For Peer Review

Characterization of purine catabolic pathway genes in

coelacanths

Journal: JEZ Part B: Molecular and Developmental Evolution

Manuscript ID: JEZ-B-2012-12-0078.R2

Wiley - Manuscript type: Research Article

Date Submitted by the Author: n/a

Complete List of Authors: Forconi, Mariko; Università Politecnica delle Marche, Scienze della vita e dell'ambiente Biscotti, Maria; Università Politecnica delle Marche, Scienze della vita e dell'ambiente Barucca, Marco; Università Politecnica delle Marche, Scienze della vita e dell'ambiente Buonocore, Francesco; Università della Tuscia, Dipartimento per l’Innovazione nei Sistemi Biologici, Agroalimentari e Forestali De Moro, Gianluca; Università di Trieste, Dipartimento di Scienze della Vita Fausto, Anna; Università della Tuscia, Dipartimento per l’Innovazione nei Sistemi Biologici, Agroalimentari e Forestali Gerdol, Marco; Università di Trieste, Dipartimento di Scienze della Vita Pallavicini, Alberto; Università di Trieste, Dipartimento di Scienze della Vita Scapigliati, Giuseppe; Università della Tuscia, Dipartimento per l’Innovazione nei Sistemi Biologici, Agroalimentari e Forestali Schartl, M.; Physiologische Chemie I, Biozentrum der Universitat Wuerzburg, Am Hubland, 97074 Wuerzburg, Germany., Physiological Chemistry I Olmo, Ettore; Università Politecnica delle Marche, Scienze della vita e dell'ambiente Canapa, Adriana; Università Politecnica delle Marche, Scienze della vita e dell'ambiente

Keywords: Purine catabolism, Latimeria, Uox, HIUase, PRHOXNB, ALN, ALC

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

For Peer Review

1

Characterization of purine catabolic pathway genes in coelacanths 1

Mariko Forconi1, Maria Assunta Biscotti

1, Marco Barucca

1*, Francesco Buonocore

2, Gianluca De 2

Moro3, Anna Maria Fausto

2, Marco Gerdol

3, Alberto Pallavicini

3, Giuseppe Scapigliati

2, Manfred 3

Schartl4, Ettore Olmo

1, Adriana Canapa

1. 4

5

1 Dipartimento di Scienze della Vita e dell’Ambiente, Università Politecnica delle Marche, Ancona, 6

Italy. 7

2 Dipartimento per l’Innovazione nei Sistemi Biologici, Agroalimentari e Forestali, Università della 8

Tuscia (Viterbo, Italy). 9

3 Dipartimento di Scienze della Vita, Università di Trieste (Italy). 10

4 Physiological Chemistry, Biocenter, University of Wuerzburg, Wuerzburg, Germany. 11

12

13

Figures: 5 14

15

Running headline: Purine catabolism in Latimeria 16

17

18

Correspondence to: Marco Barucca, Dipartimento di Scienze della Vita e dell’Ambiente, 19

Università Politecnica delle Marche, via Brecce Bianche, 60131 Ancona, Italy. Email: 20

m.barucca@univpm.it. Tel: 0039 071 2204 158, Fax: 0039 071 2204 609. 21

22

23

Page 1 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

2

ABSTRACT 1

Coelacanths are a critically valuable species to explore the gene changes that took place in the 2

transition from aquatic to terrestrial life. One interesting and biologically relevant feature of the 3

genus Latimeria is ureotelism. However not all urea is excreted from the body; in fact high 4

concentrations are retained in plasma and seem to be involved in osmoregulation. The purine 5

catabolic pathway, which leads to urea production in Latimeria, has progressively lost some steps, 6

reflecting an enzyme loss during diversification of terrestrial species. We report the results of 7

analyses of the liver and testis transcriptomes of the Indonesian coelacanth Latimeria menadoensis 8

and of the genome of Latimeria chalumnae, which has recently been fully sequenced in the 9

framework of the coelacanth genome project. We describe five genes, uricase, 5-hydroxyisourate 10

hydrolase, parahox neighbor B, allantoinase, and allantoicase, each coding for one of the five 11

enzymes involved in urate degradation to urea, and report the identification of a putative second 12

form of 5-hydroxyisourate hydrolase that is characteristic of the genus Latimeria. The present data 13

also highlight the activity of the complete purine pathway in the coelacanth liver and suggest its 14

involvement in the maintenance of high plasma urea concentrations. 15

16

17

INTRODUCTION 18

Urea, one of the most common excreted products of vertebrates together with urate and ammonia, is 19

produced by two pathways: the urea cycle and purine catabolism. Purine catabolism begins with 20

adenine and guanine degradation to xanthine, which is converted to urate, the central metabolite in 21

all vertebrates (Fig. 1). Urate is oxidized by urate oxidase (uricase, UOX) (E.C. 1.7.3.3) to 5-22

hydroxyisourate, which in turn is transformed by 5-hydroxyisourate hydrolase (HIUase) (E.C. 23

3.5.2.17) to 5-hydroxy-2-oxo-4-ureido-2,5-dihydro-1H-imidazole-5-carboxylate. This molecule is 24

Page 2 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

3

converted to allantoin through catalysis by parahox neighbor B (PRHOXNB, also known as 2-oxo-1

