Cheese-Derived Probiotic Shifts Gut Microbes and Behavior in Autism-Model Mice

If you've looked into research connecting gut bacteria to autism spectrum disorder (ASD), you've probably seen plenty of correlations and few concrete interventions to point to. A new mouse study adds something more specific: a probiotic strain isolated from a traditional Colombian cheese, built into a synbiotic, reshaped the gut bacteria of mice carrying human ASD-linked microbiota, and was followed by improved sociability and reduced repetitive behavior in those same mice. The effects showed up far more strongly in mice carrying microbiota from children with ASD than in mice carrying microbiota from neurotypical children, which is the detail that makes this more than a generic probiotic result.

Key takeaways

• Researchers transplanted gut microbiota from Colombian children with ASD (7 donors) and without ASD (18 donors) into germ-free mice, then treated the mice with a synbiotic built on the probiotic strain Limosilactobacillus fermentum K73.
• After 21 days, mice carrying ASD-linked microbiota showed bigger shifts in gut bacteria and short-chain fatty acids than mice carrying neurotypical microbiota, plus increased sociability and more than double the marble-burying performance.
• The catch: this is a mouse model using transplanted human microbiota, not a human clinical trial, and the donor sample was small and unbalanced (7 ASD donors vs. 18 neurotypical donors).

Limosilactobacillus fermentum K73 Synbiotic Reshapes ASD-Linked Gut Microbiota

Researchers from Universidad de La Sabana in Colombia and the University of Florida, publishing in the Journal of Functional Foods, tested a synbiotic (a pairing of a probiotic bacterium with fermentable compounds meant to help it survive digestion) built around Limosilactobacillus fermentum K73, a strain the team previously isolated from artisanal Colombian "Costeño" cheese. They used gnotobiotic mice (germ-free mice with no microbes of their own until researchers deliberately colonize them) to compare the synbiotic's effect on two human-derived gut microbiota: one transplanted from Colombian children diagnosed with ASD, the other from neurotypical (NT) children of similar age. After 21 days of treatment, mice carrying ASD-linked microbiota showed the larger response, with declines in genera tied to gut dysbiosis such as Enterococcus and Bacteroides, alongside increases in beneficial, fatty-acid-producing genera including Faecalibacterium and Sellimonas.

Why Gut Dysbiosis in Autism Spectrum Disorder Matters to Families

Gut symptoms and behavior often move together for families managing autism spectrum disorder (ASD), which affects roughly 1 in 31 8-year-olds in the United States according to CDC surveillance data; many of those families, or those managing overlapping conditions like irritable bowel syndrome (IBS), already suspect a gut connection. This study's starting point supports that instinct: before any treatment, fecal samples from the 7 children with ASD had significantly lower bacterial diversity than samples from the 18 neurotypical donors, and the two groups formed statistically distinct microbial communities that persisted after transplantation into mice. That echoes a landmark 2019 study in Cell, which found that transplanting ASD-associated human microbiota into germ-free mice alone was enough to produce autism-linked behaviors. A microbiota that reliably carries that signature gives researchers something concrete to test an intervention against.

Inside the Study: Humanized Mice and a 21-Day Synbiotic Trial

The researchers first confirmed the synbiotic was safe in ordinary (non-germ-free) mice over three weeks, with no changes in body weight, food intake, or colon health. They then colonized germ-free mice with fecal samples from 4 ASD donors and 4 neurotypical donors (3 to 5 mice per donor, delivered by oral gavage), creating two groups of "humanized" mice. Four weeks later, once donor microbiota had established, the mice received the L. fermentum K73 synbiotic mixed into their food for 21 days. The team tracked gut bacteria by 16S rRNA sequencing, measured short-chain fatty acids (SCFAs, metabolites gut bacteria make from fiber) and the signaling molecule serotonin, and ran three standard behavioral tests, open field, three-chamber sociability, and marble burying, on 12 to 13 mice per group for metabolite analysis and 15 to 17 mice per group for behavior.

