Upregulation of the ferroptosis-related STEAP3 gene is a specific predictor of poor triple-negative breast cancer patient outcomes

Abstract

Objective: This study was designed to assess ferroptosis regulator gene (FRG) expression patterns in patients with TNBC based on data derived from The Cancer Genome Atlas (TCGA). Further, it was utilized to establish a TNBC FRG signature, after which the association between this signature and the tumor immune microenvironment (TIME) composition was assessed, and relevant prognostic factors were explored.

Methods: The TCGA database was used to obtain RNA expression datasets and clinical information about 190 TNBC patients, after which a prognostic TNBC-related FRG signature was established using a least absolute shrinkage and selection operator (LASSO) Cox regression approach. These results were validated with separate data from the Gene Expression Omnibus (GEO). The TNBC-specific prognostic gene was identified via this method. The STEAP3 was then validated through Western immunoblotting, immunohistochemical staining, and quantitative real-time polymerase chain reaction (RT-qPCR) analyses of clinical tissue samples and TNBC cell lines. Chemotherapy interactions and predicted drug sensitivity studies were investigated to learn more about the potential clinical relevance of these observations.

Results: These data revealed that 87 FRGs were differentially expressed when comparing TNBC tumors and healthy tissue samples (87/259, 33.59%). Seven of these genes (CA9, CISD1, STEAP3, HMOX1, DUSP1, TAZ, HBA1) are significantly related to the overall survival of TNBC patients. Kaplan-Meier analyses and established FRG signatures and nomograms identified CISD1 and STEAP3 genes of prognostic relevance. Prognostic Risk Score values were positively correlated with the infiltration of CD4+ T cells (p = 0.001) and myeloid dendritic cells (p =0.004). Further evidence showed that STEAP3 was strongly and specifically associated with TNBC patient OS (P<0.05). The results above were confirmed by additional examinations of STEAP3 expression changes in TNBC patient samples and cell lines. High STEAP3 levels were negatively correlated with half-maximal inhibitory concentration (IC50) values for GSK1904529A (IGF1R inhibitor), AS601245 (JNK inhibitor), XMD8-85 (Erk5 inhibitor), Gefitinib, Sorafenib, and 5-Fluorouracil (P < 0.05) in patients with TNBC based on information derived from the TCGA-TNBC dataset.

Conclusion: In the present study, a novel FRG model was developed and used to forecast the prognosis of TNBC patients accurately. Furthermore, it was discovered that STEAP3 was highly overexpressed in people with TNBC and associated with overall survival rates, laying the groundwork for the eventually targeted therapy of individuals with this form of cancer.

Keywords: STEAP3; ferroptosis; overall survival; prognostic signature; triple-negative breast cancer (TNBC).

Figure 1

The schematic representation of the study workflow.

Figure 2

Identification of TNBC-related DE-FRGs in the TCGA database. DE-FRGs that were differentially expressed FRGs between TNBC tumors and healthy tissues (FDR < 0.05, |Log2 (Fold Change)| > 2) were presented in the form of Volcano plots (A), in which blue and red dots respectively correspond to up-regulated and downregulated DE-FRGs; Heat maps (B), in which each dot and its color (Red is high-expression, blue is low-expression) indicate the expression value of each DE-FRGs in different samples, the greater the expression level, the darker the color. (C) Venn diagrams were used to identify TNBC-associated FRGs.

Figure 3

Identification and assessment of prognostic TNBC-associated FRGs. (A, B) Forest plot-based identification of TNBC patient risk factors identified through univariate (A) and multivariate (B) Cox regression analyses. (C) A nomogram was established based on multivariate Cox regression analysis results. (D) Calibration plot for the established Nomogram. (E) Evaluation of the developed Nomogram based on Kaplan-Meier OS curves.

Figure 4

Development of an 87 DE-FRG-based prognostic risk signature in the TCGA-TNBC cohort. (A) LASSO coefficient profiles for 87 DE-FRGs. (B) LASSO regression analyses with 10-fold cross-validation yielded 16 prognostic DE-FRGs based on a minimum λ value. (C, E) OS distributions, OS status, and risk scores for patients in the TCGA-TNBC cohort. (D) Kaplan-Meier curves corresponding to the OS of TCGA-TNBC patients stratified into low- and high-risk groups. (F) Z-scores corresponding to the expression of the 16 prognostic DE-FRGs included in the established risk signature. (G) AUC values for time-dependent ROC curves were employed to assess the predictive utility of prognostic signature-derived risk scores.

Figure 5

Correlations between prognostic risk score values and immune cell infiltration. (A-F) Correlations between predictive risk scores and the six indicated immune cell types.

Figure 6

Analyses of the relationship between CISD1 and STEAP3 expression and the survival of BRCA and TNBC patients. (A, B) The relationship between the expression of CISD1 and the OS of BC (P<0.05) and TNBC patients (P<0.05). (C, D) The relationship between the expression of STEAP3 and the OS of BC (P>0.05) and TNBC patients (P<0.05).

Figure 7

Validation of TNBC patient STEAP3 expression levels in GEO datasets. (A, B) STEAP3 expression levels were compared between TNBC and normal tissue samples in the selected TCGA (P<0.0001) and GEO (P<0.001) datasets. (C, D) AUC analyses for the TCGA and GEO datasets.

Figure 8

Validation of STEAP3 expression in breast cancer cell lines and tissue samples. (A) STEAP3 mRNA levels were assessed in the control MCF-10A cell line and the MDA-MB-231, BT-549, and BT-468 TNBC cell lines. (B, C) STEAP3 levels were detected via Western immunoblotting in the MCF-10A, MDA-MB-468, and MDA-MB-231 cell lines and in six pairs of TNBC (T) and adjacent normal (N) tissue samples from patients. (D) Representative IHC staining results for STEAP3 in adjacent normal tissues (Scale bar: 100 μm) and TNBC samples (Scale bars:100μm and 50 μm), and non-TNBC(Luminal A) samples (Scale bar: 50μm). (E) Quantitative data from IHC staining results for STEAP3 expression in 23 TNBC vs adjacent normal breast,12 non-TNBC vs adjacent normal breast and 23 TNBC vs 12 non-TNBC are shown. (**P<0.01, ***P< 0.001, ^ns P>0.05).

Figure 9

Correlations between STEAP3 expression and drug sensitivity in patients with TNBC. (A-F) The expression of STEAP3 and IC50 values correspond to 5−Fluorouracil, GSK1904529A, AS601245, XMD8−85, and Gefitinib in TNBC patients included in the TCGA-TNBC cohort (P < 0.05). (G, H) Correlations between STEAP3 expression and Sorafenib IC50 values with further details regarding expression in the low, high, and normal groups (P<0.01). (****P< 0.0001, ns P>0.05).