Crosstalk between hypoxia-sensing ULK1/2 and YAP-driven glycolysis fuels pancreatic ductal adenocarcinoma development

Autophagy and glycolysis are two catabolic processes that manipulate pancreatic ductal adenocarcinoma (PDAC) development in response to hypoxia sensing, yet the underlying mechanism of how they are interlinked remain elusive. Methods: The functional roles of Unc-51 like kinase 1 and 2 (ULK1/2) in pyruvate kinase M2 (PKM2) transcription and glycolysis under hypoxia were assessed by chromatin immunoprecipitation, luciferase reporter, glucose consumption and lactate production assay. Co-immunoprecipitation, cellular ubiquitination, His-pulldown, in vitro protein kinase assay, immunofluorescence, immunohistochemistry, CRISPR technology, in silico studies were adopted to determine the molecular mechanism. Correlation analyses were performed in KPC (Pdx1-Cre; LSL-Kras^G12D/+; Trp53^fl/+) mice and clinical samples from PDAC patients. Therapeutic potential of ULK1/2 inhibitor and 2-deoxyglucose (2-DG) or 3-bromopyruvate (3-BP) was evaluated in cell-derived xenograft (CDX) and the patient-derived xenograft (PDX) models of nude mice. Results: ULK1/2, but not ULK3, augments hypoxic glycolysis in PDAC cells mediated by PKM2 independent of BCL2/adenovirus E1B 19 kDa interacting protein 3 (BNIP3). Mechanistically, hypoxia stimulates ULK1 to translocate into nucleus, where it interacts with and phosphorylates yes-associated protein (YAP) at Ser227, resulting in YAP stabilization through blockade of ubiquitin-proteasome system (UPS), which in turn facilitates PKM2 transcription, glycolysis, cell proliferation in vitro as well as PDAC growth in mice. ULK1/2 is positively correlated with YAP and PKM2 in tumor tissues from KPC mice and clinical samples from PDAC patients. Pharmacological deactivation of ULK1/2 potentiates the antineoplastic efficacy of 2-DG and 3-BP in CDX and PDX models. Conclusion: Our findings underscore the Ser227 autophosphorylation-dependent nuclear YAP stabilization as a central node that couples ULK1/2-initiated autophagy to hypoxic glycolysis during PDAC development and propose that targeting ULK1/2 combined with 2-DG or 3-BP might be a feasible therapeutic strategy against PDAC.