Department of Biological Sciences, State University of New York, Binghamton, NY 13902, USA1
Centre for Plant Biodiversity Research, CSIRO Plant Industry, Canberra ACT 2601, Australia2
Soybean and Alfalfa Research Laboratory, USDA, ARS, HH-4, Bldg010, BARC-West, 10300 Baltimore Blvd, Beltsville, MD 20705, USA3
Author for correspondence: Matthew A. Parker. Tel: +1 607 777 6283. Fax: +1 607 777 6521. e-mail: mparker{at}binghamton.edu
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Keywords: geographic variation, nitrogenase, nodule bacteria, Rhizobiaceae
Abbreviations: ML, maximum-likelihood; MP, maximum-parsimony; NJ, neighbour-joining
a The GenBank accession numbers for the nifD sequences determined in this work are AF484254AF484287.
b Present address: Centre dEtudes sur le Polymorphisme des Micro-Organismes, UMR CNRS-IRD 9926, 911 Avenue Agropolis BP 64501, 34394 Montpellier Cedex 5, France.
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Incongruence between phylogenetic trees for different loci also provides evidence for horizontal gene transfer (Dykhuizen & Green, 1991 ; Zhou et al., 1997
). Incongruent gene trees have been reported within the genera Rhizobium, Sinorhizobium and Mesorhizobium (Haukka et al., 1998
; Wernegreen & Riley, 1999
; Turner & Young, 2000
; Laguerre et al., 2001
; Qian & Parker, 2002
). However, it is important to investigate additional examples from a variety of geographic regions, to better characterize the prevalence of gene transfer in natural environments.
Bacteria in the genus Bradyrhizobium are important symbionts for diverse legume taxa throughout the world, but no studies have been done with a broad geographic sample of isolates to compare relationships for sequences of 16S rRNA and other gene loci. In this study, we sequenced 822 bp of the gene for the -subunit of nitrogenase (nifD) in isolates of Bradyrhizobium from Asia, North America, Central America and Australia that had previously been characterized for 16S rRNA relationships (Lafay & Burdon, 1998
, 2001
; Parker, 1999
, 2000
, 2001
; Parker & Lunk, 2000
; van Berkum & Fuhrmann, 2000
). Our objectives were (1) to evaluate the degree of phylogenetic tree congruence across the two loci and (2) to compare the geographic structure of bacterial relationships revealed by 16S rRNA vs nifD sequence variation. Our results indicate very different phylogeographic patterns for these two loci, and suggest that differences in gene transfer have greatly altered their relative genealogical history.
![]() |
METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
Phylogenetic analyses.
Sequences were aligned using CLUSTAL W (Thompson et al., 1994 ) and trees were constructed by maximum-parsimony (MP), neighbour-joining (NJ) and maximum-likelihood (ML) methods using the PAUP software (version 4.0b1, from D. L. Swofford, Smithsonian Institution, Washington, DC, USA). To determine the degree of statistical support for branches in the phylogeny (Felsenstein, 1985
), 1000 bootstrap replicates of the data were analysed (100 only for ML). NJ analysis (Saitou & Nei, 1987
) used HKY85 distances (Hasegawa et al., 1985
). Two genera of
-Proteobacteria (Azospirillum brasilense, Rhodobacter capsulatus), for which both 16S rRNA and nifD sequences were available in GenBank (Z29617, D16428, M64344, M15270), served as outgroups in all analyses.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
Seven of the ten Australian isolates, and four of the six Panamanian isolates, each formed an apparent group within the B. japonicum clade. Apart from these small groups, there was little correspondence between geographic origin of the strains and their phylogeny. For example, isolates from North America, Central America and Australia were represented in both of the major groups. To analyse the extent to which geographic origin was distributed non-randomly on the tree of Fig. 1, the location of terminal taxa was randomly permuted 1000 times while holding tree topology constant (Maddison & Slatkin, 1991
). The minimum number of migration events consistent with the phylogeny was then inferred by MP criteria (Slatkin & Maddison, 1989
). This provides a null distribution for the number of migration events that can be compared to the pattern in the actual tree. This analysis indicated that chance alone would generate as much geographic clustering as observed (Fig. 1
) in about 375 out of 1000 trials. Thus, the null hypothesis that phylogenetically related isolates are distributed at random across regions cannot be rejected for 16S rRNA sequences.
