A single gut bacterium made selectively bred mice friendlier, and raised their oxytocin

Selective breeding can turn a skittish wild mouse into an animal that willingly climbs onto a human hand. Scientists have known for years that genetics plays some role in that shift, but the low heritability of tameness hinted that something else was involved. New preliminary research published in DNA Research points to an unexpected contributor: a common gut bacterium called Limosilactobacillus reuteri.

L. reuteri is consistently enriched in selectively bred tame mice

The research team, led by Biswa, Mori, Koide, and colleagues, bred wild-derived heterogeneous stock (WHS) mice over multiple generations for a behavior called active tameness: the motivation to approach and remain on a human hand. Prior selective breeding work on WHS mice had mapped tameness-linked loci to chromosome 11, but the genetics only explained part of the story. In the new study, shotgun metagenomic sequencing of fecal samples from 80 mice across two tamed and two nonselected groups produced 374 high-quality bacterial genome assemblies, 27 of which represented novel species. Despite this resolution, overall gut microbiota diversity was not meaningfully different between tamed and wild-type groups. The single consistent signal, emerging across three independent analytical methods (Kraken2 classification, metagenome-assembled genomes, and Pearson correlation), was a significantly elevated abundance of L. reuteri in the tamed mice.

Tameness, oxytocin, and plasma pyruvate all shift together

The published findings showed that tamed mice scored higher on every component of the behavioral battery: longer contact durations, more heading and accepting behaviors, and more time staying on a human hand. Their blood oxytocin levels were also elevated, and hippocampal expression of Oxtr, the gene encoding the oxytocin receptor, was higher than in nonselected animals. Metabolomic screening of 70 plasma compounds flagged one molecule as significantly elevated after correction for multiple comparisons: pyruvic acid, or pyruvate, a metabolite that L. reuteri is known to secrete. Whether the pyruvate is the actual signal linking the bacterium to the brain, or just a correlated byproduct, remains an open question. The gut-brain axis and microbial metabolites are an active area of research, and pyruvate's precise role in neural signaling is not well established.

Dosing wild mice with a pyruvate-secreting strain changed their behavior

To test causality, the team isolated L. reuteri strains directly from the tame mice and screened them for pyruvate secretion. Strain NIG-A41 secreted high levels of pyruvate; strain NIG-23 secreted low levels but had a nearly identical chromosomal genome. A third bacterium, Lactobacillus helveticus JCM1120, also secreted pyruvate but belongs to a different species. Nonselected mice received one of these three bacteria in their drinking water for 21 days. Both NIG-A41 and NIG-23 colonized the gut successfully, increasing more than 1,500-fold in fecal samples. L. helveticus did not establish itself in the same way. After the treatment period, mice given NIG-A41 showed significantly increased active tameness and elevated serum oxytocin. Mice given NIG-23 or L. helveticus did not show the same behavioral shift. Notably, NIG-A41 administration did not raise plasma pyruvate, even though that metabolite was elevated in naturally tamed mice. The distinction between the two L. reuteri strains appears to lie in their plasmids: NIG-A41 carries plasmid-encoded genes absent from NIG-23, suggesting that a strain-specific gene repertoire, not simply pyruvate secretion, drives the behavioral and hormonal effects.

What the study cannot yet establish

This is early-stage, investigational research, and the authors are candid about its limits. The study did not directly measure whether oxytocin is the causal link between L. reuteri and tameness, or simply a correlated marker. A significant but mild Pearson correlation between oxytocin levels and active contact time appeared in both the initial characterization and the administration experiment, but correlation is not causation. Broader questions about gut microbiota's influence on host behavior remain under active investigation: the molecular and cellular pathways by which gut microbes affect neural circuits are still being mapped, and this study adds a compelling data point rather than a resolved mechanism. The team also acknowledges that the behavioral tests measured active tameness specifically, not general sociability, so the findings should not be overread as evidence about social behavior in a broader sense.

What comes next for L. reuteri and behavior research

The plasmid difference between NIG-A41 and NIG-23 is now a concrete target. Identifying which plasmid-encoded genes are responsible for the oxytocin and tameness effects would clarify whether a specific metabolite or signaling molecule is involved, or whether the mechanism is more indirect. The broader implication, tentative but striking, is that gut bacteria shaped through selective pressure could contribute to the behavioral shifts that define domestication as a biological phenomenon. Domestication's influence on gut microbiome is increasingly recognized as a two-way street, with diet and environment shaping microbiota while microbiota in turn shapes host physiology and behavior. In clinical science, growing interest in probiotics for anxiety and depression treatment reflects a parallel push to understand whether similar microbe-to-brain pathways operate in humans. None of that is established here, but the mechanistic thread running from a gut bacterium through a metabolite to oxytocin and then to approach behavior is precisely the kind of link that makes the field worth watching.

Evidence rating

INVESTIGATIONAL — 71/100

Caveats:

• Animal study only — all findings are in mice; human relevance is unknown.
• The causal chain is incomplete: L. reuteri administration raised oxytocin and tameness, but plasma pyruvate levels did not increase, so the proposed pyruvate-oxytocin mechanism is not confirmed.
• Small groups (n=20 per group) limit statistical power.
• Funding was not disclosed by the authors.

Mouse study using selectively bred tame and non-selected control groups (n=80 total). Demonstrates an association between L. reuteri abundance and tameness behaviour, with a bacterial administration experiment suggesting a causal direction, but the proposed pyruvate-mediated mechanism was not confirmed. Entirely preclinical; not yet replicated in an independent population.

References

1. Bhim B Biswa, Hiroshi Mori, Atsushi Toyoda, Kazumichi Fujiwara, Ken Kurokawa, Tsuyoshi Koide (2026). Increased abundance of Limosilactobacillus reuteri in the gut of selectively bred high-tameness mice and its association with behavioural changes. DNA Research. https://doi.org/10.1093/dnares/dsag006
2. Yuki Matsumoto, Hiromichi Nagayama, Hirofumi Nakaoka, Atsushi Toyoda, Tatsuhiko Goto, Tsuyoshi Koide (2021). Combined change of behavioral traits for domestication and gene‐networks in mice selectively bred for active tameness. Genes, Brain and Behavior. https://doi.org/10.1111/gbb.12721
3. Host–gut microbiota interactions in health and disease: mechanisms and intervention strategies – Frontiers. frontiersin.org. https://www.frontiersin.org/articles/10.3389/fmicb.2023.1118318/full
4. The Microbiome and Host Behavior – PMC – NIH. NIH. https://pmc.ncbi.nlm.nih.gov/articles/PMC4732151/
5. Erica P Ryu, Emily R Davenport (2021). Taming the beasts inside. eLife. https://doi.org/10.7554/elife.67634
6. Probiotics' Effects in the Treatment of Anxiety and Depression: A Comprehensive Review of 2014–2023 Clinical Trials – PMC. NIH. https://pmc.ncbi.nlm.nih.gov/articles/PMC10844781/