ORIGINAL PAPER

The complete mitochondrial genome of the Antarctic sea spiderAmmothea carolinensis (Chelicerata; Pycnogonida)

Antonio Carapelli • Giulia Torricelli •

Francesco Nardi • Francesco Frati

Received: 9 October 2012 / Revised: 16 December 2012 / Accepted: 10 January 2013

� Springer-Verlag Berlin Heidelberg 2013

Abstract Mitochondria are responsible for the oxidative

phosphorylation process. Accordingly, putatively adaptive

changes in their genomic features have been variously

associated with major eco-physiological shifts in animal

evolution, including increased metabolic rates and heat

adaptation. Antarctic pycnogonids offer an interesting

system to test whether the selective pressure for heat pro-

duction and increased aerobic metabolism may be driving

genomic changes like: (a) unusual compositional biases at

the nucleotide and amino acid level, possibly related to

cold adaptation; (b) an accelerated rate of mutations/

genomic rearrangements, possibly related to the mutagenic

effects of oxygen intermediates. The complete mitochon-

drial genome (mtDNA) of the Antarctic sea spider

Ammothea carolinensis Leach, 1814 (Arthropoda: Pycno-

gonida), the type species for the genus Ammothea, has been

determined and is here compared to known genomes from

Antarctic and temperate species. We describe a marked

heterogeneity in base composition skewness parameters as

well as a strong signature of purifying selection toward an

increase in thymines at second codon positions, possibly

associated with an increased stability of hydrophobic inter-

membrane domains. We further observe a fairly high rate

of genomic changes, including a possible hot spot of

recombination at the level of tRNA-Q. Nevertheless, these

features do not seem to be restricted to the two Antarctic

pycnogonids analyzed, as to suggest a causal relationship

between cold adaptation and genomic changes, and are

better interpreted as basal features shared by the entire

group. The relevance of the newly determined sequence for

the phylogeny of pycnogonids, including its base compo-

sition and genomic rearrangements, is further discussed.

Keywords Pycnogonida �Mitochondrial DNA (mtDNA) �Molecular evolution � Antarctica � Nucleotide bias

Introduction

The mitochondrial genome of animals is almost invariably

a circular double-stranded molecule of 11–20 kb in size. It

includes the genes encoding for 13 protein products

involved in the oxidative phosphorylation (OXPHOS)

process, 22 transfer RNA (trnX), and two ribosomal rRNAs

(rrnS and rrnL), as well as a noncoding region (known as

AT-rich region in arthropods) where the signals for the

initiation of replication and transcription are located. Its

high-nucleotide substitution rate and variable gene order

combined with conservative gene content (Castellana et al.

2011), as well as the strict matrilinear inheritance and lack

of introns, made mtDNA one of the most popular markers

in phylogenetic studies, only paralleled by nuclear rRNA

markers. In addition, mitochondrial sequences are widely

used to identify unknown species (i.e., cytochrome oxidase

c subunit I in barcoding) and for population genetic studies

(Min and Hickey 2007), while the arrangement of genes

(gene order) within the organelle’s genome has been used

to infer deep phylogenetic relationships. The recent avail-

ability of large nuclear datasets for phylogenomic studies

(Regier et al. 2010; Rota-Stabelli et al. 2010; Rehm et al.

2011; Giribet and Edgecombe 2012), together with the

recognition that highly divergent rates of molecular

Electronic supplementary material The online version of thisarticle (doi:10.1007/s00300-013-1288-6) contains supplementarymaterial, which is available to authorized users.

A. Carapelli (&) � G. Torricelli � F. Nardi � F. Frati

Department of Life Sciences, University of Siena,

Via Aldo Moro 2, 53100 Siena, Italy

e-mail: antonio.carapelli@unisi.it

123

Polar Biol

DOI 10.1007/s00300-013-1288-6

evolution (in primary sequence as well as gene order) may

affect specific taxa (Gissi et al. 2000; Hassanin 2006) with

undesirable outcomes on the analysis, has lowered the

interest in mitochondrial genomes as a phylogenetic mar-

ker, although attempts to improve data analysis are ongoing

(Carapelli et al. 2007; Gissi et al. 2008; Talavera and Vila

2011; Nardi et al. 2012a). Nevertheless, the accumulation

of over 2800 complete animal mitochondrial genomes from

a variety of different invertebrate and vertebrata taxa has

opened to the possibility of performing comparative studies

on the molecular features, functionality, and evolution of

the mitochondrial genome on a large scale, so far impos-

sible for other genomic compartments (nuclear and plas-

tidial). Key transitions of animal evolution (i.e., origin of

multicellularity and bilaterality) are associated with sub-

stantial modifications of the mitochondrial DNA architec-

ture (Lavrov 2011). Some of these changes represent

interesting features to be investigated in order to under-

stand the molecular evolution of the organelle genome. In

this respect, the mechanisms that lead to gene rearrange-

ments have been hypothesized following the discovery of

peculiar mtDNA organizations (Lavrov et al. 2002; San

Mauro et al. 2006; Juhling et al. 2012; Nardi et al. 2012b).

In addition, some posttranscriptional modifications (non-

templated nucleotide addition and insertion, which com-

pensate for defects in gene sequence and restore the

functionality of tRNAs), once believed to be exclusive of

plant mitochondrial and plastidial DNAs, have been

observed also in animals (Lavrov et al. 2002; Segovia et al.

2011). Eventually, sex-linked mtDNA genomes have been

discovered in some bivalves highlighting a transmission

route (Doubly uniparental inheritance) other than orthodox

matrilinear inheritance (Passamonti et al. 2011).

Six complete or almost complete mtDNAs have been

sequenced for Pycnogonida (Podsiadlowski and Braband

2006; Park et al. 2007; Masta et al. 2010; Dietz et al. 2011),

an enigmatic group of arthropods whose fossil record dates

back to the Upper-Cambrian (&490 Myr ago) (Waloszek

and Dunlop 2002) and that is characterized by a number of

peculiar morphological features that complicate the iden-

tification of their phylogenetic relationships (Budd and

Telford 2005; Maxmen et al. 2005; Edgecombe 2010).

