How Gut Bacteria Make Brain Chemicals From Your Food

For years we said the gut talks to the brain and waved our hands at how. In 2026, two papers got specific: one where bacteria build a brain chemical out of your food, and one where a bacterial molecule leaves the gut, travels to the brain, and helps it restore its own calming signal.

Lit Review Friday · How Gut Bacteria Make Brain Chemicals From Your Food · 2026 · ~15 min read

The Wire We Already Built

A while back, we spent an episode on neuropod cells: the rare gut cells that form actual synapses onto the vagus nerve and fire electrical signals to the brain in seconds, faster than any hormone in the blood. We called the vagus a fiber-optic cable, and the neuropod the gut's sensory operative. That episode built the wire.

It is worth being specific about what that wire actually is, because the specificity is the whole point. The cells doing the sensing are enteroendocrine cells, roughly 1% of the cells lining your gut. For a century they were filed under "hormone cells": the assumption was that the gut could only whisper to the brain slowly, by dribbling hormones into the bloodstream that the brain might pick up minutes to hours after a meal. Kaelberer and colleagues, in Science in 2018, showed that assumption was incomplete. A subset of these cells, the ones they named neuropod cells, do not whisper through the blood. They are electrically excitable, they carry the presynaptic hardware of a real neuron, and about two-thirds of enteroendocrine cells sit in direct contact with nerve fibers. The neuropod cells form a genuine synapse onto the vagus nerve and pass their signal using glutamate, a fast neurotransmitter, in milliseconds. One synapse connects the inside of your gut to your brainstem. Not a slow chemical rumor carried by the blood: a hardwired line, as fast as any circuit inside the brain itself.

That is the cable. This episode is about the messages traveling on it. Two 2026 papers, read side by side, describe two completely different languages the microbiome uses to talk to the nervous system. And they are more specific, and more honest, than the story most of us have been telling.

Is It True That 90% of Your Serotonin Is in Your Gut?

You have probably heard the headline: most of your serotonin is made in your gut, therefore your gut is dosing your brain with a feel-good chemical. The first half is roughly true. The conclusion is not.

Here is the actual mechanism, because it is more interesting than the headline. Most gut serotonin is made in a specialized enteroendocrine subtype called the enterochromaffin cell. Yano and colleagues showed in 2015 that these cells do not do it alone: specific spore-forming gut bacteria produce metabolites that reach into the enterochromaffin cell and turn up Tph1, the rate-limiting enzyme for building serotonin. Turn the enzyme up, and the cell makes more. The cleanest proof is the absence experiment. Mice raised germ-free, with no microbes at all, run low on gut and blood serotonin; give them back a normal microbial community and the serotonin comes back. The bacteria are setting the dial.

But look at where that serotonin goes. It works largely where it is made: on gut motility, on the platelets that manage clotting, on the local nerves in the gut wall. It does not stroll across the blood-brain barrier into your mood circuits. This is a beautiful microbe-to-host signal, and a local one.

We have run into this exact honesty problem before. In an earlier episode we covered work (Cao) where the serotonin in brain tissue was restored while the serotonin in the blood barely moved. A clinician looking at routine bloodwork would have seen nothing, even though something real was happening in the brain. Same lesson, pointed the other direction: where a signal lives matters as much as what it is. Making a neurotransmitter in the gut is not the same as delivering it to the brain. Hold that distinction, because it is exactly what separates our two papers.

Can Gut Bacteria Make a Brain Chemical From Your Food?

Paper one is Song and colleagues, published in Nature in 2026. The one-sentence version: specific gut bacteria take choline, a common nutrient in food, and convert it directly into acetylcholine, the neurotransmitter your brain uses for memory and focus and voluntary movement. Acetylcholine has a special place in the history of biology: it was the first neurotransmitter ever discovered, the molecule that proved our nerves communicate in chemistry, not just electricity. And here, gut bacteria are making it, then using it to strengthen the gut's own immune defenses and keep the gut moving.

Before anything else: this is a mouse study. A mouse is not a person. We say that early and we mean it.

