

Intron

>10kb

LCNS316

1,185

96.6

chr14

28,928,655

28,929,832

chr12

51,220,916

51,222,097

Intergene

>100kb

LCNS474

1,180

96.4

chr1

97,051,832

97,053,011

chr3

119,421,838

119,423,010

Intergene

�10kb

LCNS364

1,162

95.7

chr14

56,126,548

56,127,704

chr14

49,075,805

49,076,962

Intron

LCNS181

503

95.2

chr1

32,282,701

32,283,345

chr4

129,390,856

129,391,399

Intron

LCNS228

503

95.6

chr19

35,724,274

35,724,775

chr7

38,271,355

38,271,857

Intron

LCNS062

503

96.8

chr2

144,978,686

144,979,187

chr2

44,952,525

44,953,237

Intron

>10kb

LCNS038

503

95.8

chr9

127,775,210

127,775,707

chr2

34,022,608

34,023,110

Intergene

�10kb

196 Y. Gondo

LCNS426

503

95.6

chrX

39,829,216

39,829,717

chrX

11,645,035

11,645,535

Intron

LCNS358

502

96.2

chr10

77,543,699

77,544,199

chr14

23,468,812

23,469,312

Intron

>10kb

LCNS411

502

96.0

chr18

43,323,216

43,323,716

chr18

76,812,642

76,813,143

Intergene

>100kb

LCNS113

502

95.4

chr2

177,211,340

177,211,838

chr2

74,976,214

74,976,715

Intergene

>10kb

LCNS114

502

95.8

chr2

177,393,500

177,394,001

chr2

75,155,446

75,155,947

Intergene

>100kb

LCNS153

502

98.8

chr6

97,651,729

97,652,230

chr4

24,596,593

24,597,094

Intron

>10kb

LCNS424

502

95.2

chr9

75,991,328

75,991,829

chr19

19,511,315

19,511,816

Intergene

>100kb

LCNS433

502

95.2

chrX

147,900,026

147,900,525

chrX

67,145,332

67,145,833

Intergene

>10kb

LCNS295

501

96.2

chr17

32,268,766

32,269,265

chr11

84,424,428

84,424,928

Intergene

>10kb

LCNS403

501

99.2

chr18

22,169,478

22,169,978

chr18

15,003,269

15,003,769

Intron

LCNS291

501

95.0

chr2

57,958,773

57,959,273

chr11

26,714,982

26,715,482

Intergene

>10kb

LCNS277

501

96.2

chr2

60,709,281

60,709,781

chr11

23,915,928

23,916,426

Intergene

>10kb

LCNS197

501

95.8

chr4

85,619,277

85,619,777

chr5

102,074,332

102,074,830

Intergene

>10kb

LCNS264

500

96.2

chr11

115,737,826

115,738,325

chr9

46,468,340

46,468,838

Intergene

>10kb

LCNS035

500

96.0

chr9

127,959,649

127,960,155

chr2

33,888,639

33,889,144

Intergene

>100kb

aLength

isin

bp

bDistance

from

thenearbyprotein-codingsequence

inthemouse

genome


Table

12.3

Twenty

mostandleastsimilar

LCNS

I.D.

Conservation

Human

(hg18)

Mouse

(mm9)

Length

a%

Identity

Chr.

Start

End

Chr.

Start

End

Location

Distance

b

LCNS438

962

99.8

chrX

24,918,245

24,919,206

chrX

90,555,381

90,556,342

Intron

LCNS269

596

99.7

chr3

138,466,129

138,466,724

chr9

100,171,994

100,172,686

Intergene

>100kb

LCNS441

819

99.5

chrX

24,804,732

24,805,549

chrX

90,674,418

90,675,236

Intron

LCNS637

667

99.3

chr10

102,437,335

102,438,001

chr19

44,773,397

44,774,063

Intergene

>10kb

LCNS344

785

99.2

chr7

20,970,118

20,970,902

chr12

119,958,510

119,959,293

Intergene

>100kb

LCNS403

501

99.2

chr18

22,169,478

22,169,978

chr18

15,003,269

15,003,769

Intron

LCNS414

581

99.1

chr18

43,024,590

43,025,170

chr18

77,101,560

77,102,140

Intron

LCNS103

557

99.1

chr2

174,904,641

174,905,197

chr2

73,106,824

73,107,380

Intergene

�10kb

LCNS592

551

99.1

chr5

77,183,641

77,184,191

chr13

95,472,039

95,472,588

Intergene

>10kb

LCNS400

559

98.9

chr18

20,946,991

20,947,549

chr18

13,897,326

13,897,883

Intron

>10kb

LCNS152

538

98.9

chr6

97,769,812

97,770,349

chr4

24,471,724

24,472,261

Intron

LCNS477

616

98.9

chr3

181,919,488

181,920,103

chr3

33,781,632

33,782,247

Intergene

>10kb

LCNS472

525

98.9

chr9

134,485,104

134,485,628

chr2

28,740,382

28,740,906

Intron

LCNS039

516

98.8

chr9

127,696,600

127,697,115

chr2

34,100,300

34,100,813

Intron

>10kb

LCNS153

502

98.8

chr6

97,651,729

97,652,230

chr4

24,596,593

24,597,094

Intron

>10kb

LCNS506

739

98.8

chr7

114,117,479

114,118,217

chr6

15,388,331

15,389,069

30 U

TR

LCNS640

550

98.7

chr10

102,405,068

102,405,616

chr19

44,745,421

44,745,970

Intergene

>10kb

LCNS634

603

98.7

chr5

139,475,193

139,475,792

chr18

36,448,133

36,448,659

Intergene

�10kb

LCNS585

664

98.6

chr5

81,183,117

81,183,780

chr13

91,510,428

91,511,091

Intergene

>10kb

LCNS242

504

98.6

chr8

37,357,379

37,357,882

chr8

27,787,978

27,788,481

Intergene

>100kb

LCNS620

627

95.1

chr2

44,024,538

44,025,163

chr17

85,148,509

85,149,135

Intron

LCNS249

546

95.1

chr16

49,663,946

49,664,490

chr8

91,497,355

91,497,899

Intergene

>10kb

LCNS328

586

95.1

chr14

33,182,370

33,182,955

chr12

55,017,916

55,018,500

Intron

>10kb

LCNS342

586

95.1

chr14

98,953,026

98,953,611

chr12

109,360,726

109,361,309

Intron

198 Y. Gondo

LCNS361

889

95.1

chr10

78,060,662

78,061,550

chr14

23,913,918

23,914,806

Intergene

�10kb

LCNS140

786

95.0

chr8

59,976,155

59,976,939

chr4

6,717,506

6,718,290

Intron

>10kb

LCNS350

583

95.0

chr6

1,723,057

1,723,639

chr13

32,062,621

32,063,203

Intron

>10kb

LCNS280

603

95.0

chr2

60,370,645

60,371,246

chr11

24,212,663

24,213,264

Intergene

>100kb

LCNS381

723

95.0

chr13

99,409,271

99,409,993

chr14

122,852,051

122,852,772

Intergene

�10kb

LCNS406

522

95.0

chr18

35,155,674

35,156,195

chr18

27,787,891

27,788,411

Intergene

>10kb

LCNS072

522

95.0

chr2

147,006,391

147,006,912

chr2

47,158,094

47,158,615

Intergene

>100kb

LCNS442

522

95.0

chrX

71,478,279

71,478,798

chrX

99,489,220

99,489,740

Intergene

>10kb

LCNS067

803

95.0

chr2

146,405,657

146,406,459

chr2

46,521,971

46,522,769

Intergene

>100kb

LCNS057

542

95.0

chr2

144,471,545

144,472,086

chr2

44,480,335

44,480,871

Intron

>10kb

LCNS432

642

95.0

chrX

147,827,152

147,827,793

chrX

67,071,319

67,071,960

Intron

>10kb

LCNS291

501

95.0

chr2

57,958,773

57,959,273

chr11

26,714,982

26,715,482

Intergene

>10kb

LCNS088

601

95.0

chr2

163,813,314

163,813,913

chr2

63,413,486

63,414,085

Intergene

>100kb

LCNS549

621

95.0

chr15

65,691,994

65,692,614

chr9

63,159,161

63,159,778

Intron

LCNS621

661

95.0

chr2

44,661,079

44,661,739

chr17

85,694,024

85,694,683

Intron

>10kb

LCNS255

680

95.0

chr16

50,492,105

50,492,780

chr8

92,233,969

92,234,648

Intergene

>100kb

aLength

isin

bp

bDistance

from

thenearbyprotein-codingsequence

inthemouse

genome


12.5 Working Hypotheses for Genomic Sequence Conservation

To understand the biological function(s) of the highly conserved noncoding

sequences, it is necessary to consider plausible mechanisms of making conserved

sequences in many different species. Four working hypotheses that would create

and/or maintain highly conserved sequences in coding sequences as well as in

noncoding sequences will be discussed below. These working hypotheses are not

exclusive to each other. Two or more combinations of plausible mechanisms may

contribute to maintain the conservation of genomic sequences among various

species. Among the four working hypotheses, only the first one (Sect. 12.5.1)

requires significant biological function for the maintenance of conserved sequences

whereas the other three hypotheses do not necessary need such functions to explain

the highly identical sequences among various species.

12.5.1 Functional Constraint

As described above, the primary working hypothesis for maintenance of LCNS is

that evolutionary constraints keep the functionally important genomic DNA

sequence from changing. Such functional genomic sequences may be protected

from accumulation of spontaneous and/or induced mutations by natural selection.

Mutations usually disrupt and disturb the normal function of the gene (or genomic

sequence), since the nature of the mutation is random in terms of base-pair array in

the genome. It is why radiations, chemical mutagens, and other genotoxic agents

are usually harmful to biology and cause various genetic disorders including

tumorigenesis, genetic diseases, and predispositions of various genetic risk factors

to individuals. Such detrimental mutations are eliminated from natural populations

by Darwinian selection. Thus, having more significant function, a genomic

sequence tends to exhibit higher degree of conservation among various species

due to the evolutionary constraints.

To directly test this hypothesis, in vivo assay has been conducted (Poulin et al.

2005; Pennacchio et al. 2006; Visel et al. 2008). By using a transgenic mouse

Table 12.4 Summary of

LCNS locationsUTR 50 6 1.0% 3.6%

30 16 2.6%

Intron >100 kb 3 0.5% 41.1%

>10 kb 119 19.5%

�10 kb 129 21.1%

Intergenic >100 kb 147 24.1% 55.3%

>10 kb 132 21.6%

�10 kb 59 9.7%

Total 611

The location of 611 LCNS are classified to one of eight cate-

gories based on the distance from the nearby protein-coding

sequence in the mouse genome

200 Y. Gondo

enhancer assay with reporter genes, highly conserved elements have been experi-

mentally examined of their enhancer cis-regulatory activity. For instance, Pennacchioet al. (2006) tested 167 highly conserved sequences and found that 45% of the

sequences had tissue-specific cis-regulatory function at mouse embryonic day 11.5.

Furthermore, Visel et al. (2008) compared such enhancer activities between UCE

and highly conserved but not in 100% identity sequences by using the transgenic

approach. They confirmed the enhancer activity not only in UCE but also in the

other highly conserved sequences, suggesting UCE may be a part of a larger

enhancer family in the genome.

Derti et al. (2006) proposed another possible function of UCE. They proposed

that the UCE and/or flanking sequences might maintain the diploid karyotype by the

dosage sensitivity. Mammalian UCEs are highly depleted among segmental dupli-

cations and copy number variants. This hypothesis seems to be concordant with the

fact that UCEs were not found on Y chromosome, human chromosome 21, or in the

syntenic regions of the mouse genome. The Y chromosome is only the only non-

diploid region in mammals. Human chromosome 21, in which trisomy causes

Down syndrome, might be less tolerant of diploid constraint. We, however, found

three LCNS on human chromosome 21 and the syntenic region in the mouse

(Sakuraba et al. 2008 and Sect. 12.4.2.3).

Knockout (KO) mouse studies of UCEs have raised controversial findings

related to the functional constraint hypothesis. For instance, Ahituv et al. (2007)

disrupted four UCE independently and analyzed the KO mice. None of four KO

mouse strains exhibited any anomalies, indicating such UCE should be dispensable.

Then, McLean and Bejerano (2008) found that ultraconserved-like elements were

over 300-fold less likely than neutral DNA to have been lost during rodent evolu-

tion. If UCEs are dispensable, then they should have been lost from the population,

similar to neutral sequences. The mutagenesis analysis of highly conserved

sequences is also discussed in Sect. 12.5.2.

12.5.2 Mutational Cold Spots

If a genomic sequence is a mutational cold spot, meaning little or no mutation

occurs in a sequence, such a genomic sequence might keep the same array of base

pairs in many generations and consequently conserved in many different species.

Since many mutagens directly target genomic DNA sequences to modify or break

down DNA molecules, tightly packed chromatin structure, e.g., in heterochromatic

regions, prevent the mutagen from attacking DNA molecules, resulting in a void of

the accumulation of mutations. Alternatively (or together), an enhanced DNA

repair system in particular genomic sequences would be another mechanism to

give rise to mutational cold spots. Whatever the mechanisms of making mutational

cold spots would be, if they exist in the genome, they would be highly conserved

portions of the genome.


Bejerano et al. (2004) found much less but some SNPs in human UCE than

average. Thus, some mutations have occurred in UCEs. Several analyses of geno-

type data in human SNP projects (Drake et al. 2006; Katzman et al. 2007) indirectly

suggested that UCE and highly conserved sequences were not mutational cold

spots. We, therefore, experimentally tested if LCNS are mutational cold spots by

using ENU mutagenesis (Sakuraba et al. 2008). We have produced �10,000 ENU-

mutagenized G1 mice and extracted each DNA (Sakuraba et al. 2005). By using a

high-throughput mutation discovery system combining PCR amplification and

heteroduplex detection (Sakuraba et al. 2005), several LCNS as well as non-

LCNS were subjected to detect ENU-induced mutations (Sakuraba et al. 2008).

We found 12 and 136 ENU-induced mutations by screening a total of 16.5 and

181.0 Mb of LCNS and nonLCNS, respectively. Thus, ENU-mutations were found

one in 1.371 Mb and in 1.331 Mb of LCNS and nonLCNS, respectively. This very

equivalent ENU-induced mutation frequency was also reproduced in a new

enhanced mutation discovery system, in which we found 23 and 207 ENU-induced

mutations by screening 24.2 and 223.9 Mb of LCNS and nonLCNS, respectively

(Sakuraba et al. 2008). Thus, the mutational cold spot hypothesis is unlikely to

explain the maintenance of highly conserved sequences in vertebrates during

evolution.

All the G1 mice that were examined for the ENU mutagenesis study above were

maintained as frozen sperm (Sakuraba et al. 2005); therefore, it is possible to

analyze live mice carrying an ENU-induced mutation in the LCNS. The total of

35 mouse strains carrying an ENU-induced mutation in an LCNS are listed in our

WEB site (http://www.brc.riken.go.jp/lab/mutants/genedriven.htm) and freely

available based upon request to RIKEN BioResource Center (BRC) (http://www.

brc.riken.jp/lab/animal/en/depo.shtml).

12.5.3 Horizontal Transfer

Another mechanism to make a highly conserved genomic sequence among various

species is a recent event of DNA transfer from one species to the other. Interspecies-

active transposition and retroposition would be a plausible mechanism. If a DNA

segment horizontally transferred to many species at one time very recently, the

transmitted portion of the genomic DNA would have the very similar sequences in

the affected species. One discrepancy is that the horizontal transfer by transposon,

for instance, usually gives rise to multiple copies in the genome, comprising a part of

repetitive sequences. Also, if horizontal transfer happened very recently, the degree

of conservation should not be inversely proportional to the evolutionary distance. As

described in Sect. 12.4.2.4, however, the degree of the LCNS conservation was

inversely proportional to the evolutionary distance. A simple transposon hypothesis

does not explain syntenic localization of UCE and LCNS pairs in human and mouse.

A combination of functional constraint and horizontal transfer may have

occurred. At the beginning of adaptive radiation of vertebrate species, horizontal

202 Y. Gondo

http://(http://www.brc.riken.go.jp/lab/mutants/genedriven.htm)

http://(http://www.brc.riken.jp/lab/animal/en/depo.shtml).

http://(http://www.brc.riken.jp/lab/animal/en/depo.shtml).

transfer might have been very active via various transposons and spread out to

many radiated ancestors of vertebrates. If the transposons had been originated not

from the direct ancestor species but from e.g., fungi, viruses, and/or bacteria, it is

reasonable that neither UCE nor LCNS would be found in any invertebrate species.

In this model, various sequences could have been horizontally transmitted to

various loci in the genome of many vertebrate ancestors. Then bottleneck and

founder effects reduced the number of ancestors and a few lineages furthermore

may have undergone adaptive radiations. Each lineage, then, would maintain the

syntenic localization of highly conserved sequences like UCEs and LCNS in

human, mouse, and rat. Functional constraints might have been maintaining only

the highly conserved sequences like UCE and LCNS but flanking sequences

diversified. Bejerano et al. (2006) showed some evidence of retroposon-like origins

of UCEs.

12.5.4 Concerted Evolution and Gene Conversion

Some portions of genomic sequences have been homogenized to the identical or

similar sequences, resulting in the concerted evolution (Nenoi et al. 1998; Gondo

et al. 1998; Nei et al. 2000; Okada Y, Gondo Y, Ikeda JE, unpublished). An example

has been found in the genomic sequences of the ubiquitin gene among very diversi-

fied species of fungi, plants, and animals including human. Gene families code

ubiquitin and head-to-tail tandem structure and unequal crossing over seems to

maintain the identical genomic DNA sequence of the poly-ubiquitin gene (Nenoi

et al. 1998; Nei et al. 2000). A deubiquitinase gene coding for USP17 in human

(Gondo et al. 1998; Saitoh et al. 2000) was also found to be very conserved among

tested mammalian species (Gondo et al. 1998; Okada Y, Gondo Y, Ikeda JE

unpublished). The USP17 gene was found on human chromosomes 4 and 8 with

50–100 head-to-tail tandem copies and a few copies, respectively (Gondo et al. 1998;

Okada et al. 2002). The USP17 gene was also identified in many mammalian species

in head-to-tandem repeat structure except in the mouse (Gondo et al. 1998; Okada Y,

Gondo Y. Ikeda JE unpublished). The copy numbers on human chromosome 4 were

highly polymorphic (Gondo et al. 1998) but the 4.7 kb unit sequence of the USP17gene with the flanking sequences was very identical between copies (�99%). The

degree of homology (> 99%) between the 4.7 kb repeating units was at the level of

the UCE and LCNS. The extremely high similarity was found not only within the

tandem repeat on the chromosome 4 but also in a few copies on the chromosome 8.

Thus, simple unequal crossing over to homogenize the unit sequence may not be

enough to explain the highly conserved 4.7-kb sequences in human and other

mammalian species. Some unknown gene-conversion mechanism might have homo-

genized the 4.7-kb unit sequences between the tandemly repeated sequences on

chromosome 4 as well as between unit sequences on chromosome 4 and 8. If the

homogenization mechanism of the ubiquitin and the 4.7-kb unit including the USP17


gene is revealed, it might provide another working hypothesis to give rise to highly

conserved sequences.

12.6 Conclusions

Highly conserved sequences have been found in vertebrates. The rich accumulation

of the knowledge of highly conserved sequences in vertebrates raises various

questions and working hypotheses. The answers, however, are yet to be determined.

One of the most critical issues in this field of study is the lack of highly conserved

sequences like UCE and LCNS in invertebrate species. Invertebrates may have their

own highly conserved sequences. It is necessary to survey in the other clade if some

other classes of highly conserved sequences exist. The horizontal transfer hypothe-

sis emphasizes the importance of genomic sequence data not only from species that

are closely related to vertebrates but also from more distantly related organisms

including fungi, bacteria, and viruses. Even metagenomics of lower eukaryotes and

prokaryotes may provide key genomic sequencing data set to explain the presence

of highly conserved sequences in vertebrates. New generation sequencing technol-

ogies should enhance such surveys. Extensive surveys of highly conserved

sequences in all kingdoms may provide clues to understand the nature of highly

conserved sequences in the genome such as the origin, mechanism of conservation,

and function if any at all.

Acknowledgments Author appreciates Dr. Daniel E. Janes for constructive discussions and

critical reading of this manuscript. The author thanks Dr. Yoshiyuki Sakaki and his colleagues

at RIKEN Genomic Sciences Center and Dr. Masayuki Yamamura and his colleagues at Tokyo

Institute of Technology for the extraction of LCNS and useful discussions. The author also

acknowledges Dr. Yoshiyuki Sakuraba and the members of the Population and Quantitative

Genomics Team at RIKEN Genomic Sciences Center, where the most of the LCNS works

described in this chapter was conducted. This work is partly supported by Grants-in-Aid for

Scientific Research (A) (KAKENHI 15200032 and KAKENHI 21240043).

References

Ahituv N, Zhu Y, Visel A, Holt A, Afzal V, Pennacchio LA, Rubin EM (2007) Deletion of

ultraconserved elements yields viable mice. PLoS Biol 5(9):e234

Bantle JA, Hahn WE (1976) Complexity and characterization of polyadenylated RNA in the

mouse brain. Cell 8:139–150

Bejerano G, Pheasant M, Makunin I, Stephen S, Kent WJ, Mattick JS, Haussler D (2004)

Ultraconserved elements in the human genome. Science 304(5675):1321–1325

Bejerano G, Lowe CB, Ahituv N, King B, Siepel A, Salama SR, Rubin EM, Kent WJ, Haussler D

(2006) A distal enhancer and an ultraconserved exon are derived from a novel retroposon.

Nature 441(7089):87–90

Britten RJ, Kohne D (1968) Repeated sequences in DNA. Science 161(841):529–540

204 Y. Gondo

Chikaraishi DM, Deeb SS, Sueoka N (1978) Sequence complexity of nuclear RNAs in adult rat

tissue. Cell 13:111–120

Dermitzakis ET, Reymond A, Lyle R, Scamuffa N, Ucla C, Deutsch S, Stevenson BJ, Flegel V,

Bucher P, Jongeneel CV, Antonarakis SE (2002) Numerous potentially functional but non-

genic conserved sequences on human chromosome 21. Nature 420(6915):578–582

Dermitzakis ET, Reymond A, Scamuffa N, Ucla C, Kirkness E, Rossier C, Antonarakis SE (2003)

Evolutionary discrimination of mammalian conserved non-genic sequences (CNGs). Science

302(5647):1033–1035

Derti A, Roth FP, Church GM, Wu CT (2006) Mammalian ultraconserved elements are strongly

depleted among segmental duplications and copy number variants. Nat Genet 38(10):

1216–1220

Drake JA, Bird C, Nemesh J, Thomas DJ, Newton-Cheh C, Reymond A, Excoffier L, Attar H,

Antonarakis SE, Dermitzakis ET, Hirschhorn JN (2006) Conserved noncoding sequences are

selectively constrained and not mutation cold spots. Nat Genet 38(2):223–227

Gondo Y, Okada T, Matsuyama N, Saitoh Y, Yanagisawa Y, Ikeda JE (1998) Human mega-

satellite DNA RS447: copy-number polymorphisms and interspecies conservation. Genomics

54(1):39–49

International Human Genome Sequencing Consortium (2001) Initial sequencing and analysis of

the human genome. Nature 409:860–921

International Human Genome Sequencing Consortium (2004) Finishing the euchromatic sequence

of the human genome. Nature 431:932–945

Judd BH, Shen MW, Kaufman TC (1972) The anatomy and function of a segment of the X

chromosome of Drosophila melanogaster. Genetics 71(1):139–156

Katzman S, Kern AD, Bejerano G, Fewell G, Fulton L, Wilson RK, Salama SR, Haussler D (2007)

Human genome ultraconserved elements are ultraselected. Science 317(5840):915

McLean C, Bejerano G (2008) Dispensability of mammalian DNA. Genome Res 18(11):

1743–1751

Mouse Genome Sequencing Consortium (2002) Initial sequencing and comparative analysis of the

mouse genome. Nature 420:520–562

Mukai T (1964) The genetic structure of natural populations of Drosophila melanogaster I.

Spontaneous mutation rate of polygenes controlling vaiability. Genetics 50:1–19

Mukai T (1978) Population genetics. Kodansha Scientific, Tokyo, in Japanese

Mukai T, Chigusa SI, Mettler LE, Crow JF (1972) Mutation rate and dominance of genes affecting

viability in Drosophila melanogaster. Genetics 72(2):335–355Nei M, Rogozin IB, Piontkivska H (2000) Purifying selection and birth-and-death evolution in the

ubiquitin gene family. Proc Natl Acad Sci USA 97(20):10866–10871

Nenoi M, Mita K, Ichimura S, Kawano A (1998) Higher frequency of concerted evolutionary

events in rodents than in man at the polyubiquitin gene VNTR locus. Genetics 148(2):867–876

Nowak R (1994) Mining treasures from “junk DNA”. Science 263:608–610

O’Brien SJ, Menotti-Raymond M, Murphy WJ, Nash WG, Wienberg J, Stanyon R, Copeland NG,

Jenkins NA, Womack JE, Marshall Graves JA (1999) The promise of comparative genomics in

mammals. Science 286(5439):458–481

Ohnishi O (1977) Spontaneous and ethyl methanesulfonate-induced mutations controlling viabil-

ity in Drosophila melanogaster. II. Homozygous effect of polygenic mutations. Genetics

87(3):529–545

Okada T, Gondo Y, Goto J, Kanazawa I, Hadano S, Ikeda JE (2002) Unstable transmission of the

RS447 human megasatellite tandem repetitive sequence that contains the USP17 deubiquiti-

nating enzyme gene. Hum Genet 110(4):302–313

Pennacchio LA, Ahituv N, Moses AM, Prabhakar S, Nobrega MA, Shoukry M, Minovitsky S,

Dubchak I, Holt A, Lewis KD, Plajzer-Frick I, Akiyama J, De Val S, Afzal V, Black BL,

Couronne O, Eisen MB, Visel A, Rubin EM (2006) In vivo enhancer analysis of human

conserved non-coding sequences. Nature 444(7118):499–502


Poulin F, Nobrega MA, Plajzer-Frick I, Holt A, Afzal V, Rubin EM, Pennacchio LA (2005) In vivo

characterization of a vertebrate ultraconserved enhancer. Genomics 85(6):774–781

Saitoh Y, Miyamoto N, Okada T, Gondo Y, Showguchi-Miyata J, Hadano S, Ikeda JE (2000) The

RS447 human megasatellite tandem repetitive sequence encodes a novel deubiquitinating

enzyme with a functional promoter. Genomics 67(3):291–300

Sakuraba Y, Sezutsu H, Takahasi KR, Tsuchihashi K, Ichikawa R, Fujimoto N, Kaneko S, Nakai

Y, Uchiyama M, Goda N, Motoi R, Ikeda A, Karashima Y, Inoue M, Kaneda H, Masuya H,

Minowa O, Noguchi H, Toyoda A, Sakaki Y, Wakana S, Noda T, Shiroishi T, Gondo Y (2005)

Molecular characterization of ENU mouse mutagenesis and archives. Biochem Biophys Res

Commun 336(2):609–616

Sakuraba Y, Kimura T, Masuya H, Noguchi H, Sezutsu H, Takahasi KR, Toyoda A, Fukumura R,

Murata T, Sakaki Y, Yamamura M, Wakana S, Noda T, Shiroishi T, Gondo Y (2008)

Identification and characterization of new long conserved noncoding sequences in vertebrates.

Mamm Genome 19(10–12):703–712

Visel A, Prabhakar S, Akiyama JA, Shoukry M, Lewis KD, Holt A, Plajzer-Frick I, Afzal V, Rubin

EM, Pennacchio LA (2008) Ultraconservation identifies a small subset of extremely con-

strained developmental enhancers. Nat Genet 40(2):158–160

Wetmur J, Davidson N (1968) Kinetics of renaturation of DNA. J Mol Biol 31(3):349–370

206 Y. Gondo

Part IIIMorphological Evolution / Speciation

Chapter 13

Male-Killing Wolbachia in the Butterfly

Hypolimnas bolina

Anne Duplouy and Scott L. O’Neill

Abstract Maternally inherited insect symbionts often manipulate host reproduc-

tion for their own benefit. Symbionts are transmitted to the next host generation

through the female hosts, and as such males represent dead ends for transmission.

Natural selection therefore favors symbiont-induced phenotypes that provide a

reproductive advantage to infected females, regardless of possible negative selec-

tive effects on males.

Male-killing (MK) is one such phenotype, in which symbionts kill the male

progeny of infected females. Compared with other symbiont-associated reproduc-

tive phenotypes, MK is relatively unexplored mechanistically as well as ecologi-

cally. A male-killingWolbachia bacterium strain named wBol1 has been describedin the tropical butterfly Hypolimnas bolina. By reviewing the different features of

this association it is possible to summarize what is already known about the biology

and evolution of MK symbionts, as well as highlight the current gaps in our

understanding of this striking reproductive phenotype.

13.1 Introduction

There are numerous symbiotic associations known to occur within nature; however,

few associations are more complex than those involving endosymbiosis. The study of

endosymbionts challenges the scientific community with questions about how each

member of the symbiosis coexists and how they maximize their reproductive fitness.

Endosymbionts are extremely common and over the course of evolution

have arisen in very different taxonomic groups. In insects, although endosym-

biotic eukaryotic microorganisms are common (e.g., the yeast-like endosymbiont

A. Duplouy and S.L. O’Neill

School of Biological Sciences, The University of Queensland, Brisbane, QLD 4072, Australia



209

Symbiotaphrina buchneri infecting anobiid beetles, Noda and Kodama 1996; or the

fungal symbiont of the brown-banded cockroach species Supella longipalpa,Gibson and Hunter 2009), most described endosymbionts are bacteria including

members of the Proteobacteria (e.g., Buchnera and Wolbachia), Flavobacteria(e.g., Blattabacterium), and Mollicutes (e.g., Spiroplasma) (Werren and O’Neill

1997; Bourtzis and Miller 2003, 2006), amongst others. Insect endosymbionts also

show diversity in their modes of transmission, either vertically (maternally) trans-

mitted from mother to offspring or horizontally transmitted. In the latter case,

symbiont may be infectious within a single species or between different species.

Examples of occasional horizontal transfer of maternally transmitted symbionts

have been reported (Werren and O’Neill 1997).

“Primary endosymbionts” are usually obligate endosymbionts, needed for host

reproduction and/or survival. For example, Moran et al. (2005) showed that Buch-nera aphidicola provides essential nutrients deficient within the aphid host’s diet.

Some primary endosymbionts have been shown to display phylogenetic concor-

dance with their hosts over millions of years demonstrating long-term coevolution

(Moran et al. 1993, 1994; Bandi et al. 1994).

Facultative endosymbionts, often referred to as “secondary endosymbionts,”

infect individuals already carrying a primary symbiont. A classic example is the

pea aphid Acyrthosiphon pisum that harbors multiple secondary symbionts such as

Hamiltonella defensa, in addition to the primary symbiont Buchnera sp. (Moran

et al. 2005; Oliver et al. 2007). The functional roles of secondary symbionts within

the host are not always well defined, as any effect can be hidden by the action of the

primary symbionts (Chen et al. 2000; Moran et al. 2005; Ruan et al. 2006).

Finally, “reproductive symbionts,” also termed “guest microbes” (Bourtzis and

Miller 2003), were first described as symbionts able to enhance their own fitness by

manipulating host reproduction (Taylor and Hoerauf 1999). Some of these distortions

involve sex ratio manipulation of the host. Spiroplasma for example kills males in

Drosophila species (Hurst et al. 1999a), while Cardinium sterilizes certain males of

the wasp Encarsia pergandiella (Hunter et al. 2003). However, recent studies have

revealed additional capabilities of reproductive symbionts that enhance their fitness

without affecting the host’s reproductive system (Brownlie et al. 2009).

13.2 Wolbachia Pipientis

Wolbachia pipientis is a species of obligate intracellular alpha-Proteobacteria

closely related to Rickettsia. Wolbachia were first discovered in the early 1920s

in the ovaries of the mosquito Culex pipiens (Hertig and Wolbach 1924). Based on

genetic variationWolbachia strains were divided into eight highly divergent super-groups named A through H (Bandi et al. 1998; Zhou et al. 1998; Bourtzis and Miller

2003; Lo et al. 2007). The two most studied and describedWolbachia supergroups,A and B, diverged approximately 50–70 million years ago (Werren et al. 1995;

Werren and O’Neill 1997).Wolbachia belonging to these two groups, known as the

210 A. Duplouy and S.L. O’Neill

“arthropodWolbachia,” are mostly harbored by insects but are also described from

other host phyla such as Crustacea or Arachnida. Supergroups A and B Wolbachiaare mostly parasitic and induce a broad range of reproductive distortions in their

hosts. In comparison, Wolbachia belonging to both the C and D supergroups are

mutualistic strains required for fertility and development of their filarial nematode

hosts (Bandi et al. 1998). Within the C and D clusters, Wolbachia phylogeny is

concordant with host phylogeny, suggesting long-term coevolution. The remaining

four clusters (E–H) infect various arthropods or nematodes; however, these asso-

ciations are often poorly described and symbiont-induced effects are not always

known (Vandekerckhove et al. 1999; Lo and Evans 2007; Covacin and Barker

2007). W. pipientis, the most extensively studied reproductive endosymbiont to

date, has the greatest diversity of host interactions including mutualism and all

types of known reproductive manipulations – cytoplasmic incompatibility, femini-

zation, parthenogenesis, or male-killing (O’Neill et al. 1997).

13.2.1 Reproductive Distortions

Maternally transmitted endosymbionts, such as Wolbachia, can enhance their

transmission rate by manipulating their host’s reproduction (O’Neill et al. 1997;

Bourtzis and Miller 2003). To understand the benefits they gain from these manip-

ulations, it is worthwhile summarizing what is known about the most common

symbiont-induced reproductive phenotypes.

The first reproductive manipulation to be attributed to Wolbachia was cytoplas-

mic incompatibility (CI). In the 1950s, Ghelelovitch (1952) and Laven (1959)

described crosses between strains of the mosquito Culex pipiens that sometimes

failed to produce progeny. Later, Yen and Barr (1971) showed that Wolbachia was

the causative agent of these reproductive failures. Wolbachia-infected males when

crossed with uninfected females failed, whereas all other possible crosses (crosses

between uninfected individuals, and between infected females and either uninfected

males or males carrying the same infection) resulted in normal reproductive output.

The mechanistic basis of this reproductive incompatibility between uninfected

females and infected males has been linked to abnormalities during fertilization

by cytological studies (Tram and Sullivan 2002). Abnormal behavior of chromo-

somal material from infected males causes incompatibility with female pronuclei

and later the death of the progeny. The CI of these gametes provides an advantage

to infected females, as they can successfully mate with both infected and nonin-

fected males. As a result, the maternally transmitted symbiont spread rapidly into

the host population. CI is not unique to Wolbachia: Cardinium also induces CI in

the parasitoid wasp Encarsia pergandiella (Hunter et al. 2003; Perlman et al. 2008),

and CI has been described as the most common endosymbiont-induced reproduc-

tive manipulation in arthropods.

AsWolbachia are maternally transmitted, some strains distort the sex ratio of their

host population to favor the female sex only, creating populations where males are

sometimes extremely rare. Three mechanisms, feminization, parthenogenesis, and

13 Male-Killing Wolbachia in the Butterfly Hypolimnas bolina 211

male-killing (MK), cause imbalanced sex ratio in the host population. Feminizing

symbionts such asCardinium andWolbachia have been found in numerous arthropod

hosts including the isopod Armadillidium vulgare (Cordaux et al. 2004), the butterflyspecies Ostrinia furnacalis and Eurema hecabe (Narita et al. 2007; Kageyama et al.

2008), and the spider mite Brevipalpus phoenicis (Weeks et al. 2001). During

feminization, genetic males reproduce as functional females, which therefore trans-

mit Wolbachia to their progeny (Rigaud 1997; Stouthamer et al. 1999).

Feminization is often mistaken for parthenogenesis, as both mechanisms pro-

duce female-biased populations. Although feminization requires sexual reproduc-

tion, parthenogenesis allows the production of viable progeny without the need for

a male partner. Two types of parthenogenesis have been described: arrhenotokous

parthenogenesis (or arrhenotoky) occurs when diploid females arise from fertilized

eggs and thelytokous parthenogenesis (or thelytoky) where females are produced

from unfertilized eggs. In the wasp species Trichogramma spp., thelytoky is

induced by Wolbachia (Stouthamer and Kazmer 1994), which restores diploidy

by enhancing the fusion of the two nuclei of the first mitotic division (Stouthamer

and Kazmer 1994; Huigens et al. 2000).

Finally, a wide range of endosymbiont-infected arthropods produce only daugh-

ters as male offspring die at an early development stage. Males are usually killed

embryonically, but deaths also occur much later, typically in fourth instar larvae

(Hurst 1991). This common reproductive manipulation is known as male-killing

(MK). MK is caused by at least nine different bacteria from four taxonomic groups:

Mollicutes, Flavobacteria, Rickettsiaceae, and Enterobacteriaceae (Hurst et al.

1997, 2003). However, there are still very few studies investigating the underlying

cytogenetic and genetic mechanisms of this phenotype.

13.3 Male-KillingWolbachia in the Butterfly Hypolimnas bolina

Although MK systems are diverse, a review of the association between the MK

Wolbachia strain wBol1 and H. bolina provides a general overview of this repro-

ductive phenotype.

H. bolina, also known as the common or great egg-fly (Australia), or blue-moon

butterfly (New Zealand), was first described by Linnaeus in 1758. This species has a

vast subtropical distribution from Sri Lanka to French Polynesia and a latitudinal

range from Hong-Kong to Canberra, Australia. Occasional reports describe

H. bolina in Japan and New Zealand since the 1970s (Ramsay 1971; Clarke and

Sheppard 1975; Morishita and Kazuhiko 2002; Patrick 2004), but it is suspected

that these regions do not support endemic populations (Common and Waterhouse

1972). Individuals observed in Japan and New Zealand were probably migratory

individuals using favorable meteorological conditions (Ryan and Harris 1990;

Christensen 2004) to invade from close neighboring regions such as South East

Asia (SEA) or Australia, where stable populations exist (Gibbs 1961; Ramsay and

Ordish 1966; Ramsay 1971; Christensen 2004).


13.3.1 All-Female Broods in the Butterfly H. bolina

A strong female sex distortion has been described in numerous H. bolina popula-

tions throughout their wide geographical distribution (Simmonds, 1926, Clarke

et al. 1975; Dyson et al. 2002; Charlat et al. 2005). All-female broods were first

described in the 1920s (Poulton 1923; Simmonds 1926). This reproductive trait was

showed to be exclusive to females and therefore due to a cytoplasmic factor (Clarke

et al. 1975). It was reported not to be parthenogenesis as males were dying at early

stages of development (Clarke et al. 1975, 1983).

Dyson et al. (2002) identifiedW. pipientis as the causative agent of male rareness

in H. bolina, using PCR amplification and sequence analysis of a bacterial surface

protein gene (Zhou et al. 1998). ThisWolbachia strain termed wBol1 was shown tokill the male progeny of infected female butterflies at an early embryonic stage

before caterpillars hatch from the eggs (Dyson et al. 2002, Fig. 13.1).

First identified in Fiji, wBol1 was found to be present in most H. bolinapopulations across the South Pacific (Charlat et al. 2005). One intriguing feature

of the wBol1/H. bolina association has been a variation in wBol1 infection preva-

lence among different host populations. wBol1 infections were absent from

(1)

(2)

(5)4 days

4 days

25 days

7 days pupae

5 caterpillar instars

Death of thewBol1-infectedmale embryos

eggs

wBol1

wBol1

(4)(3)

Fig. 13.1 Life cycle of wBol1-infected Hypolimnas bolina: (1) a wBol1-infected female mates

with an uninfected male, (2) all males die during embryogenesis, only female eggs hatch 4 days

after being laid, (3) caterpillars develop in 20 days through 5 larval instars, (4) wBol1-infectedfemales emerge from 7-day old pupae, (5) and 4 days after emerging from the pupae, females are

reproductively mature


Australian and the Tubuai (French Polynesia, Austral Islands Archipelago)

H. bolina populations, while wBol1 infection frequencies of up to 50% in Fijian

populations and more than 85% in both the Independent Samoan and Tahitian

populations were recorded (Fig. 13.2; Charlat et al. 2005, 2006).

13.3.2 Competition Between Wolbachia Infections

A number of possible reasons have been suggested to explain the heterogeneity in

wBol1 infection rates (Fig. 13.2, Table 13.1). In the extreme case of Tubuai (Austral

Islands Archipelago, French Polynesia), no butterflies were found to be infected by

the male killer strain wBol1, while on the closest neighboring island of Rurutu, only210 km away, female wBol1 infection rate was more than 75% (Charlat et al. 2005,

2006). It was found that butterflies from Tubuai were infected with another Wol-bachia strain, named wBol2. The wBol2 strain is an A-group Wolbachia that is

phylogenetically distant from wBol1, a B-group Wolbachia. Crosses between

wBol1-infected females and wBol2-infected males were fully incompatible and

Island not infected by wBol 1

Low and medium infection rate

High infection rate(2)

(1)

(3)

(4)(5)

(6)

(7)(8)

(9) (10)

(12)(11)

Australia

New Zealand 1000 km

Equator

N

Fig. 13.2 Wolbachia infection frequencies in 12 H. bolina populations. (1) Philippines, (2)

Thailand, (3) Vanuatu, (4) Fiji, (5) New Caledonia: Ile des Pins, (6) Australia: Brisbane, (7)

Independent Samoa, (8) American Samoa, French Polynesia: (9) Moorea, (10) Tahiti, (11) Rurutu,

and (12) Tubuai. Less than 65% of the females are wBol1-infected in islands with low and medium

infection frequencies, and 65–100% of the females are wBol1-infected in islands with high

infection frequency


lead to unviable progeny. This phenotype was the result of wBol2-induced CI in

H. bolina (Charlat et al. 2006). The competition between wBol2 and wBol1 and thestrong CI observed between the twoWolbachia strains make the invasion of Tubuai

by the MK strain, wBol1, extremely unlikely. The presence of wBol2 was reported

in several other islands of the South Pacific where wBol1 was not shown to occur

(Charlat et al. 2006).

13.3.3 When the MK Phenotype Is Repressed, wBol1 Induces CI

At the other extreme, allH. bolina from South East Asian populations were infected

by wBol1, including males (Charlat et al. 2005; Hornett et al. 2006). Under the

strong selection pressure exerted by the wBol1 infection, butterflies have evolved

resistance to the MK phenotype. This mutation led to survival of male offspring and

restored a balanced sex ratio (Hornett et al. 2006, Table 13.1).

If wBol1 from host populations with the MK repressor gene were shown to retain

their ability to induce MK in nonresistant host, then it would suggest either (1) that

the repressor gene was the result of an extremely recent mutation in the host or (2)

that the MK character was linked to a desirable trait providing an advantage to the

repressed wBol1. Otherwise, long-term evolution in a host population that repressed

MK may result in the loss of wBol1’s MK virulence – a character no longer able to

spread in the population. Hornett and co-workers (2008) conducted crosses between

MK resistant H. bolina from SEA and nonresistant populations of French Polynesia

(Moorea and Tahiti, Society Islands Archipelago) and tested whether wBol1 from

SEA could induce MK. The SEA wBol1 infection was able to distort host

Table 13.1 Percentage of males and females in different populations naturally uninfected (column 2)

or infected by the different Wolbachia strains (columns 3–5)

Populations % Uninfected

male/female

% wBol1-a-infected male/

female

% wBol1-b-infected male/

female

% wBol2-infected male/

female

MK

repressor

gene

Philippines 0/0 100/100 0/0 0/0 Present

Thailand 0/0 100/100 0/0 0/0 Present

Ile des Pins 100/17 0/83 0/0 0/0

Fiji 100/50 0/50 0/0 0/0

Vanuatu 100/70 0/30 0/0 0/0

Australia 100/100 0/0 0/0 0/0

Ind. Samoa 0/0 100/100 0/0 0/0 Present

Am. Samoa 0/0 0/0 0/0 100/100

Moorea 98/17 0/80 0/3 2/0

Tahiti 100/4 0/90 0/6 0/0

Rurutu 98/29 0/69 0/0 2/2

Tubuai 2/2 0/0 0/0 98/98

MK repressor gene presence is shown in column 6 (Charlat et al. 2005, 2006, 2007b; Hornett et al.

2006)


reproduction when transferred into a French Polynesian background, indicating that

wBol1 from SEA can still induce theMK phenotype in nonresistant hosts. The study

also revealed a complete failure in egg hatch when SEAmales carrying both the MK

infection and MK repressor gene(s) were crossed with uninfected females. Control

crosses showed that the females were not sterile, suggesting that in addition to MK

wBol1 also induces CI in this population of H. bolina (Hornett et al. 2008).

13.3.4 MK Wolbachia Diversity in H. bolina

More recently, Charlat et al. (2009) shown that the MK phenotype in H. bolina wasinduced by two substrains, wBol1-a and wBol1-b. Although they are extremely

closely related phylogenetically, genetic variations between them have only been

found at two loci, wBol1-a and wBol1-b show phenotypic differences that make

them interesting candidates for comparative analysis (Charlat et al. 2009). wBol1-aand wBol1-b seem to differ in their sensitivity to the MK repressor from SEA.

Preliminary results suggest that wBol1-a MK was repressed when transferred into a

SEA background, while wBol1-b showed persistent MK phenotype in this novel

host background (Charlat pers. comm. 2007). These results suggest small variations

in the MK genetic bases between these two substrains.

These two substrains also differ in their transmission level. The wBol1-b infec-

tion, which has been found in only French Polynesia and Vanuatu (Charlat et al.

2009, Table 13.1), was associated with mitochondrial haplotypes (mitotypes) 3 and

6. These mitotypes were also found in Wolbachia-free butterflies, suggesting

imperfect vertical transmission of wBol1-b. In contrast, the most common strain

wBol1-a was present on all the islands where wBol1 was previously described

(Charlat et al. 2005, 2006 Table 13.1) and was strictly associated with mitotype 1.

The almost complete absence of uninfected butterflies carrying mitotype 1 suggests

a very high transmission efficiency of wBol1-a.More recent investigations into wBol1-a genetic variation have found no evi-

dence that wBol1-a prevalence was related to genetic differences between wBol1-apopulations (Duplouy et al. 2009). The age of the infection in the South Pacific

islands may vary; for example, the wBol1-a invasion of Fiji could be more recent

than that of Tahiti, where a larger proportion of females carried the infection.

13.3.5 A Rapidly Evolving System

The association between wBol1 and H. bolina has proved to be highly dynamic. In

2001, Samoan H. bolina populations were shown to have at most a single male per

hundred females. Charlat and colleagues (2007a) reported in a 2006 survey equal

sex ratios and a second case of MK repression in the South Pacific. It was not known

whether the genetic basis of MK resistance was similar in both SEA and Samoan


populations. Nonetheless, the shift in population sex ratio from 100:1 to 1:1 in less

than ten generations seemed to be one of the fastest ever recorded (Charlat et al.

2007a). A more ancient but similar evolution of a MK repressor gene has also been

described in butterflies from Malaysian Borneo (Hornett et al. 2009).

The spread of wBol1 through SEA and the South Pacific was estimated to have

taken less than 3,000 years (Duplouy et al. 2009). However, in some populations,

local invasions were suggested to have occurred more rapidly, on the scale of a

century. Museum samples from different South Pacific islands were tested for both

the infection type and prevalence in previous butterfly generations. In the 120 years

from 1883 until 2002, the infection frequencies in the French Polynesian Islands of

Ua Huka and Tahiti varied from very low prevalence (0% and less than 20%,

respectively) to very high prevalence (more than 80%) (Hornett et al. 2009).

13.4 Open Questions in wBol1 Research

13.4.1 wBol1 Biogeography

The biogeography of the wBol1-a infection in the South Pacific was one of the most

intriguing aspects of this system. The presence of butterfly populations on numer-

ous South Pacific islands provided clear evidence of natural migrations occurring

between islands; however, the range of these exchanges remained an unknown

factor. Butterfly populations infected with the CI-inducing strain wBol2 were

almost as common as wBol1-infected populations in the South Pacific. In contrast,

populations where the two infections coexist have rarely been recorded, and doubly

infected butterflies have never been found (Charlat et al. 2005). Models predicted

that, in this system, a CI-infected population would resist MK invasion. Under the

same conditions, a MK-infected population would only resist invasion by

CI-inducing Wolbachia if the latter did not reach a certain frequency threshold

(Freeland and McCabe 1997; Engelst€adter et al. 2004). If the limit was exceeded,

then the CI-inducing strain became more competitive and therefore, spread into the

population driving the former MK infection to extinction (Engelst€adter et al. 2004).Butterfly populations where wBol1 and wBol2 were in competion were rare in the

South Pacific islands (Charlat et al. 2005; Engelst€adter et al. 2008). This rarity

suggested a low migration rate between islands, allowing MK-infected populations

to resist wBol2 invasion.

13.4.2 Effects of MK Infection on Host Fitness

Endosymbiotic infections are generally costly to maintain as the symbionts exploit

resources that are destined for their host (Haine 2008). In order to be maintained


and spread within host populations, symbionts may develop strategies that enhance

the fitness of infected hosts relative to uninfected individuals. Wolbachia strains

have developed very intimate relationships with their hosts and stress treatments

have shown that some strains are beneficial to their hosts (Hedges et al. 2008;

Brownlie et al. 2009). Modeling predicts that MK fixation would lead to population

extinction because of a severe shortage of males (Hamilton 1967; Hurst 1991;

Randerson et al. 2000); however, wBol1 infection sometimes exceeds 75% of host

individuals in a population. The success of infected individuals over uninfected

ones suggests that wBol1 infection may confer a fitness advantage to its hosts, but

the nature of this benefit has not yet been characterized in H. bolina.

13.4.2.1 Benefits from the Infection in Other Host/MK WolbachiaAssociations

Direct benefits from the infection, such as an increased size, fecundity, or longevity,

have been recorded in different associations with MK Wolbachia (Ikeda 1970;

Majerus and Hurst 1997; Fry et al. 2004). These observations contrast with the

wBol1/H. bolina system where no benefit of this type has been shown (Dyson and

Hurst 2004; Charlat et al. 2007b).

Similarly, although indirect benefits from MK infection have been described in

several other MK systems, none has yet been associated with a fitness increase in

wBol1-infected butterflies. (Werren 1987) suggested that MK endosymbionts

could reduce sibling inbreeding, thereby favoring infected females. This explana-

tion makes sense for species that lay many eggs on the same plant and are not very

mobile after hatching. In the case of H. bolina, however, butterflies lay few eggs

per plant and are good migrants as individuals have frequently invaded New

Zealand from Australia, a journey of 2,000 km (Ramsay 1971; Ryan and Harris

1990; Patrick 2004). Majerus and Hurst (1997) suggested that the success of MK

strains in ladybirds (e.g., Adalia bipunctata) was correlated with different host

characteristics, including cannibalism at various developmental stages; so

infected females gain nutrition from feeding on their dead brothers, and from

large clutch sizes, as MK reduces sibling competition for food by diminishing

their numbers by half. H. bolina does not exhibit the characteristics of a host in

which a MKWolbachia would be successful. This butterfly is strictly herbivorousduring its larval stages and as adult feeds exclusively on nectar. As such, male

death would provide no direct nutritional benefit to infected sisters and as females

lay only 1–2 eggs per plant, food competition would be limited (Nafus 1993;

Kemp 1998).

13.4.2.2 Alternative Hypotheses

wBol1-a may confer a “hidden” selective advantage to infected hosts (Duplouy

et al. 2009). Insects are often infected with entomopathogenic agents (fungi, viruses


or bacteria). Phytophageous insects, such as butterflies, also have to avoid plant

defenses, such as toxic compounds, developed by their host plants to fight against

natural enemies (Lindroth 1989; Li et al. 2003; Wen et al. 2006). Caterpillars are

common prey for parasitoid or predatory wasps such as Cotesia spp. or Polistes spp.(Stamp and Bowers 1988; Nafus 1993; Beckage et al. 1994; van Nouhuys and

Hanski 2005). These selective pressures allow the survival of only resistant or

adapted individuals (Hochberg 1991; Russell and Moran 2005; Moran 2006; Haine

2008). Wolbachia may confer their host a benefit when exposed to toxins and/or

parasites and thereby increase its prevalence within host populations.

Recent studies have shown Wolbachia-infected flies delay mortality after virus

infection (Hedges et al. 2008; Teixeira et al. 2008). Investigating the effect of

wBol1 infection in a metacommunity involving the host, the symbiont, and at least a

third party such as a virus or a parasitoid wasp could provide insights into fitness

benefit(s) this infection provides the butterfly host with. Fitness benefit(s) that could

therefore help explaining the striking success of wBol1-a in H. bolina.

13.4.3 Mechanisms of MK

13.4.3.1 Cytology of MK

Two types of MK have been characterized based on the timing of male death (Hurst

1991). “Early MK” occurs during embryogenesis while “late MK” takes effect

during larval or pupal stages. Both early and late MK were observed inWolbachia-infected insects (Hurst et al. 1999b; Fialho and Stevens 2000; Jiggins et al. 2000;

Dyson et al. 2002; Jaenike 2007); however, the underlying mechanisms of either

phenomena have not yet been elucidated.

