To test the hypothesis that the proteins of thermophilic prokaryotes are subject to unusually stringent functional constraints, we estimated the numbers of synonymous and nonsynonymous nucleotide substitutions per site between 17,957 pairs of orthologous genes from 22 pairs of closely related species of Archaea and Bacteria. The average ratio of nonsynonymous to synonymous substitutions was significantly lower in thermophiles than in nonthermophiles, and this effect was observed in both Archaea and Bacteria. There was no evidence that this difference could be explained by factors such as nucleotide content bias. Rather, the results support the hypothesis that proteins of thermophiles are subject to unusually strong purifying selection, leading to a reduced overall level of amino acid evolution per mutational event. The results show that genome-wide patterns of sequence evolution can be influenced by natural selection exerted by a species' environment and shed light on a previous observation that relatively few of the mutations arising in a thermophilic archaeon were nucleotide substitutions in contrast to indels.

---

HIGHLY thermophilic species, inhabiting environments where temperatures can exceed 100°, have been described for both domains of prokaryotic life (WOESE et al. 1990), the Bacteria (or eubacteria) and the Archaea (or archaebacteria). Considerable interest has focused on understanding the mechanisms that make life possible under these extreme conditions (BROWN and LUPAS 1998). Both analyses of protein structures and genome-wide comparisons of amino acid composition have revealed widely shared characteristics of proteins encoded by the genomes of thermophiles; these include the presence of surface ion networks, location of hydrophobic residues in partly surface-exposed positions, avoidance of thermally unstable amino acid residues, and overall compact protein structure (PERUTZ 1978; BROWN and LUPAS 1998; GROMIHA et al. 1999; CAMBILLAU and CLAVERIE 2000; KUMAR et al. 2000; KUMAR and NUSSINOV 2001; CHEN et al. 2003; CRISWELL et al. 2003). In addition, there is evidence that the proteins of thermophiles are characterized by a distinct pattern of amino acid composition (KREIL and OUZOUNIS 2001; TEKAIA et al. 2002). The evident survival value conferred by adaptive protein features is expected to result in strong purifying (conservative) natural selection on protein-coding genes of thermophiles.

We tested this prediction by estimating the numbers of synonymous and nonsynonymous (amino acid-altering) nucleotide substitutions by comparing 17,957 pairs of orthologous gene pairs from closely related species of Bacteria and Archaea. We compared 16 pairs of closely related Eubacteria (usually congeners) whose genomes have been completely sequenced, completely sequenced genomes from three genera of Archaea (including the thermophiles Pyrococcus and Sulfolobus), and sequences from partially sequenced genomes of three thermophilic genera of Bacteria (Thermotoga, Thermoanaerobacter, and Thermus).

We estimated the number of synonymous nucleotide substitutions per synonymous site (d_S) and the number of nonsynonymous nucleotide substitutions per nonsynonymous site (d_N) between orthologous gene pairs. Because synonymous mutations are less likely than nonsynonymous mutations to be selectively deleterious, d_S between two related sequences is expected to reflect both the mutation rate and the time since the sequences' last common ancestor (NEI 1987). By contrast, d_N will reflect mainly the effects of purifying selection, which acts to eliminate deleterious mutations. Thus, the d_N/d_S ratio is a measure of the strength of purifying selection at the amino acid level, a lower ratio implying stronger selective constraint.

Sequences analyzed:

We compared orthologous gene pairs identified in 16 complete genomes of Bacteria and three complete genomes of Archaea (Table 1). In addition, available sequences were compared within three genera of thermophilic Bacteria for which complete genomes were not available (Table 1). We identified orthologous gene pairs (genes homologous by descent from a common ancestral gene without gene duplication) between each of these pairs of species. We applied the BLASTCLUST computer program, available from the collection of BLAST tools (ALTSCHUL et al. 1997), to cluster protein translations to define protein families (single-linkage method). Homology search between sequences was performed using an E value of 10^–50 and a minimum of 20% identity across at least 30% of the sequence lengths. These strict search criteria were used to increase our chance of identifying orthologous gene pairs rather than paralogues. Only families with one member from each species in the pair were used in analyses, to avoid the problem of paralogous comparisons. Sequences were aligned at the amino acid level using the CLUSTALW program (THOMPSON et al. 1994) and the alignment was imposed on the coding sequences.

