Identification of a ferroptosis-related lncRNA signature with prognosis for Wilms tumor

Abstract

Background: Wilms tumor (WT) is a widespread urologic tumor in children. Ferroptosis, on the other hand, is a novel form of cell death associated with tumor development. In this study, we aim to explore the predictability of ferroptosis-related biomarkers in estimating prognosis in WT patients.

Methods: To determine a link between ferroptosis-related gene expression and WT prognosis, we first collected RNA sequencing data and clinical information, involving 124 WT and 6 healthy tissue samples, from the TARGET database. Next, we screened the collected information for ferroptosis-related long non-coding RNA using Cox regression analysis, and constructed a signature model, as well as a nomogram, related to prognosis. Finally, we explored a potential link between ferroptosis-related lncRNA and tumor immunity and screened for possible immune checkpoints.

Results: We constructed a WT prognosis prediction signature containing 12 ferroptosis-related lncRNAs. The area under the curves values, from the ROC curves, predicting overall survival rates at the 1, 3-, and 5-year timepoints were 0.775, 0.867, and 0.891 respectively. Moreover, we generated a nomogram, using clinical features and risk scores, carrying a C-index value of 0.836, which suggested a high predictive value. We also demonstrated significant differences in tumor immunity between low- and high-risk WT patients, particularly in the presence of B cells, NK cells, Th1 cells, Treg cells, inflammation promoting, and type I and II IFN responses. In addition, we showed that immune checkpoints like SIRPA, ICOSLG, LAG3, PVRIG, NECTIN1, and SIRPB2 can serve as potential therapeutic targets for WT.

Conclusions: Based on our analyses, we generated a ferroptosis-related lncRNA signature that can both estimate prognosis of WT patients and may provide basis for future WT therapy.

Keywords: TARGET; Wilms tumor (WT); ferroptosis; lncRNAs; prognosis.

Figure 1

Identification of ferroptosis-related lncRNAs in WT patients. (A) Volcano plot of DElncRNAs. (B) Forest plots of the hazard ratios (HR) of the top 10 ferroptosis-related lncRNAs. WT, Wilms tumor.

Figure 2

Prognosis analysis of the ferroptosis-related lncRNA signature in TARGET database. (A) LASSO coefficient profiles of the 26 ferroptosis-related lncRNAs correlated with OS; (B) Confidence interval under each lambda; (C) Kaplan-Meier curve and (D) ROC curve of the ferroptosis-related lncRNA signature in WT patients; (E) Prognostic classifier analyses (risk scores, survival status and lncRNAs expression) in distinguishing patients into low- and high-risk groups. WT, Wilms tumor; OS, overall survival.

Figure 3

Clinical significance of ferroptosis-related lncRNAs in WT. (A) Correlations between ferroptosis-related lncRNAs and clinical characteristics; (B) Univariate and (C) multivariate Cox regression analysis of clinical characteristics and the signature-based risk score; (D) ROC curve of clinical characteristics and the signature-based risk score. WT, Wilms tumor.

Figure 4

A nomogram of clinical characteristics and ferroptosis-related lncRNAs. (A) Prediction of 1-, 3- and 5-year survival of WT patients based on the nomogram; (B) Calibration curves between actual and predicted survival at 1-, 3- and 5-year.

Figure 5

The relationship and functional enrichment between ferroptosis-related lncRNA and mRNA expression. (A) The ferroptosis-related lncRNA-mRNA co-expression network; (B) The Sankey diagram of RNAs in co-expression network; (C) GO and (D) KEGG analyses of genes included in the co-expression network.

Figure 6

Gene set enrichment analysis of low- and high-risk groups based on the ferroptosis-related lncRNA signature.

Figure 7

Heatmap of immune status in low- and high-risk groups based on ESTIMATE, MCPcounter, CIBERSORT and ssGSEA algorithms.

Figure 8

Immunological characteristics of the ferroptosis-related lncRNA signature. Boxplots of (A) immune cells and (B) immune-related functions between low- and high-risk groups by ssGSEA scores; (C) Boxplots of immune checkpoints expression of two groups. *, P<0.05; **, P<0.01; ***, P<0.001 (adjusted P values).