Killem all? Interactions of predatory Myxobacteria with soil microbes an in vitro and microcosm perspective on their role in the soil microbial food-web.

Soil is one of the most complex ecosystems, being an elementary source for food and resources needed by humankind. Despite its pivotal role in shaping and transforming this complex ecosystem, the soil microbial food-web is still poorly understood, which stands also true for matter and energy fluxes to higher trophic levels. A group of microbiome predators that has recently come more into focus are the Myxobacteria. They are famous for their predatory life style and might thus influence microbial death, growth and turnover in the soil environment via predation and an elaborate arsenal of secondary metabolites. To shed light on their role in the soil food-web, we investigated their predatory behavior and interactions with potential prey and predators using in vitro and in situ approaches. We tested the predation spectra in vitro with binary interaction assays of four different Myxobacteria (Haliangium ochraceum, Myxococcus virescens, Myxococcus fulvus, Corallococcus coralloides) with 16 different prey bacteria isolated from soils. The in vitro assay showed that each Myxobacterium had species-specific prey spectra. While Haliangium ochraceum and Myxococcus virescens showed the strongest predation effects on prey bacteria, the Corallococcus coralloides strain lysed the fewest prey bacteria. Taken together, not a single bacterium of the tested ones was resistant to lysis. Remarkably, also strains of the never before tested phyla Gemmatimonadota and Veruccomicrobiota can be lysed by Myxobacteria. To shed light on (inter)actions of Myxobacteria with microbiome members in situ, a 32 day long microcosm study with an agricultural soil was performed, manipulating (a) the carbon source and (b) the grazing pressure of higher trophic levels, via the addition of fungivorous and bacterivorous nematodes, respectively. We applied quantitative PCR and quantitative metatranscriptomics microbiome profiling of 80 samples to shed light of the impact of aforementioned manipulations of C-source and grazers on the microbiome. Three-Domain SSU rRNA profiling showed that myxobacteria were highly abundant (up to 20%) in the used agricultural soil. In contrast, fungal abundance was much lower (1.5 %), while Protozoa, i.e. dominated by Amoebozoa and Cercozoa, were much more abundant than fungi (5 %). This suggests a microbial food-web dynamic in this agricultural soil that was heavily dominated by a bacterial, and not fungal channel. As expected, the abundance and composition of Myxobacteria were not affected by addition of fungivorous Nematodes, but it was surprising that this was also true for adding bacterivorous nematodes. We speculate that Myxobacteria reduce predation pressure of nematodes by utilizing secondary metabolites, while in turn killing enough prey bacteria for their own metabolism and growth. Consequently, the high abundance of Myxobacteria suggests a substantial contribution of their (predatory) activity on matter and energy fluxes in these microcosms. In a next step, we will integrate the metatranscriptomics results with organic matter and energy fluxes via flux-web modelling. In conclusion, the in vitro assays showed that Myxobacteria killed all prey bacteria. This, together with their high abundance and resistance to predation from higher trophic levels in the in situ microcosm suggest Myxobacteria as important players in the agricultural soil food web.