nifD relationships
PCR using primers for nifD yielded an 822 bp amplification product with the exception of 10 of the 11 Australian isolates, which generated a 3 bp shorter fragment. All Australian isolates except strain 5111P had a deletion corresponding to codon 220 of the B. japonicum USDA 110 nifD gene. This insertion/deletion polymorphism was excluded from the data for phylogenetic analyses, but it does provide additional support for a close genealogical relationship of nifD genes among the Australian bradyrhizobia (see below). The fraction of polymorphic sites in first, second and third codon positions was 0·157, 0·077 and 0·901, respectively. Mutational saturation at third codon positions can cause misleading results in parsimony analysis of highly divergent sequences (Cunningham, 1997 ). However, the transition/transversion ratio for third position substitutions (2·76) was not depressed relative to first and second positions (1·72, 1·22), indicating that the phylogenetic signal associated with third position polymorphisms had not been erased by multiple substitutions. To investigate this issue, MP analyses were done on both the full dataset and on a reduced dataset where third codon positions with transitions were excluded.
MP analysis of the full nifD dataset yielded 84 trees that differed primarily in the placement of closely related strains within terminal clades. In the consensus tree, nearly all nifD sequences clustered into groups exclusively composed of taxa from a single geographic region (Fig. 2). For example, five Panamanian isolates from diverse legume hosts formed a group in 100% of bootstrap replicates. Similarly, all of the Australian isolates formed a highly supported group. All B. elkanii strains had identical nifD sequences, and were placed with all of the other North American strains (jwc91-2, th-b2, ApB16 and 10 B. japonicum strains) in 97% of bootstrap replicates. The three Asian soybean strains (USDA 6, USDA 38, B. liaoningense) each had nifD sequences that were identical to those of certain North American B. japonicum strains (e.g. USDA 4 and USDA 135). Therefore, the entire set of Bradyrhizobium strains originating from North America and Asia was labelled North Temperate (Fig. 2
). An MP analysis excluding 121 third codon sites with transitions yielded an identical tree topology as in Fig. 2
, with bootstrap values of 99, 89 and 100% for the groups formed by strains originating from North Temperate, Australia and Panama, respectively. The same three highly supported groups and precisely the same branching order within groups as in Fig. 2
were observed in NJ and ML analyses. Bootstrap values for the North Temperate, Australia and Panama groups were all 100% for the NJ tree and were 99, 98 and 100%, respectively, for the ML analysis.
|
In summary, 36 of the 38 Bradyrhizobium nifD sequences clustered into three geographic groups that were highly supported by all methods of phylogenetic analysis, and the relationships of two others (from isolates Tv2a-2 and Cj3-3) must be considered unresolved. The extent of geographic clustering was analysed by a permutation test (Maddison & Slatkin, 1991 , as described above). In the actual tree, the geographic distribution of all nifD sequences requires a hypothesis of just four changes of location throughout the entire phylogeny. However, in 1000 randomly permuted datasets, the number of inferred migration events ranged from 8 to 16 (median 13). Thus, phylogenetically related nifD sequences showed highly significant geographic clustering; there is less than one chance in a thousand of obtaining the pattern of regional clades (Fig. 2
) under the null hypothesis that geographic origin and genealogy are independent.
Are 16S rRNA and nifD trees congruent?
Comparison of 16S rRNA and nifD relationships indicated that there were several major discrepancies. The primary division between B. japonicum and B. elkanii lineages in the 16S rRNA data (Fig. 1) was missing for nifD (Fig. 2
). For example, strain Cj3-3 had a 16S rRNA sequence identical to that of B. elkanii USDA 94, but had a nifD sequence that was divergent from all B. elkanii strains. The 16S rRNA gene of Australian strain 5111P differed by only two to three substitutions from certain North American isolates in the B. elkanii group (Fig. 1
), but the nifD gene of strain 5111P was similar to those of other Australian strains (Fig. 2
) related to B. japonicum in the 16S rRNA analysis. Isolate ApB16 was a close relative of several isolates in the 16S rRNA B. japonicum group (e.g. it differed from USDA 38 and 5329H by one and two substitutions, respectively). However, the nifD sequence of ApB16 was highly similar to those of two North American isolates (jwc91-2, th-b2) that were placed with B. elkanii in the 16S rRNA analysis. The partition homogeneity test of Farris et al. (1995)
indicated that the MP trees for 16S rRNA and nifD were significantly incongruent (P<0·001). Thus, the genealogical histories of these two genes are demonstrably different for this set of Bradyrhizobium strains.