Studies on MK Spiroplasma-infected Drosophila have shown that male embryo

death was associated with abnormal mitoses, while later death was caused by

degeneration of cell nuclei (pycnosis) (Counce and Poulson 1962). In a similar

system, modification of the dosage compensation complex (DCC), which is

involved in sex differentiation, can also rescue males from MK symbionts. This

indicates that the DCC may be involved in expression of the MK phenotype (Veneti

et al. 2005). Although MK in Wolbachia-infected insects must also involve host

sex determination, similar mechanisms to those in Spiroplasma associations havenot yet been identified. One study showed that treatment of wBol1-a-infectedbutterflies with bacteriostatic antibiotics delayed the MK effect. This demon-

strates that wBol1 was able to identify male individuals and induce MK at

different time points during host development (Charlat et al. 2007c). However,

it is unknown if the basic mechanisms of the MK phenotype are identical at each

time point. As suggested, MK could be expressed through different pathways

(Hurst and Jiggins 2000), which would complicate the identification of the

mechanistic basis of these MK phenotypes.


13.4.3.2 Genomics of MK

To date the genomes of one mutualistic and three CI-inducing Wolbachia strains

have been sequenced (Wu et al. 2004; Foster et al. 2005; Klasson et al. 2008, 2009)

and several others are underway. Wolbachia’s intracellular biology has hampered

the completion of whole genome-sequencing projects. The genome sequence of the

MK strain wBol1 is nearing completion (Duplouy pers.comm.), and analysis of the

first chromosomal DNA sequence of a MK Wolbachia strain will certainly be of

great value. Comparative genomic analysis of wBol1 with the closely related and

fully sequenced wPip strain, which induces CI in Culex mosquitoes, should provide

an unique opportunity to investigate the evolution of Wolbachia genomes across

relatively short evolutionary timescales. This first genomic comparison between a

MK strain and a CI-inducing strain offers opportunities to test hypotheses

concerning the evolution and induction of the MK phenotype, such as identifying

candidate genes involved in both MK and CI.

Previous whole genome analyses have attempted to link genetic elements, such as

ankyrin coding genes, to the induction of different reproductive manipulations

(Iturbe-Ormaetxe et al. 2005; Duron et al. 2007; Walker et al. 2007; Klasson et al.

2008). Ankyrin repeat domains are believed to be involved in cellular and molecular

functions via protein–protein interactions (Caturegli et al. 2000;Mosavi et al. 2004).

Twenty-three, 29, and 60 ankyrin genes have been annotated in the Wolbachiastrains wMel, wRi, and wPip, respectively (Wu et al. 2004; Klasson et al. 2009;

Walker et al. 2007), while wBm seems to contain only 5 ankyrin coding genes

(Foster et al. 2005). wBol1 is phylogenetically close to the wPip strain, and it is

therefore expected that theMK strain also contains a large number of ankyrin coding

genes. The number and density of ankyrin coding genes in pathogenic strains make

them good candidates in the search for genes likely to play a role in the interactions

between Wolbachia and its host (Iturbe-Ormaetxe et al 2005; Duron et al. 2007;

Walker et al. 2007).

Despite intensive efforts,Wolbachia transformation is currently not an available

technique. While waiting for an efficient transformation protocol for Wolbachia,genomic comparison ofWolbachia strains may provide extremely valuable data. If

the mechanisms of MK are similar across strains, the genetic basis of this phenotype

should be conserved between these strains and putative MK genes could potentially

be identified. Genome comparisons of phylogenetically related strains such as

wBol1 and wPip or wBol1-a and wBol1-b, which induce different phenotypes in

their hosts, may identify highly variable genes or genetic features potentially

involved in the induction of the observed phenotypic differences.

To date, only protein coding genes have been investigated as potential genetic

mechanisms underlying Wolbachia-induced phenotypes (Sinkins et al 2005;

Walker et al. 2007; Duron et al. 2007). Small RNA molecules (sRNAs) are

known in other systems to act through RNA interference (RNAi) to regulate

translation of targeted genes (Tjaden et al. 2006 and including references). Simi-

larly, MK Wolbachia could use sRNAs, rather than proteins to distort their hosts’

reproductive system. Comparative projects should therefore not only focus on


protein coding genes present in Wolbachia genomes, but also on the diversity of

sRNA sequences, as they could also play a key role in the distortion of host

reproductive systems.

13.4.3.3 Role of the Host in the Expression of MK

We have already described the symbiont-induced effects on different aspects of

host biology; however, biological interactions are rarely unidirectional. Hosts can

also act to mitigate any negative fitness effects associated with the symbiont. These

interactions have been highlighted in different Wolbachia associations; however,

the molecular mechanisms that underlie these interactions are not understood.

In SEA and Samoa,H. bolina evolved resistance to theMK phenotype ofwBol1-a,saving males from embryonic death (Hornett et al. 2006; Charlat et al. 2007a).

Although the investigation of the butterfly genetics is in progress and should soon

provide answers (Hornett pers. comm. 2009), it is not yet known if the resistance

mechanism involves one or several genes, and whether this resistance is identical in

both the SEA and Samoan populations (Charlat et al. 2007a). This repression,

however, confirms the active involvement of the host in the phenotype induced

by its symbiont.

More interestingly, butterfly resistance to MK resulted in wBol1-a shifting to

inducing CI (Hornett et al. 2008). In general, the reproductive phenotype observed

in the natural host has been maintained in transfected hosts (Braig et al. 1994;

Riegler et al. 2004; Sakamoto et al. 2005; McMeniman et al. 2008); however,

immediate phenotypic shifts after transfection have been reported (Sasaki et al.

2002, 2005; Jaenike 2007). Phylogenetic studies of Wolbachia have demonstrated

that very closely related strains express different phenotypes in their native hosts

(Baldo et al. 2006), suggesting that shifts in phenotype expression are probably

more common than originally thought. It also suggests that MK and CI might share

a similar molecular basis that is differently expressed depending on host genotype

(Jaenike 2007). Both phenotypes could be mechanistically similar; however, MK

has evolved to be more extreme in its outcome than the CI.

13.5 Conclusion

Wolbachia have attracted the attention of a large scientific community, hoping to

understand the biology of this bacterium that induces such a wide range of host

phenotypes and has great potential as a biological control agent of insect pests and

human diseases (Brelsfoard et al. 2009; McMeniman et al. 2009; Moreira et al.

2009). Many discoveries have been made in the last decade, but a multitude of

questions still remain to be answered. MK is one of the least known Wolbachiaphenotypes. Although we have a relatively good understanding of how MK Wol-bachia affect host populations, genetics and dynamics, the cytology and genomics


aspects underlying the MK phenotype both remain poorly understood. We may

come closer to finding answers with projects such as whole genome comparison of

MK strains, but we are still far from having resolved all of Wolbachia’s mysteries.

Acknowledgments We would like to thank Dr. I. Iturbe-Ormaetxe, Dr. M. Woolfit and

Dr. P. Cook for very constructive comments on the manuscript. We are grateful to the Australian

Research Council (DP0772992) and to The University of Queensland (UQCS and UQIRTA) for

provision of the funds.

References

Baldo L, Hotopp JCD, Jolley KA, Bordenstein SR, Biber SA, Choudhury RR, Hayashi C, Maiden

MCJ, Tettelin H, Werren JH (2006) Multilocus sequence typing for Wolbachia. Appl EnvironMicrobiol 72(11):7098–7110

Bandi C, Damiani G, Magrassi L, Grigolo A, Fani R, Sacchi L (1994) Flavobacteria as intracellu-

lar symbionts in cockroaches. Proc Biol Sci 257:43–48

Bandi C, Anderson TJC, Genchi C, Blaxter ML (1998) Phylogeny of Wolbachia in filarial

nematodes. Proc Biol Sci 265:2407–2413

Beckage NE, Tan FF, Schleifer KW, Lane RD, Cherubin LL (1994) Characterization and

biological effects of Cotesia congregata polydnavirus on host larvae of the tobacco hornworm,

Manduca sexta. Arch Insect Biochem Physiol 26:165–195

Bourtzis K, Miller TA (eds) (2003) Insect symbiosis. CRC Press, New York, NY

Bourtzis K, Miller TA (eds) (2006) Insect symbiosis, vol 2. CRC Press, New York, NY

Braig HR, Guzman H, Tesh RB, O’Neill SL (1994) Replacement of the natural Wolbachiasymbiont of Drosophila simulans with a mosquito counterpart. Nature 367:453–455

Brelsfoard CL, StClair W, Dobson SL (2009) Integration of irradiation with cytoplasmic incom-

patibility to facilitate a lymphatic filariasis vector elimination approach. Parasit Vectors 2:38

Brownlie JC, Cass BN, Riegler M, Witsenburg JJ, Iturbe-Ormaetxe I, McGraw EA, O’Neill CL

(2009) Evidence for metabolic provisioning by a common invertebrate endosymbiont,

Wolbachia pipientis, during periods of nutritional stress. PLoS Pathog 5:6

Caturegli P, Asanovich KM, Walls JJ, Bakken JS, Madigan JE, Popov VL, Dumler JS (2000)

ankA: an Ehrlichia phagocytophila group gene encoding a cytoplasmic protein antigen with

ankyrin repeats. Infect Immun 68(9):5277–5283

Charlat S, Hornett EA, Dyson EA, Ho PPY, Thi-Loc N, Schilthuizen M, Davies N, Roderick GK,

Hurst GDD (2005) Prevalence and penetrance variation of male-killing Wolbachia across

Indo-Pacific populations of the butterfly Hypolimnas bolina. Mol Ecol 14:3525–3530

Charlat S, Engelstadter J, Dyson E, Hornett E, Duplouy A, Tortosa P, Davies N, Roderick G,

Wedell N, Hurst G (2006) Competing selfish genetic elements in the butterfly Hypolimnasbolina. Curr Biol 16:2453–2458

Charlat S, Hornett EA, Fullard JH, Davies N, Roderick GK, Wedell N, Hurst GDD (2007a)

Extraordinary flux in sex ratio. Science 317:214

Charlat S, Reuter M, Dyson EA, Hornett EA, Duplouy A, Davies N, Roderick GK, Wedell N,

Hurst GDD (2007b) Male-killing bacteria trigger a cycle of increasing male fatigue and female

promiscuity. Curr Biol 17:273–277

Charlat S, Davies N, Roderick GK, Hurst GDD (2007c) Disrupting the timing of Wolbachia-induced male-killing. Biol Lett 3:154–156

Charlat S, Duplouy A, Hornett EA, Dyson EA, Davies N, Roderick GK, Wedell N, Hurst GDD

(2009) The joint evolutionary histories of Wolbachia and mitochondria in Hypolimnas bolina.BMC Evol Biol 9:64


Chen D-Q, Montllor CB, Purcell AH (2000) Fitness effects of two facultative endosymbiotic

bacteria on the pea aphid, Acyrthosiphon pisum, and the blue alfalfa aphid, A. kondoi. Entomol

Exp Appl 95:315–323

Christensen B (2004) Tracking of migrant blue moon butterfly, Hypolimnas bolina nerina, usingweb-based software. Weta 28:47–48

Clarke C, Sheppard PM (1975) The genetics of the mimetic butterfly Hypolimnas bolina (L.).

Philos Trans R Soc Lond B Biol Sci 272(917):229–265

Clarke C, Sheppard P, Scali V (1975) All-female broods in the butterfly Hypolimnas bolina (L.).

Proc Biol Sci 189:29–37

Clarke SC, Jonhson G, Jonson B (1983) All-female broods in Hypolimnas bolina (L.). A re-survey

of West Fiji after 60 years. Biol J Linn Soc 19:221–235

Common IFB, Waterhouse DF (1972) Butterflies of Australia. Angus and Robertson, Sydney

Cordaux R, Michel-Salzat A, Frelon-Raimond M, Rigaud T, Bouchon D (2004) Evidence for a

new feminizing Wolbachia strain in the isopod Armadillidium vulgare: evolutionary implica-

tions. Heredity 93:78–84

Counce SJ, Poulson DF (1962) Developmental effects of the sex-ratio agent in embryos of

Drosophila willistoni. J Exp Zool 151:17–31

Covacin C, Barker SC (2007) Supergroup F Wolbachia bacteria parasite lice (Insecta: Phthirap-

tera). Parasitol Res 100:479–485

Duplouy A, Hurst GDD, O’Neill SL, Charlat S (2009) Rapid spread of male-killing Wolbachia in

the butterfly Hypolimnas bolina. J Evol Biol. Doi:10.1111/j.1420-9101.2009.01891.x

Duron O, Boureux A, Echaubard P, Berthomieu A, Berticat C, Fort P, Weill M (2007) Variability

and expression of ankyrin domain genes in Wolbachia infecting the mosquito Culex pipiens.J Bacteriol 189(12):4442–4448

Dyson EA, Hurst GDD (2004) Persistence of an extreme sex-ratio bias in a natural population.

PNAS 101(17):6520–6523

Dyson E, Kamath M, Hurst G (2002) Wolbachia infection associated with all-female broods in

Hypolimnas bolina (Lepidoptera: Nymphalidae): evidence for horizontal transmission of a

butterfly male killer. Heredity 88:166–171

Engelst€adter J, Telschow A, Hammerstein P (2004) Infection dynamics of different Wolbachia-

types within one host population. J Theor Biol 231:345–355

Engelst€adter J, Telschow A, Yamamura N (2008) Coexistence of cytoplasmic incompatibility and

male-killing-inducing endosymbionts, and their impact on host flow. Theor Popul Biol

73:125–133

Fialho RF, Stevens L (2000) Male-killing Wolbachia in a flour beetle. Proc Biol Sci

267:1469–1474

Foster J, Ganatra M, Kamal I, Ware J, Makarova K, Ivanova N, Bhattacharyya A, Kapatral V,

Kumar S, Posfai J, Vincze T, Ingram J, Moran L, Lapidus A, Omelchenko M, Kyrpides N,

Ghedin E, Wang S, Goltsman E, Joukov V, Ostrovskaya O, Tsukerman K, Mazur M, Comb D,

Koonin E, Slatko B (2005) TheWolbachia genome of Brugia malayi: endosymbiont evolution

within a human pathogenic nematode. PLoS Biol 3:599–614

Freeland SJ, McCabe BK (1997) Fitness compensation and the evolution of selfish cytoplasmic

elements. Heredity 78:391–402

Fry AJ, Palmer MR, Rand DM (2004) Variable fitness effects ofWolbachia infection in Drosoph-ila melanogaster. Heredity 93:379–389

Ghelelovitch S (1952) Sur le determinisme genetique de la sterilite dans les croisements entre

differentes souches de Culex autogenicus Roubaud. C R Acad Sci III 234:2386–2388

Gibbs GW (1961) New Zealand butterflies. Tuatara J Biol Soc 9:65–76

Gibson CM, Hunter MS (2009) Inherited fungal and bacterial endosymbiont of a parasitic wasp

and its cockroach host. Microb Ecol 57(3):542–549

Haine ER (2008) Symbiont-mediated protection. Proc Biol Sci 275:353–361

Hamilton WD (1967) Extraordinary sex ratios. Science 156(774):477–488


Hedges LM, Brownlies JC, O’Neill SL, Johnson KN (2008) Wolbachia and virus protection in

insects. Science 322:702

Hertig M, Wolbach SB (1924) Studies on Rickettsia-like microorganisms in insects. J Med Res

44:329–374

Hochberg ME (1991) Viruses as costs to gregarious feeding behaviors in the Lepidoptera. Oikos

61(3):291–296

Hornett EA, Charlat S, Duplouy AMR, Davies N, Roderick GK, Wedell N, Hurst GDD (2006)

Evolution of male killer suppression in natural population. PLoS Biol 4(9):e283

Hornett EA, Duplouy AMR, Davies N, Roderick GK, Wedell N, Hurst GDD, Charlat S (2008)

You can’t keep a good parasite down: evolution of a male-killer suppressor uncovers cytoplas-

mic incompatibility. Evolution 62(5):1258–1263

Hornett EA, Charlat S, Wedell N, Jiggins CD, Hurst GDD (2009) Rapidly shifting sex ratio across

a species range. Curr Biol 19:1628–1631

Huigens ME, Luck RF, Klaassen RHG, Maas MFPM, Timmermans MJTN, Stouthamer R (2000)

Infectious parthenogenesis. Nature 405:178–179

Hunter MS, Perlman SJ, Kelly SE (2003) A bacterial symbiont in the Bacteroidetes induces

cytoplasmic incompatibility in the parasitoid wasp Encarsis pergandiella. Proc Biol Sci

270:2185–2190

Hurst L (1991) The incidences and evolution of cytoplasmic male killers. Proc Biol Sci 244:91–99

Hurst GDD, Jiggins FM (2000) Male-killing bacteria in insects: mechanisms, incidence, and

implications. Emerg Infect Dis 6(4):329–336

Hurst GDD, Hurst LD, Majerus MEN (1997) Cytoplasmic sex ratio distorters. In: O’Neill SL,

Hoffmann AA, Werren JH (eds) Influential passengers, inherited microorganisms and arthro-

pod reproduction. Oxford University Press Inc, New York, pp 125–154

Hurst GDD, van der Schulenburg JHG, Majerus TMO, Bertrand D, Zakharov IA, Baungaard J,

Volkl W, Stouthamer R, Majerus MEN (1999a) Invasion of one insect species, Adaliabipunctata, by two different male-killing bacteria. Insect Mol Biol 8(1):133–139

Hurst GDD, Jiggins FM, van der Schulenburg JHG, Bertrand D, West SA, Goriacheva II,

Zakharov IA, Werren JH, Stouthamer R, Majerus MEN (1999b) Male-killing Wolbachia in

two species of insect. Proc Biol Sci 266(1420):735–740

Hurst GDD, Jiggins FM, Majerus MEN (2003) Inherited microorganisms that selectively kill male

hosts: the hidden players of insect evolution? In: Bourtzis K, Miller TA (eds) Insect symbiosis.

CRC Press, New York, NY, pp 177–197

Ikeda H (1970) The cytoplasmic-inherited ‘sex-ratio-condition’ in natural and experimental

populations of Drosophila bifasciata. Genetics 65:311–333Iturbe-Ormaetxe I, Riegler M, O’Neill SL (2005) New names for old strains?Wolbachia wSim is

actually wRi. Genome Biol 6:401

Jaenike J (2007) Spontaneous emergence of a new Wolbachia phenotype. Evolution 61

(9):2244–2252

Jiggins FM, Hurst GDD, Jiggins CD, von der Schulenburg JHG, Majerus MEN (2000) The

butterfly Danaus chrysippus is infected by a male-killing Spiroplasma bacterium. Parasitology

120:439–446

Kageyama D, Narita S, Noda H (2008) Transfection of feminizing Wolbachia endosymbionts of

the butterfly, Eurema hecabe, into the cell culture and various immature stages of the silkmoth,

Bombyx mori. Microb Ecol 56(4):733–741

Kemp DJ (1998) Oviposition behaviour of post-diapause Hypolimnas bolina (L.) (Lepidoptera:

Nymphalidae) in tropical Australia. Aust J Zool 46:451–459

Klasson L, Walker T, Sebaihia M, Sanders MJ, Quail MA, Lord A, Sanders S, Earl J, O’Neill SL,

Thomson N, Sinkins SP, Parkhill J (2008) Genome evolution of Wolbachia strain wPip from

the Culex pipiens group. Mol Biol Evol 25(9):1877–1887

Klasson L, Westberga J, Sapountzis P, Naslund K, Lutnaes Y, Darby AC, Veneti Z, Chend L,

Braig HR, Garrett R, Bourtzis K, Andersson SGE (2009) The mosaic genome structure of the

Wolbachia wRi strain infecting Drosophila simulans. PNAS 106(14):5725–5730


Laven H (1959) Speciation by cytoplasmic isolation in the Culex pipiens complex. Cold Spring

Harb Symp Quant Biol 24:166–175

Li W, Schuler MA, Berenbaum MR (2003) Diversification of furanocoumarin-metabolizing

cytochrome P450 monooxygenases in two papilionids: specificity and substrate encounter

rate. PNAS 100(Suppl 2):14593–14598

Lindroth RL (1989) Host plant alteration of detoxication activity in Papilio glaucus glaucus.Entomol Exp Appl 50:29–35

Lo N, Evans TA (2007) Phylogenetic diversity of the intracellular symbiontWolbachia in termites.

Mol Phylogenet Evol 44:461–466

Lo N, Paraskevopoulos C, Bourtzis K, O’Neill SL, Werren JH, Bordenstein SR, Bandi C (2007)

Taxonomic status of the intracellular bacteriumWolbachia pipientis. Int J Syst Evol Microbiol

57:654–657

Majerus MEN, Hurst GDD (1997) Ladybirds as a model for the study of male-killing symbionts.

Entomophaga 42(1/2):13–20

McMeniman CJ, Lane AM, Fong AW, Voronin DA, Iturbe-Ormaetxe I, Yamada R, McGraw EA,

O’Neill SL (2008) Host adaptation of a Wolbachia strain after long-term serial passage in

mosquito cell lines. Appl Environ Microbiol 74(22):6963–6969

McMeniman CJ, Lane RV, Cass BN, Fong AWC, Sidhu M, Wang Y-F, O’Neill SL (2009) Stable

introduction of a life-shorteningWolbachia infection into the mosquito Aedes aegypti. Science323:141–144

Moran NA (2006) Symbiosis. Curr Biol 16(20):866–871

Moran NA, Munson MA, Baumann P, Ishikawa H (1993) A molecular clock in endosymbiotic

bacteria is calibrated using the insect hosts. Proc Biol Sci 253:167–171

Moran NA, Baumann P, von Dohlen C (1994) Use of DNA sequences to reconstruct the history of

the association between members of the Sternorrhyncha (Homoptera) and their bacterial

endosymbionts. Eur J Entomol 91:79–83

Moran NA, Dunbar HE, Wilcox JL (2005) Regulation of transcription in a reduced bacterial

genome: nutrient-provisioning genes of the obligate symbiont Buchnera aphidicola. J Bacteriol187(12):4229–4237

Moreira LA, Iturbe-ormaetxe I, Jeffery JAL, Lu G, Pyke AT, Hedges LM, Rocha BC, Hall-

Mendelin S, Day A, Riegler M, Hugo LE, Johnson KN, Kay BH, McGraw EA, van der Hurk

AF, Ryan PA, O’Neill SL (2009) AWolbachia symbiont in Aedes aegypti limits infection with

dengue, chikungunya and Plasmodium. Cell 139(7):1268–1278Morishita and Kazuhiko (2002) A migrant from an oceanic island – Hypolimnas bolina, 6 days

stay near Zushi Beach, Kanagawa, Japan. Butterflies 32:24–26

Mosavi LK, Cammett TJ, Desrosiers DC, Peng Z-Y (2004) The ankyrin repeat as molecular

architecture for protein recognition. Protein Sci 13:1435–1448

Nafus DM (1993) Movement of introduced biological control agents onto nontarget butterflies,

Hypolimnas spp. (Lepidoptera: Nymphalidae). Environ Entomol 22(2):265–272

Narita S, Kageyama D, Nomura M, Fukatsu T (2007) Unexpected mechanism of symbiont-

induced reversal of insect sex: feminizing Wolbachia continuously acts on the butterfly

Eurema hecabe during larval development. Appl Environ Microbiol 73(13):4332–4341

Noda H, Kodama K (1996) Phylogenetic position of yeast-like endosymbionts of Anobiid beetles.

Appl Environ Microbiol 62(1):162–167

O’Neill SL, Hoffmann AA, Werren JH (1997) Influencial passengers,inherited microorganisms

and arthropod reproduction. Oxford University Press Inc., New York

Oliver KM, Campos J, Moran NA, Hunter MS (2007) Population dynamics of defensive symbionts

in aphids. Proc Biol Sci 275:293–299

Patrick BH (2004) Invasion of the blue moon butterfly in Taranaki. Weta 28:45–46

Perlman SJ, Kelly SE, Hunter MS (2008) Population biology of cytoplasmic incompatibility:

maintenance and spread of Cardinium symbionts in a parasitic wasp. Genetics 178:1003–1011

Poulton EB (1923) All female families of Hypolimnas bolina, bred in Fiji by HW Simmonds. Proc

R Ent Soc Lond 1923:9–12


Ramsay GW (1971) The blue moon butterfly Hypolimnas bolina nerina in New Zealand during

autumn, 1971. N Z Entomol 5:73–75

Ramsay GW, Ordish RG (1966) The Australian blue moon butterflyHypolimnas bolina nerina (F.)in New Zealand. NZ J Sci 9:719–729

Randerson JP, Smith NGC, Hurst LD (2000) The evolutionary dynamics of male-killers and their

hosts. Heredity 84:152–160

Riegler M, Charlat S, Stauffer C, Mercot H (2004) Wolbachia transfer from Rhagoletis cerasi toDrosophila simulans: investigating the outcomes of host-symbiont coevolution. Appl Environ

Microbiol 70(1):273–279

Rigaud T (1997) Inherited microorganisms and sex determination of arthropod hosts. In: O’Neill

SL, Hoffmann AA, Werren JH (eds) Influential passengers, inherited microorganisms and

arthropod reproduction. Oxford University Press Inc, New York, pp 81–101

Ruan Y-M, Xu J, Liu S-S (2006) Effects of antibiotics on fitness of the B biotype and a non-B

biotype of the whitefly Bemisia tabaci. Entomol Exp Appl 121:159–166

Russel JA, Moran NA (2005) Horizontal transfer of bacterial symbiont: heritability and fitness in a

novel aphid host. Appl Environ Microbiol 71(12):7987–7994

Ryan PA, Harris AC (1990) A note of recent records of Australian butterflies in New Zealand. N Z

Entomol 13:40–41

Sakamoto H, Ishikawa Y, Sasaki T, Kikuyama S, Tatsuki S, Hoshizaki S (2005) Transinfection

reveals the crucial importance ofWolbachia genotypes in determining the type of reproductive

alteration in the host. Genet Res 85:205–210

Sasaki T, Kubo T, Ishikawa H (2002) Interspecific transfer of Wolbachia between two lepidop-

teran insects expressing cytoplasmic incompatibility: a Wolbachia variant naturally infecting

Cadra cautella causes male-killing in Ephesia kuehniella. Genetics 162:1313–1319Sasaki T, Massaki N, Kubo T (2005) Wolbachia variant that induces two distinct reproductive

phenotypes in different hosts. Heredity 95:389–393

Simmonds HW (1926) Sex ratio of Hypolimnas bolina in Viti Levu, Fiji. Proc R Ent Soc Lond

1:29–32

Sinkins SP, Walker T, Lynd AR, Steven AR, Makepeace BL, Godfray HC, Parkhill J (2005)

Wolbachia variability and host effects on crossing type in Culex mosquitoes. Nature

14:257–260

Stamp NE, Bowers MD (1988) Direct and indirect effects of predatory wasps (Polistes sp.:

Vespidae) on gregarious caterpillars (Hemileuca lucina: Saturniidae). Oecologia 75:619–624

Stouthamer R, Kazmer D (1994) Cytogenetics of microbe-associated parthenogenesis and its

consequences for gene flow in Trichogramma wasps. Heredity 73:317–327

Stouthamer R, Breeuwer JAJ, Hurst GDD (1999) Wolbachia pipientis: microbial manipulator of

arthropod reproduction. Annu Rev Microbiol 53:71–102

Taylor MJ, Hoerauf A (1999) Wolbachia bacteria of filarial nematodes. Parasitol Today 15

(11):437–442

Teixeira L, Ferreira A, Ashburner M (2008) The bacterial symbiontWolbachia induces resistanceto RNA viral infections in Drosophila melanogaster. PLoS Biol 6(12):2753–2763

Tjaden B, Goodwin SS, Opdyke JA, Guillier M, Fu DX, Gottesman S, Storz G (2006) Target

prediction for small, noncoding RNAs in bacteria. Nucleic Acids Res 34(9):2791–2802

Tram U, Sullivan W (2002) Role of delayed nuclear envelope breakdown and mitosis in Wolba-chia-induced cytoplasmic incompatibility. Science 296:1124–1126

van Nouhuys S, Hanski I (2005) Metacommunities of butterflies, their host plant, and their

parasitoids. In: Holyoak M, Leibold MA, Holt RD (eds) Metacommunities spatial dynamics

and ecological communities. University of Chicago Press, USA

Vandekerckhove TTM, Watteyne S, Willems A, Swings JG, Mertens J, Gillis M (1999) Phyloge-

netic analysis of the 16 S rDNA of the cytoplasmic bacterium Wolbachia from the novel host

Folsomia candida (Hexpoda, Collembola) and its implications for Wolbachia taxonomy.

FEMS Microbiol Lett 180:179–286


Veneti Z, Bentley JK, Koana T, Braig HR, Hurst GDD (2005) A functional dosage compensation

complex required for male-killing in Drosophila. Science 307:1461–1463Walker T, Klasson L, Sebaihia M, Sanders MJ, Thomson NR, Parkhill J, Sinkins SP (2007)

Ankyrin repeat domain-encoding genes in the wPip strain ofWolbachia from the Culex pipiensgroup. BMC Biol 5(39):1–9

Weeks AR, Marec F, Breeuwer JAJ (2001) Amite species that consists entirely of haploid females.

Science 292:2479–2482

Wen Z, Rupasinghe S, Niu G, Berenbaum MR, Schuler MA (2006) CYP6B1 and CYP6B3 of the

Black Swallowtail (Papilio polyxenes): adaptative evolution through subfunctionalization.

Mol Biol Evol 23(12):2434–2443

Werren JH (1987) The coevolution of autosomal and cytoplasmic sex ratio factors. J Theor Biol

124:317–334

Werren JH, O’Neill SL (1997) The evolution of heritable symbionts. In: O’Neill SL, Hoffmann

AA, Werren JH (eds) Influential passengers, inherited microorganisms and arthropods repro-

duction. New York, Oxford University Press Inc., pp 1–41

Werren JH, Windsor D, Guo L (1995) Distribution of Wolbachia among neotropical arthropods.

Proc Biol Sci 262:197–204

Wu M, Sun LV, Vamathevan J, Riegler M, Deboy R, Brownlie JC, McGraw EA, Martin W,

Esser C, Ahmadinejad N, Wiegand C, Madupu R, Beanan MJ, Brinkac LM, Daugherty SC,

Durkin AS, Kolonay JF, Nelson WC, Mohamoud Y, Lee P, Berry K, Young MB, Utterback T,

Weidman J, Nierman WC, Paulsen IT, Nelson KE, Herve Tettelin, O’Neill SL, Eisen JA

(2004) Phylogenomics of the reproductive parasite Wolbachia pipientis wMel: a streamlined

genome overrun by mobile genetic elements. PLoS Biol 2:327–341

Yen JH, Barr AR (1971) New hypothesis of the cause of cytoplasmic incompatibility in Culexpipiens L. Nature 232:657–658

Zhou W, Rousset F, O’Neill SL (1998) Phylogeny and PCR-based classification of Wolbachiastrains using wsp gene sequences. Proc Biol Sci 265(1395):509–515


Chapter 14

Evolution of Immunosuppressive Organelles

from DNA Viruses in Insects

Brian A. Federici and Yves Bigot

Abstract Endoparasitic wasps inject particles into their lepidopteran hosts that

enable these parasitoids to evade or directly suppress the hosts’ innate immune

response, especially encapsulation by hemocytes. For decades, these particles

have been considered virions produced by DNA viruses known as polydnaviruses

(family Polydnaviridae). Structurally, there are two main types of particles, those

resembling, respectively, virions of baculoviruses or ascoviruses. These particles

contain double-stranded DNA in the form of multiple small circular molecules that

are transcribed but not replicated in cells of the lepidopteran hosts. Instead particle

DNA is replicated from the wasp genome and selectively amplified for packaging

into the particles in the reproductive tract of female wasps. Once assembled and

secreted into calyx lumen, the particles become mixed with eggs and injected

into caterpillars during wasp oviposition. Particle DNA, referred to as the “viral

genome,” has now been sequenced for several polydnaviruses. Annotation shows

that most of this DNA consists of noncoding DNA or wasp genes, not viral genes.

More significantly, recent studies have shown that particle structural proteins are

coded by the wasp genome, not by particle DNA, but are of viral origin. Together,

these findings provide strong evidence that these particles originated from viruses,

but through symbiogenesis followed by gene deletion and acquisition evolved into

transducing organelles that shuttle wasp immunosuppressive genes into their hosts,

thereby enhancing wasp progeny survival and species radiation.

B.A. Federici

Department of Entomology, University of California, Riverside 900 University Avenue, Riverside,

California 92521, USA

Laboratoire d’Etude des Parasites GenetiquesParc Grandmont, Universite de Tours, U.F.R. des

Sciences et Techniques, 37200, Tours, France


Y. Bigot

Laboratoire d’Etude des Parasites GenetiquesParc Grandmont, Universite de Tours, U.F.R. des

Sciences et Techniques, 37200 Tours, France


229

14.1 Introduction

14.1.1 Background

George Salt at the University of Cambridge published a series of pioneering studies

during the 1960s aimed at understanding how endoparasitic wasps circumvented

the innate immune response of their caterpillar hosts. Based on studies of the

ichneumonid parasitoid, Venturia (then Nemeritis) canescens and its lepidopteran

host, larvae of the Mediterranean flower moth, Ephestia kuehniella, he determined

that parasitoid eggs gained protection as they passed through the calyx (egg

storage region) of the female wasp’s reproductive tract (Salt 1965, 1966, 1968).

This protection was due to a coating added to the eggs in the calyx. Subsequently,

Susan Rotheram, one of Salt’s graduate students, determined that this coating

contained masses of enveloped virus-like particles about 130 nm in diameter. After

assembly in calyx cell nuclei, these were secreted into the calyx lumen where they

adhered to fibrillar matrix on the egg surface (Rotheram 1967). In later studies,

Rotheram 1973a, b showed that the particles contained protein and complex sugars,

but no DNA. Then another of Salt’s graduate students found that a major particle

glycoprotein was responsible for the immunoprotection (Bedwin 1979a, b). Follow-

ing on these studies, Otto Schmidt and his collaborators in Germany showed that this

protein was encoded in the wasp genome, but likely originated from basal lamina

proteins found in the caterpillar host (Schmidt and Schuchmann-Feddersen 1989;

Schmidt and Theopold 1991; Schmidt et al. 2001).

After Salt and Rotheram’s studies, Vinson and colleagues as well as others found

that particles in the calyx fluid of the endoparasitic ichneumonids Campoletissonorensis and Cardiochiles nigriceps also suppressed the immune response of

their caterpillar hosts (Vinson 1972; Vinson and Scott 1975; Vinson 1990). These

particles were also produced in the nuclei of calyx cells, but though morphologi-

cally similar to V. canescens particles, they contained DNA. These findings stimu-

lated numerous investigations of the calyx gland and secretions of many

endoparasitic wasps of the families Ichneumonidae and Braconidae, revealing

two major particle types, one in ichneumonids and another in braconids (see Stoltz

and Vinson 1979, and Vinson 1990; Webb et al. 2005). When first discovered, the

ichneumonid particles were not typical of virions of any known type of insect virus

(Fig. 14.1). They were bound by two unit membranes, were oblong to globular in

shape, and ranged from 130 to 150 nm in diameter by 300–400 nm in length, with a

fusiform nucleocapsid (Webb et al. 2005). Later, viruses of a new family, the

ascoviruses (family Ascoviridae) were discovered that attacked caterpillars, repli-

cating and produces progeny virions in various host tissues. The virions produced

by ascoviruses are structurally similar to the ichneumonid particles and are trans-

mitted by parasitic wasps (Federici 1983; Federici et al. 2005). In contrast to

the ichneumonid particles, those produced by braconid wasps resembled nudi-

virus virions and similar virions of the occluded form of baculoviruses (Burand

1998; Wang and Jehle 2009). They consisted primarily of one or more cylindrical

230 B.A. Federici and Y. Bigot

particles surrounded by a single envelope (Fig. 14.1b). The cylindrical inner

particle varied in length from 30 to 100 nm, even within the same wasp species.

Similar particles have been identified in more than 50 wasp species. In these, unlike

the genomes of most viruses of insects, the DNA does not occur as a single circular

molecule, but as numerous circular molecules. These vary in size from few to many

kbp and are referred to as segmented, polydispersed, or multipartite DNA (Stoltz

1993; Webb et al. 2005). Most evidence indicates these particles do not have a

genome per se, but rather their DNA is part of the wasp genome (Espagne et al.

2004; Webb et al. 2006; Desjardins et al. 2008). Moreover, as far as is known,

though genes contained in the particles are expressed in nuclei of the parasitoid’s

caterpillar host cells, no particle DNA replication occurs in these, nor do the

particles produce any progeny. From the standpoint of a viral life cycle, they are

a dead end.

14.1.2 Establishment of the Family Polydnaviridae

Based on the unusual physical and biological properties of these particles and their

obligate symbiotic relationship with wasps (Edson et al. 1981), a new virus family,

Polydnaviridae (“Poly” referring to the polydispersed DNA), was established to

accommodate these newly discovered viruses (Stoltz et al. 1984). Establishment of

this family formalized the recognition of two genera, the genus Ichnovirus (ichno-viruses) for particles produced by ichneumonid wasps, and genus Bracovirus(bracoviruses) for particles produced by braconids (Webb et al. 2005). At the

time these genera where erected, the particles were considered to be infective

viruses capable of replication (at least for these viruses in calyx cells), much like

that which occurs in other types of viruses. Although molecular data were not

Fig. 14.1 Transmission electron micrographs of immunosuppressive particles produced by endo-

parasitic braconid and ichneumonid wasps. (a) Bracovirus particles. (b) Ichneumonid particles.

The bracovirus particles resemble nudivirus and baculovirus virions, and molecular evidence now

indicates that these particles have their origin in an ancestral nudivirus. The ichneumonid particles

resemble ascovirus virions, but their origin remains uncertain at present. Bars ¼ 200 nm. Original

micrographs by D.B. Stoltz

14 Evolution of Immunosuppressive Organelles from DNA Viruses in Insects 231

sufficient at that time to undertake meaningful comparisons of these viruses,

available information as well as the significant structural differences between the

particles of these two virus types suggested that the association of each with its

corresponding wasp family arose independently. Thus, their similar functional roles

in parasite biology and success were and are considered a result of convergent

evolution.

14.1.3 Particle Function: General Mechanismsof Viral Immunosuppression

Detailed studies of several polydnavirus/parasitoid systems have shown that the

virus-like particles produced by these wasps in major braconid and ichneumonid

lineages (Whitfield 2002a, b) are required for suppression of the wasps’ hosts’

immune system in all species studied to date (Stoltz 1993, Vinson 1990; Webb et al.

2005, 2006). Suppression, depending on the specific system, occurs either by

molecular mimicry, where the surface of the egg and early instars are coated with

particles not recognized as foreign, by hemocyte inactivation through expression of

particle genes after oviposition, or by both mechanisms. Many of the genes encoded

by these wasp particles also inhibit components of innate immune pathways,

including the Toll and Imd pathways. Detailed knowledge of how the particle

genes of individual wasp species elude or incapacitate innate immune responses

varies considerably from one wasp species to another, and thus our understanding

of these processes is still in the early stages of development. Our purpose in this

chapter, therefore, is not to discuss specific particle functions, but rather to summa-

rize the key data that support the concept that these particles, though they originated

as virions, are a novel type of organelle that originated by lateral gene transfer/

symbiogenesis. Those interested in detailed discussions of particle functions as well

as their similarities and differences are referred to the excellent articles by Webb

et al. (2006) and Tanaka et al. (2007).

14.2 Polydnavirus Particles as Organelles Rather

Than Virions – the Concept

The structural similarity of braconid particles to baculovirus virions, and ichneu-

monid particles to ascovirus virions, made these viruses obvious choices as the

evolutionary sources of these two types of immunosuppressive particles (Federici

1991; Federici and Bigot 2003). At the time braconid particles were discovered,

the baculoviruses consisted of two main types, referred to as “occluded,” meaning

that the virions were occluded in a protein matrix, and “nonoccluded,” meaning that

they were not. Subsequently, the nonoccluded baculoviruses were reclassified into


a new type known as the nudiviruses. The nudivirus group consists of a small and

very diverse group of nonoccluded viruses from insects and crustaceans that share

33 core genes with baculoviruses (out of more than 100), but differ in host range

and pathology (Wang and Jehle 2009). Of significant evolutionary importance is

that one of these nudiviruses, HzNV-2, replicates in the reproductive tract of the

lepidopteran Heliothis zea, a host used commonly by many braconid and ichnomo-

nid wasps. Of particular significance is the recent finding that an ancestral nudivirus

is the likely source of the structural proteins encoded by braconid wasps that

compose their immunosuppressive particles (Bezier et al. 2009). While current

evidence for the origin of the ichneumonid immunosuppressive particles is not

nearly as strong as that for the braconids, recent molecular analyses suggest these

originated from ascovirus virions or a related ancestor virus (Bigot et al. 2008).

Data supporting these origins are discussed in more detail later below.

Although the braconid and ichneumonid particles clearly resemble nudivirus and

ascovirus virions, even early studies of these indicated they lacked important

properties characteristic of all viruses. For example, once within a lepidopteran

host cell, there was no replication of DNA. Moreover, in no case was there any

production of progeny virions to disseminate the virus and infect the next host or

cell. Other evidence indicating that the particles were not virions of a virus were

that the so-called infection of host cells and particle production in the wasp tissues

was strictly under control of the wasp. In all viruses, while they interact in various

ways with host cells, it is the virus that controls the synthesis of virus proteins and

replication of DNA, not the host cell, strictly speaking. Yet in the case of the

braconid and ichneumonid particles, they were only produced in female wasps, and

only in a narrow region of the reproductive tract, and only in pupal and adult tissues

as eggs were being produced (Webb et al. 2006). Adding to these problems in

classifying the particles as those of a virus was the occurrence of similar immuno-

suppressive particles that contained no DNA, such as those produced by the

ichneumonid, V. cansecens, discussed above (Rotheram 1967) and more recently

in other parasitic wasps (Barratt et al. 1999).

Given that even before the DNA in particles was sequenced there was substantial

evidence that they were not virions, the question became what are they? The most

obvious correlates were something like mitochondria and plastids, organelles that

originated from bacteria through the fusion of genomes, i.e., symbiogenesis fol-

lowed by gene loss and acquisition (Margulis and Fester 1991; Margulis 1992;

Khakhina 1992). The evidence is now indisputable that mitochondria and chlor-

oplasts, for example, originated from bacteria that became endosymbionts and

subsequently evolved into organelles. By analogy, the same evolutionary processes

occurred, although much more recently, with endoparasitic braconid and ichneu-

monid wasps and at least two different types of viruses, an ancestral nudivirus in the

case of the braconids, and for the ichneumonids, probably an ancestral ascovirus or

iridovirus (the latter being the ancestor of the ascoviruses). Whereas the molecular

evidence is still weak for the origin of the ichneumonid particles from ascoviruses,

the evidence that bracoviruses originated from an ancestral nudivirus is now very

strong (Bezier et al. 2009).


At present, polydnavirus researchers continue to refer to the braconid and

ichneumonid particles as, respectively, bracovirus or ichnovirus virions, despite

overwhelming evidence from their own studies to the contrary (Webb et al. 2006;

Tanaka et al. 2007; Bezier et al. 2009). Alternatively, based on the molecular data

regarding their evolution, current genetic complements, and functions, we argue

that these interesting immunosuppressive particles should be recognized for what

they are – organelles that evolved from viruses. Continuing to view these organelles

as viruses masks a much more interesting biological and evolutionary phenomenon

than viewing them as “symbiotic viruses.” It also contravenes the definition of such

fundamental concepts as a virus, a genome, and symbiosis. If these particles are

viruses, we have a tripartite – a virus, a wasp, and its lepidopteran host (Webb et al.

2006). Viewing the particles as organelles makes it a bipartite system, a wasp with a

novel organelle encoded in the genome and a lepidopteran host (Federici and Bigot

2003). We think that this new paradigm better explains their biological properties

and diversity and leads to better hypotheses for testing how they evolved and

facilitated the evolution of wasps and their insect hosts.

Below we elaborate on some of the key evidence for the likely evolutionary

pathways that led to these novel organelles. We move from the braconid system, for

which the most molecular data are available, to the ichneumonid system. We finish

with a description of several other types of endoparasitic wasp/insect host systems

which putatively represent various phases of the symbiotic evolutionary process

that range from (1) tripartite systems consisting of a wasp, true virus, and insect

host, to (2) bipartite systems consisting of a wasp with an organelle that has a DNA

complement, and an insect host, to (3) bipartite systems with wasp with organelle

lacking a DNA complement, and an insect host.

14.3 The Evolution of Braconid Particles from Nudiviruses

14.3.1 Early Studies of Nudiviruses in BraconidWasps and Their Hosts

Several viruses that have the structural features of nudiviruses have been known for

many years. For example, the nudivirus of the braconid, Microplitis croceipes, istransmitted vertically, replicates in hemocytes and other tissues, and causes signifi-

cant pathology and mortality in adult wasps (Hamm et al. 1988). A more interesting

nudivirus is the so-called filamentous virus (FV) of the braconid, Cotesia margin-iventris. CmFV is apparently a benign virus that is transmitted vertically by

C. marginiventris and replicates in cells of both the wasp’s lateral and common

oviduct, the latter near the calyx, and in cells of its lepidopteran hosts including

Helicoverpa zea and Spodoptera frugiperda (Hamm et al. 1990). Structurally, the

virions of these wasp-transmitted viruses resemble the nudiviruses, Hz-I, and the

Gonad-Specific Virus, that occur, respectively, in cells lines derived from H. zea


and in the gonadal tissues of this species (Burand 1998). TheMicroplitis and CmFV

nudiviruses viruses are apparently maintained in host populations by vertical

transmission. An even more interesting nudivirus is Hz-NV1, a large virus with a

genome of 228 kbp (Wang and Jehle 2009). This virus has been shown to integrate

into the chromosomes of Trichoplusia ni (TN 368) and S. frugiperda (SF21AE and

SF9) cells, in which it can establish a latent infection (Lin et al. 1999). This is

particularly relevant to symbiogenesis because it demonstrates that a large ds DNA

circular genome can integrate into the chromosomes of their insect hosts. This

provides a possible mechanism for the evolutionary entry of full or partial nudivirus

genomes into wasp genomic DNA.

The above examples are very limited but they do at least provide examples of

the types of viral/host systems that could lead over evolutionary time to the

integration of nudivirus or baculovirus genomes into those of their wasp hosts.

Fortunately, owing to the studies by Espagne et al. (2004), and more recently

Bezier et al. (2009), we now have very strong evidence that such an integration

actually occurred, and given the estimates of Whitfield (2002a), a little less than

100 mya.

14.3.2 Molecular Evidence for the Evolution of BraconidParticles from a Nudivirus

One of the predictions of a viral paradigm is that the DNA in the virions would

encode virion structural proteins and enzymes needed for the various replication

and assembly processes. An organelle paradigm, on the other hand, would predict a

significant reduction in genome size and that many, if not most of the original

genes, would be transferred to the nuclear genome or lost during evolution. Thus,

before any braconid or ichneumonid particles genomes, the so-called “viral

genomes” were sequenced, we predicted that most of the DNA in the particle

would consist of wasp genes, that is, DNA originating from wasp chromosomes

(Federici 1991; Federici and Bigot 2003). The first significant confirmation of the

organelle paradigm came from the sequencing DNA in the particles produced by

the braconid wasp, Cotesia congregata (Espagne et al. 2004). In this important

study, it was shown that fewer than 2% of the genes were related to those of any

known virus. Most of the genes encoded proteins with physiological functions,

such as protein tyrosine phosphatases, ankyrins, cysteine-rich proteins, and cysta-

tins. Some of the genes were related to the genes found in the particles produced

by other braconid species, but nevertheless, none of these was related to any

known virion structural protein. Similar findings have now been reported for the

“genomes” of particles produced by other braconids, including those of Glypta-panteles indiensis and G. flavicoxis (Desjardins et al. 2008). The DNA in all the

particles sequenced to date consists mostly of noncoding DNA of wasp origin, and

DNA that codes for wasp proteins. Some of these genes may well have originated


from viruses or bacteria, but they likely have been part of wasp genomes for

millions of years, and therefore are now in essence wasp genes. Even though the

structural characteristics of the particles made it probable they originated from a

baculovirus or nudivirus, these results made it clear that the “genomic” DNA,

unlike in the case of any other known virus, could not be used to find the viral

origin from which the particles evolved. Nor would these “genomes” be very

useful for polydnavirus systematics, because if the particle DNA is wasp DNA,

the sequences would likely reflect the relationships of the wasps. In fact, evidence

for this was already apparent years ago for braconid particle DNAs for several

Cotesia species (Whitfield 2002b).

As it had been known for many years that the braconid particles were produced

in calyx cells, a way to get at more meaningful data regarding the origin of the

braconid particles was to clone and sequence the transcripts from reproductive

tissues at the time of particle production. Thus, in another important and insightful

paper, Bezier et al. (2009) sequenced 5,000 expressed sequence tags from the

ovaries of two braconid wasps, Chelonus inanitus and C. congregata, and one

ichneumonid,Hyposoter didymator. The sequences from the ichneumonid wasp did

not show any relationship to known viral proteins, but analysis of the braconid

sequences proved very profitable. They identified 22 sequences related to nudi-

viruses, and 13 of these were core genes shared with baculoviruses. The genes

identified correlated with nudivirus and baculovirus virion structural proteins,

proteins involved in virion assembly, and subunits of viral RNA polymerases. No

polymerases involved in DNA replication were detected, indicating wasp poly-

merases were likely responsible for synthesis of braconid particle “genomes.”

Aside from providing excellent data regarding the original of crucial particle

components and proteins needed for particle assembly, these data show clearly

that these proteins are all encoded in the wasp genome and are under strict regulation

by the wasp genome, again a property not characteristic of any known virus.

14.4 Origin and Evolution of Ichneumonid Particles

As noted above for braconid particles, the DNA in ichnemonid particles consists

primarily of noncoding ichneumonid wasp DNA and genes coding for ichneumonid

proteins involved in immunosupression. Therefore, this DNA, while of some value

for suggesting the possible viral origins of these particles, as discussed below, we

do not currently have the type of information from these wasps corresponding to the

data described above for the braconid particles. The structure of the ichneumonid

particles suggests they originated from ascoviruses, and fortunately we do have

reasonably good molecular data for the evolution of ascoviruses from iridoviruses

(Stasiak et al. 2003). So we first review here pertinent key features of iridioviruses

and ascoviruses, and then review the limited molecular evidence suggesting the

ichnoviruses evolved from an ascovirus or iridovirus ancestor of these.


14.4.1 Family Iridoviridae

The family Iridoviridae is comprised of a diverse group of enveloped, double-

stranded (ds) DNA viruses which produce large icosahedral virions that typically

range 125–160 nm in diameter (Fig. 14.2). These viruses are commonly found in

invertebrates, particularly insects, but also occur among vertebrates (Chinchar et al.

2005). Iridoviruses have a broad tissue tropism in insects, and infect and replicate in

most tissues, with the unusual exception of the midgut epithelium, a tissue that most

insect viruses attack readily. Corresponding with their tissue tropism, iridoviruses

are poorly infectious per os (Federici 1993). Once within a cell, iridovirus DNA

replication, formation of the virogenic stroma, and virion assembly all take place in

the cytoplasm.

Iridoviruses have been reported from diverse lepidopteran hosts, including the

rice stem borer, Chilo suppressalis (Pyralidae), the American armyworm,

Heliothis armigera (Noctuidae), and the fall armyworm, S. frugiperda (Noctui-

dae). Relevant to the possibility that an ancestral iridovirus or ascovirus is the

source of the ichneumonid particles, the ichneumonid, Eiphosoma vitticolle,which parasitizes larvae of the fall armyworm, S. frugiperda, is also infected by

an iridovirus, and transmits this virus to fall armyworm populations in the field

(Lopez et al. 2002).

Fig. 14.2 Electron micrographs of iridovirus and ascovirus virions. Iridovirus virions observed in

negatively stained preparations (a) and by transmission electron microscopy (b), respectively.

Ascovirus virions as observed in negatively stained preparations (c) and by transmission electron

microscopy (d), respectively. Despite the marked difference in virion structure, molecular evi-

dence indicates these two types of viruses are closely related, and that the ascoviruses evolved

from iridoviruses. Bar ¼ 100 nm


14.4.2 Family Ascoviridae

The ascoviruses (family Ascoviridae) are ds DNA viruses that attack lepidopterans

and are characterized by large, enveloped virions, 130 � 400 nm, which vary,

depending on the species, from allantoid to bacilliform in shape (Federici et al.

2005). Structural studies of ascovirus virions suggest that these contain two unit

membranes, one that is part of the inner particle that surrounds the DNA core, and a

second that makes up part of the outer virion envelope (Fig. 14.2). There are

significant differences between ascovirus and ichneumonid particles, but neverthe-

less they correspond in size and general morphology (Figs. 14.1 and 14.2). Each

ascovirus virion contains a single ds DNA genome, which, depending on the species,

ranges from 138 to 180 kb. Four species of ascoviruses are recognized, S. frugiperdaascovirus (SfAV-1a), Trichoplusia ni ascovirus (TnAV-2a), Heliothis virescensascovirus (HvAV-3a), and Diadromus pulchellus ascovirus (DpAV-4a). The first

three occur in noctuid species such as the cabbage looper, T. ni, cotton budworms and

bollworms of Heliothis and Heliocoverpa species, and armyworms, Spodopteraspecies, in the United States. These viruses are pathogens that kill the wasp’s host

and as a result, wasp larvae as well. The fourth, noted earlier, occurs in France, where

it attacks the pupa of the leakmoth, Acrolepiosis assectella (family Yponomeutidae).

This ascovirus is a true symbiotic virus that enhances the parasitic success of its wasp

vector. All ascoviruses replicate genomic DNA, producing large numbers of progeny

virions in their caterpillar or pupal hosts. Ascoviruses differ from all other viruses in

that after they invade a cell, they destroy the nucleus and direct the cell to cleave into

numerous vesicles in which virion assembly proceeds. These vesicles are liberated

from tissues into the hemolymph, where female wasps acquire them mechanically

during oviposition and transmit them to new caterpillar hosts.

