Dr. Axe Wire · Study decoded · gut
Gut Bacteria Byproduct Shrinks Colon Tumors by Targeting Cholesterol
A microbial metabolite called 4-C-3 cut tumor cholesterol and shrank colorectal tumors in cell and animal models, researchers report.

- Who was studied: Colon cancer cells, lab-grown patient tumor models and mice, plus stool and tumor samples from colorectal cancer patients.
- What it doesn't show: Whether restoring 4-C-3 in the gut shrinks tumors in people.
A metabolite made by gut bacteria shrank colorectal tumors by rewiring how the tumor handles cholesterol, according to a study published November 10, 2026 in Nature Metabolism.
Colorectal tumors build up excess cholesterol, which fuels growth, but body-wide cholesterol-lowering drugs have shown limited benefit and carry safety concerns. Researchers screened cholesterol-related metabolites made by gut bacteria and found one, 4-cholesten-3-one (4-C-3), depleted in stool and tumor samples from colorectal cancer patients. Restoring 4-C-3 suppressed tumor growth in colorectal cancer cell lines, in patient-derived tumor organoids (miniature lab-grown versions of a patient's own tumor), and in animal models.
The team restored 4-C-3 two ways: with Oscillibacter ruminantium, a gut bacterium that naturally makes it and is reduced in colorectal cancer, and with a modified, non-disease-causing E. coli strain engineered to convert intestinal cholesterol into 4-C-3. Both approaches lowered cholesterol inside the tumors and caused tumor regression in the animal models, without observed overt toxicity.
The authors report 4-C-3 blocks two of the tumor's main growth drivers: it stops activation of the EGFR growth receptor, and it locks mutated KRAS, a common cancer-driving gene, into an inactive state. That second action, the authors found, overcame tumor resistance to existing anti-EGFR drugs, a known problem in KRAS-mutated colorectal cancer. In plain terms, a gut bacteria byproduct can starve tumor cells of cholesterol while jamming two growth switches at once.
4-C-3 suppresses tumour growth in CRC cell lines, patient-derived organoids and xenografts.
This work was done in cell lines, lab-grown organoids and animal models, not human patients, so whether restoring 4-C-3 would shrink tumors or improve outcomes in people is untested. Human data came only from measuring existing 4-C-3 levels, not from giving it as treatment, so larger studies testing 4-C-3 restoration directly in people are the needed next step.
Dr. Axe’s Take
Targeting cholesterol specifically inside the tumor, rather than body-wide, is a smarter approach than statin-style strategies that have fallen short for colorectal cancer. I'd expect this from gut research generally: bacteria in the colon produce compounds that talk directly to tumor cells. This is animal and lab-dish work, not a human treatment, so I would not tell anyone to change cancer care based on it. What I'd do this week is support the gut bacteria already present with fiber-rich vegetables, fermented foods like sauerkraut and kefir, and bone broth. Keep following your oncologist's treatment plan exactly as prescribed.
— Dr. AxeStudy summary
- Who
- Cell lines, patient-derived tumor organoids, animal models, and stool/tumor samples from colorectal cancer patients; size not reported.
- Design
- Laboratory screen of microbial cholesterol metabolites, followed by testing 4-C-3 restoration in cell lines, organoids and animal models.
- Dose
- 4-C-3 restored via Oscillibacter ruminantium bacteria or an engineered E. coli strain converting intestinal cholesterol into 4-C-3.
- Primary result
- 4-C-3 was depleted in stool and tumor samples from colorectal cancer patients; restoring it suppressed tumor growth in cell lines, organoids and animal models.
- Secondary
- 4-C-3 blocked EGFR activation and locked mutant KRAS in an inactive state, overcoming resistance to anti-EGFR therapy; both restoration methods lowered intratumoral cholesterol and caused tumor regression without overt toxicity.
- Funding / conflicts
- Competing interests: H.L.X.W., P.A., J.W. and L.Z. are named inventors on a patent application related to this work (U.S. Patent Application Publication No. US 2026/0077000 A1). The other authors declare no competing interests. Funding not reported in the abstract.
- Date
- Nov 10, 2026