Gene duplication could potentially lead to an incorrect inference regarding phylogenetic incongruence across loci. For example, if multiple copies of either 16S rRNA or nifD genes are present, then a mixed sample that included non-orthologous gene sequences could misrepresent the true pattern of relationships (Doyle & Davis, 1998 ; Wendel & Doyle, 1998
). However, all existing evidence suggests that Bradyrhizobium possesses a single rRNA gene region (Kundig et al., 1995
; van Berkum et al., 1998
; Klappenbach et al., 2000
). In strain USDA 110 of B. japonicum, nifD also appears to be a single-copy gene (Kundig et al., 1993
; Gottfert et al., 2001
), but information is very limited for other strains. Therefore, Southern hybridization analysis was done with ten of the Bradyrhizobium strains using HindIII and EcoRI digests of genomic DNA and 822 bp of the B. japonicum USDA 110 nifD gene as probe. For HindIII digests, a single hybridization signal was observed in all isolates, whereas in EcoRI digests six of the ten isolates had a single band and the remaining four isolates (jwc91-2, th-b2, ApB16 and Cj3-3) displayed two hybridization signals. Inspection of nifD sequences for these four strains indicated that they had an EcoRI site in the middle of the gene, thus accounting for the presence of two genomic restriction fragments with nifD sequence homology. These results imply that all strains most likely have a single copy of nifD. Thus, the incongruence of tree topology in Figs 1
and 2
is unlikely to be an artefact caused by the inclusion of paralogous nifD genes in the analysis.
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The only cases where identical nifD sequences were found among strains originating from different regions involved symbionts of soybean (Glycine max). USDA 6 and USDA 38, both from Japan, had nifD sequences that were identical to several North American B. japonicum strains (USDA 4, USDA 123, USDA 127). Also, the nifD sequence of B. liaoningense from China was identical to B. japonicum USDA 135 from Iowa. These possibly represent cases of recent human-mediated dispersal, since there are historical records documenting the deliberate introduction of bacterial symbionts from Asia during the early years of soybean cultivation in North America (e.g. Wilson, 1934 ). Apart from these two cases, the complete absence of any closely related nifD sequences from different regions was noteworthy. This suggests that inter-regional migration of Bradyrhizobium has been negligible over recent evolutionary time.
Host specificity could be a factor contributing to the strong regional differentiation of nifD sequences. If bradyrhizobial strains tended to be narrowly specialized for nodulation of specific legume taxa, then the restricted geographic range shown by most legume hosts would limit the distribution of bacterial lineages. For example, the hosts for three Panama isolates were legumes in the tribe Dalbergieae (Machaerium, Platypodium), which is largely restricted to tropical America, and all but one of the Australian isolates came from legume genera in the tribes Mirbelieae and Bossiaeae which are endemic to Australia and nearby south-east Asia. Existing data on host specificity are too limited to fully analyse this problem. All of the Australian and Panamanian strains have the ability to form nodules on Macroptilium atropurpureum (tribe Phaseoleae; Parker & Lunk, 2000 ; Parker, 2000
, 2001
; B. Lafay, unpublished data), which indicates that they are not strictly specialized on their original legume host. Among the Panama isolates, apparently identical bacterial genotypes are shared by legume hosts from three separate legume tribes (Parker & Lunk, 2000
; Parker, 2001
), again suggesting that these strains are not highly host specific. An important direction for future research will be to analyse bacteria sampled from related legumes in different regions, so that host taxonomic group is not confounded with geographic origin.
