Comprehensive analysis of ferroptosis-related genes and prognosis of cutaneous melanoma
Changjiang Liu,Yuhang Liu,Yifeng Yu,Yong Zhao,Aixi Yu
DOI: https://doi.org/10.1186/s12920-022-01194-z
2022-03-01
Abstract:Background: Cutaneous Melanoma (CM) is a malignant disease with increasing incidence and high mortality. Ferroptosis is a new kind of cell death and related to tumor blood and lymphatic metastasis. This study aims at using bioinformatics technology to construct a prognostic signature and identify ferroptosis-related biomarkers to improve the prognosis and treatment of cutaneous melanoma. Methods: We used bioinformatics tools to analyze RNA sequencing expression data with clinical information from multiple databases, utilized varieties of statistical methods to construct a ferroptosis-related prognostic signature of cutaneous melanoma and screened out specific genes with independent prognostic ability. Results: We obtained 22 ferroptosis-related (P < 0.05) prognostic DEGs in the uniCox regression analysis, among which 10 high-expressed genes (ATG5, CHAC1, FANCD2, FBXL5, HMOX2, HSPB1, NQO1, PEBP1, PRNP, SLC3A2) were screened out by LASSO regression analysis to establish a predictive model. Meanwhile, the ferroptosis-related signature and the nomogram we drew performed an excellent performance based on Kaplan-Meier (K-M), Receiver operating characteristic (ROC) and calibration curves. Univariate and multivariable cox analyses displayed that our model was greater than other prognostic features. GO and KEGG analyses revealed that 10-biomarker signature was mainly related to epidermis differentiation and immunity. ssGSEA analysis indicated that the immune status between the two risk groups was highly different. Besides, we found that two genes (CP, ZEB1) had independent prognostic ability and can be applied for drug research. Both genes were highly related to immunity. GSEA illustrated that ZEB1 may be involved in cellular functions such as proliferation, apoptosis, and migration, while CP was closely connected to immune cell related functions. Conclusion: The present study suggested a 10-biomarker signature can be clinically used to predict the prognosis of cutaneous melanoma, which was better than conventional factors. CP and ZEB1 were independent prognostic genes and can be applied to guide treatment. In addition, ZEB1 mutation was highly related to overall survival in cutaneous melanoma, while CP may be associated with tumor progression. Our study comprehensively analyzed the relationship between iron metabolism, ferroptosis-related genes, and the prognosis of cutaneous melanoma, provided new insight for molecular mechanisms and treatment of ferroptosis and cutaneous melanoma.