Complete genome sequence of the Methanogenic archaeon, Methanococcus jannaschii

Science

Volumn 273, Issue 5278, 1996, Pages 1058-1073

  Bult, Carol J ^f   White, Owen ^f   Olsen, Gary J ^a   Zhou, Lixin ^f   Fleischmann, Robert D ^f   Sutton, Granger G ^f   Blake, Judith A ^f   FitzGerald, Lisa M ^f   Clayton, Rebecca A ^f   Gocayne, Jeannine D ^f   Kerlavage, Anthony R ^f   Dougherty, Brian A ^f   Tomb, Jean Francois ^f   Adams, Mark D ^f   Reich, Claudia I ^a   Overbeek, Ross ^b   Kirkness, Ewen F ^f   Weinstock, Keith G ^f   Merrick, Joseph M ^c   Glodek, Anna ^f   Scott, John L ^f   Geoghagen, Neil S M ^f   Weidman, Janice F ^f   Fuhrmann, Joyce L ^f   Nguyen, Dave ^f   Utterback, Teresa R ^f   Kelley, Jenny M ^f   Peterson, Jeremy D ^f   Sadow, Paul W ^f   Hanna, Michael C ^f   Cotton, Matthew D ^f   Roberts, Kevin M ^f   Hurst, Margaret A ^f   Kaine, Brian P ^a   Borodovsky, Mark ^d   Klenk, Hans Peter ^f   Fraser, Claire M ^f   Smith, Hamilton O ^e   Woese, Carl R ^a   Venter, J Craig ^f   more..

^a UNIVERSITY OF ILLINOIS AT URBANA CHAMPAIGN   (United States)

^b ARGONNE NATIONAL LABORATORY   (United States)

^c UNIVERSITY AT BUFFALO   (United States)

^d GEORGIA INSTITUTE OF TECHNOLOGY   (United States)

^e JOHNS HOPKINS UNIVERSITY SCHOOL OF MEDICINE   (United States)

^f J CRAIG VENTER INSTITUTE   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

DNA POLYMERASE; RIBOSOME RNA; RNA POLYMERASE; TRANSFER RNA;

ARTICLE; BACTERIAL GENETICS; BACTERIAL METABOLISM; BASE PAIRING; DNA REPLICATION; GENE MAPPING; GENOME; METHANOBACTERIUM; MULTIGENE FAMILY; NONHUMAN; NUCLEOTIDE SEQUENCE; PRIORITY JOURNAL;

ARCHAEA; BACTERIA (MICROORGANISMS); EUKARYOTA; METHANOBACTERIUM; METHANOCALDOCOCCUS JANNASCHII; METHANOCOCCUS;

EID: 16044367245     PISSN: 00368075     EISSN: None     Source Type: Journal    
DOI: 10.1126/science.273.5278.1058     Document Type: Article

References (68)

