Coelacanth Research South African Journal of Science 102, September/October 2006 479

Molecular biology studies on the : a review

Keoagile W. Modisakenga, Chris T. Amemiyab, Rosemary A. Dorringtona* and Gregory L. Blatcha*

Tanzania and Kenya.8 Investigations into the relatedness of the The discovery of the African coelacanth in 1938 and subsequently East African using mitochondrial DNA sequence the Indonesian coelacanth in 1998 has resulted in a keen interest in and microsatellite DNA from 47 coelacanth individuals suggested molecular studies on the coelacanth. A major focus has been on the that the scattered groups of African coelacanths might have phylogenetic position of the coelacanth. Lobe-finned fish such as come from a single remote population, possibly the Comoros or 8 the coelacanth are thought to be at the base of the evolutionary other, unknown habitats in the . branch of fish leading to tetrapods. These studies have further aimed to resolve the phylogenetic relationship of extant lobe-finned Preservation of tissue material fish (two coelacanth species and the lungfishes) to vertebrates. The population size of the Comoran coelacanths was estimated Notwithstanding the lack of readily accessible good-quality coela- at 200 individuals in 1994.9 An independent survey of the canth tissue, several major contributions to coelacanth molecular Comoran population between 1991 and 1994 showed a decrease studies and biology have been possible. The mitochondrial from an average of 20.5 to 6.5 individuals in all surveyed under- genome sequences of both species of the coelacanth suggest that water caves.10 Despite the capture of over 180 coelacanths,1 they diverged from one another 40–30 million years ago. A number proper tissue preservation has been problematic. The main of large gene families such as the HOX, protocadherin and heat difficulty is that the habitat and distribution of both African and shock protein clusters have been characterized. Furthermore, the Indonesian coelacanths coincides with some of the poorest and recent successful construction of a large-insert (150–200 kilobase) most undeveloped communities in the world. Consequently, genomic library of the Indonesian coelacanth will prove to be an most of the stored tissue specimens are unsuitable for molecular invaluable tool in both comparative and functional genomics. Here research because of the time lag between an accidental catch and we summarize and evaluate the current status of molecular the preservation of the tissue, and the use of unsuitable preser- research, published and databased, for both the African ( vation methods. Some of the best-preserved material was chalumnae) and the Indonesian (Latimeria menadoensis) coelacanth. reported in 1992, which included tissue from a gravid female with 26 late-term fetuses caught off the coast of Mozambique. Frozen tissue samples from this catch were donated to the J.L.B. Smith Institute of Ichthyology in Grahamstown, South Africa, From discovery of individuals to population biology and the Max-Plank-Institut für Verhaltensphysiologie in Coelacanths were thought to have become extinct about 80–70 Seewiesen, Germany.11 million years ago (Myr), when they disappeared from the fossil Another difficulty facing functional studies on the coelacanth record. Preceding this date, there appears to have been around is the lack of a coelacanth-specific biological assay system. The 80 species of coelacanths in 40 genera and four families during impact of the lack of such an experimental system is further the period extending from 350 to 70 Myr.1 In 1938, a coelacanth discussed under the section on genes encoding heat shock was trawled near East London, South Africa, and named protein 70 (Hsp70 proteins are encoded by the hsp70 genes). Latimeria chalumnae after its discoverer, Marjorie Courtney- Latimer.2 A second was caught off Anjouan Island in the Etymology and evolutionary significance Comoros in 1952 followed by more than one hundred subse- The genus name Coelacanthus means ‘hollow spine’ and refers quent catches in the region.1,3 to the hollow neural and haemal spines of the vertebrae that Coelacanths have also been discovered off the coast of Manado connect to the tubular bones supporting the upper and lower Tua Island, , . In September 1997, Mark caudal-fin rays.1 Notably all coelacanths, extinct and extant, Erdmann and his wife saw and photographed a fish that was possess these hollow spines, as well as other unique features later identified as a coelacanth in the fish market in Manado Tua. compared to other living fishes. These include the extra lobe on Subsequent to this sighting, the second Indonesian coelacanth their tail, paired lobed fins, a vertebral column that is not was caught north of Manado Tua and was named Latimeria completely developed, and a fully functional intracranial joint.12 menadoensis.4,5 Two further Asian coelacanths were later sighted The coelacanth is a good example of a species that has evolved by Fricke and co-workers southwest of Manado Tua.6 and adapted while retaining the characteristic features of its The first discovery of a live coelacanth population off the ancestors. Evidence of the minor evolutionary change observed South African coast occurred in November 2000, when recre- in the coelacanth can be found in its skeleton. The skeleton of the ational divers found three coelacanths in the underwater Jesser extant coelacanth is almost identical to that of Macropoma (Upper 1 Canyon at Sodwana Bay at a depth of 117 m. Surveys by the Cretaceous period: ~80 Myr) and Laugia (Lower Triassic period: African Coelacanth Ecosystem Programme (ACEP) have identi- ~230 Myr) (Fig. 1).12 Several ad hoc hypotheses have sought to fied 25 coelacanth individuals.7 Other African coelacanth catches explain the slow change throughout the history of the coela- have been reported off the coasts of Mozambique, Madagascar, canth. One such explanation came from the Darwinian view aDepartment of Biochemistry, Microbiology and Biotechnology, Rhodes University, that high-level adaptation to certain environments and lack of Grahamstown 6410, South Africa. 12 b competition may have led to morphological uniformity. Molecular Genetics Program, Benaroya Research Institute at Virginia Mason, 1201 1 Ninth Avenue, Seattle, Washington 98101, U.S.A. Heemstra suggests that Latimeria fits Darwin’s description of a *Authors for correspondence. E-mail: [email protected] and [email protected] living fossil, since it is the only extant member of the coelacanths, 480 South African Journal of Science 102, September/October 2006 Coelacanth Research

Fig. 2. Three possible phylogenetic relationships of tetrapod, lungfish and coelacanth lineages. (A) The lungfish is the sister group to tetrapods; (B) the coelacanth is the sister group to tetrapods; and (C) the coelacanth and lungfish are sister groups and equally distant from tetrapods.13–15

