Sister Group Relationship of Turtles to the Bird-Crocodilian Clade Revealed by Nuclear DNA–Coded Proteins
http://www.100md.com
《分子生物学进展》
* Department of Biophysics, Graduate School of Science, Kyoto University, Kyoto, Japan; and Division of Material Science, Graduate School of Science, Nagoya University, Nagoya, Japan
Correspondence: E-mail: kkatoh@kuicr.kyoto-u.ac.jp.
Abstract
The phylogenetic position of turtles is a currently controversial issue. Recent molecular studies rejected a traditional view that turtles are basal living reptiles (Hedges, S. B., and L. L. Poling. 1999. A molecular phylogeny. Science 83:998–1001; Kumazawa, Y., and M. Nishida. 1999. Complete mitochondrial DNA sequences of the green turtle and blue-tailed mole skink, statistical evidence for archosaurian affinity of turtles. Mol. Biol. Evol. 16:784–792). Instead, these studies grouped turtles with birds and crocodiles. The relationship among turtles, birds, and crocodiles remained unclear to date. To resolve this issue, we have cloned and sequenced two nuclear genes encoding the catalytic subunit of DNA polymerase and glycinamide ribonucleotide synthetase–aminoimidazole ribonucleotide synthetase–glycinamide ribonucleotide formyltransferase from amniotes and an amphibian. The amino acid sequences of these proteins were subjected to a phylogenetic analysis based on the maximum likelihood method. The resulting tree showed that turtles are the sister group to a monophyletic cluster of archosaurs (birds and crocodiles). All other possible tree topologies were significantly rejected.
Key Words: amniote phylogeny ? anapsid ? archosaur ? lepidosaur ? turtle
Introduction
Unlike other reptiles and birds, turtles have no temporal holes in their skull. Most morphologists traditionally believed that turtles are the only survivors of anapsids, a primitive reptile group lacking temporal holes (e.g., Caroll 1988; Lee 1997), and that turtles are a phylogenetically isolated group in living reptiles, like figure 1a (hypothesis a). A recent morphological analysis, however, did not support the traditional view and placed turtles as the sister group to lepidosaurs (squamates [lizards and snakes] and tuatara), as shown in figure 1b (hypothesis b), with a moderate support value (Rieppel and deBraga 1996; deBraga and Rieppel 1997).
FIG. 1.— Four possible hypotheses about the phylogenetic position of turtles. (a) Caroll (1988) and Lee (1997). (b) Rieppel and deBraga (1996) and deBraga and Rieppel (1997). (c) Zardoya and Meyer (1998), Hedges and Poling (1999), Kumazawa and Nishida (1999), and Rest et al. (2003). (d) Hedges and Poling (1999) and Mannen and Li (1999).
Recent molecular phylogenetic studies showed the archosaurian (birds and crocodiles) affinities of turtles (Zardoya and Meyer 1998; Hedges and Poling 1999; Kumazawa and Nishida 1999; Mannen and Li 1999; Rest et al. 2003), as shown in figure 1c and d (hypotheses c and d). Mitochondrial data tended to support hypothesis c, whereas nuclear data favored hypothesis d. Using both mitochondrial and nuclear data, Cao et al. (2000) conducted an extensive analysis and concluded that hypotheses a and b are significantly rejected and that more data are needed to discriminate between hypotheses c and d.
Multiple nuclear DNA–coded protein data were considered to be more reliable than mitochondrial data in recovering phylogenetic relationships among major groups of vertebrates (Hedges 2001; Kikugawa et al. 2004; Takezaki et al. 2004). Even when each single gene gives an ambiguous result, a statistically solid inference is possibly obtained by using multiple orthologous gene sequences. Considerable amount of amino acid sequence data are currently available for inferring phylogenetic position of turtles, as shown in the supplementary table. These genes are collectively referred to as "previously available" genes in this paper. Each of these genes is, however, generally short in amino acid alignment length (138–372 aa). We could not completely exclude the possibility of paralogous comparison for these genes because some of them, such as globin and lactate dehydrogenase, have multiple copies in a vertebrate genome. In order to avoid paralogy, large and single-copy genes should be used for phylogenetic inference.
