Construction And Validation Of A Prognostic Model For Anoikis-related Genes In Colorectal Cancer

Background:Colorectal cancer(CRC)is the third most common malignant tumor worldwide,causing nearly a million deaths each year.Due to the lack of obvious symptoms in the early stages,it is often not detected until later stages,resulting in poor prognosis for late-stage patients.In addition,due to the heterogeneity of CRC,there are significant differences in individual prognosis.Therefore,finding suitable CRC biomarkers is of great clinical significance.Anoikis is a special form of cell apoptosis caused by the loss or inappropriate cell adhesion.Currently,multiple studies have shown that anoikis is closely related to the occurrence and development of CRC,but the role of anoikis-related genes in CRC still needs further exploration.Objective:This study aimed to use bioinformatics methods to screen for anoikis-related genes(ARGs)with prognostic value and construct a prognostic risk model to evaluate the prognosis of CRC patients.At the same time,we investigate the role of core genes that make up the prognostic model in the occurrence and development of CRC.Methods:1.We used the GeneCards website to retrieve ARGs.Differential expression analysis and univariate Cox regression analysis were performed on the TCGA dataset to screen for differentially expressed ARGs that are associated with prognosis,followed by GO and KEGG enrichment analysis.2.We used LASSO and multivariate Cox regression analysis to screen for ARGs and construct a prognostic model.Based on the median of the risk score from the prognostic model,CRC patients were divided into high-and low-risk groups.The accuracy of the prognostic model was evaluated using survival curves(Kaplan-Meier,KM),time-dependent receiver operating characteristic(ROC)curves,and an external validation cohort.Furthermore,we explored the differences in enriched pathways,immune infiltration levels,and clinical characteristics between the high-and low-risk groups of the model.3.We constructed a nomogram that includes the risk score and multiple clinical features based on univariate and multivariate Cox regression analysis.The discrimination and accuracy of the model were evaluated using the C-index,calibration curve,decision curve analysis(DCA),and ROC curve.4.We validated the expression and prognostic significance of six core genes in CRC using both the TCGA and GSE39582 datasets.In addition,we used qRT-PCR technology to validate the differential expression of these six core genes in colon normal mucosal cell(HIEC)and CRC cells(SW480,HCT116).5.We conducted single-gene bioinformatics analysis on BDNF and LTB4R2,and assessed their effects on CRC cell proliferation,migration,and apoptosis using CCK-8 assay,colony formation assay,Transwell migration assay,cell scratch assay,and flow cytometry.Results:1.Based on the TCGA database,we screened out 21 differentially expressed ARGs that were also related to prognosis.Enrichment analysis showed that they were enriched in cellmatrix adhesion and pathways such as Ras,MAPK,and HIF-1.2.A linear prognostic model containing six ARGs(CDKN2A,BDNF,LTB4R2,CD36,KL,and NAT1)was constructed.KM curves,ROC curves,and external validation results showed that the model had good accuracy in predicting overall survival in CRC patients.GSEA enrichment analysis indicated that multiple cancer-related pathways were enriched in the high-risk group.Immune infiltration analysis revealed that immune responses were suppressed in patients of the high-risk group.Clinical feature analysis showed significant differences in pathological stage,mortality rate,and survival time between the high-and lowrisk groups(P<0.05).3.We integrated the risk score and clinical features to construct a nomogram for predicting the overall survival of CRC patients.The C-index(0.776±0.028),calibration curve,decision curve analysis,and ROC curve results indicated that the nomogram had good discriminative ability and accuracy.4.The results from TCGA and HPA databases showed that LTB4R2,BDNF,and CDKN2A were highly expressed in CRC tissues,while CD36,KL,and NAT1 were lowly expressed.In addition,qRT-PCR experiment results indicated that LTB4R2,BDNF,and CDKN2A were highly expressed in CRC cells,while CD36,KL,and NAT1 were lowly expressed.5.The results of single-gene bioinformatics analysis for BDNF and LTB4R2 showed differential expression in multiple cancer tissues,positively correlated with some immune cells,and enriched in multiple cancer-related pathways.In vitro cell functional experiments showed that knocking down BDNF and LTB4R2 can inhibit the proliferation and migration of CRC cells and promote apoptosis of CRC cells.Conclusions:1.21 differentially expressed ARGs with prognostic value were screened from public databases and could serve as potential biomarkers for CRC.A six-gene prognostic risk model was constructed based on CDKN2A,BDNF,LTB4R2,CD36,KL,and NAT1,which had predictive value for the prognosis of CRC patients.2.CDKN2A,BDNF,and LTB4R2 were highly expressed in CRC cells,while CD36,KL,and NAT1 were lowly expressed in CRC cells.BDNF and LTB4R2 were found to promote CRC cell proliferation and migration while inhibiting apoptosis,indicating their potential as new biological markers and therapeutic targets for CRC.