Identification Of Potential Therapeutic Targets And Prognostic Markers For Ferroptosis-associated Enzalutamide-Resistant Prostate Cancer

Background:Prostate cancer(PCa)is the most commonly diagnosed solid malignancy in men.Because of the lack of obvious symptoms in the early stage and no widely application of tumor screening,many patients have already progressed to a high-risk stage or even metastasis by the time of diagnosis.Currently,endocrine therapy is the standard treatment for prostate cancer,however,the disease usually progresses to castration-resistant prostate cancer(CRPC)after treatment among 18 to 24 months.Enzalutamide,a second-generation androgen receptor inhibitor,would prolong the progression of CRPC to some extent.But it can only extend the overall survival time around 4～6 months,due to the enzalutamide resistance.Therefore,exploring the specific mechanisms of enzalutamide resistance in prostate cancer and identifying new diagnostic and therapeutic targets as well as prognostic biomarkers is an urgent task at present.Objective:Bioinformatic approach was applied to investigate the prognostic biomarkers base on the analysis of potential molecular mechanisms of ferroptosis in enzalutamide resistant prostate cancer.The biomarkers could provide theoretical support for precisive treatment of prostate cancer.Methods:The gene sets(GSE104935,GSE78201)of enzalutamide-resistant and non-resistant prostate cancer cells were downloaded from the gene expression database(GEO)to screen differentially expressed genes.Genes related with ferroptosis were collected from FerrDb database.The differentially expressed genes in GEO database and ferroptosis genes in FerrDb database were intersected by ImageGP Venn diagram tool to obtain genes related with ferroptosis and enzalutamide resistance.The selected differentially expressed genes were enriched by R software for GO function(including biological process,molecular function,cellular component)and KEGG action pathway analysis.The selected genes were performed with univariate and multifactorial Cox regression analysis to investigate the potential prognostic value by R software based on the sequencing data and clinical data in prostate cancer with the American Institute for Cancer Research and the American Human Genome Research Institute(TCGA-PRAD).The expression levels of selected genes were analyzed in the TCGA-PRAD cohort using R software,and the Kaplan-Meier method was used to compare overall survival(OS),progression-free survival(PFS),and disease-specific survival(DSS)in those with high and low expression of these genes.The subject operating characteristic(ROC)curves predicting patients’ OS and PFS at 1,3 and 5 years were plotted using R software,and the predictive value was evaluated by the area under the curve(AUC).The association of independent prognostic genes with clinicopathological factors of prostate cancer was further explored and their ability to jointly predict patient prognosis was evaluated.The correlation of independent prognostic genes with the immune microenvironment of prostate cancer was subsequently clarified.Finally,the expression of marker genes in different clinicopathological features of prostate cancer was detected by immunohistochemistry.Results:A total of 31 differentially expressed genes in ferroptosis-associated enzalutamide-resistant prostate cancer were found.GO functional analysis showed that the main biological processes enriched in the differentially expressed genes were nutrient level response,cellular response to external stimuli,oxidative stress response,etc.The main cellular components enriched were autophagosomal membrane.As well,the significant enriched cellular components were autophagic membrane,secondary lysosomes,autophagosomes,etc.The main enriched molecular functions were steroid dehydrogenase activity,aldol NADP+1-oxidoreductase,acyl-CoA ligase activity,etc.;KEGG action pathway analysis showed that the significant enriched action pathways were arachidonic acid metabolism,autophagy,Kaposi’s sarcoma-associated herpesvirus infection,etc.Univariate and multivariate Cox regression analysis showed that LAMP2 was a prognostic biomarker for ferroptosis-associated enzalutamide-resistant prostate cancer.in the TCGA-PRAD cohort,the relative expression of LAMP2 was lower in prostate cancer tissues than in normal tissues(P<0.01),OS and PFS were lower in those with low LAMP2 expression than in those with high expression(both P<0.05),and the DSS and high The ROC curves showed that the AUCs of LAMP2 for predicting OS at 1,3 and 5 years were 0.825,0.747 and 0.770 for prostate cancer patients,and the AUCs for predicting PFS at 1,3 and 5 years were 0.539,0.601 and 0.568 for prostate cancer patients,respectively.immunohistochemical results showed that the expression levels of LAMP2 in CRPC tissues showed significantly lower expression levels than normal prostate tissues as well as hormone-dependent prostate cancer tissues.Conclusion:LAMP2 is a prognostic biomarker in ferroptosis-associated enzalutamide-resistant prostate cancer.The abnormally low expression of LAMP2 is an independent risk factor for poor prognosis in prostate cancer patients,and it has a good predictive value for patient prognosis.