What They Did

To find out which bacteria were talking to which human receptors, the team ran a screen called PRESTO-Salsa, a kind of molecular speed-dating. They took the chemical output of about 100 diverse gut commensal strains and tested it against more than 300 human cellular receptors, looking for the moment a bacterial product turns a human lock. Acetylcholine lit up the board, produced by strains including Bifidobacterium breve and the probiotic Pediococcus pentosaceus.

Then they proved it three ways. They changed the diet by supplementing choline. They changed the bacteria, comparing a normal B. breve against a mutant engineered to lack the single enzyme it needs to make acetylcholine. And they blocked the host receptors to see what happened when the body could not hear the signal, measuring gut motility, IgA antibody production, and resistance to Salmonella.

What They Found

The chain they drew is clean: dietary choline enters the gut, specific microbes grab it and convert it to acetylcholine, and that acetylcholine docks on host muscarinic and nicotinic receptors in the gut tissue. Gut motility was pinned to one specific muscarinic receptor, CHRM3: block that lock, and the motility effect stops entirely. The immune arm, the rise in protective IgA antibodies, is likely a nicotinic receptor, though the exact one is still inferred rather than nailed down. Good science tells you which parts it proved and which parts it suspects.

One clarification matters, because acetylcholine is famous as a nerve signal. The bacteria are not wiring into your neurons or firing them. They are making the nervous system's own molecule and using it to speak directly to two kinds of gut cells: the smooth muscle that drives motility, and the immune cells that make antibodies. One molecule, two local jobs, no nerve required. That is also why the effect stays in the gut, which we come back to below.

The Number That Stopped Me: Breast Milk

Here is the detail I keep turning over. Human breast milk is flooded with choline. And a newborn's gut is dominated by the acetylcholine-producing Bifidobacterium species that turn choline into this signal. Read those two facts together and you get an evolutionary handshake: the parent's milk delivers the raw material, the infant's resident microbes turn it into the signal, and that signal helps build the newborn's gut immune system. Nobody designed that on a whiteboard. It emerged.

The Same Egg, Two Destinations

Here is a callback that gives me chills. A couple of episodes ago, we followed choline down a darker road: a different gut microbe converts the same nutrient into TMA, which the liver turns into TMAO, a molecule tied to inflammation and cardiovascular risk. Same nutrient. One community of microbes makes a molecule linked to harm. Another makes a brain chemical that builds your immune defenses.

This is the through-line of almost everything we cover: the poison is the context, not the compound. The food is raw material. The microbes are the translators. Take care of the translators.

Does Fiber Actually Reach the Brain?

Paper two is the one that leaves the gut. Zou and colleagues, in Microbiome, 2026. They fed mice the prebiotic fiber inulin and found it reduced cognitive impairment by feeding bacteria that make short-chain fatty acids, molecules that then physically traveled to the brain and restored the brain's own calming signal.

I will admit I pushed back on this when I first read it. I had always filed short-chain fatty acids under local gut business: fuel for the colon, a nudge to inflammation. The idea that they ride the bloodstream to the brain and dock on receptors there was not my mental model. The data changed my mind.

What They Did, and How They Closed Every Loophole

The model matters, so read it carefully: this is a surgical model called bilateral carotid artery stenosis, which restricts blood flow to the brain to mimic vascular cognitive impairment. It is not natural aging, and it is not Alzheimer's. The mice were adult and male only.

What makes the paper compelling is the layering of proof. The chain runs: inulin feeds fatty-acid-producing microbes (Lactobacillus and Akkermansia), those microbes pump out acetate, propionate, and butyrate, the fatty acids enter the blood, travel to the medial prefrontal cortex, and dock on two receptors, GPR41 and GPR43. That docking restores GABA signaling, the brain's main inhibitory, calming transmission, and cognition and anxiety both improve.