To infer the phylogenetic position of turtles based on statistically reliable data, we have cloned and sequenced two large genes encoding the catalytic subunit of DNA polymerase (DPLA) and glycinamide ribonucleotide synthetase–aminoimidazole ribonucleotide synthetase–glycinamide ribonucleotide formyltransferase (GAG) from amniotes and an amphibian. These genes are single copy in the human, mouse, and fugu genomes. The amino acid sequences coded by these genes of human, mouse, chicken, caiman (crocodile), iguana (squamate), turtle, and axolotl (amphibian) were subjected to a phylogenetic analysis based on the maximum likelihood (ML) method.
The ML tree inferred from the concatenated alignment of DPLA and GAG (DPLA + GAG; 2,195 aa) is shown in figure 2, in which turtles are the sister group to the monophyletic cluster of chicken and caiman, corresponding to hypothesis c (fig. 1c). All other tree topologies, including those corresponding to hypotheses a, b, and d, were significantly rejected (P(KH) [P value based on two-sided Kishino-Hasegawa test] < 0.010, P(AU) [P value based on approximately unbiased test] < 0.044), as shown in table 1. The same tree topology was obtained independently from each of the GAG and DPLA data sets, although not statistically significant; the minimum log-likelihood differences between the ML tree and the second best tree were 12.3 ± 7.71 (P(KH) = 0.060, P(AU) = 0.072) and 8.02 ± 9.38 (P(KH) = 0.20, P(AU) = 0.39) for DPLA and GAG, respectively.
FIG. 2.— The ML tree inferred from DPLA and GAG. Branch lengths were calculated from the concatenated alignment of DPLA and GAG. RELL BP values calculated from the concatenated alignment/DPLA/GAG are shown in this order for each branch.
Table 1 Comparison of Log-Likelihood Values Based on the DPLA and GAG Proteins
When the DPLA and GAG sequences were concatenated together with the previously available data (total length was 5,189 aa), hypothesis c was strongly supported, whereas all other tree topologies were significantly rejected (P(KH) < 0.026, P(AU) < 0.032), as shown in table 2.
Table 2 Comparison of Log-Likelihood Values Based on GAG, DPLA, and Previously Available Data
We also carried out separate analyses, in which the log-likelihood values were separately calculated for 15 proteins and then summed, and obtained similar results to those from concatenated alignment analyses (the right three columns of tables 1 and 2).
We analyzed the concatenated alignment of previously available amino acid sequences (2,994 aa), although the possibility of paralogy could not be completely excluded. This data also supported hypothesis c, but the difference from the second best tree was only 1.69 ± 9.03 (P(KH) = 0.56, P(AU) = 0.57). This result disagreed with Hedges and Poling (1999), which supported hypothesis d based on the nuclear DNA data available at that time. This discrepancy was probably caused by the amount of available data and/or unrecognized paralogous comparisons.
In summary, we conducted a phylogenetic analysis of amniotes based on large amount of nuclear DNA–coded protein sequence data and obtained a single tree topology supporting hypothesis c, in which turtles are the sister group to a monophyletic cluster of archosaurs (birds and crocodiles). All other tree topologies, including the traditional one, were significantly rejected.
Materials and Methods
Total RNAs were extracted from liver of Caiman crocodilus (spectacled caiman), tail of Trachemys scripta (red-eared slider), tail of Iguana iguana (green iguana), tail of Ambystoma mexicanum (Mexican axolotl), and embryo of Gallus gallus (chicken) using TRIZOL reagent (Invitrogen, Carlsbad, Calif.). These total RNAs were reverse-transcribed to cDNAs using SMART RACE cDNA Amplification Kit (Clontech, Palo Alto, Calif.). These cDNAs were used as templates for PCR amplification with Expand High-Fidelity PCR System (Roche, Basel, Switzerland). The sense and antisense degenerate primers were designed from conserved amino acid residues of each gene. As fragments of chicken DPLA gene were found in the dbEST database of the National Center for Biotechnology Information, gene specific primers were used for this gene. The PCR products were purified and cloned into the pT7Blue vector (Novagen, Darmstadt, Germany). More than three independent clones were isolated and sequenced using ABI 3100 DNA Sequencer (Applied Biosystems, Foster City, Calif.). The 3' ends of cDNAs were amplified using 3'RACE (GIBCO BRL, Invitrogen) and sequenced in the same way as above. The following sequence data were taken from the GenBank database: human and mouse DPLA; human, mouse, and chicken GAG; and previously available data listed in supplementary table.