1. G. E. Fox et al., Proc. Natl. Acad. Sci. U.S.A. 74, 4537 (1977); C. R. Woese and G. E. Fox, ibid., p. 5088; C. R. Woese et al., ibid. 87, 4576 (1990).
2. G. E. Fox et al., Proc. Natl. Acad. Sci. U.S.A. 74, 4537 (1977); C. R. Woese and G. E. Fox, ibid., p. 5088; C. R. Woese et al., ibid. 87, 4576 (1990).
3. G. E. Fox et al., Proc. Natl. Acad. Sci. U.S.A. 74, 4537 (1977); C. R. Woese and G. E. Fox, ibid., p. 5088; C. R. Woese et al., ibid. 87, 4576 (1990).
4. N. Iwabe et al, ibid. 86, 9355 (1989); J. P. Gogarten et al., ibid., p. 6651; W. Zillig et al., Endocytobiosis Cell Res. 6, 1 (1989); J. R. Brown and W. F. Doolittle, Proc. Natl. Acad. Sci. U.S.A. 92, 2441 (1995).
5. N. Iwabe et al, ibid. 86, 9355 (1989); J. P. Gogarten et al., ibid., p. 6651; W. Zillig et al., Endocytobiosis Cell Res. 6, 1 (1989); J. R. Brown and W. F. Doolittle, Proc. Natl. Acad. Sci. U.S.A. 92, 2441 (1995).
6. N. Iwabe et al, ibid. 86, 9355 (1989); J. P. Gogarten et al., ibid., p. 6651; W. Zillig et al., Endocytobiosis Cell Res. 6, 1 (1989); J. R. Brown and W. F. Doolittle, Proc. Natl. Acad. Sci. U.S.A. 92, 2441 (1995).
7. N. Iwabe et al, ibid. 86, 9355 (1989); J. P. Gogarten et al., ibid., p. 6651; W. Zillig et al., Endocytobiosis Cell Res. 6, 1 (1989); J. R. Brown and W. F. Doolittle, Proc. Natl. Acad. Sci. U.S.A. 92, 2441 (1995).
8. R. D. Fleischmann et al., Science 269, 496 (1995); C. M. Fraser et al., ibid. 270, 397 (1995).
9. R. D. Fleischmann et al., Science 269, 496 (1995); C. M. Fraser et al., ibid. 270, 397 (1995).
10. N. Williams, ibid. 272, 481 (1996).
11. M. D. Adams et al., Nature 377, 3 (1995); R. Wilson et al., ibid. 368, 32 (1994).
12. M. D. Adams et al., Nature 377, 3 (1995); R. Wilson et al., ibid. 368, 32 (1994).
13. W. Jones et al., Arch. Microbiol. 136, 254 (1983).
14. G. Sutton et al., Genome Sci Tech. 1, 9 (1995).
15. The statistical prediction of M. jannaschii genes was performed with GeneMark [M. Borodovsky and J. McIninch, Comput. Chem. 17, 123 (1993)]. Regular GeneMark uses nonhomogeneous Markov models derived from a training set of coding sequences and ordinary Markov models derived from a training set of noncoding sequences. Only a single 16S ribosomal RNA sequence of M. Jannaschii was available in the public sequence databases before the whole genome sequence described here. Thus, the initial training set to determine parameters of a coding sequence Markov model was chosen as a set of ORFs >1000 nucleotides (nt). As an initial model for noncoding sequences, a zero-order Markov model with genome-specific nucleotide frequencies was used. The initial models were used at the first prediction step. The results of the first prediction were then used to compile a set of putative genes used at the second training step. Alternate rounds of training aria predicting were continued until the set of predicted genes stabilized and the parameters of the final fourth-order model of coding sequences were derived. The regions predicted as noncoding were then used as a training set for a final model for noncoding regions. Cross-validation simulations demonstrated that the GeneMark program trained as described above was able to correctly identify coding regions of at least 96 nt in 94% of the cases and noncoding regions of the same length in 96% of the cases.: These values assume that the self-training method produced correct sequence annotation for complied control sets. Comparison with the results obtained by searches against a nonredundant protein database (3) demonstrated that almost all genes identified by sequence similarity were predicted by the GeneMark program as well. This observation provides additional confidence in genes predicted by GeneMark whose protein translations did not snow significant similarity to known protein sequences. The predicted protein-coding regions were searched against the Blocks database [S. Henikoff and J. G. Henikoff, Genomics 19, 97 (1994)) by means of BLIMPS [J. C. Wallace and S. Henikoff, Comput. Appl. Biosci. 8, 249 (1992)] to verify putative identifications and to identify potential functional motifs in predicted protein-coding regions that had no database match. Genes were assigned to known metabolic pathways. When a gene appeared to be missing from a pathway, the unassigned ORFs and the complete M. jannaschii genome sequence were searched with specific query sequences or motifs from the Blocks database. Hydrophobicity plots were performed on all predicted protein-coding regions by means of the Kyte-Doolittle algorithm [J. Kyte and R. F. Doolittte, J. Mol. Biol. 157, 105 (1982)] to identify potentially functionally relevant signatures in these sequences. The results of the Blocks and Kyte-Doolittle analyses are available on the World Wide Web (http:/ /www.tigr.org/tdo/mdb/mjdb/mjdb.html).
