A bioinformatics analysis, pre-clinical and clinical conception of autophagy in pancreatic cancer: Complexity and simplicity in crosstalk.

Pancreatic cancer (PC) is a serious gastrointestinal tract disease for which the 5-year survival rate is less than 10%, even in developed countries such as the USA. The genomic profile alterations and dysregulated biological mechanisms commonly occur in PC. Macroautophagy/autophagy is a cell death process that is maintained at a basal level in physiological conditions, whereas its level often changes during tumorigenesis. The function of autophagy in human cancers is dual and can be oncogenic and onco-suppressor. Autophagy is a potent controller of tumorigenesis in PC. The supportive autophagy in PC escalates the growth rate of PC cells and its suppression can mediate cell death. Autophagy also determines the metastasis of PC cells, and it can control the EMT in affecting migration. Moreover, starvation and hypoxia can stimulate glycolysis, and glycolysis induction can be mediated by autophagy in enhancing tumorigenesis in PC. Furthermore, protective autophagy stimulates drug resistance and gemcitabine resistance in PC cells, and its inhibition can enhance radiosensitivity. Autophagy can degrade MHC-I to mediate immune evasion and also regulates polarization of macrophages in the tumor microenvironment. Modulation of autophagy activity is provided by silibinin, ursolic acid, chrysin and huaier in the treatment of PC. Non-coding RNAs are also controllers of autophagy in PC and its inhibition can improve therapy response in patients. Moreover, mitophagy shows dysregulation in PC, which can enhance the proliferation of PC cells. Therefore, a bioinformatics analysis demonstrates the dysregulation of autophagy-related proteins and genes in PC as biomarkers.