Modelling interactions that determine core gut microbiome stability to predict microbiome perturbation by opportunistic pathogens

Abstract

The gut microbiome plays a crucial role in host homeostasis, with implications for nutrition, immune development, metabolism, and protection against pathogens. Disturbance of the microbiome by microbial invasion can be negative or positive: invasions of opportunistic pathogens can cause disease while dysbiotic states need invasions to recover. However, the complexity of the microbiome challenges our understanding of what factors determine the ability of microbes to invade. In this study we measure interactions between members of a SynCom of prominent gut bacteria with supernatant assays. We measure relative abundances of growth of co-cultures up to four species to validate a generalised Lotka-Volterra model parameterized with these supernatant assays. We predict differential invasion profiles of the opportunistic pathogens Escherichia coli and Bacteroides ovatus based on their monoculture growth profiles and interactions with other species. With experimental data of both predicted invasion and resistant communities we confirm model predictions of invasion success. Our model shows that both negative interactions within a community and neutral to positive interactions with the invading species promote invasion success, but the interactions towards the invading species dominate. Our validated approach opens the way for testing of interactions of human gut microbiome species, thereby developing inventions to avoid pathogenic overgrowth and therapies to enhance health-benefitting invasions. ImportanceThe stability of the human gut microbiome is crucial for host health, with opportunistic pathogen invasions causing diseases and healthy strain replacements needed for recovery. The microbiotas complexity complicates understanding its stabilizing mechanisms. This study uses a 10-species synthetic community of common gut microbiota to predict stable communities and invasion success. We employ supernatant experiments to parameterize a computational model, accurately predicting community composition and invasion success of Escherichia coli and Bacteroides ovatus. Our findings show that interactions within the resident community and with the invader are important, but the latter dominate. These results pave the way for larger-scale studies to characterize gut microbiome interactions and properties that resist invasions, potentially benefiting health through improved probiotics and faecal microbiota transplants.