- Email: [email protected]
Molecular Evolution of Myelin Proteolipid Protein
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.
237, 559–561 (1997)
RC977179
Molecular Evolution of Myelin Proteolipid Protein Tadashi Kurihara,1 Mitsuhiro Sakuma, and Takashi Gojobori* Institute of Life Science, Soka University, Hachioji, Tokyo 192, Japan; and *National Institute of Genetics, Mishima, Shizuoka 411, Japan
Received June 12, 1997
We show that the major membrane protein of central nervous system myelin, proteolipid protein, evolved much more rapidly than it does now more than 300 million years ago. We reason that myelin proteolipid protein evolved rapidly just after its appearance in vertebrates and that its evolutionary rate then gradually decreased. Comparison of the rates between the synonymous and nonsynonymous nucleotide substitutions for the cDNA suggests the possibility that positive selection operated on myelin proteolipid protein at least when it appeared in vertebrates. q 1997 Academic Press
The amino-acid sequences of human, dog, rat, mouse, chicken, and Xenopus laevis PLPs from the DDBJ/EMBL/GenBank databases (accession numbers M15026-32, X55317, M11185, M15442, X61661, and Z19522-3) and the amino-acid sequences of bovine (8) and zebra finch (9) PLPs were used. The amino-acid sequence of shark DM20 (DMa) (10) was compared with the DM20 region of amphibian, avian, and mammalian PLPs. The cytochrome c sequences (11) for human, bovine, dog, rabbit, chicken, emu, penguin, duck, pigeon, turtle, snake, bullfrog, tuna, bonito, shark, and lamprey were used. Number of synonymous and nonsynonymous nucleotide substitutions per site was estimated by the procedure of Nei and Gojobori (12). The cDNA sequences of human, dog, rat, mouse, chicken, and Xenopus laevis PLPs from the above databases and the cDNA sequence of zebra finch PLP (9) were used.
RESULTS Myelin is a multilamellar membrane structure ensheathing axons and is found in all vertebrates except the phylogenically oldest lampreys (1). Myelin accounts functionally for fast nerve conduction (2). Proteolipid protein (PLP) is the major membrane protein constituting about 50% of the protein component of mammalian central nervous system (CNS) myelin (1). Several dysmyelinating disorders result from mutations in PLP gene, which dramatically alter oligodendrocyte (myelin forming cell) differentiation and myelin development (3). Since the proposal of the ‘‘molecular evolutionary clock’’ (4), it has been generally accepted that the evolutionary rate of each protein in terms of amino-acid substitution is approximately constant (5-7). We show here however that PLP evolved much more rapidly than it does now more than 300 million years ago. METHODS Number of amino-acid substitutions per site was estimated by the empirical equation of Kimura for Dayhoff correction (6), K Å 0ln(1 0 p 0 0.2p2), where p represents the fraction of observed amino-acid differences. Two amino-acid sequences were aligned by maximum matching (Genetyx) and gaps were excluded. If the sequence was determined chemically, initiator methionine was added to the amino terminus. 1 Corresponding author. Fax: 81-426-91-9315. E-mail: kurihara@ t.soka.ac.jp. Abbreviations: PLP, proteolipid protein; CNS, central nervous system.
