HGF-driven plasticity keeps colorectal cancer cells in a resistant state
Colorectal cancer does not stay biologically still. Under pressure from the tumor microenvironment, malignant cells can shift into a more plastic, drug-tolerant state that helps them survive treatment, seed recurrence, and sustain metastatic behavior. This study centers on a particularly troublesome axis: hepatocyte growth factor, or HGF, and the cell-surface stemness marker CD44v6.
Investigators show that AMC303, a highly selective CD44v6 inhibitor, is not enough on its own when microenvironmental signals remain active. HGF stimulates c-Met phosphorylation, increases CD44v6 expression, and boosts clonogenicity, a pattern consistent with expansion of the cancer stem cell compartment. In practical terms, the cells do not simply die when one pathway is blocked; they rewire themselves toward survival.
CD44v6 blockade alone leaves a molecular escape route open
The most important biology here is compensatory signaling. Even when canonical CD44v6/c-Met signaling is blocked with AMC303, HGF still drives phenotypic reprogramming in colorectal cancer stem cell mouse avatars. The escape route runs through c-Met co-receptors and downstream PI3K/AKT and MAPK pathways, which preserve growth and support a hybrid epithelial-mesenchymal state.
That hybrid state matters because it is not a simple on-off switch between epithelial and mesenchymal identity. Cells in this intermediate condition tend to retain enough adhesion and survival machinery to persist, while also gaining mobility, invasiveness, and resistance to stress. In this setting, the tumor does not merely resist therapy; it adapts to it.
A single pathway block therefore risks becoming a temporary delay rather than a durable therapeutic interruption.
Dual inhibition brought the proliferative program under control
The addition of Apitolisib, a dual PI3K/mTOR inhibitor, changed the response profile by suppressing those compensatory survival signals. When PI3K/AKT and MAPK activation were reduced, the proliferation rate fell and cancer cells became more vulnerable to exposure to standard chemotherapy in vitro. The combination strategy attacked both the surface trigger and the intracellular circuitry that allows the cell to recover after the first insult.
This is the logic that cancer therapy often forces clinicians and researchers to confront: blocking the dominant receptor is not enough if the downstream network can reroute around the obstruction. Here, the tumor microenvironment supplies the fuel, c-Met provides the gateway, and PI3K/mTOR helps keep the cellular engine running. Shutting down two of those nodes at once creates a more difficult environment for adaptive escape.
Cancer stem cells and the hybrid EMT state explain the resistance phenotype
The study places cancer stem cells at the center of this resistance program. HGF expanded stem-like populations and pushed transcriptional reprogramming toward a hybrid epithelial-mesenchymal transition, or EMT, state with greater resistance and metastatic potential. This is clinically relevant because stem-like cells are often the reservoir from which relapse emerges after treatment appears to be working.
Hybrid EMT is especially troubling because it supports both plasticity and persistence. Cells can alternate between phenotypes as conditions change, which makes them harder to eradicate with regimens designed for static tumor populations. In this framework, CD44v6 is not just a marker; it marks a biologically active population capable of responding to environmental cues and shifting into a more aggressive mode.
The paper’s mouse avatar model strengthens that interpretation by recreating the influence of HGF in a living system rather than in isolated cell culture alone. Even with one arm of the pathway restrained, the microenvironment still preserved enough signaling strength to maintain reprogramming unless PI3K/mTOR was also inhibited.
Why this combination strategy is mechanistically credible
The appeal of pairing AMC303 with Apitolisib comes from pathway logic rather than from a single dramatic readout. HGF activates c-Met, c-Met supports CD44v6 expression, and downstream PI3K/AKT and MAPK signaling preserve survival, proliferation, and resistance. A therapy that only targets the receptor leaves the downstream circuitry ready to absorb the blow and continue transmitting growth signals.
By contrast, combining a CD44v6 inhibitor with a PI3K/mTOR inhibitor addresses both the initiating signal and the adaptive fallback program. That design is especially important in colorectal cancer, where the tumor microenvironment can actively instruct malignant cells to become more invasive and less treatment-sensitive. The result is a strategy aimed not only at killing tumor cells, but at preventing the reprogramming state that makes them harder to kill in the first place.
The study also suggests why standard chemotherapy may perform better when this adaptive circuitry is interrupted. If the cells lose their ability to maintain a resilient stem-like phenotype, cytotoxic exposure is more likely to tip the balance toward cell death rather than survival.
Resistance biology now points to a narrower therapeutic target set
The most actionable lesson from this work is the breadth of the resistance mechanism. HGF-driven reprogramming did not depend on one isolated signal; it used a network that connected the microenvironment, c-Met activation, CD44v6 expression, stem-cell enrichment, and downstream kinase cascades. That complexity explains why monotherapy against a single node can be outmaneuvered.
For colorectal cancer research, the findings sharpen the case for combination regimens that are built around tumor plasticity rather than around bulk tumor shrinkage alone. The data remain preclinical, and the work does not establish patient-level benefit, dosing strategy, or clinical safety. What it does establish is a biologically coherent route through which colorectal cancer stem cells can survive environmental stress, and a plausible way to interrupt that route before the cells complete their reprogramming program.
The next clinical hurdle is straightforward to state and difficult to solve: whether a dual blockade of CD44v6 and PI3K/mTOR can suppress adaptive escape in patients without adding prohibitive toxicity or simply shifting resistance to another pathway.