A new ferroptosis-related gene model for prognostic prediction of papillary thyroid carcinoma

Abstract

Papillary thyroid carcinoma (PTC) is a highly heterogeneous malignancy with diverse prognoses. Ferroptosis is a new type of cell death dependent on iron. Nevertheless, the predictive ability of ferroptosis-related genes for PTC is unclear. Based on the mRNA expression information from The Cancer Genome Atlas, we compared tumor and normal tissues in terms of the gene expression, for identifying differentially expressed genes (DEGs). Then, the risk score of a 5-gene signature was calculated and a prognostic model was established to test the predictive value of this gene signature by virtue of the LASSO Cox regression. The 5 genes were validated in PTC tissues by RT-qPCR.At last, functional analysis was implemented to investigate the underlying mechanisms. We found a total of 45 ferroptosis-related genes expressed differentially between tumor and normal tissues. 6 DEGs exhibited a significant relevance to the overall survival (OS) (P< 0.05). We classified patients into group with high risk and group with low risk based on the median risk score of a 5-gene signature. Patients in the group with low risk presented a remarkably higher OS relative to the group with high risk (P< 0.01). The Cox regression analysis displayed the independent predictive ability of the risk score. The receiver operating characteristic analysis helped to validate the predictive power owned by the gene signature. After validation, the 5 genes were abnormally expressed between PTC and normal tissues. Functional analysis showed two groups had different immune status. A new ferroptosis-related gene signature can predict the outcomes of PTC patients.

Keywords: Ferroptosis; gene model; overall survival; papillary thyroid carcinoma; prognosis.

Graphical abstract

Figure 1.

Identification of prognostic ferroptosis-related differentially expressed genes.a.Venn plot of the differentially expressed genes between tumor and normal tissue that were correlated with OS.b. Forest plot of the results of the univariate Cox regression analysis between gene expression and OS.c.Heatmap of the differentially expressed genes associated with OS.d. The correlation of the differentially expressed genes associated with OS

Figure 2.

Prognostic analysis of the 5-gene signature model. a. The distribution and median value of the risk scores. b. The distributions of OS status, OS and risk score. c. PCA analysis of the TCGA cohort. d. t-SNE analysis of the TCGA cohort. e. Kaplan–Meier curves of the OS in the two groups. f.AUC of time-dependent ROC curves evaluated the prognostic capacity of the risk score

Figure 3.

Results of univariate and multivariate Cox regression analysis on OS

Figure 4.

Functional enrichment analysis of DEGs. a. Top 10 biological process (BP) terms, cellular components (CC) terms, molecular functions (MF) terms. b. Top 6 Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways

Figure 5.

Comparison of the ssGSEA scores between different risk groups.a. The scores of 16 immune cells.b.The scores of 13 immune-related functions. Adjusted P values were showed as: ns, not significant; *, P< 0.05; **,P< 0.01; ***, P< 0.001

Figure 6.

AUC of time-dependent ROC curves of the internal validation

Figure 7.

RT-qPCR detecting the mRNA expression levels of DPP4 (a), GSS (b), HMGCR (c), PGD (d), TFRC (e) in PTC and normal tissues