Tuatara, one of the four major groups of reptiles (crocodiles, turtles, squamates, and tuatara), is thought to be closely related to squamates based on morphological data (Caroll 1988) and mitochondrial sequence data (Rest et al. 2003). Thus, tuatara was excluded from the present analysis.
Chicken has a pair of duplicated GAG genes (GAG-A and -B) (Smith et al. 2000). Chicken GAG-B was estimated to be more closely related to chicken GAG-A than to any of the reptile GAG genes, in distance measured by synonymous substitutions (Miyata and Yasunaga 1980). According to a comparison of amino acid substitution rates of amniote GAG genes using amphibian sequences as an out-group, the chicken GAG-B gene was estimated to evolve approximately 3.5–4.5 times faster than the others. These observations suggest that the GAG gene was duplicated on the avian lineage and that one (GAG-B) of the duplicated genes accumulated amino acid changes at an extremely rapid rate. Thus, the chicken GAG-B gene was excluded from the present analysis, in order to avoid the long-branch attraction artifact (Felsenstein 1978).
Each protein data set was multiply aligned by MAFFT (Katoh et al. 2002) and manually inspected. Unambiguously aligned amino acid positions were subjected to phylogenetic tree analyses based on the ML method (Felsenstein 1981; Kishino, Miyata, and Hasegawa 1990). For each of 945 possible topologies consisting of seven taxonomic groups (six amniote groups and an out-group), the log-likelihood values based on GAG (856 aa), DPLA (1,339 aa), each of the previously available 13 proteins, and three types of concatenated alignments (DPLA + GAG, 2,195 aa; DPLA + GAG + previously available data, 5,189 aa; and previously available data only, 2,994 aa) were calculated using the GAMT (Katoh, Kuma, and Miyata 2001) program, assuming the JTT-F model (Jones, Taylor, and Thornton 1992; Cao et al. 1994; Adachi and Hasegawa 1996). Heterogeneity of evolutionary rates among sites was modeled by a discrete distribution (Yang 1994) with the shape parameter optimized for each data set.
Two-sided KH test and the AU test were carried out using the CONSEL package (Shimodaira and Hasegawa 2001). Bootstrap probability (BP) based on the resampling of estimated log-likelihoods (RELL) approximation (Kishino, Miyata, and Hasegawa 1990) was also computed. The bootstrap probability for a cluster is calculated by totaling the RELL BP values of the tree topologies having the cluster.
Supplementary Materials
The sequences reported in this paper have been deposited in the GenBank database (accession numbers AB178525–AB178532).
Acknowledgements
We thank K. Kuma for comments and suggestions. This work was supported in part by a Grant-in-Aid for Creative Scientific Research and a Grant for the Biodiversity Research of the 21st Century COE (A14) from the Ministry of Education, Culture, Sports, Science and Technology of Japan.
References
Adachi, J., and M. Hasegawa. 1996. MOLPHY Version 2.3: programs for molecular phylogenetics based on maximum likelihood. Comput. Sci. Monogr. 28:1–150.
Cao, Y., J. Adachi, A. Janke, S. Paabo, and M. Hasegawa. 1994. Phylogenetic relationships among eutherian orders estimated from inferred sequences of mitochondrial proteins: instability of a tree based on a single gene. J. Mol. Evol. 39:519–527.
Cao, Y., M. D. Sorenson, Y. Kumazawa, D. P. Mindell, and M. Hasegawa. 2000. Phylogenetic position of turtles among amniotes: evidence from mitochondrial and nuclear genes. Gene 259:139–148.
