Identification of a Hypoxia-Related Molecular Classification and Hypoxic Tumor Microenvironment Signature for Predicting the Prognosis of Patients with Triple-Negative Breast Cancer.

Xiaoli Sun,Yu Zhang,Cunli Yan,Huan Luo,Chenbo Han

doi:10.3389/fonc.2021.700062

Abstract

PurposeThe hypoxic tumor microenvironment was reported to be involved in different tumorigenesis mechanisms of triple-negative breast cancer (TNBC), such as invasion, immune evasion, chemoresistance, and metastasis. However, a systematic analysis of the prognostic prediction models based on multiple hypoxia-related genes (HRGs) has not been established in TNBC before in the literature. We aimed to develop and verify a hypoxia gene signature for prognostic prediction in TNBC patients.MethodsThe RNA sequencing profiles and clinical data of TNBC patients were generated from the TCGA, GSE103091, and METABRIC databases. The TNBC-specific differential HRGs (dHRGs) were obtained from differential expression analysis of hypoxia cultured TNBC cell lines compared with normoxic cell lines from the GEO database. Non-negative matrix factorization (NMF) method was then performed on the TNBC patients using the dHRGs to explore a novel molecular classification on the basis of the dHRG expression patterns. Prognosis-associated dHRGs were identified by univariate and multivariate Cox regression analysis to establish the prognostic risk score model.ResultsBased on the expressions of 205 dHRGs, all the patients in the TCGA training cohort were categorized into two subgroups, and the patients in Cluster 1 demonstrated worse OS than those in Cluster 2, which was validated in two independent cohorts. Additionally, the effects of somatic copy number variation (SCNV), somatic single nucleotide variation (SSNV), and methylation level on the expressions of dHRGs were also analyzed. Then, we performed Cox regression analyses to construct an HRG-based risk score model (3-gene dHRG signature), which could reliably discriminate the overall survival (OS) of high-risk and low-risk patients in TCGA, GSE103091, METABRIC, and BMCHH (qRT-PCR) cohorts.ConclusionsIn this study, a robust predictive signature was developed for patients with TNBC, indicating that the 3-gene dHRG model might serve as a potential prognostic biomarker for TNBC.

Highlights

Breast cancer is reported to be one of the most common causes of cancer-related deaths among females around the world, which is a heterogeneous tumor, resulting in variable clinical features [1]
A total of 492 differentially expressed genes (DEGs) in GSE104193 were screened between normoxia and hypoxia cultured breast cancer cells, including 242 upregulated genes and 150 downregulated genes (Supplementary Table S3), and were displayed in volcano plots and heat maps (Figures 1A, B)
A total of 555 DEGs in GSE33950 were identified in GSE33950, namely, 219 upregulated genes and 336 downregulated genes (Figures 1C, D and Supplementary Table S3)

Summary

Introduction

Breast cancer is reported to be one of the most common causes of cancer-related deaths among females around the world, which is a heterogeneous tumor, resulting in variable clinical features [1]. TNBC lacks the expression of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) and constitutes 12%– 18% of breast cancer patients [3]. Exploration of new therapeutic target and investigations of clinically applicable predictors are indispensable for TNBC. Targeting hypoxia will inhibit several traits of tumor progression, metastasis, radioresistance, and chemoresistance [10, 11], which has been an important focus of TNBC therapy. According to a previous study, certain hypoxiarelated genes (HRGs) and their mediators, hypoxia-inducible factors (HIFs), may serve as prognostic predictors and therapeutic targets in breast cancer [9]. A systematic analysis of the prognostic prediction models based on multiple HRGs have not been established in TNBC before in the literature

Objectives

Methods

Results