16. The statistical prediction of M. jannaschii genes was performed with GeneMark [M. Borodovsky and J. McIninch, Comput. Chem. 17, 123 (1993)]. Regular GeneMark uses nonhomogeneous Markov models derived from a training set of coding sequences and ordinary Markov models derived from a training set of noncoding sequences. Only a single 16S ribosomal RNA sequence of M. Jannaschii was available in the public sequence databases before the whole genome sequence described here. Thus, the initial training set to determine parameters of a coding sequence Markov model was chosen as a set of ORFs >1000 nucleotides (nt). As an initial model for noncoding sequences, a zero-order Markov model with genome-specific nucleotide frequencies was used. The initial models were used at the first prediction step. The results of the first prediction were then used to compile a set of putative genes used at the second training step. Alternate rounds of training aria predicting were continued until the set of predicted genes stabilized and the parameters of the final fourth-order model of coding sequences were derived. The regions predicted as noncoding were then used as a training set for a final model for noncoding regions. Cross-validation simulations demonstrated that the GeneMark program trained as described above was able to correctly identify coding regions of at least 96 nt in 94% of the cases and noncoding regions of the same length in 96% of the cases.: These values assume that the self-training method produced correct sequence annotation for complied control sets. Comparison with the results obtained by searches against a nonredundant protein database (3) demonstrated that almost all genes identified by sequence similarity were predicted by the GeneMark program as well. This observation provides additional confidence in genes predicted by GeneMark whose protein translations did not snow significant similarity to known protein sequences. The predicted protein-coding regions were searched against the Blocks database [S. Henikoff and J. G. Henikoff, Genomics 19, 97 (1994)) by means of BLIMPS [J. C. Wallace and S. Henikoff, Comput. Appl. Biosci. 8, 249 (1992)] to verify putative identifications and to identify potential functional motifs in predicted protein-coding regions that had no database match. Genes were assigned to known metabolic pathways. When a gene appeared to be missing from a pathway, the unassigned ORFs and the complete M. jannaschii genome sequence were searched with specific query sequences or motifs from the Blocks database. Hydrophobicity plots were performed on all predicted protein-coding regions by means of the Kyte-Doolittle algorithm [J. Kyte and R. F. Doolittte, J. Mol. Biol. 157, 105 (1982)] to identify potentially functionally relevant signatures in these sequences. The results of the Blocks and Kyte-Doolittle analyses are available on the World Wide Web (http:/ /www.tigr.org/tdo/mdb/mjdb/mjdb.html).
17. The statistical prediction of M. jannaschii genes was performed with GeneMark [M. Borodovsky and J. McIninch, Comput. Chem. 17, 123 (1993)]. Regular GeneMark uses nonhomogeneous Markov models derived from a training set of coding sequences and ordinary Markov models derived from a training set of noncoding sequences. Only a single 16S ribosomal RNA sequence of M. Jannaschii was available in the public sequence databases before the whole genome sequence described here. Thus, the initial training set to determine parameters of a coding sequence Markov model was chosen as a set of ORFs >1000 nucleotides (nt). As an initial model for noncoding sequences, a zero-order Markov model with genome-specific nucleotide frequencies was used. The initial models were used at the first prediction step. The results of the first prediction were then used to compile a set of putative genes used at the second training step. Alternate rounds of training aria predicting were continued until the set of predicted genes stabilized and the parameters of the final fourth-order model of coding sequences were derived. The regions predicted as noncoding were then