genetic position of the coelacanth within the sarcopterygian lineage (Fig. 2): (a) the lungfish is the sister group to tetrapods; (b) the coelacanth is the sister group to tetrapods; or (c) the coelacanth and lungfish are sister taxa and equally distant from tetrapods.13–15 Most molecular biology research on the coelacanth has been aimed at resolving the phylogenetic classification of the coela- canth. These include studies on genes encoding coelacanth α and β haemoglobin proteins, 18S rRNA and 28S rRNA, cytochrome oxidase subunit 1, major histocompatibility complex (MHC) class 1, 12S rRNA and 16S rRNA, the enolase protein family and cytochrome b (see Table 1 in online supplementary material).16–24 A close tetrapod–coelacanth relationship was initially surmised on the basis of phylogenetic analysis of haemoglobin sequences. Among 522 haemoglobin sequences analysed, the β chains of the coelacanth and the α chains of the Rana catesbeiana tadpoles had the highest match. These data supported a sister-group relation- ship between the coelacanth and tetrapods [hypothesis (b), Fig. 2B],16 but this conclusion was vigorously rejected by other researchers who re-analysed the data and concluded that the coelacanth–tetrapod relationship was not supported.25 An attempt to solve the coelacanth phylogenetic position using comparative analysis of 18S rRNA sequences resulted in a tree with none of its relationships significantly supported at 95% Fig. 1. The skeletons of coelacanths depicting their conserved morphology. Differ- confidence. The 18S rRNA was therefore deemed as insufficient ences in the morphologies include the increased lobulation in the second dorsal fin to address the question of coelacanth relations.24 In contrast, a and anal fin. Latimeria (extant coelacanth) has a fat-filled organ in place of the comparative study of the genes encoding 12S rRNA and gas-filled bladder found in Macropoma. However, the basic skeletal structure of the extant coelacanth remains unchanged relative to Macropoma and Laugia (figure cytochrome b from the coelacanth, tetrapods, lungfish and frog adapted from ref. 12). supported the sister-group relationship of the lungfish and tetrapods [hypothesis (a), Fig. 2A].24,26 Using vertebrate 28S rRNA which was abundantly found 290–208 Myr ago. Furthermore, sequences, Hillis and co-workers18 provided evidence for a the coelacanth inhabits rocky caves at 100–300 m below sea level, sister-group relationship between the coelacanth and tetrapods; 1 where competition is presumably not severe. An alternative however, this analysis was incomplete since it did not include approach to investigating the slow evolutionary change might the lungfish 28S rRNA. The re-analysis of the 28S rRNA data set 12 be through molecular genetics. However, the application of including the lungfish 28S rRNA supported the hypothesis (c), genetics to derive a clear relationship between coelacanth evolu- that the coelacanth and lungfish are sister taxa and equally tion to generation time or differential lineage-specific mutation distant from tetrapods.19 It was obvious at this stage that a much 12 rate necessitates more research on the coelacanth genome. larger data set would be needed to resolve the phylogenetic classification of the coelacanth. Phylogenetic studies The complete mitochondrial genome sequences of both coela- The evolutionary relationships of gnathostomes (jawed verte- canth species have been successfully used to provide evidence brates) have been hotly debated for over a century. Gnathostomes for their relatedness to one another. In 1997, Zardoya and fall into two major taxa: the Chondrichthyes (cartilaginous Meyer27 reported a complete DNA sequence of the L. chalumnae fishes) and the Osteichthytes (bony fishes). Bony fishes include mitochondrial genome. This study revealed that the general actinopterygians (ray-finned fish) and sarcopterygians (coela- codon usage of the coelacanth was similar to that of the lamprey, canths, lungfish and tetrapods).13 bichir, carp, trout, lungfish and tetrapods with adenosines Currently, there are three contested hypotheses on the phylo- and cytosines preferentially used at the third codon position. Coelacanth Research South African Journal of Science 102, September/October 2006 481

Comparison of the recently completed L. menadoensis mitochon- evolutionary history of tetrapod genomes.36 The protocadherins drial sequence with that of L. chalumnae revealed a 4.28% differ- are synaptic cell adhesion molecules thought to provide a molec- ence between the two coelacanth species and provided an ular code involved in the generation of synaptic complexity in estimate of the divergence time between the two species at the developing brain. The copy number of the protocadherin 40–30 Myr.28 This estimate of divergence time was different from genes and their sequence in vertebrates could be an indication of the 6.3–4.7 Myr estimate by Holder and co-workers.29 The latter adaptive differences in protocadherin function with reference to study, however, was only based on mitochondrial fragments level of organism complexity.37 The coelacanth has been found containing the cyt b and two rRNA genes. to have 49 protocadherin-encoding genes clustered in the same Phylogenetic analyses using the complete L. chalumnae mito- manner as the 54 protocadherin-encoding genes in humans. chondrial DNA sequence could not clearly exclude any of the These results, including statistical comparisons of rates of molec- three hypotheses on the tetrapod, lungfish and coelacanth ular evolution, strongly suggest that the coelacanth proto- relationship; however, it did place the coelacanth closer to cadherin genes have not undergone accelerated rates of tetrapods and other sarcopterygians than to ray-finned fish.27 molecular evolution as is the case with teleost fishes. If the The inability of the coelacanth mitochondrial genomes to protocadherin results are indicative of the entire genome, the resolve the tetrapod, lungfish and coelacanth phylogenetic coelacanth could be used as a genome outgroup in tetrapod relationship intensified the search for answers in the chromo- comparative sequence analyses and, thus, is a valid candidate somal genome. Phylogenetic analysis using the amino-acid for whole-genome sequencing.36 sequence of the myelin DM20 protein supported hypothesis (a) (Fig. 2A), that the lungfish and not the coelacanth is the closest Hox genes relative of tetrapods.30 Phylogenetic analysis of molecular The Hox genes encode transcription factors engaged in speci- markers, RAG1 and RAG2 genes, also supported a sister relation- fying body plans of metazoans.35 The Hox genes are important ship between the lungfish and tetrapods.14 not only because of their central developmental function, but Another recent attempt to clarify the tetrapod, lungfish, and also because they offer insights into genome duplication events coelacanth relationship came from Takezaki and co-workers15 that occurred during the evolution of modern tetrapods and who analysed amino-acid sequences from 44 protein coding ray-finned fishes.38 All invertebrates so far investigated have nuclear genes. Data from this analysis, however, did not conclu- been shown to have a single HOX cluster as opposed to three to sively support any one of the three hypotheses, leading to the eight clusters in the major taxa of vertebrates.38,39 It has also been conclusion that the divergence between the coelacanth, lungfish shown that mammals and tetrapods have four Hox gene clusters and tetrapods probably occurred in a very short period, thereby and the sea lamprey (Petromyzon marinus), a jawless vertebrate making their phylogenetic resolution difficult. The authors fur- (Agnatha), has three or four HOX clusters, each cluster contain- ther proposed that the short divergence period between the ing up to 13 paralogous groups (Fig. 3).40–43 Comparative analysis three lineages necessitated an increase in the number of loci of these genes and gene clusters has often been used to exem- used for such analyses. To resolve these phylogenetic relation- plify the notion that at least two rounds of genome duplication ships with more confidence there is a need to conduct in the preceded modern mammals.44,45 Given the evolutionary signifi- short to medium term an analysis of other informative molecular cance of the coelacanth, Koh and co-workers used an extensive markers (loss and gain of introns, indels, genomic rearrange- PCR survey with degenerate primers to characterize the L. mena- ments such as inversions, etc.), and in the long term, whole- doensis HOX clusters and their gene complement in an attempt genome sequencing.13,14 to partially answer the questions regarding genome duplication in the coelacanth and to probe the origins of tetrapod limbs.46 The nuclear genome L. menadoensis was found to have 33 genes out of a possible 52 The coelacanth has a 48-chromosome karyotype that contains genes organized into four clusters. Although the gene comple- metacentric, subtelocentric and telocentric chromosomes, as ment was lower than the 39 genes normally found in mammals, well as microchromosomes. This karyotype is unlike that of the the coelacanth Hox genes were more similar in sequence to lungfish (which has 36–38 large metacentric chromosomes), but mammalian Hox genes than to that of ray-finned fish. Even very similar to the 46-karyotype of the Ascaphus truei, an ancient though the strategy used by Koh and co-workers could not frog.31 Flow cytometric analysis of Comoran coelacanth (L. confirm the absence of Hox genes that were not obtained, or chalumnae) blood estimated the genome size of this species of the the absolute arrangement of the genes within the respective coelacanth at 2.75 pg per 1C (or ~2.75 × 109 bp; C.T. Amemiya, clusters, it provided good support for the notion of two rounds unpubl. data). This estimate is substantially smaller than the of genomic duplication during vertebrate evolution.46 In a sepa- 3.6 pg per 1C (or ~3.6 × 109 bp) estimated from prior studies.31,32 rate report, the L. menadoesis Hoxc8 early enhancer was isolated A coelacanth (L. menadoensis) genomic library with average and shown to bear a close resemblance to the mouse early insert size of 171 kb has been successfully constructed using the enhancer in both structure and function.34,35 bacterial artificial chromosome (BAC) system. The representa- Recent sequence analysis of the L. menadoensis HOX-A cluster tion of the library is estimated at ≥7 genome-equivalents.32 The and the horn shark (Heterodontus francisci) HOX-D cluster successful construction of this Indonesian coelacanth genomic revealed the presence of additional paralogous Hox14 genes, library is expected to bring new insight to both comparative and named Hoxa14 and Hoxd14, respectively.33 Although the functional genomics. Data and lines of investigation that have cephalochordate amphioxus (closest extant relative of vertebrates) come out of this resource include the discovery of the L. has a single HOX cluster with a group 14 gene, the discovery of menadoensis Hoxa14 gene33, the comparative and functional the horn shark and coelacanth paralogous Hox14 genes is the analysis of the L. menadoensis Hoxc8 early enhancer34,35 and the first in vertebrates (Fig. 3).33 While the coelacanth and horn shark characterization of L. menadoensis hsp70 genes (hsp70-containing Hox14 genes occur in different clusters (A and D, respectively), BAC identification32; Modisakeng et al., unpubl.). The study of the similarity between them is strongly suggestive of the the protocadherin-encoding gene has been used to evaluate the existence of the group 14 genes in the common ancestor of the possible utility of the coelacanth genome sequence to infer the gnathostomes.33 This finding strongly challenges the traditional 482 South African Journal of Science 102, September/October 2006 Coelacanth Research