Pycnogonids have been considered, in turn, as the sister

taxon to the chelicerates (Dunlop and Arango 2005) or

basal to all extant arthropods (Giribet et al. 2001). Unfor-

tunately, due to the heterogeneity of nucleotide composi-

tion compared to other arthropods, mitogenomic data failed

to provide a robust phylogenetic hypothesis and all phy-

logenetic reconstructions based on mtDNA recovered a

highly problematic placement of the group as branching

from within the Chelicerate group (mostly as sister group

to Acarida) that would imply a terrestrial origin of the

group (Hassanin 2006; Podsiadlowski and Braband 2006;

Park et al. 2007; Masta et al. 2010). Nevertheless, some

interesting molecular features that may in turn be deployed

as phylogenetic markers at a lower taxonomic level have

been described from the mitochondrial genomes of pyc-

nogonids. An extensive description of some of these

characters has been proposed by Masta et al. (2010),

although others are still being analyzed in more detail.

In this work, we describe the newly determined complete

mitochondrial genome of the Antarctic species Ammothea

carolinensis, provide a description of major molecular fea-

tures of the genome, and compare this with other published

pycnogonid mtDNAs.

Materials and methods

One living specimen of A. carolinensis was sampled off the

Antarctic coast in the vicinity of the Italian Research Station

Mario Zucchelli (Victoria Land) (74�410S, 164�070E), during

the XXI Italian-Antarctic expedition (2005/2006). The

specimen was identified by Dr. L. Dietz and colleagues from

the University of Bochum (Germany) and a small part of a leg

was frozen and preserved at -80 �C for the molecular anal-

yses. The remaining of the DNA extraction and of the ani-

mal’s body are preserved at the Department of Life Sciences

(University of Siena). Total genomic DNA was purified using

the Wizard SV genomic DNA purification system (Promega)

and used in all subsequent amplifications. Partial cox1 and

cob sequences were initially produced using primer pairs

C1-J-1718 (50-GGAGGATTTGGAAATTGATTAGTTCC-

30)/C1-N-2191 (50-CCCGGTAAAATTAAAATATAAAC

TTC-30) and CB-J-10933 (50-TATGTACTACCATGAGG

ACAAATATC)/CB-N-11367 (50-ATTACACCTCCTAA

TTTATTAGGAAT-30 (Simon et al. 1994) and standard

amplification procedures. Four specific primers were then

designed on these sequences to amplify the rest of the genome

in two segments of about 9 and 6 Kb, respectively: couples

ACA-cox1-1753J (50-GCTCATTCAGGATCCTCAGTAG

ATTTAAC-30)/ACA-cox1-1773N (50-GCCCCTAAAATT

GAAGATACTCCCGC-30) (fragment 1) and ACA-cob-

10682J (50-TTCTCAGTAAATAATGCCACCTTAAAT

CG-30)/ACA-cob-10906N (50-GTTAGCTGGAATGAAG

TTTTCTGGGTC-30) (fragment 2). Long-PCR conditions

were the following: fragment 1) 94 �C for 1 min, 60� for

1 min, and 68� for 12 min, 35 cycles; fragment 2) 94 �C

for 1 min, 50� for 1 min, and 68� for 8 min, 35 cycles.

Amplifications were performed in a 25 ll reaction volume

composed of 10.75 ll of sterilized distilled water, 2.5 ll

of LA PCR Buffer II (Takara), 2.5 ll of 25 mM MgCl2,

4 ll of dNTPs mix, 1.25 ll of each primer (10 lM),

2.5 ll of DNA template, and 0.25 ll (1.25 U) of TaKaRa

LA Taq polymerase (Takara). Long-PCR fragments were

purified using the Kit Wizard SV Gel and PCR Clean-up

Polar Biol

123

System (Promega) and completely sequenced on both

strands using a primer walking approach (primer list

available on request). The final sequence assembly was

produced using Sequencher 4.4.2 (Gene Codes) and

deposited in GenBank under accession number

GU065293.

Protein-coding, rRNA, and tRNA genes were identified

based on their amino acid translation or secondary structure

features, respectively. The AT-rich region was identified as

the single noncoding region of significant length and

according to its peculiar structure and base composition.

Start and stop codons of protein-encoding genes (PCGs)

were identified and gene boundaries were inspected for the

presence of gene overlaps. Secondary structures of tRNAs

were manually reconstructed and rendered using RnaViz

2.0 (de Rijk and de Wachter 1997). Basic sequence sta-

tistics and codon usage were calculated using PAUP* 4b10

(Swofford 2002) and codonw (Peden 1999), respectively.

Strand asymmetry was described using the formulas:

AT-skew = [A %-T %]/[A %?T %] and CG-skew =

[C %-G %]/[C %?G %] (Perna and Kocher 1995; Hass-

anin et al. 2005). The graphical representation of the

percentage of A?C and G?T nucleotides across the whole

J-strand was calculated using the program MacVector�

7.2.3 (Accelrys). The presence of repeated sequences

within the AT-rich region was assessed using the mreps

software (Kolpakov et al. 2003) and associated secondary

structure motifs were reconstructed using the software

Mfold (Zuker 2003) under default settings.