The differential rate of evolution of 16S rRNA and nifD may be partially responsible for variation in phylogeographic patterns. The 16S rRNA gene evolves slowly (Woese, 1987 ), with an estimated substitution rate per site of 0·0120·018 per 100 million years in non-endosymbiotic bacteria (Clark et al., 1999
). Assuming that substitutions occur according to a Poisson process (Werren et al., 1995
), the upper 95% confidence limit for divergence time in cases where strains showed no nucleotide differences throughout the entire 1410 bp 16S rRNA fragment would range from 5·9 to 8·9 million years. Thus, pairs of isolates in different regions that descended from a common ancestor more than 5 million years in the past would often retain identical 16S rRNA sequences. No estimates of evolutionary rate for nifD have been made. However, a substantial difference in evolutionary rate is implied since nifD had more than a sixfold higher frequency of polymorphic sites than 16S rRNA (0·376 vs 0·059). If a sixfold higher evolutionary rate is assumed for nifD, then there is less than a 5% chance that pairs of isolates separated for more than 1·5 million years would retain identical sequences. A higher evolutionary rate for nifD would contribute to the evolution of distinctive regional clades in geographically isolated populations (Fig. 2
), despite a lack of strong phylogeographic structure in the more slowly evolving 16S rRNA gene (Fig. 1
).
Nevertheless, differences in evolutionary rate alone cannot account for the significantly different tree topologies for the two genes. Figs 1 and 2
reveal numerous examples where strains that were closely related for one gene had very divergent sequences in the other, and the overall structure of strain relationships was significantly discordant across loci. For example, the most basal split in the 16S rRNA tree involved divergence into two lineages related to B. japonicum and B. elkanii, which were both represented in most of the regions sampled (Fig. 1
). By contrast, in the nifD tree there was no indication of a basal split into two lineages each shared by multiple geographic areas (Fig. 2
).
Horizontal gene transfer at one or both loci is the most likely process responsible for these conflicting phylogenies (Spratt & Maiden, 1999 ). For bacterial ribosomal genes, several cases have been reported for transfer events involving entire genes (Yap et al., 1999
) or portions of genes (Eardly et al., 1996
; Wang & Zhang, 2000
; Parker, 2001
). However, our data suggest that horizontal transfers of nifD may have been more prevalent. Isolates from the same geographic region almost invariably resembled one another for nifD sequences, even when they were not close relatives at the 16S rRNA locus. This would suggest a scenario where each region was initially colonized by multiple Bradyrhizobium lineages with divergent 16S rRNA sequences, followed by a process of lateral nifD gene transfer. More frequent transfer of nifD than 16S rRNA would account for the within-region homogeneity of nifD sequences (Fig. 2
) across lineages that have retained divergent 16S rRNA genes (Fig. 1
).
For symbiotic loci such as the common nod genes, it has been suggested that gene trees match the phylogeny of host legumes and often show discrepancies from bacterial 16S rRNA relationships (Ueda et al., 1995 ; Wernegreen & Riley, 1999
; Laguerre et al., 2001
). The extent to which nifD may be expected to match host plant phylogeny is uncertain. Differences commonly exist between nod and nif gene trees (Haukka et al., 1998
; Laguerre et al., 2001
) and our nifD tree showed no clear correspondence to the phylogenetic relationships of host legumes. For example, the legume Tachigali belongs to a lineage in the subfamily Caesalpinioideae that gave rise to the subfamily Mimosoideae (Doyle et al., 1997
). Yet the sole bradyrhizobial isolate from a mimosoid legume (strain 1808N from Acacia) did not cluster with the Tachigali isolate (Tv2a-2), but rather, grouped with isolates from distantly related legumes (Fig. 2
). Also, the host for strain Ec3-3 from Panama is in the same subtribe as that for isolate ApB16 from North America (subtribe Erythrininae, tribe Phaseoleae), yet both these strains had nifD sequences that grouped with isolates from more distantly related legumes sharing the same geographic region. The Australian strains from the legume tribe Bossiaceae (Bossiaea, Goodia) and tribe Mirbelieae (Mirbelia, Podolobium, Oxylobium, Daviesia) also did not form distinct groups within the nifD tree (Fig. 2
).