Aside from structural similarities, ascovirus virions and ichneumonid particles

depend on parasitic wasps for transmission. Much like insect iridoviruses, ascov-

iruses are very difficult to transmit per os, but are highly infectious when transmit-

ted by parasitoids or by injection (Hamm et al. 1985). Even more importantly with

respect to the organelle paradigm and symbiogenesis, the genome of the

D. pulchellus ascovirus (DpAV-4a) is carried in a nonintegrated form in the nuclei

of males and females of its ichneumonid wasp vector, D. pulchellus (Bigot et al.1997a, b). If one were looking for evolutionary intermediates between ascoviruses

and ichnoviruses, this would be a type that would be expected.

14.4.3 Molecular Evidence for the Evolution of Ascovirusesfrom Iridoviruses

As noted above, the molecular evidence that ichnovirus particles evolved from

ascoviruses is very limited. We therefore first discuss the data that exist for the

evolution of the ascoviruses from iridoviruses. These data provide an important


foundation for the ascovirus > ichneumonid particle hypothesis because ascoviruses

differ so much from iridoviruses in their cytopathology and morphology of their

virions. Thus, if ascoviruses, which recall are transmitted by parasitoids, evolved

from iridoviruses, the possibility that ichnoviruses evolved from ascoviruses, where

at least the changes in virion structure are less substantial, becomes more plausible.

The molecular evidence that ascoviruses evolved from iridoviruses is based on

analyses of four proteins that occur among a diversity vertebrate and invertebrate ds

DNA viruses. These proteins are the major capsid protein, DNA polymerase,

thymidine kinase, and ATPase III. Our analyses, performed using Parsimony and

Neighbor-Joining programs, indicate all these evolved from the same virus ancestor

(Stasiak et al. 2000, 2003). Although there are variations in the topologies of the

trees that emerged from our analyses of these proteins, two significant patterns are

apparent. First, ascoviruses and iridoviruses are more closely related to each other

than to the algal or vertebrate viruses in this viral lineage. Second and more

significantly, the TK and ATPase trees show the lepidopteran Chilo iridovirus

(CIV) clustering more closely with ascoviruses than with any of the vertebrate

iridioviruses (Stasiak et al. 2000, 2003). That the CIV and ascovirus MCP do not

cluster on the same branch is not surprising given the marked differences in virion

shape (Fig. 14.2). Another important feature that emerged from these analyses is

that the ascoviruses that are mechanically vectored by wasps, i.e., SfAV-1a,

TnAV-2a, and HvAV-3a, cluster together on one branch of the ascovirus tree,

whereas DpAV-4a, which is vertically transmitted by its wasp host, is found on a

separate branch. This difference correlates with the important difference in biology,

specifically, the more intimate association that DpAV-4a has with its wasp vector.

In summary, while the data indicating ascoviruses evolved from iridoviruses must

be considered preliminary, as the genes analyzed represent a small portion of those

encoded by these viruses, the results are nevertheless important because they reflect

patterns consistent with the biology of virus transmission by parasitic wasps.

More recent molecular studies, specifically the sequencing of the DpAV-4a

genome, suggest that in fact the ichneumonid particles may well have originated

from an ancestral iridovirus. We noted above that the ichneumonid, E. vitticolle, aparasite of noctuid caterpillars, is both capable of transmitting and being infected by

an iridovirus (Lopez et al. 2002). Annotation of the DpAV-4a genome shared more

core genes with lepidopteran iridoviruses than the more common, highly patho-

genic ascoviruses, e.g., SfAV-1, TnAV-2, and HzAV-3 (Bigot et al. 2009). These

findings again illustrate the need for more genomic sequence data on iridoviruses

and ascoviruses that infect lepidopteran insects.

14.4.4 Molecular Data Supporting an Iridovirus/AscovirusOrigin for Ichneumonid Particles

Though the molecular evidence at this stage is minimal, and despite the findings

regarding the DpAV-4 genome noted above, BLAST results obtained with several


ORFs in this genome provide evidence that certain ichnovirus ORFs have their

closest relatives in ascovirus genomes. Specifically, we identified a 13 kbp region

that contains a cluster of three genes (Fig. 14.3; ORF90, 91, and 93; Bigot et al.

2008) that have close homologs in a GfIV gene family composed of seven members

(Lapointe et al. 2007). All contain a domain similar to a conserved domain found in

the pox-D5 family of NTPases. To date, this pox-D5 domain has been identified as a

NTP binding domain of about 250 amino acid residues found only in viral proteins

encoded by poxvirus, iridovirus, ascovirus, and mimivirus genomes. These genes

seem to be specific to GfIV, as they are absent in the three sequenced genomes of

other ichnoviruses, namely CsIV, Tranosema rostrales ichnovirus (TrIV), and

Hyposoter fugitivus ichnovirus (HfIV).More specifically, in DpAV-4, ORF90 encodes a protein of 925 amino acid

residues that is 40 similar from position 140 to 925 to a protein of 972 amino acid

residues encoded by the ORF1 contained in the segment C20 in the GfIV genome.

These two proteins can therefore be considered putative orthologs. The 480

C-terminal residues of this DpAV-4 protein are also 42 similar to the C-terminal

domain of the protein homologs encoded by the ORF1 of the D1 and D4 GfIV

segments, 36 similar to the N-terminal and the C-terminal domains of the protein

encoded by the ORFs 184R and 128L of the iridovirus CIV and LCDV, and 30

similar with those encoded by ORFs 119, 99, and 78 in the ascovirus genomes of

HvAV-3e, SfAV-1a, and TnAV-2c, respectively. Overall, this indicates that this

DpAV-4 protein is more closely related to that of GfIV than to those found in other

ascovirus and iridovirus genomes currently available in databases. ORF091

encodes a protein of 161 amino acid residues similar only with the C-terminal

domain of three proteins encoded by the ORFs 1, 1, and 3, contained, respectively,

in GfIV segments D1, D4, and D3. In contrast, ORF93 is closer to iridovirus and

ascovirus genes than to GfIV genes. This protein of 849 amino acid residues is 43

similar over all its length to CIV ORF184R orthologs in all iridoviral and ascoviral

genomes and is only 36 similar over 350 amino acid residues to the C-terminal

domain of the GfIV protein homologs encoded by the ORF1, 2, 1, 1, 1, and 1 in,

respectively, the C20, C21, D1, D2, D3, and D4 segments of this virus.

Since the three DpAV-4 genes have relatives in all ascovirus and iridovirus

genomes sequenced so far, their presence in the DpAV-4 genome cannot result

Fig. 14.3 Map of the 13-kbp region of the DpAV4 genome (EMBL Acc. No. CU469068 and

CU467486) that contains the gene cluster with direct homologs in the genome of the Glyptafumiferanae ichnovirus. DpAV-4 ORF with well-characterized direct homologs among other

ascovirus and iridovirus genomes are represented by white arrows. Homologous ORF of the

GfIV genes are represented by black arrows (from Bigot et al. 2008). Below, the graph is scaled

in kbp


from a lateral transfer that occurred from an ichnovirus genome related GfIV to

DpAV-4. Thus, as these DpAV-4 genes are the closest relatives of the pox-D5 gene

family present in GfIV identified so far, they could be considered a landmark of the

symbiogenic ascovirus origin of the ichnovirus lineage to which this polydnavirus

belongs. An alternative explanation is that the presence of DpAV-4-like genes in the

genome of GfIV resulted from a lateral transfer from viral genomes closely related

to those of GfIV and DpAV-4. Indeed, this might have happened when a Glyptawasp was infected by an ancestral virus related to DpAV-4. Nevertheless, the

symbiogenic origin of GfIV from ascoviruses is also supported by morphological

features of its virions (Lapointe et al. 2007), which, aside from similarities in shape,

also show reticulations on their surface in negatively stained preparations, a charac-

teristic of the virions of all ascovirus species examined to date (Federici et al. 2005).

14.4.5 Relationships Between Ascovirus Virion and IchneumnidParticle Proteins

Because ascovirus virions and ichnovirus particles display structural similarities,

we developed an approach to search for homologs of virion structural proteins in

ichnoviruses. To date, only two virion proteins from the Campoletis sonorensisichnovirus (CsIV) have been characterized (Webb et al 2006). The first is the P44, a

structural protein that appears to be located as a layer between the out envelope and

nucleocapsid, and the second, P12, a capsid protein. Presently, there are more than

one hundred ascoviral or iridoviral MCP sequences in databases. BLAST searches

using these sequences failed to detect any similarities between CsIV virion proteins

and ascoviral or iridoviral MCPs, or any other proteins. To evaluate the possibility

that homology between ichnovirus and ascovirus virion proteins may simply not be

detectable by conventional Blastp searches, we used a different method, WAPAM

(weighted automata pattern matching). The models were designed on the basis of a

previous study (Stasiak et al. 2003) demonstrating that MCP encoded by ascovirus,

iridovirus, phycodnavirus, and asfarvirus genomes are related, and all contain seven

conserved domains separated by hinges of very variable size. We investigated these

conserved domains further using hydrophobic cluster analysis. This analysis

revealed that most conservation occurred at the level of hydrophobic residues, as

expected for structural proteins. The size variability of the hinges between con-

served domains and the conservation of hydrophobic residues might explain why

BLAST searches using iridoviral and ascoviral MCP sequences have limited ability

to detect MCP orthologs in phycodnavirus and asfarvirus genomes. We designed

two syntactic models which together were able to specifically align all MCP

sequences of the four virus families. Importantly, WAPAM aligned the CsIV

ichnovirus P44 structural protein with both models. Complementary structural

and HCA confirmed the presence of the seven conserved domains in this CsIV

structural protein (Fig. 14.4a).


In addition to the above analysis, ten syntactic models were developed using

proteins conserved in the three sequenced ascovirus species (SfAV-1a, TnAV-2c,

and HvAV-3a) and twelve iridoviruses. None of these models detected homologs

among ichnovirus proteins available in databases, except for one, developed from

small proteins encoded by the DpAV-4 ORF041, SfAV1a ORF061, HvAV-3a

ORF74, and TnAV-2c ORF118 in the ascovirus genomes, and iridovirus CIV

ORF347L and mimivirus MIV ORF096R genomes, respectively. Importantly,

these proteins have orthologs in vertebrate iridoviruses, phycodnaviruses, and

asfarvirus. In SfAV1a, the peptide encoded by ORF061 is one of the virion

components. In ascoviruses, iridoviruses, phycodnaviruses, and the asfarvirus,

Fig. 14.4 Sequence (lanes 1–3) and secondary structure (lanes 4–6) comparisons among (a) MCP

and (b) SfAV1a ORF061 orthologs fromCsIV (lanes 1 and 4, typed in black), DpAV4 (lanes 2 and 5,typed in blue), and SfAV1a (lanes 3 and 6, typed in purple). Conserved positions among the amino

acid sequence of CsIV and those of DpAV4 and SfAV1a are highlighted in gray. Secondarystructures in the three SfAV1a ORF061 orthologs were calculated with the Network Protein

Sequence Analysis at http://npsa-pbil.ibcp.fr/ website and the statistical relevance of the secondarystructures were evaluated with Psipred at http://bioinf.cs.ucl.ac.uk/psipred/ website. C, E, and H in

lanes 4–6 respectively indicated for each amino acid that it is involved in a coiled, b sheet, or a

helix structure. Using default parameters of Psipred, upper case letters indicate that the predicted

secondary structure is statically significant in Psipred results. Significant secondary structures are

highlighted in yellow. In (a), the comparisons were limited to three of the seven conserved

domains, 2, 5, and 7. Indeed, classical in silico methods appeared to be inappropriate to predict

statistically significant secondary structures in conserved structural protein rich in b strand such as

iridovirus and ascovirus major capsid proteins. In contrast, a complete and coherent domain

comparison was obtained by HCA profiles (see Bigot et al. 2008)


http://npsa-pbil.ibcp.fr/

http://bioinf.cs.ucl.ac.uk/psipred/

they have been annotated as thioredoxines, proteins that play a role in initiating

viral infection. Database mining with our model revealed four hits with CsIV

sequences (Acc N�. M80623, S47226, AF236017, AF362508) each a homolog

2a. Conservation, translocations and losts of the Ascovirus genes

2b. Translocation, duplication and diversification of host genes in the proviral genome of Ascoviral origine.

3a. Resulting proviral Ichnovirus genomes (monolocus solution)

3b. Resulting proviral Ichnovirus genomes (multilocus solution obtainedafter fragmention of the proviral genome by recombination)

1. Chromosomal integration of an Ascovirus genome in ancestors wasp genome of the Banchinae and Campopleginae lineages.

Fig. 14.5 Hypothetical mechanism for the integration and evolution of ascovirus genomes in

endoparasitic wasps. Schematic representation of the three-step process of symbiogenesis, and

DNA rearrangements that putatively occurred in the germ line of the wasp ancestors in the

Banchinae and Campopleginae lineages, from the integration of an ascoviral genome to

the proviral ichnoviral genome. Sequences that originate from the ascovirus are in blue, those ofthe wasp host and its chromosomes are in pink. Genes of ascoviral origin are surrounded by a thinblack or white line, depending on their final chromosomal location. Two solutions can account for

the final chromosomal organization of the proviral ichnovirus genome, monolocus or multilocus,

since this question is not fully understood in either wasp lineage. More complex alternatives to this

three-step process might also be proposed and would involve, for example, the complete de novo

creation of a mono or multi locus proviral genome from the recruitment by recombination or

transposition of ascoviral and host genes located elsewhere in the wasp chromosomes. This model

for the chromosomal organization of proviral DNA in polydnaviruses is consistent with published

data (Desjardins et al. 2007)


ORF of SfAV-1a ORF061. In fact, these sequences correspond to several variants

of a single region contained in the B segment of the CsIV genome. To date, these

have not been annotated in the final CsIV genome, probably because they overlap a

recombination site. HCA analyses confirmed that the hydrophobic cores were

conserved (Fig. 14.4b).

Confirmation of the apparent relationship of iridoviruses, ascoviruses, and the

ichneumonid particles awaits the sequencing of more of the viral genomes and

sequencing of the wasp genes that code for at least the structural proteins that make

up the ichneumonid particles. Nevertheless, the significant biological relationships

of endoparasitc ichneumond wasps with iridoviruses, ascoviruses, and their cater-

pillar hosts, and especially the unique relationship of DpAV-4 with its vector,

provide all the reagents for the development of symbiotic relationships that lead

to symbiogenesis. The evolutionary progression of these relationships, and the

benefits certain lineages of symbiotic viruses provided the wasps, and the likely

account for the origin of ichneumonid (and braconid) particles. In Fig. 14.5, we

illustrate a possible evolutionary scenario and mechanism that may have yielded the

interesting immunosuppressive organelles.

Table 14.1 Examples of viruses vertically transmitted by parasitoids and their possible viral

origins

Virus Evolutionary

origin

Parasitoid

family

Parasitoid

host

Reference

Produce virions in parasitoid’s host

Diadromus pulchellusascovirusa

Iridovirusc Ichneumonidae Lepidoptera Bigot et al. 1997a

Diachasmimorphalongicaudata poxvirusa

Poxvirus Braconidae Diptera Lawrence 2002

Microctonus aethiopoidesvirusa

Ascovirusc Braconidae Coleoptera Barratt et al.

1999

Cotesia melonoscela virus Ascovirusc Braconidae Lepidoptera Stoltz et al. 1988

Cotesia marginiventrisnudivirus

Nudivirus Braconidae Lepidoptera Hamm et al. 1990

Microplitis croceipes nudivirus Nudivirus Braconidae Lepidoptera Hamm et al. 1988

Diadromus pulchelluscypovirusb

Reovirus Ichneumonidae Lepidoptera Rabouille et al.

1994

Diachasmimorphalongicaudata rhabdovirusb

Rhabdovirus Braconidae Diptera Lawrence and

Akin 1990

No virions produced in parasitoid’s host

Campoletis sonorensisichnovirus

Ascovirusc Ichneumonidae Lepidoptera Webb et al. 2000

Cotesia marginiventrisbracovirus

Nudivirusc Braconidae Lepidoptera Webb et al. 2000

Bathyplectes anurus virus Poxvirusc Ichneumonidae Coleoptera Hess et al. 1980aInvolved in immunosuppressionbRNA viruscAncestral viruses from which the respective parasitic particles originated


14.5 Examples of the Diversity of Immunosuppressive

Wasp Viruses and Organelles

While the focus here has been on the origin and evolution of braconid and

ichneumonid particles, there are several other known endoparasitic wasp/virus

associations that range from symbiotic (i.e., involving true viruses) to organelles

that likely originated from viruses. These associations, along with several others

that have been discussed above, are listed in Table 14.1 to show the diversity of

these relationships, most of which have received very little study. Of particular

interest are the ascoviruses and poxviruses that replicate in both the parasitoid and

its insect host, produce progeny virions, and play a role in immunosuppression.

These include the D. pulchellus ascovirus, D. longicaudata entomopoxvirus, the

pox-like particles of Bathyplectes anurus, an ichneumonid parasite of a coleop-

teran, and the asco-like “virus” of M. aethiopoides, a braconid parasite of a

coleopteran.

14.6 Summary

During the last 100 million years, the genomes of at least two different types of

DNA viruses were integrated into the genomes of, respectively, endoparasitic

braconid and ichneumonid wasps. These viral genes thus became part of the wasp

genome. Over time, many of the original viral genes were deleted from the DNA

packaged into the virions and replaced by wasp genes involved in suppressing the

immune response of their caterpillar hosts, thereby transforming the original virions

into a novel type of transducing immunosuppressive organelle that enhanced the

survival of wasp progeny. The principal original viral genes that were selectively

maintained in a functional state in the wasp genomes were those involved in

producing critical structural proteins and enzymes essential for organelle assembly

and trafficking wasp immunosuppressive genes into caterpillar host cells and nuclei

for transcription. There are marked structural differences between the braconid and

ichneumonid organelles and their transducing wasp DNAs, yet their common role

in immunosuppression demonstrates a high degree of convergent evolution. This

relatively recent example of symbiogenesis through which two DNA viruses

evolved into immunosuppressive organelles likely accounts for much of the species

radiation characteristic of endoparasitic braconids and ichneumonids, two of the

largest groups of higher eukaryotic organisms.

Acknowledgments This research was supported by grants from the CNRS and the N.A.T.O. to

Y. Bigot, and U.S. National Science Foundation Grant INT-9726818 to B. A. Federici. The

photographs used in Fig. 14.1 are by D.B. Stoltz, of Dalhouise University, Halifax, Canada.


References

Barratt BIP, Evans AA, Stoltz DB, Vinson SB, Easingwood R (1999) Virus-like particles in the

ovaries of Microctonus aethiopoides Loan (Hymenoptera: Braconidae), a parasitoid of adult

weevils (Coleoptera: Curculionidae). J Invertebr Pathol 73:182–188

Bedwin O (1979a) The particulate basis of the resistance of a parasitoid to the defense reaction of

its insect host. Proc Biol Sci 205:267–270

Bedwin O (1979b) An insect glycoprotein; a study of the particles responsible for the resistance of

a parasitoids egg to the defense reactions of its insect hosts. Proc Biol Sci 205:271–286

Bezier A, Annaheim M, Herbiniere J, Wetterwald C, Gyapay G, Bernard-Samain S, Wincker P,

Roditi I, HellerM, BelghaziM, Pfister-WilhemR, Periquet G, DupuyC, Juguet E, Volkoff A-N,

Lanzrein B, Drezen J-M (2009) Polydnaviruses of braconid wasps derive from an ancestral

nudivirus. Science 323:926–930

Bigot Y, Rabouille A, Sizaret P-Y, Hamelim M-H, Periquet G (1997a) Particle and genomic

characterisation of a new member of the Ascoviridae, Diadromus pulchellus ascovirus. J GenVirol 78:1139–1147

Bigot Y, Rabouille A, Doury G, Sizaret P-Y, Delbost F, Hamelim M-H, Periquet G (1997b)

Biological and molecular features of the relationships between Diadromus pulchellus ascov-

irus, a parasitoid hymenopteran wasp (Diadromus pulchullus) and its lepidopteran host,

Acrolepiosis assectella. J Gen Virol 78:1149–1163

Bigot Y, Samain S, Auge-Gouillou C, Federici BA (2008) Molecular evidence for the evolution of

ichnoviruses from ascovirsues by symbiogenesis. BMCEvol Biol. doi:10.1186/1471-2148-8-253

Bigot Y, Renault S, Nicolas J, Moundras, C, Demattei MV, Semain S, Bideshi DK, Federici BA

(2009) Symbiotic virus at the evolutionary intersection of three types of large DNA viruses:

Iridoviruses, Ascoviruses, and Ichnoviruses. PloS One doi:10.1371/journal.pone.000639

Burand JP (1998) Nudiviruses. In: Miller LK, Bell LA (eds) The insect viruses. Plenum Press,

New York, pp 69–90

Chinchar VG, Essbauer S, He JG, Hyatt A, Miyazaki T, Seligy V, Williams T (2005) Family

Iridoviridae. In: Fauquet CM, Mayo MA, Maniloff J, Desselberger U, Ball LA (eds) Virus

taxonomy: eight report of the international committee on virus taxonomy. Elsevier/Academic

Press, London, pp 145–162

Deng L, Stoltz DB, Webb BA (2000) A gene encoding a polydnavirus structural polypeptide is not

encapsidated. Virology 269:440–450

Desjardins CA,Gundersen-RindalDE,Hostetler JB, TallonLJ, Fuester RW, SchatzMC, PedroniMJ,

Fadrosh DW, Haas BJ, Toms BS, Chen D, Nene V (2007) Structure and evolution of a proviral

locus of Glyptapanteles indiensis bracovirus. BMC Microbiol. doi:10.1186/1471-2180-7-61

Desjardins CA, Gundersen-Rindal DE, Hostetler JB, Tallon LJ, Fadrosh DW, Fuester RW,

Pedroni MJ, Haas BJ, Schatz MC, Jones LM, Crabtree J, Forberger H, Nene V (2008)

Comparative genomics of mutualistic viruses of Glyptapanteles parasitic wasps. Genome

Biol. doi:10.1186/gb-2008-9-12-r183

Edson KM, Vinson SB, Stoltz DB, Summers MD (1981) Virus in a parasitoid wasp: supression of

the cellular immune response in the parasitoid’s host. Science 211:582–583

Espagne E, Dupuy C, Huguet E, Cattolico L, Provost B, Martins N, Poire M, Periquet G, Drezen

JM (2004) Genome sequence of a polydnavirus: insights into symbiotic virus evolution.

Science 306:286–289

Federici BA (1983) Enveloped double stranded DNA insect virus with novel structure and

cytopathology. Proc Natl Acad Sci USA 80:7664–7668

Federici BA (1991) Viewing polydnaviruses as gene vectors of endoparasitic hymenoptera. Redia

74:387–392

Federici BA (1993) Viral pathology in relation to insect control. In: Beckage NE, Thompson SN,

Federici BA, (eds) Parasites and Pathogens of Insects, Vol 2, Academic Press, New York,

pp 81–101


Federici BA, Bigot Y (2003) Origin and evolution of polydnaviruses by symbiogenesis of insect

DNA viruses in endoparasitic wasps. J Insect Physiol 49:419–432

Federici BA, Bigot Y, Granados RR, Hamm JJ, Miller LK, Newton I, Stasiak K, Vlak JM (2005)

Family Ascoviridae. In: Fauquet CM, Mayo MA, Maniloff J, Desselberger U, Ball LA (eds)

Taxonomy of virus taxonomy: eight report of the international committee on virus taxonomy.

Elsevier/Academic Press, London, pp 269–274

Hamm JJ, Nordlung DA, Marti OG (1985) Effects of a nonoccluded virus of Spodoptera frugi-perda (Lepidoptera: Noctuidae) on the development of a parasitoid, Costesia marginiventris(Hymenoptera: Braconidae). Environ Entomol 14:258–261

Hamm JJ, Styer EL, Lewis WJ (1988) A baculovirus pathogenic to the parasitoid Microplituscroceipes (Hymenoptera: Braconidae). J Invertebr Pathol 52:189–191

Hamm JJ, Styer EL, Lewis WJ (1990) Comparative virogenesis of filamentous virus and poly-

dnavirus in the female reproductive track of Cotesia marginiventris (Hymenoptera: Braconi-

dae). J Invertebr Pathol 55:357–360

Hess RT, Poinar GO Jr, Etzel L, Merritt CC (1980) Calyx particle morphology of Bathyplectesanurus and B. curculionis (Hymenoptera: Ichneumonidae). Acta Zoo (Stockholm)

61:111–114

Khakhina LN (1992) Concepts of symbiogenesis. In: Margulis L, McMenamin M (eds) Historical

and critical study of the research of Russian botanists. Yale University Press, New Haven

Lapointe R, Tanaka K, Barney WE, Whitfield JB, Banks JC, Beliveau C, Stoltz D, Webb BA,

Cusson M (2007) Genomic and morphological features of a banchine oplydnavirus: compari-

son with bracoviruses and ichnoviruses. J Virol 81:6491–6501

Lawrence P (2002) Purification and partial characterization of an entomoposvirus (DLEPV) from

a parasitic wasp of tephritid fruit flies. J Insect Sci 2:10

Lin C-L, Lee JC, Chen SS, Wood HA, Li M-L, Li C-F, Chao Y-C (1999) Persistent Hz-1 virus

infection in insect cells: evidence for insertion of viral DNA into host chromosomes and viral

infection in a latent status. J Virol 73:128–139

Lopez M, Rojas JC, Vandame R, Williams T (2002) Parasitoid mediated transmission of an

iridescent virus. J Invertebr Pathol 80:160–170

Margulis L (1992) Biodiversity: molecular biological domains, symbiosis and kingdom origins.

Biosystems 27:39–51

Margulis L, Fester R (1991) Symbiosis as a source of evolutionary innovation. MIT Press,

Cambridge Massachusetts

Rabouille A, Bigot Y, Drezen JM, Sizaret P-Y, Hamelin M-H, Periquet G (1994) A member of the

reoviridae (DpRV) has a ploidy-specific genomic segment in the wasp Diadromus pulchellus(Hymenoptera). Virology 205:228–237

Rotheram S (1967) Immune surface of eggs of a parasitic insect. Nature 214:700

Rotheram S (1973a) The surface of the egg of a parasitic insect. I. The surface of the egg and first

instar larvae of Nemeritis. Proc Biol Sci 183:179–194Rotheram S (1973b) The surface of the egg of a parasitic insect. IL. The ultrastructure of the

particulate coat on the egg of Nemeritis. Proc Biol Sci 183:195–204Salt G (1965) Experimental studies in insect parasitism XIII. The haemocytic reaction of a

caterpillar to the eggs of its habitual parasite. Proc Biol Sci 162:303–318

Salt G (1966) Experimental studies in insect parasitism XIII. The haemocytic reaction of a

caterpillar to the eggs of its habitual parasite. Proc Biol Sci 165:155–178

Salt G (1968) The resistance of insect parasitoids to the defense reactions of their hosts. Biol Rev

43:200–232

Schmidt O, Schuchmann-Feddersen I (1989) Role of virus-like particles in parasitoid-host inter-

action of insects. Subcell Biochem 15:91–119

Schmidt O, Theopold U (1991) Immune defense and suppression in insects. BioEssays 13:343–346

Schmidt O, Theopold U, Strand M (2001) Innate immunity and its evasion and suppression by

hymenopteran endoparasitoids. BioEssays 23:344–351


Stasiak K, Demattei M-V, Federici BA, Bigot Y (2000) Phylogenetic position of the DpAV-4a

ascovirus DNA polymerase among viruses with a large double-stranded DNA genome. J Gen

Virol 81:3059–3072

Stasiak K, Renault S, Demattei MV, Bigot Y, Federici B (2003) Evidence for the evolution of

ascoviruses from iridoviruses. J Gen Virol 84:2999–3009

Stoltz DB (1993) The polydnavirus life cycle. In: Beckage NE, Thompson SN, Federici BA (eds)

Parasites and pathogens of insects, vol 1. Academic Press, New York, pp 167–187

Stoltz DB, Faulkner G (1978) Apparent replication of an unusual virus-like particle in both a

parasitoid wasp and its host. Can J Microbiol 24:1509–1514

Stoltz DB, Vinson SB (1979) Viruses and parasitism in insects. Adv Virus Res 24:125–171

Stoltz DB, Krell P, Summers MD, Vinson SB (1984) Polydnaviridae – a proposed family of insect

viruses with segmented, double-stranded, circular DNA genomes. Intervirology 21:1–4

Stoltz DB, Krell PJ, Cook D, MacKinnon EA, Lucarotti CJ (1988) An unusual virus from the

parasitic wasp Cotesia melanoscela. Virology 162:311–320

Tanaka K, Lapointe R, Narney WE, Makkay AM, Stoltz D, Cusson M, Webb BA (2007) Shared

and species-specific features among ichnovirus genomes. Virology 263:26–35

Vinson SB (1972) Factors involved in successful attack on Heliothis virescens by the parasitoid

Cardiochiles nigriceps. J Invertebr Pathol 20:118–123Vinson SB (1990) How parasitoids deal with the immune system of their host: an overview. Arch

Insect Biochem Physiol 13:2–27

Vinson SB, Scott JR (1975) Particles containing DNA associated with the oocyte of an insect

parasitoid. J Invertebr Pathol 25:375–378

Wang Y, Jehle JA (2009) Nudiviruses and other large, double-stranded circular DNA viruses of

invertebrates: new insights into an old topic. J Invertebr Pathol 101:187–193

Webb BA, Beckage NE, Hayakawa Y, Lanzrein B, Stoltz DB, Strand MR, Summers MD (2005)

Family Polydnaviridae. In: Fauquet CM, Mayo MA, Maniloff J, Desselberger U, Ball LA (eds)

Virus taxonomy: eight report of the international committee on virus taxonomy. Elsevier/

Academic Press, London, pp 255–265

Webb BA, Strand MR, Dickey SE, Beck MH, Hilgarth RS, Barney WE, Kadash K, Kromer JA,

Lindstrom KG, Rattanadechakul E, Shelby KS, Thoetkiattikul H, Turnbull MS, Witherell RA

(2006) Polydnavirus genomes reflect their dual roles as mutualists and pathogens. Virology

347:160–174

Whitfield JB (2002a) Estimating the age of the polydnavirus/braconid wasp symbiosis. Proc Natl


Whitfield JB (2002b) Phylogeny of microgastroid braconid wasps, and what it tells us about

polydnavirus evolution. In: Austin AD, Dowton M (eds) Hymenoptera, evolution, biodiversity,

and biological control. CSIRO Publishing, Collingswood, Australia, pp 97–105


Chapter 15

The Neogastropoda: Evolutionary Innovations

of Predatory Marine Snails with Remarkable

Pharmacological Potential

Maria Vittoria Modica and Mande Holford

Abstract The Neogastropoda include many familiar molluscs, such as cone snails

(Conidae), purple dye snails (Muricidae), mud snails (Nassariidae), olive snails

(Olividae), oyster drills (Muricidae), tulip shells (Fasciolariidae), and whelks (Bucci-

nidae). Due to their amazing predatory specializations, neogastropods are often

dominantmembers of the benthic community at the top of the food chain. In a dazzling

display that ranges from boring holes to darting harpoons, neogastropods have

developed several prey hunting innovations with specialized compounds pharmaceu-

tical companies could only dream about. It has been hypothesized that evolutionary

innovations related to feeding were the main drivers of the rapid neogastropod

radiation in the late Cretaceous. The anatomical, behavioral, and biochemical specia-

lizations of neogastropod families that are promising targets in drug discovery

and development are addressed within an evolutionary framework in this chapter.

15.1 Introduction

15.1.1 The Neogastropoda

Neogastropoda is an order of gastropod molluscs that are well characterized mor-

phologically and are traditionally viewed as monophyletic (Ponder 1973; Taylor

and Morris 1988; Ponder and Lindberg 1996, 1997; Kantor 1996; Strong 2003).

M.V. Modica

Dipartimento di Biologia Animale e dell’Uomo, “La Sapienza”, University of Rome, Viale

dell’Universita 32, 00185 Rome, Italy


M. Holford

The City University of New York – York College & Graduate Center, and The American Museum

of Natural History, 94–20 Guy R. Brewer Blvd, Jamaica, NY 11451, USA



249

This characterization of the Neogastropoda persists even after contrasting inter-

pretations have been proposed (see e.g., Colgan et al. 2007; Kantor and Fedosov

2009). Strong (2003) has recently provided the most updated report of potential

neogastropod synapomorphies. Anatomical characteristics of neogastropods

include a very peculiar anterior foregut with a proboscis (pleurembolic or intraem-

bolic), a valve of Leiblein, a gland of Leiblein (or a venom gland in Toxoglossa),

paired primary and accessory salivary glands, an anal gland, and several radular

peculiarities (Ponder 1973; Kantor 2002; Strong 2003). Figure 15.1 illustrates a

generalized scheme of neogastropod anatomy.

The order Neogastropoda includes up to 25 families (Bouchet and Rocroi 2005)

traditionally split into three superfamilies, Cancellarioidea, Conoidea, and Muri-

coidea, on the basis of anatomical features of the anterior foregut, including the

radula. Cancellarioidea, also called Nematoglossa, comprised of the single family

Cancellariidae, is perceived to be the basal offshoot of neogastropods (Kantor 1996;

Strong 2003; Oliverio and Modica 2009; Modica et al. 2009). They are character-

ized by a nematoglossan radula with a complex mechanism of interlocking of the

distal cusps (viewed as an adaptation to suctorial feeding: Petit and Harasewych

1986) and a mid-oesophageal gland that is generally not separated from the

oesophagous (Fig. 15.2a). Conoidea, also referred to as Toxoglossa, include Con-

idae, Terebridae, and the “turrid” which are estimated to have more than 10,000

extant species, and whose taxonomy is under revision (Puillandre et al. 2008). In

Conoidea, the radula is modified in various degrees until forming a harpoon

(toxoglossan radula), and the dorsal mid-oesophageal gland is separated from the

oesophagous and develops into a venom apparatus, with a muscular bulb and a

secretory tubule producing neurotoxins (Fig. 15.2b). Muricoidea (also termed

Rachiglossa) include the vast majority of neogastropod families, whose monophyly

is currently debated (Kantor 1996, 2002; Oliverio and Modica 2009). The muri-

coidean radula is rachiglossate (Fig. 15.2c) and their anatomy is similar to the

generalized model proposed in Fig. 15.1, but there are many modifications at

different taxonomic levels. Variations include the presence/absence of radula,

accessory salivary glands, valve and gland of Leiblein, anal gland and a number

of other foregut, renal, and reproductive features.

According to the fossil record, the adaptive radiation of neogastropods has been

particularly rapid (Taylor et al. 1980) and may be attributed to the evolution of a

predatory lifestyle and diversification in a number of different trophic strategies.

Such attributes allowed neogastropods to fully diversify their niches and to effici-

ently exploit their alimentary resources. In this scenario, the evolutionary role

played by chemical innovations in feeding is unquestionable.

The Cancellarioidea, Conoidea, and Muricoidea possess a bountiful reservoir of

bioactive compounds routinely used to sedate or capture prey. These compounds

are the building blocks for future drug discovery targets. Outlined in this chapter are

the anatomical features, specialty feeding strategies, and potential bioactive com-

pounds found in the families of the Neogastropoda. Specific attention is given to the

discovery and characterization of bioactive compounds from the Conoidea. Based

on the successful characterization and implementation of cone snail toxins in

250 M.V. Modica and M. Holford

250

pharmacological approaches (Favreau and Stocklin 2009; Twede 2009; Olivera and

Teichert 2007; Fox and Serrano 2007), several groups within the Neogastropoda are

highlighted as potential biodiversity targets for drug discovery.

15.1.2 Discovery and Characterization of Cone Snail Toxins

The gold standard for investigating toxins from marine snails is the discovery

and characterization of neurotoxins from cone snails (Conus) (Fig. 15.2b). Thisextremely diversified group of marine snails comprises active predators that use

biochemical substances to subdue their prey. Characterization of cone snail toxins

begun almost a half century ago (Kohn 1956; Kohn et al. 1960; Endean et al. 1974),

starting from empirical observations of envenomation episodes, and has blossomed

into a successful research field (review; Norton and Olivera 2006). The characteri-

zation of conotoxins provides scientists with new, powerful tools to manipulate the

function of ion channels and receptors governing the physiology of the nervous

Fig. 15.1 Generalized

scheme of neogastropod

anatomy (male). Mantle

longitudinally dissected, body

wall not shown.

Abbreviations are as follows:

a anus; ag anal gland; asgaccessory salivary gland; ctctenidium; dg digestive

gland; ft foot; hghypobranchial gland; lg glandof Leiblein; lv valve ofLeiblein; mo mouth; opoperculum; os osphradium;

pe penis; pg prostate gland;

pr proboscis; sd salivary duct;sg salivary gland; st stomach;

t testis. Modified after Ponder

(1998a)

15 The Neogastropoda: Evolutionary Innovations of Predatory Marine Snails 251

251

system. The pharmacological usage of ion channels and receptors as drug develop-

ment targets for the treatment of neurological and cardiovascular diseases is rapidly

gaining momentum. The discovery of Prialt (Ziconotide) (Miljanich 2004), the

synthetic form of the Conus magus peptide o-conotoxin MVIIA, an N-type calcium

channel blocker, significantly highlight the potential of toxins from marine snails.

Prialt was approved by the Food and Drug Administration of the United States in

December 2004 for analgesic use in HIV and cancer patients.

Although Prialt is a significant breakthrough, Conus represents only a very small

fraction of the diversity of Neogastropoda. Conus is one of the 20–30 recognized

neogastropod families and includes ca 4–500 species out of 10–15,000 estimated in

the Conoidea (Bouchet and Rocroi 2005). The pharmacological potential of neo-

gastropods as a source for bioactive compounds is largely unrealized. Similar to

cone snails, several other neogastropods have evolved specialized compounds as

a result of their feeding ecology that may have potential in pharmacological

applications.

Fig. 15.2 The Neogastropoda radiation. Three major families of the Neogastropoda are shown:

(a) Cancellarioidea, (b) Conoidea, and (c) Muricoidea. The grey triangles shown are proportional

to the number of species included in each lineage. Shown for each superfamily are radula, scheme

of the foregut, and some shell representatives. Shells shown, from left to right, by genus: (a)

Scalptia. (b) Conus, Terebra, Thatcheria, Gemmula. (c) Murex, Oliva, Vexillum, Melongena,Cymbiola, Fusinus, Volutopsius. (d) Schematic arrangement of the foregut (modified after Kantor

1996). Shell images courtesy of Guido and Philippe Poppe. Radula pictures courtesy of Yuri

Kantor (b) and Alisa Kosyan (c).


252

15.2 Feeding Strategies in the Neogastropoda

From what is known about the diets of neogastropod families, the vast majority of

neogastropods are carnivorous, with a degree of predatory activity that varies

from actively seeking prey to grazing on sessile invertebrates, to scavenging.

Some neogastropod families, such as Buccinidae and Muricidae, include many

generalist species, which can feed on a variety of living and dead organisms. Most

Muricidae feed on living bivalves, gastropods, polychaetes, bryozoans, sipuncu-

lids, barnacles, and other small crustaceans, but there are a few that also feed on

carrions. A species of Drupa has been observed feeding also on holothurians (Wu

1965), while Drupella (Ergalataxinae) and all Coralliophilinae feed on corals

(Taylor 1976; Ward 1965; Haynes 1990) (Fig. 15.4a). Some neogastropod

families appear to be highly specialized, such as the Mitridae, which feed exclu-

sively on sipunculids (Taylor et al 1980) and possess peculiar anatomical adapta-

tions to this kind of prey (Harasewych 2009). An interesting feeding strategy is

also displayed by the Volutidae, which has been reported for feeding on bivalves,

gastropods, and in some deep-water species, on echinoderms (Darragh and Ponder

1998). Members of the Volutidae use their large foot to engulf the prey in a

semiclosed environment, in which anesthetic substances are apparently released

(Bigatti et al. 2009). Described in the following paragraphs are neogastropod

feeding strategies that involve bioactive substances that may have pharmacologi-

cal utility.

15.2.1 Harpooning

Cone snails, terebrids, and turrids make up the superfamily Conoidea (or Tox-

oglossa, “poisoned tongued”). Toxoglossans are a megadiverse group of hunting

snails where the rapid evolution of venom peptide genes has led to an amazing

molecular diversity. They feed on molluscs, polychaetes, acorn worms, and fish

(Kohn 1959, 1968; Kohn and Nybakken 1975; Leviten 1980). The key evolution-

ary innovations enabling conoideans to hunt preys are a conspicuous venom

apparatus made up of highly modified radular teeth (harpoon), a venom duct

(a glandular duct connected to the oesophagous), and a muscular venom bulb

(Fig. 15.2b). The radular tooth, held at the proboscis tip, is inserted into the

prey and dispensed similar to a hypodermic needle (Olivera 2002). The mecha-

nism of envenomation involves the contraction of the muscular venom bulb,

which forces the secretion of the venom duct through the proboscis, until reaching

the tooth. A single cone snail specimen may produce between 50 and 200 dif-

ferent peptides, which are known to target different ion channels (Terlau and

Olivera 2004).


253

15.2.2 Shell Drilling

Shell drilling is the most common feeding technique in muricids, and it is achieved

by the concerted action of the radula and a specialized glandular pad (the accessory

boring organ) placed on the foot sole (Carriker 1961) (Fig. 15.3a). The drilling

process may last up to 1 week (Palmer 1990; Dietl and Herbert 2005). Drilling is not

restricted to muricids and has been observed in other rachiglossans, such as the

marginellid genus Austroginella (Ponder and Taylor 1992), the buccinid Cominella(Peterson and Black 1995), and the nassariid Nassarius festivus (Morton and Chan

1997). Other feeding strategies developed by the muricids include the opening of

the prey shell with the foot (Wells 1958), the cracking of the shells close to the

apertural margin followed by proboscis insertion (Radwin and D’Attilio 1976) and

the use of shell projections on outer lip (labial spines) to force the opening of the

valves (Marko and Vermeij 1999).

15.2.3 Shell Wedging and Proboscis Insertion

As noted above, drilling has been reported for a few species of Buccinidae, but the

majority of buccinids use the strengthened margin of their shells to wedge open

bivalve shells (Nielsen 1975), in order to insert their proboscis (Fig. 15.3b).

Buccinidae eat polychaetes, small crustaceans, and some species have been

observed feeding on peculiar preys, e.g., Neptunea antiqua on priapulids, Taylor

1978). Buccinds can also insert their proboscis into the aperture of gastropod

shells. Similar strategies of proboscis insertion with mild radular rasping or use

of shell margins have been reported in families related to buccinids, such as: the

Nassariidae, which feed on polychaetes, barnacles and carrion; the Fasciolariidae,

which feed on bivalves, gastropods, sedentary polychaetes, and carrions; the

Melongenidae, which feed on gastropods and bivalves; and the Columbellidae,

which feed on ascidians, hydroids, small crustaceans, polychaetes, and algae

(Taylor et al. 1980).

15.2.4 Suctorial Feeding

Suctorial feeding, or sucking the innards of prey organisms, is an evolutionary

advanced feeding technique demonstrated by several neogastropod families. This

form of feeding does not always result in the death of the prey, and several

neogastropod species coexist with the prey. Two kinds of suctorial feedings are

described: haematophagy and corallivory.


254

15.2.4.1 Haematophagy

Three different neogastropod families, Cancellariidae, Marginellidae, and Colu-

brariidae, have independently evolved haematophagous feeding on fish

(Fig. 15.3c). The buccinoidean family Colubrariidae includes at least six species

involved in a parasitic association with different species of fish, mainly belonging

to the family Scaridae (Johnson et al. 1995; Bouchet and Perrine 1996). Colubrariaspecimens can extend their proboscis to a length exceeding three times the shell

length. When the extended Colubraria proboscis is in contact with the skin of the

prey, a scraping action with its minute radula allows access to the blood vessels of

the fish. The snail then apparently takes advantage of the blood pressure of the fish

to ingest its meal (Oliverio and Modica 2009). Experimental observations on

different Colubraria species (Modica and Oliverio, unpublished) suggest that

adaptation to haematophagy involves the use of anesthetic and anticoagulant

compounds. In fact, the fish appears to be anesthetized when the snail is feeding.

Anesthetization is reversible, and the fish usually recovers its full mobility in a few

minutes after the interruption of the contact with the snail. The anesthetic com-

pounds used are not lethal as the prey recovers, in agreement with field observations

Fig. 15.3 Examples of neogastropod feeding strategies. (a) An ocinebrine Muricidae drilling the

shell of a venerid bivalve (photo G. Herbert). (b) A Muricanthus sp. (Muricidae) using the shell

margin to wedge open a bivalve shell (photo G. Herbert). (c) Colubraria muricata (Colubrariidae)feeding on a clownfish in aquarium; the proboscis is inserted under the pectoral fins (photo

M. Oliverio). (d) Coralliophila meyendorffi (Coralliophilinae) feeding on Actinia equina (photo

P. Mariottini)


255

that Colubraria usually feed on fish sleeping in crevices of the reef (M. Oliverio

pers. comm.; Bouchet and Perrine 1996; Johnson et al. 1995).

A similar strategy has been reported for the cancellariid Cancellaria cooperi(Cancellarioidea), which has been observed using its proboscis to ingest blood from

open injuries on the body of the electric ray Torpedo californica (O’Sullivan et al.

1987). Cancellariidae are likely to include exclusively suctorial feeders, as inferred

from foregut and radular characteristics. Dissection of Cancellaria cooperi evi-denced a peculiar oesophageal structure (M.V. Modica, J. Biggs, and M. Holford,

unpublished observations). In fact, the mid oesophagous is extremely long (up to 5

times the shell length) and glandular, similar to what is found in Colubraria,suggesting a convergent adaptation to haematophagy. Other examples of haemato-

phagous feeding are the very minute species of Marginellidae, Kogomea ovata,Hydroginella caledonica, and Tateshia yadai, that live attached to the pectoral fins

of their host (Kosuge 1986; Bouchet 1989).

15.2.4.2 Corallivory

Feeding on the living tissues of corals and other Anthozoans is reported in

Muricidae for Drupella (Ergalataxinae) and for the subfamily Coralliophilinae

(Taylor 1976; Ward 1965; Haynes 1990). Coralliophilinae includes over 200

marine tropical to temperate species, from shallow to deep waters. The few species

for which alimentary preferences are known (about 10% of the shallow water

species, Oliverio et al. 2008) feed exclusively on anthozoans (Fig. 15.3d). A variety

of feeding strategies and preferences are displayed for this group. Some species are

stenophagous, with very strict host specificity; they are mostly sessile on corals,

and many groups have developed interesting eco–morphological adaptations. In

fact, while Quoyula has a limpet-like shell suitable for external life on stony corals,

Rhizochilus lives and feeds on anthipatharians with the shell deformed to adhere to

the black coral branch. A second group lives embedded in the host skeleton: Rapalives inside alcyonarian octocorals, Magilopsis and Leptoconchus have ovoid

shells and bore holes into corals, while Magilus is sessile inside corals and

possesses an uncoiled adult shell (Robertson 1970). Some others are mobile as

Latiaxis, which is probably associated with deep-water gorgonians, or Babelo-murex that mostly feeds on shallow water hexacorals. In a few cases mobile

euryphagous species can feed on anthozoans belonging to different orders, such

as some species of Coralliophila associated with sea anemones, scleractinians,

and zoanthids (M. Oliverio, unpublished observations). Among coralliophilines

some anatomical modifications related to parasitism on corals are widespread, such

as the loss of the radula and jaws, viewed as an adaptation to suctorial feeding, and

brooding of embryos in capsules kept in the pallial cavity (Richter and Luque

2002).

The amazing display of feeding strategies developed by neogastropods is possi-

ble due to the diversity of innovative anatomical features and chemical compounds

that can be readily employed to overcome their prey.


256

15.3 Neogastropod Specialized Anatomy and Predatory

Chemical Substances

Most neogastropod snails have developed specialized glands or other anatomical

features that enable them to produce and use chemical substances to subdue their

prey. It can be argued that the development of specialized foregut glands, such as

the venom gland in Conoidea, or salivary and accessory salivary glands in other

neogastropod groups, has lead to the successful radiation of neogastropods. The

biochemical weaponry developed in the foregut and other glands is an evolutionary

advantage that has enabled neogastropods to thrive.

15.3.1 Foregut Glands

The foregut glands described here include the venom gland, primary, and accessory

salivary glands (Figs. 15.1 and 15.2). Toxins may be produced in a specific venom

gland, as is the case with most Conoideans, or in primary and/or accessory salivary

glands (Andrews 1991) for species that do not have a venom gland. In some cases,

the production of toxins might involve other foregut organs/tissues, such as the

glandular mid-oesophagous of the haematophagous Colubraria and Cancellaria.

15.3.1.1 Venom Gland

The presence of a venom apparatus is characteristic of the Conoidea (Fig. 15.2b).

Generally it is a conspicuous organ, constituted by a proximal muscular bulb and a

very long, convolute duct (the gland itself). The tubular gland always passes

through the nerve ring and opens into the buccal cavity, posterior to the radular

sac opening. The active exocrine secretion of the venom is due to a single cell type:

cuboidal ciliated cells, accumulating venom granules at their apex, until they are

discharged into the lumen (Smith 1967). The venom gland may be lined with such

secretory cells for its whole length or, as happens in some species, the secretory

tissue may be confined to the region posterior to the nerve ring, while the anterior-

most region is a simple ciliated duct (Taylor et al. 1993). The terminal muscular

bulb is usually constituted by two muscular layers, internal and external, separated

by connective tissue; the relative thickness and development of these layers is

variable between species. According to Ponder (1973) the tubular venom gland

originated from the dorsal glandular folds of the oesophagous while the gland of

Leiblein gave rise to the muscular bulb. Some conoideans, mostly radula-less

species, do not possess a venom apparatus.

All cone snails (Conus) have a venom apparatus and the toxins found in their

venom glands have led the field in characterizing peptide toxins from marine snails.

When venom is injected into a prey, the conotoxins work in a concerted manner to


257

shut down the prey’s nervous system. Conotoxins are potent neurotoxins that target

ion channels and receptors. The complement of peptides found in any one Conusvenom is strikingly different from that found in the venom of any other Conusspecimens (Romeo et al. 2008). Thus, in the whole genus, many tens of thousands

of distinct active peptides have evolved. A question that immediately arises is why

individual cone snails should need somany different peptides. It has been speculated

that the complement of peptides in a venom may be used for at least three general

purposes: An individual peptide may play a role in (1) prey capture, directly or

indirectly; (2) defense and escape from predators; or (3) other biological processes,

such as interaction with potential competitors. Not all terebrids and turrids have a

venom apparatus, but those that do also produce toxins to subdue their prey. Unlike

conotoxins, less is known about terebrid and turrid toxins, teretoxins and turritoxins,

respectively. Preliminary characterization of terebrid and turrid toxins (Imperial

et al. 2003, 2007; Watkins et al. 2006; Heralde et al. 2008) indicate a similar three-

domain conotoxin structure consisting of a highly conserved signal sequence, amore

variable pro-region, and a hypervariable mature toxin sequence. While conotoxins

have been identified as potent neuropeptides, no known molecular target has been

identified for teretoxins or turritoxins. However, given their similarities to conotox-

ins it is expected they will also be effective modifiers for ion channels and receptors

in the nervous system.

15.3.1.2 Primary Salivary Glands

Primary salivary glands are usually acinous, with a very small lumen and a system

of narrow branched ducts (Fig. 15.1). In some species, the paired glands may be

fused together in a single glandular mass, but two salivary ducts are always present

and run along the oesophagous (or, in some groups, embedded in the oesophageal

walls) until opening into the roof of the buccal cavity. Two cell types have been

identified in the secretory epithelium, mixed with one another: (1) basal cells with

apocrine secretion and (2) superficial ciliated cells secreting mucus (Andrews

1991). Ciliary movement is responsible for delivering the secretion, as the outer

layer of muscle fibers is poorly developed (Andrews 1991). Acinous salivary glands

are present in all neogastropod, although their role in toxin production may be

variable, depending on whether other secreting structures, such as venom gland or

accessory salivary glands, are present.

Only acinous salivary glands are present in Buccinidae and related families, such

as Nassariidae, Melongenidae, Fasciolariidae, and Columbellidae (accessory sali-

vary glands are missing). Species of the buccinid genus Neptunea (as e.g.,

N. antiqua) have very large salivary glands containing high quantity of tetramine

(F€ange 1960; Asano and Itoh 1959, 1960; Saitoh et al. 1983; Fujii et al. 1992;

Shiomi et al. 1994; Watson-Wright et al. 1992; Power et al. 2002), which blocks

nicotinic acetylcholine receptors (Emmelin and F€ange 1958). A number of human

intoxication has been reported so far, caused by consumption of snails of these

species (Fleming 1971; Millar and Dey 1987; Reid et al. 1988). Further studies have


258

shown the presence of three additional unidentified toxins in the salivary glands of

N. antiqua that appear to inhibit neuronal Ca2+ channels (Power et al. 2002). Other

whelks are known to produce histamine, choline, and choline esters (Endean 1972).

Nassariidae possess three types of secreting cells in their salivary glands, one of

which secretes a glycoprotein rich in disulphide groups like the accessory salivary

glands of the muricid Nucella lapillus (Fretter and Graham 1994; Minniti 1986;

Martoja 1964).

The finding that conopeptides are expressed in the salivary gland of Conuspulicarius (Biggs et al. 2008) suggests that salivary glands may play a role in the

envenomation process. Crude extracts of salivary glands of the haematophagous

Colubraria reticulata have been observed to increase coagulation time of human

blood (S. Rufini, M.V. Modica, and M. Oliverio, unpublished). Current research by

Modica and colleagues is underway to identify the anticoagulant transcript using

cDNA analysis.

