The nitrogen-fixing gene (nifH) of Rhodopseudomonas palustris: a case of lateral gene transfer?
Jose Jason L. Cantera
,
Hiroko Kawasaki and
Tatsuji Seki
The International Center for Biotechnology, Osaka University, 2-1 Yamada-oka, Suita-shi 565-0871, Japan
Correspondence
Hiroko Kawasaki
ICBKawasakiNakagawa{at}icb.osaka-u.ac.jp
 |
ABSTRACT
|
---|
Nitrogen fixation is catalysed by some photosynthetic bacteria. This paper presents a phylogenetic comparison of a nitrogen fixation gene (nifH) with the aim of elucidating the processes underlying the evolutionary history of Rhodopseudomonas palustris. In the NifH phylogeny, strains of Rps. palustris were placed in close association with Rhodobacter spp. and other phototrophic purple non-sulfur bacteria belonging to the
-Proteobacteria, separated from its close relatives Bradyrhizobium japonicum and the phototrophic rhizobia (Bradyrhizobium spp. IRBG 2, IRBG 228, IRBG 230 and BTAi 1) as deduced from the 16S rRNA phylogeny. The close association of the strains of Rps. palustris with those of Rhodobacter and Rhodovulum, as well as Rhodospirillum rubrum, was supported by the mol% G+C of their nifH gene and by the signature sequences found in the sequence alignment. In contrast, comparison of a number of informational and operational genes common to Rps. palustris CGA009, B. japonicum USDA 110 and Rhodobacter sphaeroides 2.4.1 suggested that the genome of Rps. palustris is more related to that of B. japonicum than to the Rba. sphaeroides genome. These results strongly suggest that the nifH of Rps. palustris is highly related to those of the phototrophic purple non-sulfur bacteria included in this study, and might have come from an ancestral gene common to these phototrophic species through lateral gene transfer. Although this finding complicates the use of nifH to infer the phylogenetic relationships among the phototrophic bacteria in molecular diversity studies, it establishes a framework to resolve the origins and diversification of nitrogen fixation among the phototrophic bacteria in the
-Proteobacteria.
The GenBank accession numbers for the nifH and 16S rDNA sequences reported in this paper are AB079615 to AB079635, AB079675 to AB079683 and AB079690.
Present address: Environmental Sciences Department, University of California, Riverside, CA 92521, USA.
 |
INTRODUCTION
|
---|
Rhodopseudomonas palustris, a phototrophic purple non-sulfur bacterium, is phylogenetically close to the nodule-forming, non-phototrophic Bradyrhizobium japonicum (Wong et al., 1994
; Woese et al., 1984
) and the bacteriochlorophyll a-synthesizing stem-nodulating phototrophic rhizobia (Eaglesham et al., 1990
; Inui et al., 2001
; Fleischman et al., 1995
; Wong et al., 1994
; Young et al., 1991
) in the
-2 Proteobacteria based on sequence analysis of the 16S rRNA. However, this bacterium shares a number of phenotypic characteristics with the Rhodobacter spp. and other purple non-sulfur bacteria. In fact, some of the presently known species of Rhodobacter were formerly included in the genus Rhodopseudomonas (Imhoff et al., 1984
). The close juxtaposition of Rps. palustris and B. japonicum merits further study, since the evolutionary history of the former, particularly its phototrophic ability, could not be accurately inferred from the 16S rRNA data alone. Phylogenetic studies of genes that are common and indispensable to both organisms are needed to elucidate the evolutionary history of Rps. palustris and B. japonicum.
Rps. palustris and other phototrophic purple non-sulfur bacteria are able to fix molecular nitrogen (Hennecke et al., 1985
), a characteristic shared with B. japonicum and the phototrophic rhizobia, and other physiologically diverse groups of Bacteria and Archaea (Young, 1992
). There is evidence that nitrogen fixation activity is coupled to photosynthesis in another phototrophic bacterium, Rhodobacter capsulatus (Rhodopseudomonas capsulata) (Meyer et al., 1978
). Moreover, the nitrogenase iron protein (NifH, encoded by nifH) shares structural features with the chlorophyll iron protein subunits (chlorin reductase, encoded by the bchL and bchX genes) in Rba. capsulatus, suggesting their evolution from an ancient gene duplication event; this led Xiong et al. (1998)
to use nifH as an outgroup to root the tree of the bacteriochlorophyllchlorophyll genes. In addition, nifH shares a common evolutionary history with 16S rRNA (Hennecke et al., 1985
; Ueda et al., 1995
), and thus has been used as molecular marker in diversity studies (Widmer et al., 1999
; Ohkuma et al., 1999
).