Caroll, R. L. 1988. Vertebrate paleontology and evolution. Freeman, New York.
deBraga, M., and O. Rieppel. 1997. Reptile phylogeny and the interrelationships of turtles. Zool. J. Linn. Soc. 120:281–354.
Felsenstein, J. 1978. Cases in which parsimony or compatibility methods will be positively misleading. Syst. Zool. 27:401–410.
———. 1981. Evolutionary trees from DNA sequences: a maximum likelihood approach. J. Mol. Evol. 17:368–376.
Hedges, S. B. 2001. Molecular evidence for the early history of living vertebrates. Pp. 119–134 in P. E. Ahlberg, ed. Major events in early vertebrate evolution. Taylor & Francis, London.
Hedges, S. B., and L. L. Poling. 1999. A molecular phylogeny of reptiles. Science 283:998–1001.
Jones, D. T., W. R. Taylor, and J. M. Thornton. 1992. The rapid generation of mutation data matrices from protein sequences. Comput. Appl. Biosci. 8:275–282.
Katoh, K., K. Kuma, and T. Miyata. 2001. Genetic algorithm-based maximum-likelihood analysis for molecular phylogeny. J. Mol. Evol. 53:477–484.
Katoh, K., K. Misawa, K. Kuma, and T. Miyata. 2002. MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic Acids Res. 30:3059–3066.
Kikugawa, K., K. Katoh, S. Kuraku, H. Sakurai, O. Ishida, N. Iwabe, and T. Miyata. 2004. Basal jawed vertebrate phylogeny inferred from multiple nuclear DNA-coded genes. BMC Biol. 2:3.
Kishino, H., T. Miyata, and M. Hasegawa. 1990. Maximum likelihood inference of protein phylogeny and the origin of chloroplasts. J. Mol. Evol. 31:151–160.
Kumazawa, Y., and M. Nishida. 1999. Complete mitochondrial DNA sequences of the green turtle and blue-tailed mole skink: statistical evidence for archosaurian affinity of turtles. Mol. Biol. Evol. 16:784–792.
Lee, M. S. Y. 1997. Pareiasaur phylogeny and the origin of turtles. Zool. J. Linn. Soc. 120:197–280.
Mannen, H., and S. S. Li. 1999. Molecular evidence for a clade of turtles. Mol. Phylogenet. Evol. 13:144–148.
Miyata, T., and T. Yasunaga. 1980. Molecular evolution of mRNA: a method for estimating evolutionary rates of synonymous and amino acid substitutions from homologous nucleotide sequences and its application. J. Mol. Evol. 16:23–36.
Rest, J. S., J. C. Ast, C. C. Austin, P. J. Waddell, E. A. Tibbetts, J. M. Hay, and D. P. Mindell. 2003. Molecular systematics of primary reptilian lineages and the tuatara mitochondrial genome. Mol. Phylogenet. Evol. 29:289–297.
Rieppel, O., and M. deBraga. 1996. Turtles as diapsid reptiles. Nature 384:453–455.
Shimodaira, H., and M. Hasegawa. 2001. CONSEL: for assessing the confidence of phylogenetic tree selection. Bioinformatics 17:1246–1247.
Smith, J., J. Reboul, G. Lutfalla, and D. W. Burt. 2000. Human chromosomes 3 and 21 are the products of an ancestral gene arrangement that is at least 300 million years old. Mamm. Genome 11:806–807.
Takezaki, N., F. Figueroa, Z. Zaleska-Rutczynska, N. Takahata, and J. Klein. 2004. The phylogenetic relationship of tetrapod, coelacanth, and lungfish revealed by the sequences of forty-four nuclear genes. Mol. Biol. Evol. 21:1512–1524.
Yang, Z. 1994. Maximum likelihood phylogenetic estimation from DNA sequences with variable rates over sites: approximate methods. J. Mol. Evol. 39:306–314.
