New Gut Bacterium C. immunis Cut Visceral Fat in Obese Mice

Researchers at Duke University School of Medicine have identified a previously unknown member of the healthy human gut microbiome and shown that a sugar it secretes stripped visceral fat from obese mice. The bacterium has been named Clostridium immunis, and the work was published on September 23, 2026 in the journal Cell Host & Microbe. Duke described the findings publicly the following day.

The molecule at the center of the study is an exopolysaccharide, a complex sugar released by the bacterium, carrying a chemical tag called phosphocholine. The team built genetically modified versions of C. immunis that could no longer make that tag. Without phosphocholine, the protection against metabolic disease disappeared, which is how the researchers concluded the tag is the essential part of the molecule.

In obese mice, the bacterial sugar reduced visceral fat -- the deposits that gather around internal organs and that Duke notes are associated with diabetes and heart disease -- and increased metabolic activity, most clearly in visceral adipose tissue itself. Two details make the result unusual: the animals did not eat less, and they did not lose muscle mass.

That separates the mechanism from the GLP-1 drugs that now dominate obesity treatment, which work mainly by cutting how much a patient eats. "GLP-1 drugs have transformed treatment of obesity and its associated comorbidities, but they do not work for everyone and often cause significant side effects," said Neeraj Surana, MD, PhD, the Duke assistant professor of pediatrics and of molecular genetics and microbiology who led the work. He said the bacterium and its sugar "could be used as either an adjunct or alternative to GLP-1 medications."

The team also traced an immune route for the effect. The bacterial molecule lowered levels of the signaling protein IL-22 in the small intestine and in visceral fat, and that drop was followed by the rise in metabolic activity in visceral adipose tissue.

Whether any of this carries over to people is unsettled. When the researchers examined microbiome data from thousands of humans, genes for making phosphocholine turned out to be less abundant in people with obesity or with high triglyceride levels. That is a correlation, not evidence that the molecule treats anything in humans. Duke states plainly that the findings do not yet show the bacterium or its sugar can treat obesity in people, and that the group is seeking funding for clinical studies.

The idea has been in public view far longer than the announcement suggests: the same team posted its first description of the phosphocholine mechanism as a bioRxiv preprint on June 12, 2024, roughly 27 months before the peer-reviewed paper appeared. The study was supported by the National Institutes of Health, The Hartwell Foundation, the Duke Microbiome Center, Duke's Gilhuly Accelerator Fund, the American Heart Association and the National University of Singapore.