4-hydroxy-4-carboxy-5-ureidoimidazoline decarboxylase, E.C.4.1.1.-). The latter two steps have 2

recently been discovered through phylogenetic comparison of whole genomes (Ramazzina et al., 3

2006). Allantoin is degraded by allantoinase (ALN) (E.C. 3.5.2.5) to allantoate, which is hydrolyzed 4

to urea and ureidoglycolate by allantoicase (ALC) (E.C. 3.5.3.4). Finally urea is converted by 5

urease (E.C. 3.5.1.5) to ammonia and carbon dioxide (Fig. 1). 6

Some of these enzymes have been lost during vertebrate evolution, with a truncation of the pathway 7

in higher animals and a consequent change in the excreted end product (Fig. 1). Loss of the final 8

steps of the pathway may be considered to be an adaptation to terrestrial life, since high plasma 9

ammonia can lead to toxemia when water is scarce (Keilin, ’59; Shambaugh, ’77); in such 10

conditions urate excretion is more convenient. The study of Latimeria, a lobe-finned fish and one of 11

the nearest extant relatives to the tetrapod ancestor, could provide insights into the evolution of 12

uricolysis. Furthermore, the high plasma urea levels described in Latimeria (Schmidt-Nielsen, ’64) 13

seem to be involved in osmoregulation (Brown and Brown, ’67), thus providing an intriguing 14

example of the evolutionary convergence with elasmobranchs (Wood et al., ’95). Investigation of 15

the carbamoyl phosphate synthetase (CPS I)–arginase (ARG2) system, the other source of urea, in 16

the African coelacanth Latimeria chalumnae has uncovered potentially adaptive novelties in the 17

hepatic urea cycle that arose during the evolution of terrestrial vertebrates (Amemiya et al., 2013). 18

The Latimeria gene inventory of purine catabolism enzymes reported in this study documents a 19

complete pathway including five enzymes. An additional form of HIUase in testis identified in this 20

work suggests neofunctionalization. Phylogenetic analyses and microsynteny conservation of the 21

five enzymes indicate a closer relationship of coelacanths to tetrapods than to teleosts. This 22

suggested to us that the enzyme pattern found in Latimeria is probably an ancestral condition and 23

that its characterization may be a fundamental starting point to understand the evolutionary changes 24

involved in purine catabolism during the radiation of terrestrial vertebrates. 25

Page 3 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

4

1

2

MATERIALS AND METHODS 3

L. menadoensis sampling is described in Makapedua et al. (2011), liver and testis transcriptome 4

sequencing and assembly are described in the work Pallavicini et al. (2013). L. chalumnae 5

sampling, genome sequencing, and assembly are described in Amemiya et al. (2013). 6

Nucleotide sequences of UOX, PRHOXNB, ALN, ALC, HIUase A, HIUase B, and transthyretin 7

(TTR) (accession numbers HF678434 - HF678440) were obtained from high-quality transcripts 8

generated by RNA-sequencing analysis of L. menadoensis liver and testis (Canapa et al., 2012; 9

Pallavicini et al., 2013). The same set of genes was investigated in the genome of L. chalumnae 10

(Amemiya et al., 2013). Each retrieved sequence was validated using tBLASTn on the NCBI 11

database, in order to allow annotation of the putative protein, and analyzed using ProtParam 12

(Gasteiger et al., 2005, http://web.expasy.org/protparam/) for the biophysical parameters: molecular 13

weight, isoelectric point, and estimated half-life (based on the “N-end rule”). The genetic distance 14

between respective nucleotide sequences of the two Latimeria species was calculated as p-distance 15

percentage with PAUP (Swofford et al., 2002). Latimeria coding sequences (CDS) were aligned 16

with CDS from zebrafish, puffer fish, Xenopus, platypus, and mouse as in Keebaugh and Thomas 17

(2009). The ω rate (non-synonymous/synonymous mutations) was obtained with KaKs_calculator 18

(Zhang et al., 2006) according to Goldman and Yang (’94). The expression levels of purine 19

catabolism pathway enzymes and of the housekeeping gene phosphoglycerate kinase (PGK) were 20

computed in L. menadoensis liver as Fragments Per Kilobase of exons per Million sequenced 21

fragments (FPKM) using CLC Genomics Workbench 4.5.1. Expression levels for testis were also 22

calculated for comparison (data not shown). The sequences of other vertebrate species used in 23

phylogenetic analyses and multiple alignments are listed in Table S1. 24

Page 4 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

5

Vertebrate UOX, HIUase-TTR, PRHOXNB, ALN, and ALC amino acid alignments were 1

performed with CLUSTALW2 using default parameters (Larkin et al., 2007). HIUase-TTR 2

phylogenetic analysis was performed using Bayesian Inference (BI) with MrBayes program 3

(Huelsenbeck et al., 2001) by applying the WAG amino acid model (Whelan and Goldman, 2001). 4