Zardoya, R., and A. Meyer. 1998. Complete mitochondrial genome suggests diapsid affinities of turtles. Proc. Natl. Acad. Sci. USA 95:14226–14231.(Naoyuki Iwabe*, Yuichiro )
Correspondence: E-mail: kkatoh@kuicr.kyoto-u.ac.jp.
Abstract
The phylogenetic position of turtles is a currently controversial issue. Recent molecular studies rejected a traditional view that turtles are basal living reptiles (Hedges, S. B., and L. L. Poling. 1999. A molecular phylogeny. Science 83:998–1001; Kumazawa, Y., and M. Nishida. 1999. Complete mitochondrial DNA sequences of the green turtle and blue-tailed mole skink, statistical evidence for archosaurian affinity of turtles. Mol. Biol. Evol. 16:784–792). Instead, these studies grouped turtles with birds and crocodiles. The relationship among turtles, birds, and crocodiles remained unclear to date. To resolve this issue, we have cloned and sequenced two nuclear genes encoding the catalytic subunit of DNA polymerase and glycinamide ribonucleotide synthetase–aminoimidazole ribonucleotide synthetase–glycinamide ribonucleotide formyltransferase from amniotes and an amphibian. The amino acid sequences of these proteins were subjected to a phylogenetic analysis based on the maximum likelihood method. The resulting tree showed that turtles are the sister group to a monophyletic cluster of archosaurs (birds and crocodiles). All other possible tree topologies were significantly rejected.
Key Words: amniote phylogeny ? anapsid ? archosaur ? lepidosaur ? turtle
Introduction
Unlike other reptiles and birds, turtles have no temporal holes in their skull. Most morphologists traditionally believed that turtles are the only survivors of anapsids, a primitive reptile group lacking temporal holes (e.g., Caroll 1988; Lee 1997), and that turtles are a phylogenetically isolated group in living reptiles, like figure 1a (hypothesis a). A recent morphological analysis, however, did not support the traditional view and placed turtles as the sister group to lepidosaurs (squamates [lizards and snakes] and tuatara), as shown in figure 1b (hypothesis b), with a moderate support value (Rieppel and deBraga 1996; deBraga and Rieppel 1997).
FIG. 1.— Four possible hypotheses about the phylogenetic position of turtles. (a) Caroll (1988) and Lee (1997). (b) Rieppel and deBraga (1996) and deBraga and Rieppel (1997). (c) Zardoya and Meyer (1998), Hedges and Poling (1999), Kumazawa and Nishida (1999), and Rest et al. (2003). (d) Hedges and Poling (1999) and Mannen and Li (1999).
Recent molecular phylogenetic studies showed the archosaurian (birds and crocodiles) affinities of turtles (Zardoya and Meyer 1998; Hedges and Poling 1999; Kumazawa and Nishida 1999; Mannen and Li 1999; Rest et al. 2003), as shown in figure 1c and d (hypotheses c and d). Mitochondrial data tended to support hypothesis c, whereas nuclear data favored hypothesis d. Using both mitochondrial and nuclear data, Cao et al. (2000) conducted an extensive analysis and concluded that hypotheses a and b are significantly rejected and that more data are needed to discriminate between hypotheses c and d.
Multiple nuclear DNA–coded protein data were considered to be more reliable than mitochondrial data in recovering phylogenetic relationships among major groups of vertebrates (Hedges 2001; Kikugawa et al. 2004; Takezaki et al. 2004). Even when each single gene gives an ambiguous result, a statistically solid inference is possibly obtained by using multiple orthologous gene sequences. Considerable amount of amino acid sequence data are currently available for inferring phylogenetic position of turtles, as shown in the supplementary table. These genes are collectively referred to as "previously available" genes in this paper. Each of these genes is, however, generally short in amino acid alignment length (138–372 aa). We could not completely exclude the possibility of paralogous comparison for these genes because some of them, such as globin and lactate dehydrogenase, have multiple copies in a vertebrate genome. In order to avoid paralogy, large and single-copy genes should be used for phylogenetic inference.
