Modeling the emergent metabolic potential of soil microbiomes in Atacama landscapes

Abstract

Soil microbiomes harbor complex communities and exhibit important ecological roles resulting from biochemical transformations and microbial interactions. Difficulties in characterizing the mechanisms and consequences of such interactions together with the multidimensionality of niches hinder our understanding of these ecosystems. The Atacama Desert is an extreme environment that includes unique combinations of stressful abiotic factors affecting microbial life. In particular, the Talabre Lejia transect has been proposed as a unique natural laboratory for understanding adaptation mechanisms. We propose a systems biology-based computational framework for the reconstruction and simulation of community-wide and genome-resolved metabolic models, in order to provide an overview of the metabolic potential as a proxy of how microbial communities are prepared to respond to the environment. Through a multifaceted approach that includes taxonomic and functional profiling of microbiomes, simulation of the metabolic potential, and multivariate analyses, we were able to identify key species and functions from six contrasting soil samples across the Talabre Lejia transect. We highlight the functional redundancy of whole metagenomes, which act as a gene reservoir from which site-specific functions emerge at the species level. We also link the physicochemistry from the puna and the lagoon samples to specific metabolic machineries that could be associated with their adaptation to the unique environmental conditions found there. We further provide an abstraction of community composition and structure for each site that allows to describe them as sensitive or resilient to environmental shifts through putative cooperation events. Our results show that the study of community-wide and genome-resolved metabolic potential, together with targeted modeling, may help to elucidate the role of producible metabolites in the adaptation of microbial communities. Our framework was designed to handle non-model microorganisms, making it suitable for any (meta)genomic dataset that includes nucleotide sequence data and high-quality environmental metadata for different samples.