Analyses were run for 2,000,000 generations with sampling every 100 generations; the first 5,000 5

were discarded as burn-in, and obtainment of stationarity was considered when the average standard 6

deviation of split frequencies reached a value < 0.007. A Maximum Parsimony (MP) analysis was 7

performed with PAUP using tree bisection–reconnection (TBR) branch swapping and random 8

stepwise additions with 100 replications. Only minimal trees were retained. Bootstrap values refer 9

to 10,000 replications. Parsimony informative sites: 104. 10

UOX, PRHOXNB, ALN, and ALC phylogenetic trees were obtained with MrBayes (1,000,000 11

generations, sampling every 100 generations, burn-in 2,500 for each analysis). The amino acid 12

models applied are annotated in each tree legend (S-Figs. 1-4). 13

Conservation of microsynteny was analyzed for UOX, HIUase A, HIUase B, PRHOXNB, ALN, and 14

ALC by comparing Ensembl gene annotations (release 69-Oct. 2012) for L. chalumnae, Anolis 15

carolinensis, Bos taurus, Callithrix jacchus, Danio rerio, Gallus gallus, Gasterosteus aculeatus, 16

Homo sapiens, Macaca mulatta, Mus musculus, Oreochromis niloticus, Ornithorhynchus anatinus, 17

Oryzias latipes, Petromyzon marinus, Tetraodon nigroviridis, and Xenopus tropicalis. 18

19

20

RESULTS 21

Transcripts for the genes of the five enzymes involved in the degradation of urate to urea were 22

identified from the analysis of the L. menadoensis liver transcriptome. The genetic distance between 23

Page 5 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

6

L. menadoensis and L. chalumnae nucleotide sequences ranges from 0.053% to 0.182% (Table 1) 1

and is in line with data reported for other genes (Amemiya et al., 2013; Forconi et al., 2013; 2

Pallavicini et al., 2013). Such low genetic distance allows combining the sequence information from 3

the two coelacanths. Alignment of L. menadoensis transcripts to L. chalumnae genomic scaffolds 4

permitted better definition of transcript features and genomic annotation. 5

A second putative HIUase transcript, that we designated HIUase B, was identified in addition to the 6

sequences coding for the 5 enzymes making up the pathway. HIUase B appears to be composed by 7

a partial cadherin-like sequence at the 5’ terminus (eight inferred exons, 1,215 nt) and three HIUase 8

exons (493 nt). The alignment of the two HIUases shows that two exons, transcribing the 5’ UTR 9

and the 5’ end of CDS in HIUase A, are missing in the B form. The 19% divergence calculated 10

between the two HIUase forms in their comparable portion indicates the existence of two different 11

genes. Moreover, analysis of microsynteny conservation in the flanking regions of the two forms 12

suggests that the localization of the A form in the genome of L. chalumnae, where it is linked 13

downstream to a region containing genes GAS8 and DNDBB1, is similar to the tetrapod pattern 14

(Fig. 2). The synteny of genes downstream of HIUase A is not conserved in piscine species (data 15

not shown). A sequence similar to cadherin 1 (CDH1-like) was identified about 30 kb upstream the 16

coelacanth HIUase A gene. O. anatinus, X. tropicalis, some cetartiodactyla, and carnivora also have 17

a similar CDH1-like region upstream of the HIUase gene (Fig. 2). The B form of the HIUase gene 18

is linked, upstream of its CDS, to a region containing CDH1 and TMCO7, which are arranged in the 19

opposite direction. CDH3, the gene commonly found upstream of CDH1 and TMCO7 in tetrapods, 20

is not annotated in L. chalumnae, nor in G. gallus. However chicken CDH1 is found close to the 21

single HIUase gene on chromosome 11, whereas in coelacanths it is more than 100 kb away from 22

HIUase B. Moreover it was not possible to infer the relative proximity of the HIUase A and HIUase 23

B loci as the genes were found on two separate scaffolds (Fig. 2). TTR sequences, i.e., the other 24

members of the TTR/HIUase family, were included in the phylogenetic analyses together with 25

Page 6 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

7

several representative vertebrate HIUases, to clarify the relationships between the HIUase forms of 1

the two coelacanths. Phylogenetic trees show two main clades corresponding to HIUase and TTR 2

proteins. In both MP and BI trees, which share a similar topology, the two putative HIUase forms of 3

Latimeria are nested together in the same node in a sister clade to the tetrapod HIUases (Fig. 3). 4

The expression pattern of the A form in the Indonesian coelacanth indicates higher transcription in 5

liver (14.10 FPKM) than testis (2.66 FPKM), whereas the B form is more expressed in testis than 6

liver (3.59 FPKM vs. 0.18 FPKM) (Fig. 4A). 7

With regard to the genes of other enzymes, allantoinase was the most highly expressed (95.31 8

FPKM) while PRHOXNB showed the lowest expression (4.96 FPKM) (Fig. 4B). UOX, PRHOXNB, 9

ALN, and ALC expression in testis was below the 1.00 FPKM threshold (data not shown). 10

Molecular phylogenetic analyses (Fig. 3, S-Figs. 1-4) reveal a number of different relationships 11

among the coelacanth proteins: in the PRHOXNB, ALN, and ALC trees the Latimeria sequences 12

form a sister clade to the tetrapod sequences, whereas in the UOX tree they cluster with teleost 13

sequences. 14

Microsyntenies are reported in supplementary Figures 5-8. The flanking gene arrangement suggests 15

that coelacanth PRHOXNB and ALC adhere to a tetrapod pattern: the CDX2 gene, which in 16

sarcopterygian genomes is found downstream PRHOXNB, is missing in teleosts; COLEC11 and 17