To infer the phylogenetic position of turtles based on statistically reliable data, we have cloned and sequenced two large genes encoding the catalytic subunit of DNA polymerase (DPLA) and glycinamide ribonucleotide synthetase–aminoimidazole ribonucleotide synthetase–glycinamide ribonucleotide formyltransferase (GAG) from amniotes and an amphibian. These genes are single copy in the human, mouse, and fugu genomes. The amino acid sequences coded by these genes of human, mouse, chicken, caiman (crocodile), iguana (squamate), turtle, and axolotl (amphibian) were subjected to a phylogenetic analysis based on the maximum likelihood (ML) method.
The ML tree inferred from the concatenated alignment of DPLA and GAG (DPLA + GAG; 2,195 aa) is shown in figure 2, in which turtles are the sister group to the monophyletic cluster of chicken and caiman, corresponding to hypothesis c (fig. 1c). All other tree topologies, including those corresponding to hypotheses a, b, and d, were significantly rejected (P(KH) [P value based on two-sided Kishino-Hasegawa test] < 0.010, P(AU) [P value based on approximately unbiased test] < 0.044), as shown in table 1. The same tree topology was obtained independently from each of the GAG and DPLA data sets, although not statistically significant; the minimum log-likelihood differences between the ML tree and the second best tree were 12.3 ± 7.71 (P(KH) = 0.060, P(AU) = 0.072) and 8.02 ± 9.38 (P(KH) = 0.20, P(AU) = 0.39) for DPLA and GAG, respectively.
FIG. 2.— The ML tree inferred from DPLA and GAG. Branch lengths were calculated from the concatenated alignment of DPLA and GAG. RELL BP values calculated from the concatenated alignment/DPLA/GAG are shown in this order for each branch.
Table 1 Comparison of Log-Likelihood Values Based on the DPLA and GAG Proteins
When the DPLA and GAG sequences were concatenated together with the previously available data (total length was 5,189 aa), hypothesis c was strongly supported, whereas all other tree topologies were significantly rejected (P(KH) < 0.026, P(AU) < 0.032), as shown in table 2.
Table 2 Comparison of Log-Likelihood Values Based on GAG, DPLA, and Previously Available Data
We also carried out separate analyses, in which the log-likelihood values were separately calculated for 15 proteins and then summed, and obtained similar results to those from concatenated alignment analyses (the right three columns of tables 1 and 2).
We analyzed the concatenated alignment of previously available amino acid sequences (2,994 aa), although the possibility of paralogy could not be completely excluded. This data also supported hypothesis c, but the difference from the second best tree was only 1.69 ± 9.03 (P(KH) = 0.56, P(AU) = 0.57). This result disagreed with Hedges and Poling (1999), which supported hypothesis d based on the nuclear DNA data available at that time. This discrepancy was probably caused by the amount of available data and/or unrecognized paralogous comparisons.
In summary, we conducted a phylogenetic analysis of amniotes based on large amount of nuclear DNA–coded protein sequence data and obtained a single tree topology supporting hypothesis c, in which turtles are the sister group to a monophyletic cluster of archosaurs (birds and crocodiles). All other tree topologies, including the traditional one, were significantly rejected.
Materials and Methods
Total RNAs were extracted from liver of Caiman crocodilus (spectacled caiman), tail of Trachemys scripta (red-eared slider), tail of Iguana iguana (green iguana), tail of Ambystoma mexicanum (Mexican axolotl), and embryo of Gallus gallus (chicken) using TRIZOL reagent (Invitrogen, Carlsbad, Calif.). These total RNAs were reverse-transcribed to cDNAs using SMART RACE cDNA Amplification Kit (Clontech, Palo Alto, Calif.). These cDNAs were used as templates for PCR amplification with Expand High-Fidelity PCR System (Roche, Basel, Switzerland). The sense and antisense degenerate primers were designed from conserved amino acid residues of each gene. As fragments of chicken DPLA gene were found in the dbEST database of the National Center for Biotechnology Information, gene specific primers were used for this gene. The PCR products were purified and cloned into the pT7Blue vector (Novagen, Darmstadt, Germany). More than three independent clones were isolated and sequenced using ABI 3100 DNA Sequencer (Applied Biosystems, Foster City, Calif.). The 3' ends of cDNAs were amplified using 3'RACE (GIBCO BRL, Invitrogen) and sequenced in the same way as above. The following sequence data were taken from the GenBank database: human and mouse DPLA; human, mouse, and chicken GAG; and previously available data listed in supplementary table.