Fig. 3. Evolution of vertebrate HOX clusters. (A) The structure of the amphioxus HOX cluster. The cluster has 14 Hox genes and two Evx genes. The genes at the 3’ end of the cluster pattern the anterior of the organism’s body while the 5’ and central genes pattern the posterior and central regions of the body, respectively.38(B) Duplication of HOX clusters and genes in the vertebrate lineage. The occurrence of one cluster in cephalochordates and four in chondrichytes, crossopterygians and tetrapods supports the hypothesis of two rounds of duplication. The occurrence of more than four clusters in higher teleosts is thought to be the result of an independent genome duplication within the teleost radiation.39 When assuming the 14-gene HOX cluster in the vertebrate stem lineage, each vertebrate HOX cluster can have a maximum of 14 genes. The amphioxus cluster possesses all 14 genes whereas vertebrate clusters have lost some genes. Assuming that the phylogeny is correct, teleosts and tetrapods are thought to have independently lost their Hox14 genes.38 view that the protovertebrates possessed only 13 Hox genes per different Hsp70-encoding genes.50 Using degenerate primers HOX cluster.33 In an attempt to answer questions on the possible three hsp70 partial fragments were amplified from L. chalumnae occurrence of more Hox genes (e.g. Hox15), an analysis genomic DNA. One of the three fragments was found to be a of the amphioxus HOX cluster (650 kb), from Hox1 to Evx, was 1840-bp intronless open reading frame encoding most of an L. conducted. The data from this study suggested that, based on chalumnae Hsp70 (LcHsp70).50 Amemiya’s laboratory screened the archetypal character of a single amphioxus HOX cluster (14 the Indonesian coelacanth genomic BAC library with a 200-bp genes per cluster, Fig. 3A), it is improbable that more paralogous African coelacanth hsp70 probe and identified 26 putative Hox genes will be discovered in .38 Recent analyses of hsp70-containing BACs.32 Preliminary analysis of these BACs complete HOX cluster sequences from the coelacanth and shark suggested that L. menadoensis also contained an intronless hsp70 have borne this out (C.T. Amemiya, unpubl.). gene that was almost identical to the intronless L. chalumnae hsp70 gene (Modisakeng, unpubl. data). Since intronless hsp70 Genes encoding heat shock protein 70 genes from eukaryotes, including fish such as tilapia, have been Molecular chaperones are a class of proteins that have evolved shown to be heat inducible,51 the coelacanth very likely contains to help other proteins fold, and to refold misfolded or unfolded at least one inducible hsp70 gene. The elucidation of the number, proteins.47 Some heat shock proteins, such as the well-character- organization and regulation of coelacanth hsp70 genes will ized heat shock protein 70 (Hsp70), are molecular chaperones. enable a comparison to other fish hsp70 gene clusters (e.g. that of Hsp70 facilitates the folding of nascent polypeptides, and the Fugu rubripes),52 and an analysis of the functional importance of refolding of denatured proteins after stress in an ATP-dependent these genes.53 manner. The genes encoding Hsp70s occur in multiple copies in The lack of a coelacanth experimental system (e.g. tissue eukaryotic genomes and their expression is induced by a num- culture cell lines) will necessitate the use of heterologous ber of stress stimuli including heat, pathophysiological stress, systems such as bacterial (e.g. Escherichia coli), yeast (e.g. growth, and development.48,49 The diverse roles of Hsp70s in cell Saccharomyces cerevisiae) or mammalian (human) systems to and developmental biology, and the fact that the coelacanth has carry out functional studies. The heterologous production of survived millions of years without significant morphological LcHsp70 in E. coli was not possible without the co-production of changes, makes the study of the genomic organization and regu- tRNAs for codons rarely used in E. coli (tRNAs for AGA and AGG lation of the coelacanth hsp70 genes interesting. The African coe- for Arg (R), AUA for Ile (I) and GGA for Gly (G) using the RIG lacanth’s genome has been shown to contain at least three plasmid developed by Baca and co-workers;54 K.W.Modisakeng, Coelacanth Research South African Journal of Science 102, September/October 2006 483

