Determining divergence times of the major kingdoms of living organisms with a protein clock

Science

Volumn 271, Issue 5248, 1996, Pages 470-477

  Doolittle, Russell F ^a   Feng, Da Fei ^a   Tsang, Simon ^a   Cho, Glen ^a   Little, Elizabeth ^a  

^a UNIVERSITY OF CALIFORNIA   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

ENZYME;

AMINO ACID SEQUENCE; ARCHAEBACTERIUM; ARTICLE; BIODIVERSITY; EUBACTERIUM; EUKARYOTE; EVOLUTION; FUNGUS; NONHUMAN; PHYLOGENY; PLANT; PRIORITY JOURNAL; PROKARYOTE;

ANIMALIA; ARCHAEA; BACTERIA (MICROORGANISMS); EUKARYOTA; FUNGI; PROKARYOTA; PROTISTA;

AMINO ACID; ANIMAL KINGDOM; DIVERGENCE TIME; ENZYME; FUNGUS; PLANT KINGDOM; PRECAMBRIAN;

EID: 0030023247     PISSN: 00368075     EISSN: None     Source Type: Journal    
DOI: 10.1126/science.271.5248.470     Document Type: Article

References (68)

1. J W Schopf, Science 260, 640 (1993).
2. A H Knoll, ibid 256, 622 (1992).
3. E Zuckerkandl and L. Pauling, in Evolving Genes and Proteins, V Bryson and H. J. Vogel, Eds. (Academic Press, New York, 1965), pp. 97-166
4. R. F. Doolittle and B. Blomback, Nature 202, 147 (1964)
5. As an example an entire issue of J. Mol Evol. [26, 1-164 (1987)] was devoted to molecular clocks and their use in different settings.
6. T. H. Jukes, Space Life Sci. 1, 469 (1969).
7. P. J. McLaughlin and M. O Dayhoff, Science 168, 1469 (1970), M. Kimura and T Ohta, Nature New Biol. 243, 199 (1973), H Hon and S Osawa, Proc. Natl Acad Sci. U.S.A 76, 381 (1979); M O Dayhoff, Atlas of Protein Sequence and Structure (National Biomedical Research Foundation, Washington, DC, 1978), vol. 5, suppl. 3.
8. P. J. McLaughlin and M. O Dayhoff, Science 168, 1469 (1970), M. Kimura and T Ohta, Nature New Biol. 243, 199 (1973), H Hon and S Osawa, Proc. Natl Acad Sci. U.S.A 76, 381 (1979); M O Dayhoff, Atlas of Protein Sequence and Structure (National Biomedical Research Foundation, Washington, DC, 1978), vol. 5, suppl. 3.
9. P. J. McLaughlin and M. O Dayhoff, Science 168, 1469 (1970), M. Kimura and T Ohta, Nature New Biol. 243, 199 (1973), H Hon and S Osawa, Proc. Natl Acad Sci. U.S.A 76, 381 (1979); M O Dayhoff, Atlas of Protein Sequence and Structure (National Biomedical Research Foundation, Washington, DC, 1978), vol. 5, suppl. 3.
10. P. J. McLaughlin and M. O Dayhoff, Science 168, 1469 (1970), M. Kimura and T Ohta, Nature New Biol. 243, 199 (1973), H Hon and S Osawa, Proc. Natl Acad Sci. U.S.A 76, 381 (1979); M O Dayhoff, Atlas of Protein Sequence and Structure (National Biomedical Research Foundation, Washington, DC, 1978), vol. 5, suppl. 3.
11. J. W. Schopf and D. Z Oehler, Science 193, 47 (1976)
12. C R Woese, in Evolution from Molecules to Men, D. S. Bendall, Ed. (Cambridge Univ. Press, London, 1983), pp. 209-233; M. L Sogin, in New Perspectives on Evolution, L Warren and H. Koprowski, Eds. (Wiley-Liss, New York, 1991), pp. 175-188.
13. C R Woese, in Evolution from Molecules to Men, D. S. Bendall, Ed. (Cambridge Univ. Press, London, 1983), pp. 209-233; M. L Sogin, in New Perspectives on Evolution, L Warren and H. Koprowski, Eds. (Wiley-Liss, New York, 1991), pp. 175-188.
14. Long-term amino acid replacement in proteins is subject to a large number of independently acting factors, each of which may have only a small effect on the overall process In our study we have assumed that the replacement process is approximately uniform, an assumption that is open to question [see J H. Gillespie, The Causes of Molecular Evolution (Oxford Univ. Press, New York, 1991)].
15. R. F Doolittle, K. L Anderson, D. F. Feng, in The Hierarchy of Life. B. Femholm, K. Bremer, H Jomvall, Eds. (Elsevier, Amsterdam, 1989), pp. 73-85.
16. Some enzymes that have the same name and Enzyme Commission (E C.) numbers, but are derived from independent origins, include superoxide dismutases, carbonic anhydrases, aldolases, and senne proteases.
17. W. M. Fitch, Syst Zool 19, 99 (1970)