There is a deeper layer under the GABA end of this story that is worth naming. Zou's mice restored their own brain GABA after the fatty-acid signal arrived. But bacteria are already fluent in GABA directly. Strandwitz and colleagues showed in 2019 that the human gut runs an actual GABA economy: some bacteria produce it, and others depend on consuming it to grow. They found one such GABA-eater so hungry for it that it would only grow in a dish when a GABA-producing neighbor was feeding it. This is not a brain result, it is gut microbiology, and that is exactly why it matters here. It says the microbiome is not a passive supplier of a raw ingredient. It is a community actively making and trading the calming molecule, setting the baseline the rest of the system inherits. Zou shows the brain restoring its GABA signal; Strandwitz shows how deep the microbial fluency in that same molecule runs.

Two Languages, One Conversation

Put the papers side by side and a beautiful distinction appears. In Song, bacteria build an actual neurotransmitter, acetylcholine, and use it locally in the gut. In Zou, bacteria build a molecule that is not a neurotransmitter at all, a short-chain fatty acid, and that molecule travels to the brain and tells the neurons to repair their own signaling. One makes the chemical and keeps it home. The other sends a molecule that prompts the brain to restore its own signaling.

Local production versus long-distance repair. That is a far more honest, and far more interesting, picture than the flat claim that the gut makes brain chemicals. Some messages stay home. Some make the trip. The whole game is knowing which, and this is the year the field started to actually know.

It also sharpens the neurotransmitter story we have told before. Gut bacteria have long been known to make the precursors and building blocks: tryptophan on the way to serotonin, glutamate that partners into GABA. Song is a genuine step past that. It is a microbe making the finished neurotransmitter itself. And Zou shows a separate class of molecule commandeering a brain receptor to restore a neurotransmitter the host makes. The vocabulary of the gut-brain conversation is bigger than we thought.

How This Builds on What We've Covered

If you have followed these letters for a while, you can feel the picture assembling piece by piece. The neuropod episode gave us the wire: the physical, millisecond-fast synapse from gut to brainstem. An earlier episode gave us the honest caution about serotonin: a signal restored in brain tissue that a blood test would have missed, a reminder that location is not a detail. And we have said, more than once, that the microbes are the translators in the middle, that the active ingredient is the signal and not the cell.

Song and Zou are not a new story. They are the next specific paragraph of the same one. Song hands us a microbe building a finished neurotransmitter out of food, going a step past the precursor chemistry we described before. Zou hands us one of these molecules using a known door into the brain to reach a named receptor. Same thesis we keep landing on, now with the receptors labeled and the doors identified. The gut-brain axis was never mystical. It is molecules, receptors, and a wire, and the field is finally specific enough to point at each one.

Four Roads Out of the Gut

Step back from the two papers for a second and look at the whole map. When Cryan and colleagues wrote the definitive review of the microbiota-gut-brain axis in 2019, they were not describing one channel. They were describing a switchboard. Broadly, the gut and brain talk over four kinds of road: the neural road (the vagus wire we just walked), the immune road (bacterial signals tuning inflammation, which the brain reads), the endocrine road (hormones and the stress axis), and the metabolite road (the actual small molecules bacteria make, moving through the blood). Most real signals use more than one road at once.

That map is what makes our two papers legible. They are not competing theories of "how the gut talks to the brain." They are two different roads, correctly labeled.

Song's acetylcholine is a metabolite-road signal that never leaves town. The bacteria make the molecule and it acts on local receptors in the gut wall. Real, specific, and local.

Zou's short-chain fatty acids are a metabolite-road signal that goes the distance. And here is where the supporting literature earns its place, because "a molecule travels to the brain" is the exact claim that deserves scrutiny. How would a short-chain fatty acid even get there? Reviews by Silva and colleagues (2020) and Dalile and colleagues (2019) lay out the honest answer: there is more than one route, and the field is still weighing how much each one carries. Some of the signal is indirect, short-chain fatty acids nudging the same vagal wire, or shifting immune and hormonal tone that the brain then reads. And some of it looks direct: short-chain fatty acids can cross the blood-brain barrier using dedicated shuttles called monocarboxylate transporters, the same class of doorway the brain uses to import other fuels. So when Zou reports the fatty acids reaching the prefrontal cortex, that is not a mechanistic miracle. It is a known door. What was new in Zou was catching a specific molecule using it to reach a specific receptor and change a specific circuit.