RPS7 synteny upstream of ALC appears to be a feature of the tetrapod lineage. 18

On the other hand the microsynteny relationships of UOX and ALN appear to be conserved across 19

all the vertebrates analyzed, even though the latter locus displays a gene loss in amniota with the 20

exception of A. carolinensis. In this representative of reptiles the ALN sequence is retained in the 21

genome under purifying selection (Keebaugh and Thomas, 2010) even though purine catabolism 22

ends with production of urate in this taxon. 23

A putative peroxisomal translocation signal 1 (PTS1) characterized by the sequence SKL or by 24

conservative variants (S/A/C)(K/R/H)(L/M), identified at the UOX C-terminus (Fig. 5), may 25

Page 7 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

8

indicate a subcellular localization of the enzyme (Rachubinski and Subramani, ’95); no 1

translocation signals were identified in the two HIUases, PRHOXNB, ALN, or ALC. 2

The ω rates (non-synonymous/synonymous substitutions) of L. menadoensis and other 3

representative vertebrates (zebrafish, puffer fish, Xenopus, platypus, and mouse) are reported for 4

each gene in Table 2. All enzymes appear to be under purifying selection (Ka/Ks < 1). 5

6

7

DISCUSSION 8

Among the vertebrate adaptations associated with the passage from aquatic to terrestrial life, the 9

evolution of purine catabolism is particularly intriguing: comparison of data from the various 10

phylogenetic lineages shows a complex picture due to several gene losses and pseudogene 11

conversion events that led to shortening of the pathway in "higher vertebrates" (Remy et al., ’51; 12

Keilin, ’59; Yeldandi et al., ’91; Wu et al., ’92; Urich, ’94; Oda et al., 2002). One such instance is 13

the loss of UOX activity in hominoids, birds, and reptiles (Remy et al., ’51; Keilin, ’59; Urich, ’94). 14

Over the years, several hypotheses have been advanced to explain the adaptive role of high serum 15

uric acid resulting from UOX inactivity in hominoids (Haldane, ’55; Orowan, ’55; Ames et al., ’81; 16

Watanabe et al., 2002). 17

In this scenario Latimeria, which has seemingly ancient morphological and physiological 18

characteristics and represents one of the nearest relatives to the tetrapod ancestor, is a particularly 19

interesting species for exploring some pre-adaptations that facilitated the transition to land. Urate 20

oxidase catalyzes the first step in the uricolytic pathway and its presence in prokaryotes and 21

eukaryotes suggests an ancient origin (Oda et al., 2002). Surprisingly, while in hominoids UOX has 22

been classified as a pseudogene, in reptiles and birds a defective gene has been reported to evolve 23

under purifying selection, suggesting that in these taxa the gene, no longer involved in purine 24

catabolism, may have acquired new functions (Keebaugh and Thomas, 2010). The phylogenetic 25

Page 8 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

9

analysis indicates that the UOX sequences of the two coelacanths and of Protopterus annectens 1

(lungfish) are more closely related to UOX sequences of the teleost clade than to those of other 2

sarcopterygian clades (S-Fig. 1). The ω ratio values show strong purifying selection (Table 2). 3

The C-terminal sequence of the Latimeria UOX enzyme (Fig. 5) encodes a peroxisome 4

translocation signal (PTS1, Gould et al., ’89; Miura et al., ’92) that suggests translocation to the 5

organelle as shown for all homologous functional vertebrate enzymes (Hayashi et al., 2000). PTS1 6

is absent in the reptile A. carolinensis and in the chicken G. gallus, which encode a slightly larger 7

defective protein (Wu et al., ’89; Yeldandi et al., ’90, ’91; Oda et al., 2002). Despite the purifying 8

selection reported by Keebaugh and Thomas (2010) in such lineages, the long bird/reptile branches 9

shown in Supplementary Figure 1 could be explained by the presence of small regions under 10

positive selection hidden by larger regions under purifying selection, which thus go undetected 11

(Keebaugh and Thomas, 2010). 12

The second step of the uric acid degradation pathway is catalyzed by HIUase. This step and 13

PRHOXNB activity have been described recently (Ramazzina et al., 2006). The HIUase enzymes 14

show high sequence similarity with TTR, which arose by gene duplication in vertebrates (both 15

belong to the TTR-like family). TTRs are a group of plasma proteins that carry thyroid hormones to 16

the brain; they are highly conserved in vertebrates, whereas HIUases are found in a wider range of 17

organisms including bacteria, plants, fungi, and vertebrates (Eneqvist et al., 2003). In TTRs the 18

conserved “TAVV” domain at the C-terminus is replaced by “YRGS” (Eneqvist et al., 2003; Lee et 19

al., 2005). 20

HIUase-TTR phylogenetic analysis highlighted two genes closely related to HIUases in the two 21

coelacanths. Their expression in liver and microsynteny conservation indicate that HIUase A may 22

have maintained its original function. The assembled transcript of HIUase B is characterized by a 23

portion encoding a cadherin-like sequence at the 5’ end and by three exons of HIUase in frame. A 24

similar association with a cadherin-like sequence has been found in the single HIUase of Mustela 25

