Abstract: Objective To study the expression of Xklp2 target protein (targeting protein for Xenopus kinesin-like protein2, TPX2) in pancreatic carcinoma and its potential value in predicting prognosis. Methods Based on 200 genes related to glycolysis, the information of pancreatic patients was obtained from TCGA and GEO databases, and the characteristics of glycolysis metabolism of the latter were evaluated. At the same time, the metabolic characteristics and differential genes of the patients were evaluated, the risk model was established by Cox regression and Lasso regression analysis, and the key glycolysis related gene TPX2 was obtained. The prognostic validity of the model was verified by GEO database based on the expression of prognostic genes, the overall survival (OS) of patients with pancreatic cancer, the difference in the expression of prognostic genes between normal pancreatic tissue and pancreatic cancer was analyzed and finally the function of prognostic genes was verified at the cellular level. Results The case cohort was divided into three characteristic groups: high, medium and low glycolysis metabolism. Patients with low-rate glycolysis metabolism had certain survival advantages. TPX2 was the key gene of this model. Pancreatic cancer patients showed high expression of TPX2 and had poor prognosis. Knocking down TPX2 at cellular level can inhibit the proliferation and colony formation of pancreatic cancer cells. Conclusions The model based on glycolysis metabolism can effectively predict the prognosis of patients with pancreatic cancer. TPX2 is expected to become a potential marker gene for predicting the prognosis of pancreatic cancer.

Key words: TPX2, glycolysis, pancreatic carcinoma, proliferation