Results

Gene order

The gene order of six of the seven published mtDNA

genomes of pycnogonids is similar—although not identi-

cal—to that of Limulus polyphemus (Lavrov et al. 2000),

presumably the ancestral state for arthropods. In these

genomes, nevertheless, trnQ is never observed in its

putative original position between trnI and trnM (Masta

et al. 2010), suggesting either a ‘‘hot spot’’ for independent

rearrangements or the occurrence of an ancestral translo-

cation followed by independent repositioning in at least

three different lineages (Achelia ? Ammothea ? Tanysty-

lum, Colossendeis, and Nymphon). Notably, the actual

position of trnQ is unknown for two of the three incom-

plete genomes (Ammothea hilgendorfi and Nymphon

unguiculatum, erroneously reported as Colossendeis sp. in

Masta et al. 2010; Dietz et al. 2011), although this infor-

mation could be deduced based on the gene order of the

congeneric species A. carolinensis (Fig. 1) and Nymphon

gracile, where trnQ is located on the N-strand between the

AT-rich region and rrnS and between rrnS and trnN,

respectively. Furthermore, in Colossendeis megalonyx,

trnQ is not observed on the original position, yet it has not

been found in the sequenced fragment of the genome

(Dietz et al. 2011). Interestingly, in the taxa where this

gene has been sequenced, trnQ is always positioned at the

50-end of the rrnS gene, with the only exception of the

highly rearranged N. gracile (where it is found at the 30-end

of rrnS). This latter evidence suggests an ancestral trans-

location of trnQ at the base of the pycnogonid lineage,

either at the 30- or 50-end of rrnS, followed by additional

independent relocation in some lineages. The two Nym-

phon species further share the translocations of trnA, trnN,

and trnSgcu and the switch of position between trnE and

trnR. The observation that in all pycnogonid genomes, the

30-end of cox2 directly abuts the 50-end of cox1 without any

intergenic spacers while the two tRNA genes encoding

Leucine (trnLuag and trnLuaa) are located between rrnL

and nad1 unambiguously supports the placement of the

group outside the pancrustacean lineage (Boore et al.

1998).

Gene initiation and termination, intergenic spacers

Mitochondrial PCGs are notorious for the use of nonstan-

dard initiation and truncated stop codons (Boore 2006;

Carapelli et al. 2008). In A. carolinensis, nine genes start

Fig. 1 The mitochondrial genome organization of A. carolinensis.

Genes for proteins and rRNAs are indicated with standard abbrevi-

ations, whereas those for tRNAs are designated by a single letter for

the corresponding amino acid. Arrows indicate the direction of coding

regions. In black are genes encoded on the J-strand, in gray those with

opposite polarity

Polar Biol

123

with canonical codons (ATG or ATA) encoding for

Methionine, whereas others show alternative codons

(Leucine or Isoleucine): nad2 and nad3 (ATT; Ile), cox1

(TTA; Leu), and nad1 (ACT; Leu) (Table 1). As in the

majority of other arthropods, most protein-encoding genes

(8/13) display truncated stop codons (either T- or TA-). These

are generally believed to be converted in complete TAA

termination codons by posttranscriptional polyadenylation of

mRNAs (Table 1), although EST data in the lepidopteran

species Maruca vitrata (Margam et al. 2011) did not find

evidence for polyadenylation in transcripts with truncated 30-ends.

The overall structure of the genome is very compact,

with frequent overlaps of genes encoded on different

strands (trnW/trnC, trnY/cox1, trnE/trnF, and trnP/nad6).

In two instances (trnK/trnD and atp8/atp6), overlaps are in

Table 1 Annotation, nucleotide composition, and other features of the mitochondrial genome of A. carolinensis

%A %C %G %T Gene length (bp) Positions Strand Start/stop codons

trnI 42.0 7.2 13.0 37.7 69 1–69 J

trnM 34.8 16.7 10.6 37.9 66 70–135 J

nad2 42.1 10.6 3.9 43.4 989 141–1,129 J ATT(I)/TA–

trnW 49.3 7.5 6.0 37.3 67 1,130–1,196 J

trnC 38.2 5.9 5.9 50.0 68 1,189–1,256 N

trnY 42.9 11.1 7.9 38.1 63 1,264–1,326 N

cox1 32.5 14.4 13.7 39.4 1,537 1,323–2,859 J TTA(L)/T–

cox2 39.3 14.6 8.4 37.8 670 2,860–3,529 J ATG(M)/T–

trnK 40.0 12.3 12.3 35.4 65 3,530–3,594 J

trnD 42.6 6.6 8.2 42.6 61 3,593–3,653 J

atp8 45.3 10.7 3.8 40.3 159 3,664–3,822 J ATG(M)/TAA

atp6 38.8 10.6 6.3 44.3 668 3,816–4,483 J ATG(M)/TA–

cox3 33.7 14.6 11.8 39.8 786 4,484–5,269 J ATG(M)/TAA

trnG 45.3 7.8 6.3 40.6 64 5,274–5,337 J

nad3 33.0 14.3 6.6 46.1 349 5,338–5,686 J ATT(I)/T–

trnA 36.8 12.3 12.3 38.6 57 5,687–5,743 J

trnR 38.8 9.0 7.5 44.8 67 5,745–5,811 J

trnN 40.0 12.3 12.3 35.4 65 5,820–5,884 J

trnSgcu 41.1 5.4 7.1 46.4 56 5,885–5,940 J

trnE 46.9 6.3 4.7 42.2 64 5,941–6,004 J

trnF 33.3 16.7 4.2 45.8 72 6,004–6,075 N

nad5 48.9 10.9 6.6 33.6 1,723 6,076–7,798 N ATG(M)/T–

trnH 40.6 13.0 2.9 43.5 69 7,799–7,867 N

nad4 50.7 11.1 7.1 31.1 1,368 7,869–9,236 N ATG(M)/TAA

nad4L 51.0 10.2 3.7 35.0 294 9,237–9,530 N ATG(M)/TAA

trnT 39.7 6.3 9.5 44.4 63 9,533–9,595 J

trnP 41.5 12.3 3.1 43.1 65 9,596–9,660 N

nad6 40.0 10.6 3.8 45.6 502 9,636–10,137 J ATA(M)/T–

cob 35.4 14.2 9.7 40.6 1,132 10,150–11,281 J ATG(M)/T–

trnSuga 45.5 4.5 7.6 42.4 66 11,282–11,347 J

nad1 49.3 15.8 6.6 28.3 936 11,364–12,299 N ACT(L)/TAA

trnLuaa 35.9 21.9 7.8 34.4 64 12,300–12,363 N

trnuag 37.5 17.2 6.3 39.1 64 12,367–12,430 N

rrnL 44.7 12.2 6.2 36.9 1,213 12,431–13,643 N

trnV 41.4 10.0 5.7 42.9 70 13,644–13,713 N

rrnS 43.5 12.7 5.3 38.6 830 13,714–14,543 N

trnQ 43.3 13.4 4.5 38.8 67 14,544–14,610 N

AT-rich 46.8 3.9 2.2 47.0 491 14,611–15,101

J-strand 42.2 12.2 7.4 38.3 1–15,101

Polar Biol

123

turn observed between genes encoded on the same strand.