View this table: In this window In a new window   TABLE 1 Sequences used in analyses

 

Evolutionary analyses:

The maximum-likelihood (ML) method (YANG and NIELSEN 2000) implemented in the PAML program (YANG 1997) was used to estimate the number of synonymous nucleotide substitutions per synonymous site (d_S) and the number of nonsynonymous nucleotide differences per nonsynonymous site (d_N). We used the F3X4 model, which incorporates the nucleotide composition and transition/transversion ratios estimated from the compared sequences and assumes equal likelihood of each possible codon pathway. We also estimated d_S and d_N by the modified Nei-Gojobori method (ZHANG et al. 1998); because the results of both methods were similar, only the ML results are presented here. We excluded from analyses those cases in which there were no synonymous differences or where d_S was undefined; of 19,606 comparisons between putative orthologs, 1649 (8.4%) were excluded on these grounds.

We estimated d_S and d_N between putative orthologs in pairs of closely related species (Table 1). Note that because these comparisons were between orthologous genes from pairs of closely related species, each nucleotide difference between pair members must have arisen since the most recent common ancestor of the two species. Thus, each comparison is phylogenetically and statistically independent of each other comparison (FELSENSTEIN 1985). In statistical analyses, we assumed that each orthologous gene pair evolves independently of other genes in the genome. However, because we used the orthologous gene pair as the unit of statistical analysis, our analyses did not require the assumption that each nucleotide site evolves independently, a biologically unrealistic assumption that is routinely made in studies of molecular evolution.

We also used phylogenetically independent comparisons in the analysis of nucleotide content. Nucleotide content (both G + C content and A + G content) was measured at third codon positions in each of the genes compared, and then nucleotide content was averaged between the two members of each orthologous pair. Thus values of G + C at the third position (GC3) and of A + G at the third position (AG3) were obtained for each orthologous gene pair.

Factorial analysis of variance in mean d_N/d_S ratio showed no significant difference between the two prokaryotic domains (Bacteria and Archaea) but a significant difference between thermophiles and nonthermophiles (F_1,17956 = 19.09; P < 0.001; Figure 1a). There was no significant interaction between domain and thermophily, indicating a similar difference between thermophiles and nonthermophiles in the two prokaryotic domains.

View larger version (16K): In this window In a new window Download PPT slide   FIGURE 1.— (a) Mean dN/dS ratio in comparisons of orthologous genes between closely related species pairs of thermophilic (T) and nonthermophilic (NT) Bacteria and Archaea. Numbers of comparisons in each category are shown. (b) Mean dN/dS in comparisons of orthologous genes excluding cases with dS or dN 1.5.

 
Although the ML method (YANG and NIELSEN 2000) corrects for nucleotide content bias, it is well known that under simple models, nucleotide content bias at third-codon positions can affect the observed numbers of synonymous substitutions (WOLFE et al. 1989). If nucleotide-content bias had a similar effect in our analyses, the higher d_N/d_S ratio we observed in nonthermophiles might be simply due to a reduction in observed d_S resulting from compositional bias in nonthermophile species. To test whether such a phenomenon could be responsible for our results, we examined the percentage of G + C at third-codon positions (GC3) in the genes in our data set (Figure 2a). Factorial analysis of variance showed a significant difference in mean GC3 between the two prokaryotic domains (F_1,17956 = 142.27; P < 0.001) and a significant difference between thermophiles and nonthermophiles (F_1,17956 = 5.76; P = 0.016). There was also a significant interaction between domain and thermophily (F_1,17956 = 40.28; P < 0.001).