15.3.1.3 Accessory Salivary Glands

Accessory salivary glands are considered to be an informative synapomorphy of

Neogastropoda, although they are missing in several families. Accessory salivary

glands are present in the basal family Cancellariidae (Fig. 15.2a) and in several

Toxoglossa, where in some vermivorous cones they coexist with the venom gland

(Marsh 1971). Two pairs of accessory salivary glands are also found in Muricidae,

Mitridae, Costellariidae, Volutidae, and Olividae, while in Volutomitridae only

one gland is found. In Marginellidae, Harpidae, and in the buccinoideans, acces-

sory salivary glands are generally missing, but are present in Busycon (Andrews

1991). A common anatomical organization of the glands is shared by all neogas-

tropods. The paired glands are tubular in shape, with a lumen lined by a columnar

secretory epithelium surrounded by a subepithelial muscular coat richly inner-

vated. External to the muscle layer there is an outer layer of gland cells, with long

necks opening in the central lumen of the gland (Ponder 1973; Andrews 1991)

producing a peculiar granular secretion (Andrews 1991). Exceptions to this model

include olives, volutids, and some mitriform species (Marcus and Marcus 1959;

Ponder 1970, 1972). The structure is very similar to the venom gland of Conoidea

(West et al. 1996). The glandular accessory salivary glands open at the tip of the

buccal cavity with nonciliated ducts.

In Muricidae, accessory salivary glands are usually large and well developed. In

Nucella lapillus and Stramonita haemastoma, the only muricids studied so far at the

biochemical level, accessory salivary glands produce a glycoprotein rich in

cysteines (Martoja 1971; McGraw and Gunter 1972), similar to conotoxins. Extracts

of the glands are able to elicit flaccid paralysis inMytilus eduliswhich can be drilledor not, and, in the case of S. haemastoma, in barnacles, which are never drilled

(Carriker 1981; Huang and Mir 1972; Andrews 1991; West et al. 1996; Andrews

et al. 1991). S. haemastoma also produces a toxic secretion in the primary salivary

glands that decreases cardiac activity in mammals and induces vasodilatation,


259

hypotension, and smooth muscle contraction (Huang and Mir 1972). A similar

response was demonstrated in a combined primary/accessory salivary glands extract

of another muricid, Acanthina spirata (Hemingway 1978). N. lapillus extracts alsodisrupt neuromuscular transmission in rat phrenic nerve–hemidiaphragm prepara-

tions (West et al. 1996). In some Volutidae, the accessory salivary glands have

been reported to produce a narcotizing compound, with a very low pH, inducing

muscular relaxation in the preys (Bigatti et al 2009).

15.3.2 Hypobranchial Gland

The hypobranchial gland is constituted by a thickening of the epithelium in the roof

of the pallial cavity and produces large amounts of mucus. Its primary function is

currently viewed to be the cleaning of the mantle cavity; the mucous secretion binds

together the particulate matter, which is then eliminated from the mantle cavity.

However, the hypobranchial gland comprises at least three different cell types that

may correspond to distinct chemical activities, which have only been partially

identified (Naegel and Aguilar-Cruz 2006). In many muricid species, the hypo-

branchial gland produces chromogens, which, exposed to light and oxygen, develop

into a purple pigment that has been used for centuries as a dye (Tyrian purple).

Similarly, in the Mitridae, the hypobranchial secretion once exposed to air becomes

yellowish, then purple, and finally dark brown (Harasewych 2009), while in Cost-

ellariidae it remains predominantly yellow-green (Ponder 1998b). The production

of small compounds, mainly choline esters, but also biogenic amines, has been

detected in the hypobranchial gland of several species of muricids and buccinids.

These substances elicit neuromuscular blocking, with paralyzing effects both in

invertebrates and vertebrates (Roseghini et al. 1996). Due to the low concentra-

tions in which these toxic compounds are found in the snails, it is not sure how

effective they are in prey hunting (West et al. 1996). The functions of the hypo-

branchial gland and the role it played in the evolution and diversification of the

Neogastropoda are still to be clarified; nevertheless, hypobranchial secretions may

have useful pharmacological properties.

15.4 Neurotoxins, Anesthetics, and Anticoagulants: Prominent

Bioactive Compounds from Neogastropod Snails

As stated in the introduction of this chapter, conotoxins, with the approval of the

analgesic drug Prialt, have demonstrated the utility of translating basic research of

marine snail compounds into drug development targets. The identification of novel

neurotoxins, anesthetics, and anticoagulants are three areas in which harvesting the

bioactive compounds of the Neogastropoda could prove very fruitful. The following


260

section highlights the success of conotoxins as neurotoxins and outlines the potential

of identifying anesthetic and anticoagulant compounds from neogastropod snails.

15.4.1 Neurotoxins

In the Conoidea, the best-characterized venom components are small, highly

structured disulfide peptides, individually encoded by a separate gene. Every

Conus species has its own distinct repertoire of 50–200 venom peptides, with

each peptide presumably having a physiologically relevant target in prey or poten-

tial predators/competitors (Olivera 2002). Most conotoxins are small peptides

(6–40 amino acids in length), with the majority being in the size range of 12–30

amino acids (Olivera et al. 1990; Terlau and Olivera 2004). Conotoxins are

comprised of a highly conserved precursor structure including a signal sequence,

followed by a propeptide region and then a mature toxin that is cleaved from the

prepro-structure. The mature toxins are highly disulfide rich and are classified

according to their cysteine framework. Cone snails practice combinatorial drug

therapy in that it is not one conotoxin that attacks the prey, but instead a cocktail of

the 50–200 venom peptides working together to shut down the prey’s nervous

system. The conotoxin cocktail contains ion channel and receptor modifiers that

can affect neuronal signaling. For example, conotoxins that inhibit Na+ channel

function prevent the formation of action potential, while conotoxins that target Ca2+

prevent vesicle fusion, which impedes the release of neurotransmitters. There are

presently more than 3,000 different Conus venom proteins reported in the literature

(Conoserver: http://research1t.imb.uq.edu.au/conoserver/). Less than 10% of the

described conotoxins have been functionally characterized. Of those characterized,

at least 25 different functions have been described (Olivera 2006; Conoserver).

Several conotoxins are at various stages of drug development with the more

promising examples being: MrIA (active on norepinephrine transporters), Vc1.1

(active on nicotinic receptors), and Conantokin-G (active on NMDA receptors)

(Olivera 2006). While the majority of conotoxins in therapeutic development are

analgesic compounds, conotoxins are also being considered as viable targets for

epilepsy or myocardial infarction, as well as disorders concerning neuroprotective/

cardioprotective properties (Twede et al. 2009).

Another promising group to investigate in order to discover new neurotoxins

and/or substances capable of inactivating toxins is the corallivorous subfamily

Coralliophilinae (Muricidae). The Anthozoa, such as sea anemones, and stony

and soft corals, which are included in the Cnidaria along with the jellyfishes

(Scyphozoa), sea-wasps (Cubozoa), hydrocorals, and hydromedusae (Hydrozoa),

are known to produce a neurotoxin-rich venom as well as other toxic defensive

compounds, from which the Coralliophilinae appear to be immune. Envenomation

by cnidarians represents a remarkable sanitary problem for humans. An estimated

40,000–50,000 marine envenomations occur annually due to several species of

Cnidaria. Cubozoan alone have been responsible for over 5,000 human deaths in


261

http://research1t.imb.uq.edu.au/conoserver/

the last 130 years (Brinkman and Burnell 2009). Antivenom is available only for a

very limited number of species. If, as is suggested by reported observations,

coralliophilines have antivenom-type compounds, they may potentially be useful

in cases of cnidarian envenomations. The immunity of Coralliophilinae raises a

number of interesting evolutionary questions, such as: What are the physiological

adaptations related to corallivory? Do corallivorous species secrete bioactive com-

pounds interacting with and inactivating anthozoans’ toxin? Are there specialized

organs involved in the production of the antivenom (e.g., salivary glands)? Is host

switching in euryphagous and host specificity in stenophagous correlated with

biochemical variations in the secretion? The answers to these questions may

translate into a modern physiological and biochemical understanding of gastropod

innovations related to feeding.

15.4.2 Anesthetic and Anticoagulant Compounds

As pointed out in Sect. 15.3, three different neogastropod families have haemato-

phagous species, which produce anesthetic and anticoagulant compounds that may

be useful in elucidating cellular communication in the nervous system and as

antithrombotic agents.

In Colubrariidae, anticoagulants are produced in the salivary glands, but the

anatomical structures responsible for anesthetic secretion are not yet known. In

addition to the salivary glands, it might be worthy to investigate the glandular

mid-posterior oesophagous, a peculiar derived structure that may be related to the

haematophagous lifestyle (Oliverio and Modica 2009). Furthermore, the peculiar

mid-oesophagous of Cancellaria cooperi is a very advantageous tissue to test for

bioactive compounds production, as cancellariid mid-oesophagous may be

homologous to toxoglossan venom glands (Ponder 1973; Kantor 1996, 2002).

Another issue of interest is the presence in Cancellariidae of both primary and

accessory salivary glands. The roles these anatomical structures play in prey

subduction and in the production of bioactive substances, as well as their inter-

actions, are still to be investigated. Are the bioactive substances the same in the

different haematophagous lineages? Intriguing evolutionary questions may be

addressed studying and comparing anticoagulant and anesthetic molecules in

Colubrariidae and Cancellariidae.

15.5 Investigating Genetic Evolution and Expression

of Neogastropod Toxins

The early evolution, and the first diversification of venom toxins, has been inter-

preted as the result of a process of neofunctionalization in which strong positive

selection acts on redundant genes produced in duplication events, originating new


262

functions (Ohno 1970). This evolutionary mechanism was reported also for con-

otoxins (Duda and Palumbi 1999). The evolutionary pressure promoting the varia-

bility of these “specialty genes” (also called exogenes, as their products act outside

the organism; Olivera 2006) is related with a predator–prey arms race process in

which the availability of a particular kind of prey may produce an evolutionary

force acting on ecologically important genetic loci. Conotoxins are particularly

prone to rapid genetic variations, due to their extremely reduced size. It is still

unclear at which level the results reported for Conus might be generalized in the

neogastropods, but it is plausible at least to hypothesize that the same organs

produce the same type of bioactive substances across the entire order Neogastro-

poda. According to the amount of variation that will be detected at the different

taxonomic levels in neogastropods, it will be possible to clarify the evolutionary

patterns acting at each level. In snakes, where the same neofunctionalization

mechanism is responsible for the evolution of the toxin gene families, the genes

that have been recruited to constitute the venom proteome have been partly

identified (Fry 2005). In neogastropods, including cone snails, the origin of the

toxin sequences has yet to be investigated.

The role of differential gene expression and posttranscriptional modifications

in modulating toxin diversity is also an intriguing area requiring further investi-

gation. This line of research could be addressed at different taxonomic levels: (1)

Between different species – a particular focus should be dedicated to host speci-

ficity, to verify if the inverse correlation between the degree of specialization and

the diversity of the venom in Conus leopardus (Remigio and Duda 2008) can be

generalized to other neogastropod groups. (2) In individuals of the same species –

the high levels of intraspecific variability observed in Conus ventricosus (Romeo

et al. 2008) raise the possibility that fine-scale modulatory mechanisms may act

in response to environmental and ecological variations. And (3) at different

ontogenetic stages – juvenile neogastropods have often a largely different diet

from the adults, implying a different suite of toxins. How and under which

mechanisms does venom composition change during ontogenesis? To address

these and other toxin evolution and expression topics, a robust phylogenetic

hypothesis and an integrated strategy for the characterization of bioactive com-

pounds are required.

15.6 Conclusion: Integrated Strategies for Building a More

Robust Evolutionary Framework and Effective Drug

Development Methods

The major challenges in characterizing bioactive compounds in snails are the

complexity of sampling, the scarcity of the biological material, and the absence

of databases for determination of peptide and protein sequences. Venom profiling

may thus prove an elusive target, unless molecular biology techniques are coupled


263

with biochemical analysis of polypeptide composition. A multidisciplinary plat-

form, combining modern genomic and proteomic techniques, as well as phylogeny

and descriptive approaches to ecology and anatomy, is necessary to increase the

rate of pharmacological characterization of new bioactive compounds. Genomic

libraries can be obtained from tissues of interest and their analysis can be integrated

with proteomic techniques, such as venom fractionation, peptide purification, mass

spectrometry, and sequence analysis using automated Edman degradation. Spider

venoms have recently been analyzed by a three-dimensional approach, combining

calculated, predicted, and measured data obtained with different techniques such as

cDNA sequences and LC-MALDI analysis (Escoubas et al. 2006). The use of such

“venom landscapes” may constitute a significant improvement in venom profiling

and can also be effective as molecular markers in taxonomic and phylogenetic

studies. A similar strategy has been applied to snake venoms (Nascimento et al.

2006). Molecular phylogeny, combined with anatomical and ecological data, can

guide us through the maze of snail biodiversity, toward the species or group of

species which are likely to possess bioactive compounds worthy to investigation to

find new therapeutics (Fig. 15.4). This strategy was successfully applied to the

Terebridae, outlining particular genera/species important for teretoxin discovery

(Holford et al. 2009a, b).

Research fields

Integrative approach

Output

Ecology

Integratedevolutionaryframework

Enhanceddrug

development

Anatomy &Physiology

Chemicalecology

Comparative phylogeny

Genomics & Proteomics

Phylogeny Pharmacology

Fig. 15.4 Integrated research strategies for investigating biodiversity. The integration of different

approaches to diversity may lead to a more complete evolutionary framework and enhance the rate

of drug discovery and development


264

Interestingly, the relationship between drug discovery and phylogeny is a two-

way street. In fact, exogenes mostly belong to gene superfamilies with highly

conserved sequence elements, enabling the use of standard molecular techniques.

In what has been called a “concerted discovery strategy” venom toxins are revealed

to be useful characters for the taxonomy and phylogenetic relationships of their

producers (Olivera 2006; Olivera and Teichert 2007; Bulaj 2008). This integrated

approach has been used in non-molluscan toxin-producing groups such as snakes to

garner insight into the molecular evolution of snake venoms and to correlate the

appearance of other morphological evolutionary novelties (Fry and W€uster 2004).For the Neogastropoda, whose phylogeny cannot be readily elucidated using

standard taxonomic approaches, an integrated approach has several possibilities.

Proteomics of the venom as well as the characterization of its biochemical and

functional properties successfully separated two closely related, morphological

indistinguishable pit-viper species (Angulo et al. 2007).

The use of genomic analysis and venom profiling techniques, along with more

traditional approaches such as anatomical and physiological studies, will allow a

better understanding of the correlation between venom composition, trophic pre-

ferences, and adaptive radiation of the Neogastropoda, creating the basis for a

modern integrated evolutionary framework and an effective drug discovery strategy

(Fig. 15.4).

Acknowledgments The authors thank Marco Oliverio for invaluable advice and helpful com-

ments on the manuscript. Yuri Kantor, Alisa Kosyan, Gregory Herbert, Paolo Mariottini, Marco

Oliverio, and Guido and Philippe Poppe are acknowledged for images used in the figures. MH

acknowledges support from NIH grant GM088096-01.

References

Andrews EB (1991) The fine structure and function of the salivary glands of Nucella lapillus(Gastropoda: Muricidae). J Moll Stud 57:111–126

Andrews EB, Elphick MR, Thorndyke MC (1991) Pharmacologically active constituents of the

accessory salivary and hypobranchial glands of Nucella lapillus. J Moll Stud 57:136–138

Angulo Y, Escolano J, Lomonte B, Gutierrez JM, Sanz L, Calvete JJ (2007) Snake venomics of

Central American pitvipers: clues for rationalizing the distinct envenomation profiles of

Atropoides nummifer and Atropoides picadoi. J Proteome Res 7(2):706–719

Asano M, Itoh M (1959) Occurrence of tetramine and choline compounds in the salivary gland of a

marine gastropod Neptunea arthritica (Bernardi). J Agric Res 10:209

Asano M, Itoh M (1960) Salivary poison of a marine gastropod, Neptunea arthritica Bernardi, andthe seasonal variation of its toxicity. Ann N Y Acad Sci 90:675–688

Bigatti G, Sanchez Antelo CJM, Miloslavich P, Penchaszadeh PE (2009) Feeding behavior of

Adelomelon ancilla (Lighfoot, 1786): a predatory neogastropod (Gastropoda: Volutidae) in

Patagonian benthic communities. The Nautilus 123(3):159–165

Biggs JS, Olivera BM, Kantor YI (2008) a-Conopeptides specifically expressed in the salivary

gland of Conus pulicarius. Toxicon 52:101–105

Bouchet P (1989) A marginellid gastropod parasitize sleeping fishes. Bull Mar Sci 45:76–84

Bouchet P, Perrine D (1996) More gastropods feeding at night on parrotfishes. Bull Mar Sci 59

(1):224–228


265

Bouchet P, Rocroi JP (2005) Classification and nomenclator of gastropod families. Malacologia 47

(1–2):1–397

Brinkman DL, Burnell JN (2009) Biochemical and molecular characterisation of cubozoan protein

toxins. Toxicon 54:1162–1173

Bulaj G (2008) Integrating the discovery pipeline for novel compounds targeting ion channels.

Curr Opin Chem Biol 12:441–447

Carriker MR (1961) Comparative functional morphology of boring mechanisms in gastropods.

Am Zool 1(2):263–266

Carriker MR (1981) Shell penetration and feeding by naticacean and muricacean predatory

neogastropods: a synthesis. Malacologia 20:403–422

Colgan DJ, Ponder WF, Beacham E, Macaranas JM (2007) Molecular phylogenetics of Caeno-

gastropoda (Gastropoda: Mollusca). Mol Phylogenet Evol 42(3):717–737

Conoserver: http://research1t.imb.uq.edu.au/conoserver/

Darragh TA, Ponder WF (1998) Family Volutidae. In: Beesley PL, Ross JGB, Wells A (eds)

Mollusca: the Southern synthesis. Fauna of Australia, vol 5. CSIRO Publishing, Melbourne,

pp 833–835, part B

Dietl GP, Herbert GS (2005) Influence of alternative shell-drilling behaviours on attack duration of

the predatory snail Chicoreus dilectus. J Zool 265:201–206Duda TFJ, Palumbi SR (1999) Molecular genetics of ecological diversification: duplication and

rapid evolution of toxin genes of the venomous gastropod Conus. Proc Natl Acad Sci USA

96:6820–6823

Emmelin N, F€ange R (1958) Comparison between biological effects of neurine and a salivary

glands extract of Neptunea antiqua. Acta Zool 39:47–52Endean R (1972) Aspects of molluscan pharmacology. In: Florkin M, Scheer BT (eds) Chemical

zoology, vol 7, Mollusca. Academic Press, New York, pp 421–466

Endean R, Parrish G, Gyr P (1974) Pharmacology of the venom of Conus geographus. Toxicon12:131

Escoubas P, Sollod B, King GF (2006) Venom landscapes: mining the complexity of spider

venoms via a combined cDNA and mass spectrometric approach. Toxicon 47:650–663

F€ange R (1960) The salivary gland of Neptunea antiqua. Ann N Y Acad Sci 90:689–694

Favreau P, Stocklin R (2009) Marine snail venoms: use and trends in receptor and channel

neuropharmacology. Curr Opin Pharmacol 9:594–601

Fleming C (1971) Case of poisoning from red whelks. Br Med J 3:250–251

Fox JW, Serrano SM (2007) Approaching the golden age of natural product pharmaceuticals from

venom libraries: an overview of toxins and toxin-derivatives currently involved in therapeutic

or diagnostic applications. Curr Pharm Res 13:2927–2934

Fretter V, Graham A (1994) British prosobranch molluscs. Revised and updated edition, Ray

Society, London

Fry BG (2005) From genome to “venome”: molecular origin and evolution of the snake venom

proteome inferred from phylogenetic analysis of toxin sequences and related body proteins.


Fry BG, W€uster W (2004) Assembling an arsenal: origin and evolution of the snake venom

proteome inferred from phylogenetic analysis of toxin sequences. Mol Biol Evol 21

(5):870–883

Fujii R, Moriwaki N, Tanaka K, Ogawa T, Mori E, Saitou M (1992) Spectrophotometric determi-

nation of tetramine in carnivorous gastropods with tetrabromophenolphthalein ethyl ester. J

Food Hyg Soc Japan 33(3):237–240

Harasewych MG (2009) Anatomy and biology of Mitra cornea Lamarck, 1811 (Mollusca,

Caenogastropoda, Mitridae) from the Azores. Acoreana 6:121–135

Haynes JA (1990) Distribution movement and impact of the corallivorous gastropod Coralliophilaabbreviata (Lamarck) in a Panamanian patch. J Exp Mar Biol Ecol 142:25–42

Hemingway GT (1978) Evidence for a paralytic venom in the intertidal snail Acanthina spirata(Neogastropoda: Thaisidae). Comp Biochem Physiol 60C:79–81


266

http://research1t.imb.uq.edu.au/conoserver/

Heralde FM, Imperial J, Bandyopadhyay P, Olivera BM, Concepcion GP, Santos AD (2008) A

rapidly diverging superfamily of peptide toxins in venomous Gemmula species. Toxicon

51:890–897

Holford M, Puillandre N, Modica MV, Watkins M, Collin R, Bermingham E, Olivera BM (2009a)

Correlating molecular phylogeny with venom apparatus occurrence in panamic auger snails

(Terebridae). PLoS ONE 4(11):e7667. doi:10.1371/journal.pone.0007667

Holford M, Puillandre N, Terryn Y, Cruaud C, Olivera BM, Bouchet P (2009b) Evolution of the

Toxoglossa venom apparatus as inferred by molecular phylogeny of the Terebridae. Mol Biol

Evol 26(1):15–25

Huang CL, Mir GN (1972) Pharmacological investigation of salivary gland of Thais haemastoma(Clench). Toxicon 10:111–117

Imperial JS, Watkins M, Chen P, Hillyard DR, Cruz LJ, Olivera BM (2003) The augertoxins:

biochemical characterization of venom components from the toxoglossate gastropod Terebrasubulata. Toxicon 42:391–398

Imperial JS, Kantor YI, Watkins M, Heralde FM, Stevenson B, Chen P, Hansson K, Stenflo J,

Ownby J-P, Bouchet P, Olivera BM (2007) Venomous auger snail Hastula (Impages) hectica(Linnaeus, 1758): molecular phylogeny, foregut anatomy and comparative toxinology. J Exp

Zool 308B:744–756

Johnson S, Johnson J, Jazwinski S (1995) Parasitism of sleeping fish by gastropod mollusks in the

Colubrariidae and Marginellidae at Kwajalein, Marshall Islands. The Festivus 27(11):121–126

Kantor YI (1996) Phylogeny and relationships of Neogastropoda. In: Taylor J (ed) Origin and

evolutionary radiation of the Mollusca. Oxford University Press, Oxford, pp 221–230

Kantor YI (2002) Morphological prerequisite for understanding neogastropod phylogeny. Boll

Malacol Suppl 4:161–174

Kantor YI, Fedosov A (2009) Morphology and development of the valve of Leiblein: possible

evidence for paraphyly of the Neogastropoda. The Nautilus 123(3):73–82

Kohn AJ (1956) Piscivorous gastropods of the genus Conus. Proc Natl Acad Sci USA 42:168–171

Kohn AJ (1959) The ecology of Conus Hawaii. Ecol Monogr 29:47–90

Kohn AJ (1968) Microhabitats, abundance and food of Conus (Gastropoda) on atoll reefs in the

Maldive and Chagos islands. Ecology 49:1046–1062

Kohn AJ (1978) Ecological shift and release in an isolated reefs: the significance of prey size.

Ecology 59:614–631

Kohn AJ, Nybakken JW (1975) Ecology of Conus on eastern Indian ocean fringing reefs: diversityof species and resource utilization. Mar Biol 29:211–234

Kohn AJ, Saunders PR, Wiener S (1960) Preliminary studies on the venom of the marine snail

Conus. Ann N Y Acad Sci 90:706–725

Kosuge S (1986) Description of a new species of ecto-parasitic snail on fish. Bull Inst Malacol 2

(5):77

Leviten PJ (1980) The foraging strategy of vermivorous conid gastropods. Ecol Monogr

46:157–178

Marcus E, Marcus E (1959) Studies on Olividae. Bol Fac Fil Cienc Let Univ S Paulo Zool

22:99–188

Marko PB, Vermeij GJ (1999) Molecular phylogenetics and the evolution of labral spines among

eastern pacific ocenebrine gastropods. Mol Phylogenet Evol 13(2):275–288

Marsh M (1971) The foregut glands of some vermivorous cone shells. Aust J Zool 19:313–326

Martoja M (1964) Contribution a l’etude de l’appareil digestif et la digestion chez les gasteropodes

carnivores de la famille Nassarides. Cell 64:237–334

Martoja M (1971) Donnees histologiques sur les glandes salivaires et oesophagiennes de Thaislapillus (L.) (¼ Nucella lapillus. Prosobranche Neogastropode) Arch Zool Exp Gen

112:249–291

McGraw KA, Gunter G (1972) Observations on killing of the Virginia oyster by the Gulf oyster

borer, Thais haemastoma, with evidence for a paralytic secretion. Proc Natl Shellfish Assoc

62:95–97


267

Miljanich GP (2004) Ziconotide: neuronal calcium channel blocker for treating severe chronic

pain. Curr Med Chem 11:3029–3040

Millar JG, Dey A (1987) Food poisoning due to the consumption of red whelks Neptunea antiqua.Comm Dis Scotl Wkly Rep 21(38):5–6

Minniti F (1986) Morphological and histochemical study of pharynx of Leiblein, salivary glands

and gland of Leiblein in the carnivorous Gastropoda Amyclina tinei Maravigna and Cyclopeneritea Lamarck (Nassariidae: Prosobranchia Stenoglossa). Zool Anz 217:14–22

Modica MV, Kosyan A, Oliverio M (2009) The relationships of the enigmatic gastropod Trito-noharpa: new data on early neogastropod evolution? The Nautilus 123(3):177–188

Morton B, Chan K (1997) The first report of shell-boring predation by a representative of the

Nassariidae (Gastropoda). J Moll Stud 63:480–482

Naegel LCA, Aguilar-Cruz CA (2006) The hypobranchial gland from the purple snail Plicopur-pura pansa (Gould, 1853) (Prosobranchia, Muricidae). J Shellfish Res 25(2):391–394

Nascimento DG, Rates B, Santos DM, Verano-Braga T, Barbosa-Silva A, Dutra AAA, Biondi I,

Martin-Euclaire MF, De Lima ME, Pimenta AMC (2006) Moving pieces in a taxonomic

puzzle: venom 2D-LC/MS and data clustering analyses to infer phylogenetic relationships in

some scorpions from the Buthidae family (Scorpiones). Toxicon 47:628–639

Nielsen C (1975) Observations on Buccinum undatum L. attacking bivalves and on prey responses,

with a short review on attacking methods of other prosobranchs. Ophelia 13:87–108

Norton RS, Olivera BM (2006) Conotoxins down under. Toxicon 48:780–798

O’Sullivan JB, McConnaughey RR, Huber ME (1987) A blood-sucking snail: the Cooper’s

nutmeg Cancellaria cooperi Gabb, parasitizes the California electric ray, Torpedo californicaAyres. Biol Bull 172:362–366

Ohno S (1970) Evolution by gene duplication. Springer, Berlin

Olivera BM (2002) Conus venom peptides: Reflections from the biology of clades and species.

Annu Rev Ecol Syst 33:25–47

Olivera BM (2006) Conus peptides: biodiversity-based discovery and exogenomics. J Biol Chem

281:31173–31177

Olivera BM, Teichert RW (2007) Diversity of the neurotoxic Conus peptides: a model for

concerted pharmacological discovery. Mol Interv 7(5):253–262

Olivera BM, Rivier J, Clark C, Ramilo CA, Corpuz GP, Abogadie FC, Mena EE, Woodward SR,

Hillyard DR, Cruz LJ (1990) Diversity of Conus neuropeptides. Science 249:257–263Oliverio M, Modica MV (2009) Relationships of the haematophagous marine snail Colubraria

(Rachiglossa, Colubrariidae), within the neogastropod phylogenetic framework. Zool J Linn

Soc. 158:779–800

Oliverio M, Barco A, Modica MV, Richter A, Mariottini P (2008) Ecological barcoding of

corallivory by ITS2 sequences: hosts of coralliophiline gastropods detected by the cnidarian

DNA in their stomach. Mol Ecol Resour 9(1):94–103

Palmer AR (1990) Effect of crab effluent and scent of damaged conspecifics on feeding, growth,

and shell morphology of the Atlantic dogwhelk, Nucella lapillus (L.). Hydrobiologia

193:155–182

Peterson CH, Black R (1995) Drilling by buccinid gastropods of the genus Cominella in Australia.The Veliger 38:37–42

Petit RE, HarasewychMG (1986) New Philippine Cancellariidae (Gastropoda: Cancellariacea), with

notes on the fine structure and function of the nematoglossan radula. The Veliger 28(4):436–443

Ponder WF (1970) The morphology of Alcithoe arabica (Mollusca: Volutidae). Malacol Rev

3:127–165

Ponder WF (1972) The morphology of some mitriform gastropods with special reference to their

alimentary and reproductive system (Neogastropoda). Malacologia 11(2):295–342

Ponder WF (1973) The origin and evolution of the Neogastropoda. Malacologia 12:295–338

Ponder WF (1998a) Infraorder Neogastropoda. In: Beesley PL, Ross JGB, Wells A (eds) Mol-

lusca: the Southern synthesis. Fauna of Australia, vol 5. CSIRO Publishing, Melbourne, p 819

part B


Ponder WF (1998b) Family Costellariidae. In: Beesley PL, Ross JGB, Wells A (eds) Mollusca: the

Southern synthesis. Fauna of Australia, vol 5. CSIRO Publishing, Melbourne, pp 843–845,

part B

Ponder WF, Lindberg DR (1996) Gastropod phylogeny – challenges for the 90s. In: Taylor J (ed)

Origin and evolutionary radiation of the Mollusca. Oxford University Press, London,

pp 135–154

Ponder WF, Lindberg DR (1997) Towards a phylogeny of gastropod molluscs: an analysis using

morphological characters. Zool J Linn Soc 119:83–265

Ponder WF, Taylor JD (1992) Predatory shell drilling by two species of Austroginella (Gastro-

poda: Marginellidae). J Zool 228:317–328

Power AJ, Keegan BF, Nolan K (2002) The seasonality and role of the neurotoxin tetramine in the

salivary glands of the red whelk Neptunea antiqua L. Toxicon 40:419–425

Puillandre N, Samadi S, Boisselier M-C, Sysoev AV, Kantor YI, Cruaud C, Couloux A, Bouchet P

(2008) Starting to unravel the toxoglossan knot: molecular phylogeny of the “turrids” (Neo-

gastropoda: Conoidea). Mol Phylogenet Evol 47:1122–1134

Radwin GE, D’Attilio A (1976) Murex shells of the world. Stanford University Press, Stanford

Reid TMS, Gould IM, Mackie IM, Ritchie AH, Hobbs G (1988) Food poisoning due to the

consumption of red whelks Neptunea antiqua. Epidemiol Infect 101:419

Remigio EA, Duda TFJ (2008) Evolution of ecological specialization and venom of a predatory

marine gastropod. Mol Ecol 17:1156–1162

Richter A, Luque AA (2002) Current knowledge on Coralliophilidae (Gastropoda) and phyloge-

netic implication of anatomical and reproductive characters. Boll Malacol 38:5–19

Robertson R (1970) Review of the predators and parasites of stony corals, with special reference to

symbiotic prosobranch gastropods. Pac Sci 24:43–54

Romeo C, Di Francesco L, Oliverio M, Palazzo P, Raybaudi Massilia G, Ascenzi P, Polticelli F,

Schinina ME (2008) Conus ventricosus venom peptides profiling by HPLC-MS: a new insight

in the intraspecific variation. J Sep Sci 31:488–498

Roseghini M, Severini C, Falconieri Erspamer G, Erspamer V (1996) Choline esters and biogenic

amines in the hypobranchial gland of 55 molluscan species of the neogastropod Muricoidea

superfamily. Toxicon 34(1):33–55

Saitoh H, Oikawa K, Takano T, Kamimura K (1983) Determination of tetramethylammonium ion

in shellfish by ion chromatography. J Chromatogr 281:397

Shiomi K, Mizukami M, Shimakura K, Nagashima Y (1994) Toxins in the salivary gland of some

marine carnivorous gastropods. Comp Biochem Physiol 107B:427–432

Smith EH (1967) The neogastropod midgut, with notes on the digestive diverticula and intestine.

Trans R Soc Edinburgh 67:23–42

Strong EE (2003) Refining molluscan characters: morphology, character coding and a phylogeny

of the Caenogastropoda. Zool J Linn Soc 137:447–554

Taylor JD (1976) Habitats, abundance and diets of muricacean gastropods at Aldabra Atoll. Zool J

Linn Soc 59:155–193

Taylor JD (1978) Habitats and diet of predatory gastropods at Addu Atoll, Maldives. J Exp Mar

Biol Ecol 31:83–103

Taylor JD, Morris NJ (1988) Relationships of neogastropoda. Malacol Rev 4:167–179

Taylor JD, Morris NJ, Taylor CN (1980) Food specialization and the evolution of predatory

prosobranch gastropods. Palaentology 23(2):375–409

Taylor JD, Kantor YI, Sysoev AV (1993) Foregut anatomy, feeding mechanisms, relationships

and classification of the Conoidea (¼Toxoglossa) (Gastropoda). Bull Br Mus Nat Hist

59:125–170

Terlau H, Olivera BM (2004) Conus venoms: a rich source of novel ion channel-targeted peptides.

Pysiol Rev 84:41–68

Twede VD, Miljanich GP, Olivera BM, Bulaj G (2009) Neuroprotective and cardioprotective

conopeptides: an emerging class of drug leads. Curr Opin Drug Discov Dev 12:231–239


Ward J (1965) The digestive tract and its relation to feeding habits in the stenoglossan prosobranch

Coralliophila abbreviata (Lamarck). Can J Zool 43:447–464

Watkins M, Hillyard DR, Olivera BM (2006) Genes expressed in a turrid venom duct: divergence

and similarity to conotoxins. J Mol Evol 62:247–256

Watson-Wright WM, Sims GG, Smyth C, Gillis M, Maher M, Trottier T, Van Sinclair DE,

Gilgan M (1992) Identification of tetramine as toxin causing food poisoning in Atlantic

Canada following consumption of whelks Neptunea decemcostata. In: GopalakrishnakoneP, Tan CK (eds) Recent advances in toxinology research, vol 2. University of Singapore,

Singapore, pp 551–561

Wells HW (1958) Feeding habits of Murex fulvescens. Ecology 39:556–558

West DJ, Andrews EB, Bowman D, McVean AR, Thorndyke MC (1996) Toxins from some

poisonous and venomous marine snails. Comp Biochem Physiol 113C:l–10

Wu SK (1965) Comparative functional studies of the digestive system of the muricid gastropods

Drupa ricina and Morula granulata. Malacologia 3:211–233


Chapter 16

Antennal Hammers: Echos of Sensillae Past

Nina Laurenne and Donald L.J. Quicke

Abstract Many hosts of parasitoids live in concealed environments such as within

plants tissue and wood, and therefore they are difficult to find. This is likely to be

especially true when concealed hosts are in the pupal stage and thereby silent and

immobile. Cryptine ichneumonids collectively have a wide host range including

members of several insect orders with different degrees of concealment. Many

cryptine genera show a morphological adaptation to finding concealed hosts; their

antennal tips are modified into a hammer-like structures that are used to tap the

substrate. This vibrational sounding (¼echolocation though solid media) is typical

to the tribe Cryptini and it has multiple origins within the subfamily. We show that

vibrational sounding is associated with antennal modification and the usage of

wood-boring buprestid and cerambycid beetles, and suggest, based on an apparent

transition series, that the hammers are derived from mechano-sensilla within the

Cryptinae.

16.1 Introduction

The Ichneumonidae is one of the largest insect families with more than 20,000

described species (Yu et al. 2005), though, according to Gaston and Gauld (1993),

the real number of species may reach more than half of a million. Ichneumonid

wasps are cosmopolitan and whereas most species are parasitoids of other insects

N. Laurenne

Museum of Natural History, Entomology Division, University of Helsinki, P.O. Box 17,

(P. Arkadiankatu 13), 00014 Helsinki, Finland


D.L.J. Quicke

Department of Life Sciences, Imperial College London, Silwood Park Campus, Ascot, Berkshire

SL5 7PY, UK

Department of Entomology, Natural History Museum, London SW7 5BD, UK


271

and in some cases spiders, their way of life varies remarkably. Unlike simple

parasites, parasitoids always kill their hosts that are typically larvae or pupae of

various Lepidoptera, Coleoptera and Diptera.

Parasitoid life history strategies are commonly divided into two classes, the

koinobionts and idiobionts (Askew and Shaw 1986; Godfray 1994; Quicke 1997).

These two life strategies differ from each other considerably, but the defining

difference between them is that idiobionts do not permit their host to carry on

developing after parasitisation. In those cases in which the host is a larval stage, it is

typically paralysed by the female wasp’s venom. In contrast to idiobionts, the hosts

of koinobionts are allowed to continue their development after parasitisation until

they reach a suitable stage to be consumed by parasitoid larvae. Several other

features are associated with these life strategies, for example, koinobionts are

most usually endoparasitoids with relatively narrow host ranges as they have to

be able to adapt to the host’s immunological defenses. Idiobionts are typically

ectoparasitoids and generalists with a wide host range, though those that attack

pupal hosts are usually endoparasitoids. For idiobionts, host range is often largely

determined by the potential hosts that are encountered. The hosts of koinobionts are

often exposed or very little concealed. Paralysed hosts would be very prone to

predation if they were exposed, and therefore hosts of idiobionts tend to live in

concealed conditions (i.e. leaf-rolls or leaf mines, plant stems, under bark or inside

wood). The trait of exploiting concealed hosts is regarded as the ancestral state in

the Ichneumonidae and transitions from idiobiosis to koinobionsis appear to have

happened multiple times within the family (Belshaw et al. 1998; Whitfield 1998).

16.1.1 Host Location

According to Vinson (1988), host location consists of several stages beginning with

finding a suitable habitat. Then, a parasitoid must locate a potential host therein,

followed by examining it for suitability (species and the developmental stage) and

finally, oviposition.

Parasitoids use many modalities in host location; scent, vision, sound and

vibration are involved (Wertheim et al. 2003; Fischer et al. 2004; Fatouros et al.

2005). Wasps lead to a host by several cues, for example, parasitoids can recognise

shape, colour and a movement of a host (Fischer et al. 2001, 2004). Volatile

chemicals from host frass and damaged plant material are shown to be attractive

to parasitoids (Gohole et al. 2003; Bukovinszky et al. 2005). Some species have

even evolved to detect host sex pheromones and other kairomones and use them in

host searching (Wertheim et al. 2003; Jumean et al. 2005). In general, multiple cues

are involved in host-searching process and their efficiency is affected by environ-

mental factors, such as temperature (Fischer et al. 2001; Kroder et al. 2007a, b).

The female ovipositor and antennae are both important for host examination and

acceptance as they have various sensillae for detecting the suitability of a potential

272 N. Laurenne and D.L.J. Quicke

host (Mackauer et al. 1996; Ignacimuthu and Dorn 2000; Isidoro et al. 2001;

Romani et al. 2002).

Many mobile hosts of ichneumonid wasps live in concealed places such as

within wood. Such host larvae cause vibration when they chew wood and move,

and some parasitoid groups have evolved an ability to detecting these host-

generated vibrations. However, not all potential concealed hosts create their own

vibrations, e.g. pupal and prepupal stages or larva shortly about to moult. To locate

these, some parasitic wasps have evolved an active, vibration-based, method called

vibrational sounding. This form of echolocation occurs in one non-apocritan group,

the Orussidae which have highly modified antennae and massively enlarged sub-

genual (hearing) organs in the forelegs (Vilhelmsen et al. 2001). Females tap with

their antenna the substrate and detect the echoes with their subgenual organs. This

idea was originally suggested by Cooper (1953) and later Powell and Turner

(1975) made similar observations of female behaviour supporting Cooper’s

conjecture.

Use of vibrational sounding as a means of host location has also evolved on a

number of separate occasions within the Ichneumonidae. Amongst the parasitic

apocritan wasps vibrational sounding has been most thoroughly investigated in the

pimpline ichneumonid genus Pimpla and relatives (Henaut and Guerdoux 1982;

Henaut 1990; Meyhofer and Casas 1999; Fischer et al. 2001, 2003). The success of

echolocation is dependent of several factors, and Kroder et al. (2006, 2007b) have

shown it to be more efficient in warmer conditions and the role of vision to be more

important in cooler conditions. Parasitoids can adjust the intensity of echolocation

according to the temperature which shows adaptation to environmental conditions

in temperate regions. The ability to adjust to the microhabitat and its varying

environmental factors involves a complicated interaction. According to Otten

et al. (2001), females with larger size are better in finding concealed hosts in

comparison with smaller ones: a larger body mass is capable of transmitting

vibration better than smaller one.

Apart from in the pimplines, females of a number of other ichneumonid genera

are hypothesised to use vibrational sounding based on their morphology: with

antennal tips modified into a hammer-like structures suitable for “hammering”

the substrate and enlarged subgenual organs in their fore tibiae for detecting

substrate-borne vibrations (Broad and Quicke 2000). Additionally, the antennal

pegs of female Xorides (Xoridinae) are solid (Quicke unpublished observations)

and therefore likely to act as antennal hammers.

The largest subfamily of Ichneumonidae is the Cryptinae with 4,659 species

belonging to 394 genera (Yu et al. 2005). The cryptines are appropriate model

group as the vibrational sounding has multiple origins and losses and there is a

detailed molecular phylogenetic analysis (Laurenne et al. 2006). We tested the

association between the occurrence of hammer-like terminal antennal segments

within the Cryptinae and the explotation of wood-boring buprestids and ceramby-

cids within a comparative phylogenetic framework.

Traditionally, the Cryptini has been divided into three tribes: Cryptini, Phyga-

deuontini and Hemigasterini, and molecular studies largely support this classification

16 Antennal Hammers: Echos of Sensillae Past 273

(Laurenne et al. 2006; Quicke et al. 2009). Most cryptines are idiobiont ectoparasitoids

and their hosts usually belong the largest insect orders (Coleoptera, Lepidoptera,

Hymenoptera and Diptera), but spider egg predation occurs in some cryptine genera,

and a few other insect orders are occasionally attacked. Despite their host groups

covering several orders as a whole, individual cryptine species can be quite host

specific or have a narrow host range (Askew and Shaw 1986; Gauld 1988; Schwarz

and Shaw 1998, 2000).

16.2 Material and Methods

We examined the terminal antennal flagellomeres of species representing 122

genera of the subfamily Cryptinae, six of Ichneumoninae and one species each of

the Alomyinae, Eucerotinae and Pedunculinae. Scanning electron microscopy

(SEM) was used for the vast majority, though light microscope was occasionally

relied upon for larger sized specimens of some groups. For males we included 32

genera (26 cryptines, 2 hemigasterines and 4 phygadeuontines). Female antennal

tips were classified into five categories according to the degree of modification from

unmodified antennae with a tapered tip to ones forming a large flat surface. The

intermediate stages show structures of individual setae becoming thicker and

forming a cluster (Laurenne et al. 2009).

16.2.1 Comparative Analysis

Comparative analysis (CAIC) was carried out to test the statistical significance of

association between antennal modification and the use of wood-boring beetles

(buprestids and cerambycids) (Purvis and Rambaut 1995). The degree of antennal

modification was treated as a continuous variable and the coleopteran hosts were

treated as a categorical variable. Evolutionary rate was assumed to be the same for

each taxon.

The trees used in the comparative analysis were based on Laurenne et al.’s

(2006) molecular study of cryptine phylogeny based on the length-variable D2

(þD3) variable region of the nuclear 28S rDNA gene, but taxa without the host

record information were pruned from the tree as missing values are not allowed in

CAIC. Two cryptine genera (Mallochia and Schreineria) with host records were

added into the tree and, in the absence of molecular data, their placements were

based on Townes’s (1969) classification. To avoid biased results, the comparative

analyses were carried out using five different gap cost ratios and with two different

alignment methods (POY and Clustal W þ PAUP*). Details of the methods are

described in Laurenne et al. (2009).


16.3 Results

The percentages of the degree of antennal modification are shown in Fig. 16.1.

Figure 16.2 presents the occurrence of antennal development on a phylogenetic tree.

Figures 16.3 and 16.4 show the transformation series from a simple antennal tip

with no especially modified sensilla to a large united structure with a virtually

uniform surface. Surculus (Fig. 16.3a) displays a simple antennal tip without

obvious modification. Figure 16.3b,c shows thickening of some apical setae in

the genera Latibulus (Fig. 16.3b) and Hidryta (Fig.16.3c). Setae are modified into

truncate structures forming a cluster in genera Camera (Fig.16.3d) and in Crypta-nura (Fig.16.3f) modified structures have started to fuse in the middle. In Fig. 16.4,

fused structures form a more or less flat surface in females of Acrorichnus(Fig. 16.4a), and Buathra (Fig. 16.4b) shows a smooth face of modified and fused

structures. The antennal tip of Osprynchotus (Fig. 16.4c) forms a large uniform flat

surface, a truly hammer-like antenna. Some genera have different types of specia-

lisation of the antennal tip, for example, Meringopus (Fig. 16.4d) has thickened

“setae” originating from sockets inside the antennal surface.

Terminal antennal structures of cryptines are often sexually dimorphic charac-

ters as males typically do not display any particular antennal modification. How-

ever, some specialisations do occur in males of a few genera. For example, males of

Gabunia (Fig. 16.4e) have two peg-like structures on their antennal tip and those ofEurycryptus have one smaller structure (Fig. 16.4f).

Fig. 16.1 The precentage of occurrence of each degree of antennal hammer development in each

tribe of Cryptinae


Fig. 16.2 The phylogeny of cryptine waps (Laurenne et al. 2009). The black circles indicate

attacking buprestid/cerambycid beetles and having strongly modified antennae (category 4–5).

Grey circles indicate the occurrence of slightly modified antenna (categories 1–3)


The CAIC analysis showed a significant association between the degree of

antennal development and the usage of wood-boring buprestid and cerambycid

beetles in the Cryptini. Thirteen genera of the tribe Cryptini exploit wood-boring

beetle larvae and have modified antennae. Within the Phygadeuontini, only

five genera have this association. p-Values showed a significant association

(0.0080–0.0397) in all analysis except with the alignment obtained with the highest

gap:substitution cost (4:1, p-value ¼ 0.0707). Detailed results are presented in

the Laurenne et al. (2009).

16.4 Discussion

Possession of an antennal hammer is a clearly homoplastic character at an higher

level as it is found also in other ichneumonid subfamilies (Labeninae, Xoridinae,

Claseinae and Pimplinae) (Broad and Quicke 2000) as well as in the Orussidae

(Cooper 1953; Broad and Quicke 2000; Vilhelmsen et al. 2001). This structure is

a b

c d

e f

Fig. 16.3 Female antennal tips showing antennal modification. (a) Surculus, not modified. (b

and c) Some thickened setae on a tip – (b) Latibulus and – (c) Hidryta. (d) Diapetimorpha,thickened structures form a cluster. (e) Camera, dense cluster of truncate structures form a patch.

(f) Cryptanura, a cluster of short apically flattened structures with a fusion in the middle


associated with deeply concealed cerambycid and buprestid beetle hosts and we

have shown by comparative analysis that it is also highly homoplastic within the

single but large subfamily Cryptinae.

Behavioural observations of Echthrus and of a Gabunia sp. (Quicke et al. 2003)

support the hypothesis that antennal hammers in the Cryptini are associated with

host searching. In 2004, we video recorded the host-searching behaviour of a

female Echthrus reluctator on a pile of pine logs in Hungary (Quicke 2001). The

wasp walked along the log tapping the substrate with the antennae repeatedly

sweeping symmetrically in inwardly directed arcs. Similar behaviour was also

observed in an unidentified Afrotropical species of Gabunia (tribe Cryptini) in

Kibale Forest National Park in Uganda.

16.4.1 Hosts of Cryptine Wasps

Most cryptine wasps are ectoparasitoids and they do not need to adapt to host’s

immunological defense. This may explain why some genera attack hosts from

a b

c d

e f

Fig. 16.4 Antennal tips of female and males. (a) Acrorichnus female, apical structures form

a clear patch. (b) Buathra female, structures form a smooth patch. (c) Osprychotus female,

a hammer-like antennal tip. (d) Meringopus female, thickened antennal setae originating from

deep sockets. (e) Gabunia male, two pegs on antennal tips, (f) Eurycryptus, one antennal peg


several insect orders. The essential ability in host usage might be to find concealed

hosts of suitable sized.

16.4.1.1 Hosts of the Phygadeuontini

Species of the tribe Phygadeuontini typically parasitise exposed or weakly con-

cealed hosts and this is considered to be a ground-plan biology for the Cryptinae

(Gokhman 1996). The comparative analysis using the phylogeny (Laurenne et al.

2006) shows that modified antennal tips have multiple origins within the Phyga-

deuontini and host range covers several insect orders. Antennal modification was

found in three genera, all of which attack wood-boring beetles (Fig. 16.4).

16.4.1.2 Hosts of the Cryptini

In the tribe Cryptini, all the taxa that exploit wood-boring beetles have antennal

hammers. This is probably the ground-plan for the tribe. Strongly modified antennal

structures are also found in genera that attack other insect groups such as aculeate

Hymenoptera larvae in their nests. Parasitoids probably locate cells with suitable

host using vibrational sounding. Aculeate larvae are probably largely silent and do

not chew wood, though they move inside a cell when they need a feed by adults.

Members of the genera Acroricnus, Eurycryptus, Messatoporus, Osprynchotus andPhotocryptus exploit aculeate larvae (Genaro 1996) and they all have modified

antennal tips. According to Gauld (1988) there may be a host shift from Coleoptera

hosts to the young of nest-building aculeate Hymenoptera, but this is only a

hypothesis and cannot be tested at present due to the lack of sufficient detailed

host information for the vast majority of Cryptinae genera.

Unlike most other subtribes, the Gabuniini form a well-supported monophyletic

group (Laurenne et al. 2006) comprising 12 genera. Ten of these have strong

antennal modifications and the four available host records indicate that these

species exploit cerambycid or buprestid beetle hosts. The cylindrical body shape

of gabuniines and their long ovipositors probably enable them to reach their hosts

and are perhaps constrained by host boring shape (Townes and Townes 1962); the

enlarged subgenual organs found in the forelegs of females are assumed to be for

detecting echos during host location (Broad and Quicke 2000).

Most of available host records concerning the cryptine wasps concern phyga-

deuontines, many of which attack rather weakly concealed hosts, especially ones in

cocoons, or spider egg masses. The spider egg “parasitoids” attack exposed egg

masses, and therefore, vibrational sounding probably has no role in locating them,

and the antennal tips of the spider egg “parasitoids” examined are typically simple.

Hyperparasitism of cocooned parasitoid hosts occurs more commonly in the Cryp-

tini than in the Phygadeuontini, though there are numerous examples within the

latter. Some genera have modified antennal tips, but that could possibly be

explained by the adaptation to exploit other insect groups as well.

Within the Cryptini, males of six out of the ten genera examined had either one

or two terminal flagellomere pegs. The females of the same genera also had


antennal modifications except for the case of Chrysocryptus. Structures of male

terminal flagellomeres are probably not related to the echolocation role of female

antennal hammers. Their co-occurrence suggests that there might be homologous

genetic control in the tribe Cryptini. Whether, and in what way, they may be

functional has yet to be determined. Field observations of mate-location and mating

are sadly largely lacking.

Considering the size of the subfamily, very few host records are available for

cryptine genera, and when records exist, they are often vague. Records typically

especially lack information about the host’s precise developmental stage. Field

records are largely lacking, and the host-location behaviour is usually referred to as

“antennation” without describing what part of the antennae is used. We hope that

this paper will encourage more detailed observation and reporting in the future.

16.4.2 Postulated Derivation of Hammers from Sensilla

If the states shown in Fig. 16.3a–f represent various stages in the evolution of

antennal hammers as seems likely, then the individual components of the hammer

surface would appear to be derived from sensilla. The unmodified terminal flagel-

lomere of Surculus has many thin curved sensilla chaetica, with a lower number of

more erect obliquely ended chaetica (on right), and one visible blunt sensillum. In

Latibulus (Fig. 16.3b), there are numerous blunt trichoid sensilla in relatively small

sockets plus several longer more pointed chaetica in rather large sockets. In

Fig. 16.3c, there is a similar grouping of socketed and less conspicuously socketed

blunt sensilla but with their apices curving towards the antennal tip and interspersed

with small trichoid sensilla. In Fig. 16.3d, the apical cluster comprises a dense

central area of T-shaped pegs that lack sockets at least on the basal side though on

the side of the antennal apex there appears to be a well-developed basal socket;

these are surrounded by curved, socketed robust trichoid sensilla. Socketed trichoid

sensilla are typically involved in mechanoreception.

If, as the above suggests, the antennal hammers of cryptines, and possibly other

ichneumonid wasps, are evolved from mechanoreceptory sensilla, it begs the

question as to what the intermediate evolutionary stages did, and what substrates,

the hosts during those intermediate phases occupied. Certainly more detailed

behaviour, microscopic and ultrastructural observations of living representatives

of apparent intermediate stages are needed.

References

Askew RR, Shaw MR (1986) Parasitoid communities: their size, structure and development. In:

Waage J, Greathead D (eds) Insect parasitoids. Academic, London, pp 225–264

Belshaw R, Fitton M, Herniou E, Gimeno C, Quicke DLJ (1998) A phylogenetic reconstruction of

the Ichneumonoidea (Hymenoptera) based on the D2 variable region of 28S ribosomal RNA.