Fig. 4. Comparison of codon usages from Latimeria menadoensis (Lm), Homo sapiens (Hs), Saccharomyces cerevisiae (Sc) and E. coli (Ec). In general, the L. menadoensis codon usage patterns followed similar trends to those of the two other eukaryotes, but showed significant differences to that of E. coli (see text for details). L. menadoensis, Homo sapiens, S. cerevisiae and E. coli codon usage tables were derived from http://www.kazusa.or.jp/codon (accessed on 14 August 2005). unpubl. data). Analysis of the codon usage of L. chalumnae and known predators, and very little competition; and lastly, the L. menadoensis genes (Fig. 4, Table 2 in online supplementary availability of data from prior studies on vertebrate MHC genes. material), revealed that in general coelacanth genes used the These feature make the MHC genes good candidates for codons AGA, AGG, AUA and GGA with greater frequency than comparative studies when attempting to determine the tempo did genes from E. coli, and that one of the most rarely used and nature of changes influencing the slow evolution of the codons in E. coli (AGG for R) was a frequently used codon in coelacanth.21 The coelacanth has at least one functional MHC coelacanth genes. When comparing the codon usage patterns of gene encoding a class I complex, suggesting that the unique human and yeast genes to that of coelacanth genes, there did not environment of the coelacanth has not led to the underdevelop- appear to be any major disparities in codon usage, and in those ment of its immune system.21 The authors further showed that cases where there was rare codon usage in human and yeast the coelacanth MHC exon-intron layout was similar to that of genes (e.g. CGA for R), coelacanth genes also showed low or rare mammals. In terms of DNA sequence similarity, the coelacanth usage (Fig. 4). This suggests that coelacanths, like other MHC sequence was much closer to amphibian MHC genes than eukaryotes, may utilize codon usage-based translational pausing actinopterygian class I genes. These findings supported prior for regulated and appropriate protein synthesis kinetics. findings that the coelacanth possesses functional sets of genes encoding immunoglobulin proteins.55 However, the data on its Immune system immunoglobulin heavy chain gene locus suggest a unique The MHC proteins are the first receptors in the specific arrangement that, while utilizing the cardinal components of the immune system that come into contact with peptides of immunoglobulin system, is organized in a completely different processed foreign protein.21 Benz and co-workers conducted fashion from any other gnathostome.55,56 studies on coelacanth genes encoding MHC class I proteins in view of three characteristics of coelacanths: first, its interesting Adaptation of vision to the deep-sea environment evolutionary and phylogenetic position; second, the fact that the In their natural habitat, coelacanths receive only a narrow Comoran coelacanths live at depths of 200 m where they have no band of light (at approximately 480 nm).57–59 Both species of the 484 South African Journal of Science 102, September/October 2006 Coelacanth Research

coelacanth have been shown to use rhodopsin (RH) 1 and 2 Concluding remarks pigments to detect colour within a narrow range. RH1 and RH2 Contrary to the belief that the discovery of the living coela- genes have been cloned and characterized from both species of canth was going to solve the mysteries of the origin of tetrapods, the coelacanth.57–59 These studies demonstrated that both species studies have revealed the divergence of the coelacanth, lungfish have identical RH1 and RH2 pigments with optimum light and tetrapod lineages occurred in a short time interval, making λ sensitivities ( max) at 485 and 479 nm, respectively, in keeping their phylogenetic relationships difficult to resolve. Lack of with a narrow colour range.57–59 Compared to corresponding coelacanth tissue material that has been properly preserved has orthologous pigments, the maximum absorption of coelacanth limited the quality of molecular biology research. Several major RH1 and RH2 pigments is shifted by 20 nm towards the blue.57 contributions to the molecular biology of the coelacanth have, Ten amino-acid changes have been shown to shift the maximum however, been achieved. Although complete mitochondrial absorption of visual pigments by more than 5 nm. Among these, genome sequences could not conclusively resolve the relation- D83N, E122Q, M207L and A292S are prevalent in shifting ship of the coelacanth with tetrapods and lungfishes, the data maximum absorption of RH1 and RH2 pigments in vertebrates. suggested placement of the coelacanth closer to tetrapods E122Q/A292S and E122Q/M207L can be used to explain the than to ray-finned fish. The shortcomings of the mitochondrial 20 nm shift in the maximum absorption of coelacanth RH1 and genome analyses necessitated the usage of large data sets of RH2 pigments, respectively.57–58 Most use ultraviolet nuclear genes to investigate the coelacanth phylogenetic rela- (UV) light for activities such as foraging, mate selection, and tionships further. These analyses did not yield any unequivocal communication.60 Short-wavelength-sensitive type 1 (SWS1) phylogenetic hypotheses. Having failed to resolve the tetrapod, pigments absorbing light maximally at 360 nm mediate UV lungfish and coelacanth relationship using 44 nuclear genes, vision.60 When UV light is unavailable or not necessary, the gene Takezaki and co-workers15 proposed the use of whole-genome encoding the SWS1 pigments can be lost or become nonfunc- sequences. The successful construction of the Indonesian coela- tional.60 The coelacanth appears to have lost the SWS1 pigments canth BAC genomic library is undoubtedly one of the major and further modified the RH1 and RH2 pigments to detect light stepping-stones towards a revolution in coelacanth molecular between 479–485 nm.57 biology. The discovery of the Hox14 gene in the coelacanth exem- plifies the importance of the coelacanth genome resource.33 In addition, the finding that the coelacanth has 49 protocadherin- The endocrine system encoding gene clusters organized in the same manner as the Proopiomelanocortin (POMC) is the precursor of a group of 54-protocadherin-encoding gene clusters in humans, strengthens β proteins, such as adrenocorticotropin (ACTH), -lipotropin the case for the coelacanth as a candidate for whole-genome β β ( -LPH), melatropin and -endorphin (EP), which are closely sequencing whereby it would serve as an effective outgroup for associated with response to stress and environmental adapta- tetrapod comparative studies.36 The growing need for coelacanth 61 tion. The diversity of POMC is suggested to have occurred as a primary cell lines was highlighted by differences in codon usage result of the POMC gene evolution events including POMC as a potential major problem associated with the heterologous gene internal duplication, insertion and deletion in the region production of coelacanth proteins in E. coli. The development of encoding melanocyte-stimulating hormone (MSH) domains. coelacanth cell lines will greatly enhance functional genomics The evolutionary significance of POMC also comes from the fact and proteomics studies, especially the isolation of mRNA for the that it is the first gene of the adenohypophysial hormones construction of cDNA libraries, and mapping the functional 61 shown to be present in all classes of vertebrates. genes on the coelacanth genome. However, the ultimate Little is known about the components of the coelacanth’s resource would be mRNA (and the cDNA derivatives) extracted endocrine system. However, structural characterization of from various tissues of the coelacanth. hormones derived from the POMC of L. chalumnae has been reported.62 The interest in the structure of the hormones derived This research was funded by the South African Department of Science and Technology (administered through the South African Institute for Aquatic from the coelacanth POMC was sparked by observations that Biodiversity [SAIAB]). We are grateful to the German Student Exchange Services the coelacanth’s pituitary basal structure was significantly (DAAD) and the National Research Foundation for study scholarships to K.W.M. different from that of the ray-finned fish, cartilaginous fish and C.T.A. has been funded by grants from the National Science Foundation and the 62 lampreys. Reverse-phase high performance liquid chromatog- National Institutes of Health. raphy, amino-acid sequencing and mass spectroscopic analysis of an extract from the pituitary gland identified several MSHs 1. Heemstra P.(2001). New coelacanth discovery: Sodwana’s living fossils. Archi- including alpha-MSH, N-Des-acetyl-alpha-MSH, beta-MSH, medes 43, 13–17. and an N-terminal peptide containing gamma-MSH. In addi- 2. Smith J.L.B. (1939). A living fish of Mesozoic type. Nature 3620, 455–456. tion, a corticotrophin-like intermediate lobe peptide (CLIP) and 3. Smith J.L.B. (1953). The second coelacanth. Nature 171, 99–101. N-acetyl-beta-endorphin (END) were identified.62 Phylogenetic 4. Erdmann M.V.,Caldwell R.S. and Moosa M.K. (1998). The Indonesian ‘king of the sea’ discovered. Nature 395, 335. analysis of these peptides revealed the coexistence of putative 5. Pouyaud L., Wirjoatmodjo S., Rachmatica I., Tjakrawidjaja A., Hadiaty R. tetrapod-type and fish-type molecules in the coelacanth POMC and Hadie W. (1999). A new species of the coelacanth. C. R. Acad. Sci. III 322, perhaps indicative of the coelacanth’s phylogenetic intermediate 261–267. position between fish and tetrapods. For instance, the 6. Fricke H., Hissmann K., Schauer J., Erdmann M., Moosa M.K. and Plante R. (2000). Biogeography of the Indonesian coelacanth. Nature 403, 38. gamma-MSH and CLIP molecules showed high similarity to am- 7. Sink K.J. and Scott-Williams N. (2005). African Coelacanth Ecosystem Programme phibians and birds, while the beta-END was most similar to the ROV Trial Report. 25 May 2005 (http://acep.co.za/content/view/276/118/). sturgeon peptide and the alpha-MSH was highly similar to 8. Schartl M., Hornung U., Hissmann K., Schauer J. and Fricke H. (2005). Related- mammalian alpha-MSH. The conclusion drawn from the coela- ness among east African coelacanths. Nature 435, 901. canth POMC study was that the molecular design of the coela- 9. Fricke H. and Hissmann K. (1994). Home range and migrations of the living coelacanth Latimeria chalumnae. Mar. Biol. 120, 171–180. canth putative POMC closely resembled the tetrapod’s POMC 10. Fricke H. and Hissmann K. (1990). Natural habitat of the coelacanths. Nature whereas its amino-acid sequences are a mixture of types found 346, 323–324. in fish and tetrapods.62 11. Fricke H. (1992). Coelacanth tissue bank. Nature 357, 102. Coelacanth Research South African Journal of Science 102, September/October 2006 485