It has been frequently stated that the amino-acid sequence of PLP is extremely conservative among mammalian species (1, 3, 13). The PLP sequence consisting of 277 amino-acid residues is 100% conserved among rat, mouse, and human. There is one amino-acid difference in dog PLP and there are four amino-acid differences in bovine PLP relative to the rat, mouse, and human PLPs. The evolutionary rate of PLP in the recent 80 million years was below 0.1 1 1009 amino-acid substitutions per site per year from Fig. 1A, indicating that PLP indeed changes very slowly now. However, PLP evolved much more rapidly than it does now more than 300 million years ago (before the divergence of reptiles/ aves from the mammalian line) (Fig. 1A). Schliess and Stoffel (14) have already pointed out that the sequence of Xenopus PLP deviates considerably from those of mammalian PLPs. The DM20 isoform of PLP in bony and cartilaginous fishes was recently identified as a highly conserved component of fish CNS myelin (15) (see the last paragraph of the Discussion section). When the substitution of shark DM20 (DMa) (10) was plotted, DMa was found at a position extrapolated from the amphibian PLP (Fig. 1A). Fig. 1B shows the typical clock-like substitution of cytochrome c calculated by the same procedure. The evolutionary change of PLP (Fig.1A) seems to be markedly different from that of cytochrome c. Reptile sequence for PLP is unavailable at present, and we used
559
0006-291X/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AID
BBRC 7179
/
6936$$$101
08-12-97 13:49:29
bbrcg
AP: BBRC
Vol. 237, No. 3, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
tution values of avian PLP. The substitution of avian PLP before and after this correction was plotted in Fig. 1A and the line was drawn using the corrected substitution. The rapid evolution of PLP more than 300 million years ago was still evident after the correction of the avian substitution (Fig. 1A). Vertebrates acquired PLP after the divergence of bony fishes (380-400 million years ago) and DM20 after the divergence of lampreys (420-450 million years ago). We reason from Fig. 1A that PLP/DM20 evolved rapidly just after it appeared in vertebrates, and that the evolutionary rate then gradually decreased. Fig. 2 shows the evolutionary change of PLP/DM20 cDNA. Synonymous (silent) and nonsynonymous (aminoacid altering) nucleotide substitutions were estimated separately. The synonymous substitution occurred rapidly and its rate was roughly constant in the recent 300 million years. The rate of the synonymous substitution more than 300 million years ago, however, could not be determined due to the great variation of the substitution values; this situation seems to be inevitable,
FIG. 1. Evolutionary changes of PLP (A) and cytochrome c (B). Divergence time indicated is based on fossil records (7). The time after the divergence of amphibia from the mammalian line was assumed to be 380 million years, because a large amphibian fossil has been discovered from the zone of 370 million years ago (19). The bar indicates the time when PLP or DM20 appeared in vertebrates. (A) s, amphibian (Xenopus) or mammalian PLP; m, avian (chicken and zebra finch) PLP; n, avian (chicken and zebra finch) PLP, the substitution values corrected by the avian factor (0.70) of cytochrome c; h, fish DM20 (shark DMa) or DM20 region of amphibian (Xenopus) PLP. Each point represents a single determination or mean value of 9 to 14 determinations { SD. (B) m, avian cytochrome c; s, cytochrome c of other vertebrates. Each point represents a mean value of 6 to 20 determinations { SD.
avian (chicken and zebra finch) PLP sequences instead. The substitutions of avian proteins, however, are lower than those of reptile proteins (7), which might affect the evolutionary profile of PLP constructed from the limited number of points. The substitution values of avian PLP are about the same as those of avian cytochrome c. We used therefore the avian factor of cytochrome c (0.70; the deviation of the avian substitution from the linear slope, Fig. 1B) for correcting the substi-
FIG. 2. Evolutionary change of PLP cDNA. L, synonymous substitution of PLP cDNA; s, nonsynonymous substitution of PLP cDNA; h, nonsynonymous substitution of fish DM20 cDNA or corresponding sequence of amphibian PLP cDNA. The nonsynonymous substitution of avian PLP cDNA was corrected by the avian factor (0.70). Each point represents a single determination or mean value of 5 to 12 determinations { SD.
