Importancia del segundo mensajero C-DI-GMP en la simbiosis rizobio-leguminosa

Resumen

Biological Nitrogen Fixation (BNF) is a prokaryotic-exclusive, fundamental process by which the atmospheric molecular nitrogen is reduced to ammonia and made accessible to other organisms, particularly plants. Rhizobia are able to carry out BNF after establishing a symbiotic interaction with legume hosts. Root nodules are the visible expression of this symbiosis, and represent plant organs where specialized forms of the bacteria reduce molecular nitrogen at the expense of carbohydrates provided by the plant. A permanent exchange of signals between plant and bacteria is necessary for the establishment of an efficient nitrogen-fixing symbiosis. These signals must be integrated by the bacteria to modify their lifestyles and coordinate the expression of essential determinants for colonization of the roots and infection of the nodules. The second messenger cyclic diguanylate (c-di-GMP) is a key molecule for the transition between different lifestyles in bacteria. Since rhizobia undergo dramatic changes during their life cycles, from free-living to intimate interaction with their hosts, and back to free life, it seems reasonable that the second messenger c-di-GMP could play an important role in coordination of those changes. The large number of c-di-GMP metabolizing (for synthesis and degradation) enzymes encoded in plant-associated bacterial genomes suggests likely functional redundancies, which usually hinder genetic approaches. Therefore, in a first approach we evaluated the effects of artificial increases of intracellular c-di-GMP levels in diverse plant-interacting bacteria, including the phytopathogenic Pseudomonas syringae pv. tomato DC3000 and several mutualistic rhizobia, Rhizobium etli CFN42, R. leguminosarum bv. viciae UPM791 and Sinorhizobium meliloti 8530. In all of them, elevated c-di-GMP contents resulted in reduction of motility, increased production of exopolysaccharides and enhanced biofilm formation. Regarding the interaction with the host legumes, we found some differences depending on the type of interaction and the symbiotic stage concerned. In the three symbiotic relationships evaluated, high c-di-GMP levels seem to favour the early stages since enhanced attachment to plant roots, but the symbiotic efficiency was negatively affected in some cases. The most affected association was the R. etli-P. vulgaris (common bean), and bean plants inoculated with elevated c-di-GMP R. etli showed decreased number of nodules, less aerial biomass and reduced nitrogen contents. The elevation of c-di-GMP intracellular contents was achieved by the overexpression of the diguanylate cyclase PleD* from Caulobacter crescentus from the pJBpleD* plasmid (Pérez-Mendoza et al., 2014). However, the low stability of this pJBpleD* was a serious experimental limitation when a selective pressure for the plasmid could not be applied, particularly in planta. This prompted us to construct a set of tools for the stable integration of the pleD* gene into the genomes of plant-interacting bacteria. The suitability of the new constructions, based on the Tn7 transposon properties, was demonstrated in all the above plant-interacting bacteria. The mini-Tn7pleD* delivery vehicles allow increases of the intracellular c-di-GMP levels in a similar way as vector pJBpleD*. However, mini-Tn7 constructs resulted far more stable in the absence of antibiotic pressure than the plasmid-based pleD* constructs. This high stability ensures experimental homogeneity in time and space with regard to enhancing c-di-GMP intracellular levels in bacteria of interest. Furthermore, we have also implemented a system based on the repressor/inductor LacIq/IPTG, to modulate pleD* expression and intracellular c-di-GMP rises on demand. EPS overproduction was one of the strongest and most visible responses to elevated c-di-GMP in all bacteria. At the time this thesis was started, cellulose was the only EPS known to be activated by c-di-GMP in rhizobia. We demonstrated that R. etli produces at least two c-di-GMP regulated EPS, cellulose and the cellulose-like EPS Mixed-Linkage β-Glucan (MLG), previously reported in S. meliloti (Pérez-Mendoza et al., 2015). With the help of bacterial mutants, we concluded that both EPS promote red stained colonies and fluorescence in medium added with the dyes Congo red (CR) and calcofluor, respectively, at high intracellular c-di-GMP levels. However under our experimental conditions, cellulose was the main EPS involved in R. etli adhesion to abiotic (polystyrene) and biotic surfaces (bean root), as well as biofilm formation. Concerning the interaction with the host plant, overproduction of cellulose and MLG were not found responsible for the reduced symbiotic efficiency observed in plants inoculated with high c-di-GMP strains, since mutants lacking both EPS still formed impaired symbiosis under these conditions. Microscopy studies indicated that nodules formed by high c-di-GMP bacteria have an altered infection pattern, with a smaller number of infected cells, which could explain the reduced symbiotic efficiency. Far from restoring this defect, mutants unable to produce both cellulose and MLG displayed an even more drastic phenotype, with plant cells distributed in sectors within the nodule, indicative of altered infection thread growth or ramification. This result suggests putative roles of cellulose and/or MLG during nodule infection. The impact of c-di-GMP on the transcriptome of R. etli was also investigated with the use of microarrays. Under the conditions tested, high c-di-GMP determined mainly downregulation of a relatively small number of genes, which did not allow reaching clear conclusions about additional functions affected by high c-di-GMP levels. In plant-interacting bacterial genomes, the number of genes encoding putative c-di-GMP metabolizing enzymes (diguanylate cyclases and phosphodiesterases) is rather high. This contrasts with the reduced number of predicted c-di-GMP binding proteins, which suggests that most c-di-GMP effectors are yet to be identified. In order to identify new c-di-GMP binding proteins in R. etli, we followed a novel chemical proteomics approach which combines affinity chromatography with a c-di-GMP analogue (2´- AHC-c-di-GMP) and subsequent mass spectrometry (Düvel et al., 2012). We identified 66 putative c-di-GMP binding proteins. Following a second screening based on putative protein function and domain composition, nine of those proteins were overexpressed in E. coli and tested for binding to a c-di-GMP fluorescent analogue (2´-Fluo-AHC-c-di-GMP). Dgt and GlmU proteins both gave positive signals, although they do not harbour any known c-di-GMP binding motif. To corroborate binding, we used other methodologies more sensitive and accurate, such as fluorescence polarization (FP). However, after protein purification Dgt did not show any interaction with the florescent analogue. GlmU interacted specifically with the 2´-Fluo-AHC-c-di-GMP but not with other nucleotide fluorescent analogues. The KD value of this interaction was determined to be 2.37 ± 0.32 µM. However, a competition assay with free, unmodified c-di-GMP, did not result in significant fluorescent signal reduction. Finally, we performed an isothermal titration calorimetry (ITC) assay for GlmU, which was also inconclusive. Therefore, despite the many efforts, we were not able to validate GlmU as a c-di-GMP binding protein. This thesis represents one of the first dedicated studies on c-di-GMP regulation in rhizobia. Whereas other laboratories have followed mainly genetic approaches (Wang et al. 2010; Gao et al. 2014; Schäper et al. 2016), here we have used an alternative approach based on the artificial alteration of c-di-GMP contents and the observation of phenotypical changes, to discern important processes regulated by this second messenger. Also different to other groups, we have evidenced that c-di-GMP regulates important properties of rhizobia for the symbiotic interaction with legumes.