The nifD locus occurs sufficiently near to the common nod gene cluster (220 kb apart on the B. japonicum USDA 110 chromosome; Kundig et al., 1993
) that evolution of host specificity conditioned by nod genes may affect nifD sequence variation. For example, a 500 kb transmissible chromosomal symbiotic element (symbiosis island) has been characterized in Mesorhizobium loti (Sullivan & Ronson, 1998
). If similar transmissible elements existed in Bradyrhizobium, then nod gene transfer events associated with selection for symbiotic compatibility with novel legume hosts (Sullivan et al., 1995
) may also redistribute nifD sequences across bacterial lineages. It will be important in future work to clarify mechanisms of gene transfer that have acted in Bradyrhizobium.
The prevalence of lateral transfer is an important issue both for bacterial population genetics and systematics. In the presence of lateral transfer, the true genealogical history will vary among genes, or among different portions of the same gene (Dykhuizen & Green, 1991 ; Zhou et al., 1997
; Spratt & Maiden, 1999
). Within the genus Bradyrhizobium, the basic division between the two primary taxa B. japonicum and B. elkanii is widely accepted (Kuykendall et al., 1992
; Young & Haukka, 1996
; Barrera et al., 1997
; van Berkum & Fuhrmann, 2000
; Willems et al., 2001
). However, most existing studies of Bradyrhizobium have focused mainly on ribosomal gene sequences. Our nifD results provide no support for the interpretation that the genus Bradyrhizobium is divided into two primary lineages affiliated with B. japonicum or B. elkanii, and instead suggest that the genus may be composed of an unknown number of regionally endemic lineages. To avoid biases associated with the idiosyncratic genealogical history of any one gene, it will be important in future work to extend analyses to additional loci (of both housekeeping and symbiotic genes) to provide a broad comparative basis for understanding Bradyrhizobium evolution.
![]() |
ACKNOWLEDGEMENTS |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Clark, M. A., Moran, N. A. & Baumann, P. (1999). Sequence evolution in bacterial endosymbionts having extreme base composition. Mol Biol Evol 16, 1586-1598.[Abstract]
Cunningham, C. W. (1997). Can three incongruence tests predict when data should be combined? Mol Biol Evol 14, 733-740.[Abstract]
Doyle, J. J. & Davis, J. I. (1998). Homology in molecular phylogenetics: a parsimony perspective. In Molecular Systematics of Plants. II. DNA Sequencing, , pp. 101-131. Edited by D. E. Soltis, P. S. Soltis & J. J. Doyle. Boston:Kluwer.
Doyle, J. J., Doyle, J. L., Ballenger, J. A., Dickson, E. E., Kajita, T. & Ohashi, H. (1997). A phylogeny of the chloroplast gene rbcl in the Leguminosae: taxonomic correlations and insights into the evolution of nodulation. Am J Bot 84, 541-554.[Abstract]
Dykhuizen, D. E. & Green, L. (1991). Recombination in Escherichia coli and the definition of biological species. J Bacteriol 173, 7257-7268.[Medline]
Eardly, B. D., Wang, F.-S. & van Berkum, P. (1996). Corresponding 16S rRNA gene segments in Rhizobiaceae and Aeromonas yield discordant phylogenies. Plant Soil 186, 69-74.
Farris, J. S., Kallersjo, M., Kluge, A. G. & Bult, C. (1995). Testing significance of incongruence. Cladistics 10, 315-319.
Felsenstein, J. (1985). Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39, 783-791.
Gottfert, M. S., Rothlisberger, S., Kundig, C., Beck, C., Marty, R. & Hennecke, H. (2001). Potential symbiosis-specific genes uncovered by sequencing a 410-kilobase DNA region of the Bradyrhizobium japonicum chromosome. J Bacteriol 183, 1405-1412.
Hasegawa, M., Kishino, H. & Yano, T. (1985). Dating the human-ape split by a molecular clock of mitochondrial DNA. J Mol Evol 22, 160-174.[Medline]
Haukka, K., Lindstrom, K. & Young, J. P. W. (1998). Three phylogenetic groups of nodA and nifH genes in Sinorhizobium and Mesorhizobium isolates from leguminous trees growing in Africa and Latin America. Appl Environ Microbiol 64, 419-426.
Klappenbach, J. A., Dunbar, J. M. & Schmidt, T. M. (2000). rRNA operon copy number reflects ecological strategies of bacteria. Appl Environ Microbiol 66, 1328-1333.