560
AID
BBRC 7179
/
6936$$$101
08-12-97 13:49:29
bbrcg
AP: BBRC
Vol. 237, No. 3, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
since the reliability of the estimates declines as the number of nucleotide substitutions increases (7). The nonsynonymous substitution occurred much more slowly than the synonymous substitution in the recent 300 million years. However, if approximate linearity of the synonymous substitution is assumed, the rate (slope) of the nonsynonymous substitution 380-420 million years ago seems to have been at the same level as, or even higher than, that of the synonymous substitution (Fig. 2). The higher rate of the nonsynonymous substitution than that of the synonymous substitution is regarded as evidence for positive selection (16, 17), though positive selection does not necessarily cause the higher rate of the nonsynonymous substitution. The results suggest the possibility that positive selection operated on PLP/DM20 at least when it appeared in vertebrates. DISCUSSION We suspect that 1) more than 300 million years ago the function of PLP had not been established; 2) the more advanced function of PLP gave advantages to animals; 3) the rapid substitution of PLP more than 300 million years ago contained, probably minor, adaptive mutations; 4) the effect of purifying selection on PLP gradually increased as its function gradually became established; 5) the evolutionary change of PLP represents the process that a young revolutionary gene becomes a conservative gene; this transition has been assumed, but has not been traced so far. The results are based on the assumption that the sequences used are those of the homologous PLP molecules. In mammals it has been established that PLP gene is a single-copy gene located in X chromosome whose mutations cause dysmyelinating disorders (3, 13). In chicken and zebra finch only a single PLP gene is known and no other similar gene is detectable (9, 14). Xenopus PLP gene is duplicated, as indicated by Southern hybridization (K. A. Nave, personal communication); the two Xenopus PLPs are different from each other in eight amino-acid residues (see the DDBJ/ EMBL/GenBank databases under accession numbers Z19522-3). The two Xenopus PLP sequences show the same mean substitution values (0.36 substitutions per site) when compared with those of mammalian and avian PLPs, suggesting that the two Xenopus PLPs are functionally equivalent. In bony and cartilaginous fishes, P0 protein (an adhesive protein belonging to the immunoglobulin superfamily) in place of PLP has been supposed to be the intrinsic membrane protein of CNS myelin (1). P0 protein seems to have been replaced by PLP in the terrestrial vertebrate CNS. Yoshida and Colman (15), however, recently showed that DM20 isoform, where a positively charged 35 amino-acid sequence is deleted from
PLP, coexists with P0 protein in bony and cartilaginous fishes. Fish DM20 is probably the ancestor of terrestrial vertebrate PLP (15), and shark DM20 (DMa) sequence (10) was used in the present study. PLP in amphibia coexists with P0 protein (18); fish DM20 and amphibian PLP may be functionally incomplete, requiring the presence of P0 protein. The above authors suggest that the acquisition of adhesive properties by PLP/DM20, coupled with the introduction of the positively charged PLP segment, was permissive for the drop-out of P0 protein from the vertebrate CNS (15). Our data suggest that fish DM20 and terrestrial vertebrate PLP evolved actively before the drop-out of P0 protein from the vertebrate CNS (Fig. 1A). ACKNOWLEDGMENTS We thank Dr. T. Endo for his help and Drs. H. Kojima and Y. Tamai for critical reading of the manuscript. This study was supported in part by a grant (08680873) from the Ministry of Education, Science, and Culture of Japan.
