Prognostic Model Construction Of Basal Membrane Related Genes In Bladder Cancer Based On Bioinformatics Analysis

Background and Objective:Bladder cancer is the 10th largest malignant tumor in the world.Though the incidence rate of bladder cancer is not as high as that of prostate cancer in male tumors,it poses a great threat to human life and health with 510000 new cases and nearly 200000 deaths every year.According to the latest statistics of the World Health Organization,bladder cancer is the second largest urinary tumor in the world.Although the incidence rate of bladder cancer rate varies greatly between different geographical regions,it is expected to continue to rise in the next decade.Risk factors for bladder cancer include men,the elderly,white people,occupational exposure to certain chemicals,pelvic radiation,use of cyclophosphamide drugs,chronic bladder infection or irritation,personal or family history of bladder cancer and smoking.In addition to geographical location and age,the risks vary between sexes.Smoking is the most common type of exposure that is severely affected by many carcinogens.Mortality rates in developed countries have begun to decline,while mortality rates in low-income regions around the world have tended to rise.Because the incidence of bladder cancer is hidden,most patients generally go to the hospital after having more obvious clinical symptoms,resulting in illness delay.In addition,there is no specific marker with high accuracy for bladder cancer at present,so large-scale screening in the population cannot be achieved,which is also an important reason why patients with early disease cannot be effectively treated.Therefore,it is of great value to actively search for potential genes and other biomarkers related to bladder cancer.With the development of high throughput sequencing technology,more and more researchers are engaged in gene related research.As the starting point of transcription and translation,genes are closely related to the occurrence and development of tumors.Basal membranes(BMs)are extracellular matrix of cell adhesion that are widely distributed in metazoan tissues.BMs were first discovered in skeletal muscle 176 years ago.They are mainly derived from extracellular matrix transformation,including proteoglycans,collagen,and other components,as well as variations in basement membrane genes.It is also the foundation of many diseases,and abnormalities in the expression and turnover of basement membrane proteins are key pathogenic factors for diseases,including cancer.Research has shown that the network of BM related components provides BM with astonishing complexity,which may be the basis for the functional diversity of BM that is crucial to human health.Through bioinformatics analysis,the construction of a prognosis model of basement membrane related genes can provide a reliable scheme for early diagnosis,prognosis judgment and potential therapeutic targets of bladder cancer patients.Methods:Basement membrane related genes(BMGs)were downloaded from the BM database,and GSE13507 and GSE32894 chips were downloaded based on the GEO database as training sets and validation sets,respectively.In the training set,differentially expressed BMGs were screened and analyzed for functional enrichment of KEGG and GO.Prognostically significant BMGs were selected through single factor Cox regression and LASSO regression analysis.Then,LASSO regression was used to model prognostic differential genes.According to the median risk score,BC patients were divided into low-risk groups and high-risk groups,and K-M survival analysis was conducted.The ROC curve was used to test the predictive effectiveness of the model.In addition,the GSE32894 chip is used to validate the risk model.The R package was used to evaluate the immune cell score and stromal cell score of bladder cancer patients,and the ssGSAE analysis was used to analyze the difference of immune scoring between the training set and the verification set,to explore the differential expression of BMGs in immune cells of bladder cancer.Finally,GSEA analysis was conducted to reveal the potential function of gene differential expression between low risk and high-risk groups.Results:The expression profile data and clinical information of bladder cancer in GSE13507 dataset were downloaded from GEO database,and the clinical data of patients with basement membrane related genes were obtained from GSE32894 chip.BMGs with differences between bladder cancer samples and normal subjects were screened from the GSE13507 dataset,and 109 basement membrane related differential genes were screened,including 18 up-regulated genes and 91 down regulated genes.KEGG pathway and GO function enrichment analysis were performed on differentially expressed BMGs.The GO function enrichment analysis results showed that differential genes were highly enriched in the regulation of extracellular matrix tissue,collagen fiber tissue,vascular development,and positive regulation of cell adhesion,including integrin mediated cell adhesion,integrin mediated signal pathways,and extracellular structural tissue.KEGG results showed that differential genes were significantly enriched in the ECM receptor interaction and protein digestion and absorption pathways.After further univariate Cox regression and LASSO regression analysis,we finally screened out 3 genes that were significantly correlated with prognosis from the 9 genes that were obtained above,namely FBN1,TNC,COL13A1,COL18A1,NELL2,GPC2,COL5A1,ADAMTS1,and ITGA5,forming a basement membrane related prognostic risk scoring model,namely FBN1,COL13A1,and GPC2,According to the gene expression level of these BMGs and the relationship with the prognosis of bladder cancer,a prognostic risk model was established.The calculation formula of Risk-Score is:Risk-Score=(1.08716567079229*FBN1 expression)+(0.404634614078234*COL13A1 expression)+(GPC2*0.27196846828786 expression).According to the median value of RiskScore,patients in the training set and the validation set were divided into high and low risk groups,and K-M survival curves,survival state charts,and ROC curves were plotted.The results showed that as the risk score increased,the number of dead patients increased,and the overall survival rate of the high-risk group was lower than that of the low-risk group.The ROC curve results of the training set showed that the 1-year AUC was 0.694,the 3-year AUC was 0.724,and the 5-year AUC was 0.683;The 1-year,3-year and 5-year AUC values of the validation set are 0.707,0.725 and 0.761.The above evaluation results show that the bladder cancer risk score model has a good predictive effect on its prognosis.The results of independent prognostic analysis suggest that the prognostic risk model is independent of gender,age and tumor invasion.The prognostic risk model we established is an independent factor in predicting the prognosis of bladder cancer.We used R package to score the immune cells and stromal cells of patients with bladder cancer,and ssGSEA typing results to score 22 kinds of immune cells between the training group and the high-risk group of the test group.The results showed that as the patient’s risk score increased,the content of immune cells and stromal cells increased,and there was a difference in the content of dendritic cells and inflammatory factors between the high and low risk groups(p<0.05).After GSEAS enrichment analysis,it was suggested that TP binding cassette(ABC)transporters,glycerol phospholipid metabolism,glycosaminoglycan biosynthesis-heparan sulfate,glycosylphosphatidylinositol(GPI)anchored biosynthesis,O-glycan biosynthesis,ribosomes,etc.were significantly enriched in the low expression group.Chemokine signaling pathways,cytokine cytokine receptor interactions,interstitial nodes,neural signaling pathways,oocyte meiosis,and progesterone mediated oocyte maturation are significantly enriched in high expression.Conclusion:1.The prognosis model of basement membrane related genes in bladder cancer based on bioinformatics was successfully constructed and validated,and the results showed that the prediction accuracy was good.2.Age,tumor aggressiveness,and the model’s risk score were all independent prognostic risk factors for patients with bladder adenocarcinoma.3.Screening out 109 differentially expressed genes related to basement membrane,which are highly enriched in extracellular matrix tissue and cell matrix adhesion related pathways.Analysis of functional enrichment of KEGG and GO shows that basement membrane related genes are related to tumor migration and invasion4.The tumor microenvironment and ssGSAE analysis showed that the risk score was significantly positively correlated with the expression of immune cells,and the contents of dendritic cells and proinflammatory proteins were different between high-risk groups(p<0.05),which suggested that the tumor microenvironment and immune status might also be one of the factors affecting the prognosis of bladder cancer patients.