In this study we used the nifH gene from a set of reference species belonging to the
-Proteobacteria to provide more substantive information on the evolutionary relationship of Rps. palustris and B. japonicum. The aberrant position of Rps. palustris in the phylogenetic tree derived from NifH strongly suggested the occurrence of lateral gene transfer. Further evidence for the lateral transfer of nifH was provided by the Rps. palustris CGA009 genome sequence data. This study helps to establish a theoretical framework for understanding the diversification of the phototrophic bacteria in the Proteobacteria by identifying the factors (i.e. lateral gene transfer) involved in the phenotypic diversity of these species.
 |
METHODS
|
---|
Bacterial strains and growth conditions.
The species names and strain numbers of the taxa investigated, and the accession numbers of the sequences analysed in this study, are given in Table 1
. Phototrophic purple non-sulfur bacteria were grown as described previously (Cantera et al., 2002
; Kawasaki et al., 1993
); the phototrophic rhizobia were cultivated in yeast extract-glucose (YGA) medium (Ladha & So, 1994
).
View this table:
[in this window]
[in a new window]
|
Table 1. Bacterial strains and GenBank/DDBJ/EMBL accession/protein ID numbers of nifH and 16S rRNADNA sequences used in this study
|
|
DNA preparation and sequencing.
Chromosomal DNA was extracted using the method of Ausubel et al. (1995)
. PCR amplifications were performed using Ex Taq polymerase (TaKaRa Shuzo) in a GeneAmp PCR system 9700 (PE Applied Biosystems). PCR primers used for nifH amplification were nifH-1c (5'-CARATCGCVTTYTACGG-3') (Torok & Kondorosi, 1981
) and nifH-GEM-R (5'-ADNGCCATCATYTCNCC-3') (Ohkuma et al., 1996
), located at positions 340356 and 783800 of the B. japonicum USDA 110 gene (accession no. K01620), respectively. PCR products were cloned using the Original TA cloning kit (InVitrogen), and plasmids containing the insert were extracted using the QIAprep Spin Miniprep kit (Qiagen). M13F and M13R primers were used for sequencing cloned nifH PCR products. The 16S rDNA gene sequences were determined using the primers as described by Lisdiyanti et al. (2000)
. PCR products were purified by using either the Qiagen Gel Extraction Kit or the Qiagen PCR Purification Kit. DNA sequences were determined by the dideoxy-chain-termination method (Sanger et al., 1977
) with a BigDye Terminator cycle sequencing kit (PE Applied Biosystems) applied to an ABI PRISM 310 Genetic analyser (Perkin-Elmer). Sequence data were analysed by the ABI PRISM (Perkin-Elmer) sequence analysis program and assembled using the ABI Auto Assembler (Perkin-Elmer).
In silico analyses.
Homology searches were performed via BLAST (Altschul et al., 1997
) either at the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/) or at the DNA Database of Japan (DDBJ, http://www.ddbj.nig.ac.jp/). Other bacterial gene sequences used for the tree construction were obtained from the EMBL/GenBank databases. Sequence data for Rps. palustris CGA009 and Rhodobacter sphaeroides 2.4.1 were obtained from the DOE Joint Genome Institute (JGI) at http://spider.jgi-psf.org/. Putative amino acid sequences were deduced from the nucleotide sequences using Genetix-Win (version 3.1.0). Unaligned regions and gaps were excluded from the analyses. A total of 153 amino acids and 1247 nucleotide positions were used in the NifH and 16S rDNA analyses, respectively. Phylogenetic analyses were performed using the neighbour-joining (NJ) method (Saitoh & Nei, 1987
) of the CLUSTAL X program (Jeanmougin et al., 1998
; Thompson et al., 1997
), the maximum-likelihood (ML) method (Jones, Taylor and Thornton's method) of the MOLPHY 2.3 program (Adachi & Hasegawa, 1996
), and the maximum-parsimony (MP) of PHYLIP version 3.6a2 (default parameters) after manual refinement of the alignments. Evolutionary distance (ED) analyses were conducted on the 16S rDNA dataset using the Kimura two-parameter model in CLUSTAL X. Graphical representations of the resulting trees were made using NJPlot (Perrière & Gouy, 1996
) and TreeView (Page, 1996
).
