It is logical to assume that rhizobia, in addition to being able to catabolize a wide variety of nutrient sources, would benefit from the ability to detect and move towards various nutrients, as well as toward plant roots. Several reports have indicated that chemotactic ability and/or motility confers a competitive advantage on rhizobia (Ames, Bergman 1981; Bauer, Caetano-Anolles 1990), and there is also evidence that rhizobia are chemotactic towards the same compounds which induce nod gene expression (Dharmatilake, Bauer 1992). Because rhizobia are able to use such a variety of different carbon and nitrogen sources, we set out to look for genes homologous to MCP type chemoreceptors, with the aim of finding receptors specific for compounds whose catabolism was plasmid-encoded. A family of at least 17 genes, and probably over 20, exhibiting homology to MCP genes was detected in R. leguminosarum VF39 (Yost et al. 1998). The complete sequences of three genes, mcpB, mcpC and mcpD, were reported by Yost et al. (1998) as was the fact that mcpB was required for a generalized chemotactic response, and that both mcpC (which is carried on plasmid pRleVF39f) and mcpB were important for competitive nodulation. Further studies were carried out on the mcpG gene, which is located on pRleVF39d, the pSym of VF39. The mcpG gene encodes a typical MCP; it is flanked by mono-oxygenase genes and is located next to an alanine transporter. Mutants in mcpG exhibit no obvious chemotactic defects. Homologs of mcpG are plasmid encoded in all strains of both R. leguminosarum and R. etli that we have examined, but are absent in S. meliloti and R. tropici. Gene fusions have been constructed to a number of our MCP genes and have been used to show that there is no obvious regulation during the free-living growth of VF39, but that MCP genes are down-regulated during nodulation.
While it is as yet unclear that all of the MCP gene homologs present in VF39 code for genuine MCPs, the four which we have sequenced completely, and the other three for which extensive sequencing has been done all seem to code for canonical MCPs possessing methylation sites and transmembrane domains. However, a search of the S. meliloti 1021 genome sequence reveals only nine genes potentially coding for MCP-like receptors, and none of these is particularly similar to any of the VF39 MCPs. This suggests that there may be important differences between the two genera in the way environmental signals are sensed. Further evidence for this comes from our observation that R. leguminosarum VF39 exhibits swimming patterns (i.e. tumbling) similar to those of E. coli, whereas S. meliloti has unidirectional flagella and does not tumble (Greek et al. 1995). We are interested in further elucidating these differences, and have isolated a number of chemotactic and motility mutants of VF39 which are currently being characterized.
Was this article helpful?