Alexandre G et al. (2000) J. Bacteriol. 182, 6042-6048 Cevallos M et al. (1996) J. Bacteriol. 178, 1646-1654 Dunn et al. (1996) J. Bacteriol. 178, 5960-5970 Ekblad A, Huss-Danell K (1995) New Phytol. 131, 453-459 Taylor BL et al. (1999) Annu. Rev. Microbiol. 53, 103-128 Yost CK etal. (1996) Microbiol. 144, 1945-1956 Zhulin et al. (1996) J. Bacteriol. 178, 5199-5202


M.F. Hynes1, C.K. Yost1, I.J. Oresnik2, A. Garcia de los Santos3, S.R.D. Clark1, J.E. Macintosh1

'Dept. of Biological Sciences, University of Calgary, 2500 University Dr. NW, Calgary

AB, T2N 1A9 Canada 2Dept. of Microbiology, University of Manitoba, Winnipeg MB, R3T 2N2 Canada 3CIFN, UNAM, Cuernavaca, Morelos, Mexico

1. Introduction

Although much effort has been devoted to the study of the interaction of rhizobia with their legume host plants, there are still many things about the life of rhizobia in the soil, and about the early interactions of the bacteria with the plant, which remain unclear. Specifically, we have little information about what carbon, nitrogen and other nutrient sources are important to the survival and proliferation of the bacteria during free-living growth, and during early stages of legume infection, and we also do not understand the role of motility and chemotaxis in the infection process and in the ability of the bacteria to thrive in legume and non-legume rhizospheres.

2. Plasmid-Encoded Catabolic Genes

Up to 40% of the genome of Rhizobium and Sinorhizobium species can consist of plasmids. The number and size of plasmids varies significantly from one strain to another, so we predicted that plasmid encoded properties would have a significant effect on interstrain differences in behavior. For this reason we have had a long-term interest in determining the function of cryptic plasmids in the rhizobia, and mapping plasmid encoded genes and investigating their importance in symbiosis and in inter-strain competition. The approach taken was to generate plasmid-cured derivatives using a Tn5 derivative carrying the sacB gene which is lethal to gram-negative bacteria in the presence of sucrose (Hynes et al. 1989). This strategy has allowed the isolation of plasmid-cured derivatives of strains of R. leguminosarum biovars viciae and trifolii (Hynes et al. 1989; Hynes, McGregor 1990; Baldani et al. 1992) and Sinorhizobium meliloti (Hynes et al. 1989; Oresnik et al. 2000). Examination of the properties of cured strains has resulted in the identification of a number of plasmid encoded phenotypes, the majority of which appear to involve catabolic genes. Initially, we screened for loss of catabolic activity in cured strains by growing the bacteria in minimal media supplemented with a variety of readily available and inexpensive carbon sources, but more recently, we have adopted more rapid screens based on BioLog™ plates, which simultaneously screen, through a colorimetric test, for use of a large number of carbon, nitrogen, sulfur and phosphorus sources. This latter approach has resulted in identification of a number of novel plasmid encoded catabolic phenotypes, some of which are summarized in the tables below. It should be mentioned that in some cases confirmation of a phenotype in a BioLog test is rendered difficult by the fact that the wild-type (wt) strain does not grow well enough on the C or N source as sole source in minimal media to distinguish it from the cured strain. Only results that seemed very clear, and which in many cases have been confirmed, are listed in the tables. Some of these results, using only BioLog GN plates, have previously been reported (Oresnik et al. 2000).

Identification of plasmid encoded catabolic phenotypes associated with plasmids is only one step in determining the importance of plasmids in soil ecology. Ideally, genes should be cloned, characterized, and mutated, and the mutants compared to wt for survival and proliferation in various environments as well as for competition for nodulation. We have initiated such experiments for a number of carbon catabolic genes, and are continuing to do so for new genes isolated from

R. leguminosarum and S. meliloti (Tables 1 and 2). We have already established a role for rhamnose catabolism and uptake genes in competition for clover nodulation (Oresnik et al. 1998) while at the same time showing that sorbitol and adonitol catabolism genes appear to be unimportant.

Table 1. Putative catabolic genes on plasmids in R. leguminosarum strain VF39.

Plasmid Approximate size Carbon catabolism genes on plasmid


220 kb

gluconate, malonate, glucuronate


300 kb

glycerol, adonitol*, melibiose


350 kb

alanine, adonitol*, trigonelline, hydroxyproline


500 kb

rhamnose, sorbitol, histidine, serine


600 kb

erythritol, ornithine, citrate, proline

* Only strains cured of both pRleVF39c and pRleVF39d are unable to grow on adonitol, suggesting there are genes for its use on both plasmids

Table 2. C and N utilization genes on plasmid pRmel 021a (pSymA megaplasmid of S. meliloti), determined by comparison of wt with cured strain on BioLog GN and PM1, PM2 and PM3 plates.

Carbon utilization genes: lyxose, psicose, acetate, acetoacetate, propionate, gluconate, arabinose, arbutin, sorbose, allose, lactate, xylitol, lactate, serine, alanine, gamma aminobutyrate, arginine, leucine, valine

Nitrogen utilization genes: adenosine, xanthine, cytidine, guanine, thymidine alanine, serine, valine, arginine, ammonium, various dipeptides

An alternative approach to looking for catabolic genes important in the soil and rhizosphere is to look for genes induced by root exudates, or by specific compounds known to be abundant in the rhizosphere and plant root. We have generated several banks of mutants made with different promoter probe transposons in order to screen for inducible genes. One such mutant library was made with Tn5B22 (Simon et al. 1989), which generates lacZ fusions. This library was examined by plating on various media with X-gal, for transposons located in genes, which were induced by a variety of substrates, including arabinose. One arabinose inducible fusion turned out to be located in the malate synthase gene of R. leguminosarum VF39. This gene is homologous to the E. coli malate synthase G gene, and in addition to being strongly induced by arabinose, is inducible by glycolate and glyoxylate. It appears to be required for symbiotic nitrogen fixation on peas.

Was this article helpful?

0 0

Post a comment