The screening and analysis of protein signatures and signaling associated with chemoresistance based on Protein Pathway Array technology in gastric cancer.

Abstract

The present study was aimed to identify proteins associated with signaling pathways involved in chemoresistance, and establish a predictive model for chemoresistance in gastric cancer patients after radical surgery. A total of 140clinically-stagedIII gastric cancer samples from patients after D2radical gastrectomy were enrolled in the present study. Protein Pathway Array (PPA) and 286antibodies were used to assess the protein expression in tumor tissues of patients. The Significance Analysis of Microarray (SAM) software and clustering and discriminant analysis were used to identify differentially expressed proteins between chemosensitive and chemoresistant subsets, and a predictive model for chemoresistance was established using the independent predictive factors. The Ingenuity Pathway Analysis (IPA) software was also used to investigate the relationship between proteins and the signaling transduction network. A total of 23proteins were differentially expressed between 67chemosensitive and 73chemoresitant tumor tissues. Six proteins including PLK1 and DACH1 were independent risk factors for chemoresistance. A predictive model for chemoresistance by these proteins was established, and the accuracy, the sensitivity, and the specificity of this modal was 89.3, 90.3 and 88.2%, respectively. In addition, the present study revealed that differentially expressed proteins were closely related to cellular activity, DNA methylation and DNA damage and repair, and also involved in the ERK/MAPK, Wnt/β-catenin, PI3K/AKT, apoptosis and p53 signaling pathways. In conclusion, the predictive model established by PPA may be an effective detection system for predicting the chemosensitivity of gastric cancer patients after D2 gastrectomy.