Syst Entomol 23:109–123


Broad GR, Quicke DLJ (2000) The adaptive significance of host location by vibrational sounding

in parasitoid wasps. Proc R Soc Lond B Biol 267:2403–2409

Bukovinszky T, Gols R, Posthumus MA, Vet LEM, van Lenteren JC (2005) Variation in plant

volatiles and attraction of the parasitoid Diadegma semiclausum (Hellen). J Chem Ecol

31:461–480

Cooper KW (1953) Egg gigantism, oviposition, and genital anatomy: their bearing on the biology

and phylogenetic position of Orussus (Hymenoptera: Siricoidea). Proc R Acad Sci 10:38–68

Fatouros NE, Huigens ME, van Loon JJA, Dicke M, Hilker M (2005) Butterfly antiaphrodisiac

lures parasitic wasps. Nature 433:704

Fischer S, Samietz J, W€ackers FL, Dorn S (2001) Interaction of vibrational and visual cues in

parasitoid host location. J Comp Physiol A 187:785–791

Fischer S, Samietz J, Dorn S (2003) Efficiency of vibrational sounding in parasitoid host location

depends on substrate density. J Comp Physiol A 189:723–730

Fischer S, Samietz J, W€ackers FL, Dorn S (2004) Perception of chromatic cues during host

location by the pupal parasitoid Pimpla turionellae (L.) (Hymenoptera: Ichneumonidae).

Environ Entomol 33:81–87

Gaston KJ, Gauld ID (1993) How many species of pimplines (Hymenoptera: Ichneumonidae) are

there in Costa Rica? J Trop Ecol 9:491–499

Gauld ID (1988) Evolutionary patterns of host utilization by ichneumonoid parasitoids hymenop-

tera Ichneumonidae and Braconidae. Biol J Linn Soc 35:351–378

Genaro JA (1996) Nest parasites (Coleoptera, Diptera, Hymenoptera) of some wasps and bees

(Vespidae, Sphecidae, Colletidae, Megachilidae, Anthophoridae) in Cuba. Caribb J Sci

32:239–240

Gohole LS, Overholt WA, Khan ZR, Vet LEM (2003) Role of volatiles emitted by host and non-

host plants in the foraging behaviour of Dentichasmias busseolae, a pupal parasitoid of the

spotted stemborer Chilo partellus. Entomol Exp Appl 107:1–9

Godfray HCJ (1994) Parasitoids: behavioral and evolutionary ecology. Princeton University Press,

Princeton, NJ

Gokhman VE (1996) Trends of biological evolution in the subfamily Ichneumoninae and related

groups (Hymenoptera Ichneumonidae): an attempt of phylogenetic reconstruction. Russ

Entomol J 4:91–103

Henaut A, Guerdoux J (1982) Location of a lure by the drumming insect Pimpla instigator(Hymenoptera, Ichneumonidae). Experientia 38:346–347

Henaut A (1990) Study of the sound produced by Pimpla instigator (Hymenoptera, Ichneumoni-

dae) during host selection. Entomophaga 35:127–139

Ignacimuthu S, Dorn S (2000) Mechano- and chemoreceptors and their possible role in host

location behaviour of parasitoid Anisopteromalus calandrae Howard (Hymenoptera: Pteroma-

lidae). Entomon 25:179–184

Isidoro N, Romani R, Bin F (2001) Antennal multiporous sensilla: their gustatory features for host

recognition in female parasitic wasps (Insecta, Hymenoptera: Platygastroidea). Microsc Res

Tech 55:350–358

Jumean Z, Unruh T, Gries R, Gries G (2005)Mastrus ridibundus parasitoids eavesdrop on cocoon-spinning codling moth, Cydia pomonella, larvae. Naturwissenschaften 92:20–25

Kroder S, Samietz J, Dorn S (2006) Effect of ambient temperature on mechanosensory host

location in two parasitic wasps of different climatic origin. Physiol Entomol 31:299–305

Kroder S, Samietz J, Dorn S (2007a) Temperature affects interaction of visual and vibrational cues

in parasitoid host location. J Comp Physiol 193:223–231

Kroder S, Samietz J, Schneider D, Dorn S (2007b) Adjustment of vibratory signals to ambient

temperature in a host-searching parasitoid. Physiol Entomol 32:105–112

Laurenne NM, Broad GR, Quicke DLJ (2006) Direct optimization and multiple alignment of 28S

D2–D3 rDNA sequences: problems with indels on the way to a molecular phylogeny of the

cryptine ichneumon wasps (Insecta: Hymenoptera). Cladistics 22:442–473


Laurenne NM, Karatolos N, Quicke DLJ (2009) Hammering homoplasy: multiple gains and losses

of vibrational sounding in cryptine wasps (Insecta: Hymenoptera: Ichneumonidae). Biol J Linn

Soc 96:82–102

Meyhofer R, Casas J (1999) Vibratory stimuli in host location by parasitic wasps. J Insect Physiol

45:967–971

Mackauer M, Michaud JP, Volkl W (1996) Host choice by aphidiid parasitoids (Hymenoptera:

Aphidiidae): host recognition, host quality, and host value. Can Entomol 128:959–980

Otten H, W€ackers F, Battini M, Dorn S (2001) Efficiency of vibrational sounding in the parasitoid

Pimpla turionellae is affected by female size. Anim Behav 61:671–677

Powell JA, Turner WJ (1975) Observations on oviposition behaviour and host selection in Orussusoccidentalis (Hymenoptera: Siricoidea). J Kans Entomol Soc 48:299–307

Purvis A, Rambaut A (1995) Comparative analysis by independent contrasts (CAIC): an Apple

Macintosh application for analysing comparative data. Comput Appl Biosci 11:247–251

Quicke DLJ (1997) Parasitic wasps. Chapman & Hall, London, New York

Quicke DLJ (2001) Movie of host searching Echthrus. http://www.imperial.ac.uk/imedia/vid/fons/

biology/quicke//Echthrus.mp4. Accessed 7 Dec 2009

Quicke DLJ, Laurenne NM, Broad GR, Barclay MVL (2003) Host location behaviour and a new

host record for Gabunia aff. togoensis Krieger (Hymenoptera: Ichneumonidae: Cryptinae) in

Kibale Forest National Park, West Uganda. Afr Entomol 11:308–310

Quicke DLJ, Laurenne NM, Fitton MG, Broad GR (2009) A thousand and one wasps: a 28S rDNA

and morphological phylogeny of the Ichneumonidae (Insecta: Hymenoptera) with an investi-

gation into alignment parameter space and elision. J Nat Hist 43:1305–1421

Romani R, Isidoro N, Bin F, Vinson SB (2002) Host recognition in the pupal parasitoid Trichopriadrosophilae: a morpho-functional approach. Entomol Exp Appl 105:119–128

Schwarz M, Shaw MR (1998) Western Palaearctic Cryptinae (Hymenoptera: Ichneumonidae) in

the National Museums of Scotland, with nomenclatural changes, taxonomic notes, rearing

records and special reference to the British check list. Part 1. Tribe Cryptini. Entomologist’s

Gaz 49:101–127

Schwarz M, Shaw MR (2000) Western Palaearctic Cryptinae (Hymenoptera: Ichneumonidae) in

the National Museums of Scotland, with nomenclatural changes, taxonomic notes, rearing

records and special reference to the British check list. Part 3. Tribe Phygadeuontini, subtribes

Chiroticina, Acrolytina, Hemitelina and Gelina (excluding Gelis), with descriptions of new

species. Entomologist’s Gaz 51:147–186

Townes H (1969) The genera of Ichneumonidae, part 1. Mem Am Entomol Inst 11:1–300

Townes H, Townes M (1962) Ichneumon-flies of America north of Mexico: 3. Subfamily Gelinae,

tribe Mesostenini. United States National Museum Bulletin 216:1–602

Vinson SB (1988) Comparison of host characteristics that elicit host recognition behavior of

parasitoid Hymenoptera. In: Gupta VK (ed) Advances in parasitic Hymenoptera research:

proceedings of the II conference on the taxonomy and biology of parasitic Hymenoptera. E. J.

Brill, Leiden, pp 285–291

Vilhelmsen L, Isidoro N, Romani R, Basibuyuk HH, Quicke DLJ (2001) Host location and

oviposition in a basal group of parasitic wasps: the subgenual organ, ovipositor apparatus

and associated structures in the Orussidae (Hymenoptera, Insecta). Zoomorphology 121:63–84

Wertheim B, Vet LEM, Dicke M (2003) Increased risk of parasitism as ecological costs of using

aggregation pheromones: laboratory and field study of Drosophila–Leptopilina interaction.

Oikos 100:269–282

Whitfield JB (1998) Phylogeny and evolution of host–parasitoid interactions in Hymenoptera. Ann

Rev Entomol 43:129–151

Yu D, van Achtenberg K, Horstmann K (2005) World Ichneumonoidea 2004. Taxonomy, biology,

morphology and distribution. CD/DVD, Taxapad, Vancouver, Canada


http://www.imperial.ac.uk/imedia/vid/fons/biology/quicke//Echthrus.mp4

http://www.imperial.ac.uk/imedia/vid/fons/biology/quicke//Echthrus.mp4

Chapter 17

Adaptive Radiation of Neotropical Emballonurid

Bats: Molecular Phylogenetics and Evolutionary

Patterns in Behavior and Morphology

Burton K. Lim

Abstract A phylogenetic analysis of loci from the four genetic transmission path-

ways in mammals (mitochondrial, autosomal, X, and Y sex chromosomes) was

used to investigate the evolution of bats in the pantropically distributed family

Emballonuridae. The nuclear data sets support a monophyletic clade of species

found in the New World. Character optimization of distributional areas suggests

that the most recent common ancestor colonized South America from Africa.

Molecular dating with fossil calibrations estimated that a basal split occurred

approximately 27 million years ago followed by primary intergeneric diversifica-

tion 19.4–18.0 million years ago. An analysis of historical biogeography identified

the northern Amazon as the ancestral area where there was speciation by taxon

pulses from a stable core area in the Guiana Shield. Range contractions followed by

expansions during the Early Miocene suggest an adaptive radiation in cluttered

forest and open savannah habitats. A correlation of ear morphology, echolocation,

and foraging behavior indicates a phylogenetic basis for these complex character

systems.

17.1 Introduction

South America was an insular continent from the Late Cretaceous to the Early

Pliocene but nevertheless, it has high levels of biodiversity for many groups of

organisms compared with other parts of the world. For example, bats account for

20% of the mammalian faunal diversity (Wilson and Reeder 2005) and are unique

in being the only order of mammals that can fly. This gives bats an advantage for

over-water dispersal but there have been no studies investigating the evolutionary

B.K. Lim

Department of Natural History, Royal Ontario Museum, 100 Queen’s Park, Toronto, Ontario M5S

2C6, Canada



283

mechanisms for the successful radiation of bats, especially in the rainforests of the

Amazon. As with most taxa, this has been hindered by a lack of comprehensive

species-level phylogenies, a dearth of fossils in the paleontological record, and a

paucity of ecological data. Herein, I synthesize data on New World emballonurid

bats in the tribe Diclidurini as one of the first detailed studies of an adaptive

radiation of mammals in the Neotropics.

I begin by giving general background information on the biology of the family

Emballonuridae. The primary objective of this study is to hypothesize the processes

involved in the biotic diversification in NewWorld emballonurid bats by inferring a

robust phylogeny of New World emballonurid bats using a molecular phylogenetic

approach, estimating times of divergence based on molecular dating with fossil

calibration points, examining the historical biogeography with the incorporation of

both temporal and spatial information, and investigating patterns of evolution in

morphology and behavior as inferred from the phylogeny.

17.1.1 Emballonurid Bats

The family Emballonuridae is characterized by a tail that emerges mid-dorsally

from the interfemoral membrane, which is the origin of its common name of sheath-

tailed bats. They are found pantropical in distribution, and the New World embal-

lonurids occur from Mexico through Central America into South America to

southeastern Brazil, including the off-shore islands of Trinidad, Tobago, and

Grenada (Koopman 1994). Most species are uncommonly encountered in Neotrop-

ical rainforests using traditional methods of capture such as mesh mist nets set in the

understory because they typically fly in or over the canopy. Consequently, New

World emballonurid bats are typically poorly studied and incompletely sampled in

terms of taxonomic and geographic coverage. However, this apparent rarity is

associated with a sampling bias that may be partially corrected by supplemental

surveying by novel methods such as flap trapping (Borissenko 1999; Lim 2009),

acoustic monitoring (Jung et al. 2007), and systematically searching for roosts

(Simmons and Voss 1998).

17.1.2 Taxonomy

There are 16 genera of emballonurid bats with 13 extant (eight in the New World

and five in the Old World) and three extinct (all Old World) that are represented by

63 species with 52 extant (22 NewWorld and 30 Old World) and 11 extinct (all Old

World; McKenna and Bell 1997; Simmons 2005; Lim et al. 2010). Four previous

phylogenies have been proposed for Emballonuridae including studies on cranial

morphology (Barghoorn 1977), protein electrophoresis and immunology (Robbins

and Sarich 1988), hyoid morphology (Griffiths and Smith 1991), and morphology

284 B.K. Lim

and behavior (Dunlop 1998). All of these studies were at the taxonomic rank of

genus except for the species-level analysis of Dunlop (1998). However, the only

taxonomic congruence among the topologies is the higher-level recognition of

subfamilies (Emballonurinae and Taphozoinae). The lack of consensus in other

parts of these trees was confounded by a combination of incomplete taxonomic

sampling and poor resolution. A recent molecular phylogenetic analysis of DNA

sequence variation supported this taxonomic classification (Lim et al. 2008).

Although the New World emballonurid species were comprehensively surveyed,

there were only exemplar samples of the two Old World tribes, which are still

poorly represented by tissue collections.

17.2 Molecular Phylogenetic Analyses

The data set for New World emballonurid bats included 99 specimens representing

all of the eight recognized genera and 21 of the 22 species (Simmons 2005; Lim

et al. 2010). The only missing species is Saccopteryx antioquensis, which is

endemic to the northern Andes of Colombia and known by only two specimens

without tissue samples (Munoz and Cuartas 2001). Outgroup taxa included nine

specimens representing two genera of Old World emballonurids and four genera of

other bat species (Lim et al. 2008).

Loci from the four genomic components of mammalian transmission genetics

were used to hypothesize the evolutionary history of NewWorld emballonurid bats.

Each of these genetic transmission pathways has different properties associated

with effective population size, mutation rate, and recombination that should be

conducive for recovering a robust estimate of phylogeny. The mitochondrial

marker was the complete protein-coding gene cytochrome b (Cytb); the autosomal

marker was intron 26 of the protein-coding gene Chd1 (found on chromosome 5 in

humans); the Y sex chromosome marker was intron 7 of the protein-coding gene

Dby; and the X sex chromosome marker was intron 18 of the protein-coding gene

Usp9x (Lim et al. 2008). There were a total of 3,176 aligned basepairs (bp)

including 1,140 bp of Cytb, 624 bp of Chd1, 750 bp of Dby, and 662 bp of Usp9x.The phylogenetic analyses of individual and combined nucleotide data sets

incorporated both an explicit model of DNA evolution using a statistical Bayesian

approach and a model-free methodology using a maximum parsimony approach as

corroboration of topological robustness. Bayesian inference was implemented in

the programMrBayes (Ronquist and Huelsenbeck 2003) and parsimony reconstruc-

tion was implemented in the program PAUP* (Swofford 2001) as outlined by Lim

et al. (2008). Branch supports of the resultant trees were calculated by the posterior

probability distribution in the Bayesian analysis and by 1,000 bootstrap replications

in the parsimony analysis. The trees were compared for topological congruence

using the Approximately Unbiased (AU) test (Shimodaira and Hasegawa 2001).

Each data set was reciprocally constrained to the individual gene trees to determine

if one was better than another.

17 Adaptive Radiation of Neotropical Emballonurid Bats: Molecular Phylogenetics 285

17.2.1 Tree Topology

Parsimony and Bayesian analyses of each of the individual data sets gave congruent

topologies with high bootstrap proportions and posterior probabilities for mono-

phyletic clades representing the currently recognized genera and species of New

World emballonurid bats (Fig. 17.1; Lim et al. 2008). However, the mitochondrial

Fig. 17.1 Phylogenetic tree from a Bayesian analysis of combined DNA sequences of three

nuclear genes for New World emballonurid bats, tribe Diclidurini (Lim et al. 2008). The first

number along the branch is the Bayesian posterior probability percentage, and the second number

is the bootstrap percentage from a parsimony analysis. Numbers in parentheses are the

corresponding branch-support values from a phylogenetic analysis after the removal of the out-

group taxon Nycteris javanicus, which was missing data for two of the genes. Intrageneric support

values are the same for both analyses and branches with an asterisk (*) have 100% support.

Peropteryx macrotis has two divergent populations from Central America (CA) and South

America (SA)

286 B.K. Lim

gene had significantly faster rates of nucleotide substitution, higher levels of

homoplasy, and a greater degree of saturation of transitions than any of the three

nuclear genes. These factors contributed to the loss of phylogenetic signal at deeper

branches of the cytochrome b tree including the monophyly of the New World

emballonurids. In contrast, there was better resolution and branch support for the

more slowly evolving nuclear introns. However, the intergeneric relationships

within the two subtribes were poorly resolved and supported by only a few nucleo-

tide changes. This suggests a hard polytomy resulting from a lack of phylogenetic

signal in each of the different genetic transmission pathways because of rapid

speciation as opposed to a soft polytomy due to conflicting phylogenetic signal.

Based on topological congruence, linear accumulation of substitutions, and high

consistency index, the three nuclear genes were combined to lessen the effects of

random sequence errors among nucleotide sites and ensure the recovery of phylo-

genetic signal from a robust species tree. A monophyletic New World clade was

recovered in the individual and combined nuclear data sets indicating a single

origin of emballonurid bats in the Neotropics (Fig. 17.1). Similarly, there was a

basal split in the New World tribe Diclidurini that was congruent and well sup-

ported in the nuclear trees.

17.3 Divergence Times

The combined nuclear data set for the tribe Diclidurini was used in a Bayesian

relaxed clock approach to approximate the times of divergence (Thorne and

Kishino 2002). Two fossil constraints were used as calibration points including

a minimum age of 13 million years ago (mya) for the split of Cyttarops and

Diclidurus based on the only pre-Pleistocene record of an extant New World

emballonurid genus (Czaplewski 1997). The second constraint was a maximum

age of 30 million years ago for the split of the Old and New World emballonurids

based on a molecular dating analysis with fossil calibrations for all families of bats

(Teeling et al. 2005). The basal split in the New World emballonurids occurred in

the Late Oligocene approximately 27 million years ago and six of the eight

currently recognized genera diversified relatively rapidly in the Early Miocene

19.4–18.0 million years ago, and most intrageneric differentiation (16 of 21 species)

occurred before the Pliocene 5 million years ago (Fig. 17.2; Lim 2007).

17.4 Historical Biogeography

Character optimization (Farris 1970) of distributional areas onto the phylogeny for

the superfamily Emballonuroidea indicates that the ancestor of New World emballo-

nurid bats has its origins in Africa (Fig. 17.3; Lim 2007). This biogeographic scenario

was previously suggested from phylogenetic studies of interfamilial relationships of


bats (Eick et al. 2005; Teeling et al. 2005). The paleoenvironment during the Early

Oligocene was drier than today with more open habitats such as woodlands and

savannahs as suggested by the prevalence of large hypsodont mammals in the fossil

record (Flynn and Wyss 1998). Colonization of South America by trans-Atlantic

dispersal and subsequent speciation in allopatry has been reported for three other

groups of placental mammals based on fossil records from the Oligocene including

molossid bats (Legendre 1984), caviomorph rodents (Wyss et al. 1993), and platyr-

rhine primates (Takai et al. 2000). These range expansions probably occurred earlier

in the Eocene (Poux et al. 2006).

The phylogenies of each of the eight genera of New World emballonurid bats

were incorporated in an historical biogeographic analysis using the algorithm

Phylogenetic Analysis for Comparing Trees (PACT; Wojcicki and Brooks 2005).

In constructing the area cladogram, temporal information from the molecular dating

Fig. 17.2 Molecular dating based on a relaxed clock Bayesian analysis with fossil calibrations of

New World emballonurid bats (Lim 2007). Nodes are labeled with divergence time estimates

(millions of years ago) and standard deviations. Intergeneric and most intrageneric diversification

occurred in the Miocene (shaded). Peropteryx macrotis has two divergent populations from

Central America (CA) and South America (SA)

288 B.K. Lim

analysis (Lim 2007) was also used in conjunction with spatial information based

on the current distribution of each species (Table 17.1). There were nine biogeo-

graphic areas identified in Central and South America for New World emballonur-

ids (Fig. 17.4). The final area cladogram identified the Northern Amazon as the

Fig. 17.3 Phylogenetic tree for the superfamily Emballonuroidea with the ancestral areas mapped

onto each node (AF Africa, EU Europe, NA North America, SA South America) following Lim

(2007). Lineage splits, other than the extant New World emballonurids (tribe Diclidurini), are

based on the minimum age of the fossil record (black bars). The basal divergence at 52 million

years ago (mya) of the families Nycteridae and Emballonuridae is the molecular approximation by

Teeling et al. (2005). Extinct taxa are indicated by an asterisk (*)


ancestral area for the basal node and for most internal nodes based on character

optimization (Fig. 17.5). This indicates that most lineage splits were within-area

speciation events. However, there were three range expansions from the Northern

Amazon followed by vicariant contractions including (1) a peripheral isolation in

the Pacific slope of northwestern South America and subsequent colonization of

Proto-Central America during the Middle Miocene; (2) colonization of northern

Colombia and vicariant isolation after the uplift of the Andes during the Late

Miocene; and (3) overland dispersal into Central America during the Pleistocene

after the establishment of the Panamanian land bridge connection, which was

followed by extinction in the intervening area of the northern Andes in Colombia,

which resulted in allopatric speciation (Lim 2008).

As is the case for most species of New World emballonurid bats, widely

distributed species typically are not conducive for recovering biogeographic pat-

terns. However, the optimization of the Northern Amazon at most nodes of the area

cladogram indicates repeated within-area speciation events. Tectonic uplifting of

Table 17.1 Biogeographic areas identified for species of New World emballonurid bats based on

current species distributions (Lim 2008)

Species Biogeographic area

A B C D E F G H I

Balantiopteryx infusca C

Balantiopteryx io B

Balantiopteryx plicata A

Centronycteris centralis B C D H

Centronycteris maximiliani F G I

Cormura brevirostris B D E F G H

Cyttarops alecto B F G

Diclidurus albus A B D E F G I

Diclidurus ingens E F G

Diclidurus isabellus F

Diclidurus scutatus F G I

Peropteryx kappleri B C D E F G H I

Peropteryx leucoptera F G H I

Peropteryx macrotis (Central America) A B

Peropteryx macrotis (South America) D E F G H I

Peropteryx pallidoptera F G

Peropteryx trinitatis E F

Rhynchonycteris naso A B C D E F G H I

Saccopteryx antioquensis D

Saccopteryx bilineata A B C D E F G H I

Saccopteryx canescens D E F G H

Saccopteryx gymnura F G

Saccopteryx leptura A B C D E F G H I

A ¼ Pacific versant of Central America; B ¼ Atlantic versant of Central America; C ¼ Choco

region of northwestern South America; D ¼ northern Andes and valleys of Colombia; E ¼ north

coast of Venezuela and offshore islands; F ¼ north of the Amazon River; G ¼ south of the

Amazon River; H ¼ eastern slope of the Andes in the western Amazon basin; and I ¼ southeast-

ern South America (Fig. 17.4)

290 B.K. Lim

the northern Andes (Hoorn et al. 1995) combined with fluctuations in temperature

and sea levels (Haq et al. 1987; Miller et al. 2005), and changes in vegetation (Janis

1993) contributed to a heterogeneous paleoenvironment in South America during

the Miocene (Lundberg et al. 1998). This scenario is similar to the taxon-pulse

hypothesis of biotic diversification with recurring adaptive shifts over time to

different habitats centered on a stable core area (Erwin 1979, 1981). For New

World emballonurid bats, there were repeated episodes of range expansions and

contractions from a stable core area such as the ancient Guiana Shield of the

Northern Amazon.

Mapping the area cladogram (Fig. 17.5) onto the chronogram (Fig. 17.3) sug-

gests that other than an earlier colonization in the Miocene that was associated with

the genus Balantiopteryx (Lim 2008; Lim et al. 2004), range expansion from South

America into Central America probably did not occur until later in the Pliocene.

Although Centronycteris split vicariantly in the Late Miocene with Centronycterismaximiliani speciating in the Northern Amazon and Centronycteris centralis in theNorthern Andes, C. centralis did not colonize Central America until a later date.

Similarly, Saccopteryx bilineata and Saccopteryx leptura split during the Late

Miocene in the North Amazon before both species became widely distributed

throughout the continental mainland. Even more recently, Diclidurus albus and

Diclidurus ingens split during the Early Pleistocene in the North Amazon before

D. albus dispersed into Central America. Although the topology forms a trichotomy

with Peropteryx kappleri, the allopatrically distributed Central and South American

populations of Peropteryx macrotis split in the Late Pleistocene. Three other

Fig. 17.4 Map of the nine

biogeographical areas in

Central America and South

America that were identified

based on current species

distributions in Table 17.1

(Lim 2008): (A) Pacific

versant; (B) Atlantic versant;

(C) Choco; (D) Northern

Andes; (E) North Coast; (F)

Northern Amazon; (G)

Southern Amazon; (H)

Western Amazon; (I)

Southeastern South America


species (Cormura brevirostris, Cyttarops alecto, and Rhynchonycteris naso) arealso widely distributed but their range expansions cannot be discerned from the area

cladogram. Likewise, patterns of range expansion from the Northern Amazon

southwards are not explicitly discernible because no speciation events involve the

Southern Amazon. However, C. maximiliani, S. bilineata, S. leptura, Saccopteryxcanescens, Saccopteryx gymnura, Diclidurus scutatus, and Peropteryx pallidopteradispersed from the Northern to the Southern Amazon sometime after they speciated

in the late Miocene. This timing coincides with the uplifting of the eastern

Fig. 17.5 Final area cladogram from an historical biogeographic analysis of New World embal-

lonurid bats (Lim 2008). Ancestral areas at nodes are derived from character optimization. Three

nodes marked with roman numerals in parentheses identify biotic expansions followed by vicari-

ant isolation. All other nodes are within-area taxon pulses of biotic diversification in the Northern

Amazon (F)

292 B.K. Lim

cordillera of the Andes, which created the Amazon River and primary drainage of

South America east toward the Atlantic Ocean as we know it today (Hoorn et al.

1995).

17.5 Evolutionary Patterns

17.5.1 Morphological Data

The most comprehensive morphological study of the family Emballonuridae

incorporated 141 external, cranial, and skeletal characters from 43 of 52 extant

species including 18 of 22 New World species (Dunlop 1998; Lim and Dunlop

2008). However, the phylogeny was poorly supported with the exception of the

genera within the tribe Diclidurini. Topological congruence using the KH (Kishino

and Hasegawa 1989), Wilcoxon signed ranks (Templeton 1983), and winning sites

(Prager and Wilson 1988) tests indicated that the morphological data set con-

strained to each of the molecular trees was significantly worse than its own tree

(p < 0.02), except for Usp9x (p < 0.07). Similarly, all three of the molecular data

sets were significantly worse (p < 0.01) when constrained to the morphological

tree as opposed to their own tree. In terms of character congruence, the incongru-

ence length difference test (Farris et al. 1995) identified the morphological data set

as significantly different from the molecular data sets. Taxonomic congruence

summarizes these topological and character differences because the three nuclear

gene trees corroborate the split of the NewWorld taxa into the subtribes Diclidurina

and Saccopterina, which are clades not recovered by the morphological tree. Except

for a collapse to a polytomy at the basal node of the subtribe Saccopterina in the

parsimony tree, combining the morphological and molecular data sets resulted in

the same topology as the nuclear tree for both Bayesian and parsimony analyses.

This indicates that the morphological dataset has a lot of homoplasy with very little

phylogenetic signal.

17.5.2 Ecological Data

The most comprehensive ecological study incorporated 28 characters primarily

associated with roosting and foraging behavior; however, data for most of the

species were unknown (Dunlop 1998; Lim and Dunlop 2008). A phylogenetic

analysis of this incomplete dataset resulted in a largely unresolved topology.

A combined analysis of morphological and behavioral characters resulted in a

slightly better but still poorly resolved consensus tree of 509 equally parsimonious

trees. The only higher level relationships recovered were the subfamilies Tapho-

zoinae and Emballonurinae.


Although there is a lack of resolving power because of high levels of homoplasy

and large amounts of missing data, characters can be optimized onto the robust

molecular phylogeny to hypothesize evolutionary patterns in morphology and

behavior. Three examples are detailed herein that are associated with the diversifi-

cation of genera of New World emballonurid bats.

17.5.3 Wing Sacs

Species of Balantiopteryx, Cormura, Peropteryx, and Saccopteryx have a sac-like

structure in the propatagium between the shoulder and forearm that is uniquely

structured in each of the genera in terms of location in the wing membrane,

direction of the opening, and size. However, only the wing sac in S. bilineata has

been thoroughly studied. It is well developed in males and acts as a storage

container without glandular cells (Scully et al. 2000) for bodily secretions used in

a salting behavior to mark females in the harem (Voigt and von Helversen 1999).

Based on both a parsimony and likelihood method of ancestral state reconstruction

as implemented in Mesquite (Maddison and Maddison 2006), wing sac character

states mapped independently onto the molecular phylogeny (Fig. 17.6; Lim and

Dunlop 2008). An alternative hypothesis of a single origin of wing sacs for New

World emballonurid bats is less parsimonious with two additional losses and it is

also not supported by the likelihood method of ancestral state reconstruction, which

predicts no wing sac at the base of this clade. However, because of multiple

occurrences of sac-like structures in different genera, there is a possibility of a

phylogenetic predisposition (Soltis et al. 1995) whereby the genetic components

underlying the structure originated once on the tree (Lim and Dunlop 2008).

17.5.4 Roosts and Pelage

Most species of emballonurids and many bats in general have brown fur but some

genera have atypical appearances including paler pelage that is white, as in the

ghost bat Diclidurus, gray as in the smoky bat Cyttarops, or a pelage pattern with

two dorsal pale lines as in Rhynchonycteris and Saccopteryx. In terms of primary

roosting sites, most emballonurid bats occupy relatively sheltered areas such as

caves and crevices in rocky outcrops, or in man-made structures such as tombs and

buildings. Some species are primarily found in other forms of concealed roosts

including tree hollows and rotted-out logs. A few genera, however, predominately

roost in more exposed situations including in leaves at the tops of palm trees

(Cyttarops and Diclidurus), or on sloping tree trunks overhanging rivers

(R. naso), vertical tree trunks within forest (S. leptura), and within exposed cavitieson the outside of buttressed roots of trees (S. bilineata). Although Saccopteryx is

also known to roost in other places such as tree hollows, caves, and man-made

294 B.K. Lim

structures, they regularly use the exposed surfaces of trees, unlike other genera that

occupy sheltered areas (Bradbury and Emmons 1974; Bradbury and Vehrencamp

1976). Pelage and roosting behavior map consistently and are correlated on the

phylogeny suggesting a phylogenetic basis to these character systems and an

association of camouflage for genera that roost on exposed substrate such as tree

trunks and leaves at the tops of palm trees (Fig. 17.6; Lim and Dunlop 2008).

Fig. 17.6 Chronogram of New World emballonurid bats with the primary characters defining

the basal diversification during the Late Oligocene and Early Miocene. Echolocation call design:

C1 – frequency high (41.3–98.2 kHz), call duration low (4.8–7.6 ms), and pulse interval low

(58–119 ms); C2 – frequency low (23.5–42.6 kHz), call duration high (8.1–9.7 ms), and pulse

interval high (100–317 ms). Ear morphology: E1 – medial edge of ears arise from between the

eyes; E2 – medial edge of ears are connected between the eyes; E3 – medial edge of ears arise

above the inner portion of the eyes; E4 – medial edge of ears arise above the middle portion of the

eyes; E5 – medial edge of ears arise above the outer portion of the eyes. Pelage pattern: P1 – fur

typically a uniformly medium or dark brown color; P2 – fur has 2 wavy pale lines on the dorsum;

P3 – fur is pale gray; and P4 – fur is brownish white or white. Roost site: R1 – lives in shelter area;

R2 – lives in exposed areas on tree trunks; and R3 – lives in exposed areas under palm leaves.

Wing sacs: W1 – no wing sacs; W2 – large-sized wing sacs located along the forearm of the

propatagia; W3 – medium-sized wing sacs located in the middle of the propatagia; W4 – small-

sized and conspicuous wing sacs located near the leading edge of the propatagia; and W5 – small-

sized and inconspicuous wing sacs located near the leading edge of the propatagia


17.5.5 Ear Morphology and Echolocation

Although bats are not the only mammals that echolocate, they have the most

sophisticated system of high frequency emission, sound reception, and neural

processing for navigating and foraging in the dark. Ear shape and position are

important factors for receiving returning echoes. The position of the medial edge of

the ear in relation to the eye dictates the degree of forward or lateral orientation of

the ear on the head of the bat. The direction of the ear may in turn influence the

ecological adaptation of flying behavior. The more basal nodes for extant bats are

equivocal for ear position because of polymorphic states in most families and the

lack of comprehensive intrafamilial phylogenies (Lim and Dunlop 2008). Nonethe-

less, a possible accelerated character transformation is an ancestral state reconstruc-

tion of the ear directed more forward with the medial edge located between the eyes

at the base of the New World emballonurid tree (Fig. 17.6). More laterally directed

ears as seen in the subtribe Saccopterygina would be considered derived states.

New World emballonurid bats are all aerial insectivores with an echolocation

search call consisting of a central quasi-constant frequency band with short fre-

quency modulated components and multiharmonics with most of the energy in the

second harmonic. There is a negative correlation of a decrease in flying distance to

forest clutter with an increase in peak echolocation frequency and a positive

correlation of a decrease in pulse interval and call duration with a decrease in

distance to clutter (Jung et al. 2007). These acoustic parameters map consistently on

the phylogeny suggesting that foraging habitat and echolocation call design reflect

phylogenetic relationships. Species within the subtribe Saccopterygina (Centronyc-teris, Rhynchonycteris, and Saccopteryx) fly in more cluttered environments within

the forest or near the edge of forest and have higher frequencies, shorter pulse

intervals, and shorter call durations (Fig. 17.6). In contrast, the subtribe Diclidurina

(Balantiopteryx, Cormura, Cyttarops, Diclidurus, and Peropteryx) fly in less clut-

tered environment in open spaces near the forest or above the canopy and have

lower frequencies, longer pulse intervals, and longer call durations. If ear position-

ing is linked to echolocation parameters and flying behavior, foraging near to forest

clutter would be considered a derived ecological adaptation for Saccopterygina

because forward directed ears are considered ancestral for New World emballonur-

ids and are also found in Diclidurina.

17.6 Conclusions

The most recent common ancestor of New World emballonurid bats colonized an

insular South America from Africa during the Early Oligocene 30 million years ago

when savannah was more prevalent than today. A basal split occurred approxi-

mately 27 million years ago in the Northern Amazon with the speciation of the

subtribes Saccopterygina in forested habitats and Diclidurina in savannah. There

296 B.K. Lim

was relative stasis until a rapid differentiation of genera 19.4–18.0 million years ago

during the Early Miocene when marine incursions from the Caribbean into the

northwestern Amazon region resulted in heterogeneous environments in a forest-

savannah mosaic. The uplands of the Guiana Shield acted as a stable core area

during range contractions. Subsequent range expansions back into favorable low-

land habitats completed episodes of taxon pulses of biotic diversification. These

changing paleoenvironments in the Early Miocene resulted in an adaptive radiation

occurring in forested habitats that gave rise to the differentiation of the genera in

Saccopterygina. The association of ear morphology and echolocation call design

suitable for foraging within cluttered environments supports a phylogenetic basis to

the evolution of these complex character systems. A similar radiation occurred in

savannah habitats giving rise to the diversification of genera in Diclidurina that

were adapted to foraging in more open environments. More detailed study of

morphology, ecology, and echolocation of emballonurids at the species-level in a

phylogenetic context will give further insights into the remarkable evolutionary

history and adaptive radiation of bats.

Acknowledgments I thank Mark Engstrom for critical comments throughout the formulation of

the ideas presented herein. Primary funding for fieldwork and research was secured through the

generous support of the Royal Ontario Museum Governors and Department of Natural History.

References

Barghoorn SF (1977) New material of Vespertiliavus Schlosser (Mammalia, Chiroptera) and

suggested relationships of emballonurid bats based on cranial morphology. Am Mus Novit

2618:1–29

Borissenko AV (1999) A mobile trap for capturing bats in flight. Plecotus et al 2:10–19

Bradbury JW, Emmons LH (1974) Social organization of some Trinidad bats: 1. Emballonuridae.

Z Tierpsychol 36:137–183

Bradbury JW, Vehrencamp SL (1976) Social organization and foraging in emballonurid bats.

Behav Ecol Sociobiol 1:337–381

Czaplewski NJ (1997) Chiroptera. In: Kay RF, Madden RH, Cifelli RL, Flynn JJ (eds) Vertebrate

paleontology in the neotropics: the Miocene fauna of La Venta, Colombia. Smithsonian

Institution Press, Washington, DC, pp 410–431

Dunlop JM (1998) The evolution of behavior and ecology in Emballonuridae (Chiroptera). PhD

dissertation, York University, North York, Ontario

Eick GN, Jacobs DS, Matthee CA (2005) A nuclear DNA phylogenetic perspective on the

evolution of echolocation and historical biogeography of extant bats (Chiroptera). Mol Biol

Evol 22:1869–1886

Erwin TL (1979) Thoughts on the evolutionary history of ground beetles: hypotheses generated

from comparative faunal analyses of lowland forest sites in temperate and tropical regions. In:

Erwin TL, Ball GE, Whitehead DR (eds) Carabid beetles: their evolution, natural history, and

classification. Dr W. Junk, The Hague, pp 539–592

Erwin TL (1981) Taxon pulses, vicariance, and dispersal: an evolutionary synthesis illustrated by

carabid beetles. In: Nelson G, Rosen DE (eds) Vicariance biogeography: a critique. Columbia

University Press, New York, pp 159–196

Farris JS (1970) Methods for computing Wagner trees. Syst Zool 19:83–92


Farris JS, Kallersjo M, Kluge AG, Bult C (1995) Testing significance of incongruence. Cladistics

10:315–319

Flynn JJ, Wyss AR (1998) Recent advances in South American mammalian paleontology. Trends

Ecol Evol 13:449–454

Griffiths TA, Smith AL (1991) Systematics of emballonuroid bats (Chiroptera: Emballonuridae

and Rhinopomatidae) based on hyoid morphology. Bull Am Mus Nat Hist 206:62–83

Haq BU, Hardenbol J, Vail PR (1987) Chronology of fluctuating sea levels since the Triassic.

Science 235:1156–1167

Hoorn C, Guerrero J, Sarmiento GA, Lorente MA (1995) Andean tectonics as a cause for changing

drainage patterns in Miocene northern South America. Geology 23:237–240

Janis CM (1993) Tertiary mammal evolution in the context of changing climates, vegetation, and

tectonic events. Ann Rev Ecol Syst 24:467–500

Jung K, Kalko EKV, von Helversen O (2007) Echolocation calls in Central American emballo-

nurid bats: signal design and call frequency alternation. J Zool 212:125–137

Kishino H, Hasegawa M (1989) Evaluation of the maximum likelihood estimate of the evolution-

ary tree topologies from DNA sequence data, and the branching order in Hominoidea. J Mol

Evol 29:170–179

Koopman KF (1994) Chiroptera: systematics Part 60 of Mammalia, vol 8, Handbook of Zoology.

Walter de Gruyter, New York

Legendre S (1984) Etude odontologique des representants actuels du groupe Tadarida (Chiroptera,

Molossidae): implications phylogeniques, systematiques et zoogeographiques. Rev Suisse

Zool 91:399–442

Lim BK (2007) Divergence times and origin of neotropical sheath-tailed bats (tribe Diclidurini) in

South America. Mol Phylogenet Evol 45:777–791

Lim BK (2008) Historical biogeography of New World emballonurid bats (tribe Diclidurini):

taxon pulse diversification. J Biogeogr 35:1385–1401

Lim BK (2009) Environmental assessment at the Bakhuis Bauxite Concession: small-sized

mammal diversity and abundance in the lowland humid forests of Suriname. Open Biol J

2:42–57

Lim BK, Dunlop JM (2008) Evolutionary patterns of morphology and behavior as inferred from a

molecular phylogeny of New World emballonurid bats (tribe Diclidurini). J Mammal Evol

15:79–121

Lim BK, Engstrom MD, Simmons NB, Dunlop JM (2004) Phylogenetics and biogeography of

least sac-winged bats (Balantiopteryx) based on morphological and molecular data. Mamm

Biol 69:225–237

Lim BK, Engstrom MD, Bickham JW, Patton JC (2008) Molecular phylogeny of New World

emballonurid bats (Tribe Diclidurini) based on loci from the four genetic transmission systems

in mammals. Biol J Linn Soc 93:189–209

Lim BK, Engstrom MD, Reid FA, Simmons NB, Voss RS, Fleck DW (2010) A new species of

Peropteryx (Chiroptera: Emballonuridae) from western Amazonia with comments on phylo-

genetic relationships within the genus. Am Mus Novit 3686:1–20

Lundberg JG, Marshall LG, Guerrero J, Horton B, Malabarba MCSL, Wesselingh F (1998) The

stage for Neotropical fish diversification: a history of tropical South American rivers. In:

Malabarba LR, Reis RE, Vari RP, Lucena ZMS, Lucena CAS (eds) Phylogeny and classifica-

tion of Neotropical fishes. Edipucrs, Porto Alegre, Brazil, pp 13–48

Maddison WP, Maddison DR (2006) Mesquite: a modular system for evolutionary analysis,

version 1.12. http://mesquiteproject.org. Accessed 23 Sept 2006

McKenna MC, Bell SK (1997) Classification of mammals above the species level. Columbia

University Press, New York

Miller KG, Kominz MA, Browning JV, Wright JD, Mountain GS, Katz ME, Sugarman PJ,

Cramer BS, Christie-Blick N, Pekar SF (2005) The Phanerozoic record of global sea-level

change. Science 310:1293–1298

298 B.K. Lim

http://mesquiteproject.org

Munoz J, Cuartas CA (2001) Saccopteryx antioquensis n. sp. (Chiroptera: Emballonuridae) del

noroeste de Colombia. Actual Biol 23:53–61

Poux C, Chevret P, Huchon D, de Jong WW, Douzery EJP (2006) Arrival and diversification of

caviomorph rodents and platyrrhine primates in South America. Syst Biol 55:228–244

Prager EM, Wilson AC (1988) Ancient origin of lactalbumin from lysozyme: analysis of DNA and

amino acid sequences. J Mol Evol 27:326–335

Robbins LW, Sarich VM (1988) Evolutionary relationships in the family Emballonuridae (Chir-

optera). J Mammal 69:1–13

Ronquist F, Huelsenbeck JP (2003) MrBayes 3: Bayesian phylogenetic inference under mixed

models. Bioinformatics 19:1572–1574

Scully WMR, Fenton MB, Saleuddin ASM (2000) A histological examination of the holding sacs

and glandular scent organs of some bat species (Emballonuridae, Hipposideridae, Phyllosto-

midae, Vespertilionidae, and Molossidae). Can J Zool 78:613–623

Shimodaira H, Hasegawa M (2001) CONSEL: for assessing the confidence of phylogenetic tree

selection. Bioinformatics 17:1246–1247

Simmons NB (2005) Order Chiroptera. In: Wilson DE, Reeder DM (eds) Mammal species of the

world: a taxonomic and geographic reference, 3rd edn. Johns Hopkins University Press,

Baltimore, pp 312–529

Simmons NB, Voss RS (1998) The mammals of Paracou, French Guiana: a neotropical lowland

rainforest fauna. Part 1, bats. Bull Am Mus Nat Hist 237:1–219

Soltis DE, Soltis PS, Morgan DR, Swensen SM, Mullin BC, Dowd JM, Martin PG (1995)

Chloroplast gene sequence data suggest a single origin of the predisposition for symbiotic

nitrogen fixation in angiosperms. Proc Natl Acad Sci USA 92:2647–2651

Swofford DL (2001) PAUP*: phylogenetic analysis using parsimony (*and other methods),

version 4.0b10. Sinauer Associates, Sunderland, MA

Takai M, Anaya F, Shigehara N, Setoguchi T (2000) New fossil materials of the earliest New

World onkey, Branisella boliviana, and the problem of platyrrhine origins. Am J Phys

Anthropol 111:263–281

Teeling EC, Springer MS, Madsen O, Bates P, O’Brien SJ, Murphy WJ (2005) A molecular

phylogeny for bats illuminates biogeography and the fossil record. Science 307:580–584

Templeton AR (1983) Phylogenetic inference from restriction endonuclease cleavage site maps

with particular reference to the evolution of humans and the apes. Evolution 37:221–244

Thorne JL, Kishino H (2002) Divergence time and evolutionary rate estimation with multilocus

data. Syst Biol 51:689–702

Voigt CC, von Helversen O (1999) Storage and display of odour by male Saccopteryx bilineata(Chiroptera, Emballonuridae). Behav Ecol Sociobiol 50:29–40

Wilson DE, Reeder DM (eds) (2005) Mammal species of the world: a taxonomic and geographic

reference, 3rd edn. Baltimore, Johns Hopkins University Press

Wojcicki M, Brooks DR (2005) PACT: an efficient and powerful algorithm for generating area

cladograms. J Biogeogr 32:755–774

Wyss AR, Flynn JJ, Norell MA, Swisher CC, Charrier R, Novacek MJ, McKenna MC (1993)

South America’s earliest rodent and recognition of a new interval of mammalian evolution.

Nature 365:434–437


Chapter 18

Trends in Rhizobial Evolution and Some

Taxonomic Remarks

Julio C. Martınez-Romero, Ernesto Ormeno-Orrillo, Marco A. Rogel,

Aline Lopez-Lopez, and Esperanza Martınez-Romero

Abstract Bacteria that establish nitrogen-fixing symbiosis in specialized plant

structures belong to only three of over 100 bacterial phyla. Among these, rhizobial

symbioses are the best known and nodulation genes (nod) have been described in

many species. nodA phylogenies revealed a larger diversity in Bradyrhizobium than

in other genera and suggest that bradyrhizobial nod genes are the oldest in agree-

ment to the proposal that nod genes evolved in Bradyrhizobium (Plant Soil

161:11–20, 1994). In many cases, rhizobial symbiotic and housekeeping genes

have different evolutionary histories in relation to the lateral transfer of symbiotic

genes among bacteria. Misclassified Rhizobium strains were identified, to properly

identify rhizobial species we propose the use of fragments of the rpoB and dnaKgenes, which according to probability analyses reflect the behavior of whole genes.

With these analyses several rhizobial species related to Agrobacterium tumefaciensmay be reclassified to a genus other than Rhizobium.

18.1 Introduction

Legume plants are widespread and diverse with a large number of species; they

profit from symbiosis with nitrogen-fixing bacteria (collectively designated as

rhizobia and comprising different, not closely related genera, such as Bradyrhizo-bium, Mesorhizobium, Azorhizobium, Sinorhizobium, Rhizobium, and others) that

induce the formation of nodules on roots and rarely on stems and provide nitrogen

that allows the plants to grow in nitrogen poor soils. Rhizobia are used as inoculants

in agriculture, a practice that has been in use for over a hundred years, substituting

fertilizers and saving millions of dollars in some cases (Hungria et al. 2000, 2005).

J.C. Martınez-Romero, E. Ormeno-Orrillo, M.A. Rogel, A. Lopez-Lopez, and E. Martınez-

Romero

Centro de Ciencias Genomicas, UNAM, Av. Universidad, Cuernavaca, Morelos 62210, Mexico



301

Rhizobial evolution and diversity (reviewed in Terefework et al. 2000; Wang

and Martınez-Romero 2000; Sprent 2001; Sessitsch et al. 2002; Provorov and

Vorobyov 2008; Martinez-Romero 2009) and molecular mechanisms mediating

their interaction with legume hosts (Barnett and Fisher 2006; Jones et al. 2007) have

been studied for a small proportion of legume-rhizobial symbioses (Lopez-Lopez

et al. 2010). The coevolution of Rhizobium and legumes in symbiosis has been

critically analyzed (Sprent 1997; Martinez-Romero 2009).

18.2 Nitrogen-Fixing Symbioses with Plants

In plants with nitrogen-fixing symbiosis, special structures are involved (Fig. 18.1)

indicating a sort of “convergent evolution” and suggesting a need to contain (in

specialized structures) large numbers of selected bacteria to provide enough nitro-

gen for plants and/or to confine, control, or protect bacteria. Few bacterial genera

belonging to only three phyla (out of over 100 current bacterial phyla) are capable

of forming these nitrogen-fixing symbioses with plants (Fig. 18.1). There are more

phyla with nitrogen-fixing bacteria than with nodulating bacteria, suggesting that

nodulating bacteria evolved from nitrogen-fixing bacteria. Other bacteria out of the

complex community found associated with plants, such as Azoarcus (Hurek and

Reinhold-Hurek 2003) and Herbaspirillum (Roncato-Maccari et al. 2003), fix low

levels of nitrogen not in nodules but as endophytes (inside plants); maybe rhizobial

nitrogen fixation started similarly, as low-level nitrogen fixation. It is the aim of

applied research with some of the plant-associated bacteria to achieve similar levels

of nitrogen fixation with rice, corn, sugar-cane, and potatoes, as those obtained with

the well recognized nitrogen-fixing symbioses of plants.

Bacteria induce the formation of nodules on actinorrhizal plants and in legumes

(including the nonlegume Parasponia) while cyanobacteria do not induce coralloidroots in cycads (an older symbiosis than those of legumes and actinorrhizal plants),

Firm

icute

sSp

iroch

aete

sCya

noba

cter

ia*

Actin

obac

teria

*Chl

orob

i

Prot

eoba

cter

ia*

Fig. 18.1 Bacterial phyla, names correspond to phyla containing nitrogen-fixing species. Asterisks(*) indicate phyla containing bacteria that establish symbiosis in specialized structures

302 J.C. Martınez-Romero et al.

and seemingly neither the specialized cavities in Azolla and Gunnera, such struc-

tures formed normally by plants are subsequently colonized by cyanobacteria.

Rhizobia and actinobacteria become intracellular in nodules as do cyanobacteria

in Gunnera. Interestingly, in Casuarina glauca, an actinorhizal plant, a legume

symbiotic gene (symRK) has been found that is required for nodulation suggesting

a common genetic basis for nodule formation in legumes and actinorrhizal plants

(Gherbi et al. 2008).

A landmark in symbiotic research in Rhizobium was the discovery of the

inducing molecules (Lerouge et al. 1990), Nod factors (produced by enzymes

encoded by nod genes), which have a unique structure in biology, are active at

nanomolar concentrations and are capable of inducing nodules in the absence of

bacteria (Denarie et al. 1996; Relic et al. 1994). Great interest and much effort have

been devoted toward identifying nodulation factors in actinobacteria but results

have not been reported yet. Genetic approaches in the 1980s led to the discovery of

nodulation mutants in Rhizobium and nod genes were described then (Long et al.

1983; Kondorosi et al. 1984). With the exception of photosynthetic Bradyrhizobiumnodulating some Aeschynomene species on stems (Giraud et al. 2007), all other

rhizobial species use Nod factors to induce nodules on legume roots. Furthermore,

the acquisition of nod genes in some nonsymbiotic bacteria makes them form

nodules (see later). The nodABC genes constitute an operon in most rhizobia.

Exceptions are Rhizobium etli biovar phaseoli with nodA separated from nodBC(Vazquez et al. 1991) andMesorhizobium loti where nodB does not form an operon

with nodA and C (Sullivan et al. 2002). nodABC genes encode the enzymes that

synthesize the core of the Nod factor: nodC encodes an N-acetylglucosaminyltrans-

ferase, nodB a chitooligosaccharide deacetylase, and nodA specifies the N-acylationof the aminosugar backbone by different fatty acids (Atkinson et al. 1994; Debelle

et al. 1996a; Roche et al. 1996). Other nod gene products act to add chemical

modifications to the Nod factor (Relic et al. 1994; Ferro et al. 2000), mediate its

secretion (Evans and Downie 1986), provide precursors (Baev et al. 1991), or

regulate nod gene expression (Mulligan and Long 1985; Kondorosi et al. 1991).

18.3 nod Gene Evolution

Where do nod genes originally come from? A hyaluronate synthase (hyaluronic

acid is an polymer of alternative N-acetylglucosamine and glucuronic acid) from

Streptococcus has sequence similarities to NodC, DG42 from Xenopus, and chitin

synthases from yeast. Some bacterial xylanases (that catalyze the hydrolysis of

linked xylose oligomeric and polymeric substrates) contain domains homologous to

NodB proteins (Laurie et al. 1997). A Bacillus strain produces a molecule

seemingly structurally related to Nod factors that stimulates plant proliferation

(Lian et al. 2001).

Interestingly, some plant mutants affecting rhizobial nodulation are defective in

the mycorrhization process (Oldroyd et al. 2005) and it is suggested that a common

18 Trends in Rhizobial Evolution and Some Taxonomic Remarks 303

signaling pathway exists for Nod factor perception and mycorrhizal symbiosis

(Catoira et al. 2000; Gianinazzi-Pearson and Denarie 1997). Mycorrhizal symbiosis

occurs in around 80% of all plants and is considered as old as the first plants that

evolved on Earth. The Nod factor may be considered as a very small chitin

molecule that subsequently acquired other chemical modifications, some of them

involved in protecting the molecule from plant chitinases (Staehelin et al. 1994).

