Construction Of A Prognostic Risk Model For Breast Cancer Based On Ferritinophagy-Associated Genes

Objective:Breast cancer is one of the most common malignancies worldwide and it has high mortality in women.The prognosis of breast cancer patients is poor and heterogeneous,and there is an urgent need for effective therapeutic targets and more accurate prognostic assessments.There is growing evidence that ferritinophagy is strongly associated with many human diseases by regulating ferroptosis.However,the molecular mechanism and prognostic role of ferritinophagy-associated genes in breast cancer are unclear.In this study,the prognostic model of breast cancer patients was established by screening ferritinophagyassociated genes related to the prognosis of breast cancer patients.The prognostic value of prognostic models was assessed by survival analysis.It aims to achieve accurate diagnosis and treatment of breast cancer patients and improve their prognosis.Methods:The m RNA expression profile and clinical data of breast cancer(TCGA-BRCA)were downloaded from TCGA and ferritinophagy-associated genes were obtained from Gene Cards database.These genes were performed in breast cancer and normal samples and finally obtained differentially expressed ferritinophagy-associated genes(DEFGs).Gene Ontology(GO)analysis and Kyoto Encyclopedia of Genes and Genomes(KEGG)analysis were performed using the cluster Profiler software package to identify the biological functions and related signaling pathways of differential ferritinophagy-associated genes.A protein-protein interaction network of differential ferritinophagy-associated genes was constructed using the STRING database and visualized by Cytoscape software.Univariate and multivariate Cox regression analysis was used to identify candidate genes for the prognostic risk model.According to the median multivariate regression risk score,the patients included in the model were divided into high and low score groups.Receiver operating characteristic(ROC)curve analysis and Kaplan-Meier(K-M)analysis were used to evaluate the predictive efficacy of the prognostic model.Calibration analysis and the C index to evaluate the accuracy of the prognostic model.The probability of individual survival was predicted by drawing a nomogram of the combined risk score of clinical features.The contribution of the model to the clinical patient benefit is validated using the clinical decision curve(DCA).Gene set enrichment analysis for high and low score groups was performed using gene set enrichment analysis(GSEA).Immunoinfiltration analysis was performed using the CIBERSORT.The mi Rwalk website and the Ch EA3 website were used to build a prognostic ferritinophagyassociated genes regulatory network for further clinical research.The key prognostic genes were further verified by q RT-PCR experiments.Results were visualized using the R software.Results:Expression profile data of 113 paracancerous tissue samples and 1109 breast tumor samples were downloaded from the TCGA database.Excluding incomplete data,645 breast cancer patients with complete information were included in the analysis.Fifteen ferritinophagy-associated genes were obtained through the Gene Cards database.These genes were performed in breast cancer and normal samples and finally obtained 9 differentially expressed ferritinophagy-associated genes.Further functional enrichment analyses revealed that ferritinophagy-associated genes were significantly enriched in iron metabolism-related processes and autophagy-related functions.The gene expression data of TCGA-BRCA breast cancer tumor tissue were randomly divided into training set(n=962)and validation set(n=107)according to the ratio of 9:1.The first six ferritinophagy-associated genes(NCOA4,ELAVL1,FTH1,FBXW7,USP24 and ATG16L1)of high importance were extracted for inclusion in the construction of the final random forest model.Clinical subgroup survival analysis showed that all six prognostic genes were significantly correlated with the survival time of breast cancer and its subgroups.Cox regression analysis was performed by random forest model and clinical parameters,and a prognostic risk score model was constructed to predict the prognosis of breast cancer patients.Using Kaplan–Meier analysis,time-dependent ROC curves and Cox regression analysis to evaluate the survival prediction ability of the model.The Kaplan–Meier analysis showed that patients in the high-risk group tend to have a worse prognosis.The timedependent ROC curves showed that the AUC values predicted by the model for 3-year,5-year and 7-year survival rates of breast cancer patients were 0.941,0.946 and 0.892,respectively.Multivariate Cox regression analysis showed that the risk score was an independent risk factor for overall survival(OS)in breast cancer patients.A nomogram was constructed by combining risk scores with traditional clinical prognostic factors,showing that the prognosis of tumor patients could be better predicted by combining multiple indicators.The regulatory relationship network between prognostic ferritinophagy-associated gene-mi RNA and prognostic ferritinophagy-associated gene-TF was constructed.It has been shown that the prognostic iron autophagy gene is involved in a wide range of signal regulation processes in the body.In addition,q RT-PCR detection was performed on 6 human breast cancer cell lines and human normal breast epithelial cell lines.The results showed that ELAVL1 and NCOA4 genes were highly expressed in breast cancer cell lines,and FBXW7,FTH1 and USP24 genes were low expressed in breast cancer cell lines compared to normal breast epithelial cells.Conclusions:Our study constructed a novel six ferritinophagy-associated genes risk model to predict the prognosis of breast cancer,which may provide new directions for the diagnosis and treatment of breast cancer.