Elucidation of hypoxia-induced metastatic reprogramming through the regulation of KDM8 function in pancreatic cancer

ABSTRACT Pancreatic ductal adenocarcinoma (PDAC) is one of the deadliest malignancies. Full transcriptome defines clinically relevant “subtypes” of disease and the basal-like subtype of PDAC is associated with advanced stage, metastasis, resistance to treatment, and poor patient survival. The mechanism by which the tumor microenvironment drives PDAC subtype differentiation remains uncharacterized. Understanding the cancer- extrinsic factors in the tumor microenvironment that promotes the basal-like state could uncover cancer vulnerabilities and lead to the development of new therapeutic approaches. We found that inactivation of the histone demethylase gene Kdm8 reprograms PDAC cells into a highly metastatic state. Morphologically, Kdm8 inactivation reduces the expression of genes defining the classical PDAC subtype and drives a profound loss of differentiation in a genetically engineered PDAC mouse model. Importantly, the enzymatic function of Kdm8 requires molecular oxygen and hypoxia diminished the phenotypic changes induced by Kdm8 inactivation. We noted a global upregulation of histone 3 lysine 27 (H3K27) trimethylation upon the inactivation of Kdm8, supporting the involvement of its demethylase function in regulating the chromatin and cell state. In human PDAC, KDM8-regulated gene signatures are an exceptionally good predictor of the disease subtype. Our preliminary results suggest that hypoxia within the tumor microenvironment restricts the demethylase function of Kdm8, thereby suppressing the classical PDAC subtype while promoting a basal-like state and metastasis through epigenetic reprogramming. We hypothesize that a hypoxia-KDM8-chromatin axis plays a critical role in PDAC subtype determination and metastatic ability. In the proposed study, we aim to determine the mechanism by which Kdm8 controls PDAC subtypes and metastatic progression and the mechanism by which hypoxia suppresses Kdm8 function to promote chromatin reprogramming and PDAC metastasis. Overall, our study will fill the major gap in our understanding of how hypoxia within the PDAC tumor microenvironment promotes the basal-like molecular subtype and metastasis. Our mechanistic study will uncover fundamental insights into the role of KDM8 in subtype determination and may discover novel vulnerabilities of the basal-like PDAC state. The long-term objective is to gain a better understanding of the key cell-extrinsic factors in the tumor microenvironment that promote the aggressive or basal-like PDAC subtypes.