When plant nutrients run low, some gut bacteria may turn on their host and begin consuming the intestines’ protective lining.
The health effects of a plant-based diet may depend not only on what gut bacteria eat but also on whether they begin feeding on the gut itself.
New research suggests that fiber and previously overlooked plant proteins can steer intestinal microbes away from consuming the protective mucus lining of the intestines. At the same time, these nutrients encourage the production of metabolites associated with better gut health and healthy body weight.
The findings reveal a more complicated relationship between food, microbes, and human metabolism than scientists once recognized. They also challenge the widespread assumption that many biologically important compounds found in the body come exclusively from gut bacteria.
Diet Reshapes Microbial Chemistry
Two studies led by Ludwig Princeton researchers Jenna AbuSalim and Director Joshua Rabinowitz examined how these compounds are produced and how diet changes the balance between potentially beneficial and harmful forms. One study appears in the Proceedings of the National Academy of Sciences, while the other was published in Nature Metabolism.
“There’s growing interest across medical disciplines in manipulating the human microbiome or using its metabolic products themselves for therapy,” said Rabinowitz. “Diet holds great promise for controlling the microbiome and its outputs. But to devise effective therapeutic interventions, we need to understand what aspects of the diet control which microbial outputs.”
The PNAS study focused on phenols, a broad group of metabolites created as amino acids are broken down. Although these compounds share a chemical family, they can have sharply different relationships with health.
When gut bacteria process the amino acid phenylalanine, they can produce phenylpropionate and hippuric acid. These compounds have been associated with intestinal health and healthy body weight.
Bacterial metabolism of another amino acid, tyrosine, can instead generate p-cresol sulfate and phenol sulfate. These metabolites have been linked to systemic toxicity in people with kidney disease and poorer outcomes among cancer patients.
The Overlooked Power of Plant Protein
The researchers discovered that plant fiber and indigestible plant proteins shifted microbial activity toward the phenylalanine pathway, increasing the favorable metabolites while reducing those associated with harm.
“Our studies showed that both the fiber and indigestible proteins from plants—which we call ‘proteins imitating fiber,’ or Prif—shift the balance of phenol metabolites from the harmful kind made from tyrosine to the healthful variety derived from phenylalanine,” said AbuSalim.
Fiber has long been recognized as an important food source for intestinal microbes. Indigestible plant proteins, however, have received far less attention because proteins are generally expected to be broken down and absorbed earlier in digestion.
The new work suggests that some plant proteins escape digestion and reach the colon, where bacteria use them much like they use fiber. AbuSalim and her colleagues call these proteins Prif, short for “proteins imitating fiber.”
Tracing What Gut Bacteria Consume
To follow these nutrients through the digestive system, the researchers labeled proteins with stable (non-radioactive) isotopes. These chemical markers allowed them to determine which metabolites came from food and which originated in the mouse’s own tissues.
The results exposed a clear divide.
Beneficial phenols were made almost entirely from indigestible proteins in the diet. Harmful phenols, by contrast, were produced largely when bacteria broke down proteins belonging to the host, including those in the mucus that coats and protects the intestinal wall.
That mucus layer does more than lubricate the gut. It forms part of the barrier separating trillions of microbes from the cells beneath them. When bacteria lack preferred dietary nutrients, some can turn to this host-supplied material as an alternative food source.
How Fiber and Prif Protect the Gut
Fiber reduced that behavior. It suppressed bacterial breakdown of the intestinal mucus layer, which lowered production of the harmful phenols. Prif supplied additional dietary protein to microbes farther along the digestive tract, increasing production of the more favorable compounds.
“We think Prifs represent an emerging class of dietary nutrients that shape the composition of the gut microbiome and could have a far-reaching influence on metabolic health,” said AbuSalim.
“Food packaging may eventually list Prif right below fiber,” said Rabinowitz.
Challenging a Microbiome Assumption
The Nature Metabolism study addressed another basic question: Who is actually making these metabolites?
Scientists have often treated certain phenols and indoles as chemical fingerprints of microbial activity. Indoles are produced from the amino acid tryptophan and have attracted interest because of their possible roles in inflammatory bowel disease, neurodegenerative disorders, cancer spread, and antitumor immune responses.
This assumption has helped drive attempts to raise desirable indole levels through dietary changes, probiotics, or other methods aimed at altering gut bacteria.
The Body Makes These Metabolites Too
But those strategies depend on microbes being the main source. If human cells produce much of the same material, changing the microbiome may have a smaller effect than expected.
AbuSalim, Rabinowitz, and their colleagues tested the assumption using isotope tracing in mice, rats, and human cells. They found that mammalian metabolism could independently produce many indole and phenol compounds, including the physiologically important metabolites indole-3-lactate and indole-3-acetate.
The distinction became clearer after antibiotic treatment disrupted the microbiome. Many of the metabolites remained abundant in the blood of mice, indicating that the animals’ own cells continued producing them.
Patient samples showed a similar pattern, including samples from cancer patients taking antibiotics. Compounds made exclusively by microbes behaved differently. Indole-3-propionate and p-cresol sulfate, for example, declined after antibiotic treatment.
Toward More Precise Nutrition and Therapies
The findings show that the same metabolite can sometimes come from more than one source. Its concentration in the bloodstream may reflect a combination of diet, microbial activity, and metabolism in the body’s own tissues.
That matters for therapies designed to change metabolite levels. A probiotic may be useful when a compound is made primarily by microbes, while a host-produced metabolite may require an entirely different approach.
Together, the studies provide a clearer map of how food, gut bacteria, and mammalian cells contribute to the body’s chemical environment. They also suggest that broad advice to “eat more fiber” may eventually be refined to account for the specific fibers and indigestible proteins that promote desirable microbial activity.
“Beyond that,” said Rabinowitz, “a clearer picture of how different foods interact with the microbiome to modulate the production of bacterial metabolites will help sharpen the guidance nutritionists and doctors can give to people for disease prevention and therapy.”
References:
“Digestion-resistant proteins support the healthy metabolite profiles associated with plant-based diets” by Jenna E. AbuSalim, Michael M. MacArthur, Meera Gupta, Asael Roichman, Craig J. Hunter, Felix C. Keber, Seema Chatterjee, Mahta Moussavi, Javad Barouei, Martin Wühr, Dinanath Sulakhe, Christopher J. Lehmann, Mohamed S. Donia and Joshua D. Rabinowitz, 27 July 2026, Proceedings of the National Academy of Sciences.
DOI: 10.1073/pnas.2605226123
“Host metabolism can produce many indoles and phenols independently of the microbiome” by Jenna E. AbuSalim, Kellen Olszewski, Salma Youssef, Sarah J. Mitchell, Craig J. Hunter, Jessica Little, Ashley Sidebottom, Jacob A. Boyer, Steve D. Knutson, Laith Z. Samarah, Michael R. MacArthur, Alessa L. Henneberg, Rolf-Peter Ryseck, Christiane A. Opitz, David W. C. MacMillan, Mohamed S. Donia, Eric G. Pamer, Sabrina Imam, Christopher J. Lehmann, Olatoyosi Odenike and Joshua D. Rabinowitz, 23 June 2026, Nature Metabolism.
DOI: 10.1038/s42255-026-01550-8
These studies were funded by the Ludwig Institute for Cancer Research, the National Institutes of Health, the National Institute of Diabetes and Digestive and Kidney Diseases, the Princeton Alliance for Collaborative Research and Innovation, Princeton University.
Never miss a breakthrough: Join the SciTechDaily newsletter.
Follow us on Google and Google News.