used as a training set for a final model for noncoding regions. Cross-validation simulations demonstrated that the GeneMark program trained as described above was able to correctly identify coding regions of at least 96 nt in 94% of the cases and noncoding regions of the same length in 96% of the cases.: These values assume that the self-training method produced correct sequence annotation for complied control sets. Comparison with the results obtained by searches against a nonredundant protein database (3) demonstrated that almost all genes identified by sequence similarity were predicted by the GeneMark program as well. This observation provides additional confidence in genes predicted by GeneMark whose protein translations did not snow significant similarity to known protein sequences. The predicted protein-coding regions were searched against the Blocks database [S. Henikoff and J. G. Henikoff, Genomics 19, 97 (1994)) by means of BLIMPS [J. C. Wallace and S. Henikoff, Comput. Appl. Biosci. 8, 249 (1992)] to verify putative identifications and to identify potential functional motifs in predicted protein-coding regions that had no database match. Genes were assigned to known metabolic pathways. When a gene appeared to be missing from a pathway, the unassigned ORFs and the complete M. jannaschii genome sequence were searched with specific query sequences or motifs from the Blocks database. Hydrophobicity plots were performed on all predicted protein-coding regions by means of the Kyte-Doolittle algorithm [J. Kyte and R. F. Doolittte, J. Mol. Biol. 157, 105 (1982)] to identify potentially functionally relevant signatures in these sequences. The results of the Blocks and Kyte-Doolittle analyses are available on the World Wide Web (http:/ /www.tigr.org/tdo/mdb/mjdb/mjdb.html).
18. The statistical prediction of M. jannaschii genes was performed with GeneMark [M. Borodovsky and J. McIninch, Comput. Chem. 17, 123 (1993)]. Regular GeneMark uses nonhomogeneous Markov models derived from a training set of coding sequences and ordinary Markov models derived from a training set of noncoding sequences. Only a single 16S ribosomal RNA sequence of M. Jannaschii was available in the public sequence databases before the whole genome sequence described here. Thus, the initial training set to determine parameters of a coding sequence Markov model was chosen as a set of ORFs >1000 nucleotides (nt). As an initial model for noncoding sequences, a zero-order Markov model with genome-specific nucleotide frequencies was used. The initial models were used at the first prediction step. The results of the first prediction were then used to compile a set of putative genes used at the second training step. Alternate rounds of training aria predicting were continued until the set of predicted genes stabilized and the parameters of the final fourth-order model of coding sequences were derived. The regions predicted as noncoding were then used as a training set for a final model for noncoding regions. Cross-validation simulations demonstrated that the GeneMark program trained as described above was able to correctly identify coding regions of at least 96 nt in 94% of the cases and noncoding regions of the same length in 96% of the cases.: These values assume that the self-training method produced correct sequence annotation for complied control sets. Comparison with the results obtained by searches against a nonredundant protein database (3) demonstrated that almost all genes identified by sequence similarity were predicted by the GeneMark program as well. This observation provides additional confidence in genes predicted by GeneMark whose protein translations did not snow significant similarity to known protein sequences. The predicted protein-coding regions were searched against the Blocks database [S. Henikoff and J. G. Henikoff, Genomics 19, 97 (1994)) by means of BLIMPS [J. C. Wallace and S. Henikoff, Comput. Appl. Biosci. 8, 249 (1992)] to verify putative identifications and to identify potential functional motifs in predicted protein-coding regions that had no database match. Genes were assigned to known metabolic pathways. When a gene appeared to be missing from a pathway, the unassigned ORFs and the complete M. jannaschii genome sequence were searched with specific query sequences or motifs from the Blocks database. Hydrophobicity plots were performed on all predicted protein-coding regions by means of the Kyte-Doolittle algorithm [J. Kyte and R. F. Doolittte, J. Mol. Biol. 157, 105 (1982)] to identify potentially functionally relevant signatures in these sequences. The results of the Blocks and Kyte-Doolittle analyses are available on the World Wide Web (http:/ /www.tigr.org/tdo/mdb/mjdb/mjdb.html).
