Cancer Biology · global
How Do Cancer Cells Borrow Repair Programs to Travel? ZFP36L2 Links Intestinal Regeneration and Colorectal Cancer Metastasis
Researchers have identified a molecular switch that coordinates stress responses, RNA clearance, and cellular dedifferentiation; however, shutting it down may not only suppress cancer, but could also force tumors toward more heterogeneous cell fates.
After intestinal injury, mature cells can temporarily revert and regain stem cell capacity to repair tissue. A study published in *Nature* shows that colorectal cancer may appropriate the same survival program to rebuild tumors after leaving the primary tumor and reaching distant organs; coordinating this transition is the RNA-binding protein ZFP36L2.
The research team analyzed single-cell data from normal human colon, primary tumors, and metastatic lesions, and found that ZFP36L2 was prominently expressed in LGR5-positive intestinal stem cell-like cells. Previous sequencing studies have shown that approximately 5% to 10% of colorectal cancers harbor ZFP36L2 mutations, many of which may cause loss of protein function, suggesting that its relationship with tumor cell fate is not merely incidental.
In two mouse models of intestinal injury, removing Zfp36l2 hindered mature cells from returning to an LGR5-positive stem cell state, resulting in incomplete intestinal repair; organoids established from mouse tissue were also less able to regenerate. These results position ZFP36L2 as a switch required for post-stress “dedifferentiation,” rather than merely a stem cell marker.
Cancer experiments presented a more complex picture. The team reduced ZFP36L2 in four patient-derived colorectal cancer organoid lines and transplanted them into mice: metastatic burden in the liver and lungs declined in some models, indicating that cancer cells unable to return to the typical LGR5-positive stem cell state had greater difficulty completing metastatic seeding and growth. However, metastatic cells with greater baseline plasticity could still take alternative, atypical differentiation routes, such as squamous or neuroendocrine-like states; in clinical data, these heterogeneous states are associated with worse outcomes.
Mechanistically, ZFP36L2 recognizes AU-rich sequences within the 3′ untranslated regions of stress-related messenger RNAs, forming RNA-dependent biomolecular condensates that can dynamically assemble and disassemble, thereby helping degrade RNA and terminate the stress response. Tissue dissociation, growth factor deprivation, and low-dose irinotecan could all promote condensate formation; truncating mutations commonly found in patient tumors lost this ability, linking genetic mutations, RNA metabolism, and cell-state transitions into a single mechanistic chain.
This work builds on preliminary findings presented by the research team at a cancer research conference in 2025, and the related human RNA sequencing data have also been deposited in the GEO database. It offers a new therapeutic entry point, but has not yet demonstrated that directly inhibiting ZFP36L2 can safely treat cancer: the current evidence comes mainly from mice, organoids, and correlative analyses of human specimens, and ZFP36L2 also participates in normal intestinal repair; blocking it could both reduce typical metastasis and select for more heterogeneous alternative cell states.