Using the B. japonicum USDA 110 genes available from the DNA databases as query sequence, the single most similar Rps. palustris CGA009 ORF for each B. japonicum USDA 110 and Rba. sphaeroides 2.4.1 ORF was obtained by BLAST (Altschul et al., 1997
) using an expected cutoff value of 10·0. Each BLAST result produced an inferred Rps. palustris CGA009 amino acid sequence and an alignment with either B. japonicum USDA 110 or Rba. sphaeroides 2.4.1. The length of the alignment was noted, as it gave an indication of how extensive the similarity between the sequences might be. For example, a significant hit extending over just a few amino acids might indicate a similar motif but not the identification of a homologue. The alignment lengths were used to further evaluate the similarity between the genes. The similarity values of Rps. palustris CGA009 genes and their corresponding genes in either Rba. sphaeroides or B. japonicum were plotted on a graph for comparison.
 |
RESULTS AND DISCUSSION
|
---|
NifH phylogeny
Using degenerate PCR primers, DNA fragments (461 bp) with strong sequence homologies to previously published nifH sequences were amplified from five strains of Rps. palustris, seven species of the phototrophic purple non-sulfur bacteria belonging to the RhodobacterRhodovulum group, four strains of photosynthetic rhizobia and two strains of B. japonicum, all belonging to the
-Proteobacteria. The partial sequences contained both conserved and variable regions, and are long enough to determine sequence variations among the nifH gene from different strains. The sequences were translated into amino acids to avoid bias caused by the degeneracy in the third codon position, and were compared with other sequences already available in the database. Two highly similar nifH sequences (97 % amino acid similarity) were found in the Rps. palustris CGA009 genome, and were included in the subsequent analyses. The Rps. palustris ATCC 17001 partial nifH gene sequence determined in this study is 100 % similar to that of the NifH homologue from Rps. palustris CGA009 located at position 52048505205746; those of strains HMD88 and HMD89 were 100 % similar to the gene homologue located at 15224881523378 (see genome sequence of Rps. palustris CGA009). For most of the species analysed, only one copy of nifH was randomly detected by PCR, although there is a possibility that another highly similar copy of nifH is present in these species, particularly in Rps. palustris strains. Nevertheless, the two copies of this gene in strain CGA009 share high similarity with each other.
A phylogenetic tree derived from NifH amino acid sequences that included the two NifH sequences from the Rps. palustris CGA009 genome sequence was constructed to assess the relationship of the sequences investigated (Fig. 1
). For comparison, a dendrogram constructed from the 16S rRNA of an almost identical set of species was constructed to avoid sampling artifacts (Fig. 2
). Since some of the 16S rDNA sequences were either incomplete or not available, they were determined in this study (Table 1
). In both trees, sequences from Nostoc (Anabaena) sp. PCC 7120 were used as outgroup since cyanobacteria are located in a deeper branch in the 16S rRNA phylogeny than the strains investigated (Stackebrandt et al., 1996
). Overall, highly similar orderings of taxa were found between the NifH and the 16S rRNA trees with all treeing methods (neighbour-joining, maximum-likelihood and maximum-parsimony), except for the position of Rps. palustris strains in the NifH tree. The NifH sequences of Rps. palustris strains (including both copies in strain CGA009) were unusually placed into a highly supported cluster (92 % bootstrap value) with the other phototrophic purple non-sulfur bacteria belonging to different phylogenetic lineages that consisted of Rba. capsulatus, Rba. sphaeroides, Rhodobacter blasticus, Rhodobacter azotoformans, Rhodobacter sp. AP-10, Rhodovulum sulfidophilum, Rhodovulum strictum, Rhodovulum sp. CP-10 (
-3 Proteobacteria) and Rhodospirillum rubrum (
-1 Proteobacteria), suggesting the close evolutionary relationship of their NifH.