View larger version (15K): In this window In a new window Download PPT slide   FIGURE 2.— (a) Mean percentage of G + C at third codon positions (GC3) in genes used in analyses. (b) Mean percentage of A + G at third codon positions (AG3) in genes used in analyses. T, thermophilic species; NT, nonthermophilic species.

 
These results are explained by the observation that mean GC3 was higher in thermophiles than in nonthermophiles in the case of Bacteria, but lower in thermophiles than in nonthermophiles in the case of Archaea (Figure 2a). In both Bacteria and Archaea, GC3 was closer on average to 50% in nonthermophiles than in thermophiles (Figure 2a). In the entire data set, there was a modest but significant negative correlation between d_S and GC3 (r = –0.101; P < 0.001). Thus, on the basis of G + C content alone, the d_N/d_S ratio would be expected to be highest in thermophilic Bacteria, intermediate in nonthermophiles of both domains, and lowest in thermophilic Archaea. The fact that this pattern was not observed (Figure 1) supports the hypothesis that G + C content bias was not a major factor in yielding the observed pattern in d_N/d_S ratios. When GC3 was included as a covariate in the analysis of variance in d_N/d_S, there was a significant effect of the covariate (F_1,17956 = 5.10; P = 0.024), but the significant effect of thermophily remained (F_1,17956 = 19.44; P < 0.001). This result indicates that the association between thermophily and a reduced d_N/d_S was statistically independent of the linear relationship between d_N/d_S and GC3.

In the case of thermophiles, some data suggest a preference for purines in mRNAs (LAO and FORSDYKE 2000; LAMBROS et al. 2003). On the hypothesis that a bias in nucleotide content is most likely to be expressed at third-codon positions, we examined the percentage of A + G at third-codon positions (AG3) in the genes in our data set (Figure 2b). Factorial analysis of variance showed a significant difference in mean AG3 between the two prokaryotic domains (F_1,17956 = 60.53; P < 0.001) and a significant difference between thermophiles and nonthermophiles (F_1,17956 = 250.718; P < 0.001). There was also a significant interaction between domain and thermophily (F_1,17956 = 13.99; P < 0.001). Consistent with previous results (LAO and FORSDYKE 2000; LAMBROS et al. 2003), in both Bacteria and Archaea, mean AG3 was higher in thermophiles than in nonthermophiles (Figure 2b). This difference was more pronounced in Archaea than in Bacteria (Figure 2b).

In contrast to the case of GC3, there was a modest but significant positive correlation between d_S and AG3 (r = 0.090; P < 0.001). As with GC3, when AG3 was included as a covariate in the analysis of variance in d_N/d_S, there were significant effects of both the covariate (F_1,17956 = 25.21; P < 0.001) and thermophily (F_1,17956 = 24.31; P < 0.001). Thus, the association between thermophily and a reduced d_N/d_S was statistically independent of the linear relationship between d_N/d_S and AG3.

In certain comparisons, the ML method estimated that multiple substitutions had occurred per site, especially in the case of synonymous sites. Although some authors have argued that such estimates are reliable (BLANC et al. 2003), the accuracy of estimation may be problematic because the sites have been saturated with changes. To test whether our observed results were affected by saturation, we reanalyzed the data after excluding all comparisons for which d_S or d_N was estimated to be 1.5. The results (Figure 1b) were similar to those observed with the complete data set. In a factorial analysis of variance the only significant effect was that of thermophily (F_1,9966 = 7.49; P = 0.006). When GC3 was included in the model as a covariate, there was no significant effect of the covariate, but there was a significant effect of thermophily (F_1,9965 = 7.41; P = 0.007). Similarly, when AG3 was included as a covariate, there was no significant effect of the covariate, but there was a significant effect of thermophily (F_1,9965 = 7.61; P = 0.006).

