BackgroundSwarming is a multicellular phenomenom characterized by the coordinated and rapid movement of bacteria across semisolid surfaces. In Sinorhizobium meliloti this type of motility has been described in a fadD mutant. To gain insights into the mechanisms underlying the process of swarming in rhizobia, we compared the transcriptome of a S. meliloti fadD mutant grown under swarming inducing conditions (semisolid medium) to those of cells grown under non-swarming conditions (broth and solid medium).ResultsMore than a thousand genes were identified as differentially expressed in response to growth on agar surfaces including genes for several metabolic activities, iron uptake, chemotaxis, motility and stress-related genes. Under swarming-specific conditions, the most remarkable response was the up-regulation of iron-related genes. We demonstrate that the pSymA plasmid and specifically genes required for the biosynthesis of the siderophore rhizobactin 1021 are essential for swarming of a S. meliloti wild-type strain but not in a fadD mutant. Moreover, high iron conditions inhibit swarming of the wild-type strain but not in mutants lacking either the iron limitation response regulator RirA or FadD.ConclusionsThe present work represents the first transcriptomic study of rhizobium growth on surfaces including swarming inducing conditions. The results have revealed major changes in the physiology of S. meliloti cells grown on a surface relative to liquid cultures. Moreover, analysis of genes responding to swarming inducing conditions led to the demonstration that iron and genes involved in rhizobactin 1021 synthesis play a role in the surface motility shown by S. meliloti which can be circumvented in a fadD mutant. This work opens a way to the identification of new traits and regulatory networks involved in swarming by rhizobia.

Swarming is a mode of translocation dependent on flagellar activity that allows bacteria to move rapidly across surfaces. In several bacteria, swarming is a phenotype regulated by quorum sensing. It has been reported that the swarming ability of the soil bacterium Sinorhizobium meliloti Rm2011 requires a functional ExpR/Sin quorum-sensing system. However, our previous published results demonstrate that strains Rm1021 and Rm2011, both known to have a disrupted copy of expR, are able to swarm on semisolid minimal medium. In order to clarify these contradictory results, the role played by the LuxR-type regulator ExpR has been reexamined. Results obtained in this work revealed that S. meliloti can move over semisolid surfaces using at least two different types of motility. One type is flagellum-independent surface spreading or sliding, which is positively influenced by a functional expR gene mainly through the production of exopolysaccharide II (EPS II). To a lesser extent, EPS II-deficient strains can also slide on surfaces by a mechanism that is at least dependent on the siderophore rhizobactin 1021. The second type of surface translocation shown by S. meliloti is swarming, which is greatly dependent on flagella and rhizobactin 1021 but does not require ExpR. We have extended our study to demonstrate that the production of normal amounts of succinoglycan (EPS I) does not play a relevant role in surface translocation but that its overproduction facilitates both swarming and sliding motilities.

Sinorhizobium meliloti can exhibit diverse modes of surface translocation whose manifestation depends on the strain. The mechanisms involved and the role played by the different modes of surface motility in the establishment of symbiosis are largely unknown. In this work, we have characterized the surface motility shown by two S. meliloti reference strains (Rm1021 and GR4) under more permissive conditions for surface spreading and analyzed the symbiotic properties of two flagella-less S. meliloti mutants with different behavior on surfaces. The use of Noble agar in semisolid minimal medium induces surface motility in GR4, a strain described so far as non-motile on surfaces. The motility exhibited by GR4 is swarming as revealed by the non-motile phenotype of the flagella-less flaAB mutant. Intriguingly, a flgK mutation which also abolishes flagella production, triggers surface translocation in GR4 through an as yet unknown mechanism. In contrast to GR4, Rm1021 moves over surfaces using mostly a flagellaindependent motility which is highly reliant on siderophore rhizobactin 1021 production. Surprisingly, this motility is absent in a flagella-less flgE mutant. In addition, we found that fadD loss-of-function, known to promote surface motility in S. meliloti, exerts different effects on the two reference strains: while fadD inactivation promotes a flagella-independent type of motility in GR4, the same mutation interferes with the surface translocation exhibited by the Rm1021 flaAB mutant. The symbiotic phenotypes shown by GR4flaAB and GR4flgK, non-flagellated mutants with opposite surface motility behavior, demonstrate that flagella-dependent motility positively influences competitiveness for nodule occupation, but is not crucial for optimal infectivity.