View larger version (56K):
[in this window]
[in a new window]
|
Fig. 1. Phylogenetic tree based on partial sequences (approx. 153 amino acids) of NifH showing the phylogenetic positions of Rps. palustris and B. japonicum relative to other phototrophic purple non-sulfur and nodule-forming bacteria. NifH from a cyanobacterium was used as outgroup. The tree was constructed using the neighbour-joining method; bootstrap values derived from 1000 replicates are shown as percentages. Branches supported by 90 % bootstrap values or more using the neighbour-joining method are shown as bold lines; bootstrap values greater than 90 % are shown above the branch. Bootstrap values greater than 90 % from the maximum-likelihood method are shown below the branch. The lengths of the horizontal lines are proportional to the evolutionary distances; the lengths of the vertical lines have no significance. The mol% G+C of the nifH gene and the length of sequence analysed are shown in parentheses. Species sequenced in this study are in bold. Common phenotypic characters of members of each cluster are also indicated. The bar represents 0·1 changes per amino acid.
|
|

View larger version (33K):
[in this window]
[in a new window]
|
Fig. 2. Phylogenetic tree based on 16S rRNA showing the phylogenetic positions of strains of Rps. palustris and B. japonicum relative to other purple non-sulfur and nodule-forming bacteria. The phylogenetic tree was constructed using the neighbour-joining method. Bootstrap values derived from 1000 replicates are shown as percentages; bold lines indicate branches supported by bootstrap values greater than 90 %. The lengths of the horizontal lines are proportional to the evolutionary distances, while the lengths of the vertical lines have no significance. Species sequenced in this study are in bold. The 16S rRNA sequence of Rps. palustris CGA009 is 100 % similar to that of Rps. palustris ATCC 17001 and was not included in the tree. The bar represents 0·1 changes per nucleotide.
|
|
The highly supported clustering of the Rps. palustris strains with other purple non-sulfur bacteria included in this study was highlighted by the mol% G+C of their nifH gene (Fig. 1
), wherein no significant variation was observed. It was also apparent that the mol% G+C was almost the same for all the phototrophic strains (including the phototrophic Bradyrhizobium spp. IRBG 2, IRBG 228, IRBG 230 and BTAi 1). However, a variation of about 5 % between the nifH G+C content of Rps. palustris strains (6263 mol%) and two strains of B. japonicum (IAM 12608T and USDA 100) (56 mol%) was evident; and at least 3 % difference from that of B. elkanii (59 mol%). Furthermore, manual examination of the aligned partial NifH sequences (Fig. 3
) showed at least 17 amino acid substitutions unique among the phototrophic anaerobes, five strains of Rps. palustris and ten strains of the phototrophic Rhodobacter and Rhodovulum spp. belonging to the
-3 Proteobacteria, and Rsp. rubrum (which is included in the
-1 Proteobacteria). On the other hand, no amino acid substitution unique to the strains of Rps. palustris, B. japonicum and the phototrophic rhizobia was detected. These observations strongly suggest that the NifH proteins of Rps. palustris strains are highly related to those of species of Rhodobacter and Rhodovulum, as well as Rsp. rubrum, which belong to the
-3 and
-1 Proteobacteria, respectively. This is in contrast to the relationship inferred from the 16S rRNA analysis of Wong et al. (1994)
. Although both Rps. palustris and the phototrophic rhizobia are phototrophic species and belong to the same phylogenetic cluster in the 16S rRNA tree, their NifH proteins were apparently less related as regards their position in the NifH tree.

View larger version (69K):
[in this window]
[in a new window]
|
Fig. 3. CLUSTALX (1.81) multiple sequence alignment of the NifH showing the amino acid sites common between strains of Rps. palustris and B. japonicum, and other phototrophic purple non-sulfur bacteria. Dots represent the amino acids similar to the Rps. palustris ATCC 17001T sequence at the corresponding positions; only the amino acids that are different from the Rps. palustris ATCC 17001 partial NifH sequence are shown. The start of each sequence is at position 6 of the B. japonicum USDA 110 amino acid sequence (K01620).
|
|
The incongruence of these dendrograms as regards the position of Rps. palustris would be difficult to explain without invoking the idea of lateral gene transfer, as deduced from the radically aberrant position of these species in the NifH tree in comparison with the conventional 16S rRNA phylogeny. Lateral transfers of nifH among rhizobia belonging to the
-Proteobacteria (Haukka et al., 1998
; Eardly et al., 1992
), as well as transfer of nifH from a common donor in the
-Proteobacteria to Azoarcus sp., which belongs to the
-Proteobacteria (Hurek et al., 1997
), have already been described. The unusual position of Frankia and Anabaena within the Proteobacteria clade in the nifH tree of Normand & Bousquet (1989)
, supported by the nifK tree of Hirsch et al. (1995)
, strengthened the idea of lateral transfer of nifH. However, to our knowledge such occurrence has never been reported for the phototrophic bacteria. Although Hennecke et al. (1985)
undermined the case for nifH lateral transfer, our sequence and phylogenetic analyses provide strong support for the lateral transfer of this gene.