12. Forey P.L.(1988). Golden jubilee for the coelacanth Latimeria chalumnae. Nature Amemiya C.T. and Garcia-Fernàndez J. (2005). No more than 14: the end of the 336, 727–732. amphioxus Hox cluster. Int. J. Biol. Sci. 1, 19–23. 13. VenkateshB., Erdmann M.V.and Brenner S. (2001). Molecular synapomorphies 39. Wagner G.P., Amemiya C. and Ruddle F. (2003). Hox cluster duplications and resolve evolutionary relationship of extant jawed vertebrates. Proc. Natl Acad. the opportunity for evolutionary novelties. Proc. Natl Acad. Sci. USA 100, Sci. USA 98, 11382–11387. 14603–14606. 14. Brinkmann H., Venkatesh B., Brenner S. and Meyer A. (2004). Nuclear protein 40. Garcia-Fernàndez J. and Holland P.W.H. (1994). Archetypal organization of the coding genes support lungfish and not the coelacanth as the closest living amphioxus Hox gene cluster. Nature 370, 565–566. relatives of land vertebrates. Proc. Natl Acad. Sci. USA 101, 4900–4905. 41. Holland P.W.H. and Garcia-Fernàndez J. (1996). Hox genes and chordate evolu- 15. Takezaki N., Figueroa F., Zaleska-Rutczynska Z., Takahata N. and Klein J. tion. Dev. Biol. 173, 382–395 (2004). The phylogenetic relationship of tetrapod, coelacanth, and lungfish 42. Irvine S.Q., Carr J.L., Bailey W.J.,Kawasaki K., Shimizu N., Amemiya C.T. and revealed by the sequences of forty-four nuclear genes. Mol. Biol. Evol. 21, Ruddle F.H. (2002). Genomic analysis of Hox clusters in the sea lamprey, 1512–1524. Petromyzon marinus. J. Exp. Zool. (Mol. Dev. Evol.) 294, 47–62. 16. Gorr T.,Kleinschmidt T.and Fricke H. (1991). Close tetrapod relationships of the 43. Force A., Amores A. and Postlethwait J.H. (2002). Hox cluster organization in the coelacanth Latimeria indicated by haemoglobin sequences. Nature 351, 394–397. jawless vertebrate Petromyzon marinus. J. Exp. Zool. (Mol. Dev. Evol.) 294, 30–46. 17. Stock, D.W.,Moberg, K.D., Maxson, L.R. and Whitt, G.S. (1991). A phylogenetic 44. Aparicio S. (2000). Vertebrate evolution: recent perspectives from fish. Trends analysis of the 18S ribosomal RNA sequence of the coelacanth Latimeria Genet. 16, 54–56. chalumnae. Environ. Biol. Fishes 32, 99–117. 45. VandepoeleK., De Vos W.,Taylor J.S., Meyer A. and Van der Peer Y.(2004). Major 18. Hillis D.M, Dixom M.T. and Ammerman L.K. (1991). The relationship of the events in the genome evolution of vertebrates: Paranome age and size differ coelacanth Latimeria chalumnae: evidence from sequences of vertebrate 28S considerably between ray-finned fishes and land vertebrates. Proc. Natl Acad. ribosomal RNA genes. Environ. Biol. Fishes 32, 119–130. Sci. USA 101, 1638–1643. 19. Zardoya R. and Meyer A. (1996). Evolutionary relationships of the coelacanth, 46. Koh E.G.L., Lam K., Christoffels A., Erdmann M.V., Brenner S. and VenkateshB. lungfishes, and tetrapods based on the 28S ribosomal RNA gene. Proc. Natl (2003). Hox gene clusters in the Indonesian coelacanth, Latimeria menadoensis. Acad. Sci. USA 93, 5449–5454. Proc. Natl Acad. Sci. USA 100, 1084–1088. 20. Yokobari S., Hasegawa M., Ueda T., Okada N., Nishikawa K. and Watanabe K. 47. Ellis J. (1987). Proteins as molecular chaperones. Nature 328, 378–379. (1994). Relationship among coelacanths, lungfishes, and tetrapods: a 48. Morimoto R.I., Sarge K.D. and Abravaya K. (1992). Transcriptional regulation of phylogenetic analysis based on mitochondrial cytochrome oxidase I gene heat shock genes. J. Biol. Chem. 267, 21987–21990. sequences. J. Mol. Evol. 38, 602–609. 49. Frydman J. (2001). Folding of newly translated proteins in vivo: the role of 21. Benz U.A.K., Mayer W.E.and Klein J. (1994). Major histocompatibility complex molecular chaperones. Ann. Rev. Biochem. 70, 603–647. class I genes of the coelacanth Latimeria chalumnae. Proc. Natl Acad. Sci. USA 91, 11065–11069. 50. Modisakeng K.W, Dorrington R.A. and Blatch G.L. (2004). Isolation of genes 22. Hedges S.B., Hass C.A. and Maxon L.R. (1993). Relations of fish and tetrapods. encoding heat shock protein 70 (hsp70s) from the coelacanth, Latimeria chalum- Nature 363, 501–502. nae. S. Afr. J. Sci. 100, 683–686. 23. Tracy M.R. and Hedges S.B. (2000). Evolutionary history of the enolase gene 51. Molina A., Biemar F., Müller F., Iyengar A., Prunet P., Maclean N., Martial J.A. family. Gene 259, 129–138. and Muller M. (2000). Cloning and expression analysis of an inducible HSP70 gene from tilapia fish. FEBS Lett. 474, 5–10. 24. Meyer A. (1995). Molecular evidence on the origin of tetrapods and the relationships of the coelacanth. Trends Ecol. Evol. 10, 111–116. 52. Lim E.H. and Brenner S. (1999) Short-range linkage relationships, genomic organization and sequence comparisons of a cluster of five HSP70 genes in 25. Stock D.W. and Swofford D.L. (1991). Coelacanth’s relationships. Nature 353, Fugu rubripes. Cell. Mol. Life Sci. 55, 668–678. 