Faecalibacterium prausnitzii: the microbe that fears air

Most bacteria get famous for what they do to people. This one got famous for going missing.

In 2008, a team led by Harry Sokol looked at the gut lining of people with Crohn's disease at the moment of surgery, and again six months later ¹. One species stood out by its absence. Where there was less of it on the resected ileal mucosa, endoscopic recurrence at 6 months was more likely ¹. The organism was Faecalibacterium prausnitzii — and it is not some exotic passenger. In a healthy adult it is one of the most common bacteria in the body.

What it is, in one sentence

Faecalibacterium prausnitzii is an obligately anaerobic, non-motile, butyrate-producing bacterium of the human large intestine — a normal resident, not a pathogen, and one of the most abundant single species in a healthy adult gut ²³.

Where it lives

In the colon, and essentially nowhere else in us. It is isolated from human faeces, and the 2002 description that gave it its current name called it one of the most abundant colonizers of the human colon ³. A 2013 review put a figure on that: more than 5 % of the total bacterial population of the healthy adult intestinal microbiota ².

Two details of its diet explain a lot about it. It is an obligate anaerobe — oxygen is not merely unhelpful to it, it is the end of it. And it requires acetate in its growth medium ³. Acetate is a waste product of other gut bacteria. So this organism sits downstream of its neighbours, eating what they discard, in the one place in the body where there is no oxygen and a steady supply of leftovers.

What it does for us

It makes butyrate. That sounds like a small chemical footnote until you ask what butyrate is for.

In 2011, Dallas Donohoe and colleagues compared germ-free mice with normal ones and found that colonocytes — the cells lining the colon — use bacterially produced butyrate as their primary energy source ⁵. Germ-free colonocytes sat in an energy-deprived state: reduced oxidative phosphorylation, lower ATP, and autophagy switched on. Adding butyrate rescued their mitochondrial respiration and stopped the autophagy, and it did so by acting as fuel rather than as an enzyme inhibitor ⁵.

So a butyrate producer is not a vague "good bacterium". It is a supplier to a specific tissue that has outsourced its own catering.

There is a second job, and here the evidence is more interesting than the slogan. Sokol's group did not only count the species; they tested it. Stimulating human peripheral blood mononuclear cells with F. prausnitzii produced significantly lower IL-12 and IFN-γ and higher IL-10 — a calmer, less inflammatory profile ¹. In a reporter cell line, the bacterium itself did nothing to IL-1β-driven NF-κB activity, but its cell-free supernatant abolished it. Feeding mice either the live bacterium or its supernatant markedly reduced the severity of chemically induced colitis ¹. In other words, part of the anti-inflammatory effect is a molecule the organism secretes, not the organism. One such molecule was later identified in that supernatant: a 15 kDa protein named MAM — microbial anti-inflammatory molecule — which, expressed in intestinal epithelial cells, decreased NF-κB activation, and which prevented chemically induced colitis in mice when delivered by a food-grade Lactococcus lactis ⁶.

A 2017 study that isolated fresh strains from healthy volunteers added a useful link: among those isolates, butyrate production correlated with the capacity to induce IL-10 ⁸. Fuel and calm may not be two separate stories.

What it does against us — and where the proof runs out

Now the part that usually gets left off the infographic. Not harm — nothing cited here shows it attacking a healthy person — but the four places where the story is weaker than the label.

The human evidence is associative. Sokol's human data are observational; the causal work is cell lines and mice ¹. Depletion travelling with disease is not the same as depletion causing disease, and this species is depleted in a long list of conditions, which is exactly what you would expect from a fragile organism responding to illness in general.

And it is fragile — that cuts both ways. Because it cannot tolerate oxygen ³, anything that lets oxygen reach the colonic lumen removes it. A 2016 study of the faecal microbiota in HIV infection found F. prausnitzii critically less abundant than in controls, in a community that had shifted decisively towards oxygen-tolerant species: 80.8 % of the species enriched in the HIV group were aerotolerant (42 of 52), against 16.1 % of the species enriched in the control group (14 of 87) ⁷. That is a picture of a damaged, oxygenated gut, and the missing anaerobe is plausibly the consequence. Read a low value as a signal, not a verdict.

The name on the label promises more than it delivers. Isolates of this species from healthy people differ from one another in enzyme production, antibiotic resistance and immunomodulatory properties — the effects are strain-dependent, and the authors describe their strains as candidates for a next-generation probiotic, not as a proven one ⁸.

And it is a metabolic machine, not a benefit. Alongside butyrate and formate, all four strains in the original description produced D-lactate rather than L-lactate ³. A bacterium has outputs; we choose which one to call the point.

What none of the work cited here shows is that deliberately raising F. prausnitzii — by any means — makes a person healthier. That trial is the one to ask for.

The name keeps moving

It was described as Fusobacterium prausnitzii. In 2002, Sylvia Duncan and colleagues showed by 16S rRNA sequencing that these strains are only distantly related to Fusobacterium in the strict sense and belong instead with Clostridium cluster IV, and proposed a new genus: Faecalibacterium, with Faecalibacterium prausnitzii as the type species and strain ATCC 27768 as its type strain ³. Fusobacterium prausnitzii is the synonym you will still meet in older papers.

Then it moved again. In 2022, a genome-based re-examination split the group: three novel species were proposed, and the strain A2-165 — the workhorse behind much of the anti-inflammatory literature above — became the type strain of Faecalibacterium duncaniae ⁴. The genus has kept growing since; Faecalibacterium taiwanense was described from human faeces in 2024 ⁹.

This is not pedantry. It means a good deal of what is published under the words F. prausnitzii describes an organism that now has a different name, and any test or product quoting the species should be read with that in mind.

The honest summary

The most abundant bacterium in a healthy colon is an anaerobe that dies in air, feeds the gut lining with the fuel that lining prefers, secretes something that calms an inflammatory signal, disappears when the gut is inflamed, and has been renamed twice while we were studying it. Every one of those statements has a paper behind it. The statement that it will make you well does not — yet.