Mycorrhiza, being fungi, have chitin. Maybe rhizobia mimicked micorrhizal sym-

biosis (Debelle et al. 1996b).

nod gene phylogenies have been reported in Bradyrhizobium, Rhizobium,Mesorhizobium, and Sinorhizobium (Moulin et al. 2004; Steenkamp et al. 2008;

Stepkowski et al. 2007; Han et al. 2008; Rincon-Rosales et al. 2009). A host

correlation to nod genes has been recognized (Suominen et al. 2001) and Nod

factor fucosylation and acetylation have been correlated to bacterial phylogenies

and specificities (Moulin et al. 2004); bacteria with sulfate modifications are

scattered in rhizobial phylogenies (Martınez et al. 1995). We constructed a phylo-

genetic tree with available reported nodA sequences (Fig. 18.2). There seems to be a

larger diversity of nodA sequences in Bradyrhizobium compared with the diversity

in b-Proteobacteria or Sinorhizobium. In 1994, we proposed the hypothesis that nodgenes evolved in Bradyrhizobium and that they were later transferred to other

genera such as Rhizobium (Martinez-Romero 1994). In Bradyrhizobium, an ances-

tral nod group has been identified from bacteria nodulating several diverse legumes

(indicated in Fig. 18.2), supposedly this group of legumes extended over many parts

of the world during the Eocene after the origin of legumes north of the Tethys Sea

(Steenkamp et al. 2008). Bradyrhizobium are the main nodule bacteria of tropical

tree legumes (Qian et al. 2003; Moreira et al. 1998; Parker 2004; Ormeno-Orrillo

et al. 2006) with a low degree of specificity and tropical legumes are considered

older than temperate legumes. We found 23 novel lineages of Bradyrhizobium in

the rain forest of Los Tuxtlas in Veracruz, Mexico, and they exhibited low speci-

ficity (Ormeno-Orrillo submitted). Specificity is a characteristic of many temperate

legumes and few tropical legumes and may have been acquired later in bacteria

(Perret et al. 2000; Young et al. 2003).

Most nodule forming bacteria belong to the a-Proteobacteria and few to

b-Proteobacteria (Moulin et al. 2001; Chen et al. 2003). Lateral transfer of nod genesto b-Proteobacteria was considered to account for the existence of nodulation in

Burkholderia and Cupriavidus nodulating species (Moulin et al. 2001; Amadou et al.

2008), inDevosia (Rivas et al. 2002), and in Phyllobacterium (Valverde et al. 2005).

18.4 Different Evolutionary Histories of Chromosomal

and Symbiotic Genes

In Rhizobium, Sinorhizobium, and in b-Proteobacteria, symbiotic genes including

nod and nif (nitrogen fixation) genes are located on plasmids (Amadou et al. 2008)

that may be transferred among species both in the laboratory and in nature.


Sin

orhi

zobi

um

Mes

orhi

zobi

um

B. t

uber

um

Rhi

zobi

um/

Sin

orhi

zobi

um

M. n

odul

ans

Bradyrhizobium

Mes

orhi

zobi

um

Rhi

zobi

um/

Sin

orhi

zobi

umA

zorh

izob

ium

Mes

orhi

zobi

um

Bur

khol

deria

/ C

upria

vidu

s

Rhi

zobi

um/

Sin

orhi

zobi

um

Fig.18.2

NodA

genephylogenyin

differentrhizobialgenera


In Mesorhizobium except Mesorhizobium amorphae (Wang et al. 1999b), in

Azorhizobium, in Methylobacterium, and in Bradyrhizobium, symbiotic genes are

on the chromosome. Symbiotic islands have been found to be transferable among

mesorhizobia in the environment (Sullivan et al. 1995; Sullivan and Ronson 1998;

Nandasena et al. 2007). Evidence that transfer and recombination occurs in nature is

obtained by comparing housekeeping and nod gene phylogenies revealing differentevolutionary histories in symbiotic and housekeeping genes (Haukka et al. 1998;

Steenkamp et al. 2008). In the laboratory plant pathogens such as Agrobacteriumtumefaciens and opportunistic human pathogens as Ochrobactrum may become

fully symbiotic by acquiring symbiotic plasmids from Rhizobium tropici,albeit with reduced levels of nitrogen fixation (Martinez et al. 1987; Rogel et al.

2006). Two highly diverging lineages of R. tropici (type A and B) harbor very

similar symbiotic plasmids that we suppose are exchanged among these lineages

(Martınez-Romero 1996).

Biovars were defined in Rhizobium as the different symbiotic specificities

(mainly plasmid encoded) that could be exhibited in a single chromosomal back-

ground (species). As such three biovars were recognized in Rhizobium legumino-sarum (viciae, trifolii, and phaseoli) (Jordan 1984); however, recently a more

complicated situation has been revealed and some R. leguminosarum strains have

been assigned to different species: Rhizobium pisi (Ramırez-Bahena et al. 2008)

and Rhizobium fabae (Tian et al. 2008). The symbiotic plasmid from biovar

phaseoli in R. etli is highly conserved (Gonzalez et al. 2010) may be in relation

to a recent evolutionary origin (Martinez-Romero 2009) maybe as recent as Pha-seolus vulgaris, dating of around 2–3 million years ago (Delgado-Salinas et al.

2006). We identified a new biovar in R. etli, biovar mimosae, and supposed that it

was a more ancient plasmid than the phaseoli plasmid (Wang et al. 1999a); nodgene phylogenies seem to support this hypothesis.

Nonrandom association between plasmid and chromosome markers (Young

et al. 2003) and limited plasmid transfer have been observed in nodule bacteria

(Wernegreen and Riley 1999); however, different evolutionary histories of symbi-

otic and metabolic genes or chromosomal markers have been recognized in some

cases in rhizobia (Silva et al. 2005; Tian et al. 2007; Han et al. 2008; Rincon-

Rosales et al. 2009). Two sympatric species of Sinorhizobium nodulating wild

Acaciellas in Mexico seem to contain the same symbiotic plasmid, and incon-

gruencies in symbiotic and housekeeping phylogenies have been repeatedly

observed in sinorhizobia (Haukka et al. 1998; Toledo et al. 2003; Lloret et al.

2007). African Sinorhizobium terangae is a close relative to these American

sinorhizobia but not on the basis of symbiotic genes (Rincon-Rosales et al. 2009)

(Fig 18.3). In symbionts of Galega orientalis and Galega officinalis (two native

legumes from the Caucasus), there is evidence of transfer of symbiotic information

(Andronov et al. 2003). In Bradyrhizobium japonicum, a biovar with symbiotic

genes specific for genistoid wild legumes is also found in another species

B. canariense (Vinuesa et al. 2005). Lateral transfer of symbiotic genes is recog-

nized to have occurred in Bradyrhizobium nodulating a diversity of wild legumes

(Steenkamp et al. 2008).


Symbiotic plasmids in rhizobia are repABC plasmids. repABC plasmids are

characteristic of a-Proteobacteria and differences in repA, repB, and repC gene

evolution have been reported (Castillo-Ramirez et al. 2009), supporting the occur-

rence of large recombination rates in plasmids. Genomic analyses have revealed

mosaicism in symbiotic plasmids (Gonzalez et al. 2006). Genetic information in

plasmids has been described as accessory or the mobile genome (Young et al.

2006). Plasmid (and maybe also genomic island) plasticity may have been instru-

mental for the adaptation of rhizobia to legume evolution and specificity (Martinez-

Romero 2009).

18.5 Chromosomal Evolution and Molecular Markers

Rhizobial lineages have been estimated to be nearly as old as plants, for example,

Rhizobium and Bradyrhizobium last common ancestor was dated as being over 400

million years old but legumes evolved around 100–65 million years ago (Sprent

2001). Nodulation seemingly evolved (Young and Johnston 1989), in only one

group of bacteria that were associated with plants (maybe as endophytes, Martinez-

Romero 2009). Further spread of nod genes by lateral gene transfer may have

conferred to diverse genera their nodulating capacity.

S. americanum

S. fredii

S. saheli

S. mexicanum

S. terangae

S. chiapanecum

S. kostiense

S. arboris

S. meliloti

S. medicae

S. adhaerens

S. morelense

rpoBS. americanum

S. fredii bv. mediterranense

S. mexicanum

S. chiapanecum

Mesorhizobium de acacias

S. kostiense

S. saheli

S. arboris

S. terangae

nodA

Fig. 18.3 Schematic comparison of chromosomal and symbiotic gene phylogenies in Sinorhizobium


In 1989, it was suggested that “We will eventually need many genera to

accommodate all the root-nodule bacteria” (Young and Johnston 1989), up to

now 13 genera and over 50 species have been described establishing symbioses

with a small sample of legumes analyzed. Small subunit ribosomal (16S rRNA)

gene sequences have been commonly used to identify and propose species in

rhizobia (Wang and Martınez-Romero 2000). It is remarkable that in spite of the

large divergence of nod gene sequences found in Bradyrhizobium, this genus

exhibits only a very limited diversity of 16S rRNA genes (Barrera et al. 1997;

Vinuesa et al. 2005) and species delineation is not clear with this marker. Several

molecular markers have been used to establish phylogenies and identify new

species not only in Bradyrhizobium but in rhizobia in general. Genomic information

provides large numbers of genes for these analyses (Young et al. 2006; Gonzalez

et al. 2006; Crossman et al. 2008) and congruent bacterial relationships have been

reported using indel analyses (Gupta 2005). Alternative phylogenetic relationships

are encountered in multiple gene analyses from reported complete genomes of

Agrobacterium, Rhizobium, and Sinorhizobium (Young et al. 2006); this suggests

that the divergence of these lineages occurred within a very short time as has been

concluded for other a-Proteobacteria (Castillo-Ramırez and Gonzalez 2008).

18.6 Probability Estimates to Distinguish Rhizobial Species

Representative molecular markers are being searched to better reflect species

phylogenies and not single gene phylogenies, in this regard dnaJ was found to

reproduce accepted phylogenetic relationships (Alexandre et al. 2008). rpoB gene

sequences have been considered for diversity studies in very different habitats or

communities (Planet et al. 1995; Dahlloef et al. 2000; Case et al. 2007; Sachman-

Ruiz et al. 2009). We have used partial sequences of rpoB as part of the phylo-

genetic studies to characterize new Sinorhizobium species (Lloret et al. 2007;

Rincon-Rosales et al. 2009) and a new species of Klebseilla (Rosenblueth et al.

2004). rpoB is a large gene (more than 4,140 bp in Rhizobium) and usually, only

fragments of the gene sequence are available. Different studies report sequences of

different fragments, hampering direct comparisons. Sequencing a common frag-

ment will facilitate comparisons and diminish misclassifications. Up to now several

genomes of species within the Rhizobium genus have been completely sequenced.

A practical utility for defining gene divergence ranges is to facilitate proper

identification of novel species and of species belonging to a single species. When

describing Sinorhizobium (Ensifer) mexicanum (Lloret et al. 2007) and Sinorhizo-bium chiapanecum (Rincon-Rosales et al. 2009), we proposed a probability range

of inter- and intraspecies gene differences that allowed the distinction of different

species and bacteria belonging to the same species. Comparing full rpoB gene

sequences from seven Rhizobium genomes, we calculated that the 95% confidence

interval for identities ranges from 0.898 to 1.000 for the sequences within this

genus. The 0.898 threshold provides a useful criterion to determine if a new isolate


belongs to this genus: an identity of less than 0.898 excludes it from being a

Rhizobium. Nevertheless, this is not a practical approach to classify new isolates

due to the large size of rpoB gene, which can hardly be expected to be totally

sequenced in diversity studies considering a large number of strains. Thus, we

examined 700 bp fragments that covered the entire 4,140 bp sequence and found

that the identities of a 700 bp fragment, ranging from positions 2,800 to 3,500,

closely match the distribution of the entire gene sequence (Kolmogorov Smirnoff,

p ¼ 0.05), in contrast to all other fragments analyzed. This fragment would provide

not only a dependable molecular marker to study the phylogenies of rhizobia, but

also a performable one. In both the full gene and the 700 bp (position 2,800–3,500)

fragment, with a 95% confidence it can be stated that while Agrobacterium radio-bacter is within the ranges of Rhizobium, A. tumefaciens, and Agrobacterium vitisidentities to the members of the group do not fall within the limits of the genus in

the distribution that described the dispersion of their differences.

The same analysis was performed for dnaK. For this gene, the 95% confidence

interval for identities ranges from 0.896 to 1.000 for the sequences within Rhizo-bium. Considering this interval, A. radiobacter and Agrobacterium rhizogenes arewithin the ranges of Rhizobium (therefore should be considered Rhizobium radio-bacter and Rhizobium rhizogenes as has been proposed by Young et al. 2001),

whereas A. tumefaciens and A. vitis identities to the members of the group do not

fall within the limits of the genus (Fig. 18.4). Thus, by rpoB and by dnaK analyses,

Agrobacterium could stand as an independent genus from Rhizobium as has been

claimed before (Farrand et al. 2003), in consequence Rhizobium galegae, Rhizo-bium huautlense, Rhizobium cellulosilyticum, Rhizobium selenireducens, and Rhi-zobium daejeonense, all related to A. tumefaciens should be reclassified. It is clear

from many published phylogenetic trees that Rhizobium is not monophyletic.

We encountered several examples of misclassified Rhizobium strains in a 16S

rRNA gene phylogenetic tree (Fig. 18.5), probably because many new isolates are

only recognized by 16S rRNA genes and designation is done based on the closest

relative frequently identified only as the best Blast hit, without further characteri-

zation. Rhizobium mongolense and Rhizobium lusitanum are polyphyletic

(Fig. 18.5). Emendments to such misclassifications should be done.

Agrobacteriumtumefaciens

Agrobacteriumrhizogenes

Agrobacteriumtumefaciens

rpoB

dnaK

Rhizobium

Rhizobium

Fig. 18.4 95% Confidence

intervals for identities of

species within Rhizobiumgenus for rpoB and dnaKgenes. The arrows indicatethe average identity of

Agrobacterium tumefaciensor A. rhizogenes to the

members of Rhizobium genus


Acknowledgments To PAPIIT IN200709 and Michael Dunn for reading the manuscript. Partial

financial support for this project was from GEF PNUMA, TSBF-CIAT. E.M. is grateful to

DGAPA UNAM for a postdoctoral fellowship during her sabattical year at UC Davis in California.

References

Alexandre A, Laranjo M, Young JPW, Oliveira S (2008) dnaJ is a useful phylogenetic marker for

alphaproteobacteria. Int J Syst Evol Microbiol 58:2839–2849

Amadou C, Pascal G, Mangenot S, Glew M, Bontemps C, Capela D, Carrere S, Cruveiller S,

Dossat C, Lajus A, Marchetti M, Poinsot V, Rouy Z, Servin B, Saad M, Schenowitz C, Barbe V,

Batut J, Medigue C, Masson-Boivin C (2008) Genome sequence of the beta-RhizobiumCupriavidus taiwanensis and comparative genomics of rhizobia. Genome Res 18:1472–1483

Andronov EE, Terefework Z, Roumiantseva ML, Dzyubenko NI, Onichtchouk OP, Kurchak ON,

Dresler-Nurmi A, Young JPW, Simarov BV, Lindstroem K (2003) Symbiotic and genetic

diversity of Rhizobium galegae isolates collected from the Galega orientalis gene center in theCaucasus. Appl Environ Microbiol 69:1067–1074

Atkinson EM, Palcic MM, Hindsgaul O, Long SR (1994) Biosynthesis of Rhizobium melilotilipooligosaccharide Nod factors: NodA is required for an N-acyltransferase activity. Proc NatlAcad Sci USA 91:8418–8422

Baev N, Endre G, Petrovics G, Banfalvi Z, Kondorosi A (1991) Six nodulation genes of nod box

locus 4 in Rhizobium meliloti are involved in nodulation signal production: nodM codes for

D-glucosamine synthetase. Mol Gen Genet 228:113–124

EU399697 Rhizobium mongolense CCBAU 05122

AF008130 Rhizobium gallicum R602sp

U89819 Rhizobium mongolense USDA 1844T

U89817 Rhizobium mongolense USDA 1877

U89822 Rhizobium mongolense USDA 2377

AY509212 Rhizobium mongolense S110*

EU256432 Rhizobium sullae CCBAU 85011

DQ196418 Rhizobium leguminosarum bv. viciae PEPSM13

EF141340 Rhizobium leguminosarum bv. phaseoli ATCC 14482

AY998046 Rhizobium etli bv. phaseoli IE4804

DQ648575 Rhizobium etli bv. mimosae Mim 7-4

U28916 Rhizobium etli CFN 42

AY509209 Rhizobium mongolense S152*

EU074200 Rhizobium lusitanum CCBAU 03301*

X67234 Rhizobium tropici IIA LMG9517

EF035070 Rhizobium multihospitium CCBAU 83435

U89832 Rhizobium tropici CIAT899

AY738130 Rhizobium lusitanum P1-7

CP000628 Agrobacterium radiobacter K84

AY945955 Agrobacterium rhizogenes ATCC 11325

EF522124 Agrobacterium rhizogenes CU10

9697

100

6377

96

62

99

90

6181

100

70

100

98

0.002

Fig. 18.5 Rhizobium 16S rRNA gene phylogenies.Misclassified strains are indicated by asterisks (*)


Barnett MJ, Fisher RF (2006) Global gene expression in the rhizobial-legume symbiosis.

Symbiosis 42:1–24

Barrera LL, Trujillo ME, Goodfellow M, Garcia FJ, Hernandez-Lucas I, Davila G, van Berkum P,

Martinez-Romero E (1997) Biodiversity of bradyrhizobia nodulating Lupinus spp. Int J SystBacteriol 47:1086–1091

Case RJ, Boucher Y, Dahlloef I, Holmstroem C, Doolittle WF, Kjelleberg S (2007) Use of 16S

rRNA and rpoB genes as molecular markers for microbial ecology studies. Appl Environ


Castillo-Ramırez S, Gonzalez V (2008) Factors affecting the concordance between orthologous

gene trees and species tree in bacteria. BMC Evol Biol 8:300

Castillo-Ramirez S, Vazquez-Castellanos JF, Gonzalez V, Cevallos MA (2009) Horizontal gene

transfer and diverse functional constrains within a common replication-partitioning system in

Alphaproteobacteria: the repABC operon. BMC Genomics 10:536

Catoira R, Galera C, De Billy F, Penmetsa RV, Journet E-P, Maillet F, Rosenberg C, Cook D,

Gough C, Denarie J (2000) Four genes of Medicago truncatula controlling components of a

Nod factor transduction pathway. Plant Cell 12:1647–1666

Chen W-M, Moulin L, Bontemps C, Vandamme P, Bena G, Boivin-Masson C (2003) Legume

symbiotic nitrogen fixation by b-Proteobacteria is widespread in nature. J Bacteriol

185:7266–7272

Crossman LC, Castillo-Ramırez S, McAnnula C, Lozano L, Vernikos GS, Acosta JL, Ghazoui ZF,

Hernandez-Gonzalez I, Meakin G,Walker AW, HynesMF, Young JPW, Downie JA, Romero D,

Johnston AWB, Davila G, Parkhill J, Gonzalez V (2008) A common genomic framework for a

diverse assembly of plasmids in the symbiotic nitrogen fixing bacteria. PLoS ONE 3(7):e2567

Dahlloef I, Baillie H, Kjelleberg S (2000) rpoB-based microbial community analysis avoids limita-

tions inherent in 16s rRNA gene intraspecies heterogeneity. Appl Environ Microbiol

66:3376–3380

Debelle F, Plazanet C, Roche P, Pujol C, Savagnac A, Rosenberg C, Prome J-C, Denarie J (1996a)

The NodA proteins of Rhizobium meliloti and Rhizobium tropici specify the N-acylation of

Nod factors by different fatty acids. Mol Microbiol 22:303–314

Debelle F, Yang GP, Ferro M, Truchet G, Prome JC, Denarie J (1996b) Rhizobium nodulation

factors in perspective. In: Legocki A, Bothe H, P€uhler A (eds) Biological fixation of nitrogen

for ecology and sustainable agriculture. Springer, Heidelberg, Germany, pp 15–24

Delgado-Salinas A, Bibler R, Lavin M (2006) Phylogeny of the genus Phaseolus (Leguminosae): a

recent diversification in an ancient landscape. Syst Bot 31:779–791

Denarie J, Debelle F, Prome JC (1996) Rhizobium lipo-chitooligosaccharide nodulation factors:

signaling molecules mediating recognition and morphogenesis. Annu Rev Biochem 65:503–535

Evans IJ, Downie JA (1986) The nodI gene product of Rhizobium leguminosarum is closely related

to ATP-binding bacterial transport proteins; nucleotide sequence analysis of the nodI and nodJgenes. Gene 43:95–101

Farrand SK, van Berkum PB, Oger P (2003) Agrobacterium is a definable genus of the family

Rhizobiaceae. Int J Syst Evol Microbiol 53:1681–1687

Ferro M, Lorquin J, Ba S, Sanon K, Prome JC, Boivin C (2000) Bradyrhizobium sp. strains that

nodulate the leguminous tree Acacia albida produce fucosylated and partially sulfated Nod

factors. Appl Environ Microbiol 66:5078–5082

Gherbi H, Markmann K, Svistoonoff S, Estevan J, Autran D, Giczey G, Auguy F, Peret B,

Laplaze L, Franche C, Parniske M, Bogusz D (2008) SymRK defines a common genetic

basis for plant root endosymbioses with arbuscular mycorrhiza fungi, rhizobia, and Frankia-

bacteria. Proc Natl Acad Sci USA 105:4928–4932

Gianinazzi-Pearson V, Denarie J (1997) Red carpet genetic programmes for root endosymbioses.

Trends Plant Sci 2:371–372

Giraud E, Moulin L, Vallenet D, Barbe V, Cytryn E, Avarre J-C, Jaubert M, Simon D, Cartieaux F,

Prin Y, Bena G, Hannibal L, Fardoux J, Kojadinovic M, Vuillet L, Lajus A, Cruveiller S, Rouy

Z, Mangenot S, Segurens B, Dossat C, Franck WL, Chang W-S, Saunders E, Bruce D,


Richardson P, Normand P, Dreyfus B, Pignol D, Stacey G, Emerich D, Vermeglio A,

Medigue C, Sadowsky M (2007) Legumes symbioses: absence of nod genes in photosynthetic

bradyrhizobia. Science 316:1307–1312

Gonzalez V, Santamaria RI, Bustos P, Hernandez-Gonzalez I, Medrano-Soto A, Moreno-

Hagelsieb G, Janga SC, Ramirez MA, Jimenez-Jacinto V, Collado-Vides J, Davila G (2006)

The partitioned Rhizobium etli genome: genetic and metabolic redundancy in seven interacting

replicons. Proc Natl Acad Sci USA 103:3834–3839

GonzalezV,Acosta JL, Santamarıa RI, Bustos P, Fernandez JL,HernandezGonzalez IL,DıazR, Flores

M, Palacios R, Mora J, Davila G (2010) Conserved symbiotic plasmid DNA sequences in the

multireplicon pangenomic structure of Rhizobium etli. Appl Environ Microbiol 76:1604–1614

Gupta RS (2005) Protein signatures distinctive of a-Proteobacteria and its subgroups and a model

for a-proteobacterial evolution. Crit Rev Microbiol 31:101–135

Han TX, Wang ET, Han LL, Chen WF, Sui XH, Chen WX (2008) Molecular diversity and

phylogeny of rhizobia associated with wild legumes native to Xinjiang, China. Syst Appl


Haukka K, Lindstrom K, Young JPW (1998) Three phylogenetic groups of nodA and nifH genes in

Sinorhizobium andMesorhizobium isolates from leguminous trees growing in Africa and Latin

America. Appl Environ Microbiol 64:419–426

Hungria M, Vargas MAT, Campo RJ, Chueire LMO, Andrade DS (2000) The Brazilian experience

with the soybean (Glycine max) and common bean (Phaseolus vulgaris) symbioses. In:

Pedrosa FO, Hungria M, Yates G, Newton WE (eds) Nitrogen fixation: from molecules to

crop production. Kluwer Academic Publishers, Netherlands, p 515

Hungria M, Franchini JC, Campo RJ, Graham PH (2005) The importance of nitrogen fixation to

soybean cropping in South America. In: Werner D, Newton WE (eds) Nitrogen fixation in

agriculture, forestry, ecology, and the environment. Springer, Dordrecht, pp 25–42

Hurek T, Reinhold-Hurek B (2003) Azoarcus sp. strain BH72 as a model for nitrogen-fixing grass

endophytes. J Biotechnol 106:169–178

Jones KM, Kobayashi H, Davies BW, Taga ME, Walker GC (2007) How rhizobial symbionts

invade plants: the Sinorhizobium-Medicago model. Nat Rev Microbiol 5:619–633

Jordan DC (1984) Family III. RhizobiaceaeConn 1938, 321AL. In: Krieg NR, Holt JG (eds) Bergeys’s

manual of systematic bacteriology, vol 1. The Williams and Wilkins Co., Baltimore, pp 234–254

Kondorosi E, Banfalvi Z, Kondorosi A (1984) Physical and genetic analysis of a symbiotic region

of Rhizobium meliloti: identification of nodulation genes. Mol Gen Genet 193:445–452

Kondorosi E, Pierre M, Cren M, Haumann U, Buire M, Hoffmann B, Schell J, Kondorosi A (1991)

Identification of NolR, a negative transacting factor controlling the nod regulon in Rhizobiummeliloti. J Mol Biol 222:885–896

Laurie JI, Clarke JH, Ciruela A, Faulds CB, Williamson G, Gilbert HJ, Rixon JE, Millward-Sadler

J, Hazlewood GP (1997) The NodB domain of a multidomain xylanase from Cellulomonas fimideacetylates acetylxylan. FEMS Microbiol Lett 148:261–264

Lerouge P, Roche P, Faucher C, Maillet F, Truchet G, Prome JC, Denarie J (1990) Symbiotic

host-specificity of Rhizobium meliloti is determined by a sulphated and acylated glucosamine

oligosaccharide signal. Nature 344:781–784

Lian B, Prithiviraj B, Souleimanov A, Smith DL (2001) Evidence for the production of chemical

compounds analogous to nod factor by the silicate bacterium Bacillus circulans GY92.

Microbiol Res 156:289–292

Lloret L, Ormeno-Orrillo E, Rincon R, Martınez-Romero J, Rogel-Hernandez MA, Martınez-

Romero E (2007) Ensifer mexicanus sp. nov. a new species nodulating Acacia angustissima(Mill.) Kuntze in Mexico. Syst Appl Microbiol 30:280–290

Long SR, Buikema WJ, Ausubel FM (1983) Cloning of Rhizobium meliloti nodulation genes by

direct complementation of Nod-mutants. Nature 298:485–487

Lopez-Lopez A, Rosenblueth M, Martınez J, Martınez-Romero E (2010) Rhizobial symbioses in

tropical legumes and non-legumes. In: Dion P (ed) Soil biology and agriculture in the tropics.

Springer Heidelberg, pp. 163–184


Martinez E, Palacios R, Sanchez F (1987) Nitrogen-fixing nodules induced by Agrobacteriumtumefaciens harboring Rhizobium phaseoli plasmids. J Bacteriol 169:2828–2834

Martınez E, Laeremans T, Poupot R, Rogel MA, Lopez L, Garcıa F, Vanderleyden J, Prome JC,

Lara F (1995) Nod metabolites and other compounds excreted by Rhizobium spp. In: Tikho-

novich IA, Provorov NA, Romanov VI, Newton WE (eds) Nitrogen fixation: fundamentals and

applications. Kluwer Academic Publishers, Dordrecht, pp 281–286

Martinez-Romero E (1994) Recent developments in Rhizobium taxonomy. Plant Soil 161:11–20

Martinez-Romero E (2009) Coevolution in Rhizobium-legume symbiosis? DNA Cell Biol

28:361–370

Martınez-Romero E (1996) Comments on Rhizobium systematics. Lessons from R. tropici andR. etli. In: Stacey G, Mullin B, Gresshoff PM (eds) Biology of plant–microbe interactions.

International Society for Molecular Plant–Microbe Interactions, St. Paul, Minnesota,

pp 503–508

Moreira FMS, Haukka K, Young JPW (1998) Biodiversity of rhizobia isolated from a wide range

of forest legumes in Brazil. Mol Ecol 7:889–895

Moulin L, Munive A, Dreyfus B, Boivin-Masson C (2001) Nodulation of legumes by members of

the b�subclass of Proteobacteria. Nature 411:948–950

Moulin L, Bena G, Boivin-Masson C, Stepkowski T (2004) Phylogenetic analyses of symbiotic

nodulation genes support vertical and lateral gene co-transfer within the Bradyrhizobiumgenus. Mol Phylogenet Evol 30:720–732

Mulligan JT, Long SR (1985) Induction of Rhizobium meliloti nodC expression by plant exudate

requires nodD. Proc Natl Acad Sci USA 82:6609–6613

Nandasena KG, O’Hara GW, Tiwari RP, Sezmis E, Howieson JG (2007) In situ lateral transfer ofsymbiosis islands results in rapid evolution of diverse competitive strains of mesorhizobia

suboptimal in symbiotic nitrogen fixation on the pasture legume Biserrula pelecinus L.

Environ Microbiol 9:2496–2511

Oldroyd GED, Harrison MJ, Udvardi M (2005) Peace talks and trade deals. Keys to long-term

harmony in legume-microbe symbioses. Plant Physiol 137:1205–1210

Ormeno-Orrillo E, Vinuesa P, Zuniga-Davila D, Martinez-Romero E (2006) Molecular diversity

of native bradyrhizobia isolated from Lima bean (Phaseolus lunatus L.) in Peru. Syst Appl


Parker MA (2004) rRNA and dnaK relationships of Bradyrhizobium sp. nodule bacteria from four

Papilionoid legume trees in Costa Rica. Syst Appl Microbiol 27:334–342

Perret X, Staehelin Ch, Broughton WJ (2000) Molecular basis of symbiotic promiscuity.

Microbiol Mol Biol Rev 64:180–201

Planet P, Jagoueix S, Bove JM, Garnier M (1995) Detection and characterization of the African

citrus greening Liberobacter by amplification, cloning, and sequencing of the rplKAJL-rpoBCoperon. Curr Microbiol 30:137–141

Provorov NA, Vorobyov NI (2008) Equilibrium between the “genuine mutualists” and “symbiotic

cheaters” in the bacterial population co-evolving with plants in a facultative symbiosis.

Theor Popul Biol 74:345–355

Qian J, Kwon S, Parker MA (2003) rRNA and nifD phylogeny of Bradyrhizobium from sites

across the Pacific Basin. FEMS Microbiol Lett 219:159–165

Ramırez-Bahena MH, Garcıa-Fraile P, Peix A, Valverde A, Rivas R, Igual JM, Mateos PF,

Martınez-Molina E, Velazquez E (2008) Revision of the taxonomic status of the species

Rhizobium leguminosarum (Frank 1879) Frank 1889AL, Rhizobium phaseoli Dangeard

1926AL and Rhizobium trifolii Dangeard 1926AL. R. trifolii is a later synonym of R. legumi-nosarum. Reclassification of the strain R. leguminosarum DSM 30132 (¼NCIMB 11478) as

Rhizobium pisi sp. nov. Int J Syst Evol Microbiol 58:2484–2490

Relic B, Perret X, Estrada-Garcia MT, Kopcinska J, Golinowski W, Krishnan HB, Pueppke SG,

Broughton WJ (1994) Nod factors of Rhizobium are a key to the legume door. Mol Microbiol

13:171–178


Rincon-Rosales R, Lloret L, Ponce E, Martinez-Romero E (2009) Rhizobia with different symbi-

otic efficiencies nodulate Acaciella angustissima in Mexico, including Sinorhizobiumchiapanecum sp. nov. which has common symbiotic genes with Sinorhizobium mexicanum.FEMS Microbiol Ecol 68:255–255

Rivas R, Velazquez E, Willems A, Vizcaino N, Subba-Rao NS, Mateos PF, Gillis M, Dazzo FB,

Martinez-Molina E (2002) A new species of Devosia that forms a unique nitrogen-fixing root-

nodule symbiosis with the aquatic legume Neptunia natans (L.f.) Druce. Appl Environ


Roche P, Maillet F, Plazanet C, Debelle F, Ferro M, Truchet G, Prome J-C, Denarie J (1996) The

common nodABC genes of Rhizobium meliloti are host-range determinants. Proc Natl Acad Sci

USA 93:15305–15310

Rogel MA, Torres C, Lloret L, Rosenblueth M, Hernandez-Lucas I, Martınez L, Martınez J,

Martınez-Romero E (2006) Lateral transfer of Rhizobium symbiotic plasmids leading to

genomic innovation. In: Sanchez F, Quinto C, Lopez-Lara IM, Geiger O (eds) Biology of

plant–microbe interactions, vol 5. International Society for Molecular Plant–Microbe Interac-

tions, St. Paul, USA, pp 310–318

Roncato-Maccari LDB, Ramos HJO, Pedrosa FO, Alquini Y, Chubatsu LS, Yates MG, Rigo LU,

Steffens MBR, Souza EM (2003) Endophytic Herbaspirillum seropedicae expresses nif genesin gramineous plants. FEMS Microbiol Ecol 45:39–47

Rosenblueth M, Martinez L, Silva J, Martinez-Romero E (2004) Klebsiella variicola, a novel

species with clinical and plant-associated isolates. Syst Appl Microbiol 27:27–35

Sachman-Ruiz B, Castillo-Rodal AI, Lopez-Vidal Y, Martınez-Romero E, Vinuesa P (2009)

Diversity of environmental mycobacteria in Mexican rivers assessed by cultivation and

metagenomics approaches. In: 109th General Meeting, American Society for Microbiology,

May 17–21, 2009, Philadelphia, Pennsylvania

Sessitsch A, Howieson JG, Perret X, Antoun H, Martinez-Romero E (2002) Advances in

Rhizobium research. Crit Rev Plant Sci 21:323–378

Silva C, Vinuesa P, Eguiarte LE, Souza V, Martinez-Romero E (2005) Evolutionary genetics and

biogeographic structure of Rhizobium gallicum sensu lato, a widely distributed bacterial

symbiont of diverse legumes. Mol Ecol 14:4033–4050

Sprent JI (1997) Co-evolution of legume-rhizobial symbioses:is it essential for either partner? In:

Legocki A, Bothe H, P€uhler A (eds) Biological fixation of nitrogen for ecology and sustainable

agriculture. Springer, Heidelberg, Germany, pp 313–316

Sprent JI (2001) Nodulation in legumes. Royal Botanic Gardens, Kew, UK

Staehelin C, Schultze M, Kondorosi E, Mellor RB, Boller T, Kondorosi A (1994) Structural

modifications in Rhizobium meliloti Nod factors influence their stability against hydrolysis by

root chitinases. Plant J 5:319–330

Steenkamp ET, Stepkowski T, Przymusiak A, Botha WJ, Law IJ (2008) Cowpea and peanut in

southern Africa are nodulated by diverse Bradyrhizobium strains harboring nodulation genes

that belong to the large pantropical clade common in Africa. Mol Phylogenet Evol

48:1131–1144

Stepkowski T, Hughes CE, Law IJ, Markiewicz L, Gurda D, Chlebicka A, Moulin L (2007)

Diversification of lupine Bradyrhizobium strains: evidence from nodulation gene trees. Appl

Environ Microbiol 73:3254–3264

Sullivan JT, Ronson CW (1998) Evolution of rhizobia by acquisition of a 500-kb symbiosis island

that integrates into a phe-tRNA gene. Proc Natl Acad Sci USA 95:5145–5149

Sullivan JT, Patrick HN, Lowther WL, Scott DB, Ronson CW (1995) Nodulating strains of

Rhizobium loti arise through chromosomal symbiotic gene transfer in the environment. Proc

Natl Acad Sci USA 92:8985–8989

Sullivan JT, Trzebiatowski JR, Cruickshank RW, Gouzy J, Brown SD, Elliot RM, Fleetwood DJ,

McCallum NG, Rossbach U, Stuart GS, Weaver JE, Webby RJ, de Bruijn FJ, Ronson CW

(2002) Comparative sequence analysis of the symbiosis island of Mesorhizobium loti strainR7A. J Bacteriol 184:3086–3095


Suominen L, Roos C, Lortet G, Paulin L, Lindstroem K (2001) Identification and structure of the

Rhizobium galegae common nodulation genes: evidence for horizontal gene transfer. Mol Biol

Evol 18:907–916

Terefework Z, Lortet G, Suominenl LK (2000) Molecular evolution of interactions between

rhizobia and their legume hosts. In: Triplett E (ed) Prokaryotic nitrogen fixation: a model for

analysis of a biological process. Horizon Scientific Press, Norfolk, England, pp 187–206

Tian CF, Wang ET, Han TX, Sui XH, Chen WX (2007) Genetic diversity of rhizobia associated

with Vicia faba in three ecological regions of China. Arch Microbiol 188:273–282

Tian CF, Wang ET, Wu LJ, Han TX, Chen WF, Gu CT, Gu JG, Chen WX (2008) Rhizobium fabaesp. nov., a bacterium that nodulates Vicia faba. Int J Syst Evol Microbiol 58:2871–2875

Toledo I, Lloret L, Martınez-Romero E (2003) Sinorhizobium americanum sp. nov., a new

Sinorhizobium species modulating nativeAcacia spp. inMexico. Syst ApplMicrobiol 26:54–64

Valverde A, Velazquez E, Fernandez-Santos F, Vizcaino N, Rivas R, Mateos PF, Martinez-Molina

E, Igual JM, Willems A (2005) Phyllobacterium trifolii sp. nov., nodulating Trifolium and

Lupinus in Spanish soils. Int J Syst Evol Microbiol 55:1985–1989

Vazquez M, Davalos A, de las Penas A, Sanchez F, Quinto C (1991) Novel organization of the

common nodulaiton genes in Rhizobium leguminosarum bv. phaseoli strains. J Bacteriol

173:1250–1258

Vinuesa P, Leon-Barrios M, Silva C, Willems A, Jarabo-Lorenzo A, Perez-Galdona R, Werner D,

Martınez-Romero E (2005) Bradyrhizobium canariense sp. nov., an acid-tolerant endosymbi-

ont that nodulates endemic genistoid legumes (Papilionoideae: Genisteae) from the Canary

Islands, along with Bradyrhizobium japonicum bv. genistearum, Bradyrhizobium genospecies

alpha and Bradyrhizobium genospecies beta. Int J Syst Evol Microbiol 55:569–575

Wang ET, Martınez-Romero E (2000) Phylogeny of root- and stem-nodule bacteria associated

with legumes. In: Triplett E (ed) Prokaryotic nitrogen fixation: a model for analysis of a

biological process. Horizon Scientific Press, Norfolk, England, pp 177–186

Wang ET, Rogel MA, Garcıa-De los Santos A, Martınez-Romero J, Cevallos MA, Martınez-

Romero E (1999a) Rhizobium etli bv. mimosae, a novel biovar isolated from Mimosa affinis.Int J Syst Bacteriol 49:1479–1491

Wang ET, van Berkum P, Sui XH, Beyene D, Chen WX, Martinez-Romero E (1999b) Diversity of

rhizobia associated with Amorpha fruticosa isolated from Chinese soils and description of

Mesorhizobium amorphae sp. nov. Int J Syst Bacteriol 49:51–65Wernegreen JJ, Riley MA (1999) Comparison of the evolutionary dynamics of symbiotic and

housekeeping loci: a case for the genetic coherence of rhizobial lineages. Mol Biol Evol

16:98–113

Young JPW, Johnston AWB (1989) The evolution of specificity in the legume-Rhizobium symbi-

osis. Trends Ecol Evol 4:341–349

Young JM, Kuykendall LD, Martinez-Romero E, Kerr A, Sawada H (2001) A revision of

Rhizobium Frank 1889, with an emended description of the genus, and the inclusion of all

species of Agrobacterium Conn 1942 and Allorhizobium undicolade Lajudie et al. 1998 as newcombinations: Rhizobium radiobacter, R. rhizogenes, R. rubi, R. undicola and R. vitis. Int JSyst Evol Microbiol 51:89–103

Young JPW, Mutch LA, Ashford DA, Zeze A, Mutch KE (2003) The molecular evolution of host

specificity in the Rhizobium-legume symbiosis. In: Hails R, Godfray HJC, Beringer JE (eds)

Genes in the environment. Blackwell Science, Oxford, pp 245–257

Young JPW, Crossman LC, Johnston AWB, Thomson NR, Ghazoui ZF, Hull KH, Wexler M,

Curson ARJ, Todd JD, Poole PS, Mauchline TH, East AK, Quail MA, Churcher C, Arrowsmith

C, Cherevach I, Chillingworth T, Clarke K, Cronin A, Davis P, Fraser A, Za H, Hauser H,

Jagels K, Moule S, Mungall K, Norbertczak H, Rabbinowitsch E, Sanders M, Simmonds M,

Whitehead S, Parkhill J (2006) The genome of Rhizobium leguminosarum has recognizable

core and accessory components. Genome Biol 7:R34


Chapter 19

Convergent Evolution of Morphogenetic

Processes in Fungi

Sylvain Brun and Philippe Silar

Abstract Eumycetes fungi are a diverse group of organisms whose evolution is

characterized by frequent changes in nutritional strategy and the corresponding

developmental programs. The reasons for this versatility are unknown. We previ-

ously discovered that the NADPH oxidase Nox2 and the tetraspanin Pls1 are used in

two radically different cell types to achieve the same purpose: exiting from a

reinforced cell, suggesting that convergent evolution of morphogenetic processes

could account for the repetitive switches in trophic modes during fungal evolution.

However, we recently observed that saprobic fungi are also able to differentiate

appressorium-like structure closely resembling those of phytopathogenic species,

arguing that the ability to differentiate such cells is an ancient property of filamen-

tous fungi. Adaptation of parasitic and mutualistic fungi to plant may thus not solely

reside in their ability to penetrate their host.

19.1 Introduction

Fungi belonging to the Eumycetes (Opisthokonta) are a great success of evolution.

Their ancestors switched from phagotrophy, the original eukaryotic trophic mode,

to osmotrophy likely a billion years ago (McLaughlin et al. 2009). Since then they

have diversified into hundreds of thousands species and possibly much more

(Hawksworth 1991). They have invaded nearly all biotopes, from the deepest

depths of the oceans to the top of the highest mountains all around the globe.

They are even found in the arctic soils that remain frozen most of the years (Schadt

et al. 2003). Their total biomass is huge and they greatly impact on their environ-

ment. They live either in parasitic or in mutualistic symbiosis with other organisms,

S. Brun and P. Silar

UFR des Sciences du Vivant, Universite de Paris 7 – Denis Diderot, 75205 Paris Cedex 13, France

Institut de Genetique et Microbiologie, UMRCNRS – Universite de Paris 11, UPS Bat. 400, 91405

Orsay cedex, France



317

or as free living saprobes. The saprobes participate in the global carbon cycle,

especially they degrade highly recalcitrant materials that no other organism may

and regulate soil health by producing humic acids. As mutualistic symbionts, the

mycorhizal and endophytic fungi increase plant fitness and those present inside the

digestive tract enable many insects and mammalian herbivores to use the hard-to-

digest plant materials as food. Similarly, the mutualistic lichens are an important

component of many extreme biotopes. Parasitic fungi are known for nearly all

organisms (even fungi!), but they are especially important for plants and insects.

These have a tremendous impact on the dynamics of natural populations but also on

domesticated plants and animals. The feeding, dispersal, and “behavioral” diversity

of fungi is such that complete books are required to describe it (Webster 2007).

Because of their importance, scientific programs aimed at better understanding

the evolution and biology of fungi have been launched. The aftol (Assembling the

Fungal Tree of Life) used multigene trees to resolve their phylogeny (James et al.

2006) and proposed a new classification (Hibbett et al. 2007). Numerous genomic

programs have established sequences from a great diversity of fungi (see, for example,

http://genome.jgi-psf.org/, http://www.broadinstitute.org/science/projects/fungal-

genome-initiative/current-fgi-sequence-projects, http://www.genoscope.cns.fr/

spip/Fungi-sequenced-at-Genoscope.html). The data show that fungi are highly

diverse (McLaughlin et al. 2009). For example, the genetic diversity of fungi

belonging to related families or even to the same family may exceed that of

animals from different classes (Dujon 2005; Espagne et al. 2008).

19.2 The Versatility of Fungal Development

An important point that emerges from phylogenetic studies is the versatility with

which fungi may switch their trophic modes and “invent” repeatedly the same

structures (James et al. 2006). For instance, saprobic and symbiotic fungi may exist

within the same genus, and within the same class, saprotrophy, plant pathogeny,

lichen symbiosis, and other trophic modes may evolve. Similarly, plant pathogens

and mutualists invade their host plant by many means, one of which involves the

in-force breaking of the plant cuticule and/or cell wall. To do this, fungi differenti-

ate special cells called appressoria (Deising et al. 2000). These come in different

sizes and shapes and their origin may be quite different. For example, in Magna-porthe grisea, a hemi-biotrophic parasite of rice and barley, the appressorium

develops at the extremity of a dedicated hypha that is produced by a three-celled

spore issued from asexual reproduction. In this species, appressoria are heavily

melanized round cells with a very well-defined structure, from which the penetra-

tion peg emerges (Fig. 19.1). In Botrytis cinerea, appressorium-like structures are

also produced at the extremity of an hypha that originates from a spore issued from

asexual reproduction, but this spore has only one cell and the appressorium is no

more than a specialized hypha slightly reinforced at its tip, which is able to orient its

growth toward plant wall and to penetrate it, thanks to a penetration peg (Fig. 19.1).

318 S. Brun and P. Silar

http://genome.jgi-psf.org/

http://www.broadinstitute.org/science/projects/fungal-genome-initiative/current-fgi-sequence-projects

http://www.broadinstitute.org/science/projects/fungal-genome-initiative/current-fgi-sequence-projects

http://www.genoscope.cns.fr/spip/Fungi-sequenced-at-Genoscope.html

http://www.genoscope.cns.fr/spip/Fungi-sequenced-at-Genoscope.html

Fig. 19.1 Ontogeny of ascospores, appressorium, and appressorium-like structures. Sexual repro-

duction results in one-celled hyaline ascospores in B. cinerea, four-celled hyaline ascospores in

M. grisea, and two-celled melanised ascospores with a germ pore in P. anserina. In this latter

species, a cell death has occurred during ascospore differentiation. Appressorium is a roundish

heavily melanized structure in M. grisea, while it is no more than a reinforced hyphae in

B. cinerea. The similarity between P. anserina ascospore and M. grisea appressorium ontogenies

are highlighted by arrows

19 Convergent Evolution of Morphogenetic Processes in Fungi 319

These structures are thus qualified as “appressoria-like” rather than as true appres-

soria. M. grisea and B. cinerea belong to two different classes of ascomycetes,

the Sordariomycetes and Leotiomycetes, respectively. In these classes, numerous

species are known to live as saprobes, which seemingly do not differentiate

appressoria as they do not need to penetrate host plants. Thus, the question raised

is whether the utilization of appressoria to penetrate plants is the result of conver-

gent evolution by plant pathogens or whether it reflects an ancient ability of fungi to

differentiate penetration structures that would have been lost in saprobes.

Spore is another fungal structure (along with the fruiting body) that exhibits

many convergent evolutions. Spores are issued either from sexual (basidiospores,

ascospores. . .) or from asexual (conidia. . .) reproduction and constitute an impor-

tant part of the life cycle, since they enable fungi to disperse efficiently and to resist

to adverse conditions. They come in many shapes, sizes, and colors and have been

used in the past to classify fungi. For example, Podospora anserina, a model

ascomycete produces heavily melanized ascospores that germinate in a regulated

manner through a germ pore (Fig. 19.1). These are in fact constituted of two cells,

one of which has undergone a cell death. Neurospora crassa produces one-celled

striated ascospores with two germ pores located at the opposite poles, while

M. grisea ascospores are composed of four hyaline cells and lack a germ pore

(Fig. 19.1). Those of B. cinerea are composed of a single hyaline cell (Fig. 19.1).

Yet, spore evolution appears filled with convergence. For example, in some Sor-

dariomycetes, the fruiting body wall is a better descriptor of evolution than asco-

spore shape (Miller and Huhndorf 2005). Similarly, a germ pore is present in some

species for both basidiomycetes and ascomycetes and is absent in others.

The molecular basis for the versatility of fungi in switching trophic modes and

developments is unknown. The only documented instance is for a change from

mycoparasitism to saprotrophy in the genus Trichoderma. Indeed, there is evidencefor a horizontal transfer of a cluster of genes involved in nitrate assimilation from a

basidiomycete related to Ustilago maydis to the ascomycete Trichoderma reesei,whereas the other members of the Trichoderma genus appear to lack the cluster. Thishas been correlated with the fact that T. reesei is the only Trichoderma living as a

saprobe in woody materials, while the other members of the genus are mycoparasites

(Slot and Hibbett 2007). The nitrate assimilation cluster would enable T. reesei toefficiently scavenge nitrogen in wood, while the other Trichodermas must obtain it

from their host, accounting for the trophic change. Trichoderma may parasitize

basidiomycetes, favoring perhaps the gene transfer in the ancestors of T. reesei.

19.3 Are Appressoria and Appressorium-Like Structures

the Result of Convergent Evolution?

We discovered serendipitously a possible convergent evolution of morphogenetic

processes impacting on trophic strategy in filamentous fungi by studying the role

of the Pls1 tetraspanin and the Nox2 NADPH oxidase (Nox) in the saprobic fungus


P. anserina. Tetraspanin are membrane-bound proteins, whose roles are not yet

completely clear (Veneault-Fourrey et al. 2006b). In fungi, tetraspanin of the

Pls1 family have been at first unraveled as virulence factors in three different

plant pathogenic species. In M. grisea, B. cinerea, and Colleototrichum linde-muthianum, the Pls1 mutants are blocked at the penetration step; the appressorium

appears normal but penetration pegs are not produced (Clergeot et al. 2001;

Gourgues et al. 2004; Veneault-Fourrey et al. 2005). This was taken as the indica-

tion for a specific role of Pls1 tetraspanin in phytopathogenic fungi. Yet, ortholo-

gues of Pls1 are present in saprobic fungi, including P. anserina (Lambou et al.

2008). Tetraspanins share the same membrane localization as Nox. Nox are mem-

brane-bound enzymes that generate superoxide ions in exchange of consumption of

NADPH. Several years ago, we proposed that the ancient role of Nox (and of the

ROS they produce) was the sensing of the environment and cell-to-cell communi-

cation (Lalucque and Silar 2003). And indeed, these enzymes have now been

shown to play key roles in development, pathogeny, symbiosis, and defense in a

broad range of Eukaryotes (Lara-Ortiz et al. 2003; Malagnac et al. 2004; Aguirre

et al. 2005; Silar 2005; Takemoto et al. 2007). There is presently three Nox

isoforms known in fungi (see Table 19.1 for an update on Nox genes in fungal

genomes) and all data argue that they do not fulfill redundant roles (Takemoto et al.

2007). In particular, in two saprobic fungi, P. anserina and N. crassa, the Nox2

isoform seems to be more specifically dedicated to regulate melanized ascospore

germination (Malagnac et al. 2004; Cano-Dominguez et al. 2008). Indeed,

both fungi produce melanized ascospores and, in both species, Nox2 mutant

ascospores do not germinate. Furthermore, when P. anserina ascospore melanin

is removed, the Nox mutant ascospores germinate efficiently but in a nonregulated

manner (Malagnac et al. 2004). Accordingly, Nox2 appears dispensable for

the germination of B. cinerea ascospores, which are not melanized (Segmuller

et al. 2008).

When we deleted the PaPls1 gene of P. anserina, we discovered that the

DPaPls1 mutants had the same ascospore germination defects as the PaNox2mutants (Lambou et al. 2008). Again, removal of melanin in PaPls1 mutant

ascospores suppressed the germination default, leading to unregulated germination.

Interestingly, the Nox2 isoforms are necessary for plant penetration in M. griseaand B. cinerea (Egan et al. 2007; Segmuller et al. 2008). Additionally, Pls1 is

dispensable for the germination of theM. grisea nonmelanized ascospores (Lambou

et al. 2008). These data suggest that Pls1 and Nox2 may act together. This finding is

supported by the fact that both proteins are either present or absent in fungal

genomes (Table 19.1, Fig. 19.2). In lower fungi, the coevolution is not clear.

