Colorectal cancer (CRC) is the third most common cancer worldwide and the second leading cause of cancer death. While immunotherapy - specifically drugs that block the PD-1 checkpoint - has been a breakthrough for some patients, it still fails in many cases.
Patients whose tumors have a molecular feature called high microsatellite instability (MSI-H) are considered the best candidates for PD-1 immunotherapy. Even so, only about 44% of MSI-H patients respond to treatment, meaning more than half do not benefit.
The key question this study addresses is: what molecular programs inside tumor cells drive this resistance? The answer could unlock new ways to help the many patients who currently do not respond.
The researchers used single-cell RNA sequencing (scRNA-seq) - a technology that reads the activity of thousands of genes in individual cells - to compare tumors from patients whose cancer resisted PD-1 therapy versus those who responded.
Tumor samples from three PD-1-resistant and three PD-1-sensitive MSI-H patients were analyzed. This cell-by-cell approach revealed differences in gene regulatory networks (GRNs) - the control systems that determine which genes are switched on or off inside each cell.
The analysis used a specialized tool called SCENIC to reconstruct these regulatory networks and identify which transcription factors (proteins that switch genes on or off) were most active in resistant tumors. Machine learning algorithms were then applied to build a prognostic model from this data.
Among all the transcription factors analyzed, ETV4 stood out as the most significantly elevated in PD-1-resistant tumors. ETV4 was particularly active in malignant epithelial cells - the cancer cells themselves - rather than in immune or support cells.
Using large public cancer databases (TCGA), the team found that patients with high ETV4 activity had worse survival outcomes. They also built a machine learning prognostic model using ETV4 target genes that reliably sorted CRC patients into high-risk and low-risk groups.
A striking finding was a strong correlation between ETV4 expression and macrophage markers. This hinted that ETV4 in cancer cells was somehow communicating with - and reprogramming - nearby immune cells called macrophages.
The team identified MC1R (melanocortin 1 receptor) as the key gene through which ETV4 drives immune suppression. Normally associated with skin pigmentation, MC1R was found to be directly activated by ETV4 in colorectal cancer cells.
Laboratory experiments confirmed that ETV4 protein physically binds to the MC1R gene promoter (the gene's on-switch) and activates it. This was verified using two complementary techniques: dual-luciferase reporter assays and chromatin immunoprecipitation, which together prove direct molecular binding.
This discovery revealed a previously unknown connection between a transcription factor driving cancer progression and an immune-regulating receptor - the ETV4-MC1R signaling axis.
Macrophages are immune cells that can act either as tumor fighters (called M1 macrophages) or as tumor helpers (called M2 macrophages). Tumors that are good at evading the immune system tend to have more M2 macrophages.
Using co-culture experiments - growing cancer cells and macrophages together in the lab - the team showed that cancer cells with high ETV4 activity pushed macrophages toward the immunosuppressive M2 phenotype. This effect was dependent on the MC1R pathway.
When MC1R was blocked or silenced, ETV4-high cancer cells lost their ability to reprogram macrophages. This confirmed that the ETV4-MC1R axis is the mechanism by which cancer cells recruit macrophages to shield them from immune attack.
The researchers tested their findings in living mice using the CT26 colon cancer model. Seven experimental groups were created, each receiving different combinations of treatments: anti-PD-1 antibodies, ETV4 overexpression, MC1R knockdown, and a drug called PLX3397 that depletes tumor-associated macrophages.
The most striking result: combining ETV4-MC1R pathway targeting with anti-PD-1 therapy reduced tumor volume by 55.6% more than anti-PD-1 treatment alone. This difference was statistically significant (P less than 0.05).
Animals where MC1R was silenced and PD-1 was blocked simultaneously showed the best tumor control, confirming that disrupting the ETV4-MC1R axis restores the immune system's ability to fight the tumor when combined with checkpoint therapy.
This study identifies a concrete, targetable mechanism that explains why some MSI-H colorectal cancer patients fail immunotherapy. For the roughly 56% of patients who do not respond to PD-1 blockers, this research points to a specific pathway - the ETV4-MC1R axis - as a potential drug target.
The prognostic model built from ETV4 target genes could potentially help clinicians identify which patients are at highest risk of resistance before treatment begins, enabling more personalized therapy planning.
Drugs that block MC1R or target its upstream regulator ETV4 could be developed as combination partners for existing immunotherapy drugs. Such combinations would aim to strip away the tumor's immune shield before or during treatment, potentially converting non-responders into responders.