Abstract A024: A plastic, MYC -dependent KRT17 +/ GATA6 + hybrid state drives chemoresistance and therapeutic

Abstract

Pancreatic ductal adenocarcinoma (PDAC) progression is driven by cellular heterogeneity not fully captured by the classical (GATA6+) versus basal-like (KRT17+) model. Intermediary states are linked to aggression, but their characteristics remain unclear. Here, we deconstruct this heterogeneity using a KRT17/GATA6 framework to define four distinct states: Classical-like (GATA6+), Basal-like (KRT17+), a KRT17+/GATA6+ Hybrid, and a Double-Negative (DN) state. We employed an integrated multimodal approach, including single-cell RNA sequencing, expanded multiplex IHC on 40 PDACs (>18,000 cells), and deep spatially-resolved proteomics (LCM-MS) on all four populations. This revealed a branching model of tumor evolution where proteomic analysis confirms the Hybrid state is a bona fide intermediate with a unique program enriched in aggressive tumor components, not a simple admixture. The DN state is also proteomically distinct, immunologically quiescent, and more proximal to the classical state, challenging linear progression models. Crucially, the transition to the aggressive Hybrid state is governed by a MYC-driven oncogenic program. Functional validation using CRISPR and in-situ dose-response studies established a direct KRT17-MYC regulatory axis. We confirmed that KRT17-expressing states (hybrid and basal-like) are resistant to standard chemotherapy (5-FU) but are exquisitely sensitive to MYC inhibition. This work redefines PDAC progression as a continuum of plastic cell states, establishes a mechanistic link between a biomarker (KRT17) and an oncogene (MYC), and provides a clear, biomarker-driven rationale for using KRT17/GATA6 status to guide MYC-targeted therapies to overcome resistance.