In the complete data set, the correlation coefficient between d_N and d_S in thermophiles (r = 0.485; P < 0.001) was similar to that in nonthermophiles (r = 0.454; P < 0.001). However, analysis of covariance indicated that there was a significant difference between thermophiles and nonthermophiles with respect to the slope of the linear relationship between d_N and d_S (Figure 3). We fitted linear regression lines through the origin on the assumption that d_N and d_S are both initially zero at the moment of lineage divergence for thermophiles and nonthermophiles. The slope of the former line was 0.077, while that of the latter was 0.129. These results suggest that d_N increases for a given increase in d_S 1.67 times faster in nonthermophiles than in thermophiles.

View larger version (22K): In this window In a new window Download PPT slide   FIGURE 3.— Plot of dN vs. dS for comparisons of orthologous gene pairs from thermophilic (red dots) and nonthermophilic (green dots) species of Bacteria and Archaea. Analysis of covariance in dN with dS as covariate showed a significant effect of the covariate (F1,17956 = 1072.16; P < 0.001), a significant difference between thermophiles and nonthermophiles (F1,17956 = 7.15; P = 0.007), and a significant interaction between thermophily and the covariate (F1,17956 = 133.60; P < 0.001).

 
Our results are consistent with the hypothesis that the proteins of thermophiles are subject to unusually strong functional constraints in amino acid sequence, which is reflected in a reduced level of nonsynonymous nucleotide substitution for a given level of synonymous substitution. As a further test of the hypothesis that this constraint arises from the high-temperature environment, we made pairwise comparisons of shared orthologs between the hyperthermophilic archaeal genus Pyrococcus and both the mesophilic genus Methanosarcina and the moderately thermophilic genus Sulfolobus (MADIGAN et al. 2003).

Examining pairs of orthologs present in both genera, we found that d_N/d_S ratios for the same genes were significantly higher in Methanosarcina than in Pyrococcus (Figure 4a). The mean for Methanosarcina (0.074 ± 0.003 SEM) differed significantly from that for Pyrococcus (0.044 ± 0.002; paired t = 12.29; two-tailed P < 0.001). Similarly, d_N/d_S ratios were significantly higher in Sulfolobus than in Pyrococcus (Figure 4b). The mean for Methanosarcina (0.0584 ± 0.002) differed significantly from that for Pyrococcus (0.037 ± 0.002; paired t = 7.94; two-tailed P < 0.001). This result provided evidence of an unusually reduced rate of nonsynonymous substitution in a hyperthermophile. This, in turn, provides additional support for the hypothesis that life at high temperatures imposes unusual functional constraints on the primary structure of proteins.

View larger version (22K): In this window In a new window Download PPT slide   FIGURE 4.— Plots of dN/dS for orthologous gene pairs shared by Archaeal genera. In each case the line has a slope of 1. (a) dN/dS for Methanosarcina acetivorans vs. M. mazei plotted against dN/dS for the orthologous genes in Pyrococcus abyssi vs. P. horokoshii (N = 192). (b) dN/dS for Sulfolobus solfataricus vs. S. tokodaii plotted against dN/dS for the orthologous genes in P. abyssi vs. P. horokoshii (N = 152).

 

An analysis of mutation in a thermophilic archaeon (GROGAN et al. 2001) revealed a mutation rate close to or slightly less than that characteristic of all DNA-based microbes examined to date. However, the fraction of mutations that were base substitutions was half or less of that observed in eubacterial and eukaryotic microbes and in mammals. This observation generated the hypothesis that the average amino acid substitution is more deleterious in a thermophile than in a nonthermophile. On this hypothesis, the especially deleterious nature of mutation in thermophiles would have favored the evolutionary fixation of modifiers that specifically decrease the rates of base-substitution mutagenesis. Strong selection against amino acid replacements in thermophiles would impact their molecular evolution in two ways. First, because most selectively neutral mutations that reach fixation are base substitutions and not indels, the overall rate of molecular evolution would be reduced. Second, within this generally reduced rate of molecular evolution, the fixation of nonsynonymous substitutions would be reduced compared to that of synonymous substitutions. This study provides support for the latter prediction.

