Frédéric Moens

From the estimated 2.2 to 3.8 million fungal species existing on Earth, only a minor fraction actively colonizes the human gastrointestinal tract. In fact, these fungi only represent 0.1% of the gastrointestinal biosphere. Despite their low abundance, fungi play dual roles in human health—both beneficial and detrimental. Fungal infections are often associated with bacterial dysbiosis following antibiotic use, yet our understanding of gut fungi–bacteria interactions remains limited. Here, we used the SHIME^® gut model to explore the colonization of human fecal-derived fungi across gastrointestinal compartments. We accounted for the high inter-individual microbial diversity by using fecal samples from healthy adults, healthy babies, and Crohn’s disease patients. Using quantitative Polymerase Chain Reaction and targeted next-generation sequencing, we demonstrated that SHIME^®-colonized mycobiomes change upon loss of transient colonizers. In addition, SHIME^® reactors from Crohn’s disease patients contained comparable bacterial levels as healthy adults but higher fungal concentrations, indicating unpredictable correlations between fungal levels and total bacterial counts. Our findings rather link higher bacterial α-diversity to limited fungal growth, tied to colonization resistance. Hence, while healthy individuals had fewer fungi engrafting the colonic reactors, low α-diversity in impaired (Crohn’s disease patients) or immature (babies) microbiota was associated with greater fungal abundance. To validate, antibiotic-treated healthy colonic microbiomes demonstrated increased fungal colonization susceptibility, and bacterial taxa that were negatively correlated with fungal expansion were identified. In summary, fungal colonization varied individually and transiently, and bacterial resistance to fungal overgrowth was more related with specific bacterial genera than total bacterial load. This study sheds light on fungal–bacterial dynamics in the human gut.

Clostridioides difficile (formerly Clostridium difficile) infection (CDI) is one of the most common hospital-acquired infections, which is often triggered by a dysbiosed indigenous gut microbiota (e.g., upon antibiotic therapy). Symptoms can be as severe as life-threatening colitis. The current study assessed the antipathogenic potential of human milk oligosaccharides (HMOs), i.e., 2′-O-fucosyllactose (2′FL), lacto-N-neotetraose (LNnT), and a combination thereof (MIX), against C. difficile ATCC 9689 using in vitro gut models that allowed the evaluation of both direct and, upon microbiota modulation, indirect effects. During a first 48 h fecal batch study, dysbiosis and CDI were induced by dilution of the fecal inoculum. For each of the three donors tested, C. difficile levels strongly decreased (with >4 log CFU/mL) upon treatment with 2′FL, LNnT and MIX versus untreated blanks, coinciding with increased acetate/Bifidobacteriaceae levels. Interindividual differences among donors at an intermediate time point suggested that the antimicrobial effect was microbiota-mediated rather than being a direct effect of the HMOs. During a subsequent 11 week study with the Pathogut^TM model (specific application of the Simulator of the Human Intestinal Microbial Ecosystem (SHIME^®)), dysbiosis and CDI were induced by clindamycin (CLI) treatment. Vancomycin (VNC) treatment cured CDI, but the further dysbiosis of the indigenous microbiota likely contributed to CDI recurrence. Upon co-supplementation with VNC, both 2′FL and MIX boosted microbial activity (acetate and to lesser extent propionate/butyrate). Moreover, 2′FL avoided CDI recurrence, potentially because of increased secondary bile acid production. Overall, while not elucidating the exact antipathogenic mechanisms-of-action, the current study highlights the potential of HMOs to combat CDI recurrence, help the gut microbial community recover after antibiotic treatment, and hence counteract the adverse effects of antibiotic therapies.