217–218. 53. Basu N., Todgham A.E., Ackerman P.A.,Bibeau M.R., Nakano K., Schulte P.M. 26. Meyer A. and Wilson A.C. (1990). Origin of tetrapods inferred from their and Iwama G.K. (2002). Heat shock protein genes and their functional signifi- mitochondrial DNA affiliation to lungfish. J. Mol. Evol. 31, 359–364. cance in fish. Gene 295, 173–183. 27. Zardoya R. and Meyer A. (1997). The complete DNA sequence of the 54. Baca A.M. and Hol W.G.J. (2000). Overcoming codon bias: a method for high- mitochondrial genome of a ‘living fossil,’ the coelacanth (Latimeria chalumnae). level overexpression of Plasmodium and other AT-rich parasite genes in Esche- Genetics 146, 995–1010. richia coli. Int. J. Parasitol. 30, 113–118. 28. Inoue J.G., Miya M., Venkatesh B. and Nishida M. (2005). The mitochondrial 55. Amemiya C.T., Ohta Y., Litman R.T., Rast J.P., Haire R.N. and Litman G.W. genome of Indonesian coelacanth Latimeria menadoensis (: (1993). V gene organization in a relict species, the coelacanth Latimeria Coelacanthiformes) and the divergence time estimation between the two H chalumnae: evolutionary implications. Proc. Natl Acad. Sci. USA 90, 6661–6665. coelacanths. Gene 349, 227–235. 56. Stavnezer J. and Amemiya C.T. (2004). Evolution of isotype switching. Semin. 29. Holder M.T., Erdmann M.V., Wilcox T.P., Caldwell R.L. and Hillis D.M. (1999) Immunol. 16, 257–275. Two living species of coelacanths. Proc. Natl Acad. Sci. USA 96, 12616–12620. 57. Yokoyama S. and Tada T.(2000). Adaptive evolution of the African and Indone- 30. Tohyama Y., Ichimiya T., Kasama-Yoshidi H., Cao Y., Hasegawa M., Kojima H., sian coelacanths to deep-sea environments. Gene 261, 35–42. Tamai Y. and Kurihara T.(2000). Phylogenetic relation of lungfish indicated by the amino acid sequence of myelin DM20. Mol. Brain Res. 80, 256–259. 58. Yokoyama S. (2000). Color vision of the coelacanth (Latimeria chalumnae) and adaptive evolution of rhodopsin (RH1) and rhodopsin-like (RH2) pigments. 31. Bogard J.P.,Balon E.K. and Bruton M.N. (1994). The chromosomes of the living J. Hered. 91, 215–220. coelacanth and their remarkable similarity to those of one of the most ancient frogs. J. Hered. 85, 322–325. 59. Yokoyama S., Zhang H., Radlwimmer B.F.and Blow N.S. (1999). Adaptive evo- 32. Danke J., Miyake T.,Powers T., Schein J., Shin H., Bosdt I., Erdmann M, Caldwell lution of color vision of the Comoran coelacanth (Latimeria chalumnae). Proc. R. and Amemiya C.T. (2004). Genome resource for the Indonesian coelacanth, Natl Acad. Sci. USA 96, 6279–6284. Latimeria menadoensis. J. Exp. Zool. 301A, 228–234. 60. Shi Y. and Yokoyama S. (2003). Molecular analysis of the evolutionary signifi- 33. Powers T.P. and Amemiya C.T. (2004). Evidence for a Hox14 paralog group in cance of ultraviolet vision in vertebrates. Proc. Natl Acad. Sci. USA 100, 8308– vertebrates. Curr. Biol. 14, R183–R184. 8313. 34. Shashikant C., Bolanowki S.A., Danke J. and Amemiya C.T.(2004). Hoxc8 early 61. Takahashi A., Amemiya Y., Nozaki M., Sower S.A. and Kawauchi H. (2001). enhancer of the Indonesian coelacanth, Latimeria menadoensis. J. Exp. Zool. 302B, Evolutionary significance of the proopiomelanocortin in agnatha and 557–563. chondrichthyes. Comp. Biochem. Physiol. [B] 129, 283–289. 35. Anand S., Wang W.C.H., Powell D.R., Bolanowski S.A., Zhang J., Ledje C., 62. Takahashi A., Yashuda A., Sullivan C.V.and Kawauchi H. (2003). Identification Pawashe A.B., Amemiya C.T. and Shashikant C.S. (2003). Divergence of the of proopiomelanocortin-related peptides in the rostral pars distalis of the pitu- Hoxc8 early enhancer parallels diverged axial morphologies between mammals itary in coelacanth: evolutional implications. Gen. Comp. Endocr. 130, 340–349. and fishes. Proc. Natl Acad. Sci. USA 100, 15666–15669. 63. Freitag J., Ludwig G., Andreini I., Rössler P. and Breer H. (1998). Olfactory 36. Noonan J.P.,Grimwood J., Danke J., Schmutz J., Dickson M., Amemiya C.T.and receptors in aquatic and terrestrial vertebrates. J. Comp. Physiol. [B] 183, Myers R.M. (2004). Coelacanth genome sequence reveals the evolutionary his- 635–650. tory of vertebrate genes. Genome Res. 14, 2397–2405. 64. Meyer A. and Dolven S.I. (1992). Molecules, fossils, and the origin of tetrapods. 37. Noonan J.P,Grimwood J., Schmutz J., Dickson M. and Myers R.M. (2004). Gene J. Mol. Evol. 35, 102–113. conversion and the evolution of protocadherin gene cluster diversity. Genome Res. 14, 354–366. Note added in proof: The US National Institutes of Health have approved the African 38. Minguillón C., Gardenyes J., Serra E., Castro L.F.C., Hillforce A., Holland P., coelacanth for whole-genome sequencing. Supplementary material to:

