Key Takeaways
- Deoxycholic acid suppresses CD8+ T cells, while 3-oxo-delta-4,6-lithocholic acid activates them, affecting tumor progression.
- The gut microbiome sends molecular messengers that influence immune responses, which can either promote or inhibit diseases.
- Gut microbial metabolites, such as short-chain fatty acids and bile acids, significantly affect immune cell behavior.
- Future treatments may target the entire metabolite profile, rather than individual molecules, to avoid shifting other metabolite levels.
- Research shows strong links between gut metabolites and various diseases, including cancers and inflammatory conditions.
Take a single molecule. A secondary bile acid called deoxycholic acid, made when certain gut bacteria chew through a cholesterol-derived compound your liver has secreted. This molecule, produced in the colon, ends up in the bloodstream. It reaches CD8+ T cells, the immune system’s dedicated tumour killers. And it damps them down, binding to a pump on the cell surface and interrupting the calcium signalling that would otherwise trigger them to release their payload of immune weaponry. In people with colorectal cancer, the result is a measurable increase in tumour progression.
Now take the same class of molecule. A structurally related bile acid, 3-oxo-delta-4,6-lithocholic acid, works on the same CD8+ T cells. Same cell type, same tissue. This one does the opposite: it activates androgen receptors and drives those cells deeper into tumours, improving the odds that immunotherapy drugs will work.
That’s the puzzle at the centre of a comprehensive new review from researchers at Peking University, published this month in Immunity & Inflammation. The gut microbiome isn’t just a digestive organ that’s tangentially connected to health. It runs what amounts to a chemical diplomacy operation, dispatching hundreds of molecular messengers that land on immune cells throughout the body and can either inflame or pacify, protect or undermine, depending on which cell they find, in what concentration, under what disease conditions. And right now, we’re only beginning to map the territory.
They are chemical compounds produced by bacteria in the gut, either by fermenting food you eat or by modifying compounds your own body produces. Short-chain fatty acids from dietary fibre, bile acids transformed from cholesterol-based compounds, and tryptophan derivatives are among the most studied. Many can enter the bloodstream and act on organs far from the gut.
They bind to receptors on immune cells and alter how those cells behave. Depending on the metabolite, the concentration, the immune cell type, and the disease context, the effects can range from dampening inflammation to boosting anti-tumour killing to amplifying autoimmune responses. The same molecule can have opposite effects on different cell types.
Potentially, though not yet in clinical practice. Current evidence is largely from mouse models. The challenge is that gut metabolites form an interconnected web, so targeting one tends to shift others. Researchers at Peking University suggest that future therapies may need to target the whole metabolite profile rather than individual molecules.
The review covers inflammatory bowel disease, colorectal cancer, liver cancer, pancreatic cancer, tuberculosis, psoriasis, polycystic ovary syndrome, autoimmune conditions including uveitis and hepatitis, and multiple sclerosis, among others. The breadth reflects how widely gut bacteria communicate with the rest of the body.
The messengers fall into a few main families. Short-chain fatty acids, made when gut bacteria ferment dietary fibre, are probably the most studied. Bile acids, modified by gut bacteria from cholesterol-derived compounds the liver produces, are emerging as equally significant. Derivatives of the amino acid tryptophan constitute a third class, doing much of their work through a receptor called the aryl hydrocarbon receptor, which turns out to be expressed on a surprising number of immune cells. Beyond these, the microbiome produces vitamins (folate, in particular, keeps a population of regulatory T cells alive in the colon), trimethylamine N-oxide from dietary choline and carnitine, and an assortment of odd-chain fatty acids that researchers are still trying to characterise. The library, Changtao Jiang’s team notes, is expanding rapidly, driven by newer metabolomic tools like click chemistry-based approaches that can tag and track previously invisible molecules.
