Transcriptomic approach for the identification of genes and signals playing a role in swarming motility of sinorhizobium meliloticonnection with biofilm formation and symbiosis

Resumen

Soil bacteria known as rhizobia establish a mutualistic symbiotic association with leguminous plants. The success of the interaction depends on the mechanisms that Rhizobium uses to colonize, invade and establish the chronic infection in its host. Swarming and biofilm formation are two surface-associated bacterial processes involved in the infectiveness and invasiveness ability of pathogenic microorganisms. Swarming is a form of bacterial translocation which involves a process of cellular differentiation and is characterized by the rapid and coordinated migration of a population over a semisolid surface. Biofilms consist of groups of bacteria attached to surfaces and encased in a self-produced polymeric matrix. The few studies that have been performed concerning these multicellular processes in beneficial bacteria, such as Rhizobium, suggest that, as in pathogenic microorganisms, essential components of swarming motility and/or biofilm formation, or factors that are expressed during both processes are likely to be important in the interaction with their host. It is known that the lack of the fadD gene (encoding a long-chain fatty acyl-coenzyme A ligase) in Sinorhizobium meliloti, the endosymbiont of alfalfa, results in multicellular swarming behavior. Additionally, this mutation leads to symbiotic defects. The fadD mutant is less infective and less competitive for nodulation than the wild-type strain GR4 (Soto et al., 2002). These results suggest that swarming components could determine the infectiveness capacity of S. meliloti. With the aim of elucidating new bacterial genes which may play a role in the interaction of Rhizobium with the plant, we sought to identify genetic determinants of swarming in the model bacteria S. meliloti using a transcriptomic approach. Some of the genes identified in this way were furthered characterized by analyzing their importance in biofilm formation and for the establishment of symbiosis with alfalfa plants. As a first approach to identify the mechanisms underlying the process of swarming in Rhizobium, in chapter 1 of this study, we compared the transcriptome of a S. meliloti fadD mutant grown under swarming inducing conditions (semisolid Minimal Medium (MM)) with cells grown under non-swarming conditions (MM broth and solid MM). This analysis revealed that the genetic expression of cells growing on a surface differed from that of cells growing in broth. More than a thousand genes were identified to be differentially expressed under these conditions including genes required for carbon, proteins and energy metabolism as well as genes involved in macromolecule synthesis, motility, chemotaxis and stress response. These results revealed that important changes take place in the cellular physiology of S. meliloti in response to growth on surfaces. Additionally, it was observed that the response to swarming specific conditions is mainly characterized by the induction of iron metabolism and uptake related genes. We have demonstrated that iron levels in the MM, the pSymA plasmid, and specifically the genes required for the biosynthesis of the siderophore rhizobactin 1021 (Rhb1021), are essential for swarming by a S. meliloti wild-type strain but not in a fadD mutant. The presence of high concentrations of iron inhibits swarming motility in the wild-type but does not affect swarming in mutants lacking either the iron limitation response regulator RirA or FadD. These results have been published in 2010 in BMC Genomics (Nogales et al., 2010). Studies carried out on several bacteria have revealed that swarming motility and biofilm formation are closely related multicellular processes. In bacteria such as Pseudomonas aeruginosa, these social behaviours seem to be inversely regulated through a common genetic pathway (Caiazza et al., 2007). It has been described that S. meliloti is capable of forming biofilms in biotic and abiotic surfaces (Fujishige et al., 2006). Flagella, EPSII and the oligosaccharide component of Nod factors have been identified as important determinants of biofilm formation in this bacterium (Fujishige et al., 2006; Fujishige et al., 2008; Rinaudi and González, 2009). In order to identify the existence of a possible connection between this phenomenon and swarming motility in S. meliloti, in chapter 2 we analysed the biofilm formation ability of strains showing a different swarming motility phenotype. There is evidence suggesting that swarming regulation in the commonly used laboratory strain Rm1021, and the strain isolated from Granada¿s soils, GR4, occurs differently. Even though a fadD mutation provokes swarming in both strains, the conditions that trigger swarming in Rm1021 do not elicit swarming of GR4 (Nogales et al., 2010; Soto et al., 2002). Therefore, we analyzed the biofilm formation ability of the wild-type strains Rm1021 and GR4, as well as that of the fadD mutants of each of these strains. Additionally, we evaluated the biofilm formation phenotype of two other mutants from Rm1021 that show an altered swarming pattern. We studied the biofilm formation phenotype of a mutant in rirA, which constitutively produces Rhb1021 and exhibits swarming even in presence of high concentrations of iron in the medium, and in a mutant lacking rhbA, which is unable to produce Rhb1021 and is incapable of moving over surfaces. As iron showed to be an important signalling molecule for swarming, we have also studied the relevance of the concentration of this compound for the establishment biofilms by S. meliloti. The ability of the different strains to develop biofilms was evaluated by using abiotic surfaces such as polyvinyl chloride (PVC) and glass, as well as biotic surfaces such as the surface of alfalfa roots. Despite the fact that