REFERENCES 1. Kirschner, D. A., and Blaurock, A. E. (1992) in Myelin: Biology and Chemistry (Martenson, R. E., Ed.), pp. 3–78, CRC Press, Boca Raton, FL. 2. Rogart, R. B., and Ritchie, J. M. (1977) in Myelin (Morell, P., Ed.), pp. 117–159, Plenum Press, New York. 3. Hudson, L. D., and Nadon, N. L. (1992) in Myelin: Biology and Chemistry (Martenson, R. E., Ed.), pp. 677–702, CRC Press, Boca Raton, FL. 4. Zuckerkandl, E., and Pauling, L. (1965) in Evolving Genes and Proteins (Bryson, V., and Vogel, H. J., Eds.), pp. 97–166, Academic Press, New York. 5. Dickerson, R. E. (1971) J. Mol. Evol. 1, 26–45. 6. Kimura, M. (1983) The Neutral Theory of Molecular Evolution, Cambridge Univ. Press, Cambridge. 7. Nei, M. (1987) Molecular Evolutionary Genetics, Columbia Univ. Press, New York. 8. Lees, M. B., Lin, L.-F. H., Samiullah, M., and Laursen, R. A. (1983) Arch. Biochem. Biophys. 226, 643–656. 9. Campagnoni, C. W., Kampf, K., Mason, B., Handley, V. W., and Campagnoni, A. T. (1994) Neurochem. Res. 19, 1061–1065. 10. Kitagawa, K., Sinoway, M. P., Yang, C., Gould, R. M., and Colman, D. R. (1993) Neuron 11, 433–448. 11. Dickerson, R. E., and Timkovich, R. (1975) in Enzymes (Boyer, P. D., Ed.), Vol. 11, 3rd ed., pp. 397–547, Academic Press, New York. 12. Nei, M., and Gojobori, T. (1986) Mol. Biol. Evol. 3, 418–426. 13. Macklin, W. B. (1992) in Myelin: Biology and Chemistry (Martenson, R. E., Ed.), pp. 257–276, CRC Press, Boca Raton, FL. 14. Schliess, F., and Stoffel, W. (1991) Biol. Chem. Hoppe-Seyler 372, 865–874. 15. Yoshida, M., and Colman, D. R. (1996) Neuron 16, 1115–1126. 16. Hughes, A. L., and Nei, M. (1988) Nature 335, 167–170. 17. Endo, T., Ikeo, K., and Gojobori, T. (1996) Mol. Biol. Evol. 13, 685–690. 18. Takei, K., and Uyemura, K. (1993) Comp. Biochem. Physiol. 106B, 873–882. 19. Cowen, R. (1994) History of Life, 2nd ed., Blackwell Scientific, Boston.
561
AID
BBRC 7179
/
6936$$$101
08-12-97 13:49:29
bbrcg
AP: BBRC
237, 559–561 (1997)
RC977179
Molecular Evolution of Myelin Proteolipid Protein Tadashi Kurihara,1 Mitsuhiro Sakuma, and Takashi Gojobori* Institute of Life Science, Soka University, Hachioji, Tokyo 192, Japan; and *National Institute of Genetics, Mishima, Shizuoka 411, Japan
Received June 12, 1997
We show that the major membrane protein of central nervous system myelin, proteolipid protein, evolved much more rapidly than it does now more than 300 million years ago. We reason that myelin proteolipid protein evolved rapidly just after its appearance in vertebrates and that its evolutionary rate then gradually decreased. Comparison of the rates between the synonymous and nonsynonymous nucleotide substitutions for the cDNA suggests the possibility that positive selection operated on myelin proteolipid protein at least when it appeared in vertebrates. q 1997 Academic Press
The amino-acid sequences of human, dog, rat, mouse, chicken, and Xenopus laevis PLPs from the DDBJ/EMBL/GenBank databases (accession numbers M15026-32, X55317, M11185, M15442, X61661, and Z19522-3) and the amino-acid sequences of bovine (8) and zebra finch (9) PLPs were used. The amino-acid sequence of shark DM20 (DMa) (10) was compared with the DM20 region of amphibian, avian, and mammalian PLPs. The cytochrome c sequences (11) for human, bovine, dog, rabbit, chicken, emu, penguin, duck, pigeon, turtle, snake, bullfrog, tuna, bonito, shark, and lamprey were used. Number of synonymous and nonsynonymous nucleotide substitutions per site was estimated by the procedure of Nei and Gojobori (12). The cDNA sequences of human, dog, rat, mouse, chicken, and Xenopus laevis PLPs from the above databases and the cDNA sequence of zebra finch PLP (9) were used.