Evidence from the Rps. palustris CGA009 genome sequence for the lateral transfer of nifH
To further determine whether the nifH of Rps. palustris was acquired by lateral gene transfer, we examined the nitrogen fixation regulon of Rps. palustris CGA009 available from the genome sequence data. Three nitrogen fixation regulons (large clusters of genes involved in the same biosynthetic process) containing the two highly similar copies of nifH were found. The first regulon, located at approximately position 51880405215364 (Fig. 4a
), contains almost all structural genes encoding the nitrogen fixation enzymes. The second regulon includes the vanadium nitrogenase genes located at position 15155761527905, while the third system is located at 15904061596854 (numbering based on the completed genome sequence of Rps. palustris CGA009 available at NCBI). Two nifH-flanking genes from the first regulon, nifK and nifA, were chosen for phylogenetic comparison with that of nifH. In trees derived from the amino acid sequences encoded by these genes (Fig. 4b, c
), Rps. palustris was placed in the same cluster with B. japonicum; NifK and NifA of Rps. palustris are more related to those of B. japonicum than to those of Rba. sphaeroides. Assuming that the 16S rRNA phylogeny shows the correct evolutionary history of Rps. palustris, we propose that the nifH of this species came from a different ancestral origin than the other nitrogen fixation genes.

View larger version (32K):
[in this window]
[in a new window]
|
Fig. 4. (a) Map of the nitrogen fixation gene region drawn from the Rps. palustris CGA009 genome sequence data available at NCBI (NC_005296), showing the positions of nifK and nifA (indicated by arrows) flanking nifH. Genes shown in light grey are fix, while those in dark grey are nif. Unshaded arrows are ORFs with unknown functions. Orientations of the ORFs are indicated by arrows. (b, c) Phylogenetic trees based on amino acid sequences of nifK and nifA, respectively, showing the positions of Rps. palustris CGA009, B. japonicum USDA 110 and Rba. sphaeroides 2.4.1. Phylogenetic trees were constructed using the neighbour-joining method; bootstrap values derived from 1000 replicates are shown as percentages; bold lines indicate branches supported by bootstrap values greater than 90 %. Almost exactly the same tree topologies were produced using the maximum-likelihood and maximum-parsimony methods. The lengths of the horizontal lines are proportional to the evolutionary distances; the lengths of the vertical lines have no significance. Sequence accession numbers are given in parentheses, except for Rps. palustris CGA009 and Rba. sphaeroides, which are not available from the genome sequence data. The bar represents 0·1 changes per amino acid.
|
|
To further support the inferred evolutionary relationship, we compared the genome of Rps. palustris CGA009 with that of Rba. sphaeroides 2.4.1 and B. japonicum USDA 110 to attest the phylogenetic affiliation of Rps. palustris and B. japonicum, with an explicit assumption that the genome content of the species under study will show us its real affiliation. We thus presupposed that if these representative genes from Rps. palustris CGA009 are more similar to the corresponding genes in B. japonicum USDA 110, then nifH was laterally transferred to Rps. palustris. We randomly extracted a number of B. japonicum USDA 110 informational and housekeeping genes as well as ORF sequences from the database and used these as query sequences in searching for their corresponding sequence in the Rps. palustris CGA009 genome. The single most similar Rps. palustris CGA009 sequence for each B. japonicum USDA 110 ORF was then used as query sequence to search the Rba. sphaeroides 2.4.1 genome. The selected genes are located in different loci (data not shown) of their genomes, to give a good representation of the entire genome of these species. These genes/ORFs/molecules included the 16S, 23S and 5S rRNA, genes encoding or involved in the biosyntheses of cytochromes, ATP synthase, gyrases, polymerases, sigma factors, heat-shock proteins, nitrogen fixation and nitrate reduction, among others (Table 2
). The similarity values of each gene between Rps. palustris CGA009 and B. japonicum USDA 110, and between Rps. palustris CGA009 and Rba. sphaeroides 2.4.1, were plotted as shown in Fig. 5
. Most of the randomly selected genes of Rps. palustris CGA009 were more similar to that of B. japonicum USDA 110 than to the equivalent gene of Rba. sphaeroides 2.4.1, signifying that the genome of Rps. palustris is more closely related to that of B. japonicum. This comparison further highlighted the lateral transfer event, since the NifH of strains of Rps. palustris is highly related to other purple non-sulfur bacteria, but its genome content is more similar to that of B. japonicum.
