Disease-related protein co-expression networks are associated with the prognosis of resectable node-positive pancreatic ductal adenocarcinoma

Abstract Pancreatic ductal adenocarcinoma (PDAC) is a multifactorial disease, the molecular profile of which remains unclear. This study aimed at unveiling the disease-related protein networks associated with different outcomes of resectable, node-positive PDAC cases. We assessed laser-microdissected cancerous cells from PDAC tissues of a poor outcome group (POG; n = 4) and a better outcome group (BOG; n = 4). Noncancerous pancreatic duct tissues (n = 5) were used as the reference. We identified four representative network modules by applying a weighted network correlation analysis to the obtained quantitative PDAC proteome datasets. Two network modules that were significant for POG were associated with the heat shock response to hypoxia-related stress; in the latter, a large involvement of the non-canonical Hedgehog pathway (regulated by GLI1), the internal ribosome entry site-mediated cap-independent translation, the inositol requiring enzyme 1-alpha (IRE1α)/X-box binding protein 1 pathway of the unfolding protein response (UPR), and the aerobic glycolysis was observed. By contrast, the BOG characteristic module was involved in the inactivation of the UPR pathway via the synoviolin 1-dependent proteasomal degradation of IRE1α, the activation of SOX2, and the loss of PALB2 (partner and localizer of BRCA2) function, all potentially suppressing malignant tumor development. Our findings might facilitate future therapeutic strategies for PDAC.