Mapping Gut Microbiota Interactions That Are Robust to C. Difficile strain Variability and Nutrient Landscapes

The deleterious human gut pathogen Clostridioides difficile strains display extreme genetic variability and confront a changeable nutrient landscape in the gut. By combining high-throughput experiments to build human gut communities with data-driven computational models, we map inter-species interactions shaping the growth of diverse C. difficile strains and toxin production in different nutrient environments. We demonstrate C. difficile strain-specific variation in the inferred inter-species interaction network and reveal differences in gene expression between C. difficile strains. Besides, toxin production of C. difficile strains displays large community-context dependent variability, and they are not correlated with growth-mediated interactions. Although C. difficile growth is inhibited by gut bacteria in media that promotes high resource competition, most of them cannot inhibit C. difficile in media that mimic post-antibiotic perturbation where there are multiple preferred carbohydrates for C. difficile to consume. We identify C. hiranonis to robustly inhibit the growth and toxin production of diverse C. difficile strains across different nutrient environments. Guided by our computational model, we design an anti-C. difficile community containing C. hiranonis that protects mice from C. difficile. Our study highlights strain and nutrient variation as important factors to consider when designing anti-C. difficile consortia.