**Background:** The evolutionary relationships between gut microbes and their vertebrate hosts remain poorly understood. Limosilactobacillus reuteri has been established as a model for studying host-microbe symbiosis, with prior work identifying host-specific phylogenetic lineages. However, sampling bias toward humans and domesticated animals has limited understanding of the true evolutionary history and host specificity of this species.
**Methods:** The authors isolated 94 new L. reuteri strains from zoo and wild animals (6 bird species, 9 non-human primate species, 8 rodent species) and humans from industrialized and non-industrialized societies. Combined with 88 publicly available genomes and 25 genomes of closely related Limosilactobacillus species, the final dataset comprised 182 L. reuteri genomes from over 40 vertebrate host species. Core-genome phylogeny was constructed using 490 recombination-free core genes. Gene content analysis, pan-genome-wide association studies (Pan-GWAS), ancestral genome reconstruction, and prophage exchange analysis were performed. Bayesian evolutionary analysis (BEAST) estimated divergence times and ancestral host states. Colonization experiments in germ-free mice tested adherence to forestomach epithelium for 18 L. reuteri strains and 13 strains of related species.
**Key Results:** Phylogenetic analysis revealed ten lineages, including four novel lineages (VII–X). Rodents were identified as the ancestral hosts of L. reuteri, with divergence from the last common ancestor estimated at ~9,100 years ago (95% HPD: 9,000–9,300). Three major host transitions occurred 7,000–5,000 years ago: rodents to humans/herbivores (lineage II), rodents to birds (lineage VI), and rodents to primates (lineage IX). Rodent-associated lineages (I, III, VII, VIII, X) clustered together in gene content analysis as Gene Content Group 4, sharing ancestral genes for biofilm formation, acid resistance, and urease activity that were lost in non-rodent lineages. Gene gain and loss analysis showed >90% (507/555) of lineage-specific genes were acquired during evolution, while 48 genes were ancestral—28 shared by all rodent lineages and 20 retained in some but not all rodent lineages. Prophage exchange was >10 times higher within lineages than between lineages, with rodent lineages showing the highest inter-lineage prophage sharing. In mouse experiments, all strains from Gene Content Group 4 showed high epithelial colonization (≥10^6.9 CFU/g, ≥20% relative to control), while strains from non-rodent lineages showed low colonization (≤10^6.6 CFU/g, ≤10% relative to control). Human isolates clustered with poultry (lineage VI) or herbivore (lineage II) strains and lacked human-specific genes. L. reuteri was not isolated from any wild or zoo pig samples despite extensive sampling.
**Clinical Implications:** This study challenges the assumption that L. reuteri is autochthonous to humans, instead suggesting that human strains are acquired from domesticated animals through zoonotic transmission. The finding that human isolates cluster with poultry and herbivore strains, combined with low prevalence in modern human populations, indicates that urbanization and reduced contact with farm animals may decrease exposure to potentially beneficial microbes. The authors note that L. reuteri has documented health-promoting, immunomodulating, and anti-inflammatory effects, and that zoonotic transmission of gut symbionts may contribute to the health benefits associated with farm living (e.g., reduced allergies and autoimmune diseases). Understanding these transmission dynamics could inform probiotic development and microbiome-based interventions.