Kundig, C., Hennecke, H. & Gottfert, M. (1993). Correlated physical and genetic map of the Bradyrhizobium japonicum 110 genome. J Bacteriol 175, 613-622.[Abstract]
Kundig, C., Beck, C., Hennecke, H. & Gottfert, M. (1995). A single rRNA gene region in Bradyrhizobium japonicum. J Bacteriol 177, 5151-5154.[Abstract]
Kuykendall, D., Saxena, B., Devine, T. E. & Udell, S. E. (1992). Genetic diversity in Bradyrhizobium japonicum Jordan 1982 and a proposal for Bradyrhizobium elkanii sp. nov. Can J Microbiol 38, 501-505.
Lafay, B. & Burdon, J. J. (1998). Molecular diversity of rhizobia occurring on native shrubby legumes in southeastern Australia. Appl Environ Microbiol 64, 3989-3997.
Lafay, B. & Burdon, J. J. (2001). Small-subunit rRNA genotyping of rhizobia nodulating Australian Acacia spp. Appl Environ Microbiol 67, 396-402.
Laguerre, G., Nour, S. M., Macheret, V., Sanjuan, J., Drouin, P. & Amarger, N. (2001). Classification of rhizobia based on nodC and nifH gene analysis reveals a close phylogenetic relationship among Phaseolus vulgaris symbionts. Microbiology 147, 981-993.
Maddison, W. P. & Slatkin, M. (1991). Null models for the number of evolutionary steps in a character on a phylogenetic tree. Evolution 45, 1184-1197.
Parker, M. A. (1999). Relationships of bradyrhizobia from the legumes Apios americana and Desmodium glutinosum. Appl Environ Microbiol 65, 4914-4920.
Parker, M. A. (2000). Divergent Bradyrhizobium symbionts on Tachigali versicolor from Barro Colorado Island, Panama. Syst Appl Microbiol 23, 585-590.[Medline]
Parker, M. A. (2001). Case of localized recombination in 23S rRNA genes from divergent Bradyrhizobium lineages associated with neotropical legumes. Appl Environ Microbiol 67, 2076-2082.
Parker, M. A. & Lunk, A. (2000). Relationships of bradyrhizobia from Platypodium and Machaerium (Papilionoideae tribe Dalbergieae) on Barro Colorado Island, Panama. Int J Syst Evol Microbiol 50, 1179-1186.[Abstract]
Qian, J. & Parker, M. A. (2002). Contrasting nifD and ribosomal gene relationships among Mesorhizobium from Lotus oroboides in northern Mexico. Syst Appl Microbiol (in press).
Rigottier-Gois, L., Turner, S. L., Young, J. P. W. & Amarger, N. (1998). Distribution of repC plasmid-replication sequences among plasmids and isolates of Rhizobium leguminosarum bv. viciae from field populations. Microbiology 144, 771-780.[Abstract]
Saitou, N. & Nei, M. (1987). The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol 4, 406-425.[Abstract]
Shimodaira, H. & Hasegawa, M. (1999). Multiple comparisons of log-likelihoods with applications to phylogenetic inference. Mol Biol Evol 16, 1114-1116.
Slatkin, M. & Maddison, W. P. (1989). A cladistic measure of gene flow inferred from the phylogenies of alleles. Genetics 123, 603-613.
Spratt, B. G. & Maiden, M. C. J. (1999). Bacterial population genetics, evolution and epidemiology. Philos Trans R Soc Lond B 354, 701-710.[Medline]
Sullivan, J. T. & Ronson, C. W. (1998). Evolution of rhizobia by acquisition of a 500-kb symbiosis island that integrates into a phe-tRNA gene. Proc Natl Acad Sci USA 95, 5145-5149.
Sullivan, J. T., Patrick, H. N., Lowther, W. L., Scott, D. B. & Ronson, C. W. (1995). Nodulating strains of Rhizobium loti arise through chromosomal symbiotic gene transfer in the environment. Proc Natl Acad Sci USA 92, 8885-8889.
