When Microbes 'Rewire' Behavior: The Role of Brain CD4+ T Cells in Autism

A paper by a Korean team published in Nature Communications linked three “nodes” into a single chain: gut microbes → brain immune cells → behavioral symptoms in a model of autism spectrum disorders (ASD). The scientists showed that in BTBR mice (a classic genetic model of ASD), the absence of microbiota alleviates autism-like behavioral manifestations and reduces the number of inflammatory brain T cells. And targeted depletion of CD4+ T cells normalizes neuroinflammation and behavior. In parallel, they found a “harmful” inhabitant of the intestine that increases the excitatory shift in the metabolism of neurotransmitters (↑glutamate/GABA and ↑3-hydroxyglutaric acid), and identified the probiotic strain Limosilactobacillus reuteri IMB015, which is capable of shifting metabolism in the opposite direction and improving a number of behavioral tests. The result is a functionally confirmed gut-immune-brain axis in the context of ASD.

Background of the study

Autism spectrum disorders (ASD) are a heterogeneous group of conditions in which behavioral features (social communication, repetitive behaviors, sensory hypersensitivity) are often combined with gastrointestinal symptoms and signs of immune activation. It is this “triangle” - gut, immunity, brain - that has attracted particular attention in recent years: more and more data link the composition of the microbiota and its metabolites with neurodevelopment, neuroinflammation, and the balance of excitatory/inhibitory signals in the central nervous system.

The gut-brain axis concept includes several pathways. Neuronal - via the vagus nerve and enteric nervous system; immune - via cytokines, microglial status and lymphocyte migration/residence; metabolic - via short-chain fatty acids, tryptophan derivatives, bile acids and amino acids (including glutamate/GABA). In ASD models, the key hypothesis remains the excitation/inhibition (E/I) imbalance, which can be maintained by both altered synaptic plasticity and the "background" inflammatory environment.

A separate topic was the involvement of adaptive immunity in the brain. If the brain was previously considered “immune-privileged,” today it has been shown that meningeal and parenchymatous T cells (including CD4+) are able to modulate the work of microglia, synaptic pruning, and behavior. At the intersection with the microbiota, this opens up a simple but powerful scenario: intestinal microbes restructure the pool of metabolites and immune signals → the profile of brain T cells and microglia changes → behavioral phenotypes shift.

Practical interest in strain-specific interventions has grown after a number of preclinical studies where individual lactobacilli affected social tests in mice, and transplantation of microbiota from “healthy” animals mitigated autism-like manifestations. However, full mechanistic links “specific microbe → specific metabolites → specific immune cells in the brain → behavior” are still scarce. Recent studies are filling this gap by constructing a causal chain and proposing testable targets - from “harmful” taxa to candidate probiotics and immune nodes (CD4+, IFN-γ), which can be validated in future clinical trials.

How was this tested?

The authors created a germ-free version of BTBR and systematically compared it with standard animals (SPF). Behavior was assessed by “social” tests (three-chamber setup with novelty test), repeated manipulations (ball burial) and anxiety/hyperactivity (open field). Next, immunology (CD4+ antibody depletion, profiling of brain lymphocytes and microglia), microbiology (16S sequencing, colonization with isolated strains) and targeted fecal metabolomics were applied. Finally, a probiotic candidate was selected through genome-scale metabolic models (flux-balance) and tested in mice.

Key Findings

The bottom line is that there are four main results:

• Microbiota ↔ behavior. In germ-free BTBR males, some of the autism-like phenotypes disappeared: better social novelty, less repetitive behavior, signs of anxiety normalized; a decrease in neuronal activity in the amygdala and dentate gyrus (c-Fos) also coincided.
• The crucial role of CD4+ T cells. Selective depletion of CD4+ in the brain reduced proinflammatory signals, affected microglia, and improved behavioral tests (social memory, repetition, anxiety) without changing overall motor activity.
• "Harmful" and "beneficial" microbes. Lactobacillus murinus was isolated from the BTBR gut, mono-association of which in germ-free mice increased repetitiveness, increased glutamate/GABA and 3-hydroxyglutaric acid, as well as the proportion of IFN-γ+ T cells in the brain - a picture of neuroinflammation. In contrast, transplantation of "healthy" microbiota from regular B6 reduced excitatory shift and neuroinflammation.
• Probiotic candidate. In a computational screen for “GABA-producing and glutamate-scavenging capacity,” the L. reuteri strain IMB015 stood out. Its course: reduced glutamate and the glutamate/GABA ratio, reduced 3-hydroxyglutaric acid, attenuated neuroinflammation (↓IFN-γ+ CD4+ T cells), and improved behavior (less repetition; better social novelty). The effect on “sociability” per se was incomplete.

How it can work

The study brought together three well-studied mechanisms and showed that they “link” with each other: (1) Gut microbes set metabolite pools - “harmful” strains have predominantly glutamate and 3-hydroxyglutaric acid, increasing the excitatory background (E/I imbalance). (2) These signals - both through the vagus/circulating mediators and through border immune links - shift the state of brain CD4+ T cells to a pro-inflammatory profile with the participation of IFN-γ, affecting microglia. (3) Neuroinflammation and E/I imbalance in specific structures (amygdala, hippocampus) are translated into social and perseverative manifestations. The reverse intervention - removing the “harmful” strain or adding a strain that reduces Glu/GABA and 3-OH-glutaric - weakens the symptoms.

Why is this important?

The work translates the debate about the “gut-brain axis” in ASD into the language of specific cells and metabolites: brain CD4+ T cells are the critical mediator, and glutamate/GABA and 3-hydroxyglutaric acid are measurable “arrows” of the state. In addition, these are not just correlations, but functional tests: deplete CD4+ → behavior changes; add L. murinus → worse; give L. reuteri IMB015 → better. This strengthens the argument for targeted microbial therapy as a complement to behavioral and pharmacological approaches, albeit only in the preclinical setting.

What does this mean in practice?

• It doesn't "treat autism," but it finds targets. We're talking about mice and machines; transferring it to humans will require staged RCTs.
• Biomarkers for tracking: The glutamate/GABA ratio and fecal 3-OH-glutaric acid levels appear to be candidates for monitoring the effects of microbial interventions.
• The "subtract plus add" strategy. It is promising to simultaneously reduce "harmful" taxa and maintain protective ones (strain-specific), focusing on the metabolic profile.

Limitations that the authors themselves talk about

This is an animal model with a focus on male BTBR; the mouse behavior is only an approximation of human symptoms. “Bad” and “good” effects are shown in individual strains and under controlled colonization conditions; in a real microbiome, the interactions are orders of magnitude greater. Finally, even for IMB015, not all tests improved at once - “sociability” responded weaker than social memory and perseveration. Clinical steps are needed - from safety to doses and duration, and careful stratification (gender, age, ASD phenotype, concomitant GI symptoms).

What will science do next?

The authors outline practical tracks:

• Strain-specific RCTs in people with ASD with behavioral and neuroinflammatory endpoints, plus microbiota and metabolite 'omics'.
• Immune-driven approaches: targeting CD4+ T cells/their cytokines in the brain (without systemic immunosuppression) as a possible adjuvant strategy.
• Microbial consortia optimized for Glu/GABA and 3-OH-glutaric acid reduction with proven colonization and stability.

Source: Park JC et al. Gut microbiota and brain-resident CD4+ T cells shape behavioral outcomes in autism spectrum disorder. Nature Communications 16, 6422 (2025). https://doi.org/10.1038/s41467-025-61544-0