However, Pls1 tetraspanins are small proteins that evolve rapidly, impairing their

detection in very divergent genomes by using ordinary tools. In the “higher fungi”,

i.e., Ascomycetes and Basidiomycetes, the repartition of Pls1 and Nox2 is best

accounted for by at least nine independent losses of both genes during evolution

(Fig. 19.2). As the Pls1 and Nox2 genes are not linked in the genomes, these data

provide a strong argument for their acting in the same processes (Loganantharaj and

Atwi 2007). Both proteins may act together in a complex located at the plasma


Table 19.1 Occurrence of Nox1, Nox2, Nox3, and Pls1 in Eumycota

Fungal species Nox1/

NoxA

Nox2/

NoxB

Nox3/

NoxC

Pls1

AscomycotaPezizomycotina

Sordariomycetes Podospora anserina 1 1 1 1

Sporotrichum thermophile 1 1 0 1

Thielavia terrestris 1 1 0 1

Chaetomium globosum 1 1 0 1

Neurospora tetrasperma 1 1 0 1

Neurospora discreta 1 1 0 1

Neurospora crassa 1 1 0 1

Magnaporthe grisea 1 1 1 1

Cryphonectria parasitica 1 1 0 1

Grosmannia clavigera 1 1 0 1

Fusarium graminearum 1 1 1 1

Fusarium verticillioides 1 1 1 1

Fusarium oxysporum 1 1b 1 1

Haematonectria (Nectria)haematococca

2 1 1 1

Epichloe festucae 1 1 0 1

Trichoderma atroviride 1 1 0 1

Trichoderma reesei 1 1 0 1

Trichoderma virens 1 1 0 1

Verticillium dahliae 1 1 1 1

Verticillium albo-atrum 2 1 1 1

Colletotrichum graminicola 1 1 1 1

Leotiomycetes Sclerotinia sclerotiorum 1 1 0 1

Botrytis cinerea 1 1 0 1

Blumeria graminis 1b 1 0 1

Eurotiomycetes Aspergillus oryzae 1 0 0 0

Aspergillus flavus 1 + 1b 0 0 0

Aspergillus terreus 1 0 1 0

Aspergillus carbonarius 1 0 0 0

Aspergillus niger 1 0 0 0

Aspergillus fumigatus 1 0 0 0

Neosartorya fischeri 1 0 0 0

Aspergillus clavatus 1 0 0 0

Aspergillus nidulans 1 0 0 0

Penicillium chrysogenum 1 0 0 0

Talaromyces stipitatus 1 1 0 1

Penicillium marneffei 1 1 0 1

Histoplasma capsulatum 1 1 0 1

Paracoccidioides brasiliensis 1 1 0 1

Blastomyces dermatitidis 1 1 0 1

Uncinocarpus reesii 1 1 0 1

Coccidioides posadasii 1 1 0 1

Coccidioides immitis 1 1 0 1

Arthroderma gypseum 1 1 0 1

Microsporum canis 1 1 0 1

Trichophyton tonsurans 1 1 0 1

Trichophyton rubrum 1 1 0 1

Trichophyton equinum 1 1 0 1

Ascosphaera apis 0 0 0 0

Dothideomycetes Mycosphaerella graminicolla 1 0 1 0

Mycosphaerella fijiensis 1 0 1 0

Cochliobolus heterostrophus 1 1 1 1

Alternaria brassicola 1 1 0 1

(continued)


membrane and despite varying fungal habitat and/or physiological diversity, the

function of this complex might have been conserved in the different lineages.

The second striking conclusion is that melanized ascospore germination requires

the same proteins as the formation of the penetration peg from appressoria. When

compared (Fig. 19.1), these two processes appear noticeably similar in P. anserina

Table 19.1 (continued)

Fungal species Nox1/

NoxA

Nox2/

NoxB

Nox3/

NoxC

Pls1

Pyrenophora tritici 1 1 1 1

Stagonospora nodorum 1b 1 1 1

Saccharomycotina Saccharomyces cerevisiae 0 0 0 0

Candida glabrata 0 0 0 0

Zygosaccharomyces rouxii 0 0 0 0

Saccharomyces kluyveri 0 0 0 0

Kluyveromyces thermotolerans 0 0 0 0

Kluyveromyces lactis 0 0 0 0

Ashbya gossypii 0 0 0 0

Candida albicans 0 0 0 0

Debaryomyces hansenii 0 0 0 0

Yarrowia lipolytica 0 0 0 0

Taphrinomycotina Schizosaccharomyces japonicus 0 0 0 0

Schizosaccharomyces pombe 0 0 0 0

Schizosaccharomyces octosporus 0 0 0 0

Pneumocystis carinii 0 0 0 0

BasidiomycotaUstilaginomycotina Ustilago maydis 0 0 0 0

Malassezia globosa 0 0 0 0

Agaricomycotina

Agaricomycetes Heterobasidion annosum 1 1 0 1

Schizophyllum commune 1 1 0 1

Coprinopsis cinerea 1 1 0 1

Laccaria bicolor 1 1 0 1

Postia placentaa 1 1 0 1

Pleurotus ostreatus 1 1 0 1

Phanerochaete chrysosporium 1 1 0 1

Tremellomycetes Cryptococcus neoformans 0 0 0 0

Tremella mesenterica 1 0 0 0

Pucciniomycotina Sporobolomyces roseus 1 0 0 0

Melampsora larici-populina 3 2 0 1

Puccinia graminis 1 1 0 1

“Lower fungi”

Mucoromycotina Rhizopus oryzae 0 0 0 1?

Mucor circinelloides 0 0 0 1?

Phycomyces blakesleeanus 0 0 0 1?

Microsporidia Encephalitozoon cuniculi 0 0 0 0

Antonospora locustae 0 0 0 0

Nosema ceranea 0 0 0 0

Blastocladiomycetes Allomyces macrogynus 1(1) 1 (4) 0 ?

Chytridiomycetes Spizellomyces punctatus 1 1 0 ?

Batrachochytrium dendrobatidis 1 1 0 ?aBLAST analysis detects two very similar copies for this species. However, the P. placenta projectsequenced the genome of a dikaryon (http://genome.jgi-psf.org/Pospl1/Pospl1.home.html). The

two copies are likely the different alleles present in each haploid genomebGenome sequence with an incomplete or erroneous gene sequence. Pseudogenes are in parenthesis


http://genome.jgi-psf.org/Pospl1/Pospl1.home.html

andM. grisea. Indeed, during the ontogeny of appressoria and ascospores, there is aprogrammed cell death event (Beckett et al. 1968; Veneault-Fourrey et al. 2006a).

When the structures are formed they are both heavily melanized and both contain a

pore from which a peg is produced (Beckett et al. 1968; Deising et al. 2000).

We thus speculated that the same program was used by the two species to achieve

the same mean (exiting from a melanized structure). This provides a nice example

of the reutilisation of the same proteins to achieve a similar morphogenetic goal in

two different cell types. We also speculated that this process could be recruited

repeatedly during evolution to achieve the same mean, i.e., penetrate plants. If so,

appressoria from different fungi would be due to convergent evolution. However,

we recently obtained data that call off this statement. Indeed, we recently discov-

ered that Nox2 and Pls1 are involved in a novel developmental stage in P. anserina:the development of appressorium-like cells involved in plant material penetration

(Brun et al. 2009).

P.c

R.o

Blastocladiomycota

Mucoromycotina

Basidiomycota

Agaricomycotina

Ustilaginomycotina

Agaricomycetes

Tremellomycetes

PucciniomycotinaPucciniomycetes

Microbotryomycetes

Saccharomycotina

Ascomycota

Taphrinomycotina

Sordariomycetes

Leotiomycetes

Eurotiomycetes

Eurotiales

Dothideomycetes

Capnodiales

Pleosporales

Pezizomycotina

Sordariales

Magnaporthales

Diaporthales

Ophiostomatales

Hypocreales

Microsporidia

Chrytridiomycota

Ascosphaera

Onygenales

**

app

ress

ori

um

-lik

e st

ruct

ure

s

?

Lower Fungi

Fig. 19.2 Phylogenetic tree of Eumycetes. The tree shows the fungal groups for which complete

genome sequences are available. The nine vertical arrows locate the loss of Pls1 and Nox2.

Asterisks (*) indicate the two groups for which the Pls1 and Nox2 proteins have been recruited

for the same goal (exiting a reinforced structure) in two cell types: the ascospores in Sordariales

(P. anserina and N. crassa) and the appressorium in Magnaporthales (M. grisea). Possible

appearance of appressorium-like structures occurred very early during fungal evolution, however,

at a yet undefined moment. Fungi unable to differentiate appressorium-like structure are indicated

by P.c (Penicillium chrysogenum) and R.o (Rhizopus oryzae)


19.4 Differentiating Appressorium-Like Structures Could

Be an Ancient Property of Fungi

During our studies on Nox2 and Pls1, we noticed that in addition to their ascospore

germination default, the null mutants of both genes presented a defect in the

production of fruiting bodies, specifically when grown on cellulose as sole carbon

source (Malagnac et al. 2008). This prompted us to investigate in more details the

cellulose degradation process in P. anserina (Brun et al. 2009). When cellophane is

provided as food source, P. anserina is able to orient its growth toward the cello-

phane layer. Upon contacting cellophane, it differentiates a structure that greatly

resembles B. cinerea pseudo-appressorium. Even more striking is the similarity

between the appressorium-like phenotypes of B. cinerea and P. anserina Pls1 and

Nox2mutants (Segmuller et al. 2008; Brun et al. 2009). In both species, thesemutants

are impaired at the reorientation step toward the substrate (onion skin and cellophane,

respectively), which is a prerequisite for penetration. In both species, mutant hyphae

tend to “hesitate” in the direction to grow. Then, they establish loose contacts with

the substrate and finally are completely defective in penetrating it. Nonetheless, the

setting up of fully functional penetration structures is not only under the control of

Nox2 and Pls1 but also require the Nox1 isoform (Egan et al. 2007; Giesbert et al.

2008; Brun et al. 2009). In the view of this new finding, we speculate that the ability

to differentiate cellular structure dedicated to penetrate plant materials might be an

ancient property of filamentous fungi (at least ascomycetes and basidiomycetes),

which is used in saprobes to efficiently degrade dead plants, andmore aggressively in

phytopathogens to penetrate their hosts. To test this possibility, we have evaluated

the ability of several additional fungi to differentiate penetration structures on

cellophane (see Fig. 19.3 for an example). A variety of structures permitting to

breach the cellophane were indeed produced by a wide spectrum of fungi (several

Sordariomycetes and Agaricomycetes; S. Brun and P. Silar, unpublished data).

Presently, we did not detect such structures in two species, Penicillium chrysogenumand Rhizopus oryzae (Fig. 19.3). Significantly, both fungi lack Nox2 and Pls1

(Table 19.1, Fig. 19.2), confirming the crucial role of the two proteins in the

differentiation of appressorium-like cells. Therefore, a wide range of fungi seem to

possess the toolkit necessary to breach the plant cell wall. The patchy phylogenetic

repartition of species known to produce appressoria and related structure could thus

be due to biased sampling toward parasitic and mutualist plant symbionts in studies

dealing with appressorium formation. However, some species may truly be unable to

differentiate these structures: those that have lost Pls1 and Nox2.

In other words, there is no need to invoke complex convergent evolution of

fungal structures to explain the recurrent change in trophic lifestyle. Evidence

is arising which confirms a role of ROS and Nox in polarized hyphal growth

(Semighini and Harris 2008) and we believe that the ability of fungi to attack and

penetrate plant materials may simply rely on sensing the glucose gradient created by

the enzymatic degradation of the polysaccharides composing the plant cell wall, i.e.,

cellulose and hemicellulose (Brun et al. 2009). More generally, we believe that if


this simple model is true, penetration structures under the control of Nox2/Pls1

should be found not only for phytopathogens and saprobes, but also for entomo-

pathogens (for cuticle breaching) as well as for fungal parasites such as Trichodermasp. (for chitin-based cell walls penetration) and possibly for human pathogens. We

thus now need to confirm on a larger sample if the correlation between the ability to

build these structures and the conservation of Nox2/Pls1 holds true.

Acknowledgments This work was supported by ANR grant n�ANR-05-Blan-0385-02.

Fig. 19.3 Cellophane breach. Four days old mycelia of P. anserina (P. a), Trichoderma species(T. sp), Penicillium chrysogenum (P. c), and Rhizopus oryzae (R. o) were observed as described

(Brun et al. 2009). Numbers indicate the distance from the first picture in mm as depicted by the

arrows on the schemes on the right. In the first column, mycelia of all the strains are growing

horizontally on the cellophane layer. In the second column, mycelia of P. anserina and T. speciesreorient their growth toward the cellophane and establish bulging contacts (some examples are

indicated by arrows). In P. chrysogenum and R. oryzae, there is no reorientation toward cellophane,though rare contact may occur. In the third column, needle-like hyphae (some examples are indicated

by asterisk) are emitted in P. anserina and T. species, which allow both fungi to penetrate into the

cellophane layer. In contrast, P. chrysogenum and R. oryzae cannot penetrate cellophane. In the

fourth column, schematic representation of the structures; the arrows points toward the approximate

focal plan of the first three columns and the eye indicates the direction of the observation


References

Aguirre J, Rios-Momberg M, Hewitt D, Hansberg W (2005) Reactive oxygen species and

development in microbial eukaryotes. Trends Microbiol 13:111–118

Beckett A, Barton R, Wilson IM (1968) Fine structure of the wall and appendage formation in

ascospores of Podospora anserina. J Gen Microbiol 53:89–94

Brun S, Malagnac F, Bidard F, Lalucque H, Silar P (2009) Functions and regulation of the Nox

family in the filamentous fungus Podospora anserina: a new role in cellulose degradation. Mol


Cano-Dominguez N, Alvarez-Delfin K, Hansberg W, Aguirre J (2008) NADPH oxidases NOX-1

and NOX-2 require the regulatory subunit NOR-1 to control cell differentiation and growth in

Neurospora crassa. Eukaryot Cell 7:1352–1361Clergeot PH, Gourgues M, Cots J, Laurans F, Latorse MP, Pepin R, Tharreau D, Notteghem JL,

LebrunMH (2001) PLS1, a gene encoding a tetraspanin-like protein, is required for penetration

of rice leaf by the fungal pathogen Magnaporthe grisea. Proc Natl Acad Sci USA

98:6963–6968

Deising HB, Werner S, Wernitz M (2000) The role of fungal appressoria in plant infection.

Microbes Infect 2:1631–1641

Dujon B (2005) Hemiascomycetous yeasts at the forefront of comparative genomics. Curr Opin

Genet Dev 15:614–620

Egan MJ, Wang ZY, Jones MA, Smirnoff N, Talbot NJ (2007) Generation of reactive oxygen

species by fungal NADPH oxidases is required for rice blast disease. Proc Natl Acad Sci USA

104:11772–11777

Espagne E, Lespinet O, Malagnac F, Da Silva C, Jaillon O, Porcel BM, Couloux A, Aury JM,

Segurens B, Poulain J, Anthouard V, Grossetete S, Khalili H, Coppin E, Dequard-Chablat M,

PicardM, Contamine V, Arnaise S, Bourdais A, Berteaux-Lecellier V, Gautheret D, de Vries RP,

Battaglia E, Coutinho PM, Danchin EG, Henrissat B, Khoury RE, Sainsard-Chanet A, Boivin A,

Pinan-Lucarre B, Sellem CH, Debuchy R, Wincker P, Weissenbach J, Silar P (2008) The

genome sequence of the model ascomycete fungus Podospora anserina. Genome Biol 9:R77

Giesbert S, Schurg T, Scheele S, Tudzynski P (2008) The NADPH oxidase Cpnox1 is required for

full pathogenicity of the ergot fungus Claviceps purpurea. Mol Plant Pathol 9:317–327

Gourgues M, Brunet-Simon A, Lebrun MH, Levis C (2004) The tetraspanin BcPls1 is required for

appressorium-mediated penetration of Botrytis cinerea into host plant leaves. Mol Microbiol

51:619–629

Hawksworth DL (1991) The fungal dimension of biodiversity: magnitude, significance, and

conservation. Mycol Res 95:641–655

Hibbett DS, Binder M, Bischoff JF, Blackwell M, Cannon PF, Eriksson OE, Huhndorf S, James T,

Kirk PM, Lucking R, Thorsten Lumbsch H, Lutzoni F, Matheny PB, McLaughlin DJ,

Powell MJ, Redhead S, Schoch CL, Spatafora JW, Stalpers JA, Vilgalys R, Aime MC,

Aptroot A, Bauer R, Begerow D, Benny GL, Castlebury LA, Crous PW, Dai YC, Gams W,

Geiser DM, Griffith GW, Gueidan C, Hawksworth DL, Hestmark G, Hosaka K, Humber RA,

Hyde KD, Ironside JE, Koljalg U, Kurtzman CP, Larsson KH, Lichtwardt R, Longcore J,

Miadlikowska J, Miller A, Moncalvo JM, Mozley-Standridge S, Oberwinkler F, Parmasto E,

Reeb V, Rogers JD, Roux C, Ryvarden L, Sampaio JP, Schussler A, Sugiyama J, Thorn RG,

Tibell L, Untereiner WA,Walker C, Wang Z, Weir A, Weiss M,White MM,Winka K, Yao YJ,

Zhang N (2007) A higher-level phylogenetic classification of the fungi. Mycol Res

111:509–547

James TY, Kauff F, Schoch CL, Matheny PB, Hofstetter V, Cox CJ, Celio G, Gueidan C, Fraker E,

Miadlikowska J, Lumbsch HT, Rauhut A, Reeb V, Arnold AE, Amtoft A, Stajich JE,

Hosaka K, Sung GH, Johnson D, O’Rourke B, Crockett M, Binder M, Curtis JM, Slot JC,

Wang Z, Wilson AW, Schussler A, Longcore JE, O’Donnell K, Mozley-Standridge S,

Porter D, Letcher PM, Powell MJ, Taylor JW, White MM, Griffith GW, Davies DR,


Humber RA, Morton JB, Sugiyama J, Rossman AY, Rogers JD, Pfister DH, Hewitt D,

Hansen K, Hambleton S, Shoemaker RA, Kohlmeyer J, Volkmann-Kohlmeyer B, Spotts RA,

Serdani M, Crous PW, Hughes KW, Matsuura K, Langer E, Langer G, Untereiner WA,

Lucking R, Budel B, Geiser DM, Aptroot A, Diederich P, Schmitt I, Schultz M, Yahr R,

Hibbett DS, Lutzoni F, McLaughlin DJ, Spatafora JW, Vilgalys R (2006) Reconstructing the

early evolution of fungi using a six-gene phylogeny. Nature 443:818–822

Lalucque H, Silar P (2003) NADPH oxidase: an enzyme for multicellularity? Trends Microbiol

11:9–12

Lambou K, Malagnac F, Barbisan C, Tharreau D, Lebrun MH, Silar P (2008) A crucial role for the

Pls1 tetraspanin during ascospore germination of the saprophytic fungus Podospora anserina.Eukaryot Cell 7:1809–1818

Lara-Ortiz T, Riveros-Rosas H, Aguirre J (2003) Reactive oxygen species generated by microbial

NADPH oxidase NoxA regulate sexual development in Aspergillus nidulans. Mol Microbiol

50:1241–1255

Loganantharaj R, Atwi M (2007) Towards validating the hypothesis of phylogenetic profiling.

BMC Bioinformatics 8(Suppl 7):S25

Malagnac F, Bidard F, Lalucque H, Brun S, Lambou K, Lebrun MH, Silar P (2008) Convergent

evolution of morphogenetic processes in fungi: role of tetraspanins and NADPH oxidases 2 in

plant pathogens and saprobes. Commun Integr Biol 1:180–181

Malagnac F, Lalucque H, Lepere G, Silar P (2004) Two NADPH oxidase isoforms are required for

sexual reproduction and ascospore germination in the filamentous fungus Podospora anserina.Fungal Genet Biol 41:982–997

McLaughlin DJ, Hibbett DS, Lutzoni F, Spatafora JW, Vilgalys R (2009) The search for the fungal

tree of life. Trends Microbiol 17:488–497

Miller AN, Huhndorf SM (2005) Multi-gene phylogenies indicate ascomal wall morphology is a

better predictor of phylogenetic relationships than ascospore morphology in the Sordariales

(Ascomycota, Fungi). Mol Phylogenet Evol 35:60–75

Schadt CW, Martin AP, Lipson DA, Schmidt SK (2003) Seasonal dynamics of previously

unknown fungal lineages in tundra soils. Science 301:1359–1361

Segmuller N, Kokkelink L, Giesbert S, Odinius D, van Kan J, Tudzynski P (2008) NADPH

oxidases are involved in differentiation and pathogenicity in Botrytis cinerea. Mol Plant

Microbe Interact 21:808–819

Semighini CP, Harris SD (2008) Regulation of apical dominance in Aspergillus nidulans hyphaeby reactive oxygen species. Genetics 179:1919–1932

Silar P (2005) Peroxide accumulation and cell death in filamentous fungi induced by contact with a

contestant. Mycol Res 109:137–149

Slot JC, Hibbett DS (2007) Horizontal transfer of a nitrate assimilation gene cluster and ecological

transitions in fungi: a phylogenetic study. PLoS ONE 2:e1097

Takemoto D, Tanaka A, Scott B (2007) NADPH oxidases in fungi: diverse roles of reactive

oxygen species in fungal cellular differentiation. Fungal Genet Biol 44:1065–1076

Veneault-Fourrey C, Barooah M, Egan M, Wakley G, Talbot NJ (2006a) Autophagic fungal cell

death is necessary for infection by the rice blast fungus. Science 312:580–583

Veneault-Fourrey C, Lambou K, Lebrun MH (2006b) Fungal Pls1 tetraspanins as key factors of

penetration into host plants: a role in re-establishing polarized growth in the appressorium?

FEMS Microbiol Lett 256:179–184

Veneault-Fourrey C, Parisot D, Gourgues M, Lauge R, Lebrun MH, Langin T (2005) The

tetraspanin gene ClPLS1 is essential for appressorium-mediated penetration of the fungal

pathogen Colletotrichum lindemuthianum. Fungal Genet Biol 42:306–318Webster J (2007) Introduction to fungi, 3rd edn. Cambridge University Press, U.K


Chapter 20

Evolution and Historical Biogeography

of a Song Sparrow Ring in Western

North America

Michael A. Patten

Abstract The Song Sparrow, Melospiza melodia (Aves: Emberizidae), exhibits a

greater degree of geographic variation than does any other North American bird

species. Detailed morphological work has demonstrated that a subset of the 25

diagnosable subspecies forms a classic ring species in the western United States.

The ring’s center is the Sierra Nevada and Mojave Desert in California and adjacent

Nevada, and its connecting point is in southeastern California, where an olive and

black subspecies of the coastal slope interbreeds sporadically with a gray and rufous

subspecies of the arid interior. However, song differences associated with habitat

segregation lead to assortative mating between the two subspecies that meet in the

Coachella Valley at the southern base of San Gorgonio Pass. Moving clockwise

around the ring from the connecting point one finds a gradation of subspecies that

become paler, rustier, and grayer. Standard models of ring species evolution imply

the connecting point is the region occupied most recently, in this case after sparrows

would have spread southward down either side of the mountains and desert. This

scenario is plausible given molecular evidence of a glacial refugium on the Queen

Charlotte Islands, British Columbia, suggesting that ancestral birds could have

moved south in this pattern. By contrast, another postulated refugium is what is

now the arid desert of southeastern California or northeastern Baja California,

Mexico. This refugium’s location – coupled with a recent meta-analysis of North

American hybrid zones that identifies the San Gorgonio Pass region as an ancestral

contact zone of coastal and desert fauna – implies that the connecting point is the

region occupied earliest, an alternative that would mean the Song Sparrow ring differs

fundamentally from one that would have evolved via the standard model. Bio-

geographical and morphological data support the latter, more radical interpretation,

M.A. Patten

Oklahoma Biological Survey and Department of Zoology, University of Oklahoma,

111 E. Chesapeake Street, Norman, OK 73019, USA

e‐mail: [email protected]


329

but genetic, vocal, ecological, and behavioral data are needed around the ring to

determine conclusively which model is best supported.

20.1 Ring Species as a Biogeographic Pattern

A concrete bridge between microevolution and macroevolution, including specia-

tion, continues to elude evolutionary biologists (Mayr 1982; Jablonski 2000;

Reznick and Ricklefs 2009). Some researchers have concluded that macroevolution

is no more than the accumulated effects of microevolution (Hansen and Martins

1996; Simons 2002), whereas others have concluded that macroevolution requires a

fundamentally different mechanism (Stanley 1998; Erwin 2000). Ring species may

prove to be that crucial bridge (Irwin et al. 2001b).

A ring species consists of multiple subspecies whose contiguous geographic

ranges encircle a geographic barrier and whose terminal subspecies behave as good

biological species where their ranges meet (Cain 1954; Irwin and Irwin 2002; Coyne

and Orr 2004). Subspecies around the ring that connect the terminal subspecies grade

into each other to form a continuous set of intermediate forms. Because reproductive

isolation evolves in the face of gene flow, Mayr (1942:180) referred to ring species as

“the perfect demonstration of speciation”, and Cain (1954:141) referred to them as

“the clearest evidence of geographical speciation”. But as Coyne and Orr (2004:102)

noted, ring species do not demonstrate geographical (¼allopatric) speciation but

rather speciation that occurs “through the attenuation of gene flow with distance”.

Thus, ring species remain a key to understanding the evolution of reproductive

isolation and, therefore, of speciation, and they demonstrate how “small changes

can lead to species-level differences” (Irwin et al. 2001b).

Lost or conflated in this argument about whether ring species are examples of

geographic speciation is a clear distinction between pattern and process. To fit the

pattern of a ring species, three conditions must be met (Irwin and Irwin 2002;

Joseph et al. 2008; Patten and Pruett 2009): (1) geographic ranges of neighboring

subspecies must meet, (2) phenotype and genotype of neighbors must exhibit the

effects of intergradation, except for (3) the two subspecies that form the terminal

points, which must exhibit a sharp break in phenotype, genotype, ecology, and

behavior, enough so that these subspecies behave as good biological species where

their ranges meet. Few proposed ring species meet these criteria (Irwin et al. 2001b;

Coyne and Orr 2004), and even a weaker criterion, replacing (1) and (2) above, of

“a series of progressively intermediate forms must be arranged in a ring” (Patten

and Pruett 2009) still excludes many of the proposed ring species. Regardless, if a

geographically variable species was found to fit the above criteria, it would be fair

to dub it a ring species, immaterial of how the pattern came to be. It also seems fair

to conclude that the pattern of phenotypic variation exhibited by a ring species

demonstrates that the microevolutionary processes that lead to population differen-

tiation are akin to the processes that lead to speciation, whatever differences there

are being only a matter of degree (Irwin et al. 2001b).

330 M.A. Patten

20.2 The Evidence for Ring Species

Whether any claimed ring species fits all three criteria outlined above is debatable

or unlikely (Coyne and Orr 2004; Martens and P€ackert 2007; Joseph et al. 2008). Forexample, Irwin et al. (2001b) and Irwin and Irwin (2002) reviewed 23 ring species

reported in the scientific literature. Almost all were found wanting in some way,

often because reproductive isolation of the terminal points had not been studied but

sometimes because gene flow around the ring was unlikely or was known not to

occur. In the case of the tsetse fly, Glossina morsitans, the terminal points did not

meet in sympatry. Even the two most widely studied examples of putative ring

species, the salamander Ensatina eschscholtzii (Stebbins 1957; Wake and Yanev

1986; Wake 2006; Kuchta et al. 2009) and the warbler Phylloscopus trochiloides(Mayr 1942; Irwin et al. 2001a, 2005), do not meet criteria fully (Coyne and Orr

2004; Martens and P€ackert 2007), although they nonetheless display enough char-

acteristics to be considered ring species by most evolutionary biologists.

Just over half of the examples of ring species Irwin et al. (2001b) considered

pertained to bird species, although they did not consider Mayr’s (1942) examples of

the Zosterops white-eyes in the Lesser Sunda Islands nor the Pernis honeyeaters inthe Philippines, to say nothing of Stejneger’s (in Jordan 1905) speculation regard-

ing Lanius shrikes around the Baltic Sea. Perhaps, there are no additional pertinent

data on these systems. To these examples can be added two avian ring species

described recently: the Willow Warbler (Phylloscopus trochilus) complex encir-

cling the Baltic Sea (Bensch et al. 2009) and subspecies of the Song Sparrow

(Melospiza melodia) encircling the Sierra Nevada and Mojave Desert of the south-

western United States (Patten and Pruett 2009). The Willow Warbler varies in

plumage color, body size, AFLPs (amplified fragment length polymorphism),

microsatellite markers, and migratory behavior to the extent that it “shares many

features with the classic examples of ring species”, albeit one that evolved recently

relative to nearly all other examples (Bensch et al. 2009).

The Song Sparrow varies considerably in plumage color and pattern around the

ring (Table 20.1), with phenotypically intermediate populations present in all

contact zones, implying gene flow and intergradation where ranges meet

(Fig. 20.1; Patten and Pruett 2009). The terminal points are two subspecies – the

pale, rufescent M. m. fallax of the desert Southwest and the dark, olivaceous

M. m. heermanni of southern and central California – that meet in the Coachella

Valley, which lies between San Gorgonio Pass and the Salton Sea. The terminal

taxa hybridize only rarely; instead, there is evidence that females choose mates

assortatively, males respond more strongly to their own subspecies’ songs, and song

structure is shaped by habitat structure, which differs between the subspecies

(Patten et al. 2004b). Although genetic variation has not yet been studied around

the ring, the terminal taxa differ in frequency of microsatellite markers and these

differences are associated with plumage differences (Patten et al. 2004b). More-

over, a recent study of Song Sparrows along the whole of the Pacific Coast, from the

western Aleutian Islands of Alaska to southernmost California, found, in many

20 Evolution and Historical Biogeography of a Song Sparrow Ring 331

Table 20.1 Patterns of phenotypic variation around the Song Sparrow Melospiza melodia ring in

western North America

heermanni gouldii cleonensis montana fallax

Mantle color Grayish Reddish Dark reddish Grayish Brownish

olive-brown olive-brown brown brown gray

Mantle fringe Gray, thin Absent Gray, thin Gray, broad Reddish gray, broad

Underparts White White Grayish White White

Streak color Fuscous Black Dark brown Brown Warm brown

Streak fringe Ruddy Olive Brown Chestnut Chestnut

Malar Reddish fuscous Blackish Fuscous Chestnut brown Chestnut

Supercilia Ashy Ashy Grayish Whitish Whitish

Fig. 20.1 The Song Sparrow (Melospiza melodia) ring in western North America (from Patten and

Pruett 2009). The northwestern portion of center of the ring is the Sierra Nevada, the tallest

mountain range in the conterminous United States. The remainder of the gap is the Mojave Desert

(southern California) and southern Great Basin desert (southern Nevada). The large lake in

southeastern California is the Salton Sea, which sits at the southern edge of where the terminal

taxa meet, and San Gorgonio Pass lies at the northwestern edge of Coachella Valley

332 M.A. Patten

cases, that microsatellite variation and plumage variation (subspecies) were corre-

lated significantly (Pruett et al. 2008; cf. Zink 2010). This finding suggests that a

detailed genetic survey around the ring holds the promise of yielding a pattern that

corroborates the pattern evident in the analysis of plumage variation.

20.3 Models for the Evolution of Ring Species

The two recently proposed ring species need more study, but at the least the criteria

for establishing the pattern appear to have been met as convincingly as in the two

more well-studied examples of Ensatina eschscholtzii and Phylloscopus trochi-loides. But determining that a species or subspecies complex fits a ring species is

only half of the battle. How a ring pattern came to be is about the process of a ring

species, and the stringent criteria Coyne and Orr (2004:103) set forth for determin-

ing if a ring species is valid focused equally on process and pattern. Although these

authors agreed that criterion (1) above must hold, they modified (2) to state that

geographic continuity must have been present always; i.e., no geographic barriers to

gene flow could have existed in the past, during ring formation. They further

imposed two criteria related to the process by which the ring formed: (A) there

must be historical information that the ring was formed by a single population (i.e.,

not from two or more genetically distinct lines), with all subspecies around the ring

descended from that single line, and (B) one of the terminal points must be

represented by a population that expanded its range most recently. Criterion (A)

may be justified if we wish to hold up a ring species as a solid example of speciation

either in the face of gene flow or with geographic distance. Criterion (B), by

contrast, implies that the ring must have formed in a certain way, which ignores

other plausible ways in which a ring could evolve.

The model inherent in criterion (B) is consistent with the first model put forth for

the evolution of a ring species (Stejneger, in Jordan 1905), a half-century before the

term “ring species” was coined. In one of several published response to Jordan’s

review of geographic speciation, Stejneger postulated that two subspecies might

breed in sympatry, but only under specific circumstances. Using Lanius shrikes innorthern Europe as an example, he asked readers to imagine that two trajectories of

range expansion split from a common stock in Asia, with one heading west through

central Europe to reach the Scandinavian Peninsula by way of Denmark and the

other heading northwest through Finland to colonize the Scandinavian Peninsula

from the north. The ranges of these subspecies would meet in southern part of the

peninsula. Stejneger (p. 552) proposed that “it is then not unnatural to conclude that

in the specimens meeting there the characters might have become so fixed that the

two forms would react on each other as two distinct species, though at their original

dividing line they might still remain in the imperfectly differentiated stage”. This

scenario corresponds with the classic conceptual model of how a ring forms

(Fig. 20.2, “classical I”; Martens and P€ackert 2007). An alternative model

(Fig. 20.2, “classical II”) yields the same pattern and still invokes forming a ring

that would meet Criterion (B).


Using current snapshots to distinguish between various iterations of these

“classical” models can be challenging (Kuchta et al. 2009), but alternative models

that would yield the pattern of a ring species and conform to conceptual specifica-

tions of the “ring species hypothesis” (sensu Joseph et al. 2008) have not been

explored. Yet there are alternative models in which a ring pattern evolves by means

of a process that retains the concept’s emphasis on divergence with gene flow, a

possibility increasingly recognized as plausible (Nosil 2008; Thorpe et al. 2008;

Mila et al. 2009). One such model is a simple scenario invoking in situ divergence

across various ecotones around a ring (Fig. 20.2), with divergence being especially

pronounced across one moderately steep, but not too steep, environmental gradient

(Doebeli and Dieckmann 2003; Leimar et al. 2008). Taxa on either side of this

gradient diverge by the process of ecological speciation, “the evolution of repro-

ductive isolation between populations by divergent natural selection arising from

differences between ecological environments” (Schluter 2009). These taxa become

the terminal points of the ring. Because geographic ranges were always and are still

continuous, and intergradation persists at other contact points where gradients are

shallower, a ring species pattern forms in the face of gene flow.

Another model for the evolution of a ring species also invokes ecological

speciation across an environmental gradient (Fig. 20.2, “ecological divergence”).

In this case, ranges expand around a geographic barrier, just as in the classical

models; however, ranges split initially from the parent population across an ecotone

Fig. 20.2 Competing models

for the evolution of a species

ring. The “classical I” model

corresponds to Leonard

Stejneger’s (in Jordan 1905)

conception of how a ring

formed (see also Martens and

P€ackert 2007). A ring may

also form in the classical

sense by encircling the

geographic barrier back to the

starting point (see Kuchta

et al. 2009 for similar

examples). The “in situ”

model relies on repeated,

simultaneous ecological

speciation, whereas the

“ecological divergence”

model combines aspects of a

classical ring model (e.g.,

differentiation during range

expansion) with ecological

speciation

334 M.A. Patten

with a moderately steep gradient, an area conducive to divergence (Endler 1977).

As ranges expand around either side of the barrier, time elapsed at the initial branch

point is sufficient for divergence to occur there, but the expanding front does not

diverge at this same rate. Indeed, the two fronts remain undifferentiated enough that

when the fronts meet, the populations interbreed readily, forming a broad hybrid

zone of secondary contact. The end result would again be a ring species pattern in

the face of gene flow. The chief differences from the classical models are that

terminal points occur at an ecotone and are at the opposite end of the ring from

where the expanding fronts met.

It is important to note that a variety of other scenarios may lead to a ring species

pattern. For example, a species may have spread from multiple glacial refugia and in

doing so form multiple zones of secondary contact (Bensch et al. 2009). Or a set of

subspecies may have arisen by a process of vicariant (allopatric) divergence, but all

barriers between resultant forms have since eroded, leaving a ring of connected forms

with intergradation where ranges meet (Joseph et al. 2008). We therefore ought to

predict the existence of a ring species pattern in situations that cannot teach us about

speciation in the face of gene flow, an oft-cited hallmark of the ring species hypothe-

sis. Such examples only add to the abundant evidence for allopatric speciation, albeit

they will prove suitable for studies of the maintenance of geographic variation in the

face of gene flow (e.g., hybrid zone dynamics; Barton and Hewitt 1989).

20.4 Evolution of the Song Sparrow Ring

The Song Sparrow currently ranges across North America, with populations occur-

ring north to southwestern Alaska and to southern Canada east to Newfoundland

and contiguous populations south to northwestern Mexico. There are also geo-

graphically isolated populations on the Channel Islands and Islas Coronados off of

California and Baja California, respectively, and at various locales in mainland

Mexico, south to the Trans-Mexican volcanic belt (Patten and Pruett 2009). So wide

a geographic range may hinder interpretation of the evolution of geographic varia-

tion. We thus need to consider whether the species was always so widespread or,

more likely, if the species expanded its range considerably in the wake of the most

recent glaciation �12,000 ybp.

In the case of the Song Sparrow, two genetic analyses (Zink and Dittmann 1993;

Fry and Zink 1998) identified two or three Pleistocene refugia, respectively. That is,

extant populations of the Song Sparrow carry a genetic signature that implies range

expansion away from either two or three regions that harbored the species’ ances-

tors during the last glacial maximum (Fig. 20.3; see Sommer and Zachos 2009).

Two refugia identified by mtDNA restriction sites (Zink and Dittmann 1993) were

Newfoundland and the Queen Charlotte Islands, British Columbia (Fig. 20.3).

Because Newfoundland was covered by a sheet of ice, it seems an implausible

site for a refugium. This concern was alleviated by a follow-up study of mtDNA

sequence (Fry and Zink 1998), who found evidence for a “model of Song Sparrow


population history involving multiple Pleistocene refugia and colonization of some

formerly glaciated regions from multiple sources”. Their study identified three

refugia: the Queen Charlotte Islands, the Atlantic Coast of the northeastern United

States, and, likely, southern California (Fig. 20.3).

Southern California was considered a likely location for a refugium, but it could

not be identified conclusively because sample size was small. Nevertheless, a

genetic survey across a suite of terrestrial vertebrate taxa – but not including the

Song Sparrow – identified southeastern California as a Pleistocene refugium

(Waltari et al. 2007), lending support to Fry and Zink’s (1998) finding. Waltari

et al. (2007) also presented evidence for a refugium in the central or southern Baja

Fig. 20.3 Approximate extent of the North American ice sheets during the last glacial maximum

(Ehlers and Gibbard 2004). On the basis of mitochondrial DNA restriction sites and sequences

(Zink and Dittmann 1993; Fry and Zink 1998), three glacial refugia (dashed circles) for the SongSparrow (Melopsiza melodia) have been proposed. A fourth (solid circle) was proposed initially

but later discarded

336 M.A. Patten

California peninsula, a location Fry and Zink (1998) could not have detected

because they lacked samples of the Song Sparrow from the peninsula. The Baja

California peninsula nonetheless corresponds to a common Pleistocene refugium

incorporated into a meta-analysis of North American hybrid zones (Fig. 20.4;

Swenson and Howard 2005). That the sparrow occurs currently in all three (or

four, if we include Baja California as separate from southern California) putative

refugia (Fig. 20.3) raises the possibility of future screening for ancestral haplotypes,

preferably in the nuclear genome.

The issue of hybrid or contact zones is an additional crucial consideration when

piecing together the evolutionary and biogeographic history of the Song Sparrow.

The contact zone of the terminal points of the sparrow ring occurs in the Coachella

Valley, at the southeastern base of San Gorgonio Pass (Fig. 20.1). The San

Fig. 20.4 Proposed routes of range expansion away from glacial refugia (squares) in North

America (after Swenson and Howard 2005)


Gorgonio Pass divides the north end of the north–south Peninsular Ranges from the

east–west Transverse Ranges and is an area of faunal transition (Patten et al. 2004a;

Leavitt et al. 2007). It has been identified as a “hot spot” for phylogeographic breaks

(Swenson and Howard 2005), locations where there are deep splits in phylogenetic

history. The Transverse Ranges themselves figure prominently in phylogenetic

breaks: animal taxa (invertebrate and vertebrate) either north or south of that line

of mountains tend to be in separate phylogenetic clusters (Calsbeek et al. 2003;

Burns et al. 2007), further emphasizing the prominence of the San Gorgonio Pass

region as a contact zone hot spot.

That the terminal points of the Song Sparrow ring occur in this region of faunal

transition is likely not a coincidence. If we accept that the Song Sparrow’s ancestors

persisted in a glacial refugium in southern California or in Baja California and

spread north from there (Figs. 20.3 and 20.4), a cleave in the expanding fronts of the

geographic range would be at the San Gorgonio Pass. The moderately steep

environmental gradient in the pass – from a Mediterranean climate at the northwest

end to an extreme desert climate at the southeast end – is conceivably ideal for

ecological speciation. If speciation occurred while the expanding fronts differen-

tiated, via isolation by distance, enough to be recognized as subspecies but not

enough to yield reproductive isolation, then the result would be a true ring species

that evolved by a process that best fit the “ecological divergence” model (Fig. 20.2).

Conversely, Lapointe and Rissler (2005) examined congruent phylogeographies

across California of seven verebrates, an invertebrate, and a plant and found general

patterns that corresponded broadly to the ranges of the subspecies of the Song

Sparrow that constitute the ring (Fig. 20.1). If these regions, each of which has a

distinct environment (i.e., general climate and vegetation), tend to promote diver-

gence via an ecological speciation model, then the San Gorgonio Pass still might be

the site of speciation when other contact zones represent areas where locally

adapted populations meet. Such a scenario would yield a true ring species, but

one that evolved by means of the “in situ” model (Fig. 20.2).

Morphologically, the California subspecies of the Song Sparrow form a distinct

group, as do the subspecies in the desert Southwest and the mesic Pacific Northwest

(Patten and Pruett 2009). It therefore seems unlikely that postglacial range expan-

sion was solely from the Queen Charlotte refugium, a requisite for the ring to

conform to a “classical I” model (Fig. 20.2). Evolution by means of a “classical II”

model may be more likely, if the ancestral taxon expanded north to encircle the

Sierra Nevada and Mojave Desert counterclockwise, yet such a pattern would not

jibe with general tracks of postglacial expansion in other species (Fig. 20.4;

Swenson and Howard 2005). Moreover, the subspecies M. m. rivularis of Baja

California Sur is morphologically most likeM. m. fallax of the Sonoran Desert, oneof the terminal points of the ring; indeed, they are nearly identical in plumage – the

principal difference is the diagnostically longer bill of M. m. rivularis (Patten and

Pruett 2009). If phenotype corresponds to evolutionary relatedness and the Pleisto-

cene refugium was in the Baja California peninsula, then the ancestral form

expanded northward only on the east side of the peninsula, an unlikely scenario

338 M.A. Patten

given presumably spotty suitable habitat in the far more xeric portion of Baja

California east of the Peninsular Ranges.

20.5 Conclusions

Morphological variation in the Song Sparrow in the southwestern United States

creates a ring species pattern around the Sierra Nevada and Mojave Desert (Patten

and Pruett 2009). A detailed study of two subspecies that differ most strikingly in

plumage implies that they are terminal points of the ring (Patten et al. 2004b). These

subspecies meet at the base of the San Gorgonio Pass, a well-known area of faunal

transition (Leavitt et al. 2007).

Yet prima facie evidence suggests that neither of the classical models for the

evolution of a ring species (Fig. 20.2) holds in this case. A glacial refugium for

the Song Sparrow likely existed in the desert Southwest (Fry and Zink 1998),

and postglacial range expansion from this region tended to be of a northward trajec-

tory (Swenson and Howard 2005). It thus would appear that an “ecological diver-

gence” model is the most plausible. This model requires ecological speciation of

M. m. heermanni and M. m. fallax, the terminal points, across the San Gorgonio

Pass while the species expanded its range northward on either side of the Sierra

Nevada and Mojave Desert (Fig. 20.5). At this stage an “in situ” model cannot be

eliminated, and distinguishing between these models requires detailed genetic, eco-

logical, and behavioral research around the ring. Even so, Occam’s razor would

argue in favor of the “ecological divergence” model, if only because it invokes

ecological speciation (or subspeciation) at only one location instead of a minimum

of four (the number of contact zones between Song Sparrow subspecies that form

the ring).

There are additional wrinkles in the formation of the Song Sparrow ring. For

example, M. m. cleonensis is morphologically intermediate between subspecies

in the “California group” and those in the “Alaska and Pacific Northwest group”

(sensu Patten and Pruett 2009). I suggest that this intermediacy reflects a historical

merging of a northward expanding front from the refugium in southern California

and the southward expanding front from the Queen Charlotte Islands. That

M. m. montana, the northern “cap” to the species ring, shares characters of both

California and “Eastern” subspecies also implies extensive gene flow, but it remains

to be determined whether eastward and southward fronts merged to leave a ring

species pattern without divergence in the face of gene flow or by distance.

Only in-depth studies that combine morphology, genetics (especially nuclear

DNA), ecology, and geological history will be able to distinguish among various

models for the evolution of a ring species or confirmation of the “ring species

hypothesis” (Joseph et al. 2008; Bensch et al. 2009). Regardless, an important

starting point for any investigation of a putative ring species is full consideration of

all plausible models that could have led to a ring species’ evolution, not just an


expectation of conformity to classical models. Consideration of alternative models

not only promises to provide deeper insight in how ring species evolve but also

promises to build a stronger bridge between micro- and macroevolution.

San Gorgonio Pass

Fig. 20.5 Hypothesized postglacial expansion of the Song Sparrow (Melospiza melodia) from an

identified (but nonetheless postulated) glacial refugium in the Sonoran Desert (dashed circle).Such range expansion would yield a ring species pattern, but in this species’ case the terminal

points are in the vicinity of the San Gorgonio Pass, meaning the ring evolved by a combination of

“divergence by distance” and ecological speciation (the “ecological divergence” model of

Fig. 20.2), a process heretofore not considered in studies of ring species

340 M.A. Patten

Acknowledgments I thank Pierre Pontarotti for the opportunity to speak at the 13th Evolutionary

Biology Meeting and Axelle Pontarotti for her excellent guidance both pre and post meeting. John

T. Rotenberry, Leonard Nunney, and Marlene Zuk advised during early stages of this study, and

Christin L. Pruett has been a sounding board during later stages. I am grateful to Lukas F. Keller

and his research group and colleagues at Universit€at Z€urich for their feedback following my

September 2008 seminar there. Brenda D. Smith-Patten has been a limitless source of support

throughout this research; she also helped prepare Fig. 20.2 and commented on a draft of this

chapter.

References

Barton NH, Hewitt GM (1989) Adaptation, speciation, and hybrid zones. Nature 341:497–503

Bensch S, Grahn M, M€uller N, Gay L, Akesson S (2009) Genetic, morphological, and feather

isotope variation of migratory Willow Warblers show gradual divergence in a ring. Mol Ecol

18:3087–3096

Burns KJ, Alexander MP, Barhoum DN, Sgariglia EA (2007) Statistical assessment of congruence

among phylogeographic histories of three avian species in the California Floristic Province.

Ornithol Monogr 63:96–109

Cain AJ (1954) Animal species and their evolution. Princeton University Press, Princeton, NJ

Calsbeek R, Thompson JN, Richardson JE (2003) Patterns of molecular evolution and diversifica-

tion in a biodiversity hotspot: the California Floristic Province. Mol Ecol 12:1021–1029

Coyne JA, Orr HA (2004) Speciation. Sinauer Assoc, Sunderland, MA

Doebeli M, Dieckmann U (2003) Speciation along environmental gradients. Nature 421:259–264

Ehlers J, Gibbard PL (2004) Quaternary glaciations – extent and chronology, part 2: North

America. Elsevier, Amsterdam

Endler JA (1977) Geographic variation, speciation, and clines. Princeton Monogr Pop Biol

10:1–246

Erwin DH (2000) Macroevolution is more than repeated rounds of microevolution. Evol Dev

2:78–84

Fry AJ, Zink RM (1998) Geographic analysis of nucleotide diversity and Song Sparrow

(Aves: Emberizidae) population history. Mol Ecol 7:1303–1313

Hansen TF, Martins EP (1996) Translating between microevolutionary process and macroevolu-

tionary patterns: the correlation structure of interspecific data. Evolution 50:1404–1417

Irwin DE, Irwin JH (2002) Circular overlaps: rare demonstrations of speciation. Auk 119:596–602

Irwin DE, Bensch S, Price TD (2001a) Speciation in a ring. Nature 409:333–337

Irwin DE, Irwin JH, Price TD (2001b) Ring species as bridges between microevolution and

speciation. Genetica 112–113:223–243

Irwin DE, Bensch S, Irwin JH, Price TD (2005) Speciation by distance in a ring species. Science

307:414–416

Jablonski D (2000) Micro- and macroevolution: scale and hierarchy in evolutionary biology and

paleobiology. Paleobiology 26(suppl):15–52

Jordan DS (1905) The origin of species through isolation. Science 22:545–562

Joseph L, Dolman G, Donnellan S, Saint KM, Berg ML, Bennett ATD (2008) Where and when

does a ring start and end? Testing the ring-species hypothesis in a species complex of

Australian parrots. Proc Biol Sci 275:2431–2440

Kuchta SR, Parks DS, Mueller RL, Wake DB (2009) Closing the ring: historical biogeography of

the salamander ring species Ensatina eschscholtzii. J Biogeogr 36:982–995Lapointe F-J, Rissler LJ (2005) Congruence, consensus, and the comparative phylogeography of

codistributed species in California. Am Nat 166:290–299


Leavitt DH, Bezy RL, Crandall KA, Sites JW Jr (2007) Multi-locus DNA sequence data reveal a

history of deep cryptic vicariance and habitat-driven convergence in the desert night lizard

Xantusia vigilis species complex (Squamata: Xantusiidae). Mol Ecol 16:4455–4481

Leimar O, Doebeli M, Dieckmann U (2008) Evolution of phenotypic clusters through competition

and local adaptation along an environmental gradient. Evolution 62:807–822

Martens J, P€ackert M (2007) Ring species – do they exist in birds? Zool Anz 246:315–324

Mayr E (1942) Systematics and the origin of species. Columbia University Press, New York

Mayr E (1982) Speciation and macroevolution. Evolution 36:1119–1132

Mila B, Wayne RK, Fitze P, Smith TB (2009) Divergence with gene flow and fine-scale

phylogeographical structure in the wedge-billed Woodcreeper, Glyphorynchus spirurus, aneotropical rainforest bird. Mol Ecol 18:2979–2995

Nosil P (2008) Speciation with gene flow could be common. Mol Ecol 17:2103–2106

Patten MA, Pruett CL (2009) The Song Sparrow as a ring species: patterns of geographic variation,

a revision of subspecies, and implications for speciation. System Biodivers 7:33–62

Patten MA, Erickson RA, Unitt P (2004a) Population changes and biogeographic affinities of the

birds of the Salton Sink, California/Baja California. Studies Avian Biol 27:24–32

Patten MA, Rotenberry JT, Zuk M (2004b) Habitat selection, acoustic adaptation, and the

evolution of reproductive isolation. Evolution 58:2144–2155

Pruett CL, Arcese P, Chan YL, Wilson AG, Patten MA, Keller LF, Winker K (2008) Concordant

and discordant signals between genetic data and described subspecies of Pacific coast Song

Sparrows. Condor 110:359–364

Reznick DN, Ricklefs RE (2009) Darwin’s bridge between microevolution and macroevolution.

Nature 457:837–842

Schluter D (2009) Evidence for ecological speciation and its alternative. Science 323:737–741

Simons AM (2002) The continuity of microevolution and macroevolution. J Evol Biol 15:688–701

Sommer RS, Zachos FE (2009) Fossil evidence and phylogeography of temperate species: ‘glacial

refugia’ and post-glacial recolonization. J Biogeogr 36:2013–2020

Stanley SM (1998) Macroevolution: pattern and process. Johns Hopkins University Press,

Baltimore

Stebbins RC (1957) Intraspecific sympatry in the lungless salamander Ensatina eschscholtzii.Evolution 11:265–270

Swenson NG, Howard DJ (2005) Clustering of contact zones, hybrid zones, and phylogeographic

breaks in North America. Am Nat 166:581–591

Thorpe RS, Surget-Groba Y, Johansson H (2008) The relative importance of ecology and

geographic isolation for speciation in anoles. Phil Trans R Soc Lond B Biol Sci 363:3071–3081

Wake DB (2006) Problems with species: patterns and processes of species formation in salaman-

ders. Ann Mo Bot Gard 93:8–23

Wake DB, Yanev KP (1986) Geographic variation in allozymes in a “ring species”, the pletho-

dontid salamander Ensatina eschscholtzii of western North America. Evolution 40:702–715

Waltari E, Hijmans RJ, Peterson AT, Nyari AS, Perkins SL, Guralnick RP (2007) Locating

Pleistocene refugia: comparing phylogeographic and ecological niche model predictions.

PLoS ONE 2(7):e563

Zink RM (2010) Drawbacks with the use of microsatellites in phylogeography: the Song Sparrow

Melospiza melodia as a case study. J Avian Biol 41:1–7

Zink RM, Dittmann DL (1993) Gene flow, refugia, and evolution of geographic variation in the

Song Sparrow (Melospiza melodia). Evolution 47:717–729

342 M.A. Patten

Chapter 21

Cave Bear Genomics in the Paleolithic Painted

Cave of Chauvet-Pont d’Arc

Celine Bon and Jean-Marc Elalouf

Abstract Caves are reservoirs of fossils, some of which belong to species now

extinct. Paleogenetics explores ancient DNA that may have survived in these fossils

to better understand the phylogeny of Pleistocene species and the paleoenviron-

ment. The Chauvet-Pont d’Arc Cave, which displays the earliest known human

drawings, contains thousands of animal remains, setting this cave as a mine for

genetic analysis. We focused on the extinct cave bear, Ursus spelaeus, and proved

that Chauvet-Pont d’Arc samples still contain enough DNA for genetic studies. One

of them yielded well-preserved DNA and allowed sequencing the complete cave

bear mitochondrial genome. We used this molecular information to establish bear

phylogeny and the tempo of Ursidae speciation. Widening our analysis to cave

bears samples from Chauvet-Pont d’Arc and a closely located cave, we showed that

the Pleistocene ursine population was highly homogeneous at the regional level.