Page 9 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

10

putorius furo (AES10673). In addition the potential activity of HIUase B may be unaffected by the 1

absence of the HIUase first coding exon; indeed Power et al. (2009) have described alternative 2

splicing isoforms of HIUase devoid of the first coding exon in some fish species. However, despite 3

the current lack of evidence of the translation of this second form, HIUase B expression in testis, its 4

extremely low expression in liver, and the presence of a partial cadherin-like sequence capable of 5

binding Ca2+

suggest a possible different function. 6

The presence of two HIUase forms also in teleosts seems to be due to the teleost-specific whole 7

genome duplication (TGD, 3R hypothesis). Tree topology and the phylogenetic distance between 8

the two forms suggest that the gene duplication in coelacanths was due to a lineage-specific event; 9

the second form may have arisen due to recombination; indeed analysis of HIUase B microsynteny 10

placed the gene in a region that in tetrapods contains CDH3 (not annotated in the contig of L. 11

chalumnae), CDH1, and TMCO7 genes. 12

PRHOXNB, the third enzyme in the pathway, is found neither in reptiles nor in birds (S-Fig. 6) and 13

appears to be under relaxed selection in humans and hominoids (Keebaugh and Thomas, 2009, 14

2010). The low transcript expression and half-life predictions inferred in silico (Table 1) may 15

indicate a role for the protein as a rate-limiting step in the pathway. Alternatively the low amount of 16

PRHOXNB transcripts could be offset by a faster translation rate or by higher catalytic activity. 17

For ALN and ALC, the lack of PTS1 at C-terminus in Latimeria might preclude a subcellular 18

localization in peroxisomes; in contrast, ALN and ALC show a bona fide PTS1 sequence in several 19

teleost species (Hayashi et al., 2000). The coelacanth proteins might be localized in the 20

mitochondria as suggested for Xenopus (Usuda et al., ’94). 21

In conclusion, the significance of the present work lies in the identification of the components of the 22

purine catabolism enzyme pathway in the basal-most living sarcopterygian and in the clarification 23

of their evolutionary history and ancestral state in sarcopterygians. 24

Page 10 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

11

Identification of the complete functional purine pathway in Latimeria indicates that the loss of 1

function of some genes in the lineage that ultimately led to tetrapods occurred after the transition 2

from aquatic to terrestrial life, presumably in relation to adaptation to new environments. The 3

conservation of an ancestral form of the pathway in Latimeria may also be related to the need for 4

high urea production, which would be required to maintain high plasma osmolarity. 5

In addition, microsynteny and phylogenetic analysis of the five purine catabolism genes also 6

evinced a clear difference between teleost and tetrapod clades. However, their situation in Latimeria 7

(i.e., possessing features very similar to those of tetrapods) suggests that the divergence is not the 8

result of adaptation to terrestrial life, since such features are shared by the whole sarcopterygian 9

lineage. 10

11

12

ACKNOWLEDGMENTS 13

The authors belonging to Dipartimento di Scienze della Vita e dell’Ambiente of Università 14

Politecnica delle Marche are affiliated with Istituto Nazionale Biosistemi e Biostrutture (INBB). 15

M.S. is supported by Deutsche Forschungsgemeinschaft. We are grateful to A.C. Keebaugh and 16

J.W. Thomas for sharing the sequence alignment for Ka/Ks estimation, to anonymous reviewers and 17

to C.T. Amemiya for their editorial suggestions. 18

19

20

LITERATURE CITED 21

Amemiya CT, Alföldi J, Lee AP, et al. 2013. The African coelacanth genome provides insights into 22

tetrapod evolution. Nature 496:311-316. 23

Page 11 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

12

Ames BN, Cathcart R, Schwiers E, Hochstein P. 1981. Uric acid provides an antioxidant defense in 1

humans against oxidant- and radical-caused aging and cancer: a hypothesis. Proc Natl Acad 2

Sci USA 78:6858-6862. 3

Brown GW Jr and Brown SG. 1967. Urea and its formation in coelacanth liver. Science 155:570-4

573. 5

Canapa A, Olmo E, Forconi M, Pallavicini A, Makapedua MD, Biscotti MA, Barucca M. 2012. 6

Composition and phylogenetic analysis of vitellogenin coding sequences in the Indonesian 7

coelacanth Latimeria menadoensis. J Exp Zool B Mol Dev Evol 318:404-416. 8

Eneqvist T, Lundberg E, Nilsson L, Abagyan R, Sauer-Eriksson AE. 2003. The transthyretin-related 9

protein family. Eur J Biochem 270:518-532. 10

Forconi M, Canapa A, Barucca M, Biscotti MA, Buonocore

F, Fausto

AM, Makapedua

DM, 11

Pallavicini A, Gerdol

M, De Moro

G, Scapigliati

G, Olmo

E, Schartl

M. 2013. Characterization 12

of sex determination and sex differentiation genes in Latimeria. PLoS ONE 8(4):e56006. 13

Gasteiger E, Hoogland C, Gattiker A, Duvaud S, Wilkins MR, Appel RD, Bairoch A. 2005. Protein 14