What You Actually Eat to Feed These Pathways

It all starts with raw materials. The Song pathway runs on choline, concentrated in eggs and meat (and, for infants, in breast milk). The Zou pathway runs on inulin, a fiber found in everyday foods like chicory root, garlic, onions, and asparagus. These are exactly the compounds a modern ultra-processed diet strips out first.

And this is not just mouse abstraction. In people, a 2023 randomized, double-blind, placebo-controlled trial (Nishida et al.) found that Bifidobacterium plus inulin improved measures of cognitive function in older adults with mild cognitive impairment, alongside a rise in gut bifidobacteria. Human data on this exact fiber-plus-microbe idea exists, and it points the same direction, while still being early.

Why Eating the Right Food Isn't Enough

Here is the part the supplement and wellness world tends to get wrong. The precursor is necessary, but it is not sufficient. You can eat all the choline and all the fiber in the world. What actually decides what gets built is the environment those microbes are working in, and which microbes are there to do the work.

Choline is the cleanest example, and this story showed both of its endings. Same molecule. In the right microbial context, it becomes acetylcholine: immune-supportive, keeping the gut moving. In the wrong one, a different microbe turns it into TMAO, the pro-inflammatory metabolite associated with cardiovascular risk. Same input, opposite output. The environment picks which.

Fiber tells the same story. Feed the right community and you get the short-chain fatty acids that calm the brain. But if the gut is already inflamed, sitting in a stressed, oxidative metabolic state, those microbes cannot do the metabolism cleanly. They cannot break the fiber down the way you need, and instead of the calming signal you can end up with pro-inflammatory byproducts. You did the right thing, you ate the fiber, and a broken environment turned it against you.

So the real goal is not a single nutrient. It is a balanced ecosystem, one whose metabolism is healthy enough that when you feed it the right thing, it makes the right thing. How you repair that environment when it has drifted is a bigger question, and a subject for another day.

Where the Postbiotic Thesis Fits

Look at what actually carried the effect in both papers. It was not the presence of a live bacterium as an end in itself. It was the molecule. Zou made the point most cleanly: the team reproduced the brain benefit by giving the short-chain fatty acids directly, no bacteria required. The signal did the work.

That is the whole idea behind postbiotics. If the value lives in the chemistry a healthy microbial community produces, then when that community is depleted, delivering the signals directly becomes a logical intervention. A full-spectrum postbiotic derived from healthy human donor microbiomes aims to capture two things a single strain cannot: the emergent chemistry that only appears in a functional community, and the metabolic environment that lets your resident microbes do their own work. The microbiome speaks eloquently in chemistry, more than 10,000 molecular signals worth, and that vocabulary is what we are trying to preserve.

We believe this is a plausible mechanism. It is a thesis, not a proven clinical outcome. We have not run a trial of ThaenaBiotic on cognition, and we would not claim one. What these two papers give us is a reason the signal matters. ThaenaBiotic is one tool in your toolbox, a layer that sits alongside real food and a diverse microbiome, not a replacement for either.

The Honest Limitations

Both anchor papers are in mice. Song's effects are local to the gut and do not touch the brain in the data shown. Zou's model is a surgical vascular injury, male mice only, and cannot be read as human aging or dementia. The two most tempting overreaches are worth naming out loud so nobody makes them: no, this does not mean choline supplements treat Alzheimer's, and no, this does not mean eating inulin treats human dementia. The literature is consistent with a real and specific gut-brain chemistry. It is not a claim any single trial has proven in people.

Frequently Asked Questions

Do gut bacteria really make neurotransmitters?

Yes. Song et al. (2026, Nature) showed gut bacteria convert dietary choline into acetylcholine, a finished neurotransmitter, and earlier work (Strandwitz et al., 2019) showed gut bacteria that produce and consume GABA. The catch is location: making a neurotransmitter in the gut does not mean it reaches the brain.

Can eating more fiber improve brain function?

