| Objective: Cancer is one of the leading causes of death in the world and is a major problem that all mankind wants to overcome.It has been reported that the incidence of colorectal cancer(CRC)is increasing year by year and has become one of the most common malignant tumor in the world,and colon adenocarcinoma(COAD)is among the most common ones.There are many causes for cancer development,and oxidative stress(OS)is one of them.It can be present in multiple stages of cancer development,including playing a key role in the transformation of normal cells to cancer cells,proliferation of cancer cells,formation of tumor blood vessels and metastasis of tumors.Epidemiological studies have linked OS with cancer,and studies have been conducted to establish prognostic models of OS and cancers such as gastric,bladder,and pancreatic cancers,but there is a lack of prognostic models for OS associated with COAD,so this study was conducted to screen differentially expressed OS-related genes in COAD by performing bioinformatics analysis and establishing a prognostic model,aiming to investigate whether OS-related genes are associated with the prognosis of COAD patients and to predict potential therapeutic agents for COAD.Methods: Transcriptome expression data and relevant clinical information of COAD patients were obtained from the Cancer Genome Atlas(TCGA)database and OS-related genes with a score greater than 7 were obtained from the Gene Cards database.The limma R package was used for differential expression of genes to screen for differential expression of oxidative stress genes(DE-OSGs)between normal and COAD samples.Selected DE-OSGs were then subjected to gene ontology(GO)annotation and Kyoto Encyclopedia of Genes and Genomes(KEGG)analysis using the R language cluster analyzer to further understand the biological processes and signaling pathways they are involved in.The patients were then randomly divided into a training set and a test set in a5:5 ratio.The training group was used to construct prognostic models for OS-related genes,which were calculated using Cox proportional risk regression analysis.The accuracy and reliability of the prognostic model was evaluated using the area under the subject’s operating characteristic curve,and then the predictive effect of the prognostic model was tested with the test set and the total set.Immediately afterwards,univariate Cox regression and multivariate Cox regression analyses were used to analyze whether the prognostic risk model could be used as an independent predictor and to predict the survival of individual COAD patients were predicted by constructing Nomogram plots.Finally,the sensitivity of patients in high and low groups to each drug was predicted by drug sensitivity analysis,and the correlation between the prognostic risk model and each drug was also investigated by Spearman correlation analysis.Results:1.210 differentially expressed genes related associated with OS were screened using the limma package,and these differentially expressed genes were mainly enriched in the IL-17 signaling pathway,AGE-RAGE signaling pathway and P53 signaling pathway by GO and KEGG functional enrichment analysis.2.Univariate Cox regression,LASSO regression and multivariate Cox proportional risk regression analysis were used to construct a prognostic model of OS-related genes independent of age,gender and TMN tumor stage,which included eight variables of NOL3,CAV1,KCNE2,MAPK12,CPT2,ACADL,IL13 and GSTM2,and could classify patients into high risk and low risk groups,and can predict the prognosis of COAD patients.The results of Kaplan-Meier survival analysis showed that the prognosis of patients in the high risk group was worse in the training set,test set and total set(P =7.203e-07,P = 1.88e-02,P = 1.496e-07).3.In independent prognostic analyses of age,gender,Stage,T-stage,M-stage,N-stage and OS-related genetic prognostic models,in univariate regression analyses,age(P =0.014),Stage(P < 0.001),T-stage(P < 0.001),M-stage(P < 0.001),N-stage(P < 0.001)and risk Score(P < 0.001)met P < 0.05;in multivariate regression analysis,only age(P =0.012)and risk Score(P = 0.001)met P < 0.05.Taken together,age and prognostic model risk scores met P < 0.05 in both univariate and multivariate regression analyses and could be used as independent factors for the prognosis of COAD.4.The Nomogram plot was constructed by combining patients’ age,gender,stage and risk score,which can accurately predict the survival of COAD patients.The AUC values at 1,3 and 5 years showed that the Nomogram plot AUC values of the OS-related gene prognostic model were the largest,with AUC values of 0.782 at 1 year,0.854 at 3 years and 0.852 at 5 years.P < 0.001 for the Nomogram plot model in univariate Cox regression analysis and P = 0.040 for the Nomogram plot model in multivariate Cox regression analysis,all of which met the value of P < 0.05 and were shown to be independent determinants of COAD prognosis.5.Drug sensitivity analysis revealed significant differences in the sensitivity of the four drugs Erlotinib(P = 0.0014),PHA-665752(P = 0.0071),Salubrinal(P = 0.0013)and YM155(P = 0.00027)between the high and low risk groups,with values meeting P <0.05,and the high risk group patients had lower IC50 values compared to the low risk group,indicating that patients in the high risk group were more sensitive to the drug.Conclusion: In this study,we used large data screening analysis to establish a prognostic model of eight genes associated with OS including NOL3,CAV1,KCNE2,MAPK12,CPT2,ACADL,IL13 and GSTM2,which could predict the prognosis of COAD and could be used as an independent prognostic factor for COAD.Erlotinib,PHA-665752,Salubrinal and YM155 were also found to be potential therapeutic agents for COAD. |
PDF Full Text Request