View this table:
[in this window]
[in a new window]
|
Table 2. Genes used in comparing the genomes of Rps. palustris CGA009, Rba. sphaeroides 2.4.1 and B. japonicum USDA 110
Gene names and accession numbers are those of B. japonicum USDA 110 genes available in the database. Genome information on Rps. palustris CGA009 and Rba. sphaeroides 2.4.1 is available at http://spider.jgi-psf.org/JGI_microbial/html/
|
|
Concluding remarks
In this study, we made a direct comparison between the 16S rRNA and the NifH trees to infer the evolutionary history of Rps. palustris, with an explicit assumption that the 16S rRNA phylogeny reflects the organisms' phylogeny (Woese, 1987
). Based on this assumption, the strains of Rps. palustris possess a homologous copy of nifH but it came from an ancestral gene common among Rhodobacter, Rhodovulum and Rsp. rubrum, as demonstrated by the absolute conservation of nifH among strains of Rps. palustris and other phototrophic purple non-sulfur bacteria in the
-Proteobacteria regardless of their phylogenetic position in the 16S rRNA tree; the nifH gene was probably acquired by lateral transfer, as earlier considered by Ruvkum & Ausubel (1980)
and Normand & Bousquet (1989)
. Our study has shown that even the genes involved in a more conserved biological process than that of photosynthesis could be transferred between distantly related species. Lateral transfer of the photosynthesis genes in phototrophic bacteria has already been documented (Nagashima et al., 1997
; Igarashi et al., 2001
). This strengthens the idea that lateral transfer plays a major role in generating variation in the bacterial genome, thus contributing new phenotypes to the host species in the
-2 Proteobacteria, and in this case Rps. palustris. However, it cannot be ruled out that convergent evolution of nifH by selection pressures or neutral mutations might have restricted the divergence of nifH, since it was apparent that the topology of the NifH tree correlates with the physiological characteristics of the species the nifH genes of phototrophic purple non-sulfur bacteria belonging to different
-Proteobacteria lineages are closely related regardless of their phylogenetic affiliation based on 16S rDNA analysis, separated from the other phototrophic species like the Bradyrhizobium spp. IRBG 2, IRBG 228, IRBG 230 and BTAi 1. However, this alternative needs further studies to confirm the inferred relationship. As a marker, the nifH gene is a good molecule for differentiating the phototrophic Rps. palustris from the phototrophic rhizobia as well as from the non-phototrophic B. japonicum.
 |
ACKNOWLEDGEMENTS
|
---|
The authors would like to thank Rolando So of the International Rice Research Institute, Philippines, Douglas K. Jones of the National Rhizobium Germplasm Collection Center, ARS/USDA, and Dr Tohru Hamda of the Marine Biotechnology Institute, Kamaishi, Japan, for the photosynthetic strains. This work was supported by a Grant-in-Aid for Encouragement of Young Scientist (12760056) from the Ministry of Education, Culture, Sports, Science and Technology, Japan, to H. Kawasaki.
This paper represents a portion of the dissertation submitted by J. J. L. Cantera to Osaka University in partial fulfilment of the requirements for a PhD degree.
 |
REFERENCES
|
---|
Adachi, J. & Hasegawa, M. (1996). MOLPHY: Programs for Molecular Phylogeny, Version 2.3. Tokyo: Institute of Statistical Mathematics.
Altschul, S. F., Madden, T. L., Schaffer, A. A., Zhang, J., Zhang, Z., Miller, W. & Lipman, D. J. (1997). Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25, 33893402.[Abstract/Free Full Text]
Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A. & Struhl, K. (1995). Current Protocols in Molecular Biology, vol. I. New York: Wiley.