In mice, inulin fiber improved cognition through a microbe-to-short-chain-fatty-acid-to-brain-receptor pathway (Zou et al., 2026). A human trial of inulin plus Bifidobacterium also improved cognitive measures in older adults with mild cognitive impairment (Nishida et al., 2023). Promising and mechanistically specific, but still early, and not a treatment claim.

Is the 90%-of-serotonin-is-in-the-gut claim accurate?

It is widely repeated and misleading. Most gut serotonin acts locally, on motility and local nerves, and does not cross into the brain (Yano et al., 2015). It is a good example of why the location of a signal matters as much as its identity.

What foods feed the gut-brain pathways in these studies?

Choline (eggs, meat) feeds the acetylcholine pathway; inulin fiber (chicory root, garlic, onions, asparagus) feeds the short-chain fatty acid pathway. Both are among the first things a highly processed diet removes.

Do you need live bacteria, or just the molecules they make?

In Zou et al., giving the short-chain fatty acids directly reproduced the brain benefit without the bacteria. That is the postbiotic principle: the active ingredient is often the signal, not the living cell.

Is eating more fiber or choline enough to get these benefits?

Not by itself. The same precursor can become a helpful or a harmful metabolite depending on the microbial environment: choline becomes immune-supportive acetylcholine in one context (Song et al., 2026) and pro-inflammatory TMAO in another. The health and metabolic state of the gut ecosystem, not just the food on your plate, decides the output.

The Bottom Line

If your bacteria manufacture your neurotransmitters and your calming signals based on what you feed them, it is worth asking who is really in the driver's seat. Even when you do everything right, the ecosystem gets a vote. Not good bug or bad bug. Not gut over brain or brain over gut. A conversation, in a chemical language, that you get to feed.

References

1. Song D, Duncan-Lowey B, Khetrapal V, et al. Commensal-derived acetylcholine enhances mucosal immune education. Nature. 2026;655:1027-1036. https://doi.org/10.1038/s41586-026-10592-7 · FREE FULL TEXT
2. Zou Z, et al. Inulin alleviates vascular cognitive impairment by restoring prefrontal GABAergic function via the microbiota-SCFA-GPR41/43 axis. Microbiome. 2026. https://doi.org/10.1186/s40168-026-02469-8 · FREE FULL TEXT
3. Kaelberer MM, Buchanan KL, Klein ME, et al. A gut-brain neural circuit for nutrient sensory transduction. Science. 2018;361(6408):eaat5236. https://doi.org/10.1126/science.aat5236 · FREE FULL TEXT
4. Yano JM, Yu K, Donaldson GP, et al. Indigenous bacteria from the gut microbiota regulate host serotonin biosynthesis. Cell. 2015;161(2):264-276. https://doi.org/10.1016/j.cell.2015.02.047 · FREE FULL TEXT
5. Strandwitz P, Kim KH, Terekhova D, et al. GABA-modulating bacteria of the human gut microbiota. Nature Microbiology. 2019;4(3):396-403. https://doi.org/10.1038/s41564-018-0307-3 · FREE FULL TEXT
6. Nishida Y, et al. Effect of continuous ingestion of bifidobacteria and dietary fiber on improvement in cognitive function: a randomized, double-blind, placebo-controlled trial. Nutrients. 2023;15(1):234. https://doi.org/10.3390/nu15010234 · FREE FULL TEXT
7. Silva YP, Bernardi A, Frozza RL. The role of short-chain fatty acids from gut microbiota in gut-brain communication. Frontiers in Endocrinology. 2020;11:25. https://doi.org/10.3389/fendo.2020.00025 · FREE FULL TEXT
8. Dalile B, Van Oudenhove L, Vervliet B, Verbeke K. The role of short-chain fatty acids in microbiota-gut-brain communication. Nature Reviews Gastroenterology & Hepatology. 2019;16(8):461-478. https://doi.org/10.1038/s41575-019-0157-3
9. Cryan JF, O'Riordan KJ, Cowan CSM, et al. The microbiota-gut-brain axis. Physiological Reviews. 2019;99(4):1877-2013. https://doi.org/10.1152/physrev.00018.2018

This post accompanies the Lit Review Friday episode of Learn Something with Thaena.