It has been suggested that Archaea evolve more slowly than either Bacteria or Eukaryotes. A recent test of this hypothesis concluded that amino acid sequences of Bacteria and Eukaryotes evolve at indistinguishable rates but that those of Archaea evolve 10–40% more slowly (KOLLMAN and DOOLITTLE 2000). However, in that analysis, 10% of the Bacteria but 78% of the Archaea were thermophiles. By contrast, our results imply that natural selection eliminates a higher proportion of nonsynonymous mutations in thermophiles than in nonthermophiles, resulting in a lower average rate of amino acid replacement per mutational event in thermophiles, whereas nonthermophilic Archaea and nonthermophilic Bacteria are indistinguishable with respect to d_N/d_S. Likewise, thermophilic Archaea and thermophilic Bacteria are indistinguishable with respect to d_N/d_S. Thus, our results imply both that strong purifying selection against amino acid changes in thermophiles operates in a similar fashion across both prokaryotic domains and that rates of evolution are likely to be indistinguishable in Bacteria and Archaea but to depend critically on optimal growth temperature.

Examination of protein-coding sequences from a wide variety of organisms has provided evidence of the near ubiquity of "purifying" or conservative natural selection, which acts to eliminate deleterious mutations (NEI 1987). Because of their effect on protein structure, nonsynonymous mutations are more likely than synonymous mutations to be deleterious. As a result, over time the rate of synonymous nucleotide substitution per synonymous site (d_S) is predicted to exceed the rate of nonsynonymous nucleotide substitution per nonsynonymous site (d_N), and this prediction is supported in the case of most genes (KIMURA 1977; NEI 1987). There is evidence that some, if not most, prokaryote codon usage is subject to selective constraint (SHARP and LI 1986; BULMER 1991; LYNN et al. 2002). Where selection favors certain synonymous codons over others, purifying selection may occur at synonymous as well as nonsynonymous sites. However, that overall d_N/d_S ratios in prokaryotes are <<1.0 (Figure 1) indicates that purifying selection at synonymous sites is of negligible magnitude in comparison to that at nonsynonymous sites.

In a survey of 40 prokaryotic genomes, LYNN et al. (2002) found that the factor explaining the largest proportion (25%) of the variance in codon usage was G + C content in the genome. The second most important factor, related to optimal growth temperature, accounted for an additional 10% of the variance (LYNN et al. 2002). These authors found that the major difference between thermophiles and nonthermophiles related to their use of arginine and isoleucine codons. Other studies have suggested that thermophiles have a preference for purines in mRNAs, which is also expected to affect codon usage (LAO and FORSDYKE 2000; LAMBROS et al. 2003). If thermophiles are subject to greater constraint on codon usage than nonthermophiles, this would be expected to reduce d_S and thus to increase the d_N/d_S ratio in thermophiles. That we observed overall lower mean d_N/d_S in thermophiles than in nonthermophiles suggests that, even if greater constraint on codon usage is present in thermophiles than in nonthermophiles, such constraint is not of sufficient magnitude to mask the strong difference with respect to constraint on amino acid sequences.

The factors that influence variations among organisms with respect to patterns of synonymous and nonsynonymous nucleotide substitution remain poorly understood. Our results suggest that one such factor can be genome-wide differences among organisms with respect to the strength of purifying selection. Furthermore, our results demonstrate that a species' environment represents a source of selection that can leave a clear imprint on the pattern of long-term genomic evolution.

We are grateful for comments on the manuscript by D. W. Grogan. This research was supported by grant GM66710 to A.L.H. from the National Institutes of Health.

