Korean study reverts cancer cells to normal
- South Korea’s KAIST researchers reported two colorectal cancer studies showing tumor cells can be pushed toward normal-like states by rewiring cell fate, not killing them. - In Advanced Science, the team identified three regulators — MYB, HDAC2 and FOXA2 — whose simultaneous inhibition reverted colorectal cancer cells into normal-like enterocytes. - The work extends differentiation therapy beyond blood cancers into colon models, but remains preclinical in cells, organoids and mice. (wiley.com)
Cancer treatment usually aims to kill tumor cells. A KAIST team instead reported colorectal cancer experiments that pushed malignant cells back toward normal-like states. (wiley.com) (kaist.ac.kr) Cells normally mature along a differentiation path, like trainees taking on fixed jobs. Cancer cells often move the other way, acting less specialized and more aggressive. (wiley.com) (nature.com) The KAIST group used single-cell RNA data from human large intestine tissue to build a computer model of that maturation path. The model, called BENEIN, was designed to identify the gene regulators that steer cells toward a normal intestinal identity. (wiley.com) (pmc.ncbi.nlm.nih.gov) In that 2025 Advanced Science paper, the team named MYB, HDAC2 and FOXA2 as the key regulators. Simultaneous knockdown of those three factors reverted colorectal cancer cells into normal-like enterocytes and also suppressed malignant traits. (wiley.com) (pmc.ncbi.nlm.nih.gov) The researchers said they validated the effect in molecular and cellular experiments and in animal studies. KAIST described the result as converting colon cancer cells into a state resembling normal colon cells without directly killing them. (kaist.ac.kr) (pure.kaist.ac.kr) A second KAIST paper, published in 2025, looked even earlier in the process — the unstable transition point where normal cells tip into cancer. The team called that framework REVERT and used it to search for a molecular switch that could reverse the transition. (wiley.com) (kaist.ac.kr) That study focused on colorectal tumorigenesis and patient-derived matched organoids of colon cancer and normal colon. The paper said the method reconstructed a shared network landscape containing both normal and tumor cell states, then identified targets that could drive cells back. (wiley.com) (pmc.ncbi.nlm.nih.gov) This approach fits a broader idea known as differentiation therapy, which tries to make cancer cells grow up into less dangerous forms. That strategy already has a clinical track record in acute promyelocytic leukemia, where all-trans retinoic acid and arsenic trioxide transformed outcomes. (wiley.com) (nature.com) What KAIST has not shown is a ready-to-use treatment for patients with colon cancer. The published evidence is still preclinical, and the papers themselves frame the work as a platform for finding control targets rather than a proven therapy. (wiley.com 1) (wiley.com 2) So the clearest takeaway is narrower than the social-media version: these studies mapped gene switches that made colorectal cancer models look and act more like normal cells. Whether that can become a safe drug strategy is the next unanswered step. (wiley.com) (kaist.ac.kr)