Whereas the overlap between atp8 and atp6 is a common

feature in arthropod and vertebrate mitochondrial genomes

(Ojala et al. 1981; Krzywinski et al. 2006; Podsiadlowski

et al. 2006), and polycistronic transcripts of 2 or 3 genes

(usually atp8/atp6, but also atp8/atp6/cox3, nad4L/nad4,

and nad6/cob) are sporadically observed in arthropod

mtDNA transcriptomes (Margam et al. 2011), the possi-

bility of alternative splicing may be considered for the pair

trnK/trnD.

Nucleotide composition and codon usage

The nucleotide composition of the A. carolinensis mtDNA

is very similar to A. hilgendorfi, with the highest value of

A?T (80.4 %) and the lowest of G-content (7.4 % on the

J-strand) ever detected for pycnogonid mtDNAs (Table 1).

Total A?C richness of the J-strand, a common feature

of arthropod mtDNAs, is not uniformly distributed across

the genome. While A?C % and G?T % are almost equal

in genes encoded on the J-strand, A?C % is invariably

higher than G?T % (calculated on the J-strand) in genes

encoded on the N-strand (either protein or RNA encoding)

(Fig. 2 and Supplementary material 1). This latter trend,

nevertheless, differs across codon positions, with G�C at

first and third positions and T�A at second positions

(Fig. 2). Skewness parameters calculated for all sea spi-

ders’ mtDNAs (J-strand) are very heterogeneous (Masta

et al. 2010), with negative values of both AT- and

CG-skew observed in Tanystylum orbiculare and N. ungui-

culatum and positive values in A. carolinensis (Table 2),

A. hilgendorfi, and N. gracile. In addition, Achelia bituber-

culata displays a negative AT-skew and a positive CG-skew

(Masta et al. 2010). The A. carolinensis J-strand is extremely

poor in G nucleotides (7.4 %), and it is not surprising that

the highest CG-skew values are observed at third codon

positions of PCGs (both orientations) (Supplementary

material 1).

Base composition asymmetries in the other five com-

plete pycnogonid mtDNAs (excluding N. gracile due to its

highly rearranged gene order) are in line with that observed

in A. carolinensis. The occurrence of a high number of Ts

at second codon positions of both J- and N-oriented genes

is also shared by all pycnogonids with the same gene order,

with AT-skew values ranging from -0.40 (C. megalonyx)

to -0.45 (A. carolinensis).

Strand and nucleotide bias significantly affect the codon

usage of protein-encoding genes in the mtDNA of A. car-

olinensis. In this respect, PCGs encoded on either J- and

N-strand show a remarkably higher frequency of NNA

(44 %) and NNU (40 %) triplets versus NNC (12 %) and

NNG (4 %) ones, in line with the observed A?T nucleo-

tide bias. Additionally, more than half (52 %) of the triplets

from fourfold degenerate codon families of J-strand genes

are represented by NNA triplets, whereas most (57 %) of

those oriented on the opposite strand are NNU. Remark-

ably, due to second codon positions, specific bias of

J-strand PCGs, NUNs, rather than NANs, are the most

employed codons (1,110 vs. 424, respectively).

AT-rich region

The AT-rich region of A. carolinensis (491 bp) is charac-

terized by the presence of four copies of a repeated

sequence (R1–R3 and PR), with three being complete and

identical (R1–R3 = 125 bp, corresponding to positions

14,611–14,735; 14,736–14,860; and 14,861–14,985 in the

annotated genome) and one incomplete and slightly dif-

ferent in primary sequence (PR = 42 bp, positions

14,986–15,027) (Fig. 3). Each complete repeat starts with a

Fig. 2 Graphical representation of AT- and CG-skew values in

different sets of J- and N-oriented protein-encoding sequences

(calculated on the J-strand): all = all positions of PCGs; first, second,

and third positions only

Table 2 Skew values

calculated for different sets of J-

and N-oriented nucleotides

All sites J-strand N-strand

PCGs only Third position only PCGs only Third position only

AT-skew 0.050 -0.060 0.032 -0.220 -0.138

CG-skew 0.241 0.197 0.603 -0.291 -0.645

Polar Biol

123

(AT)5 dinucleotide. At the 30-end of R3, there is a partial

repeat (PR) corresponding to the first 40 bases of the

complete repeats plus an additional ‘‘AT’’ resulting in a

(AT)6 sequence. This repeated region corresponds to the

majority (417/491 bases) of the noncoding sequence and is

followed by a 74-long stretch of non-repeated bases at the

30-end of PR. The whole AT-rich region can be folded into

a 8-stems secondary structure, mostly composed by paired

sequences of the aforementioned repeats (Fig. 3). While

the initial duplication may have originated through differ-

ent processes, including incorrect termination during rep-

lication, the maintenance and amplification of repeats is

probably due to replication slippage (Li et al. 2012). The

primary sequence of A. carolinensis AT-rich region does

not show significant similarity with the three other known

AT-rich regions, from A. bituberculata, N. gracile, and

T. orbiculare, whereas repeated stretches of sequences that

can be folded into secondary structure motifs were also

found in the congeneric A. bituberculata (Park et al. 2007).

Manual search for plausible regulatory sequences, includ-

ing the T-stretches commonly observed in other vertebrate

and arthropod lineages (Clayton 1982; Saito et al. 2005),

was also inconclusive. In addition, sequence length among

AT-rich regions is substantially variable in pycnogonid

mtDNAs, ranging from 191 bp in N. gracile to 977 in

A. bituberculata.

tRNAs

The complete mitochondrial genome of A. carolinensis was

searched for the presence of tRNA genes using the soft-

ware ARWEN (Laslett and Canback 2008). The complete

set of 22 mt-tRNA genes expected for Metazoa was found

and their putative secondary structure reconstructed (Sup-

plementary material 2). Paired stems of 2–5 bp are present

in DHU and TWC arms. Only three out of 22 tRNA genes

(trnN, trnC, and trnI) have a complete set of Watson and

Crick bonds, whereas all the others are characterized by

non-Watson and Crick bonds (G�U = 28) and/or mis-

matches (10).

