Quantifying Sources of Variation in the Laboratory Mouse Gut Microbiome (2025-US-PO-2516)

Growing evidence shows that microbiomes influence diverse aspects of host physiology. As a consequence, taking a holobiont perspective toward animal models represents a potentially powerful approach to improve understanding of how environment, microbiome, and host genetics combine to impact disease onset and development.

However, a major challenge facing animal models is that, beyond exclusion of specific pathogens, most microbiomes are uncontrolled and largely unmonitored across different animal facilities. There is therefore an urgent need to understand 1) sources of microbiome variation in animal models, and 2) the impact of such variation on host phenotype.

Here we sought to address the first of these two challenges. Using mouse chromosome substitution strains (CSSs) combined with different multivariate variance partitioning approaches, we quantified the relative impact of environmental and genetic sources of variation on taxonomic composition of the fecal microbiome.

To do so, we applied Uniform Manifold Approximation and Projection (UMAP) for dimensionality reduction, followed by variance component analysis on the resulting ordination axes (UMAP-VC). All analyses and visualizations were performed using JMP, which enabled interactive exploration of clustering patterns and efficient estimation of variance contributions from factors such as facility, strain, sex, and diet.

Our conclusion is that animal facility exerts an effect on microbiome composition that eclipses the influence of host genetics, as well as other environmental factors such as diet.