ALTSCHUL, S. F., T. L. MADDEN, A. A. SCHAFFER, J. ZHANG, Z. ZHANG et al., 1997 Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25: 3389–3402.[Abstract/Free Full Text]

BLANC, G., K. HOKAMP and K. H. WOLFE, 2003 A recent polyploidy superimposed on older large-scale duplications in the Arabidopsis genome. Genome Res. 13: 137–144.[Abstract/Free Full Text]

BROWN, J. R., and A. N. LUPAS, 1998 What makes a thermophile? Trends Microbiol. 6: 349–351.[CrossRef][Medline]

BULMER, M., 1991 The selection-mutation-drift theory of synonymous codon usage. Genetics 129: 897–907.[Abstract]

CAMBILLAU, C., and J. M. CLAVERIE, 2000 Structural and genomic correlates of hyperthermostability. J. Biol. Chem. 275: 32383–32386.[Abstract/Free Full Text]

CHEN, Y. W., M. BYCROFT and K.-B. WONG, 2003 Crystal structure of ribosomal protein L30e from the extreme thermophile Thermococcus celer: thermal stability and RNA binding. Biochemistry 42: 2857–2865.[CrossRef][Medline]

CRISWELL, A. R., E. BAE, B. STEC, J. KONISKY and G. N. PHILLIPS, JR., 2003 Structures of thermophilic and mesophilic adenylate kinases from the genus Methanococcus. J. Mol. Biol. 333: 1087–1099.

FELSENSTEIN, J., 1985 Phylogenies and the comparative method. Am. Nat. 125: 1–15.

GROGAN, D. W., G. T. CARVER and J. W. DRAKE, 2001 Genetic fidelity under harsh conditions: analysis of spontaneous mutation in the thermoacidophilic archaeon Sulfolobus acidocaldarius. Proc. Natl. Acad. Sci. USA 98: 7928–7933.[Abstract/Free Full Text]

GROMIHA, M. M., M. OOBATAKE and A. SARAI, 1999 Important amino acid properties for enhanced thermostability from mesophilic to thermophilic proteins. Biophys. Chem. 82: 51–67.[CrossRef][Medline]

KIMURA, M., 1977 Preponderance of synonymous changes as evidence for the neutral theory of molecular evolution. Nature 267: 275–276.[CrossRef][Medline]

KOLLMAN, J. M., and R. F. DOOLITTLE, 2000 Determination of the relative rates of change for prokaryotic and eukaryotic proteins with anciently duplicated paralogs. J. Mol. Evol. 51: 173–181.[Medline]

KREIL, D. P., and C. A. OUZOUNIS, 2001 Identification of thermophilic species by the amino acid compositions deduced from their genomes. Nucleic Acids Res. 29: 1608–1615.[Abstract/Free Full Text]

KUMAR, S., and R. NUSSINOV, 2001 How do thermophilic proteins deal with heat? Cell. Mol. Life Sci. 58: 1216–1233.[CrossRef][Medline]

KUMAR, S., C. J. TSAI and R. NUSSINOV, 2000 Factors enhancing protein thermostability. Protein Eng. 13: 179–191.[Abstract/Free Full Text]

LAMBROS, R. J., J. R. MORTIMER and D. R. FORSDYKE, 2003 Optimum growth temperature and the base composition of open reading frames in prokaryotes. Extremophiles 7: 443–450.[CrossRef][Medline]

LAO, P. J., and D. R. FORSDYKE, 2000 Thermophilic bacteria strictly obey Szybalski's transcription direction rule and politely purine-load RNAs with both adenine and guanine. Genome Res. 10: 228–236.[Abstract/Free Full Text]

LYNN, D. J., G. A. C. SINGER and D. A. HICKEY, 2002 Synonymous codon usage is subject to selection in thermophilic bacteria. Nucleic Acids Res. 30: 4272–4277.[Abstract/Free Full Text]

MADIGAN, M. T., J. M. MARTINKO and J. PARKER, 2003 Brock Biology of Microorganisms, Ed. 10. Prentice Hall, Upper Saddle River, NJ.

NEI, M., 1987 Molecular Evolutionary Genetics. Columbia University Press, New York.