Tuatara, one of the four major groups of reptiles (crocodiles, turtles, squamates, and tuatara), is thought to be closely related to squamates based on morphological data (Caroll 1988) and mitochondrial sequence data (Rest et al. 2003). Thus, tuatara was excluded from the present analysis.
Chicken has a pair of duplicated GAG genes (GAG-A and -B) (Smith et al. 2000). Chicken GAG-B was estimated to be more closely related to chicken GAG-A than to any of the reptile GAG genes, in distance measured by synonymous substitutions (Miyata and Yasunaga 1980). According to a comparison of amino acid substitution rates of amniote GAG genes using amphibian sequences as an out-group, the chicken GAG-B gene was estimated to evolve approximately 3.5–4.5 times faster than the others. These observations suggest that the GAG gene was duplicated on the avian lineage and that one (GAG-B) of the duplicated genes accumulated amino acid changes at an extremely rapid rate. Thus, the chicken GAG-B gene was excluded from the present analysis, in order to avoid the long-branch attraction artifact (Felsenstein 1978).
Each protein data set was multiply aligned by MAFFT (Katoh et al. 2002) and manually inspected. Unambiguously aligned amino acid positions were subjected to phylogenetic tree analyses based on the ML method (Felsenstein 1981; Kishino, Miyata, and Hasegawa 1990). For each of 945 possible topologies consisting of seven taxonomic groups (six amniote groups and an out-group), the log-likelihood values based on GAG (856 aa), DPLA (1,339 aa), each of the previously available 13 proteins, and three types of concatenated alignments (DPLA + GAG, 2,195 aa; DPLA + GAG + previously available data, 5,189 aa; and previously available data only, 2,994 aa) were calculated using the GAMT (Katoh, Kuma, and Miyata 2001) program, assuming the JTT-F model (Jones, Taylor, and Thornton 1992; Cao et al. 1994; Adachi and Hasegawa 1996). Heterogeneity of evolutionary rates among sites was modeled by a discrete distribution (Yang 1994) with the shape parameter optimized for each data set.
Two-sided KH test and the AU test were carried out using the CONSEL package (Shimodaira and Hasegawa 2001). Bootstrap probability (BP) based on the resampling of estimated log-likelihoods (RELL) approximation (Kishino, Miyata, and Hasegawa 1990) was also computed. The bootstrap probability for a cluster is calculated by totaling the RELL BP values of the tree topologies having the cluster.
Supplementary Materials
The sequences reported in this paper have been deposited in the GenBank database (accession numbers AB178525–AB178532).
Acknowledgements
We thank K. Kuma for comments and suggestions. This work was supported in part by a Grant-in-Aid for Creative Scientific Research and a Grant for the Biodiversity Research of the 21st Century COE (A14) from the Ministry of Education, Culture, Sports, Science and Technology of Japan.
References
Adachi, J., and M. Hasegawa. 1996. MOLPHY Version 2.3: programs for molecular phylogenetics based on maximum likelihood. Comput. Sci. Monogr. 28:1–150.
Cao, Y., J. Adachi, A. Janke, S. Paabo, and M. Hasegawa. 1994. Phylogenetic relationships among eutherian orders estimated from inferred sequences of mitochondrial proteins: instability of a tree based on a single gene. J. Mol. Evol. 39:519–527.
Cao, Y., M. D. Sorenson, Y. Kumazawa, D. P. Mindell, and M. Hasegawa. 2000. Phylogenetic position of turtles among amniotes: evidence from mitochondrial and nuclear genes. Gene 259:139–148.
