Stochastic changes over time and not founder effects drive cage effects in microbial community assembly in a mouse model

Maternal transmission and cage effects are powerful confounding factors in microbiome studies. To assess the consequences of cage microenvironment on the mouse gut microbiome, two groups of germ-free (GF) wild-type (WT) mice, one gavaged with a microbiota harvested from adult WT mice and another allowed to acquire the microbiome from the cage microenvironment, were monitored using Illumina 16S rRNA sequencing over a period of 8 weeks. Our results revealed that cage effects in WT mice moved from GF to specific pathogen free (SPF) conditions take several weeks to develop and are not eliminated by the initial gavage treatment. Initial gavage influenced, but did not eliminate a successional pattern in which Proteobacteria became less abundant over time. An analysis in which 16S rRNA sequences are mapped to the closest sequenced whole genome suggests that the functional potential of microbial genomes changes significantly over time shifting from an emphasis on pathogenesis and motility early in community assembly to metabolic processes at later time points. Functionally, mice allowed to naturally acquire a microbial community from their cage, but not mice gavaged with a common biome, exhibit a cage effect in Dextran Sulfate Sodium-induced inflammation. Our results argue that while there are long-term effects of the founding community, these effects are mitigated by cage microenvironment and successional community assembly over time, which must both be explicitly considered in the interpretation of microbiome mouse experiments.