What the Sociability and Marble-Burying Results Could Mean

In the three-chamber sociability test, mice carrying ASD-linked microbiota spent significantly more time socializing with a cage mate after treatment, rising from 20.3 seconds to 32.3 seconds; neurotypical-colonized mice showed no significant change. In the marble-burying test, used as a proxy for repetitive behavior, ASD-colonized mice more than doubled their performance, from an average of 2.4 to 5.3 marbles buried, while neurotypical-colonized mice changed little. Fecal serotonin, tied to gut motility and broader gut-brain signaling, dropped from 8.4 to 5.0 nM/mg only in ASD-colonized mice, and acetic acid production tracked with the sociability improvement specifically in that group. None of this makes a specific probiotic a proven treatment for ASD in people: the study was conducted entirely in mice, and anyone managing an ASD diagnosis should treat findings like these as a reason to ask a doctor about ongoing research, not as grounds to start a supplement.

Limitations: A Small Donor Pool and No Proof of Cause

The authors call this a pilot study and flag its biggest constraint themselves: only 7 ASD donors were available against 18 neurotypical donors, an imbalance that limits statistical power and means the reported effect sizes should be read cautiously. It is also, unavoidably, a mouse model: gnotobiotic mice carrying transplanted human microbiota are a research tool for isolating microbiome effects, not a stand-in for human biology or behavior. Just as important, the researchers are explicit that their design shows correlation, not causation: acetic acid tracked with sociability, but the study wasn't built to prove one caused the other, and the acetic acid increase itself was transient rather than sustained.

What Comes Next for Synbiotic Research in Autism

The researchers point toward larger, more diverse donor cohorts, including regional efforts like the LatinBiota Consortium, to confirm whether these microbial signatures hold up in bigger, better-balanced groups. They also want to validate their earlier in-vitro digestion-bioreactor system as a faster screening tool for predicting synbiotic effects before committing to animal studies. Before any of this reaches people, researchers still need studies designed specifically to test whether microbiota and metabolite changes genuinely drive behavioral shifts, building on the broader microbiota-gut-brain axis research, rather than simply accompanying them, on the way to larger animal cohorts and eventually controlled human trials.

Common Questions

Does this mean a probiotic can treat autism in people? No. This study was conducted entirely in mice carrying transplanted human gut bacteria; it has not been tested in people with ASD, and no specific supplement should be considered a proven treatment based on these results.

What is a gnotobiotic mouse? A gnotobiotic mouse is born and raised germ-free, with no microbes of its own, until researchers deliberately colonize it with a known microbial community, here, human donor gut bacteria, so they can study that community's effects in isolation.

Where did the probiotic strain come from? The synbiotic was built around Limosilactobacillus fermentum K73, a bacterial strain the research team had previously isolated from artisanal "Costeño" cheese made in Colombia.

References

1. Katherine Bauer-Estrada, Karen M. Mancera Azamar, Caoyuanhui Wang, et al. (2026). Limosilactobacillus fermentum K73 synbiotic modulates gut microbiota function and behavior in gnotobiotic mice transplanted with microbiota from children diagnosed with autism spectrum disorder. Journal of Functional Foods. https://doi.org/10.1016/j.jff.2026.107508
2. Data and Statistics on Autism Spectrum Disorder. cdc.gov. https://www.cdc.gov/autism/data-research/index.html
3. Ygor Parladore Silva, Andressa Bernardi, Rudimar Luiz Frozza (2020). The Role of Short-Chain Fatty Acids From Gut Microbiota in Gut-Brain Communication. Frontiers in Endocrinology. https://doi.org/10.3389/fendo.2020.00025
4. Zhiqiang Zhang, Wenkai Kang, Yu Mi, Xin Zhong, Yulong He (2025). The Microbiota–Gut–Brain Axis in Autism: Associations, Causal Inference, and Interventions—A Narrative Review. Pathogens. https://doi.org/10.3390/pathogens14111145