Environmental modulation of metabolic function in microbial communities

University Of Chicago, Chicago IL

Investigators

Linked publications, trials & patents

Abstract

Microbial communities are complex systems whose emergent metabolic properties play a key role in determining human health. Metabolic processes enabled by host-associated microbiota play a defining role in individual health outcomes, and the emergent metabolism of microbial consortia affect environmental processes from eutrophication to climate change, impacting human health on a global scale. Therefore, humanity would benefit from a quantitative understanding of the rules by which the genomic composition of a microbial community, and the environment in which it resides, determines its emergent metabolism. Discovering the principles by which environmental variation alters community structure and determines metabolic function is a necessity if we are to manipulate or design communities to improve health outcomes. However, this task is challenging for existing methods. In preliminary work, we establish a new quantitative framework for predicting the emergent metabolism of a bacterial community from its genomic composition using denitrification as a model metabolic process. Combining quantitative bacterial phenotyping, modeling, and a simple statistical approach we demonstrated a method that quantitatively maps gene content to metabolite dynamics in microbial communities. This insight provides a route to quantitatively connecting the genes present in a community to metabolite dynamics. The next challenge is to use this insight to understand how community function and structure depend on the environment. We propose to extend this success by understanding how environmental gradients, complexity, and dynamics impact community structure and function. We accomplish this by developing denitrification as a model metabolic process. The outcomes of the proposed work will be three-fold. First, microbiome studies have documented ubiquitous associations between environmental conditions and community composition, but we do not understand the ecological or physiological origins of these emergent patterns or their metabolic consequences. Using denitrifying communities across a pH gradient I will show that such patterns emerge from ecological interactions. I will show that these interactions arise generically from the presence of physiological trade-offs on microbial traits, providing a generalizable route to understanding the functional impact of environmental variation on communities. Second, our preliminary study connected genomes to community metabolism for a simple metabolic pathway acting. I will extend this success to complex pathways and environmental conditions by constructing a method for predicting carbon utilization by communities in complex nutrient conditions directly from genomes. I will utilize a powerful blend of genome-scale metabolic modeling and multi-view machine learning, with impacts from host physiology to climate change. Third, I will use denitrifying communities to test the idea that, like cells and organisms, microbial communities exhibit predictive behaviors in dynamic environments. I propose that communities assembled in environments with distinct schedules of aerobic respiration and anaerobic respiration (denitrification) adapt to facilitate the prompt utilization of electron acceptors. I will test the hypothesis that community-level learning emerges from ecological interactions and distinct gene regulatory programs, providing a new conceptual lens through which we can view community adaptation to dynamic environments.