Modisakeng K.W., Amemiya C.T., Dorrington R.A. and Blatch G.L. (2006). Molecular biology studies on the coelacanth: a review. S. Afr. J. Sci. 102, 479–485.

Table 1. Sequenced loci from both coelacanth species. The sequences were accessed from the non-redundant nucleotide sequence database at the National Centre for Biotechnology Information (accessed on 15 September 2005).

Description of the sequence (accessed on 15 September 2005) Accession Ref.

Latimeria menadoensis mitochondrion, complete genome NC_006921 28

Latimeria chalumnae mitochondrion, complete genome NC_001804 27 Latimeria menadoensis clone VMRC4-140H19 and VMRC4-19C10, complete sequences AC151571 and AC147788 Unpublished: MAPPING INFORMATION: The sequence of this clone was established as part of a mapping and sequencing collaboration between the Stanford Human Genome Center and the Benaroya Research Institute at Virgina Mason. The VMRC-4 BAC library was constructed in the laboratory of Chris Amemiya using genomic DNA of nuclei isolated from Latimeria menadoensis heart tissue.

Latimeria chalumnae heat shock protein 70 (Hsp70) gene, partial cds. AY929184 50

Latimeria chalumnae DM20 gene for DM20, partial cds AB025933-AB025938 30 Latimeria chalumnae TATA box binding protein (TBP)-associated factor mRNA, partial cds. AY389957

Latimeria chalumnae ribosomal protein S9, S8, S7, S6, S4, S3 mRNA, partial cds. AY389955; 15 AY389953; AY389951; AY389947; AY389945

Latimeria chalumnae ribosomal protein Large P0 mRNA, partial cds. AY389943 15 Latimeria chalumnae ribosomal protein L7a-like mRNA, partial sequence. AY389941

Latimeria chalumnae ribosomal protein L8,L7, L5, L4,L19, L18, L17, L11, L10a mRNA, partial cds. AY389974; 15 AY389939; AY389937; AY389935; AY389933; AY389931; AY389929; AY389925; AY389923

Latimeria chalumnae protein tyrosine phosphatase non-receptor type 11 mRNA, partial cds. AY389921 15

Latimeria chalumnae protein kinase C beta 1 mRNA, partial cds. AY389919 15

Latimeria chalumnae protein kinase C alpha mRNA, partial cds. AY389917 15

Latimeria chalumnae chaperonin subunit 8 theta mRNA, partial cds. AY389915 15

Latimeria chalumnae cathepsin B mRNA, partial cds. AY389913 15

Latimeria chalumnae glucose-6-phosphate isomerase mRNA, partial cds. AY389911 15

Latimeria chalumnae nucleoside diphosphate kinase mRNA, partial cds. AY389909 15

Latimeria chalumnae guanine nucleotide binding protein beta polypeptide 2-like 1 mRNA, partial cds AY389907 15

Latimeria chalumnae ferritin heavy polypeptide 1 mRNA, partial cds. AY389905 15

Latimeria chalumnae fascin-like protein 1-like mRNA, partial sequence. AY389903 15

Latimeria chalumnae eukaryotic translation elongation factor 1 gamma mRNA, partial cds. AY389901 15

Latimeria chalumnae nuclear receptor subfamily 2 group F number 1 mRNA, partial cds. AY389899 15

Latimeria chalumnae calreticulin mRNA, partial cds. AY389897 15

Latimeria chalumnae mitochondrial ATP synthase H+ transporting F1 complex alpha subunit isoform 1 mRNA, AY389895 15 partial cds; nuclear gene for mitochondrial product.

Latimeria chalumnae aminolevulinic acid synthase 2 mRNA, partial cds. AY389893 15

Latimeria chalumnae aminolevulinic acid synthase 1 mRNA, partial cds. AY389891 15

Latimeria chalumnae alcohol dehydrogenase 3 mRNA, partial cds. AY389890 15

Latimeria menadoensis RAG1 (Rag1) gene, partial cds. AY442925 14

Latimeria menadoensis HoxD8-HoxD12 gene, exon 2 and partial cds. AY183751–AY183755 46

Latimeria menadoensis HoxD1, HoxD3, HoxD4 gene, partial cds. AY183748; 46 AY183749; AY183750

Latimeria menadoensis HoxC13 gene, exon 2 and partial cds. AY183747 46 Latimeria menadoensis HoxC10 gene, exon 2 and partial cds. AY183746 Latimeria menadoensis HoxC9 gene, exon 2 and partial cds. AY183745 Latimeria menadoensis HoxC8 gene, exon 2 and partial cds. AY183744 Latimeria menadoensis HoxC5 gene, exon 2 and partial cds. AY183743 Latimeria menadoensis HoxC4 gene, exon 2 and partial cds. AY183742 Latimeria menadoensis HoxC1 gene, exon 2 and partial cds. AY183741 Latimeria menadoensis HoxC12 gene, exon 2 and partial cds. AY189938

Latimeria menadoensis HoxB7 gene, exon 2 and partial cds. AY183739 46 Latimeria menadoensis HoxB6 gene, exon 2 and partial cds. AY183738 Latimeria menadoensis HoxB5 gene, complete cds. AY183737 Latimeria menadoensis HoxB4 gene, complete cds. AY183736 Latimeria menadoensis HoxB3 gene, exon 2 and partial cds. AY183735 Latimeria menadoensis HoxB2 gene, partial cds. AY183734 Latimeria menadoensis HoxB1 gene, exon 2 and partial cds. AY183733

Latimeria menadoensis HoxA13 gene, exon 2 and partial cds. AY183731 46 Latimeria menadoensis HoxA11 gene, exon 2 and partial cds. AY183730 Latimeria chalumnae Hoxa-11 gene, partial cds. AF287139 Latimeria menadoensis HoxA10 gene, exon 2 and partial cds. AY183729 Latimeria menadoensis HoxA9 gene, exon 2 and partial cds. AY183728 Latimeria menadoensis HoxA7 gene, partial cds. AY183727 Latimeria menadoensis HoxA6 gene, exon 2 and partial cds. AY183726 Latimeria menadoensis HoxA4 gene, exon 2 and partial cds. AY183725 Latimeria menadoensis HoxA2 gene, exon 2 and partial cds. AY183724 Latimeria menadoensis HoxA1 gene, exon 2 and partial cds. AY183723 Latimeria menadoensis Evx1 gene, exon 3 and partial cds. AY183732

Latimeria chalumnae beta enolase-1 mRNA, partial cds. AY005155 23 Latimeria chalumnae alpha enolase-1 mRNA, partial cds AY005154

Latimeria menadoensis tRNA-Glu gene, partial sequence; cytochrome b gene, complete cds; tRNA-Thr, AF176901 29 tRNA-Pro, and tRNA-Phe genes, complete sequence; 12S ribosomal RNA gene, complete sequence tRNA-Val gene, complete sequence; and 16S ribosomal RNA gene, complete sequence; mitochondrial genes for mitochondrial products

Latimeria chalumnae RH2 opsin (rh2) gene, exon 1??–5 complete cds. AF131258–AF131262 59

Latimeria chalumnae rhodopsin (rh1) gene, exon 1??– 5. AF131253–AF131257 59

Latimeria chalumnae gene encoding olfactory receptor, partial sequences AJ233777–AJ233783 63

Latimeria chalumnae 28S ribosomal RNA gene, partial sequence U34336 19

Latimeria chalumnae mitochondrion 12S and 16S rRNA genes, and tRNA-Val gene Z21921 22

Latimeria chalumnae IG VH GENES X57354–X57356 55

Latimeria chalumnae immunoglobulin VH pseudogene. X57353 55

Latimeria chalumnae MHC class I CDS protein (Lach-UB-03) mRNA, alpha 2 and alpha 3 domains, partial cds, U08046 21 exons 3 and 4 Latimeria chalumnae clone RNA2 MHC class I protein (Lach-UB-05) mRNA, alpha 2 and alpha 3 domains, U08040 partial cds, exons 3 and 4 Latimeria chalumnae clone Uli5 MHC class I protein (Lach-UD-01) gene, alpha 3 domain, partial cds, exon 4. U08045 Latimeria chalumnae clone RNA3 MHC class I protein (Lach-UB-06) mRNA, alpha 2 and alpha 3 domains, U08044 partial cds, exons 3 and 4 Latimeria chalumnae clone 10 MHC class I protein (Lach-UA-01) gene, alpha 2 and alpha 3 domains, partial cds, U08043 exons 3 and 4 Latimeria chalumnae clone LatB MHC class I protein (Lach-UX-02) gene, alpha 3 domain, partial cds, exon 4. U08042 Latimeria chalumnae clone LatA MHC class I protein (Lach-UX-01) gene, alpha 3 domain, partial cds, exon 4. U08041 Latimeria chalumnae clone Uli12 MHC class I protein (Lach-UC-01) gene, alpha 3 domain, partial cds, exon 4. U08039 Latimeria chalumnae clone TMC2g MHC class I protein (Lach-UB-02) mRNA, alpha 3 and transmembrane U08037 domains, partial cds. Latimeria chalumnae clone TMC2a MHC class I protein (Lach-UB-03) mRNA, alpha 3 and transmembrane U08036 domains, partial cds Latimeria chalumnae clone 7 MHC class I protein (Lach-UB-01) gene, alpha 1, alpha 2, alpha 3 and transmembrane U08034 domains, partial cds, exons 2, 3, 4, and 5 Latimeria chalumnae clone Coe7 MHC class I protein (Lach-UA-02) mRNA, alpha 2 and alpha 3 domains, U08033 partial cds, exons 3 and 4.

