At a primate facility twenty male macaques eat a mash of lard, egg yolk, sugar and fructose for ten months. Their livers fill with fat and become inflamed. Then ten of them are given a daily oral dose of a tripeptide called DT-109. The other ten are given water. Five months later the livers of the two groups no longer look alike.
The disease is called metabolic dysfunction–associated steatohepatitis, or MASH. Steatosis means fat inside liver cells, hepatitis means inflammation of the liver. Fat on its own can sit there for years. Once inflammation arrives the cells swell and die, and the tissue that replaces them is fibrous, hard like a scar. Fibrosis is the step towards cirrhosis.
The group of Pengxiang Qu and Enqi Liu, at Xi'an Jiaotong University, with Jifeng Zhang and Y. Eugene Chen at the University of Michigan, looked for the chain that carries the damage from the gut to the liver. The paper was published in The Journal of Clinical Investigation on 12 May 2026, in issue 136(13), article e200522, open access under CC BY 4.0.
The first link is a bacterium, Clostridium perfringens, which lives in the gut and was more abundant in the sick macaques. It carries a gene catalogued as K00366 and called NirA, which serves to make ammonia. Ammonia is a small, irritant molecule, the same compound that gives household cleaners their sharp smell. In the sick macaques ammonia was higher in the last stretch of the gut, in the blood and in the liver.
The second link is the gut wall. In the sick macaques the mucus layer was thinner, the proteins that hold the cells together were reduced, and more endotoxin circulated in the blood, a piece of the outer coat of gut bacteria. The wall keeps two worlds apart and when it loosens, what should stay out gets in.
The third link is in the liver. Immune cells called CD8 T cells switch on and make perforin, a protein that punches holes in neighbouring cells. The authors measured 35 inflammation molecules in liver extracts from five macaques per group. Twenty-one of the 35 were raised in the sick animals and DT-109 brought sixteen of them down. The molecule with the largest difference was CCL5, a chemical call that draws in lymphocytes.
DT-109 is a tripeptide, a chain of just three amino acids, glycine, glycine and leucine. The macaques received 150 milligrams per kilogram of body weight a day, by mouth, for five months. DT-109 was not detected in the animals' blood, and the authors infer that it acts inside the gut and not elsewhere. Body weight did not change.
In mice the chain was taken apart piece by piece, in groups of four to six male animals. Mice colonised with ordinary Clostridium perfringens had more ammonia and a damaged gut wall. Mice colonised with a strain lacking the NirA gene did not. Mice given ammonium chloride to drink for eight weeks had more CD8 T cells and more perforin in the liver, with no bacteria added at all. In mice, finally, DT-109 and a virus that attacks the bacterium both lowered the fibrosis score and liver hydroxyproline, the molecule used to weigh fibrous tissue.
People enter the work only as samples. In a first group CCL5 in blood was measured in 44 healthy people and 67 with MASH, and the bacterium in stool in 40 people. In a second group the livers of six healthy and six sick people were examined. No person was given DT-109.
The paper's title says attenuated, not reversed. In the text, though, for the 35 inflammation molecules the authors write that DT-109 reverses many of these alterations. The Michigan Medicine release of 7 July 2026 takes that verb and applies it to the whole disease. ScienceDaily picked it up on 17 October with a headline about a new drug reversing the disease by repairing the gut. In the paper the fibrosis score falls in mice, and for the macaques that comparison sits in a supplemental figure with no figures in the text.
Y. Eugene Chen, the senior author, founded Diapin Therapeutics, which provided DT-109 for the study. Chen and two other authors hold a patent application on these tripeptides. The University of Michigan has an ownership interest in the company. The authors state this in the paper and the release states it too.
The paper says nothing about blinded assessment, except for the microscope reading of fibrosis, done by at least two pathologists who did not know which group the slides came from. It reports no toxicity data. All the animals are male, which the authors put down to variability linked to oestrogen. On using CCL5 to follow the response to DT-109 the authors write that human interventional studies are needed.
About seven per cent of people worldwide have this liver disease, according to the release. For now the ammonia chain has been seen in twenty male macaques and in groups of four to six mice.
References
Qu P., Ding S., Zhang Y., Zhao Y., Song E., Hu L., et al., Zhang J., Liu E., Chen Y.E., «Metabolic dysfunction–associated steatohepatitis exacerbated by Clostridium perfringens–derived ammonia is attenuated by tripeptide DT-109», The Journal of Clinical Investigation, 136, 13, e200522, published 12 May 2026, open access under CC BY 4.0. doi: 10.1172/JCI200522
Michigan Medicine, press release of 7 July 2026, «Drug candidate treats severe fatty liver disease by protecting the gut in animal models», by Noah Fromson; republished by ScienceDaily on 17 October 2026 as «New Drug Reverses Severe Fatty Liver Disease by Repairing the Gut» (declared secondary source).
*Board Member, SRSN (Roman Society of Natural Science)
Past Editor-in-Chief Italian Journal of Dermosurgery



