A loss of the gene-regulating protein GATA6 pushes tumor cells into a primitive, highly mobile state — offering a new way to think about why colorectal cancer turns deadly
Researchers at Weill Cornell Medicine and MIT say they’ve pinned down a key mechanism behind one of colorectal cancer’s most dangerous behaviors: its tendency to spread to the liver. The team found that when a gene-regulating factor called GATA6 declines, cancer cells lose their normal identity and shift into an adaptable, fetal-like state capable of traveling through the bloodstream and seeding new tumors in the liver.
Notably, the researchers say this shift appears to be driven mainly by changes in gene activity rather than by new DNA mutations — a finding that redirects a search that has, until now, mostly focused on hunting for mutations.
GATA6: an “identity keeper” that goes quiet
In healthy intestinal tissue, GATA6 functions as a kind of molecular identity keeper, helping cells that line the gut maintain their specialized roles. But the study, published in the journal Cell Stem Cell, found something striking when comparing tumor samples: GATA6 levels were substantially lower in liver metastases from both mice and human colorectal cancer patients, and lower GATA6 expression correlated with worse outcomes for patients. Since metastasis remains the leading cause of death from the disease once it spreads beyond the original tumor site, understanding what triggers that spread has long been a priority.
Dr. Norihiro Goto, one of the study’s co-leaders, described the finding directly, saying the loss of GATA6 works as a decisive switch that turns primary tumor cells from harmless to capable of spreading, and that the results point toward epigenetic changes as a major driver of liver metastasis rather than something written into the cancer’s DNA.
Built from lab-grown “mini tumors”
Because studying tissue already in the liver only shows what happens after metastasis has occurred, the team needed another approach. Dr. Norihiro Goto noted that examining patient samples from established liver metastases misses the important signals that occur during the earliest stages of the spreading process.
To get around that, the scientists built organoids — small 3D clusters of cancer cells that mimic real tumor behavior — using cells derived from liver metastases, then implanted them into the colons of mice. Over repeated rounds, the tumors grew progressively more aggressive and eventually spread to the liver, letting researchers watch the cells gradually acquire the ability to metastasize.
The experiments showed that losing GATA6 promotes what’s called lineage plasticity — the capacity of a cell to change its identity and behavior. Without GATA6, colorectal cancer cells switched on alternative genetic programs and took on a flexible, fetal-like form that made it easier for them to travel through the blood and take root in other organs. This same kind of cellular reshaping normally helps the body repair wounds and adapt to stress — but in cancer, it appears to be co-opted to drive metastasis.
A stem-cell marker flips off
One telltale sign of the transformation involved a marker called LGR5, typically found on intestinal stem cells. Earlier work had already shown that cells lacking LGR5 can be the ones that kick off liver metastases, and the new study found that switching off GATA6 pushes cancer cells from an LGR5-positive state into an LGR5-negative one — the same fetal-like, migration-ready state described above. Restoring GATA6, or activating pathways connected to it, reduced how likely the cells were to metastasize.
In mouse models, deleting GATA6 altogether significantly increased both how often liver metastases formed and how extensive they were — while barely affecting how fast the original tumor grew, according to Dr. Norihiro Goto. That distinction matters: it suggests metastasis may hinge more on specific shifts between cell states than on the size or growth speed of the primary tumor.
Toward a biomarker — and a treatment target
The findings raise the possibility that GATA6 levels could serve as a biomarker for metastatic risk, helping doctors flag which patients might need closer monitoring or more aggressive treatment. The study also points to a possible treatment strategy centered on preserving cell identity or blocking cells from entering that flexible, pro-metastatic state — though the researchers caution that any such therapy would need to avoid disrupting the body’s normal wound-repair processes, which lean on similar biological machinery.
Looking ahead, the team plans to search for vulnerabilities specific to GATA6-deficient cancer cells that new drugs could exploit, and to study how the surrounding tumor environment — including immune cells and liver-specific signals — shapes these transitions.
-Rashmi Kumari