Identification and Analysis Tools on the ExPASy Server. In: Walker JM, editor. The 15

Proteomics Protocols Handbook. Totowa, NJ: Humana Press. p 571-607. 16

Goldman N and Yang Z. 1994. A codon-based model of nucleotide substitution for protein-coding 17

DNA sequences. Mol Biol Evol 11:725-736. 18

Gould SJ, Keller GA, Hosken N, Wilkinson J, Subramani, S. 1989. A conserved tripeptide sorts 19

proteins to peroxisomes. J Cell Biol 108:1657-1664. 20

Haldane JB. 1955. Origin of man. Nature 176:169-170. 21

Page 12 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

13

Hayashi S, Fujiwara S, Noguchi T. 2000. Evolution of urate-degrading enzymes in animal 1

peroxisomes. Cell Biochem Biophys 32:123-129. 2

Huelsenbeck JP, Ronquist F, Nielsen R, Bollback JP. 2001. Bayesian inference of phylogeny and its 3

impact on evolutionary biology. Science 294:2310-2314. 4

Keebaugh AC and Thomas JW. 2009. The genomes of the South American opossum (Monodelphis 5

domestica) and platypus (Ornithorhynchus anatinus) encode a more complete purine 6

catabolic pathway than placental mammals. Comp Biochem Physiol Part D Genomics 7

Proteomics 4:174-178. 8

Keebaugh AC and Thomas JW. 2010. The evolutionary fate of the genes encoding the purine 9

catabolic enzymes in hominoids, birds, and reptiles. Mol Biol Evol 27:1359-69. 10

Keilin J. 1959. The biological significance of uric acid and guanine excretion. Biol Rev 34:265-294. 11

Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, McWilliam H, Valentin F, 12

Wallace IM, Wilm A, Lopez R, Thompson JD, Gibson TJ, Higgins DG. 2007. ClustalW and 13

ClustalX version 2. Bioinformatics 23:2947-2948. 14

Lee Y, Lee DH, Kho CW, Lee AY, Jang M, Cho S, Lee CH, Lee JS, Myung PK, Park BC, Park 15

SG. 2005. Transthyretin-related proteins function to facilitate the hydrolysis of 5-16

hydroxyisourate, the end product of the uricase reaction. FEBS Lett 579:4769-4774. 17

Makapedua DM, Barucca M, Forconi M, Antonucci N, Bizzaro D, Amici A, Carradori MR, Olmo 18

E, Canapa A. 2011. Genome size, GC percentage and 5mC level in the Indonesian coelacanth 19

Latimeria menadoensis. Mar Genomics 4:167-172. 20

Page 13 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

14

Miura S, Kasuya-Arai I, Mori H, Miyazawa S, Osumi T, Hashimoto T, Fujiki Y. 1992. Carboxyl-1

terminal consensus Ser-Lys-Leu-related tripeptide of peroxisomal proteins functions in vitro 2

as a minimal peroxisome-targeting signal. J Biol Chem 267:14405-14411. 3

Oda M, Satta Y, Takenaka O, Takahata N. 2002. Loss of urate oxidase activity in hominoids and its 4

evolutionary implications. Mol Biol Evol 19:640-653. 5

Orowan E. 1955. The origin of man. Nature 175:683-684. 6

Pallavicini A, Canapa A, Barucca M, Alföldi J, Biscotti MA, Buonocore F, De Moro G, Di Palma 7

F, Fausto AM, Forconi M, Gerdol M, Makapedua DM, Turner-Meier J, Olmo E, Scapigliati 8

G. 2013. Analysis of the transcriptome of the Indonesian coelacanth Latimeria menadoensis. 9

BMC Genomics (accepted). 10

Power DM, Morgado I, Cardoso JCR. 2009. Evolutionary insights from fish transthyretin. In: 11

Richardson SJ, Cody V, editors. Recent advances in transthyretin evolution, structure and 12

biological functions. Berlin: Springer. p 59-64. 13

Rachubinski RA and Subramani S. 1995. How proteins penetrate peroxisomes. Cell 83:525-8. 14

Ramazzina I, Folli C, Secchi A, Berni R, Percudani R. 2006. Completing the uric acid degradation 15

pathway through phylogenetic comparison of whole genomes. Nat Chem Biol 2:144-148. 16

Remy C, Richert DA, Westerfeld WW. 1951.The determination of xanthine dehydrogenase in 17

chicken tissues. J Biol Chem 193:649-57. 18

Schmidt-Nielsen K. 1964. Desert animals: physiological problems of heat and water. New York, 19

NY: Oxford University Press. 20

Shambaugh GE. 1977. Urea biosynthesis I. The urea cycle and relationships to the citric acid cycle. 21

Am J Clin Nutr 30:2083-2087. 22

Page 14 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

15

Swofford DL. 2002. PAUP*. Phylogenetic analysis using parsimony (* and other methods) version 1

4. Sunderland, MA: Sinauer Associates. 2

Urich K. 1994. Comparative Animal Biochemistry. Berlin: Springer. 3

Usuda N, Hayashi S, Fujiwara S, Noguchi T, Nagata T, Rao MS, Alvares K, Reddy JK, Yeldandi 4