Caroll, R. L. 1988. Vertebrate paleontology and evolution. Freeman, New York.
deBraga, M., and O. Rieppel. 1997. Reptile phylogeny and the interrelationships of turtles. Zool. J. Linn. Soc. 120:281–354.
Felsenstein, J. 1978. Cases in which parsimony or compatibility methods will be positively misleading. Syst. Zool. 27:401–410.
———. 1981. Evolutionary trees from DNA sequences: a maximum likelihood approach. J. Mol. Evol. 17:368–376.
Hedges, S. B. 2001. Molecular evidence for the early history of living vertebrates. Pp. 119–134 in P. E. Ahlberg, ed. Major events in early vertebrate evolution. Taylor & Francis, London.
Hedges, S. B., and L. L. Poling. 1999. A molecular phylogeny of reptiles. Science 283:998–1001.
Jones, D. T., W. R. Taylor, and J. M. Thornton. 1992. The rapid generation of mutation data matrices from protein sequences. Comput. Appl. Biosci. 8:275–282.
Katoh, K., K. Kuma, and T. Miyata. 2001. Genetic algorithm-based maximum-likelihood analysis for molecular phylogeny. J. Mol. Evol. 53:477–484.
Katoh, K., K. Misawa, K. Kuma, and T. Miyata. 2002. MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic Acids Res. 30:3059–3066.
Kikugawa, K., K. Katoh, S. Kuraku, H. Sakurai, O. Ishida, N. Iwabe, and T. Miyata. 2004. Basal jawed vertebrate phylogeny inferred from multiple nuclear DNA-coded genes. BMC Biol. 2:3.
Kishino, H., T. Miyata, and M. Hasegawa. 1990. Maximum likelihood inference of protein phylogeny and the origin of chloroplasts. J. Mol. Evol. 31:151–160.
Kumazawa, Y., and M. Nishida. 1999. Complete mitochondrial DNA sequences of the green turtle and blue-tailed mole skink: statistical evidence for archosaurian affinity of turtles. Mol. Biol. Evol. 16:784–792.
Lee, M. S. Y. 1997. Pareiasaur phylogeny and the origin of turtles. Zool. J. Linn. Soc. 120:197–280.
Mannen, H., and S. S. Li. 1999. Molecular evidence for a clade of turtles. Mol. Phylogenet. Evol. 13:144–148.
Miyata, T., and T. Yasunaga. 1980. Molecular evolution of mRNA: a method for estimating evolutionary rates of synonymous and amino acid substitutions from homologous nucleotide sequences and its application. J. Mol. Evol. 16:23–36.
Rest, J. S., J. C. Ast, C. C. Austin, P. J. Waddell, E. A. Tibbetts, J. M. Hay, and D. P. Mindell. 2003. Molecular systematics of primary reptilian lineages and the tuatara mitochondrial genome. Mol. Phylogenet. Evol. 29:289–297.
Rieppel, O., and M. deBraga. 1996. Turtles as diapsid reptiles. Nature 384:453–455.
Shimodaira, H., and M. Hasegawa. 2001. CONSEL: for assessing the confidence of phylogenetic tree selection. Bioinformatics 17:1246–1247.
Smith, J., J. Reboul, G. Lutfalla, and D. W. Burt. 2000. Human chromosomes 3 and 21 are the products of an ancestral gene arrangement that is at least 300 million years old. Mamm. Genome 11:806–807.
Takezaki, N., F. Figueroa, Z. Zaleska-Rutczynska, N. Takahata, and J. Klein. 2004. The phylogenetic relationship of tetrapod, coelacanth, and lungfish revealed by the sequences of forty-four nuclear genes. Mol. Biol. Evol. 21:1512–1524.
Yang, Z. 1994. Maximum likelihood phylogenetic estimation from DNA sequences with variable rates over sites: approximate methods. J. Mol. Evol. 39:306–314.
Zardoya, R., and A. Meyer. 1998. Complete mitochondrial genome suggests diapsid affinities of turtles. Proc. Natl. Acad. Sci. USA 95:14226–14231.(Naoyuki Iwabe*, Yuichiro )