19. H. Zhao et al., Arch. Microbiol. 150, 178 (1988).
20. A. A. DiMarco et al., Annu. Rev. Biochem. 59, 355 (1990).
21. N. Belay et al., Nature 312, 286 (1984).
22. H. G. Wood et al., Trends Biochem. Sci. 11, 14 (1986).
23. M. Blaat, Antonie Leewenhoek 66, 187 (1994).
24. E. Hartmann and H. König, Arch. Microbiol. 151, 274 (1989).
25. X. M. Jiang et al., Mol. Microbiol. 5, 695 (1991).
26. K. Lechner et al., J. Mol. Evol. 29, 20 (1989); A. K. E. Köpke and B. Wittmann-Liebold. Can. J. Microbiol. 35,11 (1989).
27. K. Lechner et al., J. Mol. Evol. 29, 20 (1989); A. K. E. Köpke and B. Wittmann-Liebold. Can. J. Microbiol. 35,11 (1989).
28. P. Keeling et al., Syst. Appl. Microbiol., in press.
29. M. Wilcox, Eur. J. Biochem. 11, 405 (1969); N. C. Martin et al., J. Mol. Biol. 101, 285 (1976); N. C. Martin et al., Biochemistry 16, 4672 (1977); A. Schon et al., Biochimie 70, 391 (1988); D. Soll and U. Raj Bhandary, Eds. tRNA: Structure, Biosynthesis, and Function (American Society for Microbiology, Washington, DC, 1995).
30. M. Wilcox, Eur. J. Biochem. 11, 405 (1969); N. C. Martin et al., J. Mol. Biol. 101, 285 (1976); N. C. Martin et al., Biochemistry 16, 4672 (1977); A. Schon et al., Biochimie 70, 391 (1988); D. Soll and U. Raj Bhandary, Eds. tRNA: Structure, Biosynthesis, and Function (American Society for Microbiology, Washington, DC, 1995).
31. M. Wilcox, Eur. J. Biochem. 11, 405 (1969); N. C. Martin et al., J. Mol. Biol. 101, 285 (1976); N. C. Martin et al., Biochemistry 16, 4672 (1977); A. Schon et al., Biochimie 70, 391 (1988); D. Soll and U. Raj Bhandary, Eds. tRNA: Structure, Biosynthesis, and Function (American Society for Microbiology, Washington, DC, 1995).
32. M. Wilcox, Eur. J. Biochem. 11, 405 (1969); N. C. Martin et al., J. Mol. Biol. 101, 285 (1976); N. C. Martin et al., Biochemistry 16, 4672 (1977); A. Schon et al., Biochimie 70, 391 (1988); D. Soll and U. Raj Bhandary, Eds. tRNA: Structure, Biosynthesis, and Function (American Society for Microbiology, Washington, DC, 1995).
33. M. Wilcox, Eur. J. Biochem. 11, 405 (1969); N. C. Martin et al., J. Mol. Biol. 101, 285 (1976); N. C. Martin et al., Biochemistry 16, 4672 (1977); A. Schon et al., Biochimie 70, 391 (1988); D. Soll and U. Raj Bhandary, Eds. tRNA: Structure, Biosynthesis, and Function (American Society for Microbiology, Washington, DC, 1995).
34. R. de Pouplana et al., Proc. Natl. Acad. Sci. U.S.A. 93, 166 (1996).
35. E.A. Wagner et al., J. Bacteriol 177, 5179 (1995); D. T. Logan et al., EMBO J. 14, 4156 (1995).
36. E.A. Wagner et al., J. Bacteriol 177, 5179 (1995); D. T. Logan et al., EMBO J. 14, 4156 (1995).
37. C. R. Woese and R. S. Wolfe, Eds. The Bacteria (Academic Press, New York, 1985), vol. 8; D. Langer et al., Proc. Natl. Acad Sci. U.S.A. 92, 5768 (1995); M. Lanzendoerfer et al., Syst. Appl. Microbiol. 16, 656 (1994).
38. C. R. Woese and R. S. Wolfe, Eds. The Bacteria (Academic Press, New York, 1985), vol. 8; D. Langer et al., Proc. Natl. Acad Sci. U.S.A. 92, 5768 (1995); M. Lanzendoerfer et al., Syst. Appl. Microbiol. 16, 656 (1994).
39. C. R. Woese and R. S. Wolfe, Eds. The Bacteria (Academic Press, New York, 1985), vol. 8; D. Langer et al., Proc. Natl. Acad Sci. U.S.A. 92, 5768 (1995); M. Lanzendoerfer et al., Syst. Appl. Microbiol. 16, 656 (1994).
40. H.-P. Klenk and W. F, Doolittle, Curr. Biol. 4, 920 (1994).
41. A. Bernard et al., EMBO J. 6, 4219 (1987); G. Cullman et al., Mol. Cell, Biol. 15, 4661 (1995); T. Uemori et al, J. Bacteriol. 177, 2164 (1995); M. Delarue et al., Protein Eng. 3, 461 (1990); K. A. Gavin, M. Hidaka, B. Stillman, Science 270, 1667 (1995).
42. A. Bernard et al., EMBO J. 6, 4219 (1987); G. Cullman et al., Mol. Cell, Biol. 15, 4661 (1995); T. Uemori et al, J. Bacteriol. 177, 2164 (1995); M. Delarue et al., Protein Eng. 3, 461 (1990); K. A. Gavin, M. Hidaka, B. Stillman, Science 270, 1667 (1995).
43. A. Bernard et al., EMBO J. 6, 4219 (1987); G. Cullman et al., Mol. Cell, Biol. 15, 4661 (1995); T. Uemori et al, J. Bacteriol. 177, 2164 (1995); M. Delarue et al., Protein Eng. 3, 461 (1990); K. A. Gavin, M. Hidaka, B. Stillman, Science 270, 1667 (1995).