AV. 1994. Uric acid degrading enzymes, urate oxidase and allantoinase, are associated with 5

different subcellular organelles in frog liver and kidney. J Cell Sci 107:1073-1081. 6

Watanabe S, Kang DH, Feng L, Nakagawa T, Kanellis J, Lan H, Mazzali M, Johnson RJ. 2002. 7

Uric acid, hominoid evolution, and the pathogenesis of salt-sensitivity. Hypertension 40:355-8

360. 9

Whelan S and Goldman N. 2001. A general empirical model of protein evolution derived from 10

multiple protein families using a maximum-likelihood approach. Mol Biol Evol 18:691-699. 11

Wood CM, Part P, Wright PA. 1995. Ammonia and urea metabolism in relation to gill function and 12

acid-base balance in a marine elasmobranch, the spiny dogfish, Squalus acanthias. J Exp Biol 13

198:1545-1558. 14

Wu XW, Lee CC, Muzny DM, Caskey CT. 1989. Urate oxidase: primary structure and evolutionary 15

implications. Proc Natl Acad Sci USA 86:9412-9416. 16

Wu XW, Muzny DM, Lee CC, Caskey CT. 1992. Two independent mutational events in the loss of 17

urate oxidase during hominoid evolution. J Mol Evol 34:78-84. 18

Yeldandi AV, Wang XD, Alvares K, Kumar S, Rao MS, Reddy JK. 1990. Human urate oxidase 19

gene: cloning and partial sequence analysis reveal a stop codon within the fifth exon. 20

Biochem Biophys Res Commun 171:641-646. 21

Page 15 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

16

Yeldandi AV, Yeldandi V, Kumar S, Murthy CV, Wang XD, Alvares K, Rao MS, Reddy JK. 1991. 1

Molecular evolution of the urate oxidase-encoding gene in hominoid primates: nonsense 2

mutations. Gene 109:281-284. 3

Zhang Z, Li J, Zhao XQ, Wang J, Wong GK, Yu J. 2006. KaKs Calculator: Calculating Ka and Ks 4

through model selection and model averaging. Geno Prot Bioinfo 4:259-263. 5

6

Page 16 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

17

FIGURE LEGENDS 1

Figure 1. The purine catabolism pathway. The terminal purine catabolism step is not identical in all 2

taxonomic groups: the end product of this pathway in reptiles, birds, hominoid primates, humans, 3

and some platyrrhini is urate, allantoin in placental mammals (other than primates) and catarrhini, 4

and allantoate in some teleosts. In contrast, amphioxus, elasmobranchs, other teleosts, dipnoans, 5

amphibians, monotremes, marsupials, and coelacanths display a "complete" pathway that produces 6

urea. Vertebrate taxonomic groups are reported on the left of their end product (Keilin, ’59; Urich, 7

’94; Hayashi et al. 2000, Keebaugh and Thomas, 2009). Enzymes are reported on the right between 8

their substrate and product. 9

10

Figure 2. HIUase microsyntenies in L. chalumnae and comparison with orthologous loci for other 11

vertebrates. Coelacanth HIUase A and HIUase B loci are found on two separate scaffolds. CDH3, a 12

gene commonly found in the region of CDH1 and TMCO7 genes in tetrapods, is not annotated in L. 13

chalumnae. 14

15

Figure 3. Phylogenetic trees of the TTR/HIUase family. Left: Bayesian Inference (midpoint 16

rooting, posterior probabilities > 95); right: Maximum Parsimony (midpoint rooting, bootstrap 17

majority consensus values > 50; length tree = 772). The trees show two main clades corresponding 18

to HIUase and TTR proteins, respectively. HIUase gene duplication in coelacanth appears to be 19

lineage-specific. 20

21

Figure 4. Expression levels of genes of purine catabolism enzymes. A) Expression levels of the 22

HIUase A and HIUase B genes in L. menadoensis liver and testis. The different expression profiles 23

of the two forms suggest a newly acquired function for HIUase B. B) Expression levels of purine 24

catabolism enzyme transcripts in L. menadoensis liver. Values are expressed as FPKM. The 25

Page 17 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

18

expression of the housekeeping gene phosphoglycerate kinase (PGK) in liver is also reported for 1

comparison. 2

3

Figure 5. Multiple alignment of UOX C-termini in representative vertebrate species. The conserved 4

PTS1 sequences are in bold (S/A/C)(K/R/H)(L/M). Coelacanths display an intact PTS1 suggesting a 5

peroxisomal location for the UOX enzyme. 6

7

Page 18 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

19

SUPPORTING INFORMATION 1

Table S1. Accession number list of the sequences used in the phylogenetic analyses and for UOX 2

multiple sequence alignment. 3

Figure S1. Phylogenetic analysis of UOX. Bayesian inference of UOX amino acid sequences. 4

Numbers on branches indicate posterior probability values (> 95). Amino acid model: mixed (Jones 5

posterior probability: 0.892; WAG posterior probability: 0.108). The Branchiostoma floridae 6

sequence was the out-group. 7

Figure S2. Phylogenetic analysis of PRHOXNB. Bayesian inference of PRHOXNB amino acid 8

sequences. Midpoint rooting. Numbers on branches indicate posterior probability values (> 95). 9