Most of the tRNA genes can be folded into the canonical

cloverleaf structure, although two of them, trnA and

trnSgcu, have a truncated DHU arm (Podsiadlowski and

Braband 2006; Park et al. 2007; Masta et al. 2010), simi-

larly to what observed in all other pycnogonids with the

exception of C. megalonyx (Dietz et al. 2011). The two

tRNA genes trnK and trnD, both encoded on the J-strand,

overlap by 2 bp (Table 1; Supplementary material 2).

Notably, the anticodon for trnK is UUU as in all other

pycnogonids except for C. megalonyx, where the canonical

CUU 3-nucleotide sequence is, in turn, used (Masta et al.

2010; Dietz et al. 2011).

Discussion

Heterogeneity in nucleotide composition, also observed in

A. carolinensis (i.e., a higher content of A and T vs. C and

G), is a remarkable molecular feature of most arthropod

mtDNAs (Hassanin et al. 2005) and is thought to be a

consequence of different mutation and selection pressures

acting on the two strands. In addition, specific point

mutations occurring when the molecule is in a single strand

state, during the peculiar asymmetrical replication and/or

Fig. 3 Secondary structure of

the AT-rich region of

A. carolinensis (J-orientation)

with boundaries of complete

(R1–3) and partial (PR) repeats

highlighted

Polar Biol

123

transcription processes of the mtDNA, lead to deviations

from the second parity rule (PR2) (Frank and Lobry 1999)

and to a gradient of asymmetry in base composition

between J- and N-strands (Faith and Pollock 2003).

Accordingly, arthropods’ mitochondrial genomes fre-

quently have a J-strand with positive AT- and CG-skew

values (Hassanin et al. 2005; Wei et al. 2010), given that

alternative (asymmetric) mutation pressures may work on

opposite strands depending on the polarity of the origin of

replication, and therefore on the duration of time spent by

the N-strand into a single-stranded state (DssH) (Reyes

et al. 1998). This molecular bias usually leads to hetero-

geneity in nucleotide composition between J- and

N-strands, which are A?C- and G?T-rich, respectively. In

protein-encoding genes, a higher frequency of Ts in second

codon positions in genes encoded on the J- and N-strand

has already been observed for b-sheet structures of proteins

in humans and prokaryotes (Chiusano et al. 2000). This

may be associated, in mitochondrial genes, with the

necessity of assembling inter-membrane hydrophobic pro-

teins (Bradshaw et al. 2005) where 24 of 26 triplets

encoding for hydrophobic residues have T or C at second

position (Naylor et al. 1995). This observation would

suggest that the same selective pressure may be acting

similarly on the two strands, promoting the introduction of

a high number of T bases in second codon positions.

Notably, this pressure opposes, in genes encoded on the

J-strand, to the one introduced by deaminations, while is

concordant for genes encoded on the N-strand. To sum-

marize, given that the almost entire mitochondrial genome

is transcribed into mRNA or RNAs, its nucleotide content

is highly influenced by the chemical nature of the tran-

scripts. Pyrimidine richness may be for the most part driven

by the necessity to have T at second codon positions in

order to synthesize hydrophobic proteins, although the

overall nucleotide content frequently has a positive

AT-skews due to the double number of first and third

codon positions (with respect to second) and by the

occurrence of hydrophilic RNAs molecules, notoriously

purine-rich (Bradshaw et al. 2005).

The structure of tRNASgcu, lacking the DHU arm, but

with all Watson and Crick pairings in the remaining sec-

ondary structures, is in line with previous comparative

analyses that suggested that this tRNA may have lost the

DHU arm early before the radiation of Metazoa, as it has

never been detected in the mtDNA of multicellular organ-

isms (Wolstenholme 1992; Masta and Boore 2008). The

nucleotide overlap between tRNA genes oriented on the

same strand, observed in A. carolinensis, suggests that

alternative splicing of the immature mRNA or some other

mechanism may be needed in order to produce two complete

molecules. Notoriously, some tRNA genes undergo post-

transcriptional molecular manipulation (polyadenylation) to

restore their correct secondary structure (Yokobori and

Paabo 1995; Lavrov et al. 2000; Boore 2006; Segovia et al.

2011). Occurrence of similar overlaps in tRNA genes is very

frequent in several vertebrates and invertebrates, and some of

them (i.e., trnW/trnC and trnE/trnF) are almost invariably

present in Arthropoda (Juhling et al. 2012).

Conclusions

Our data support previous hypotheses on the molecular

evolution of arthropod mtDNAs (Hassanin et al. 2005;

Castellana et al. 2011). In mtDNAs of pycnogonids, a

strong purifying selection is seemingly acting on second

codon positions in order to maintain the functionality

(=hydrophobicity) of membrane proteins. Conversely,

changes at synonymous sites are characterized by the

common and widespread strand bias associated with the

peculiar asymmetrical replication system of the mtDNA.

Repeated stretches of sequences as those observed in the

A. carolinensis AT-rich region, usually including both

complete and incomplete units, are commonly found in

arthropod taxa (Nardi et al. 2001; Carapelli et al. 2006) and

are generally associated with imprecise initiation/termina-

tion mechanisms of mtDNA replication, and/or to regions

with abrupt shift of polarity between differently oriented

genes (Boore 2000; Carapelli et al. 2008).

The occurrence of truncated stop codons observed in

some A. carolinensis genes is though to be a common

feature shared by most animal lineages and probably con-

nected with the extreme compactness of the genome.

