Analysis of gastric cancer transcriptome allows the identification of histotype specific molecular signatures with prognostic potential

Abstract

Gastric cancer is the fifth most common malignancy but the third leading cause of cancer-associated mortality worldwide. Therapy for gastric cancer remain largely suboptimal making the identification of novel therapeutic targets an urgent medical need. In the present study we have carried out a high-throughput sequencing of transcriptome expression in patients with gastric cancers. Twenty-four patients, among a series of 53, who underwent an attempt of curative surgery for gastric cancers in a single center, were enrolled. Patients were sub-grouped according to their histopathology into diffuse and intestinal types, and the transcriptome of the two subgroups assessed by RNAseq analysis and compared to the normal gastric mucosa. The results of this investigation demonstrated that the two histopathology phenotypes express two different patterns of gene expression. A total of 2064 transcripts were differentially expressed between neoplastic and non neoplastic tissues: 772 were specific for the intestinal type and 407 for the diffuse type. Only 885 transcripts were simultaneously differentially expressed by both tumors. The per pathway analysis demonstrated an enrichment of extracellular matrix and immune dysfunction in the intestinal type including CXCR2, CXCR1, FPR2, CARD14, EFNA2, AQ9, TRIP13, KLK11 and GHRL. At the univariate analysis reduced levels AQP9 was found to be a negative predictor of 4 years survival. In the diffuse type low levels CXCR2 and high levels of CARD14 mRNA were negative predictors of 4 years survival. In summary, we have identified a group of genes differentially regulated in the intestinal and diffuse histo-types of gastric cancers with AQP9, CARD14 and CXCR2 impacting on patients prognosis, although CXCR2 is the only factor independently impacting overall survival. Simple summaryGastric cancer is the fifth most common malignancy and the third leading cause of cancer-associated mortality worldwide. Although several new pharmacological approaches are currently developed, surgery remains the unique valid option of treatment but survival remains very poor over the last decades. Therefore, understanding the underlying molecular mechanisms of the gastric carcinogenesis and identifying sensitive biomarkers could be helpful for the prevention and treatment of the disease. Currently, the high-throughput sequencing techniques, in particular the transcriptomic analysis (RNA-seq) represents a validated technique to obtain a molecular characterization of human cancers. Moreover, it has been established that genetic susceptibility and environmental factors, such as microbial infections may contribute to carcinogenesis. We have characterized the different patterns of gene expression, using RNA-seq analysis and correlated these findings with gastric cancer histological subtypes.