RESULTS Myelin is a multilamellar membrane structure ensheathing axons and is found in all vertebrates except the phylogenically oldest lampreys (1). Myelin accounts functionally for fast nerve conduction (2). Proteolipid protein (PLP) is the major membrane protein constituting about 50% of the protein component of mammalian central nervous system (CNS) myelin (1). Several dysmyelinating disorders result from mutations in PLP gene, which dramatically alter oligodendrocyte (myelin forming cell) differentiation and myelin development (3). Since the proposal of the ‘‘molecular evolutionary clock’’ (4), it has been generally accepted that the evolutionary rate of each protein in terms of amino-acid substitution is approximately constant (5-7). We show here however that PLP evolved much more rapidly than it does now more than 300 million years ago. METHODS Number of amino-acid substitutions per site was estimated by the empirical equation of Kimura for Dayhoff correction (6), K Å 0ln(1 0 p 0 0.2p2), where p represents the fraction of observed amino-acid differences. Two amino-acid sequences were aligned by maximum matching (Genetyx) and gaps were excluded. If the sequence was determined chemically, initiator methionine was added to the amino terminus. 1 Corresponding author. Fax: 81-426-91-9315. E-mail: kurihara@ t.soka.ac.jp. Abbreviations: PLP, proteolipid protein; CNS, central nervous system.
It has been frequently stated that the amino-acid sequence of PLP is extremely conservative among mammalian species (1, 3, 13). The PLP sequence consisting of 277 amino-acid residues is 100% conserved among rat, mouse, and human. There is one amino-acid difference in dog PLP and there are four amino-acid differences in bovine PLP relative to the rat, mouse, and human PLPs. The evolutionary rate of PLP in the recent 80 million years was below 0.1 1 1009 amino-acid substitutions per site per year from Fig. 1A, indicating that PLP indeed changes very slowly now. However, PLP evolved much more rapidly than it does now more than 300 million years ago (before the divergence of reptiles/ aves from the mammalian line) (Fig. 1A). Schliess and Stoffel (14) have already pointed out that the sequence of Xenopus PLP deviates considerably from those of mammalian PLPs. The DM20 isoform of PLP in bony and cartilaginous fishes was recently identified as a highly conserved component of fish CNS myelin (15) (see the last paragraph of the Discussion section). When the substitution of shark DM20 (DMa) (10) was plotted, DMa was found at a position extrapolated from the amphibian PLP (Fig. 1A). Fig. 1B shows the typical clock-like substitution of cytochrome c calculated by the same procedure. The evolutionary change of PLP (Fig.1A) seems to be markedly different from that of cytochrome c. Reptile sequence for PLP is unavailable at present, and we used
559
0006-291X/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.
AID
BBRC 7179
/
6936$$$101
08-12-97 13:49:29
bbrcg
AP: BBRC
Vol. 237, No. 3, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
tution values of avian PLP. The substitution of avian PLP before and after this correction was plotted in Fig. 1A and the line was drawn using the corrected substitution. The rapid evolution of PLP more than 300 million years ago was still evident after the correction of the avian substitution (Fig. 1A). Vertebrates acquired PLP after the divergence of bony fishes (380-400 million years ago) and DM20 after the divergence of lampreys (420-450 million years ago). We reason from Fig. 1A that PLP/DM20 evolved rapidly just after it appeared in vertebrates, and that the evolutionary rate then gradually decreased. Fig. 2 shows the evolutionary change of PLP/DM20 cDNA. Synonymous (silent) and nonsynonymous (aminoacid altering) nucleotide substitutions were estimated separately. The synonymous substitution occurred rapidly and its rate was roughly constant in the recent 300 million years. The rate of the synonymous substitution more than 300 million years ago, however, could not be determined due to the great variation of the substitution values; this situation seems to be inevitable,
FIG. 1. Evolutionary changes of PLP (A) and cytochrome c (B). Divergence time indicated is based on fossil records (7). The time after the divergence of amphibia from the mammalian line was assumed to be 380 million years, because a large amphibian fossil has been discovered from the zone of 370 million years ago (19). The bar indicates the time when PLP or DM20 appeared in vertebrates. (A) s, amphibian (Xenopus) or mammalian PLP; m, avian (chicken and zebra finch) PLP; n, avian (chicken and zebra finch) PLP, the substitution values corrected by the avian factor (0.70) of cytochrome c; h, fish DM20 (shark DMa) or DM20 region of amphibian (Xenopus) PLP. Each point represents a single determination or mean value of 9 to 14 determinations { SD. (B) m, avian cytochrome c; s, cytochrome c of other vertebrates. Each point represents a mean value of 6 to 20 determinations { SD.