18. M. W. Smith, D. F. Feng, R F. Doolittle, Trends Biochem. Sci 17, 489 (1992)
19. Most of the 531 amino acid sequences used in our study were taken from Release 42 (30 September 1994) of the Protein Identification Resource (PIR) although some sequences identified recently were abstracted from GenBank. First, a list was compiled of all entries in the PIR that included official E C numbers (16). Then a tally was made of how many different entries were listed for each E.C number Any enzyme with four or more entries was examined to see whether at least three major groups were represented (animal, plants or fungi, and eubacteria); if so, the enzyme was considered a possible candidate for inclusion in the study All told, Release 42 of the PIR contained 13,653 entries with E C. identification numbers Of these, 1262 E.C numbers were present, accounting for just under 40 percent of the officially declared 3196 enzymes (16). About half of these had three entries or fewer and were not considered further. The half with four or more entries was screened with regard to organismic representation. Sequences for enzymes encoded by organellar DNA (mitochondria and chloroplasts) and sequences from viruses were not included. The sequences of candidate groups were aligned and phylogenies were constructed (17-22). If the phylogenetic trees seemed reasonable, by which we mean there was no evidence of horizontal gene transfer or adulteration by paralogous comparisons (23), the sequence subset became a part of the study. The entire set (divided into the six standard enzyme groups) can be obtained by anonymous ftp from juno.ucsd edu. cd to directory pickup.
20. Enzyme Nomenclature, Recommendations of the Nomenclature Committee of the International Union of Biochemistry and Molecular Biology (Academic Press, New York, 1992).
21. Candidate groups were aligned by the progressive method (18), a procedure that uses the Needleman-Wunsch algorithm (19). Several different substitution matrices were used, including the Dayhoff PAM-250 (20), GCB matrix (21), and the BLOSUM-62 matrix (22). The GCB comparisons were not significantly different from those obtained with the Dayhoff PAM-250 scale, and those results have not been included in this study. The BLOSUM-62 scale, however, resulted in obviously improved alignments for the most distant of the relationships. We therefore used it to obtain all the final alignments, even though we then used the PAM-250 table to calculate distances for comparison with those obtained from the BLOSUM table. The comparison data were conveniently managed with the aid of the commercially available Microsoft Excel spreadsheet software Entry sheets listing species represented, lengths of sequences, and such items were prepared for each of the 57 enzymes, as were other sets of primary data sheets that included all resemblances and distances between groups. Summary "charts" of distances and percent identities were prepared from the entire data set or from designated subsets (Table 4). As new data become available, it is relatively easy to update the records and recalculate all values.
22. D. F. Feng and R. F. Doolittle, J Mol. Evol. 25, 351 (1987); Methods Enzymol. 183, 375 (1990).
23. D. F. Feng and R. F. Doolittle, J Mol. Evol. 25, 351 (1987); Methods Enzymol. 183, 375 (1990).
24. S R. Needleman and C. D. Wunsch, J. Mol. Biol 48, 443 (1970).
25. R. M. Schwartz and M O. Dayhoff in Atlas of Protein Sequence and Structure (National Biomedical Research Foundation, Washington, DC, 1978), vol. 5, supp. 3, pp. 353-358.
26. G. H. Gonnet, M. A. Cohen, S. A. Benner, Science 56, 1443 (1992)
27. S Heriikoff and J. G Henikoff, Proteins: Struct Funct. Genet. 17, 49 (1993)
