Abstract A023: Analysis of p53 function in a plastic injury-associated cell state suggests a noncanonical route to PDAC

Abstract

Although PDAC is thought to arise from premalignant lesions known as PanINs, the frequency of PanINs in adult pancreata is much higher than PDAC incidence in the population, raising the question of whether other routes to PDAC exist. While oncogenic Kras expression drives PDAC initiation, malignant progression relies on mutation of tumor suppressor genes, like TP53, which is mutated in ∼75% of PDACs. Despite this critical role of p53 as a barrier to PDAC, we have limited understanding of how p53 blocks PDAC progression. Understanding how p53 restrains PDAC may lead to important insights into early detection and therapeutic strategies. An important framework for understanding how p53 suppresses PDAC comes from our work in KrasG12D-driven lung adenocarcinoma, where we showed that p53 acts in a plastic transitional cell state to drive AT1 cell differentiation to suppress cancer. We showed that this tumor suppressive mechanism reflects repurposing of a critical p53 role in promoting lung injury repair. To test whether this paradigm is conserved in PDAC, we have leveraged an adult mouse PDAC model, in which the Ptf1aCreER allele is used to drive KrasG12D expression, in the presence or absence of p53, in acinar cells of 8-10 week old mice (KTC and KPTC mice, respectively). We have integrated scRNA-seq, spatial transcriptomics, scATAC-seq, and multiplexed IF, to deconstruct tumor evolution in these mice. scRNA-seq and spatial transcriptomics has revealed the full heterogeneity of premalignant populations in KTC and KPTC mice. Strikingly, we found that p53 activity is predominant in a plastic injury-associated transitional cell akin to that observed in lung cancer, and this cell state is conserved in human pancreas. We found further that p53 inactivation results in a dramatic shift in the chromatin landscape that is poised for PDAC progression. Analysis of transcription factor modules, inferred CNVs, and transcriptional distances suggest a strong similarity of the plastic transitional cell state with cancer. Trajectory analyses suggest further that this transitional cell could directly lead to cancer, via a noncanonical path of PDAC development, which we are currently testing by lineage tracing. Our chromatin analyses have suggested specific transcription factors that may promote cancer with p53 inactivation. Finally, analysis of the premalignant lesion microenvironment in KTC and KPTC mouse pancreata suggests significant differences in cellular signaling and composition. Together, our findings suggest an important role for p53 in a plastic injury-related transitional cell where it suppresses PDAC via effects on epithelial cells and the microenvironment.