A 24-hour fast before a potentially lethal dose of abdominal radiation was enough to save every mouse in a new study from the University of Texas MD Anderson Cancer Center, while every fed mouse died within nine days. The difference came down to what was living in their guts.
The study, published June 23 in PNAS and led by Kunal Rai and Helen Piwnica-Worms, identifies a complete molecular cascade linking fasting to intestinal regeneration: the fast enriches a specific gut microbe, whose metabolic byproduct remodels the epigenome of intestinal stem cells, producing a population of “primed persister cells” ready to repair damage before it even arrives.
The cascade
The mechanism runs through four stages. First, fasting expands goblet cells in the small intestine, which produce mucus. Second, that mucus feeds Akkermansia muciniphila, a mucin-degrading bacterium that became significantly enriched in fasted mice, confirmed by 16S rRNA sequencing of ileal contents (P = 4 × 10⁻⁵). Third, A. muciniphila produces propionic acid, a short-chain fatty acid, while the host liver produces β-hydroxybutyrate; both metabolites accumulate in the ileum. Fourth, propionate drives histone acetylation (H3K27ac and H3K9ac) and β-hydroxybutyrate drives histone β-hydroxybutyrylation (H3K9bhb) in crypt epithelial cells, remodeling the chromatin landscape.
CUT&Tag profiling identified 37,601 unique H3K27ac peaks and 88,040 unique H3K9ac peaks in the crypt cells of fasted mice, and these were dependent on A. muciniphila, disappearing when the bacterium was depleted with tetracycline.
The epigenetic remodeling expands a population of “primed persister cells”, Clu⁺Olfm4⁺ cells with open chromatin at stem cell genes including Lgr5, Ascl2, Sox9, Axin2, and Rnf43. These cells are distinct from classical revival stem cells because they lack Ly6a accessibility and are present before any injury occurs.
Survival and recovery
The survival data are striking. After 11.5 Gy of total abdominal X-ray irradiation, a dose uniformly lethal in fed mice, which all died within 7 to 9 days, 100% of fasted mice survived. When A. muciniphila was depleted with broad-spectrum antibiotics before fasting, survival dropped to approximately 50%. Reintroducing A. muciniphila by gavage restored 100% survival.
However, giving A. muciniphila to fed mice provided no benefit. The bacterium is necessary but not sufficient; fasting provides essential context, likely through host-derived β-hydroxybutyrate and goblet cell expansion.
At the tissue level, fasted mice showed significantly greater crypt depth, higher numbers of Olfm4⁺ stem cells, and by 96 hours post-irradiation, 60% of crypt cells were Clu⁺Lgr5⁺, compared to 43.7% in fed mice.
Limitations
The study was conducted entirely in male C57BL/6J mice. No human data are presented. The translational relevance is unproven, though the mechanistic chain from fasting to microbiome to epigenome to radioresistance is fully mapped at the molecular level. The authors note that other bacterial taxa, including Erysipelatoclostridium, Clostridium sensu stricto 1, and Coprococcus, also increased with fasting, and their potential contributions have not been ruled out.
A separate 2024 Nature paper (Imada et al.) showed that post-fast refeeding can increase intestinal tumorigenesis through mTORC1-driven polyamine metabolism. The current study did not examine long-term cancer risk from repeated fasting cycles.
The paper was a Contributed submission to PNAS by National Academy of Sciences member Helen Piwnica-Worms. Funding came from the NIH and the Cancer Prevention and Research Institute of Texas.
Source: Barrodia P, Saw AK, Jeter-Jones SL, et al. Fasting primes small intestinal regeneration after damage via a microbiome–metabolite–chromatin axis. PNAS. 2026;123(26):e2529215123. doi:10.1073/pnas.2529215123