Amino acid model: Jones. 10

Figure S3. Phylogenetic analysis of ALN. Bayesian inference of ALN amino acid sequences. 11

Numbers on branches indicate posterior probability values (> 95). Amino acid model: WAG. The 12

Strongylocentrotus purpuratus sequence was the out-group. 13

Figure S4. Phylogenetic analysis of ALC. Bayesian inference of ALC amino acid sequences. 14

Numbers on branches indicate posterior probability values (> 95). Amino acid model: mixed (WAG 15

posterior probability: 0.999; Jones posterior probability: 0.001). The Branchiostoma belcheri 16

tsingtauense sequence was the out-group. 17

Figure S5. Microsyntenies of UOX. Arrows indicate the transcriptional orientation, brackets 18

indicate the arrangement of distant loci on the same chromosome. 19

Figure S6. Microsyntenies of PRHOXNB. Arrows indicate the transcriptional orientation. 20

Figure S7. Microsyntenies of ALN. Arrows indicate the transcriptional orientation, brackets 21

indicate the arrangement of distant loci on the same chromosome. ALN of Anolis carolinensis (not 22

annotated in Ensembl) was identified by homology and is delineated by the coordinates reported in 23

the figure. 24

Page 19 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

20

Figure S8. Microsyntenies of ALC. Arrows indicate the gene orientation, brackets indicate the 1

arrangement of distant loci on the same chromosome. 2

Page 20 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

Table 1. L. menadoensis sequence annotation

Accession number Length p-distance1 Length

PI

2 MW

3 Half-life

4

UOX HF678434 4,180 bp 0.141 306 aa 6.90 35.33 30 h

HIUase A HF678438 2,976 bp 0.067 141 aa 9.48 15.75 30 h

PRHOXNB HF678435 2,858 bp 0.182 175 aa 6.22 19.98 30 h

ALN HF678436 1,888 bp 0.053 464 aa 6.90 51.20 30 h

ALC HF678437 1,410 bp 0.146 404 aa 6.58 45.33 30 h 1Percentage p-distance between L. menadoensis and L. chalumnae nucleotide sequences;

2PI:

inferred protein isoelectric point; 3MW: inferred protein molecular weight;.

4Half-life: estimated

half-life calculated on the “N-ter rule” through ProtParam (Gasteiger et al., 2005) on the inferred

proteins.

Page 21 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

1

Table 2. ω ratio between L. menadoensis and other vertebrate sequences

UOX HIUase PRHOXNB ALN ALC

Zebrafish 0.05 - 0.10 0.09 0.07

Puffer fish 0.05 0.28 0.11 0.08 0.08

Xenopus 0.07 0.13 0.08 0.18 0.06

Platypus 0.07 0.30 0.10 0.31 0.07

Mouse 0.06 0.11 0.10 - 0.081

1 ALC in mouse is non-functional.

Page 22 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

The purine catabolism pathway. The terminal purine catabolism step is not identical in all taxonomic groups: the end product of this pathway in reptiles, birds, hominoid primates, humans, and some platyrrhini is urate, allantoin in placental mammals (other than primates) and catarrhini, and allantoate in some teleosts. In

contrast, amphioxus, elasmobranchs, other teleosts, dipnoans, amphibians, monotremes, marsupials, and coelacanths display a "complete" pathway that produces urea. Vertebrate taxonomic groups are reported on the left of their end product (Keilin, ’59; Urich, ’94; Hayashi et al. 2000, Keebaugh and Thomas, 2009).

Enzymes are reported on the right between their substrate and product. 209x297mm (300 x 300 DPI)

Page 23 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

HIUase microsyntenies in L. chalumnae and comparison with orthologous loci for other vertebrates. Coelacanth HIUase A and HIUase B loci are found on two separate scaffolds. CDH3, a gene commonly found

in this region in tetrapods, is not annotated in L. chalumnae.

209x297mm (300 x 300 DPI)

Page 24 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

Phylogenetic trees of the TTR/HIUase family. Left: Bayesian Inference (midpoint rooting, posterior probabilities > 95); right: Maximum Parsimony (midpoint rooting, bootstrap majority consensus values >

50; length tree = 772). The trees show two main clades corresponding to HIUase and TTR proteins,

respectively. HIUase gene duplication in coelacanth appears to be lineage-specific. 209x297mm (300 x 300 DPI)

Page 25 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

Expression levels of genes of purine catabolism enzymes. A) Expression levels of the HIUase A and HIUase B genes in L. menadoensis liver and testis. The different expression profiles of the two forms suggest a newly

acquired function for HIUase B. B) Expression levels of purine catabolism enzyme transcripts in L.

menadoensis liver. Values are expressed as FPKM. The expression of the housekeeping gene phosphoglycerate kinase (PGK) in liver is also reported for comparison.

209x297mm (300 x 300 DPI)

Page 26 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960

For Peer Review

Multiple alignment of UOX C-termini in representative vertebrate species. The conserved PTS1 sequences are in bold (S/A/C)(K/R/H)(L/M). Coelacanths display an intact PTS1 suggesting a peroxisomal location for

the UOX enzyme.

209x297mm (300 x 300 DPI)

Page 27 of 27

John Wiley & Sons

JEZ Part B: Molecular and Developmental Evolution

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960