Cantera, J. J. L., Kawasaki, H. & Seki, T. (2002). Evolutionary relationship of phototrophic bacteria in the
-Proteobacteria based on farnesyl diphosphate synthase. Microbiology 148, 19231929.[Abstract/Free Full Text]
Eaglesham, A. R. J., Ellis, J. M., Evans, W. R., Fleischman, D. E., Hungria, M. & Hardy, R. W. F. (1990). The first photosynthetic N2-fixing Rhizobium: characteristics. In Nitrogen Fixation: Achievements and Objectives, pp. 805811. Edited by P. M. Gresshoff, L. E. Roth, G. Stacey & W. L. Newton. London: Chapman & Hall.
Eardly, B. D., Young, J. P. W. & Selander, R. K. (1992). Phylogenetic position of Rhizobium sp. strain Or 191, a symbiont of both Medicago sativa and Phaseolus vulgaris, based on partial sequences of the 16S rRNA and nifH genes. Appl Environ Microbiol 58, 18091815.[Abstract]
Fleischman, D. E., Evans, W. R. & Miller, I. M. (1995). Bacteriochlorophyll-containing Rhizobium species. Adv Photosynth 2, 123126.
Haukka, K., Lindström, 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, 419426.[Abstract/Free Full Text]
Hennecke, H., Kaluza, K., Thöny, B., Fuhrmann, M., Ludwig, W. & Stackebrandt, E. (1985). Concurrent evolution of nitrogenase genes and 16S rRNA in Rhizobium species and other nitrogen-fixing bacteria. Arch Microbiol 142, 342348.
Hirsch, A. M., McKhann, H. I., Reddy, A., Liao, J., Fang, Y. & Marshall, C. R. (1995). Assessing horizontal transfer of nifHDK genes in eubacteria: nucleotide sequence of nifK from Frankia strain HFPCc13. Mol Biol Evol 12, 1627.[Abstract]
Hurek, T., Egener, T. & Reinhold-Hurek, B. (1997). Divergence in nitrogenases of Azoarcus spp., Proteobacteria of the
subclass. J Bacteriol 179, 41724178.[Abstract]
Igarashi, N., Harada, J., Nagashima, S., Matsuura, K., Shimada, K. & Nagashima, K. V. P. (2001). Horizontal transfer of the photosynthesis gene cluster and operon rearrangement in purple bacteria. J Mol Evol 52, 333341.[CrossRef][Medline]
Imhoff, J. F., Trüper, H. G. & Pfennig, N. (1984). Rearrangement of the species and genera of the phototrophic "purple nonsulfur bacteria". Int J Syst Bacteriol 34, 340343.
Inui, M., Roh, J. H., Zahn, K. & Yukawa, H. (2001). Sequence analysis of the cryptic plasmid pMG101 from Rhodopseudomonas palustris and construction of stable cloning vectors. Appl Environ Microbiol 66, 5463.
Jeanmougin, F., Thompson, J. D., Gouy, M., Higgins, D. G. & Gibson, T. J. (1998). Multiple sequence alignment with Clustal X. Trends Biochem Sci 23, 403405.[CrossRef][Medline]
Kawasaki, H., Hoshino, Y. & Yamasato, K. (1993). Phylogenetic diversity of phototrophic purple non-sulfur bacteria in the Proteobacteria
group. FEMS Microbiol Lett 112, 6166.[CrossRef]
Ladha, J. K. & So, R. B. (1994). Numerical taxonomy of photosynthetic rhizobia nodulating Aeschynomene species. Int J Syst Bacteriol 44, 6273.