What the review maps, systematically, is how each metabolite family interacts with specific immune cell populations. The interactions are not simple. Short-chain fatty acids, for instance, turn up in several contradictory-seeming stories. Butyrate suppresses the antigen-presenting capacity of dendritic cells, essentially telling the immune system to stand down, which can help with inflammatory conditions like colitis but also, in cancer contexts, may weaken the immune attack on tumours. Yet butyrate also enhances the killing function of CD8+ T cells directly, by inhibiting enzymes called histone deacetylases. How those two effects balance out in any given patient probably depends on local concentrations, which part of the body we’re talking about, and what disease is already established. The authors are direct about this: the same metabolite can push the immune system in completely opposite directions depending on context.
Some of the most striking examples involve conditions not usually associated with gut bacteria at all. Psoriasis, for one. The tryptophan-derived metabolite indoxyl sulfate, produced in the gut and then modified by the liver, accumulates in skin tissue where it pries open chromatin in Th17 cells, amplifying the inflammatory cascade responsible for the disease’s characteristic plaques. It’s a chain of events that starts with which bacteria are fermenting tryptophan in the colon and ends at the surface of the skin, several organs and metabolic transformations later.
Polycystic ovary syndrome is another. GDCA, a bile acid produced by gut microbes, drives a type of innate lymphoid cell to secrete IL-22, a cytokine that then acts at the ovary and liver to correct the hormonal and metabolic disruptions of PCOS in mouse models. The gut-to-ovary signalling axis this implies isn’t something most endocrinologists would have predicted fifteen years ago.
Lung immunity, too. When mice receive BCG vaccination, the vaccine shifts gut microbial communities in ways that raise blood levels of butyryl carnitine. This metabolite, working through mechanisms still being unravelled, helps train memory macrophages in the lung, creating what amounts to a reinforced immune memory for tuberculosis. The vaccination, in other words, protects the lung partly by going through the gut first.
For cancer immunotherapy, the implications are substantial, though still largely preclinical. TMAO, derived when gut bacteria process choline and carnitine from meat and eggs, primes macrophages in pancreatic tumours to release interferon, making checkpoint inhibitor drugs more effective, at least in mice. Inosine, produced by Bifidobacterium, enters the bloodstream during immunotherapy and activates Th1 cells in ways that amplify the treatment’s effects. Formate, produced by certain Lachnospiraceae, strengthens CD8+ T-cell responses through a transcription factor called Nrf2, and this may partly explain why exercise improves cancer outcomes. The mechanism linking your morning run to tumour immunity might run, somewhat improbably, through gut fermentation.
The catch, and it is a significant one, is that clinical translation is almost entirely absent. The review maps a vast mechanistic landscape built largely on mouse models and in vitro experiments. Translating any of this into treatment is complicated by the same duality that makes the biology interesting: if you increase one metabolite, you’ll inevitably shift the levels of others, some of which may do exactly what you were trying to prevent. Targeting a single molecule is, in the authors’ framing, probably the wrong approach. The goal should be reshaping the whole metabolite profile, which is harder to do and harder to measure.
The tools for doing that are improving. Reverse metabolomics, which works backwards from a molecule’s biological effect to identify which bacteria produced it and via which enzyme, is accelerating the discovery of novel metabolites. AI-assisted prediction of enzyme function is helping identify candidates that would take years to characterise manually. The picture emerging is less a catalogue of individual messengers and more a kind of chemical grammar, with rules that the immune system has learned to read over millions of years of co-evolution with its microbial residents. We’re still working out what most of the sentences say.
DOI: 10.1007/s44466-026-00031-7
ScienceBlog.com has no paywalls, no sponsored content, and no agenda beyond getting the science right. Every story here is written to inform, not to impress an advertiser or push a point of view.
Good science journalism takes time — reading the papers, checking the claims, finding researchers who can put findings in context. We do that work because we think it matters.
If you find this site useful, consider supporting it with a donation. Even a few dollars a month helps keep the coverage independent and free for everyone.