A standardized gnotobiotic mouse model harboring a minimal 15-member mouse gut microbiota recapitulates SOPF/SPF phenotypes

Mus musculus is the classic mammalian model for biomedical research. Despite global efforts to standardize breeding and experimental procedures, the undefined composition and interindividual diversity of the microbiota of laboratory mice remains a limitation. In an attempt to standardize the gut microbiome in preclinical mouse studies, here we report the development of a simplified mouse microbiota composed of 15 strains from 7 of the 20 most prevalent bacterial families representative of the fecal microbiota of C57BL/6J Specific (and Opportunistic) Pathogen-Free (SPF/SOPF) animals and the derivation of a standardized gnotobiotic mouse model called GM15. GM15 recapitulates extensively the functionalities found in the C57BL/6J SOPF microbiota metagenome, and GM15 animals are phenotypically similar to SOPF or SPF animals in two different facilities. They are also less sensitive to the deleterious effects of post-weaning malnutrition. In this work, we show that the GM15 model provides increased reproducibility and robustness of preclinical studies by limiting the confounding effect of fluctuation in microbiota composition, and offers opportunities for research focused on how the microbiota shapes host physiology in health and disease.

fold magnification, NanoZoomer S60, Hamamatsu). Members of Bacteroidaceae, Tannerellaceae and Enterobacteriaceae stained Gram-negative, while Lactobacillaceae, Erysipelotrichaceae and Ruminococcaceae stained Gram-positive. Lachnospiraceae stained Gram-positive, except Longibacillum caecimuris MD329 and Subtilibacillum caecimuris MD335, which stained Gram-negative likely due to their cell wall structure as already reported for other Clostridiales [1]. Source data are provided as a Source Data file.
Supplementary Fig. 2. In vivo experimental design. a Facility 1. First, 5 couples of 8 weekold C57BL/6J GF mice were colonized by oral gavage with the fresh frozen GM15 bacterial community, twice at 48h interval (F0). GM15 microbiota stability was assessed by qPCR microfluidic assay from feces collected after 1, 2 and 3 weeks. Then, in order to evaluate the reproducible transfer of the GM15 microbiota by fecal microbiota transplantation, 7 week-old C57BL/6J GF mice were orally gavaged with a suspension of fresh fecal pellets from GM15 mice (F0), twice at 48h interval. Again, qPCR microfluidic assay from feces collected after 1, 2 and 3 weeks was carried out. Next, the GM15 mouse line was amplified to monitor the GM15 microbiota through nine filial generations at 6 weeks of age (F1-F9), and allow the phenotyping study from two consecutive generations (F1.2 and F2.1). Reproduction performance and perinatal mortality were recorded, 4 week-old mice were randomly selected at weaning, monitored weekly for body weight and size, and feed intake, until sacrifice at 8-9 weeks of age.
GF and SOPF mice were also studied as control groups. Besides, a comparative analysis was done on the fecal microbiota of 8 week-old GM15 mice (F1.1) either fed with the breeding diet or an alternative isocaloric diet given for 4 weeks. Finally, the fecal microbiota of 6 month-old and 12 month-old control GM15 mice fed with the breeding diet was analyzed. b Facility 2.
First, 3 trios of 8 week-old C57BL/6J GF mice were colonized by oral gavage with the fresh frozen GM15 bacterial community, twice at 48h interval (F0). GM15 microbiota stability was assessed by qPCR microfluidic assay from feces collected after 1 and 3 weeks. Next, the GM15 mouse colony was amplified to monitor the GM15 microbiota in generation F1 at 6 weeks of age, to allow the phenotyping study and to assess mice' response to diet-induced stunting.
Supplementary Fig. 3. Assessment of gut microbiota stability in GM15 founders and of reproducibility between fecal and caecal samples of individual GM15 mice. SOPF group shows the distribution of each GM15 strains in the complex gut microbiota of 8-week-old SOPF mice. The absolute quantification of each strain was determined by specific qPCR microfluidic assay. * Strains I50, MD294, MD300 and JM4-15 were at the detection limit of the qPCR microfluidic assay, and thus were not detected in all samples. ** Strains MD335, MD329 and MD308 were below detection limit of the qPCR microfluidic assay. Strain YL32, obtained from the DSMZ collection, was not detected in our SOPF colony.   Dunn's multiple comparison test applied to GM15 strains concentration quantified by microfluidic qPCR assay in GM15 mice and SOPF/SPF mice between facility 1 and facility 2.