Unravelling Micromonospora interactions with its host plant and the associated microbioma

Abstract

The European Union highly depends on soy imports (> 70%) as a protein source since local production barely covers 5% of its internal demand. Thus, it is necessary to explore alternative sources to reduce this dependency. Among legumes, Lupinus angustifolius is an important alternative given its high protein value and use for animal and human nutrition. This legume is a native plant of Europe, well adapted to the climatic conditions of many countries. It also thrives in poor soils due to its capacity to fix nitrogen. Plant adaptation may be partly due to the microorganisms associated with its roots, providing stability and resilience, in addition to plant growth promoting molecules and nutrients. Plant-associated microbial communities are influenced by several factors such as host genotype/species, soil type, plant compartment and climatic season, among others. Separating these factors to understand which are the most influential in the association of microorganisms to plants is a very complex task as they do not occur independently. In the first chapter of this doctoral thesis, this topic was addressed by studying seasonal and geographical variations in the soil microbiota, and characterizing the microbiome associated with the plant Lupinus angustifolius under different cultivation conditions using an independent culture methodology. The results of the soil samples analysed suggest that the difference in the microbial community composition observed between the two sampling locations, Cabrerizos and Salamanca, was partly due to differences in soil conditions. None of the communities analysed (bacterial and fungal) showed differences in alpha diversity (Shannon index) between the climatic seasons in which the samples were collected. Beta diversity (Bray-Curtis-based principal coordinate analysis) for both microbial communities separated the samples into two groups according to soil type. In the case of bacteria, it was observed that, in addition, subgroups were formed according to the climatic seasons for the Salamanca soil. Interestingly, this also occurred with the fungal communities, where the samples were separated by season in both soil types. These results suggest that the main difference in soil microbial communities is due to edaphic properties, although environmental factors such as temperature, humidity or rainfall also influence the diversity of soil microbial communities. In addition, the microbiome associated with the legume Lupinus angustifolius cultivated under natural and greenhouse conditions was also characterized. For this purpose, wild and greenhouse-grown plants were collected from the same locations and analysed by 16S rRNA gene and ITS-2 gene profiling. Bacterial communities were characterized in the different plant compartments (rhizosphere, roots, nodules and leaves) while ITS profiles were restricted to the soil and rhizosphere. As previously reported for other plants, the highest richness was found in the rhizosphere, followed by the roots, leaves, and nodules. Within the rhizosphere, the bacterial richness in the in Salamanca plants was lower, especially for the field samples, probably affected by a pH below 7 and high amounts of P and K. In general, the compartments from the plants grown under greenhouse conditions showed a slightly higher bacterial diversity when compared to the wild plants. Within the fungal communities, the Shannon index was significantly higher in soil than rhizosphere samples (P<0.0001). In soils, diversity was similar for all seasons, except for spring, being lower in both locations, while in the rhizosphere, the field samples from Cabrerizos registered a significantly higher diversity than the greenhouse samples while the opposite occurred in Salamanca (P<0.0001). In both growing conditions and soils, the phyla with the highest cumulative relative abundance in all plant compartments were Pseudomonadota (Alphaproteobacteria - the most abundant taxon) and Mucoromycota. It was confirmed that L. angustifolius is a plant with a high bacterial and fungal diversity associated. In the second chapter, the objective was the isolation and molecular identification of the bacterial community present in the different plant tissues of L. angustifolius, to generate a collection of strains for downstream studies. Based on the metagenomics results, we selected 52 target genera with a relative abundance >1% and designed several isolation protocols. A total of 722 bacterial strains were isolated. As expected, the highest number of isolates was obtained in the rhizosphere compartment and a similar pattern was observed with a decreasing diversity gradient starting from the rhizosphere followed by the roots, leaves and nodules. In total, 87 different genera were identified, of which 19 had more than 10 isolates. The most abundant strains were identified in the genera Pseudomonas, Streptomyces, Agrobacterium, Bacillus and Pseudoclavibacter. In this work, 51.9% of the searched genera were isolated, and 74.7% of the isolated genera were identified by metagenomics, but 19.6% could not be detected in any plant compartment by metagenomics. Plant pathogenicity assays showed that 29% of the L. angustifolius isolates were potentially pathogenic for Arabidopsis thaliana Col-0. In turn, 394 strains (55%) were found to be non-pathogenic and 116 (16%) promoted the growth of A. thaliana. Analysis of metagenomics and culturomics results identified a core microbiome of the host plant L. angustifolius that included Acidovorax, Bradyrhizobium, Caulobacter, Chitinophaga, Flavobacterium, Kribella, Massilia, Pseudomonas, Pseudoxanthomonas, Rhizobium, Sphingomonas, Streptomyces and Variovorax. The composition and diversity of the identified host plant-associated bacteriome varied slightly between sampling locations and growing conditions. The genera identified as the core microbiome were present in more than 80% of the samples analysed. In chapter three of this work, the aim was to decipher the interactions of Micromonospora with its host plant and the associated microbiome, using the information obtained in the previous two chapters. Seven different synthetic communities (SynComs) were designed using bacterial strains isolated from the rhizosphere and roots of L. angustifolius to study their effect on the root and rhizosphere of the plant. In addition, we wanted to learn if the selected strains had any effect on the host plant and the natural bacterial communities present in the cultivation soils. After obtaining the genomes of the bacterial strains included in the different SynComs, a comparative genomic analysis was carried out, confirming that all the selected strains had genes with functions related to plant association and growth promotion. Plants were grown for 8 weeks in unsterilised soil under greenhouse conditions, and several plant parameters were measured and compared against the control plants (uninoculated). The plants inoculated with SynCom_7 showed the best growth and development. Furthermore, 16S rRNA gene profiling showed that the soil samples were the most diverse, followed by rhizosphere and roots (alpha diversity). Beta diversity grouped the samples into three clusters according to compartments: soil, rhizosphere and roots. In addition, a clustering pattern was observed for the SynComs inoculated in the root samples. All consortia that contained the nitrogen fixer, Bradyrhizobium sp. in the synthetic community formed one cluster, while the rest of the SynComs were recovered in a second cluster. The analysis of the bacterial composition of the bulk soil samples confirmed that the synthetic communities did not affect the composition of the soil where the plant was growing. However, when we studied the bacterial composition in the rhizosphere, a slight variation was observed, and the bacterial community of root samples was greatly influenced by the inoculated SynComs. The second part of this chapter consisted in the evaluation of the different SynComs on L. angustifolius plants grown in sterile soil under a gnotobiotic system. As in the first experiment, several growth parameters were registered, observing that plants inoculated with SynCom_7 showed the highest growths, again. Pseudomonas sp. Strain CRA141 showed the closest association with the roots. This result is not unexpected as it is well known that many Pseudomonas strains associate to plant roots. In addition, it was found that Micromonospora sp. Lupac 08 was detected in the rhizosphere and roots, and while this actinobacterium is not part of the core microbiome, it could be considered a "satellite" microorganism with important beneficial functions for the plant. Plant gene expression was related to the effect of the SynComs inoculated. When inoculated consortia included the Bradyrhizobium strain, very little differences were found when compared to the control plants, however, when only the Micromonospora strain and/or the other members of the SynComs were added, the differential gene expression increased threefold (Fig. 62). Gene ontology enrichment analyses revealed that those functions that were enriched by inoculating the different SynComs were clearly related to plant-microbe interaction functions. The same was observed for the enriched metabolic pathways when KEGG analysis was performed.