avian (chicken and zebra finch) PLP sequences instead. The substitutions of avian proteins, however, are lower than those of reptile proteins (7), which might affect the evolutionary profile of PLP constructed from the limited number of points. The substitution values of avian PLP are about the same as those of avian cytochrome c. We used therefore the avian factor of cytochrome c (0.70; the deviation of the avian substitution from the linear slope, Fig. 1B) for correcting the substi-
FIG. 2. Evolutionary change of PLP cDNA. L, synonymous substitution of PLP cDNA; s, nonsynonymous substitution of PLP cDNA; h, nonsynonymous substitution of fish DM20 cDNA or corresponding sequence of amphibian PLP cDNA. The nonsynonymous substitution of avian PLP cDNA was corrected by the avian factor (0.70). Each point represents a single determination or mean value of 5 to 12 determinations { SD.
560
AID
BBRC 7179
/
6936$$$101
08-12-97 13:49:29
bbrcg
AP: BBRC
Vol. 237, No. 3, 1997
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
since the reliability of the estimates declines as the number of nucleotide substitutions increases (7). The nonsynonymous substitution occurred much more slowly than the synonymous substitution in the recent 300 million years. However, if approximate linearity of the synonymous substitution is assumed, the rate (slope) of the nonsynonymous substitution 380-420 million years ago seems to have been at the same level as, or even higher than, that of the synonymous substitution (Fig. 2). The higher rate of the nonsynonymous substitution than that of the synonymous substitution is regarded as evidence for positive selection (16, 17), though positive selection does not necessarily cause the higher rate of the nonsynonymous substitution. The results suggest the possibility that positive selection operated on PLP/DM20 at least when it appeared in vertebrates. DISCUSSION We suspect that 1) more than 300 million years ago the function of PLP had not been established; 2) the more advanced function of PLP gave advantages to animals; 3) the rapid substitution of PLP more than 300 million years ago contained, probably minor, adaptive mutations; 4) the effect of purifying selection on PLP gradually increased as its function gradually became established; 5) the evolutionary change of PLP represents the process that a young revolutionary gene becomes a conservative gene; this transition has been assumed, but has not been traced so far. The results are based on the assumption that the sequences used are those of the homologous PLP molecules. In mammals it has been established that PLP gene is a single-copy gene located in X chromosome whose mutations cause dysmyelinating disorders (3, 13). In chicken and zebra finch only a single PLP gene is known and no other similar gene is detectable (9, 14). Xenopus PLP gene is duplicated, as indicated by Southern hybridization (K. A. Nave, personal communication); the two Xenopus PLPs are different from each other in eight amino-acid residues (see the DDBJ/ EMBL/GenBank databases under accession numbers Z19522-3). The two Xenopus PLP sequences show the same mean substitution values (0.36 substitutions per site) when compared with those of mammalian and avian PLPs, suggesting that the two Xenopus PLPs are functionally equivalent. In bony and cartilaginous fishes, P0 protein (an adhesive protein belonging to the immunoglobulin superfamily) in place of PLP has been supposed to be the intrinsic membrane protein of CNS myelin (1). P0 protein seems to have been replaced by PLP in the terrestrial vertebrate CNS. Yoshida and Colman (15), however, recently showed that DM20 isoform, where a positively charged 35 amino-acid sequence is deleted from
PLP, coexists with P0 protein in bony and cartilaginous fishes. Fish DM20 is probably the ancestor of terrestrial vertebrate PLP (15), and shark DM20 (DMa) sequence (10) was used in the present study. PLP in amphibia coexists with P0 protein (18); fish DM20 and amphibian PLP may be functionally incomplete, requiring the presence of P0 protein. The above authors suggest that the acquisition of adhesive properties by PLP/DM20, coupled with the introduction of the positively charged PLP segment, was permissive for the drop-out of P0 protein from the vertebrate CNS (15). Our data suggest that fish DM20 and terrestrial vertebrate PLP evolved actively before the drop-out of P0 protein from the vertebrate CNS (Fig. 1A). ACKNOWLEDGMENTS We thank Dr. T. Endo for his help and Drs. H. Kojima and Y. Tamai for critical reading of the manuscript. This study was supported in part by a grant (08680873) from the Ministry of Education, Science, and Culture of Japan.
