Possible Roles Of Tumor-related Genes In Colorectal Cancer With Weighted Co-expression Network Analysis

Colorectal cancer(CRC)is one of common malignant tumors in digestive system of human being.Although the therapy of CRC has been greatly improved during last decades,the death rate is still high,and the pathogenesis is not completely clear.Thus,more studies of molecular pathogenesis of CRC have been given at present.In order to investigate the potential roles of new cancer-related genes in CRC and their effects on patient survival,161 tumor tissues/cases and 136 non-tumor tissues/cases of chip information were collected from the Gene Expression Omnibus database.Com Bat method was used for eliminating the batch effect of the microarray data from different sources,and then the ‘limma’ R package was performed to screen differentially expressed genes.At first step,we found 1776 differentially expressed cancer-related genes,including 742 up-regulated genes and 1024 down-regulated genes.Afterwards,the data from 161 tumor tissues/cases were analyzed with Weighted Gene Co-expression Network Analysis.The differentially expressed genes were clustered to seven modules,and two modules(black color and pink color)were finally selected,based on the conservative analysis.Then,using Gene Set Enrichment Analysis,cytokine-cytokine receptor interaction and chemotactic signaling pathways were found in association with CRC development at black module,and classic Ras,PI3K-Akt and MPKA signaling pathways were found directly or indirectly in regulation of this disease progress in pink one.In addition,FDCSP,FAM46 C,SLAMF7,MS4A1,NEDD9,PLAGL2,CACNA1 D and RUBCNL 8 genes were selected from the black and pink modules for their significant relationship with patient survival(p<0.05),resulting from single-gene survival analysis by using The Cancer Genome Atlas data.PLAGL2,CACNA1 D and RUBCNL 3 genes affect patient survival,because they have a cross in 5-year of survival.The panel of MS4A1 and NEDD9 may be as new biomarkers for the prognosis of CRC.by the Multi-variant Cox Proportional Hazard Regression Analysis,and MS4A1.had independent impacted on patient survival of CRC in this panel.After we verified that the protein expressive levels of CACNA1 D,RUBCNL and NEDD9 were higher in CRC tissues than those in paratumor tissues via immunohistochemistry.Especially,the expressive proteins of NEDD9 and CACNA1 D in CRC were related to TNM staging,clinical staging and pathological classification.Although the expression of NEDD9 protein was also higher in CRC tissues,which was also consistent with previous reports,its m RNA abnormal expression was opposite to its changing tendency in GEO and TCGA databases.So,it needs to reliably validate at RNA level by increasing the samples and using different methods.In conclusion,through bioinformatics analysis,we found that m RNA expressive levels of FDCSP,FAM46 C,SLAMF7,MS4A1,NEDD9,PLAGL2,CACNA1 D and RUBCNL 8 genes were associated with 5-year survival rate of patients in this study.And the panel of MS4A1 and NEDD9 two-gene might be served as new biomarkers for CRC prognosis.In addition,CACNA1 D,RUBCNL and NEDD9 expressive levels of m RNA and protein in CRC were significantly abnormal with potential physiological and pathological significance.It is worth for further research,without regard to the m RNA result was opposite to protein result of NEDD9 by immunohistochemistry.