The study team built a four-cohort system to trace how gut microbiomes shift from wild to laboratory conditions. They sampled free-living wood mice (Wild:As), a wild-derived colony kept in captivity for over a decade (Wild-like:As), and a colony produced from the Wild-like:As lineage by caesarean rederivation and cross-fostering into laboratory Mus musculus mothers (Lab-like:As). The authors compared those three Apodemus sylvaticus cohorts with conventional laboratory Mus musculus (Lab:Mm). The analysis examined alpha diversity, taxonomic community composition, and predicted functional potential to capture multiple dimensions of microbial change.
Data indicated that alpha diversity does not follow a simple monotonic decline from wild to lab. The study team reported that Shannon diversity was higher in Lab:Mm and Lab-like:As than in Wild:As and Wild-like:As. Short statement: Shannon and phylogenetic metrics disagreed. In contrast, phylogenetic diversity was similarly high in both Wild:As and Lab:Mm groups, so the choice of metric changed the apparent pattern. These findings mean that different aspects of diversity — richness plus evenness versus phylogenetic breadth — can yield divergent interpretations when comparing wild and laboratory microbiomes.
Community composition more closely reflected cohort histories than a simple wild-to-laboratory axis. Despite being held in captivity for over a decade, the Wild-like:As mice remained most similar to the Wild:As conspecifics, the authors observed. Lab-like:As mice shifted strongly away from Wild-like:As and toward Lab:Mm composition, but they nonetheless remained taxonomically distinct from Lab:Mm. That pattern suggests that the processes that govern taxonomic reassembly include both historical microbial exposure and the host rearing context — not just time in captivity.
A clearer wild-to-laboratory signal emerged when the team quantified microbiome consistency across individuals. Core ASVs (defined by the investigators as present in ≥90% of individuals in a cohort) increased stepwise across the spectrum: 9 core ASVs in Wild:As, 46 in Wild-like:As, 72 in Lab-like:As, and 138 in Lab:Mm. This progressive rise in shared taxa indicates that laboratory-associated microbiomes become more consistent across hosts, even where taxonomic composition does not converge completely. Short point: captivity and standardised host environments appear to increase the set of taxa shared across individuals.
When the team examined predicted functional potential, taxonomic and functional patterns did not fully align. The investigators reported that Wild:As and Wild-like:As microbiomes did not differ significantly in predicted functional potential, despite taxonomic differences between some cohorts. Lab-like:As microbiomes shifted toward the Lab:Mm functional state even while remaining taxonomically distinct from Lab:Mm. This decoupling suggests that different taxonomic configurations can produce similar inferred functional repertoires, and that the host rearing environment can drive functional convergence faster than taxonomic convergence.
The authors framed the Wild-like and Lab-like Apodemus microbiome states as persistent and experimentally tractable. They noted that these states provide a system for testing how microbiome history influences host physiology, resilience, and responses to environmental challenges. The source did not report specific intervention studies, mechanistic assays, or host phenotypic outcomes in this paper; those details remain for future work. Key unanswered questions documented by the authors include how microbiome history alters host health trajectories, which functional shifts are responsible for physiological effects, and whether rearing context or inherited microbial exposure plays the dominant causal role.
Limitations recorded in the preprint include that the work is not peer reviewed and that the reported comparisons rest on cohort construction and predicted functional inference rather than direct functional assays. The study therefore identifies clear experimental next steps: targeted functional validation, mechanistic testing of host–microbiome interactions, and controlled manipulations to parse the relative contributions of microbial history versus immediate environment to host outcomes.