44. A. Bernard et al., EMBO J. 6, 4219 (1987); G. Cullman et al., Mol. Cell, Biol. 15, 4661 (1995); T. Uemori et al, J. Bacteriol. 177, 2164 (1995); M. Delarue et al., Protein Eng. 3, 461 (1990); K. A. Gavin, M. Hidaka, B. Stillman, Science 270, 1667 (1995).
45. A. Bernard et al., EMBO J. 6, 4219 (1987); G. Cullman et al., Mol. Cell, Biol. 15, 4661 (1995); T. Uemori et al, J. Bacteriol. 177, 2164 (1995); M. Delarue et al., Protein Eng. 3, 461 (1990); K. A. Gavin, M. Hidaka, B. Stillman, Science 270, 1667 (1995).
46. L. A. Whitbred and S. Dalton, Gene 155, 113 (1995).
47. C. G. Eberhart and S. A. Wasserman, Development 121, 3477 (1995).
48. L. Rothfield and C.-R Zhao, Cell 84, 183 (1996); J. Lutkenhaus, Curr. Opin. Genet. Dev. 3, 783 (1993).
49. L. Rothfield and C.-R Zhao, Cell 84, 183 (1996); J. Lutkenhaus, Curr. Opin. Genet. Dev. 3, 783 (1993).
50. B.P. Kaine and V.L. Merkel, J. Bacteriol. 171, 4261 (1989); M. A. Porltz et al, Cell 55, 4 (1988).
51. B.P. Kaine and V.L. Merkel, J. Bacteriol. 171, 4261 (1989); M. A. Porltz et al, Cell 55, 4 (1988).
52. D. M. Faguy et al., Can. J. Microbiol. 40, 67 (1994): M. L. Kalmokoff et al., Arch. Microbiol. 157, 481 (1992).
53. D. M. Faguy et al., Can. J. Microbiol. 40, 67 (1994): M. L. Kalmokoff et al., Arch. Microbiol. 157, 481 (1992).
54. K. Sandman et al, Proc. Natl. Acad. Sci. U.S.A. 87, 5788 (1990).
55. 2 terminal extein-intein junction, so it was necessary to rely on consensus sequences to select the putative site. The inteins in MJ1042 and MJ0542 have previously uncharacterized COOH-terminal splice junctions, GNC and FNC, respectively.
56. 2 terminal extein-intein junction, so it was necessary to rely on consensus sequences to select the putative site. The inteins in MJ1042 and MJ0542 have previously uncharacterized COOH-terminal splice junctions, GNC and FNC, respectively.
57. 2 terminal extein-intein junction, so it was necessary to rely on consensus sequences to select the putative site. The inteins in MJ1042 and MJ0542 have previously uncharacterized COOH-terminal splice junctions, GNC and FNC, respectively.
58. 2 terminal extein-intein junction, so it was necessary to rely on consensus sequences to select the putative site. The inteins in MJ1042 and MJ0542 have previously uncharacterized COOH-terminal splice junctions, GNC and FNC, respectively.
59. 2 terminal extein-intein junction, so it was necessary to rely on consensus sequences to select the putative site. The inteins in MJ1042 and MJ0542 have previously uncharacterized COOH-terminal splice junctions, GNC and FNC, respectively.
60. 2 terminal extein-intein junction, so it was necessary to rely on consensus sequences to select the putative site. The inteins in MJ1042 and MJ0542 have previously uncharacterized COOH-terminal splice junctions, GNC and FNC, respectively.
61. 2 terminal extein-intein junction, so it was necessary to rely on consensus sequences to select the putative site. The inteins in MJ1042 and MJ0542 have previously uncharacterized COOH-terminal splice junctions, GNC and FNC, respectively.
62. 2 terminal extein-intein junction, so it was necessary to rely on consensus sequences to select the putative site. The inteins in MJ1042 and MJ0542 have previously uncharacterized COOH-terminal splice junctions, GNC and FNC, respectively.
63. P. T. Hamilton et al., Mol. Gen. Genet. 200, 47 (1985).
64. F. J. M. Mojica et al., Mol. Microbiol. 17, 85 (1995).
65. G. Felsenfeld et al., J. Am. Chem. Soc. 79, 2023 (1957); A. G. Letai et al., Biochemistry 27, 9108 (1988).
66. G. Felsenfeld et al., J. Am. Chem. Soc. 79, 2023 (1957); A. G. Letai et al., Biochemistry 27, 9108 (1988).
67. M. Riley, Microbiol. Rev. 57, 862 (1993).
68. Supported in part by Department of Energy Cooperative Agreements DE-FC02-95ER61962 (J.C.V.) and DEFC02-95ER61963 (C.R.W. and G.J.O), NASA grant NAGW 2554 (C.R.W,), and a core grant to TIGR from Human Genome Sciences. G.J.O. is the recipient of the National Science Foundation Presidential Young Investigator Award (DIR 89-57026). M.B. is supported by National Institutes of Health grant GM00783. We thank M. Heaney, C. Gnehm, R. Shirley, J. Slagel, and W. Hayes for software and database support; T. Dixon and V. Sapiro for computer system support; K. Hong and B. Stader for laboratory assistance; and B. Mukhopadhyay for helpful discussions. The M. jannaschii source accession number is DSM 2661. and the cells were a gift from P. Haney (Department of Microbiology, University of Illinois).