Vertical inhibition of the autophagy pathway promotes growth arrest and apoptotic cell death in pancreatic ductal adenocarcinoma

Pancreatic ductal adenocarcinoma (PDAC) is dependent on autophagy for growth. The sole FDA approved inhibitor of autophagy, chloroquine/hydroxychloroquine (CQ/HCQ), inhibited the growth of multiple preclinical models of PDAC. However, clinical trials using CQ alone or in combination with standards of care have shown that CQ has limited clinical efficacy. To identify genes to target as sensitizers to CQ treatment, we performed a CQ-anchored CRISPR-Cas9 mediated loss-of-function screen. We identified multiple autophagy-related genes. This result led us to hypothesize that concurrent inhibition of two nodes (i.e. vertical inhibition) of the autophagy pathway is a more effective method of inhibiting cell growth. Specifically, loss of PIK3C3 , the gene that encodes VPS34, a protein essential for autophagosome nucleation, sensitized PDAC cells to CQ treatment. We validated this finding via both genetic depletion and pharmacological inhibition with the compound SAR405. Additionally, pharmacological inhibition of ULK1/2, the serine/threonine kinases critical for autophagy initiation, further reduced CQ-mediated anti-proliferative effects. We recently demonstrated that inhibition of the lipid kinase PIKfyve with apilimod potently suppressed autophagy in PDAC. We found that both ULK1/2 and VPS34 inhibitor treatment sensitized cells to PIKfyve inhibition.To elucidate the mechanisms driving decreased cellular proliferation following combination treatment, we explored both autophagy-dependent and -independent pathways using reverse phase protein array (RPPA) based protein pathway activation mapping. We determined that vertical autophagy inhibition resulted in metabolic rewiring including enhanced receptor tyrosine kinase and PI3K-AKT-mTORC1 signaling. Furthermore, by monitoring levels of the autophagy-related protein LC3B following CQ or apilimod treatment alone or in combination with ULK1/2 inhibition, we demonstrated that vertical autophagy inhibition impairs both autophagosome formation and autophagosome processing, simultaneously. We also found that cleaved caspases accumulated more in the presence of two autophagy inhibitors. We validated this finding using Annexin V apoptosis assays and demonstrated that growth suppression was due in part to enhanced apoptotic cell death. In summary, our findings demonstrate that vertical inhibition of the autophagy pathway more effectively suppresses autophagy and increases cell death in PDAC and provides rationale for a vertical inhibition strategy to enhance the therapeutic efficacy of autophagy inhibition.