Identifying Molecular Biomarkers Of Bladder Cancer Associated With Clinicopathological Features And Prognosis By Weighted Gene Co-expression Network Analysis

Part I:Identifying molecular biomarkers of bladder carcinoma associated with pathogenesis by weighted gene co-expression network analysisObjective To investigate the molecular biomarkers of bladder carcinoma associated with pathogenesis with a system-level WGCNA and the relevant packages in R software.Methods The RNA sequencing data sets and relative clinical information of bladder carcinoma patients were downloaded from the Gene Expression Omnibus(GEO)database.Three R packages,affy,limma and WGCNA were utilized to analysis the RNA-sequencing data.The establishment of the gene expression matrix was conducted and the differentially expressed genes(DEGs)between low stage bladder cancers(Ta-T1)and high stage bladder cancers(T2-T4)in the expressing data was screened.Then,cluster analysis was done based on the DEGs.The WGCNA package in R was applied to construct co-expression network for the DEGs.Defined gene significance(GS)and module significance(MS)were used to screen hub genes among interesting module.Correlation between hub genes and corresponding clinical feature was tested with one-way ANOVA and an independent sample t-test.The stability of target module,and the correlation between hub genes and the clinical feature were confirmed by an validation cohort.In the STRING database,protein-protein interaction based on the hub module was structured via Cytoscape tool.Enriched Gene Ontology terms was noted and KEGG analysis was also conducted in the DAVID database.Results A total of 874 DEGs including 454 up-regulated or 420 down-regulated gene were selected from 93 sample data for subsequent analysis.103 hub genes was found and four of them was network node of protein-protein interaction.There was significant relationship between COL3A1 and bladder carcinoma.The analyses of enriched GO and KEGG shown that high expression of COL3A1 was associated with MAPK signal pathway and focal adhesion related pathways.Conclusion Weight co-expression analysis confirmed that COL3A1 was significantly associated with the progression of bladder cancer.Part Ⅱ:Combination of data mining and clinical sample identifying to confirm the relationship between COL3A1 expression and bladder carcinomaObjective To identify whether COL3A1 mRNA or protein expression levels was associated with bladder cancer according to histological and clinical results from database and relative clinical samples.Methods Linear-regression analysis,survival analysis and ROC curve were conducted to validate the relative relationship between hub genes(COL3A1)and clinic prognosis based on the test set.Surgical specimens as carcinoma tissue and pericarcinomatous tissue were collected following radical cystectomy,respectively.Ocomine data base was performed to analysis the COL3A1 expression difference between normal and bladder cancer tissue.Kaplan-Meier survival curve was conducted according to the GEPIA database.The expression levels of COL3A1 in bladder carcinoma and pericarcinomatous tissue were detected using qRT-PCR methods.Tissue microarray of bladder carcinoma and pericarcinomatous were made.The expression levels of COL3A1 in the tissue microarray was detected via immunohistochemical methods.Results The results of linear-regression analysis shows that there was strong relationship between COL3A1 expression levels and RC stage.Based on Ocomine database,it was found that COL3A1 expression was significantly related to the progression of baldder cnacer.Kaplan-Meier survival curve obtained GEPIA database revealed that BCa patients with higher expression of COL3A1 had a significantly shorter overall survival time and disease free survival time.The results of survival analysis and ROC curves show that COL3A1 expression was associated with relative clinic prognosis.The results of qRT-PCR in bladder carcinoma and pericarcinomatous tissue shows that the expression levels of COL3A1 mRNA in the carcinoma tissue was significantly high than pericarcinomatous tissue(p<0.01).Immunohistochemistry results on tissue microarray shown that high COL3A1 protein expression levels were associated with higher grade of bladder cancer.Conclusion High expression levels of COL3A1 was associated with the pathogenesis and progression of bladder cancer.Part III:Knock-down and overexpression of COL3A1 affect biological behavior of bladder cancer cellsObjective To investigate the correlation between COL3A1 and the biological behavior of bladder cancer cells.Methods Through transient transfection on human bladder cancer cell line EJ and UMUC3 with COL3A1shRNA plasmid respectively,different cell lines,including EJ-shNC,UMUC3-shNC,EJ-shCOL3Al and UMUC3-sh COL3A1 were constructed.The ability of proliferation in different cell lines and expression level of MAPK and EMT relative proteins were detected by MTT and westernblot,respectively.Plate clone formation assay was also performed.Flow cytometry methd was applied to detect cell cycle and apoptosis rate.Wound healing and transwell assay of different cells line was conducted to detect the migration ability of cells.Results The bladder cancer cell line EJ of low COL3A1 expression was successfully constructed by sh-RNA technique.Proliferation ability of EJ and UKUC3 cell line,as well as protein expression of MAPK and EMT relative protein expression levels were significantly decreased when COL3A1 expression was reduced.The results of flow cytometry shows that cell cycle was blocked in the G1 stage and the apoptotic rate increased in the EJ-shCOL3A1 and UMUC3-sh COL3A1 cells respectively.The migration ability of bladder cancer cell was reduced in the EJ-shCOL3A1 and UMUC3-sh COL3A1 cells from the wound healing assay.Overexpression of COL3A1 can increase the UMUC3 cell vibility.Conclusion At the protein level,COL3A1 was confirmed to be a molecular biomarker of different histological grades of bladder cancer.COL3A1can regulate the proliferation and migration of bladder cancer cell,thus affecting the malignancy of bladder cancer.