The current classification of the rhizobia (root-nodule symbionts) assigns them to six genera. It is strongly influenced by the small subunit (16S, SSU) rRNA molecular phylogeny, but such single-gene phylogenies may not reflect the evolution of the genome as a whole. To test this, parts of the atpD and recA genes have been sequenced for 25 type strains within the alpha-Proteobacteria, representing species in Rhizobium, Sinorhizobium, Mesorhizobium, Bradyrhizobium, Azorhizobium, Agrobacterium, Phyllobacterium, Mycoplana and Brevundimonas. The current genera Sinorhizobium and Mesorhizobium are well supported by these genes, each forming a distinct phylogenetic clade with unequivocal bootstrap support. There is good support for a Rhizobium clade that includes Agrobacterium tumefaciens, and the very close relationship between Agrobacterium rhizogenes and Rhizobium tropici is confirmed. There is evidence for recombination within the genera Mesorhizobium and Sinorhizobium, but the congruence of the phylogenies at higher levels indicates that the genera are genetically isolated. rRNA provides a reliable distinction between genera, but genetic relationships within a genus may be disturbed by recombination.


Article metrics loading...

Loading full text...

Full text loading...



vol. , part 6, pp. 2037-2048

The program SimPlot was used to look for possible reasons for the unstable behaviour of some sequences. Pairwise similarities were plotted for a 100 bp sliding window along the length of each gene, and bootscanning analyses were made to highlight changes in support of different tree topologies. Sequences were grouped by genus, with and combined.

The similarity plots for show an anomalous region of high divergence between and sequences centred around 180 bp into the sequence. Only in the central 100 bp of the sequence does the bootscan show strong support for the partition of with (i.e. for as the deepest branch as shown in Fig. 2). Elsewhere, or also gain support. These observations suggest that the position of these taxa may be influenced by the specific choice of as the outgroup, and indeed, when is removed, the bootscan shows unambiguously partitioned with in the 3´ half of the sequence, while the 5´ half tends to group with . The corresponding neighbour-joining phylogeny places next to while moves into the – clade, though support remains low. The position of the other taxa is unaltered. The apparently chimeric nature of the and possibly sequences was explored in a four-taxon bootscan with , , and . While the bootscan strongly indicated the pairing of with in the first half and in the second half of the sequence, an examination of the distribution of the 35 phylogenetically informative sites did not reveal a clear breakpoint indicative of recombination.

A clearer result was obtained for the SSU sequence of in a comparison with , A and ( Supplementary Fig. 1). The anomaly in these sequences has been noted previously (Eardly , 1996; Young & Haukka, 1996), but SimPlot reveals more detail. The sequence is most similar to that of throughout the 5´ half of the sequence, but this similarity suddenly plummets between positions 900 and 1100 ( Supplementary Fig. 1a). In the centre of this region is a cluster of eight phylogenetically informative sites, seven of which support the pairing of with ( Supplementary Fig. 1b) These are the only informative sites in the whole sequence that do not support the pairing of with . Unfortunately, there are no other informative sites nearby, so it is not possible to define the size of this potentially recombinant region with any precision: it could be as little as 50 bp but might be almost 500 bp.


Comparison of the SSU sequence of with those of (A), (B) and (C). (a) Similarity within a 200 bp sliding window (20 bp steps) along the alignment. (b) Distribution of phylogenetically informative sites supporting the four-taxon trees in which is paired with (A) or with (B). No sites support the third alternative. (c) Bootscan showing percentages of 100 bootstrap replicates that support each of the three trees, for a 200 bp sliding window.

Corresponding 16S rRNA gene segments in and yield discordant phylogenies. , 69–74. Diversity and phylogeny of rhizobia. , 87–94.

The sequences submitted to the EMBL database are shown in FastA format. The EMBL accession numbers for the gene sequences are AJ294386–AJ294409 and for are AJ294363–AJ294385. For phylogenetic analyses, the sequences were aligned as indicated by the gaps. The whole of the alignment was used while, for , positions 60–500 were used because this region was available for all species.


Most Cited This Month

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error