Akkermansia muciniphila: the microbe that eats your mucus

Start with the organism, not the headline. In 2004, microbiologists at Wageningen University did something deliberately unkind to a sample of human faeces: they diluted it until almost nothing survived, then offered the survivors a single food — gastric mucin, the protein that makes mucus slippery ¹. No sugar, no starch, no fibre. Eat the lining, or starve.

Something ate the lining. It was a Gram-negative, strictly anaerobic, non-motile, non-spore-forming oval cell that grew singly and in pairs, and its closest described relative shared only 92 % of its 16S ribosomal RNA gene sequence — distant enough to need a genus of its own ¹. They called it Akkermansia muciniphila: the mucus-lover. That is not a metaphor. It is the job description.

What it is, in one sentence

Akkermansia muciniphila is a strictly anaerobic, mucin-degrading bacterium of the phylum Verrucomicrobiota that lives inside the mucus layer of the human intestine and feeds on the mucin its host secretes there ¹¹⁰. The type strain is MucT, deposited as ATCC BAA-835 ¹.

Where it lives in the body

Not in the middle of the gut contents, where most famous gut bacteria are counted, but in the layer just above the cells — the mucus blanket the intestinal wall renews continuously. That address explains everything else about this organism. When the original team built a genetic probe and counted the species in stool, it came out at more than 1 % of all bacterial cells ² — which, for one species in an ecosystem of hundreds, is a great deal. And now the caveat, out loud: a stool count is a proxy. The bacterium's real home is the mucus; stool is what leaves the building.

What it does for us

Here the evidence gets interesting, and here I want named studies rather than enthusiasm.

It tracks with a leaner, healthier metabolism. In 2013, a group at UCLouvain and Wageningen showed that the abundance of A. muciniphila falls in obese and type-2-diabetic mice, that prebiotic feeding restores it, and that giving the live bacterium to mice on a high-fat diet reversed fat-mass gain, metabolic endotoxaemia, adipose-tissue inflammation and insulin resistance ³. Note the detail most summaries skip: heat-killed cells did nothing ³. That is what an honest mechanism looks like — one that can fail.

There is a molecule behind it. In 2017, an abundant outer-membrane pili-like protein called Amuc_1100 was purified and shown to signal through Toll-like receptors 2 and 4, to drive interleukin-10, and to raise trans-epithelial resistance — a laboratory measure of how tight an epithelial barrier is ⁴. Mechanism first: a bacterium sitting in the mucus, speaking to the immune receptors on the cells beneath it, and the barrier tightening in reply.

It has been given to people once, carefully. The 2019 proof-of-concept trial in Nature Medicine enrolled 40 overweight, insulin-resistant volunteers; 32 completed three months of daily supplementation — live bacteria, pasteurised bacteria, or placebo ⁵. It was safe and well tolerated. The pasteurised form improved insulin sensitivity by 28.62 % (P = 0.002), reduced insulinaemia by 34.08 % and total cholesterol by 8.68 % ⁵. Now the caveat at full volume: body weight fell by 2.27 kg against placebo and fat mass by 1.37 kg against the group's own baseline, and neither reached statistical significance (P = 0.091 and P = 0.092) ⁵. This was a single-centre exploratory pilot in 32 people whose primary endpoints were safety, tolerability and metabolic parameters ⁵. It is a reason to run a bigger trial. It is not a result to build a diet on.

It travels with cancer-immunotherapy response. In the 2018 Science study of patients treated with PD-1 blockade, the relative abundance of A. muciniphila in stool at diagnosis tracked with clinical response, and oral supplementation restored the response in mice that had received stool from non-responding patients ⁹. Correlation in humans; causation in mice. Both halves of that sentence matter.

What it does against us

The same act, read the other way. A bacterium that eats mucus is thinning the barrier that keeps your immune system and your gut contents apart. Whether that ends well depends on the neighbourhood.

With a pathogen present, it makes things worse. In gnotobiotic mice carrying a defined eight-species community, adding A. muciniphila to a Salmonella Typhimurium infection raised histopathology scores and the messenger RNA of IFN-γ, TNF-α, IL-6, IL-12 and IL-17 in caecum and colon ⁶. Salmonella counts in the mesenteric lymph nodes rose about tenfold, and mucin-filled goblet cells in caecal tissue fell to between a third and a half of the numbers seen in the other groups ⁶. The authors' own reading: it disturbs the host's mucus homeostasis ⁶.

Without fibre, the whole community turns on the lining. The 2016 Cell study gave gnotobiotic mice a synthetic human microbiota and then withdrew dietary fibre. Deprived of plant carbohydrate, the community — with mucus specialists such as Akkermansia prominent in it — switched to host mucus glycoproteins as a nutrient source, eroded the colonic mucus barrier, and let the mucosal pathogen Citrobacter rodentium reach the epithelium and cause lethal colitis ⁷. That is the practical lesson of this entire entry: a mucus eater eats mucus mainly when nothing better is on offer.

Its abundance is not a health score. A pooled, methodologically harmonised analysis of gut-microbiota studies in Parkinson's disease found Akkermansia raised in patients, alongside reduced butyrate producers ⁸. Higher, not lower. The authors propose degradation of the mucus layer as one plausible link ⁸ — but this is an association in cross-sectional data, and slow gut transit, constipation and medication all remain live alternative explanations. Nobody has shown which way the arrow points.

One organism, two spellings: why the phylum name changed in 2021

The species name has not changed since it was proposed in 2004 ¹. The phylum around it has. What every older paper calls Verrucomicrobia was validly published under the bacterial nomenclature code as Verrucomicrobiota in 2021, when 42 phylum names were formally established at once ¹⁰. Same organisms, current spelling. So if you read Verrucomicrobia in a 2013 abstract and Verrucomicrobiota in a 2023 one, nothing was reclassified — the rules caught up.

What to hold on to

Akkermansia muciniphila is a normal resident of a healthy human gut that makes its living from the mucus you produce. Trimming that layer appears to stimulate its renewal and tighten the barrier, and in mice and in one small human trial that showed up as better metabolic numbers ³⁵. Strip the diet of fibre, or add a pathogen — a microbe that causes disease — and the same appetite becomes erosion ⁶⁷. The human evidence is a 32-person pilot and a set of correlations — things that travel together, not yet cause and effect ⁵⁹. That is genuinely promising, and it is genuinely unsettled, and saying both at once is the only honest way to write this entry.