21.1 The Chauvet-Pont d’Arc Cave, a Well-Preserved

Paleolithic Site

21.1.1 The Earliest Rock Art Recorded to Date

In 1994, the three cavers Jean-Marie Chauvet, Eliette Brunel, and Christian Hillaire

made a major discovery in the field of archeology: they found a cave containing

hundreds of Paleolithic rock art pictures. This cave, located near Vallon-Pont d’Arc

(Ardeche, Southeastern France) at the entrance of the Ardeche Gorge, is now

known as Chauvet-Pont d’Arc from one of its discoverers, Jean-Marie Chauvet.

C. Bon and J-M. Elalouf

CEA, IBiTec-S, F-91191 Gif-sur-Yvette cedex, France



343

Since some of the pictures were drawn with charcoal, dating analysis was

possible using the radiocarbon method. Several paintings returned a radiocarbon

age between 30,000 and 32,000 years Before Present (BP), which sets them about

twice older than the age currently proposed for Lascaux Cave paintings. Chauvet-

Pont d’Arc rock art is the oldest Paleolithic drawing known to date (Valladas et al.

2001). The cave displays three kinds of rock art pictures: charcoal- and ochre-made

drawings and engravings. As dating is only feasible for charcoal-made pictures,

some of the other pictures might be older than 32,000 years BP.

The cave also contains other remains of human occupation. The track of a male

infant was found in a deep part of the cave, in the Gallery of the Crosshatches.

During his trip, the child regularly rubbed his torch against the wall, leaving

numerous sooty marks. These marks were radiocarbon dated back to 26,000 years

BP (Garcia 2005).

Huge hearths were found in other cave sectors and were most probably used by

Paleolithic artists for the production of charcoal pencils. The cave also contains

about 20 flint tools as well as an ivory assegai point (Geneste 2005). Other

anthropogenic processes, such as stone blocks grouped together by humans or a

cave bear skull deposited on a large rock, remain enigmatic.

Due to the rich overall archeological content and, especially, the great age of the

rock art pictures, the Chauvet-Pont d’Arc Cave is protected from the very day of its

discovery (Baffier 2005). As soon as they saw the first rock art pictures, the three

discoverers took care to protect the ancient soil. Afterwards, footbridgeswere installed

throughout the cave. The access to the cave is restricted to a handful of people that are

granted authorization from the prefect. A permanent watch was set to detect microbial

pollution as well as local climate change. Even the scientific researches are strictly

monitored to ensure preservation of the site. Thus, there are only two short campaigns

of studies each year, no more than 12 people are tolerated inside the cave, no direct

contact with the archeological remains or the walls are allowed, and retrieving of

samples rests on special curator’s authorization (Baffier 2005).

Despite these constraints, the cave provides a unique basis for scientific research

because its preserved state gives us access to a Paleolithic site untouched since the

entrance of the cave collapsed some 20,000 years ago.

21.1.2 The Chauvet-Pont d’Arc Cave, a Bear Cave

Even without such anthropogenic remains, Chauvet-Pont d’Arc would still have

been a major paleontological discovery since it displays thousands of animal

remains, most of which consist of Ursus spelaeus bones (Fig. 21.1) (Fosse and

Philippe 2005). Among the 3,844 bones dispatched all over the ground, 3,703 are

ascribed to the cave bear. The brown bear (Ursus arctos) has been identified

through a single skull, which contrasts with the 200 cave bear skulls that are present

in Chauvet-Pont d’Arc. Other species, such as the wolf, extinct cave hyena, fox,

ibex, deer, are evidenced by a few samples. Canidae coprolites and footprints are

also present in the cave.

344 C. Bon and J.-M. Elalouf

Fig. 21.1 Topography of the Chauvet-Pont d’Arc Cave. Blue areas correspond to places with cavebear wallows; purple circles indicate cave bears footprints; green thick lines on walls indicate thatcave bear claw marks are present. Radiocarbon ages are given as years BP. Topography:

Y. Le Guillou and F. Maksud. Paleontological data: P. Fosse and M. Philippe

21 Cave Bear Genomics in the Paleolithic Painted Cave of Chauvet-Pont D’Arc 345

But the cave is not only a bear grave, for it also displays many evidences of live

animal’s occupation. The ground is warped by the numerous wallows in which

bears used to hibernate; the walls are scratched by claw marks and polished by their

roaming; bear footprints can be seen in every chamber.

Whereas the brown bear is still an extant species, the cave bear became extinct

about 25,000 years ago (Pacher and Stuart 2009). Ursus spelaeus was a robustly

built bear that weighed 200 kg more than the sturdiest extant bears, i.e., the Kodiac

and polar bears. The sexual dimorphism is strong, as well as the intraspecific

variability (Kurten 1976). It is currently estimated that the cave bear was confined

to Europe, even though cave-bear-looking bears that may belong to some cave bear

subspecies were found in Crimea, Caucasus, or Siberia (Knapp et al. 2009). It has

been considered that the cave bear was mostly herbivorous, but two recent studies

(Richards et al. 2008; Peigne et al. 2009) showed that it was omnivorous at least

during the prehibernation period.

Since the cave bear is an extinct species, its phylogenetic relationship with other

bears has long been only known through paleontological data. The direct ancestor of

the cave bear is Ursus deningeri, because Ursus spelaeus succeeds continuously to

Ursus deningeri (Mazza and Rustioni 1994). It is estimated that the transition

between the two species occurred around the beginning of the last interglacial, but

to draw a limit between these two chrono-species may be awkward (Argant 2001).

Views diverge about the origins of theUrsus arctos and theUrsus spelaeus lineages.Whereas most paleontologists assume that these two lineages emerged from Ursusetruscus, Mazza and Rustioni proposed that Ursus etruscus is a dead end, and that

Ursus deningeri appeared among extremely polymorphic Ursus arctos lineages.This issuewas first questioned in 1994 by analyzingmitochondrial DNA fragments

from Pleistocene remains (Hanni et al. 1994). This initial studies and subsequent work

(Loreille et al. 2001) yielded sequence data for the mitochondrial control region and

cytochrome b (CYTB) gene. However, when we initiated our studies the information

available consisted of less than 10 % of the mitochondrial genome. As increasing

evidences suggest that long sequences are necessary to obtain robust phylogenies and

to accurately date the divergence events between lineages (Rohland et al. 2007), a

complete cave bear mitochondrial genome sequence was highly desirable (Bon et al.

2008).

21.2 Sequencing the Mitochondrial Genome of the

Extinct Cave Bear

21.2.1 The Challenge of Retrieving and Sequencing Ancient DNA

The study of ancient DNA is tricky. Although in the living cell enzymatic processes

continuously repair DNA, endogenous nucleases and exogenous fungi or bacteria

begin degrading DNA from the death of an organism. Under rare circumstances

(such as rapid desiccation or adsorption on a mineral matrix), the DNA may escape


the onslaught, its only source of deterioration being through chemical processes

(Hofreiter et al. 2001b; Paabo et al. 2004). Thus ancient DNA is scarce and displays

a number of chemical alterations. This has several consequences. The length of the

DNA molecules is reduced by strand breaks. In addition, depurination and cross-

linking between strands or between a DNA strand and another molecule result in

impeding PCR amplifications. As the initial amount of ancient DNA is extremely

low, the amplification stage is sensitive to contaminations, not only from modern

DNA but also from previously amplified products. Another problem is the deami-

nation of cytosine and adenine, leading to mutations such as T instead of C, and G

instead of A in the retrieved sequence. At last, the samples often contain a variety of

organic molecules that may act as PCR inhibitors. This prevents the use of a large

amount of extract in the PCR mix.

Considering the care taken to protect the Chauvet-Pont d’Arc Cave from con-

taminations, we turned to it to select an eligible cave bear sample for the sequencing

of the mitochondrial genome. After screening several samples, we chose US18

because of its biomolecular preservation. It still contained enough collagen for

radiocarbon dating, and the amino-acid racemization extent was quite low. After

DNA extraction, a 117 bp mitochondrial sequence was amplified over a wide range

of sample extract (from 0.1 to 2%), which shows that we retrieved large amounts of

DNA and few PCR inhibitors. Since independent replication is required in ancient

DNA studies, another group of investigators from another Institute performed

extraction and analysis. The same and another overlapping pair of primers were

used and confirmed the sequence initially obtained. Both extracts were employed in

the subsequent experiments.

21.2.2 Obtaining the Complete Cave Bear MitochondrialSequence

When this analysis began, only few fragments of the cave bear mitochondrial

genome were known: a portion of the control region had been sequenced from

several samples (Hanni et al. 1994; Hofreiter et al. 2002, 2007; Orlando et al. 2002;

Rohland et al. 2004). A single gene, namely CYTB, had been characterized

throughout its coding region from one sample found in the Balme-a-Collomb

Cave (Loreille et al. 2001).

We designed an iterative experimental strategy to determine the cave bear

mitochondrial genome. First, we aligned the mitochondrial genomes of the extant

brown bear (Ursus arctos), polar bear (Ursus maritimus), and American black bear

(Ursus americanus) (Delisle and Strobeck 2002). From this alignment, conserved

regions were identified and used to design a first series of primers for amplifying

DNA fragments ranging from 100 to 200 bp. These 147 primer pairs spanned the

entire genome.

Only 64 primer pairs out of 147 succeeded; the 83 failures may result from mis-

pairing between the template cave bear DNA and the primers. As a consequence, in


the following rounds, we used the sequence obtained from previous runs to design

cave bear specific primers. In the end, nine rounds were required and we successfully

used 245 primer pairs.

In order to avoid contaminations, prePCR steps were done in a dedicated

laboratory facility, in a building free from molecular biology research. Each primer

pair was designed to amplify DNA fragments shorter than 200 bp. For each

fragment, at least two PCR amplifications were performed. As differences caused

by ancient DNA damages were usually detected, a third amplification was often

carried out, and the consensus sequence was retained. In the worst case scenario,

this strategy is expected leading to a 0.06% error rate (Hofreiter et al. 2001a). PCR

products were cloned and a minimum of 12 colonies was sequenced on both

strands. In the end, 570 successful PCR amplifications and more than 14,000

sequencing reactions were required to cover the entire mitochondrial genome.

In order to check the accuracy of the sequence, we analyzed each fragment

individually by BLAST to validate that the best GenBank match was an Ursidaesequence. Specifically, we verified that previously analyzed cave bear mitochon-

drial sequences (control region and CYTB gene) displayed the best BLAST score

with our analogous sequences.

The control region sequence of US18 cave bear belongs to the B haplotype as

defined in Orlando et al. (2002) and is identical to Scladina cave’s samples SC3500

and SC3800. Our and the published CYTB sequences differ only on four transitions

(0.35% of all CYTB nucleotides), two of them being located at the third base

position of codons. Furthermore, as the two specimens belong to different mito-

chondrial haplotypes, these differences may highlight intraspecific polymorphism.

We obtained a 16,810 bp long mitochondrial genome, which is in the range of

the extant Ursidae mitochondrial genomes. These genomes vary in length between

16,723 bp (Ursus maritimus) (Arnason et al. 2002) and 17,044 bp (Ursus thibetanusformosanus). The variation of the mitochondrial genome length is mainly due to a

domain of the control region, which displays a highly variable number of repeat of a

10 bp motif (Yu et al. 2007). This domain is longer than 200 bp and therefore cannot

be retrieved through a single PCR from ancient cave bear extracts. Thus, we

designed two primer pairs to target the 50 and the 30 ends of the domain. Afterwards,

all fragments were assembled into a 350 bp repeat sequence.

Another group has sequenced a second cave bear mitochondrial genome from a

sample found in Gamssulzen cave, Austria (Krause et al. 2008). This sample is a

44,000-year-old bone and its sequence belongs to the D haplogroup as defined in

Orlando et al. (2002). The experimental strategy was slightly different from ours as

they used a two-step multiplex approach PCR. As we did, they confirmed their data

by at least two independent amplifications, cloning of the PCR product and

sequencing of multiple clones.

Both cave bears sequences are very similar. Without taking into account the

350 bp repeat region, 16,227 bp among 16,448 are identical. As expected, the 221

mutations are rather transitional mutations (216) than transversional (5), with a

transition/transversion ratio equal to 43.2. As these two sequences belong to

different haplogroups, it is not surprising that they display 1.3% differences.


Our aim was to determine the phylogenetic position of the cave bear, especially

with respect to the two main brown bear lineages (Taberlet and Bouvet 1994). As

only one brown bear mitochondrial genome was published, we decided to sequence

the mitochondrial genome of a brown bear belonging to the western lineage. We

analyzed a submodern bone sample from a French Pyrenean site (Guzet, Ariege,

France). This was conducted in a third building and after the cave bear mitochon-

drial genome had been obtained to avoid cross-species contaminations. The same

experimental strategy was followed, except that the first series of primers (designed

on a brown bear sample) was already highly specific, and that, as submodern DNA

is still well conserved, less primer pairs were needed (only 52 primer pairs). As for

the cave bear sequence, each PCR was performed at least twice, several clones were

sequenced, and the consensus sequence was checked using BLAST.

21.2.3 Resolving the Phylogeny of the Extinct Cave Bear

In order to obtain the Ursidae phylogeny, we aligned the cave bear and the

Pyrenean brown bear mitochondrial sequences (EU327344 and EU497665, respec-

tively) with sequences retrieved from GenBank for other bears species, using

MEGA 4.0.2 alignment tool with the default parameters. The giant panda was set

as an outgroup. The domain of the control region containing the 10 bp repeat motif

was removed prior to the phylogenetic analyses.

First, we tested the mutational saturation of our dataset, in order to check that

homoplasy keeps low and does not alter the results. We calculated the patristic

distance using Patristic software (Fourment and Gibbs 2006) and plotted the genetic

distance against the patristic distance. These distances are almost equal, indicating

that mutational saturation is weak and that few reversions affect the dataset. We

also calculated the transition/transversion ratio, which is equal to 19:1. As this ratio

is rather high, it confirms that saturation is rare.

Phylogenetic trees were reconstructed from this dataset using Neighbor Joining

(NJ), Maximum Parsimony (MP), and Maximum Likelihood (ML) using PhyML

(Guindon and Gascuel 2003) and Mega 4.0.2 (Tamura et al. 2007) softwares, as

appropriate. PhyML was implemented with a GTR þ G4 substitution model with

some invariable sites, and for the NJ reconstruction method, we used the Tamura

3-parameters and the gamma-distribution shape parameter estimated with PhyML.

The robustness of the phylogenetic trees was estimated with the bootstrap method

(1,000 replicates for NJ and MP, 100 replicates for ML).

Almost the same topology was recovered whatever the algorithm used

(Fig. 21.2). The only difference concerns Ursus thibetanus subspecies’ relation-

ships. Our results confirm the spectacled bear’s (i.e. Tremarctos ornatus) basal

position (Waits et al. 1999; Yu et al. 2004, 2007; Pages et al. 2008). Ursinae is a

monophyletic group in which Melursus ursinus is the most basal bear. Then

Ursinae split into two clades, one leading to Ursus spelaeus, Ursus arctos, andUrsus maritimus and the other leading to Ursus thibetanus, Ursus americanus, and


Helarctos malayanus. Whereas the first group is highly robust (all bootstrap values

equal 100%), the second one is less statistically supported. Besides, this clade is

not always found when analyzing shorter dataset (Talbot and Shields 1996a;

Waits et al. 1999; Yu et al. 2004, 2007; Bon et al. 2008; Pages et al. 2008).

As most of the internal branches are very short, we conclude that ursine speciation

Fig. 21.2 Molecular phylogeny inferred from complete mitochondrial genomes. Tree reconstruc-

tion was performed by NJ analysis using the giant panda (Ailuropoda melanoleuca) as an out-

group. The same tree topology was obtained using two other methods, except for the relationships

between Ursus thibetanus subspecies. Bootstrap values are indicated for NJ (regular), MP (bold),and ML (italic) analysis. The two sequences from this study are displayed in bold. GenBank

accession numbers for the other sequences are: Ailuropoda melanoleuca, FM177761, EF212882,

EF196663, and AM711896; Tremarctos ornatus, FM177764 and EF196665; Melursus ursinus,EF196662; Ursus thibetanus, EF1966362, EF667005, FM177759, EF587265, EF076773, and

EF196661; Ursus americanus, AF303109; Helarctos malayanus, FM177765 and EF196664;

Ursus maritimus, AF303111 and AJ428577; Ursus arctoseast, AF303110; Ursus spelaeus,FM177760


was very rapid. Because of this radiation, it is difficult to retrieve the branching

order, except for the brown-polar-cave bear clade.

Relationships within this group are always consistent and are supported by

maximal bootstrap values. The cave bear stands as a sister species to the brown

and polar bear clade. The brown bear species is a paraphyletic group with respect to

Ursus maritimus, as the polar bear species emerges from the western brown bear

lineage (Talbot and Shields 1996b).

Therefore, mitochondrial genome data disagree with Mazza and Rustoni’s late

speciation hypothesis and confirm that the cave bear and brown bear lineages split

before the radiation of the brown bear species.

The robust phylogeny obtained with a complete mitochondrial genome offers

the opportunity of evaluating the divergence times between species. We used the

BEAST software (Drummond et al. 2005; Drummond and Rambaut 2007) with the

complete mitochondrial genomes dataset. Calibration was performed with the

divergence between the giant panda and Ursidae, and between Ursinae and Tre-marctinidae, set at 12 � 1 MY and 6 � 0.5 MY (million years), respectively,

considering a normal distribution. We chose a relaxed uncorrelated lognormal

molecular clock, a GTR þ G4 substitution model with some invariable sites and

a Yule process of speciation. Two independent chains that each consist of

10,000,000 points were calculated and the burn-in was set to 10,000.

To highlight the benefits brought by the analysis of long DNA sequences in

molecular dating analysis, we randomly created alignments of various lengths from

whole mitochondrial genome sequences. We calculated node ages using the para-

meters described above. Obviously, short sequences yield different node ages and

wider credibility intervals than longer sequences. The alignment has to reach at

least 10 kb to stabilize the node ages. A long sequence alignment is therefore

required to obtain an accurate molecular dating (Bon et al. 2008).

According to the results obtained with complete mitochondrial genomes

(Fig. 21.3), Tremarctinae diverged from Ursinae 6.3 MY ago, shortly before the

appearance of Ursus boeckhi, the first ursine representative. The bears radiation

occurred about 4 millions years later, between 2 and 3 MY ago. The short time

while five bears groups appeared explains the difficulties in determining the

branching order of bears. These speciations happened during the Pliocene, when

Ursus minimus was the most common bear in Europe. As this fossil species is

assumed to be the last common ancestor ofUrsus spelaeus,Ursus arctos, andUrsusthibetanus, our results agree with paleontological data.

We date the divergence event between arctoid and speleoid lineages to 1.6 MY,

during the Villafranchian stage, when Ursus etruscus was the main bear in Europe.

Most paleontologists consider that Ursus etruscus was the last common ancestor of

the brown and cave bears.

In conclusion, our approach proved successful for sequencing the complete mito-

chondrial genome of a species extinct for more than 20,000 years. The cave bear

mitochondrial genome shares high similarities with other bear mitochondrial gen-

omes. In addition, the phylogenetic analysis robustly confirms that the cave bear is a

sister species to the brown and polar bear clade. The amount of data obtained made


possible to evaluate the tempo of bears’ history during Pliocene and Pleistocene and

compare our conclusions with paleontological ones.

The cave bear mitochondrial genome sequence opens up possibilities to push

forward extinct bears DNA analysis. First, this sequence will help rescuing poorly

preserved samples by targeting different regions of the mitochondrial genome. We

studied Chauvet-Pont d’Arc bear samples that failed to yield any DNA when

Fig. 21.3 Phylogeny and divergence times determined using the mitochondrial genome sequence

of the cave bear and of eight extant bears. Divergence times were calculated using BEAST

software with the splits between the giant panda and Ursidae and between Ursinae and Tremarc-tinidae set to 12 and 6 MY, respectively. Age for each node and 95% credibility intervals are, as

follows: 1, 6.3 MY (5.4–7.2); 2, 3.0 MY (2.2–3.8); 3, 2.8 MY (2.1–3.5); 4, 2.4 MY (1.7–3); 5, 2.1

MY (1.4–2.7); 6, 1.6 MY (1–2.1); 7, 0.6 MY (0.3–0.8); and 8, 0.4 MY (0.2–0.5). The extinct cave

bear is displayed by a picture from Chauvet-Pont d’Arc


analyzed for the mitochondrial control region. We targeted 112 bp in the 16 S gene

and obtained a successful amplification for 48% of the 23 samples, instead of 17%

when the control region was queried. Second, sequence data provided by extant

bears may not be sufficient to analyze DNA sequences of species that existed before

Ursus spelaeus, such as Ursus deningeri. The availability of the cave bear mito-

chondrial genome is expected to provide a better template for exploring very

ancient bear species.

21.3 Genetic Diversity Among Chauvet-Pont d’Arc Cave Bears

We explored the genetic diversity of cave bears from Chauvet-Pont d’Arc Cave by

analyzing several samples from the cave. For comparison purposes, we turned to

another cave from the same area, the Deux-Ouvertures Cave. This cave is located

by the end of the Ardeche Gorge, approximately 15 km away from Chauvet-Pont

d’Arc, and displays rock art pictures. It also contains numerous cave bear remains,

and except for Chauvet-Pont d’Arc, is the most striking bears cave in the area.

We collected 39 and 17 samples from Chauvet-Pont d’Arc and Deux-Ouvertures

caves, respectively. DNA was extracted, and we attempted to amplify a 117 bp

fragment of the mitochondrial genome control region.

Most of the Chauvet-Pont d’Arc cave samples (32/39) and some of the Deux-

Ouvertures cave ones (3/17) failed to yield the queried fragment. We conclude that

this fragment was no longer present or that the samples contain too much PCR

inhibitory compounds for being successfully amplified.

The samples that gave positive results belong to the same haplogroup (haplo-

group B) and to two different haplotypes, which we named HT1 and HT2. HT1 is

also found in Scladina (AY149268, AY149267) and Gigny (AY149264) Caves

(Orlando et al. 2002). HT2 differs from HT1 only in the position 16,550 and is

found in the Cova-Linares Cave (AY149271, AY149272) (Loreille et al. 2001). It is

not surprising to find the B haplogroup in these two caves since it is widely spread

throughout Western Europe.

HT1 and HT2 were both found in Chauvet-Pont d’Arc: two samples in Chauvet-

Pont d’Arc Cave displayed the HT2 haplotype (US08 and US21); the five samples

that yielded the HT1 haplotype are US17, US18, US19, US34, and US39. On the

other hand, all Deux-Ouvertures Cave samples gave the same haplotype, HT1. In

order to verify that this homogeneity is not due to a biased sampling with different

bones belonging to the same individual, we sampled five humerus from five

different individuals. We obtained the HT1 sequence for each of them, validating

that HT1 is widely spread in this cave.

Thus, we observed a high genetic homogeneity inside the bear population of

each cave, as well as from one cave to another. This evidences the frequent female

genetic exchange along Ardeche Gorge and contrasts with the highly subdivided

cave bear population hypothesis (Hofreiter et al. 2002, 2007).


In the same time, several Chauvet-Pont d’Arc samples were dated and returned

radiocarbon age between 37,300 � 340 years BP and 29,560 � 160 years BP.

Most of them range from 30,000 to 32,000 years BP, indicating that cave bears

were present at Chauvet-Pont d’Arc for a relatively brief period of time. It is worth

noting that Scladina and Cova-Linares samples which belong to the HT1 and HT2

haplotypes display approximately the same age as the Chauvet-Pont d’Arc samples.

Scladina’s bones belong to an archeological layer estimated to 40,000–45,000

years, and Cova-Linares’ ones are from a 35,000-year-old layer.

In conclusion, the genetic studies carried out in Chauvet-Pont d’Arc provided a

complete mitochondrial genome for the extinct cave bear, which enabled us to

obtain robust phylogenetic trees for Ursidae. The amount of data also offers the

opportunity of evaluating the divergence dates between species and to compare

genetic and paleontological results. Widening our studies to several samples from

this cave and another cave allowed us to explore the genetic diversity of the area.

We established that the mitochondrial genetic landscape in two caves 15 km away

from each other in the Ardeche Gorge is almost homogeneous. With other bear

caves along the river, extending such analysis to additional sites may allow to

describe more precisely the genetic pattern of the area.

This study also demonstrates that well-preserved DNA still remains in the

Chauvet-Pont d’Arc Cave and establishes this painted cave as a reservoir for

ancient DNA researches. Other species from the Chauvet-Pont d’Arc Cave can

now be analyzed to better characterize the Pleistocene environment.

Reference

Argant A (2001) Los antepasados del oso de las cavernas. Cad Lab Xeol Laxe 26:9

Arnason U, Adegoke JA, Bodin K, Born EW, Esa YB, Gullberg A, Nilsson M, Short RV, Xu X,

Janke A (2002) Mammalian mitogenomic relationships and the root of the eutherian tree. Proc

Natl Acad Sci USA 99:8151–8156

Baffier D (2005) La Grotte Chauvet: conservation d’un patrimoine. Bulletin de la societe pre-

historique francaise 102:11–16

Bon C, Caudy N, de Dieuleveult M, Fosse P, Philippe M, Maksud F, Beraud-Colomb E, Bouzaid E,

Kefi R, Laugier C, Rousseau B, Casane D, van der Plicht J, Elalouf JM (2008) Deciphering the

complete mitochondrial genome and phylogeny of the extinct cave bear in the paleolithic

painted cave of Chauvet. Proc Natl Acad Sci USA 105:17447–17452

Delisle I, Strobeck C (2002) Conserved primers for rapid sequencing of the complete mitochon-

drial genome from carnivores, applied to three species of bears. Mol Biol Evol 19:357–361

Drummond AJ, Rambaut A (2007) BEAST: Bayesian evolutionary analysis by sampling trees.

BMC Evol Biol 7:214

Drummond AJ, Rambaut A, Shapiro B, Pybus OG (2005) Bayesian coalescent inference of past

population dynamics from molecular sequences. Mol Biol Evol 22:1185–1192

Fosse P, Philippe M (2005) La faune de la grotte Chauvet: paleobiologie et anthropozoologie.

Bulletin de la societe prehistorique francaise 102:89–102

Fourment M, Gibbs MJ (2006) PATRISTIC: a program for calculating patristic distances and

graphically comparing the components of genetic change. BMC Evol Biol 6:1


Garcia MA (2005) Ichnologie generale de la grotte Chauvet. Bulletin de la societe prehistorique

francaise 102:103–108

Geneste JM (2005) L’archeologie des vestiges materiels dans la grotte Chauvet-Pont-d’Arc.

Bulletin de la societe prehistorique francaise 102:135–144

Guindon S, Gascuel O (2003) A simple, fast, and accurate algorithm to estimate large phylogenies

by maximum likelihood. Syst Biol 52:696–704

Hanni C, Laudet V, Stehelin D, Taberlet P (1994) Tracking the origins of the cave bear (Ursus

spelaeus) by mitochondrial DNA sequencing. Proc Natl Acad Sci USA 91:12336–12340

Hofreiter M, Jaenicke V, Serre D, von Haeseler A, Paabo S (2001a) DNA sequences from multiple

amplifications reveal artifacts induced by cytosine deamination in ancient DNA. Nucleic Acids

Res 29:4793–4799

Hofreiter M, Serre D, Poinar HN, Kuch M, Paabo S (2001b) Ancient DNA. Nat Rev Genet

2:353–359

Hofreiter M, Capelli C, Krings M, Waits L, Conard N, Munzel S, Rabeder G, Nagel D, Paunovic M,

Jambresic G, Meyer S, Weiss G, Paabo S (2002) Ancient DNA analyses reveal high mitochon-

drial DNA sequence diversity and parallel morphological evolution of late pleistocene cave

bears. Mol Biol Evol 19:1244–1250

Hofreiter M, Munzel S, Conard NJ, Pollack J, Slatkin M, Weiss G, Paabo S (2007) Sudden

replacement of cave bear mitochondrial DNA in the late Pleistocene. Curr Biol 17:R122–R123

Knapp M, Rohland N, Weinstock J, Baryshnikov G, Sher A, Nagel D, Rabeder G, Pinhasi R,

Schmidt HA, Hofreiter M (2009) First DNA sequences from Asian cave bear fossils reveal

deep divergences and complex phylogeographic patterns. Mol Ecol 18:1225–1238

Krause J, Unger T, Nocon A, Malaspinas AS, Kolokotronis SO, Stiller M, Soibelzon L, Spriggs H,

Dear PH, Briggs AW, Bray SC, O’Brien SJ, Rabeder G, Matheus P, Cooper A, Slatkin M,

Paabo S, Hofreiter M (2008) Mitochondrial genomes reveal an explosive radiation of extinct

and extant bears near the Miocene–Pliocene boundary. BMC Evol Biol 8:220

Kurten B (1976) The cave bear story: life and death of a vanished animal. Columbia University

Press, New York

Loreille O, Orlando L, Patou-Mathis M, Philippe M, Taberlet P, Hanni C (2001) Ancient DNA

analysis reveals divergence of the cave bear, Ursus spelaeus, and brown bear, Ursus arctos,

lineages. Curr Biol 11:200–203

Mazza P, Rustioni M (1994) On the phylogeny of Eurasian bears. Palaeontographica 230:38

Orlando L, Bonjean D, Bocherens H, Thenot A, Argant A, Otte M, Hanni C (2002) Ancient DNA

and the population genetics of cave bears (Ursus spelaeus) through space and time. Mol Biol

Evol 19:1920–1933

Paabo S, Poinar H, Serre D, Jaenicke-Despres V, Hebler J, Rohland N, KuchM,Krause J, Vigilant L,

Hofreiter M (2004) Genetic analyses from ancient DNA. Annu Rev Genet 38:645–679

Pacher M, Stuart AJ (2009) Extinction chronology and palaeobiology of the cave bear (Ursus

spelaeus). Boreas 38:189–206

Pages M, Calvignac S, Klein C, Paris M, Hughes S, Hanni C (2008) Combined analysis of fourteen

nuclear genes refines the Ursidae phylogeny. Mol Phylogenet Evol 47:73–83

Peigne S, Goillot C, Germonpre M, Blondel C, Bignon O, Merceron G (2009) Predormancy

omnivory in European cave bears evidenced by a dental microwear analysis of Ursus spelaeus

from Goyet, Belgium. Proc Natl Acad Sci USA 106:15390–15393

Richards MP, Pacher M, Stiller M, Quiles J, Hofreiter M, Constantin S, Zilhao J, Trinkaus E

(2008) Isotopic evidence for omnivory among European cave bears: late pleistocene Ursus

spelaeus from the Pestera cu Oase, Romania. Proc Natl Acad Sci USA 105:600–604

Rohland N, Siedel H, Hofreiter M (2004) Nondestructive DNA extraction method for mitochon-

drial DNA analyses of museum specimens. Biotechniques 36(814–816):818–821

Rohland N, Malaspinas AS, Pollack JL, Slatkin M, Matheus P, Hofreiter M (2007) Proboscidean

mitogenomics: chronology and mode of elephant evolution using mastodon as outgroup. PLoS

Biol 5:e207


Taberlet P, Bouvet J (1994) Mitochondrial DNA polymorphism, phylogeography, and conserva-

tion genetics of the brown bear Ursus arctos in Europe. Proc Biol Sci 255:195–200

Talbot SL, Shields GF (1996a) A phylogeny of the bears (Ursidae) inferred from complete

sequences of three mitochondrial genes. Mol Phylogenet Evol 5:567–575

Talbot SL, Shields GF (1996b) Phylogeography of brown bears (Ursus arctos) of Alaska and

paraphyly within the Ursidae. Mol Phylogenet Evol 5:477–494

Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4: Molecular Evolutionary Genetics Analysis

(MEGA) software version 4.0. Mol Biol Evol 24:1596–1599

Valladas H, Clottes J, Geneste JM, Garcia MA, Arnold M, Cachier H, Tisnerat-Laborde N (2001)

Palaeolithic paintings. Evolution of prehistoric cave art. Nature 413:479

Waits LP, Sullivan J, O’Brien SJ, Ward RH (1999) Rapid radiation events in the family Ursidae

indicated by likelihood phylogenetic estimation from multiple fragments of mtDNA. Mol

Phylogenet Evol 13:82–92

Yu L, Li QW, Ryder OA, Zhang YP (2004) Phylogeny of the bears (Ursidae) based on nuclear and

mitochondrial genes. Mol Phylogenet Evol 32:480–494

Yu L, Li YW, Ryder OA, Zhang YP (2007) Analysis of complete mitochondrial genome

sequences increases phylogenetic resolution of bears (Ursidae), a mammalian family that

experienced rapid speciation. BMC Evol Biol 7:198


Index

A

Accessory, 250, 251, 254, 257–260, 262

Actinobacteria, 303

Actinorhizal plants, 303

Adaptations, 8, 50, 53, 60, 82, 83, 95, 96

Adaption, 82, 84–90, 95

Adaptive radiation, 13, 283–297

Aeschynomene, 303Ag–NOR staining, 10

Agrobacteriumradiobacter, 309rhizogenes, 309tumefaciens, 306, 309vitis, 309

Allopatry, 50

Alpha, 119

Alpha-lactalbumin, 118, 121, 127

Alternative splicing, 31, 38

Amazon, 284, 289–293, 296, 297

Amines, 260

Amniotes, 3, 4, 6, 7, 12, 13

Ancestral area, 289, 290, 292

Ancestral karyotype, 144–146, 153

Ancient DNA, 346–348, 354

Andes, 285, 290, 291, 293

Anesthetic, 253, 255, 260–262

Antarctic fur seal (Arctocephalus gazella), 127Antennal modification

antennal hammer, 271–280

Anticoagulant, 255, 259–262

Aphid

Acyrthosiphon pisum, 133–136Aphis gossypii, 133, 134, 137Myzus persicae, 133, 134, 137

Apparatus, 250, 253, 257, 258

Appressorium ascospores, 319, 324

Area cladogram, 288, 289, 291, 292

Aromatase, 7

Ascoviruses

Diadromus pulchellus, 238, 244, 245Heliothis virescens, 238Spodoptera frugiperda, 237, 238Trichoplusia ni, 238

Azoarcus, 302Azolla, 303Azorhizobium, 301, 306

B

Background selection, 9

Baculoviruses, 230, 232, 233, 236

Bats, 283–297

Bayesian, 10, 12

Bayesian inference, 285

Bdelloid rotifers, 104

Behavior, 283–297

Beta, 119, 120

Beta-lactoglobin, 121

Biased incrementalism, 91–93, 95

Birth and death model, 31, 35, 40

BLAST, 192

Bootstrap, 107

Bovine (bos Taurus), 116, 127, 128Bracoviruses

Chelonus inanitus, 236Cotesia congregata, 235, 236Glyptapanteles flavicoxis, 235Glyptapanteles indiensis, 235

Bradyrhizobiumcanariense, 306japonicum, 306

357

Brown bear, 344, 346, 347, 349, 351

Buccinidae, 253, 254, 258

Buccinids, 254, 258, 260

C

California sea lion (Zalophuscalifornianus), 127

Cancellariid, 256, 262

Cancellariidae, 250, 255, 256, 259, 262

Cancellarioidea, 250, 252, 256

Cape fur seal (Arctocephalus pusillus),126, 127

Caseins, 118–122, 127, 128

Cave bear, 343–354

C-banding, 10

Charnov–Bull hypothesis, 7

Chauvet-Pont d’Arc, 343–354

Chdl, 285Chemical alterations, 347

Choline, 259, 260

Chromogens, 260

Chromosomal inversions, 52, 55

Chromosomal rearrangements, 51, 52, 55,

58, 59, 61

Chromosomal theory of speciation, 51, 61

Chromosome rearrangements, 55

CNGs. See Conserved nongenic sequences

Codon reassignments

ambiguous intermediate mechanism,

86–90

codon capture mechanism, 86, 87

Coevolution, 302

Colinearity, 55

Colubrariidae, 255, 262

Columbellidae, 254, 258

Comparative analysis

CAIC, 277

Comparative genomics, 10, 19–20, 25, 26,

29, 31, 40, 41

Complexity hypothesis, 102

Concerted evolution, 203–204

Conidae, 250

Connectivity analysis, 106, 108, 109, 111

Conoidea, 250, 252, 253, 257, 259, 261

Conopeptides, 259

Conotoxins, 251, 252, 257–261, 263

Conserved nongenic sequences (CNGs), 191

Constraints, 19–41

Convergence, 5

Convergent evolution, 302, 317–326

Coralliophilinae, 253, 255, 256, 261, 262

Corallivory, 254, 256, 262

Costellariidae, 259

Cot curve analysis, 188

Cow, 117, 121, 128

Cryptinae, 273–275, 278, 279

Cryptosporidium, 107, 109Cyanobacteria, 302, 303

Cycads, 302

Cytb, 285

D

Dby, 285Deletions, 55, 56, 58

Deux-Ouvertures Cave, 353

Developmental biology, 161

Diatoms, 107, 108, 110

Diclidurini, 284, 286, 287, 289, 293

Divergence times, 351, 352

Diversity, 252, 253, 256, 263, 264

Dmrt1, 10Dobzhansky, T., 50–52, 59

Dosage compensation, 12

Dosage sensitivity, 201

Drug targets, 106–110

Duplication

Genome duplication, 134

Lineage specific duplications, 138

Paralogs, 133, 136

E

Early lactation protein (ELP), 123

Ear morphology, 295–297

Echidnas (Tachyglossus and zaglosus), 116Echolocation, 295–297

E.C. number, 105, 109

Ecotones, 334, 335

Efficiency of sporulation, 54, 56

ELP. See Early lactation protein

EM. See Error minimization

Emballonuridae, 284, 289, 293

Embryos, 3, 7, 8

Emergence, 81–96

Endoparasitic wasps

Braconidae

Chelonus inanitus, 236Cotesia congregata, 235, 236Cotesia marginiventris, 234, 244Glyptapanteles flavicoxis, 235Glyptapanteles indiensis, 235Microplitis croceipes, 234

Ichneumonidae

Campoletis sonorensis, 230, 241, 244Cardiochiles nigriceps, 230Eiphosoma vitticolle, 237, 239Hyposoter didymator, 236

358 Index

Hyposoter fugitivus, 240Venturia canescens, 230

Endosymbiont

bacteria, 212

eukaryote, 209

facultative, 210

obligate, 210

primary, 210

reproductive, 211

secondary, 210

Endosymbiosis, 103, 104, 108

Enrichment analysis, 106, 111

ENU mutagenesis, 202

Environmental stress, 54

Enzymes, 103–109, 111

Epistasis, 36, 52

Ergalataxinae, 253, 256

Error minimization (EM), 83–91

Esters, 259, 260

Estrogen, 6

Eukaryotes, 102–108, 111

Eumycetes and Fungi

Botrytis cinerea, 318, 322Magnaporthe grisea, 322Neurospora crassa, 320, 322Penicillium chrysogenum, 322, 324–326Podospora anserina, 320, 322Rhizopus oryzae, 323–326Trichoderma reesei, 320, 322Trichoderma species, 326

Eutheria (eutherian or placentalia), 116

Evolution, 249–265

convergent, 182

divergent, 182

Evolutionary breakpoints, 144, 147–150

Evolutionary constraints, 190, 194, 200

Evolutionary rates

Divergence time, 144

Mutations, 133

Omega ratio (dN/dS), 134, 137–140

Synonymous non-substitution rate (dN),

134, 135, 137

Synonymous substitution rate (dS), 134,

135, 137

Evolvability, 95

Exogenes, 263, 265

Exons, 26, 38

Extinction, 9

Eye

camera, 182–185

compound, 181–183

mirror, 182

pinhole, 182

F

Fadrozole, 6

Fasciolariidae, 254, 258

Feeding, 250, 252–256, 262

Fitness change, 75–77

Fitness landscape, 33, 34, 36

Fluorescent in situ hybridization (FISH),

10, 11

Forest, 294, 296, 297

Functional constraints, 200–203

G

Gene architecture, 26

Gene-conversion, 203–204

Gene duplication, 29, 31

Gene expression, 160, 163, 171

Gene identity intervals

interspecies, 308

intraspecies, 308

Gene markers

dnaJ, 308dnaK, 309rpoB, 308, 309

Genes, 253, 261–263, 265

Genetic code

adaptive code hypothesis, 84–90

emergence hypothesis, 90–91

Genetic code evolution, 85, 90, 91

Genetic diversity, 353–354

Gene transfer

lateral, horizontal, 232

Genic theory of speciation, 51

Genome architecture, 19, 20, 23, 26–29, 35,

37, 38, 40

Genome 10K, 13

Genome sequence, 19, 23

Genomic, 56, 58

Genomic rearrangements, 51, 52, 55–61

Genomic structure, 188–190

Genotype � environment, 8–9

Gland, 250, 251, 257–260, 262

Goats, 121, 123, 128

Grey seal (Halichoerus grypus), 127Guiana Shield, 291, 297

Gunnera, 303

H

Haematophagous, 255, 257, 259, 262

Haematophagy, 254–256

Haplogroup, 348, 353

Haplotypes, 348, 353, 354

Harbour seal (Phoca vitulina), 127Harpidae, 259

Index 359

Harpooning, 253

Hemiplasy, 144, 150–154

Herbaspirillum, 302Heterogamety, 4–8, 10–12

Heteromorphic sex chromosomes, 4, 9

Hill–Robertson effect, 9

Histamine, 259

Historical biogeography, 284, 287–293

Hitchhiking, 9

Homoplasy, 144, 151, 153

Horizontal gene transfer (HGT), 101–104,

106–109, 111

Horizontal transfer, 202–204

Host location, 272–274, 279, 280

Hosts, 272–274, 278–280

Human chromosome 2, 195

Human chromosome 21, 191, 194, 195, 201

Hybrid fertility, 55, 57, 58

Hybridization, 4, 10

Hypobranchial gland, 251, 260

Hypolimnas bolina

Hypolimnas bolinaresistance, 221

I

Ichneumonidae, 271–273

Ichnoviruses

Campoletis sonorensis, 240–244Cardiochiles nigriceps, 230Hyposoter fugitivus, 240Tranosema rostrales, 240

Immunosuppressive genes

Imd, 232

Toll, 232

Inactivation, 12

Incipient, 50, 56, 58, 60

Incubation, 4–8

Insertions, 55, 56

Interaction, 8–9

Introns, 21, 23, 25, 26, 38, 39

Inversions, 52, 55, 56

Iridoviruses

Chilo suppressalis, 237Isolation, 49–61

J

Junk DNA, 190

K

Kappa, 119, 120

Karyotype, 4, 5

KEGG, 106, 111

L

Lactotransferin, 121

LALBA, 127

Lateral transfer, 304, 306

Legume plants

Phaseolus vulgaris, 306Leishmania, 107, 108, 111Lepidopterans

Chilo suppressalis, 237Ephestia kuehniella, 230Heliothis armigera, 237Heliothis zea, 233Spodoptera frugiperda, 234Trichoplusia ni, 235

Likelihood, 10

LINEs. See Long interspersed elements

Lipopolysaccharides, 111

LLP-A, 123

LLP-B, 123

Long conserved noncoding sequences

(LNCS), 192

Long interspersed elements (LINEs), 189

M

Mammaliaforms, 116

Mammals, 116–122, 124, 126–129

Marginellid, 254

Marginellidae, 255, 256, 259

Markov-chain Monte Carlo, 12

McDonald–Kreitman test, 23, 24

Melongenidae, 254, 258

Melospiza melodia, 331, 332, 340Mesorhizobium

amorphae, 306loti, 303

Metabolic enzymes, 103, 104, 111

Metatheria (marsupials or Marsupialia), 116

metaTIGER, 104–111

Methylobacterium, 306

Microarray

interspecies array, 183, 184

Microevolution, 8

Migration, 9

Milk proteins, 116–119, 122–128

Minimal gene set, 29, 30

Miocene, 287, 288, 290–292, 295, 297

Misfolding, 33, 34, 40

Mismatch repair, 51, 55

Mitochondria, 103, 104

Mitochondrial genome, 346–354

Mitridae, 253, 259, 260

Molecular dating, 284, 287, 288

360 Index

Molecular evolution, 67, 68, 78

Molluscs

cephalopod, 182–184

nautilus, 182–185

octopus, 182–184

pectin, 182–185

squid, 182–185

Morphogenetic gradient

dorsal gradient, 162, 163, 167, 169–171

dpp gradient, 168, 171

gradient, 160, 164, 166, 167, 169–172

Morphology, 283–297

Mouse chromosome 2, 195

Muller’s Ratchet, 9, 10

Muricidae, 253, 255, 256, 259, 261

Muricids, 254, 259, 260

Muricoidea, 250, 252

Mutation, 188, 192, 194, 200–202

beneficial, 69, 75–78

deleterious, 69, 75–77

neutral, 75

Mutational cold spot, 201–202

Mutational load, 55

Mutation robustness

error minimization (EM), 83, 90, 91

extrinsic, 94

intrinsic, 94, 95

Mutation-selection equilibrium, 73, 75, 77

Mycorrhizal symbiosis, 304

N

NADPH oxidase, 320, 321, 325

Nassariid, 254

Nassariidae, 254, 258, 259

Natural science,

Natural selection, 82–84, 91–96

Neotropics, 283–297

Nervous system

neural, 159–167, 172

neuroblast, 164–167, 172

Networks, 20, 26, 31–35, 37, 40

Neurotoxins, 250, 251, 258, 260–262

Neutral networks, 91, 93–95

New World emballonurid bats, 284–288,

290, 292, 294–296

Nitrogen fixation, 302, 304, 306

Nodulation factors

nodB, 303nodC, 303

Noncoding sequences, 20, 23

Nonorthologous gene displacement, 29, 31

Nonsynonymous substitutions, 21

Northern Amazon, 289–292, 296

Nudiviruses, 231, 233–236, 244

O

Odobenids, 125

Oligocene, 287, 288, 295, 296

Olividae, 259

One-band-one-gene hypothesis, 188

Operons, 23, 27, 28

Organelles

immunosuppressive, 229–245

Origin of life, 67, 68

Ortholog, 134, 137

Orthologous, 28–32, 35, 38

Ostreococcus, 107, 111Otariids (sea lions, fur seals), 125, 127

Oviparity, 12, 13

P

Paleolithic, 343–354

Pan-genome, 102

Paralogs, 29, 31, 35, 40

Parsimony, 10, 285, 286, 293, 294

Particles

immunosuppressive, 231–234

Patterning, 159–172

Pelage, 294–295

Peptides, 252, 253, 257, 258, 261, 263, 264

Phenomic, 19, 32–36, 39, 40

Phocids (true seals), 125, 127

Photoreceptors, 181, 183

Phylogenetic trees, 101–112

Phylogenies, 304, 306–310

Phylogeny, 284, 285, 287, 293–296, 349–353

Phytophthora, 107, 109, 111Pinniped, 125–126

Plasmodium, 107, 108Plasticity, 19–41

Plastids, 103, 104, 107–109, 111

Platypus (Ornithorhynchus anatinus), 116,118, 119, 121

Pleiotropy, 36

Pleistocene, 287, 290, 291, 346, 352, 354

Pleistocene refugia, 335, 336

Pleistocene refugium, 336, 337

Pliocene, 283, 287, 291

Polydnaviruses, 232–234, 236, 241, 243

Polygenic inheritance, 8

Polymorphisms, 151–153

Positive selection, 21–25

Poxviruses

Diachasmimorpha longicaudatapoxvirus, 244

Preferential attachment, 91, 92

Prezygotic, 50, 61

Prialt, 252, 260

PRIAM, 105, 106

Index 361

Primary, 250, 257–260, 262

Production, 257, 258, 260, 262

Profiling, 263–265

Prokaryotes, 23, 27–30, 36, 37, 39,

102–103, 107, 108, 111

Promiscuous domains, 26

Proteomics, 264, 265

Prototheria (monotreme or Monotrema), 116

Pseudaptation, 81–96

Pseudogenes, 21, 23

PSI-BLAST, 105, 106

PTMP-1, 123

PTMP-2, 123

Q

Quasispecies, 68, 72–74, 78

R

RAC 2 (myoblast fusion),

Radiation, 351

Radiocarbon age, 345, 354

Radula, 250, 253–257

Rearrangements, 51–53, 55–61

Reciprocical Best Hit, 133, 136

Recombination, 9, 12

Red kangaroo (macropus rufus), 122Regulators, 30, 31

Reinforcing mechanism, 50

Relative reproductive isolation, 50

Repeat masking, 192

Replication, 68, 71, 72, 75–78

Reproductive, 49–61

Reproductive barrier, 55, 56, 58–60

Reproductive isolation, 49–61, 331, 338

Rhizobia, 301–310

RhizobiumR. cellulosilyticum, 309R. daejeonense, 309R. etli, 303, 306R. fabae, 306R. galegae, 309R. huautlense, 309R. leguminosarum, 306R. lusitanum, 309R. mongolense, 309R. pisi, 306R. selenireducens, 309R. tropici, 306

Ringed seal (Pusa hispida), 127Ring species

Ensatina eschscholtzii, 331, 333Glossina morsitans, 331Lanius, 331, 333Melospiza melodia, 331, 332, 340

Phylloscopus trochiloides, 331Phylloscopus trochilus, 331Zosterops, 331

RNA

folding, 69–72

sequence-structure map, 68, 69, 72

world, 67–69, 71, 77

RNA complexity, 188

RNome, 25, 31

Robustness, 20, 32–34, 36–37, 40, 41

Roosts, 284, 293–295

Rot curve analysis, 188

S

Saccharomyces, 107Saccharomyces cerevisiae, 52, 54, 55Salivary glands, 250, 251, 257–260, 262

Savannahs, 288, 296, 297

Scale free networks, 91–92

Scaling, 30, 31, 40

Scaling, size, 160, 167–172

SDs. See Segmental duplications

Secretion, 253, 257–260, 262

Segmental duplications (SDs), 144,

147–150, 153

Selection

Adaptation, 133, 140

Fast-evolving genes, 137, 139–140

Positive selection, 134, 140

Relaxed selection, 134, 137, 140

Selective pressure, 68, 72, 78

Selfish operon, 28

Sequence data

Coding sequence (CDS), 135

Expressed sequence tag (ESTs), 133, 137

Pea aphid genome, 134, 136, 140

Sequences, 10, 11, 13, 14

Sequencing, 116–125, 129

Sex determination, 4–8, 10–13

SHARKhunt, 105, 106

Shell drilling, 254

Shell wedging, 254

Short interspersed elements (SINEs), 189

Signaling pathways

BMP signaling pathway/BMP signaling,

165, 166

SINEs. See Short interspersed elements

Single nucleotide polymorphisms (SNPs), 190,

191, 202

SinorhizobiumS. chiapanecum, 308S. mexicanum, 308S. terangae, 306

SNPs. See Single nucleotide polymorphisms

362 Index

Song Sparrow, 329–340

SOS, 53

South America, 283, 284, 286, 288–291,

293, 296

Spandrel, 82

Speciation, 50–56, 58, 60, 61

allopatric, 330, 335

ecological, 334, 338–340

Species, 49–58, 60, 61

Sporulation efficiency, 54, 56–61

16S rRNA, 308–310

Sry, 9

Starvation, 49–61

Stochastic approaches, 10

Subspecies, 330, 331, 333, 335, 338, 339

Symbiogenesis

genome fusion, 233

Symbiosis, 234, 301, 302, 304

Symbiotic genes

nif, 304nodA, 303–305nodABC, 303nodB, 303nodBC, 303nodC, 303

Symbiotic islands, 306

Symbiotic plasmids, 306, 307

repA, 307repABC, 307repB, 307repC, 307

Synaptid, 116

Synonymous sites, 21, 22, 23, 25, 26, 32, 33

Synonymous substitutions, 21

Syntenies, 144, 145, 151, 152

T

Tandem repeats, 147, 148

Taxon-pulse, 291, 292, 297

Terebridae, 250, 264

Terebrids, 253, 258

Teretoxins, 258, 264

TEs. See Transposable elements

Testosterone, 6

Tetramine, 258

Tetraploid, 51, 59, 60

Tetraploidization, 51, 52, 55, 59

Tetraspanin, 320, 321

Theileria, 107, 108Therapsid, 116

Theria, 116

Toxins, 250–252, 257–259, 261,

262–263, 265

Toxoplasma, 107, 108

Transferomics, 101–112

Translocations, 55, 56

Transposable elements (TEs), 25, 147,

149–150

Transpositions, 53, 55, 56

Trichosurin, 123

Trypanosoma, 107, 108Turrids, 250, 253, 258

Turritoxins, 258

U

UCEs. See Ultraconserved elements

Ultraconserved elements (UCEs), 191,

194–195, 201–204

Underdominance, 51, 52

Ursidae, 348, 349, 351, 352, 354Ursinae, 349, 351, 352Ursus spelaeus, 344, 346, 349–351, 353Usp9x, 285, 293

V

Venom, 250, 253, 257–259, 261–265

Vibrational sounding, 273, 279

Viviparity, 4, 12, 13

Volutidae, 253, 259, 260

Volutomitridae, 259

W

Wallaby (macropus eugenii), 122–124, 127Walrus, 125

WAP. See Whey acidic protein

WDC2, 121

Whey acidic protein (WAP), 118, 121, 122,

123, 128

Whole-genome, 13

Wing sac, 294, 295

Within-area specification events, 290

Wolbachiacytoplasmic incompatibility (CI), 211, 215,

217, 220

male-killing (MK), 209–222

supergroup, 210, 211

transmission, 211, 216

wBol1, 212–216, 218–221wBol2, 214, 215, 217wPip, 220

Wood boring beetles

Wood-boring, 273, 277

X

X chromosome, 5, 12

Y

Y chromosome, 195, 201

Index 363

Comparative genomics and transcriptomics of lineages I, II, and III strains of Listeria...

Documents

Transcript of Comparative genomics and transcriptomics of lineages I, II, and III strains of Listeria...