Pancreatic Cancer Cell States | New Treatment Targets Identified

Mapping Pancreatic Cancer Heterogeneity

Pancreatic ductal adenocarcinoma ranks as the third-leading cause of cancer-related mortality globally. Its resistance to cytotoxic, targeted, and immune therapies remains a primary challenge for clinicians and researchers. While DNA alterations—specifically mutations in KRAS, CDKN2A, TP53, and SMAD4—are well-documented, these hallmarks offer limited targets for effective treatment. Patients often develop chemoresistance rapidly after standard care. Cellular heterogeneity, characterized by distinct transcriptional and epigenetic states, likely accounts for this resilience.

New research clarifies how malignant subpopulations operate within the tumor. Previous bulk-based studies identified two broad transcriptional subtypes: differentiated tumors, often linked to better outcomes, and poorly differentiated tumors, frequently labeled as quasi-mesenchymal, basal-like, or squamous. However, translating these bulk signatures to single-cell data remains difficult. High gene dropout rates in single-cell RNA sequencing obscure these signatures, and averaging signals across mixed cell types hides specific malignant state dynamics. The study team bypassed these hurdles by applying gene regulatory network (GRN) analysis to quantify the activity of over 1,800 transcription factors and co-factors.

Identifying Six Stable Cell States

Using metaVIPER, an algorithm designed to infer protein activity from single-cell profiles, the researchers analyzed 110 pancreatic cancer samples. They discovered six molecularly distinct cell states conserved across nearly all samples. These states represent three developmental lineages: the gastrointestinal lineage state (GLS), the primitive lineage state (PLS), and the morphogenic state (MOS). These lineages were further stratified by their MAPK pathway activity into M+ (high) and M- (low) substates.

GLS cells show a well-differentiated appearance associated with glandular structures. MOS cells are poorly differentiated, appearing as isolated cells that infiltrate the surrounding stroma. This MOS state correlates with worse prognosis and poor patient survival. PLS cells display intermediate markers. Notably, the MAPK substates are independent of developmental lineage. They fluctuate rapidly, as shown by experiments with RAF, MEK, and ERK inhibitors. This malleability reveals how pancreatic cancer cells adapt to therapy by transitioning between states.

Therapeutic Implications and Plasticity

Spatial transcriptomics confirmed that these cell states align with distinct histological features. GLS cells reside within organized glandular structures, while MOS cells invade adjacent tissue. Unlike these structural states, MAPK activity is not fixed to specific physical locations. Most tumor cells in human samples exhibit low MAPK activity, likely due to restricted nutrient and oxygen access in the tumor microenvironment. Cultured cells, which often exist in nutrient-rich media, display higher MAPK levels, potentially biasing drug screenings performed in the laboratory.

Lineage tracing experiments with genetic barcodes confirmed that these cells transition between states spontaneously. Even in isogenic contexts, cells switch lineages and MAPK states. This rapid plasticity suggests that tumors can readily survive treatments that target only a subset of cells. Targeting these cell states requires a new strategy. The researchers utilized CRISPR screens to identify state-specific genetic dependencies. For instance, GLS cells rely on HNF1A, while MOS cells depend on BPTF, a protein involved in KRAS-mutant signaling and MYC regulation.

Reprogramming offers a potential path forward. The team demonstrated that ectopic expression of GLS-associated transcription factors, particularly OVOL2, can force MOS cells to adopt a less aggressive GLS identity. Combining OVOL2 and HNF1A yielded near-complete transdifferentiation in vivo. These findings suggest that targeting master regulator proteins—rather than just mutated genes—could lock cancer cells into more vulnerable states. A combination therapy regimen, designed to hit all six states simultaneously, may offer the most promising avenue to overcome the inherent plasticity that defines this aggressive disease.