Sullivan, J. T., Eardly, B. D., van Berkum, P. & Ronson, C. W. (1996). Four unnamed species of nonsymbiotic rhizobia isolated from the rhizosphere of Lotus corniculatus. Appl Environ Microbiol 62, 2818-2825.[Abstract]
Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994). CLUSTAL W: improving 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]
Turner, S. L. & Young, J. P. W. (2000). The glutamine synthetases of rhizobia: phylogenetics and evolutionary implications. Mol Biol Evol 17, 309-319.
Ueda, T., Suga, Y., Yahiro, N. & Matsuguchi, T. (1995). Phylogeny of Sym plasmids of rhizobia by PCR-based sequencing of a nodC segment. J Bacteriol 177, 468-472.[Abstract]
van Berkum, P. & Fuhrmann, J. J. (2000). Evolutionary relationships among the soybean bradyrhizobia reconstructed from 16S rRNA gene and internally transcribed spacer region sequence divergence. Int J Syst Evol Microbiol 50, 2165-2172.[Medline]
van Berkum, P., Ruihua, F., Campbell, T. A. & Eardly, B. D. (1998). Some issues of relevance in the taxonomy of rhizobia. In Proceedings of the 16th North American Conference on Symbiotic Nitrogen Fixation, pp. 15. Cancun, Mexico.
Wang, Y. & Zhang, Z. (2000). Comparative sequence analyses reveal frequent occurrence of short segments containing an abnormally high number of non-random base variations in bacterial rRNA genes. Microbiology 146, 2845-2854.
Wendel, J. F. & Doyle, J. J. (1998). Phylogenetic incongruence: window into genome history and molecular evolution. In Molecular Systematics of Plants. II. DNA Sequencing , pp. 265-296. Edited by D. E. Soltis, P. S. Soltis & J. J. Doyle. Boston:Kluwer.
Wernegreen, J. J. & Riley, M. A. (1999). Comparison of the evolutionary dynamics of symbiotic and housekeeping loci: a case for the genetic coherence of rhizobial lineages. Mol Biol Evol 16, 98-113.[Abstract]
Wernegreen, J. J., Harding, E. E. & Riley, M. A. (1997). Rhizobium gone native: unexpected plasmid stability of indigenous Rhizobium leguminosarum. Proc Natl Acad Sci USA 94, 5483-5488.
Werren, J. H., Zhang, W. & Guo, L. R. (1995). Evolution and phylogeny of Wolbachia: reproductive parasites of arthropods. Proc R Soc Lond B 261, 55-71.[Medline]
Willems, A., Coopman, R. & Gillis, M. (2001). Phylogenetic and DNA-DNA hybridization analyses of Bradyrhizobium species. Int J Syst Evol Microbiol 51, 111-117.[Abstract]
Wilson, J. K. (1934). Longevity of Rhizobium japonicum in relation to its symbiont on the soil. Cornell Univ Agric Exp Stn Mem 162, 1-11.
Woese, C. R. (1987). Bacterial evolution. Microbiol Rev 51, 221-271.
Xu, L. M., Ge, C., Cui, Z., Li, J. & Fan, H. (1995). Bradyrhizobium liaoningense sp. nov., isolated from the root nodules of soybeans. Int J Syst Bacteriol 45, 706-711.[Abstract]
Yap, W. H., Zhang, Z. & Wang, Y. (1999). Distinct types of rRNA operons exist in the genome of the actinomycete Thermomonospora chromogena and evidence for horizontal transfer of an entire rRNA operon. J Bacteriol 181, 5201-5209.
Young, J. P. W. & Wexler, M. (1988). Sym plasmid and chromosomal genotypes are correlated in field populations of Rhizobium leguminosarum. J Gen Microbiol 134, 2731-2739.
Young, J. P. W. & Haukka, K. E. (1996). Diversity and phylogeny of rhizobia. New Phytol 133, 87-94.
Zhou, J., Bowler, L. D. & Spratt, B. G. (1997). Interspecies recombination, and phylogenetic distortions, within the glutamine synthetase and shikimate dehydrogenase genes of Neisseria meningitidis and commensal Neisseria species. Mol Microbiol 23, 799-812.[Medline]
Received 12 March 2002;
revised 17 April 2002;
accepted 7 May 2002.