Highlights in Cancer Research: August 2023

Pancreatic cancer is typically diagnosed late, with few treatment options, motivating a search for early intervention strategies to improve outcomes. In this paper, the authors leveraged genetically engineered mouse models to obtain a high-resolution view of pancreatic tissue during the earliest stages of cancer progression. Single-cell transcriptomic sequencing identified cellular phenotypes that are absent from normal pancreatic epithelia, but emerge following oncogenic Kras mutation. Exposure of these cells to inflammation—a condition known to provoke the onset of rapid tumorigenesis—triggers a transition toward neoplastic states from several distinct cellular origins. Among these, a highly plastic, Nestin-expressing progenitor-like population serves as the precursor for neoplastic states that predominate in the pancreas weeks or months later. By integrating epigenetic data, the authors revealed that highly plastic populations undergo global chromatin reprogramming, heightened in the vicinity of cell-cell communication genes, which alter the way these cells send and receive signals from their environment. They developed a computational approach to catalogue the set of receptors and ligands that define the communication capabilities of each plastic epithelial state and linked these ‘communication modules’ with relevant modules in tissue immune cells to define a pre-malignant cell-cell interaction network. This analysis identified a feedback loop between epithelial cells and regulatory T cells in the microenvironment, which drives the exit from plastic states toward neoplastic endpoints.