Recent studies, mostly based on Drosophila species, have

demonstrated that mitochondrially encoded genes are

transcribed in polycistronic mRNA sequences, which are

secondarily cleaved into separate mRNAs. According to

the tRNA punctuated model (Ojala et al. 1981), the poly-

cistronic mt-mRNA is cleaved at gene junctions, thanks to

oligonucleolytic enzymes that recognize the secondary

structure of tRNA genes which are interspersed between

PCGs. Enzymes responsible for the endonucleolytic

cleavage at the 50- and 30-ends of tRNA within the poly-

cistronic transcripts (RNase P and RNase Z, respectively)

have been identified (Rossmanith 2012). Nonetheless, the

lack of intervening tRNA genes between some genes fre-

quently leads to the synthesis of bicistronic units that are

alternatively translated. In addition, when a truncated stop

codon is present, proteins encoded in the nucleus are

necessary to synthesize enzymes that restore termination

codon functionality. This latter process is believed to be

accomplished through two fundamental steps: (1) a yet

unknown enzyme restores termination codons via oligoa-

denylation; (2) a poly A polymerase (mtPAP) polyadeny-

lates with about 50 nt the 30 terminus of the mRNA.

Polar Biol

123

Additionally, several enzymes, such as the bicoid stability

factor (BSF) of Drosophila and homologs of the mam-

malian RNA-binding protein (SLIRP), are also believed to

contribute to the stabilization of polyadenylated mRNAs

(Sterky et al. 2010; Bratic et al. 2011).

Extreme environmental conditions, as those experienced

by Antarctic organisms, are generally associated with the

occurrence of peculiar physiological, as well as morpho-

logical, features. Polar gigantism and peculiar pigmenta-

tion patterns are only some of these morphological features

exhibited by marine and terrestrial invertebrates as an

adaptation to the environmental conditions of the Antarctic

Continent (Chapelle and Peck 1999). Considering that the

key feature of Antarctic environment is the extremely low

temperature, and that the mitochondrion is the central

power plant of the cell, there has been quite a lot of

speculation that modifications in the respiratory chain may

arise that divert part of the energy accumulated in the

respiratory chain from ATP to heat production. Neverthe-

less, neither the species here under study, nor the other

pycnogonid Antarctic species (N. unguiculatum), seem to

display molecular features of the mitochondrial genome

and/or of the genes encoding for the respiratory chain that

may be regarded as an adaptation to cold environment.

It is generally thought that large-bodied organisms of

high-latitude marine species (polar gigantism) are mostly

due to high oxygen availability combined with low meta-

bolic rates (Chapelle and Peck 1999; Woods and Moran

2008), but recent works on Antarctic pycnogonids seem to

reject, at least in this group, the oxygen hypothesis (Woods

et al. 2009). Oxygen uptake is a fundamental process of

metazoan taxa that play a key role in energy production

(oxidative phosphorylation) in the mitochondrion inner

membrane. Higher eukaryotic aerobic organisms rely on

this mechanism to generate ATP, transferring electrons to

oxygen by the mitochondrial electron transport chain to

produce water. Reactive intermediates (ROS), which are

very deleterious for the cell because they induce damages

to the milieu and can cause mutations of the DNA, are

however generated during this process (Giustarini et al.

2009). Higher levels of oxygen may therefore increase the

frequency of mutations of mt- and, to a lesser extent,

nu-DNAs. Comparison of pycnogonid mtDNAs does not

provide clear evidence for higher frequencies of mutational

incidence occurring either in the two Antarctic species of

this group (A. carolinensis and N. unguiculatum), or in the

Sub-Antarctic one (C. megalonyx). Accelerated rates in

nucleotide substitution with respect to L. polyphemus (a

basal chelicerate species) and other arthropod groups seem

to be a common molecular feature shared by most cheli-

cerates that leads to long branches in phylogenetic com-

parisons (Podsiadlowski and Braband 2006). This process

is probably dating as early as the initial stages of

diversification of the whole group, rather than being a

consequence of a higher mutational frequency limited to a

specific subgroup.

Acknowledgments This work was funded by the Italian Program of

Research in Antarctica (PNRA). The logistic support of PNRA during

the collection of the material is acknowledged. Partial support was

also provided by the University of Siena. We also thank Dr. Laura

Bianciardi for her collaboration with the analysis of data.

References

Boore JL (2000) The duplication/random loss model for gene

rearrangement exemplified by mitochondrial genomes of deu-

terostome animals. In: Sankoff D, Nadeau J (eds) Comparative

genomics, computational biology (Series vol 1). Kluwer,

Dordrecht, pp 133–147

Boore JL (2006) The complete sequence of the mitochondrial genome

of Nautilus macromphalus (Mollusca: Cephalopoda). BMC

Genomics 7:182

Boore JL, Lavrov DV, Brown WM (1998) Gene translocation links

insects and crustaceans. Nature 392:667–668

Bradshaw PC, Rathi A, Samuels DC (2005) Mitochondrial-encoded

membrane protein transcripts are pyrimidine-rich while soluble

protein transcripts and ribosomal RNA are purine-rich. BMC

Genomics 6:136

Bratic A, Wredenberg A, Gronke S, Stewart JB, Mourier A,

Ruzzenente B, Kukat C et al (2011) The bicoid stability factor

controls polyadenylation and expression of specific mitochon-

drial mRNAs in Drosophila melanogaster. PLoS Genet

7(10):e1002324

Budd GE, Telford MJ (2005) Evolution: along came a sea spider.

Nature 437:1099–1102

Carapelli A, Vannini L, Nardi F, Boore JL, Beani L, Dallai R, Frati F

(2006) The mitochondrial genome of the entomophagous endopar-

asite Xenos vesparum (Insecta: Strepsiptera). Gene 376:248–259

Carapelli A, Lio P, Nardi F, van der Wath E, Frati F (2007)

Phylogenetic analysis of mitochondrial protein coding genes

confirms the reciprocal paraphyly of Hexapoda and Crustacea.

BMC Evol Biol 7(Suppl 2):S8

Carapelli A, Comandi S, Convey P, Nardi F, Frati F (2008) The

complete mitochondrial genome of the Antarctic springtail

Cryptopygus antarcticus (Hexapoda: Collembola). BMC

Genomics 9:315

Castellana S, Vicario S, Saccone C (2011) Evolutionary patterns of

the mitochondrial genome in Metazoa: exploring the role of

mutation and selection in mitochondrial protein coding genes.