Latimeria chalumnae mitochondrion ribosomal RNA small subunit gene, internal sequence. M87534 64

Latimeria chalumnae 18S ribosomal RNA. L11288 17 Table 2. Latimeria chalumnae codon usage table.Tocompile the table, average codon usage was determined from eight L.chalumnae coding regions: -enolase (AY005155), -enolase (AY005154), MHC1 mRNAs (U08036 and U08037), olfactory receptor genes (AJ233783 and AJ233784), the Hoxa-11 gene (AF2887139) and hsp70 (AY929184).

Amino acid:L I V SWM Codon: UUA UUG CUU CUC CUA CUG AUU AUC AUA GUU GUC GUA GUG UCU UCC UCA UCG AGU AGC UGG AUG

Gene Alpha enolase 5.49 19.2 16.5 8.24 0 38.46 38.5 13.7 0 30.2 19.2 5.49 30.22 16.48 10.99 13.7 0 11 2.8 8.24 16.5 Beta enolase 0 0 16.7 8.33 8.3 25 41.7 41.7 8.33 33.3 0 16.7 33.33 16.67 8.33 0 0 0 17 25 8.33 MHC1.TMC2g mRNA 0 25.4 0 25.4 0 42.37 59.3 17 8.47 50.9 33.9 0 8.47 0 0 0 17 8.5 17 25.42 25.4 MHC1.TMC2a mRNA 0 25.4 0 25.4 0 50.85 67 8.47 17 42.4 33.9 0 8.47 0 0 8.47 8.47 8.5 25 25.42 25.4 Olifactory gene 1 4.55 31.8 22.7 59.1 4.6 22.73 18.2 45.5 9.09 9.09 27.3 4.55 45.45 13.64 36.36 22.7 0 18 14 4.55 36.4 Olifactory gene2 18.75 18.8 12.5 12.5 25 37.5 37.5 43.8 43.8 34.3 18.8 12.5 6.25 37.5 18.75 25 0 0 0 12.5 25 Hoxa-11 gene 4.95 9.9 0 4.95 15 4.95 9.9 4.95 9.9 14.9 9.9 0 9.9 54.46 34.95 19.8 19.8 20 35 4.95 4.95 Hsp70 1.63 3.26 50.57 29.4 42.4 1.63 4.89 29.6 6.52 24.47 16.31 14.68 4.89 1.63 6.5 16 3.26 19.6 Average usage 4.421 18.6 8.96 20.6 7.5 34.054 37.7 27.2 12.3 27.5 21.6 5.72 20.82 19.38 15.51 11.8 5.86 9.1 16 13.67 20.2

Amino acid:P T A QNKY Codon: CCU CCC CCA CCG ACU ACC ACA ACG GCU GCC GCA GCG CAA CAG AAU AAC AAA AAG UAU UAC

Gene Alpha enolase 8.24 6.41 22 2.74 22 5.49 11 0 55 13.7 33 2.75 5.49 16.48 30.22 24.7 60.4 30 11 19.23 Beta enolase 8.33 0 16.7 0 25 25 33.3 0 8.33 16.7 25 0 16.67 8.33 33.33 33.3 33.3 17 0 25 MHC1.TMC2g mRNA 8.47 11.3 17 5.63 8.5 50.85 33.9 0 17 8.47 17 8.47 0 25.42 16.95 25.4 25.4 17 8.5 25.42 MHC1.TMC2a mRNA 0 14.1 17 8.45 8.5 50.85 17 0 25.4 0 8.47 17 0 25.42 25.42 0 17 8.5 8.5 25.42 Olifactory gene 1 18.18 0 0 2.05 23 27.27 18.2 0 31.8 36.4 18.2 0 4.55 18.18 4.55 9.09 13.6 14 32 18.18 Olifactory gene2 31.25 10.3 25.5 0 19 18.75 25 0 0 6.25 6.25 0 18.75 0 6.25 25 18.8 13 38 25 Hoxa-11 gene 19.8 8.21 9.9 10.3 9.9 34.65 39.6 4.95 15 9.9 19.8 0 19.8 9.9 29.75 29.7 19.8 20 20 34.65 Hsp70 9.79 16.3 4.89 1.63 16 35.89 11.4 4.89 29.4 27.7 14.7 8.16 13.05 30.99 19.58 31 32.6 51 9.8 16.31 Average usage 13.01 8.32 14.1 3.85 16 31.094 23.7 1.23 22.7 14.9 17.8 4.54 9.789 16.84 20.76 22.3 27.6 21 16 23.65

Amino acid:R HGCDEF Codon: CGU CGC CGA CGG AGA AGG CAU CAC GGU GGC GGA GGG UGC UGU GAU GAC GAA GAG UUU UUC

Gene Alpha enolase 8.24 11 0 0 8.2 0 5.49 5.49 44 13.7 24.7 5.49 8.24 5.49 43.96 22 35.7 33 22 16.48 Beta enolase 8.33 8.33 0 8.33 8.3 0 41.7 0 50 16.7 16.7 8.33 8.33 16.67 50 25 16.7 33 17 25 MHC1.TMC2g mRNA 0 0 17 0 8.5 25.42 8.47 17 8.47 0 33.9 17 8.47 8.47 42.37 25.4 8.47 42 8.5 16.95 MHC1.TMC2a mRNA 0 0 8.47 0 8.5 25.42 8.47 17 0 0 33.9 17 8.47 8.47 42.37 59.4 17 51 8.5 16.95 Olifactory gene 1 0 9.09 4.55 0 0 27.27 0 27.3 22.7 9.09 9.09 18.2 27.27 22.73 18.18 18.2 9.09 4.6 18 40.91 Olifactory gene2 6.25 0 0 6.25 6.3 12.5 25 18.8 12.5 6.25 0 12.5 25 37.5 12.5 6.25 6.25 6.3 44 37.54 Hoxa-11 gene 4.95 4.95 0 4.95 20 29.7 14.9 19.8 15 19.8 4.95 9.9 24.75 0 9.9 34.7 24.8 50 20 19.8 Hsp70 3.26 8.16 1.63 1.63 15 19.58 1.63 6.58 13.1 19.6 27.7 9.79 3.26 4.89 27.73 50.6 26.1 46 0 16.31 Average usage 3.879 5.19 3.95 2.65 9.3 17.486 13.2 14 20.7 10.6 18.9 12.3 14.22 13.03 30.88 30.2 18 33 17 23.74