Lisdiyanti, P., Kawasaki, H., Seki, T., Yamada, Y., Uchimura, T. & Komagata, K. (2000). Systematic study of the genus Acetobacter with descriptions of Acetobacter indonesiensis sp. nov., Acetobacter tropicalis sp. nov., Acetobacter orleanensis (Hannenberg 1906) comb. nov., Acetobacter lovaniensis (Frateur 1950) comb. nov., and Acetobacter estunensis (Carr 1958) comb. nov. J Gen Appl Microbiol 46, 147165.[Medline]
Meyer, J., Kelley, B. C. & Vignais, P. M. (1978). Effect of light nitrogenase function and synthesis in Rhodopseudomonas capsulata. J Bacteriol 136, 201208.[Medline]
Nagashima, K. V. P., Hiraishi, A., Shimada, K. & Matsuura, K. (1997). Horizontal transfer of genes coding for the photosynthetic reaction centers of purple bacteria. J Mol Evol 45, 131136.[Medline]
Normand, P. & Bousquet, J. (1989). Phylogeny of nitrogenase sequences in Frankia and other nitrogen-fixing microorganisms. J Mol Evol 29, 436444.[Medline]
Ohkuma, M., Noda, S., Usami, R., Horikoshi, K. & Kudo, T. (1996). Diversity of nitrogen fixation genes in the symbiotic intestinal microflora of the termite Reticulitermes speratus. Appl Environ Microbiol 62, 27472752.[Abstract]
Ohkuma, M., Noda, S. & Kudo, T. (1999). Phylogenetic diversity of nitrogen fixation genes in the symbiotic microbial community in the gut of diverse termites. Appl Environ Microbiol 65, 49264934.[Abstract/Free Full Text]
Page, R. D. M. (1996). TREEVIEW: an application to display phylogenetic trees on personal computers. Comput Appl Biosci 12, 357358.[Medline]
Perrière, G. & Gouy, M. (1996). WWW-Query: an on-line retrieval system for biological sequence banks. Biochimie 78, 364369.[CrossRef][Medline]
Ruvkum, G. B. & Ausubel, F. M. (1980). Interspecies homology of nitrogenase genes. Proc Natl Acad Sci U S A 77, 191195.[Abstract]
Saitoh, N. & Nei, M. (1987). The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol 4, 406425.[Abstract]
Sanger, F., Nicklen, S. & Coulson, A. R. (1977). DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci U S A 74, 54635467.[Abstract]
Stackebrandt, E., Rainey, F. A. & Ward-Rainey, N. (1996). Anoxygenic phototrophy across the phylogenetic spectrum: current understanding and future perspectives. Arch Microbiol 166, 211223.[CrossRef][Medline]
Thompson, J. D., Gibson, T. J., Plewniak, F., Jeanmougin, F. & Higgins, D. G. (1997). The ClustalX Windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res 24, 48764882.[CrossRef]
Torok, I. & Kondorosi, A. (1981). Nucleotide sequence of the R. meliloti nitrogenase reductase (nifH) gene. Nucleic Acids Res 9, 57115723.[Abstract]
Ueda, T., Suga, Y., Yahiro, N. & Matsuguchi, T. (1995). Remarkable N2-fixing bacterial diversity detected in rice roots by molecular evolutionary analysis of nifH gene sequences. J Bacteriol 177, 14141417.[Abstract]
Widmer, R., Shaffer, B. T., Porteous, L. A. & Seidler, J. (1999). Analysis of nifH gene pool complexity in soil and litter at a Douglas fir forest site in the Oregon cascade mountain range. Appl Environ Microbiol 65, 374380.[Abstract/Free Full Text]
Woese, C. R. (1987). Bacterial evolution. Microbiol Rev 51, 221271.[Medline]
Woese, C. R., Stackebrandt, E., Weisburg, W. G. & 8 other authors (1984). Phylogeny of purple bacteria: the alpha subdivision. Syst Appl Microbiol 5, 315326.[Medline]
Wong, F. Y. K., Stackebrandt, E., Ladha, J. K., Fleischman, D. E., Date, R. A. & Fuerst, J. A. (1994). Phylogenetic analysis of Bradyrhizobium japonicum and photosynthetic stem-nodulating bacteria from Aeschynomene species grown in separated geographical regions. Appl Environ Microbiol 60, 940946.[Abstract]
Xiong, J., Inoue, K. & Bauer, C. E. (1998). Tracking molecular evolution of photosynthesis by characterization of a major photosynthesis gene cluster from Heliobacillus mobilis. Proc Natl Acad Sci U S A 95, 1485114856.[Abstract/Free Full Text]
Young, J. P. W. (1992). Phylogenetic classification of nitrogen-fixing organisms. In Biological Nitrogen Fixation, pp. 4386. Edited by G. Stacey, R. H. Burris & H. J. Evans. New York: Chapman & Hall.
Young, J. P., Downer, H. L. & Eardly, B. D. (1991). Phylogeny of phototrophic Rhizobium strain BTAi1 by polymerase chain reaction-based sequencing of a 16S rRNA segment. J Bacteriol 173, 22712277.[Medline]
Received 25 November 2003;
revised 6 April 2004;
accepted 19 April 2004.
Copyright © 2004 Society for General Microbiology.