28. Several situations in which displacement during evolution has led to functional convergence after paralogous radiations include bacterial omithine decarboxylase (E.G. 4.1.1.17), which is obviously more like lysine decarboxylase (E.C 4.1.1.18) than it is to eukaryotic omithine decarboxylases, and bacterial tyrosine transaminase (E.G. 2.6.1.5), which is more similar to bacterial aspartate transminase (E.C. 2.6.1.1) than it is to the eukaryotic tyrosine enzyme. Other enzyme sets that were not included on the basis of anomalous phylogenetic trees were catalase (E.G. 1.11 16), pyrroline carboxylate reductase (E.C 1.5.1.2), glutathione reductase (E.C. 1 6.4.2), phosphoribosylglycineamide formyl transferase (E C. 2.1 2.2), transketolase (E.C.2.2 1 1), glycogen phosphorylase (E.C. 2.4.1 1), hypoxanthine transferase (E.G. 2.4 2 8), orotate phosphate transferase (E C. 2.4.2.10), glutathione transferase (E C. 2.5 1 18), galactokinase (E.C 2.7. 1 6), adenylate kinase (E.C. 2.7.4 3), uridine 5′-diphosphate-glucose-hexose phosphate undyl transferase (E.C. 2.7.7.9 and E.C. 2.7.7 12), ornithine decarboxylase (E.C 4.1.1.17), glucose phosphate isomerase (E.C. 5.3.1.9), and phosphoglycerate mutase (E.C 5 4.2.1).
29. R. F. Doolittle, in Methods in Protein Sequence Analysis, K Imahori and F. Sakiyama, Eds. (Plenum, New York, 1993), pp. 241-246.
30. Average percent identities notwithstanding, the data were not entirely consistent. Of the nine possible comparisons, in five cases the archaebacterial sequences clustered with the eukaryotes, and in three with the eubactena. In one case (phosphoglycerate kinase, E C 2.7.2 3) the eubactena and eukaryote sequences were more similar to each other than to the archaebacterial sequence
31. D. F. Feng, M. S. Johnson, R. F. Doolittle, J. Mol. Evol. 21, 112 (1985).
32. The use of the Poisson distribution as a probabilistic model for amino acid replacement dates back to Zuckerkandl and Pauling (3). It is often used in the simple form D = -In (1 - p/n), with p/n being the fraction of changed residues. In this form, the equation mainly corrects for the unobserved occurrence of two or more replacements at the same site (28). Numerous modifications have been reported, including attempts to correct for invariant residues (29, 30) or chance occurrences (37, 26).
33. For example, M Kimura and T. Ohta, J. Mol. Evol 1, 1 (1971), R. E Dickerson. ibid., p 26; M. Kimura and T Ohta, ibid. 2, 87 (1972).
34. For example, M Kimura and T. Ohta, J. Mol. Evol 1, 1 (1971), R. E Dickerson. ibid., p 26; M. Kimura and T Ohta, ibid. 2, 87 (1972).
35. For example, M Kimura and T. Ohta, J. Mol. Evol 1, 1 (1971), R. E Dickerson. ibid., p 26; M. Kimura and T Ohta, ibid. 2, 87 (1972).
36. E. Margoliash and E. Smith, in Evolving Genes and Proteins, V Bryson and H. J Vogel, Eds (Academic Press, New York, 1965), pp. 221-242.
37. W. M. Fitch and E. Markowitz, Biochem. Genet 4, 579 (1970)
38. T. H. Jukes and C. R. Cantor, in Mammalian Protein Metabolism, H N. Munro, Ed (Academic Press, New York, 1969), pp. 21-132.
39. R. Johnson, Elementary Statistics (Duxbury Press, Boston, 1984).
40. C R. Marshall, J. Mol. Evol. 30, 400 (1990), M. J. Benton, ibid., p. 409; R. L. Carroll, Vertebrate Paleontology and Evolution (Freeman, New York, 1988).
41. C R. Marshall, J. Mol. Evol. 30, 400 (1990), M. J. Benton, ibid., p. 409; R. L. Carroll, Vertebrate Paleontology and Evolution (Freeman, New York, 1988).
42. C R. Marshall, J. Mol. Evol. 30, 400 (1990), M. J. Benton, ibid., p. 409; R. L. Carroll, Vertebrate Paleontology and Evolution (Freeman, New York, 1988).
43. S. C. Conway Morris, Nature 361, 219 (1993).
44. The "relative rate test" [V. M Sarich and A. C Wilson, Science 179, 1144 (1973)] can be used whenever the intergroup distances for three taxa (or more) are available. For example, it was clear from a consideration of the fungi-eubacteria and animal-eubactena distances that fungal sequences were changing faster than animal ones We were able to apply this simple test to all the taxa in our study.
