Buhl, Eva M.

Summary Cultivation via classical agar plate (CAP) approaches is widely used to study microbial communities, but they are time‐consuming. An alternative approach is the application of single‐cell dispensing (SCD), which allows high‐throughput, label‐free sorting of microscopic particles. We aimed to develop a new anaerobic SCD workflow to cultivate human gut bacteria and compared it with CAP using faecal communities on three rich culture media. We found that the SCD approach significantly decreased the experimental time to obtain pure cultures from 17 ± 4 to 5 ± 0 days, while the isolate diversity and relative abundance coverage were comparable for both approaches. We further tested the total captured fraction by sequencing the sorted bacteria directly after growth as bulk biomass from 2400 dispensed single cells without downstream identification of individual strains. In this approach, the cultured fraction increased from 35.2% to 52.2% for SCD, highlighting the potential for deeper cultivation projects from single samples. SCD‐based cultivation also captured species not detected by sequencing (16 ± 5 per sample, including seven novel taxa). From this work, 82 human gut bacterial species across five phyla (Actinobacteriota, Bacteroidota, Desulfobacterota, Firmicutes and Proteobacteria) and 24 families were obtained, including the first cultured member of 11 novel genera and 10 novel species that were fully characterized taxonomically.

Microbiome research is hampered by the fact that many bacteria are still unknown and by the lack of publicly available isolates. Fundamental and clinical research is in need of comprehensive and well-curated repositories of cultured bacteria from the intestine of mammalian hosts. In this work, we expanded the mouse intestinal bacterial collection (www.dsmz.de/miBC) to 212 strains, all publicly available and taxonomically described. This includes the study of strain-level diversity, small-sized bacteria, and the isolation and characterization of the first cultured members of one novel family, 10 novel genera, and 39 novel species. We demonstrate the value of this collection by performing two studies. First, metagenome-educated design allowed establishing custom synthetic communities (SYNs) that reflect different susceptibilities to DSS-induced colitis. Second, nine phylogenetically and functionally diverse species were used to amend the Oligo-Mouse Microbiota (OMM)12 model [Brugiroux et al. 2016 Nat Microbiol]. These strains compensated for differences observed between gnotobiotic OMM12 and specific pathogen-free (SPF) mice at multiple levels, including body composition and immune cell populations (e.g., T-cell subtypes) in the intestine and associated lymphoid tissues. Ready-to-use OMM stocks are available to the community for use in future studies. In conclusion, this work improves our knowledge of gut microbiota diversity in mice and enables functional studies via the modular use of isolates.