REFERENCES 1. Kirschner, D. A., and Blaurock, A. E. (1992) in Myelin: Biology and Chemistry (Martenson, R. E., Ed.), pp. 3–78, CRC Press, Boca Raton, FL. 2. Rogart, R. B., and Ritchie, J. M. (1977) in Myelin (Morell, P., Ed.), pp. 117–159, Plenum Press, New York. 3. Hudson, L. D., and Nadon, N. L. (1992) in Myelin: Biology and Chemistry (Martenson, R. E., Ed.), pp. 677–702, CRC Press, Boca Raton, FL. 4. Zuckerkandl, E., and Pauling, L. (1965) in Evolving Genes and Proteins (Bryson, V., and Vogel, H. J., Eds.), pp. 97–166, Academic Press, New York. 5. Dickerson, R. E. (1971) J. Mol. Evol. 1, 26–45. 6. Kimura, M. (1983) The Neutral Theory of Molecular Evolution, Cambridge Univ. Press, Cambridge. 7. Nei, M. (1987) Molecular Evolutionary Genetics, Columbia Univ. Press, New York. 8. Lees, M. B., Lin, L.-F. H., Samiullah, M., and Laursen, R. A. (1983) Arch. Biochem. Biophys. 226, 643–656. 9. Campagnoni, C. W., Kampf, K., Mason, B., Handley, V. W., and Campagnoni, A. T. (1994) Neurochem. Res. 19, 1061–1065. 10. Kitagawa, K., Sinoway, M. P., Yang, C., Gould, R. M., and Colman, D. R. (1993) Neuron 11, 433–448. 11. Dickerson, R. E., and Timkovich, R. (1975) in Enzymes (Boyer, P. D., Ed.), Vol. 11, 3rd ed., pp. 397–547, Academic Press, New York. 12. Nei, M., and Gojobori, T. (1986) Mol. Biol. Evol. 3, 418–426. 13. Macklin, W. B. (1992) in Myelin: Biology and Chemistry (Martenson, R. E., Ed.), pp. 257–276, CRC Press, Boca Raton, FL. 14. Schliess, F., and Stoffel, W. (1991) Biol. Chem. Hoppe-Seyler 372, 865–874. 15. Yoshida, M., and Colman, D. R. (1996) Neuron 16, 1115–1126. 16. Hughes, A. L., and Nei, M. (1988) Nature 335, 167–170. 17. Endo, T., Ikeo, K., and Gojobori, T. (1996) Mol. Biol. Evol. 13, 685–690. 18. Takei, K., and Uyemura, K. (1993) Comp. Biochem. Physiol. 106B, 873–882. 19. Cowen, R. (1994) History of Life, 2nd ed., Blackwell Scientific, Boston.
561
AID
BBRC 7179
/
6936$$$101
08-12-97 13:49:29
bbrcg
AP: BBRC