Genome Biol Evol 3:1067–1079

Chapelle G, Peck LS (1999) Polar gigantism dictated by oxygen

availability. Nature 399:114–115

Chiusano ML, Alvarez-Valin F, Di Giulio M, D’Onofrio G,

Ammirato G, Colonna G, Bernardi G (2000) Second codon

positions of genes and the secondary structures of proteins.

Relationships and implications for the origin of the genetic code.

Gene 261:63–69

Clayton DA (1982) Replication of animal mitochondrial DNA. Cell

28:693–705

de Rijk P, de Wachter R (1997) RnaViz, a program for the

visualisation of RNA secondary structure. Nucleic Acids Res

25:4679–4684

Dietz L, Mayer C, Arango CP, Leese F (2011) The mitochondrial

genome of Colossendeis megalonyx supports a basal position of

Polar Biol

123

Colossendeidae within the Pycnogonida. Mol Phylogenet Evol

58:553–558

Dunlop JA, Arango CP (2005) Pycnogonid affinities: a review. J Zool

Syst Evol Res 43:8–21

Edgecombe GD (2010) Arthropod phylogeny: an overview from the

perspectives of morphology, molecular data and the fossil

record. Arthropod Struct Dev 39:74–87

Faith JJ, Pollock DD (2003) Likelihood analysis of asymmetrical

mutation bias gradients in vertebrate mitochondrial genomes.

Genetics 165:735–745

Frank AC, Lobry JR (1999) Asymmetric substitution patterns: a

review of possible underlying mutational or selective mecha-

nisms. Gene 238:65–77

Giribet G, Edgecombe GD (2012) Reevaluating the arthropod tree of

life. Annu Rev Entomol 57:167–186

Giribet G, Edgecombe GD, Wheeler WC (2001) Arthropod phylog-

eny based on eight molecular loci and morphology. Nature

413:157–161

Gissi C, Reyes A, Pesole G, Saccone C (2000) Lineage-specific

evolutionary rate in mammalian mtDNA. Mol Biol Evol

17:1022–1031

Gissi C, Iannelli F, Pesole G (2008) Evolution of the mitochondrial

genome of Metazoa as exemplified by comparison of congeneric

species. Heredity 101:301–320

Giustarini D, Dalle-Donne I, Tsikas D, Rossi R (2009) Oxidative

stress and human diseases: origin, link, measurement, mecha-

nisms, and biomarkers. Crit Rev Clin Lab Sci 46:241–281

Hassanin A (2006) Phylogeny of Arthropoda inferred from mito-

chondrial sequences: strategies for limiting the misleading

effects of multiple changes in pattern and rates of substitution.

Mol Phylogenet Evol 38:100–106

Hassanin A, Leger N, Deutsch J (2005) Evidence for multiple

reversals of asymmetric mutational constraints during the

evolution of the mitochondrial genome of Metazoa, and

consequences for phylogenetic inferences. Syst Biol 54:277–298

Juhling F, Putz J, Bernt M, Donath A, Middendorf M, Florentz C,

Stadler PF (2012) Improved systematic tRNA gene annotation

allows new insights into the evolution of mitochondrial tRNA

structures and into the mechanisms of mitochondrial genome

rearrangements. Nucleic Acids Res 40:2833–2845

Kolpakov R, Bana G, Kucherov G (2003) mreps: efficient and flexible

detection of tandem repeats in DNA. Nucleic Acids Res

31:3672–3678

Krzywinski J, Grushko OG, Besansky NJ (2006) Analysis of the

complete mitochondrial DNA from Anopheles funestus: an

improved dipteran mitochondrial genome annotation and a

temporal dimension of mosquito evolution. Mol Phylogenet

Evol 39:417–423

Laslett D, Canback B (2008) ARWEN, a program to detect tRNA

genes in metazoan mitochondrial nucleotide sequences. Bioin-

formatics 24:172–175

Lavrov DV (2011) Key transitions in animal evolution: a mitochon-

drial DNA perspective. In: Schierwater B, DeSalle R (eds) Key

transitions in animals evolution. Science Publishers & CRC

Press, Enfield, New Hampshire, pp 35–54

Lavrov DV, Boore JL, Brown WM (2000) The complete mitochon-

drial DNA sequence of the horseshoe crab Limulus polyphemus.

Mol Biol Evol 17:813–824

Lavrov DV, Boore JL, Brown WM (2002) Complete mtDNA

sequences of two millipedes suggest a new model for mitochon-

drial gene rearrangements: duplication and non-random loss.

Mol Biol Evol 19:163–169

Li H, Liu H, Shi A, Stys P, Zhou X, Cai W (2012) The complete

mitochondrial genome and novel gene arrangement of the

unique-headed bug Stenopirates sp. (Hemiptera: Enicocephali-

dae). PLoS ONE 7(1):e29419

Margam VM, Coates BS, Hellmich RL, Agunbiade T, Seufferheld

MJ, Sun W, Ba MN et al (2011) Mitochondrial genome sequence

and expression profiling for the legume pod borer Maruca vitrata(Lepidoptera: Crambidae). PLoS ONE 6(2):e16444

Masta SE, Boore JL (2008) Parallel evolution of truncated transfer

RNA genes in arachnid mitochondrial genomes. Mol Biol Evol

25:949–959

Masta SE, McCall A, Longhorn SJ (2010) Rare genomic changes and

mitochondrial sequences provide independent support for con-

gruent relationships among the sea spiders (Arthropoda, Pycno-

gonida). Mol Phylogenet Evol 57:59–70

Maxmen A, Browne WE, Martindale MQ, Giribet G (2005)

Neuroanatomy of sea spiders implies an appendicular origin of

the protocerebral segment. Nature 437:1144–1148

Min XJ, Hickey DA (2007) DNA barcodes provide a quick preview of

mitochondrial genome composition. PLoS ONE 2(3):e325

Nardi F, Carapelli A, Fanciulli PP, Dallai R, Frati F (2001) The

complete mitochondrial DNA sequence of the basal hexapod

Tetrodontophora bielanensis: evidence for heteroplasmy and

tRNA translocations. Mol Biol Evol 18:1293–1304

Nardi F, Carapelli A, Frati F (2012a) Internal consistency as a method

to assess the quality of dating estimates using multiple markers.