45. S. L Baldourf and J D. Palmer, Proc. Natl Acad Sci U S.A. 90, 11558 (1993). P O. Wainnght, G. Hinkle, M L. Sogin, S. K Stickel, Science 260, 340 (1993).
46. S. L Baldourf and J D. Palmer, Proc. Natl Acad Sci U S.A. 90, 11558 (1993). P O. Wainnght, G. Hinkle, M L. Sogin, S. K Stickel, Science 260, 340 (1993).
47. A H Knoll, Proc. Natl Acad. Sci. U.S.A. 91, 6743 (1994)
48. M. L Sogin, J. H. Gunderson, H. J. Elwood, R. A. Alonso, D. Q A Peattie, Science 243, 75 (1989)
49. T. Cavalier-Smith, Microbiol Rev 57, 953 (1993)
50. W. F Loomis and D. W. Smith, Proc. Natl. Acad Sci. U.S.A. 87, 9093 (1990)
51. Z. Zhang, J. Micropalaeontol, 5, 9 (1986).
52. T.-H. Han and B. Runnegar, Science 257, 232 (1992).
53. N. Iwabe, K.-I. Kuma, M. Hasegawa, S. Osawa, T. Miyata, Proc. Natl. Acad. Sci., U.S.A. 86, 9355 (1989).
54. H. Ochman and A. C. Wilson, J. Mol Evol. 26, 74 (1987).
55. Matrices of pairwise distances were examined by the program BLEN (18) which uses a least squares approach (46).
56. L. C Klotz and R L Blanken, J. Theoret. Biol. 91, 261 (1981).
57. C. H. Langley and W. M. Fitch, J. Mol. Evol. 3, 161 (1974). W. M Fitch and C. H. Langley, Fed. Proc. 35, 2092 (1976). W. M. Fitch and F. J. Ayala, Proc. Natl Acad. Sci U.S.A. 91, 6802 (1994).
58. C. H. Langley and W. M. Fitch, J. Mol. Evol. 3, 161 (1974). W. M Fitch and C. H. Langley, Fed. Proc. 35, 2092 (1976). W. M. Fitch and F. J. Ayala, Proc. Natl Acad. Sci U.S.A. 91, 6802 (1994).
59. C. H. Langley and W. M. Fitch, J. Mol. Evol. 3, 161 (1974). W. M Fitch and C. H. Langley, Fed. Proc. 35, 2092 (1976). W. M. Fitch and F. J. Ayala, Proc. Natl Acad. Sci U.S.A. 91, 6802 (1994).
60. An extreme but valid example is the case of enzyme sequences in retroviruses. For example, a survey of 26 ribonuclease H sequences revealed that only 4 of 120 residues remained unchanged; R. F. Doolittle. D-F. Feng, M S. Johnson, M. A. McClure, Q. Rev Biol. 64, 1 (1989).
61. M. Nei, Molecular Evolutionary Genetics (Columbia Univ. Press, New York, 1987).
62. Although, on the average, eukaryote and eubactena sequences are 37 percent identical, the observed fraction of irreplaceable residues was, again on average, only 17 percent. There was also a natural tendency for the fraction of irreplaceable residues to be smaller; the larger the number of sequences in a set, the extrapolated fraction being about 5 percent.
63. A computer model has been constructed that follows the divergence of mutated protein sequences under various circumstances of constraint (R F Doolittle and D. F. Feng, in preparation).
64. One of the earliest estimates made about the prokaryote-eukaryote divergence concluded, on the basis of a relatively small number of transfer RNA sequences, that the split occurred about twice as long ago as the divergence of plants, animals, and fungi (6)
65. There will be some who will remind us of alternative scenarios concerning the origin of eukaryotic organisms, and especially of the possibility that some of the sequences discussed here were actually imported by an archaebacterial symbiont destined to become the nucleus. The fusion of a eubacterial "prokaryote" and an archaebacterium has been widely discussed (54). Although we are skeptical of such models on other grounds, we should point out that such an occurrence would not affect our findings, except that the time we are reporting as a divergence time for eukaryotes and eubacteria would instead chronicle the alleged fusion event.
66. R. S. Gupta and G. B. Golding, J Mol. Evol. 37, 573 (1993); G. B Golding and R S. Gupta, Mol. Biol. Evol. 12, 1 (1995).
67. R. S. Gupta and G. B. Golding, J Mol. Evol. 37, 573 (1993); G. B Golding and R S. Gupta, Mol. Biol. Evol. 12, 1 (1995).
68. We thank K. Anderson for assistance in preparing this manuscript and S Frank, J Gillespie, and two anonymous reviewers for helpful suggestions. Supported in part by NIH grant HL-26873.