PERUTZ, M. F., 1978 Electrostatic effects in proteins. Science 201: 1187–1191.[Abstract/Free Full Text]

SHARP, P. M., and W.-H. LI, 1986 An evolutionary perspective on synonymous codon usage in unicellular organisms. J. Mol. Evol. 24: 28–38.[CrossRef][Medline]

TEKAIA, F., E. YERAMIAN and B. DUJON, 2002 Amino acid composition of genomes, lifestyles of organisms, and evolutionary trends: a global picture with correspondence analysis. Gene 297: 51–60.[CrossRef][Medline]

THOMPSON, J. D., D. G. HIGGINS and T. J. GIBSON, 1994 CLUSTAL W: improvement of the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22: 4673–4680.[Abstract/Free Full Text]

WOESE, C. R., O. KANDLER and M. L. WHEELIS, 1990 Towards a natural system of organisms: proposal for the domains Archaea, Bacteria, and Eucarya. Proc. Natl. Acad. Sci. USA 87: 4576–4579.[Abstract/Free Full Text]

WOLFE, K. H., P. M. SHARP and W.-H. LI, 1989 Mutation rates differ among regions of the mammalian genome. Nature 337: 283–285.[CrossRef][Medline]

YANG, Z., 1997 PAML: a program package for phylogenetic analysis by maximum likelihood. Comput. Appl. Biosci. 13: 555–556.[Free Full Text]

YANG, Z., and R. NIELSEN, 2000 Estimating synonymous and nonsynonymous substitution rates under realistic evolutionary models. Mol. Biol. Evol. 17: 32–43.[Abstract/Free Full Text]

ZHANG, J., H. F. ROSENBERG and M. NEI, 1998 Positive Darwinian selection after gene duplication in primate ribonuclease genes. Proc. Natl. Acad. Sci. USA 98: 3708–3713.

This article has been cited by other articles:

				R. R. Mackwan, G. T. Carver, G. E. Kissling, J. W. Drake, and D. W. Grogan The Rate and Character of Spontaneous Mutation in Thermus thermophilus Genetics, September 1, 2008; 180(1): 17 - 25. [Abstract] [Full Text] [PDF]

				J. Klockgether, D. Wurdemann, O. Reva, L. Wiehlmann, and B. Tummler Diversity of the Abundant pKLC102/PAGI-2 Family of Genomic Islands in Pseudomonas aeruginosa J. Bacteriol., March 15, 2007; 189(6): 2443 - 2459. [Abstract] [Full Text] [PDF]

				E. E. Allen, G. W. Tyson, R. J. Whitaker, J. C. Detter, P. M. Richardson, and J. F. Banfield Genome dynamics in a natural archaeal population PNAS, February 6, 2007; 104(6): 1883 - 1888. [Abstract] [Full Text] [PDF]

				A. Bragonzi, L. Wiehlmann, J. Klockgether, N. Cramer, D. Worlitzsch, G. Doring, and B. Tummler Sequence diversity of the mucABD locus in Pseudomonas aeruginosa isolates from patients with cystic fibrosis. Microbiology, November 1, 2006; 152(Pt 11): 3261 - 3269. [Abstract] [Full Text] [PDF]

				J. W. Drake Chaos and Order in Spontaneous Mutation Genetics, May 1, 2006; 173(1): 1 - 8. [Full Text] [PDF]

				J. W. Drake and C. B. C. Hwang On the Mutation Rate of Herpes Simplex Virus Type 1 Genetics, June 1, 2005; 170(2): 969 - 970. [Abstract] [Full Text] [PDF]

				B. Tummler and P. Cornelis Pyoverdine Receptor: a Case of Positive Darwinian Selection in Pseudomonas aeruginosa J. Bacteriol., May 15, 2005; 187(10): 3289 - 3292. [Full Text] [PDF]

				A. L. Hughes Evidence for Abundant Slightly Deleterious Polymorphisms in Bacterial Populations Genetics, February 1, 2005; 169(2): 533 - 538. [Abstract] [Full Text] [PDF]

International Access Link

Online ISSN: 1943-2361  Print ISSN: 0016-6731
Copyright 2009 by the Genetics Society of America
phone: 412-268-1812   fax: 412-268-1813   email: genetics-gsa{at}andrew.cmu.edu