Mol Phylogenet Evol 62:874–879

Nardi F, Carapelli A, Frati F (2012b) Repeated regions in mitochon-

drial genomes: distribution, origin and evolutionary significance.

Mitochondrion 12:483–491

Naylor GJ, Collins TM, Brown WM (1995) Hydrophobicity and

phylogeny. Nature 373:565–566

Ojala D, Montoya J, Attardi G (1981) tRNA punctuation model of

RNA processing in human mitochondria. Nature 290:470–474Park SJ, Lee YS, Hwang UW (2007) The complete mitochondrial

genome of the sea spider Achelia bituberculata (Pycnogonida,

Ammotheidae): arthropod ground pattern of gene arrangement.

BMC Genomics 8:343

Passamonti M, Ricci A, Milani L, Ghiselli F (2011) Mitochondrial

genomes and Doubly Uniparental Inheritance: new insights from

Musculista senhousia sex-linked mitochondrial DNAs (Bivalvia

Mytilidae). BMC Genomics 12:442

Peden JF (1999) Analysis of codon usage. Dissertation. University of

Nottingham, UK

Perna NT, Kocher TD (1995) Patterns of nucleotide composition at

fourfold degenerate sites of animal mitochondrial genomes.

J Mol Evol 41:353–358

Podsiadlowski L, Braband A (2006) The complete mitochondrial

genome of the sea spider Nymphon gracile (Arthropoda:

Pycnogonida). BMC Genomics 7:284

Podsiadlowski L, Carapelli A, Nardi F, Dallai R, Koch M, Boore JL,

Frati F (2006) The mitochondrial genomes of Campodea fragilisand Campodea lubbocki (Hexapoda: Diplura): high genetic

divergence in a morphologically uniform taxon. Gene 381:49–61

Regier JC, Shultz JW, Zwick A, Hussey A, Ball B, Wetzer R, Martin

JW et al (2010) Arthropod relationships revealed by phyloge-

nomic analysis of nuclear protein-coding sequences. Nature

463:1079–1083

Rehm P, Borner J, Meusemann K, von Reumont BM, Simon S, Hadrys

H, Misof B et al (2011) Dating the arthropod tree based on large-

scale transcriptome data. Mol Phylogenet Evol 61:880–887

Reyes A, Gissi C, Pesole G, Saccone C (1998) Asymmetrical

directional mutation pressure in the mitochondrial genome of

mammals. Mol Biol Evol 15:957–966

Rossmanith W (2012) Of P and Z: mitochondrial tRNA processing

enzymes. Biochim Biophys Acta 1819:1017–1026

Rota-Stabelli O, Kayal E, Gleeson D, Daub J, Boore JL, Telford MJ,

Pisani D et al (2010) Ecdysozoan mitogenomics: evidence for a

common origin of the legged invertebrates, the Panarthropoda.

Genome Biol Evol 2:425–440

Polar Biol

123

Saito S, Tamura K, Aotsuka T (2005) Replication origin of

mitochondrial DNA in insects. Genetics 171:1695–1705

San Mauro D, Gower DJ, Zardoya R, Wilkinson M (2006) A hotspot

of gene order rearrangement by tandem duplication and random

loss in the vertebrate mitochondrial genome. Mol Biol Evol

23:227–234

Segovia R, Pett W, Trewick S, Lavrov DV (2011) Extensive and

evolutionarily persistent mitochondrial tRNA editing in the

velvet worms (Phylum Onychophora). Mol Biol Evol 28:2873–

2881

Simon C, Frati F, Beckenbach A, Crespi B, Liu H, Flook P (1994)

Evolution, weighting, and phylogenetic utility of mitochondrial

gene sequences and a compilation of conserved polymerase

chain reaction primers. Ann Entomol Soc Am 87:651–701

Sterky FH, Ruzzenente B, Gustafsson CM, Samuelsson T, Larsson

NG (2010) LRPPRC is a mitochondrial matrix protein that is

conserved in metazoans. Biochem Biophys Res Commun

398:759–764

Swofford DL (2002) PAUP*: phylogenetic analysis using parsimony

(* and other methods), version 4.0. Sinauer Associates,

Sunderland

Talavera G, Vila R (2011) What is the phylogenetic signal limit from

mitogenomes? The reconciliation between mitochondrial and

nuclear data in the Insecta class phylogeny. BMC Evol Biol

11:315

Waloszek D, Dunlop JA (2002) A larval sea spider (Arthropoda:

Pycnogonida) from the Upper Cambrian ‘Orsten’ of Sweden, and

the phylogenetic position of pycnogonids. Palaeontology

45:421–446

Wei SJ, Shi M, Chen XX, Sharkey MJ, van Achterberg C, Gong-Yin

Y, Jun-Hua H (2010) New views on strand asymmetry in insect

mitochondrial genomes. PLoS ONE 5(9):e12708

Wolstenholme DR (1992) Animal mitochondrial DNA: structure and

evolution. Int Rev Cytol 141:173–216

Woods HA, Moran AL (2008) Temperature–oxygen interactions in

Antarctic nudibranch egg masses. J Exp Biol 211:798–804

Woods HA, Moran AL, Arango CP, Mullen L, Shields C (2009)

Oxygen hypothesis of polar gigantism not supported by perfor-

mance of Antarctic pycnogonids in hypoxia. Proc R Soc B

276:1069–1075

Yokobori S, Paabo S (1995) Transfer RNA editing in land snail

mitochondria. Proc Natl Acad Sci USA 92:10432–10435

Zuker M (2003) Mfold web server for nucleic acid folding and

hybridization prediction. Nucleic Acids Res 31:3406–3415

Polar Biol

123