Pancreatic cancer cells use their surroundings to choose between growth and survival
Deep inside a pancreatic tumor, two populations of cancer cells are doing opposite things. Some cells are dividing rapidly - exploiting a rich local environment of structural proteins to grow. Others, further from those fibers, have switched into a self-preservation mode: breaking down their own internal components for nutrients, resisting chemotherapy, and waiting. Both populations exist simultaneously in the same tumor. This is why pancreatic cancer is so difficult to treat, and why a single drug so rarely works.
A study published in Cell by NYU Langone Health researchers reveals the molecular mechanism behind this switching behavior and identifies a specific target that, when disrupted, forces cells into a more uniform state that is significantly more vulnerable to existing treatment.
The fiber-sensing switch
The key player is a process called autophagy - a cellular self-eating mechanism in which cells degrade their own components into nutrients to sustain themselves. When autophagy is high, cancer cells focus on surviving rather than dividing. This protects them from chemotherapy drugs designed to attack fast-dividing cells. When autophagy is low, cells multiply faster but become more susceptible to those same drugs.
What determines which state a cell adopts? The research team found that it comes down to proximity to the extracellular matrix (ECM) - the fibrous scaffolding that surrounds cells in a tumor. Specifically, cells that detect ECM structural proteins such as laminin do so through a surface receptor called integrin subunit alpha3 (integrin-alpha3). Cells that can sense the ECM keep autophagy low and grow fast. Cells too far from the ECM to detect it ramp up autophagy and become chemotherapy-resistant.
"Our findings show that the sensing of the ECM by pancreatic cancer cells enables them to switch between states of active growth and autophagic survival," said study first author Mohamad Assi, PhD, a postdoctoral fellow at NYU Langone.
Forcing uniformity
The practical implication of this two-population problem is that no single drug can target all cancer cells simultaneously - one population is in growth mode, the other in survival mode, and they require different therapeutic approaches. Hydroxychloroquine, the only FDA-approved autophagy inhibitor available for patients, has shown limited success as a standalone treatment, partly because only a fraction of tumor cells are in high-autophagy mode at any given time.
The research team tested whether forcing all cells into the same state could change this picture. By genetically suppressing integrin-alpha3, they pushed cells toward uniformly high autophagy, making them all susceptible to hydroxychloroquine simultaneously. The result was a 50% reduction in cancer cell survival compared to hydroxychloroquine alone - a substantial gain from addressing tumor heterogeneity rather than just blocking a single pathway.
A second approach targeted NF2, a protein that passes along the integrin-alpha3 signal inside cells. NF2 inhibits the integrin signal, so knocking it out reduces autophagy by slowing lysosomes - cellular structures critical to the autophagic process. NF2-knockout-driven lysosomal inhibition drastically reduced pancreatic tumor growth and triggered cancer cell death across multiple experimental models.
Why current autophagy-blocking strategies fail over time
The researchers point out that current strategies designed to block autophagy work initially but fail as cancer cells adapt. Their results suggest that targeting both the ECM-mediated regulation of autophagy levels and lysosomal function together might produce longer-lasting anti-tumor responses than either approach alone.
The study was conducted using three-dimensional cancer cell spheres embedded in gel-like substances that mimic how tumors grow in the body - a more physiologically realistic model than standard flat-plate cultures. A fluorescent protein tracked which cells had high or low autophagy, allowing direct visualization of the two populations within a single tumor structure.
Important limitations
All experiments were conducted in vitro or in animal models; none involved pancreatic cancer patients. The genetic suppression of integrin-alpha3 used in the experiments is not a clinically available intervention - developing a drug that achieves a similar effect without affecting healthy tissue is a separate challenge. Hydroxychloroquine's poor tumor penetration is a known limitation that the current study does not resolve; forcing cells into high-autophagy mode may help, but delivery remains a practical barrier.
The study was funded by National Cancer Institute grants and received support from the Damon Runyon Cancer Research Foundation, the Lustgarten Foundation, and Stand Up to Cancer.