no differences were observed in biofilm formation between Rm1021 and GR4 on PVC, GR4 showed a better biofilm formation capacity than Rm1021 on glass surfaces and alfalfa roots. In general S. meliloti seems to be very inefficient in developing biofilms on a PVC surface. We found that iron levels in the media impact the biofilm structure of S. meliloti wild-type strains. While low concentrations of iron led to the formation of a biofilm with a homogenous and flat structure, the presence of high concentrations of this compound induced the formation of a structured biofilm composed by towers and valleys in both Rm1021 and GR4. Independent of the iron concentration in the media and the surface, the mutants in fadD, rhbA or rirA showed an altered biofilm formation pattern with respect to the parental strains. FadD, RirA and the iron concentration in the media seem to inversely regulate swarming and biofilm formation. Altogether, these results show that genes and signals (iron in the media) that participate in swarming control also play an essential role in S. meliloti biofilm formation. Moreover, mutants in these genes were less efficient in the colonization of alfalfa roots, which suggest that the identification of determinants of swarming could be a useful strategy to identify bacterial genes that are important in the colonization of plants. Mutations in rhb genes abolish Rhb1021 production and surface spreading in S. meliloti Rm1021 (Nogales et al., 2012; Nogales et al., 2010). On the other hand, it has been observed that an rhbA mutant is unable to form thick biofilms. These results suggest that Rhb1021 production is a key factor for S. meliloti Rm1021 surface motility and biofilm formation. Chapter 3 of this study addresses the role that could be played by Rhb1021 on both processes. In order to study this, Rhb1021 was isolated from the supernatant of cells of a rirA mutant which presents a deregulated production of the siderophore. The drop collapsing test performed with purified Rhb1021 revealed that this siderophore shows surfactant activity that could impact both surface motility and biofilm formation. In order to confirm that the function of Rhb1021 on both surface-associated processes is linked to the surfactant activity of the molecule, we obtained and characterized mutants in two genes (rhbD and rhbG) that according to in silico predictions could be responsible for incorporating the lipid tail that forms part of the Rhb1021 molecule and confers surfactant properties. The rhbD mutant proved to be incapable of producing the siderophore and its behavior on surfaces (motility and biofilm formation) was similar to that shown by the rhbA mutant. Our results concerning RhbD agree with the function designated to this compound by other authors. In contrast, data derived from the characterization of the rhbG mutant clearly demonstrate that in contrast to what has been reported by other authors, (Lynch et al. (2001) and Challis (2005)) SMa2339 (rhbG) is not responsible for incorporating the lipid tail into the Rhb1021 molecule. Thus, to date, the function of RhbG remains unknown. In addition, unlike the rest of genes belonging to the rhbA-F operon, rhbG is not essential for either siderophore synthesis or surface translocation of S. meliloti, although it influences surface associated phenotypes. Loss-of function of rhbG promotes flagella-independent surface translocation and interferes with normal biofilm development with similar effects as those displayed by the rhbA and rhbD mutants. Liquid chromatography and mass spectrophotometry analysis from the rhbG culture supernatant, demonstrated that this mutation leads to an accumulation of (2E)-N-[3-(acetylamino)propyl]-N-hydroxydec-2-enamide and also to a decreased production of Rhb1021. Finally, by using a transcriptomic approach we investigated the possibility that Rhb1021 may act as a signaling molecule. In order to do this, the global gene expression pattern of 1021rhbD cells was compared with that of Rm1021 cells grown on semisolid surfaces. Even though relevant changes in genetic expression were not detected, this analysis revealed that either Rhb1021 or some of its biosynthetic intermediates are required for maximal expression of siderophore synthesis and uptake genes. The final objective of this Thesis was to analyze the influence on the establishment of symbiosis of the different genes identified throughout this study to be involved in the control of surface motility and biofilm formation capacity of S. meliloti (rhb, rhtA, and rirA). Although it has been described that none of these genes are essential for the development of N-fixing nodules (Chao et al., 2005; Lynch et al., 2001; Viguier et al., 2005), their role in multicellular phenomena associated with surfaces may be indicative of their function at early stages of the interaction with the plant. Our results indicate that as observed for mutants rhbA and rirA (chapter 2), the rhtA mutant which is defective in siderophore Rhb1021 utilization, also shows a significant deficiency in colonizing alfalfa roots. Moreover, the study of the infective and competitive capacity of each of the strains indicated that they show some type of alteration in their symbiotic properties. These defects were more pronounced for mutant strains displaying increased surface motility (rhtA and rirA) than for those impaired in surface translocation (rhbA). Especially interesting were the results obtained with the rirA mutant which in addition to be severely affected in infectivity (delayed nodule formation and reduced number of nodules) and competiveness, also showed deficiencies in its capacity to fix nitrogen. The results obtained in this PhD Thesis study validate the strategy used which consists of characterizing the molecular bases governing surface spreading motility of S. meliloti, in order to identify new bacterial genes important for the colonization of plants.