School of Medicine, Henan Polytechnic University, Jiaozuo, China.
School of Life Science and Technology, Xinxiang Medical University, Xinxiang, China.
PeerJ. 2024 Aug 5;12:e17861. doi: 10.7717/peerj.17861. eCollection 2024.
As a heterogeneous malignancy, breast cancer (BRCA) shows high incidence and mortality. Discovering novel molecular markers and developing reliable prognostic models may improve the survival of BCRA.
The RNA-seq data of BRCA patients were collected from the training set The Cancer Genome Atlas (TCGA)-BRCA and validation set GSE20685 in the Gene Expression Omnibus (GEO) databases. The "GSVA" R package was used to calculate the glycolysis score for each patient, based on which all the patients were divided into different glycolysis groups. The "limma" package was employed to perform differentially expression genes (DEGs) analysis. Key signature genes were selected by performing un/multivariate and least absolute shrinkage and selection operator (LASSO) C regression and used to develop a RiskScore model. The ESTIMATE and MCP-Counter algorithms were used for quantifying immune infiltration level. The functions of the genes were validated using Western blot, colony formation, transwell and wound-healing assay.
The glycolysis score and prognostic analysis showed that high glycolysis score was related to tumorigenesis pathway and a poor prognosis in BRCA as overactive glycolysis inhibited the normal functions of immune cells. Subsequently, we screened five key prognostic genes using the LASSO Cox regression analysis and used them to establish a RiskScore with a high classification efficiency. Based on the results of the RiskScore, it was found that patients in the high-risk group had significantly unfavorable immune infiltration and prognostic outcomes. A nomogram integrating the RiskScore could well predict the prognosis for BRCA patients. Knockdown of PSCA suppressed cell proliferation, invasion and migration of BRCA cells.
This study developed a glycolysis-related signature with five genes to distinguish between high-risk and low-risk BRCA patients. A nomogram developed on the basis of the RiskScore was reliable to predict BRCA survival. Our model provided clinical guidance for the treatment of BRCA patients.
作为一种异质性恶性肿瘤，乳腺癌（BRCA）具有较高的发病率和死亡率。发现新的分子标志物并建立可靠的预后模型可能会提高 BRCA 患者的生存率。
从 TCGA-BRCA 训练集和 GEO 数据库中的 GSE20685 验证集中收集 BRCA 患者的 RNA-seq 数据。基于“GSVA”R 包，根据每个患者的糖酵解评分，将所有患者分为不同的糖酵解组。使用“limma”包进行差异表达基因（DEGs）分析。通过执行单变量和多变量以及最小绝对值收缩和选择算子（LASSO）C 回归，选择关键签名基因，并用于开发 RiskScore 模型。使用 ESTIMATE 和 MCP-Counter 算法来量化免疫浸润水平。使用 Western blot、集落形成、Transwell 和划痕愈合实验验证基因的功能。
糖酵解评分和预后分析表明，高糖酵解评分与 BRCA 的肿瘤发生途径和不良预后相关，因为过度活跃的糖酵解会抑制免疫细胞的正常功能。随后，我们使用 LASSO Cox 回归分析筛选了五个关键预后基因，并使用它们建立了具有高分类效率的 RiskScore。基于 RiskScore 的结果，发现高风险组患者的免疫浸润和预后结果明显不利。一个整合了 RiskScore 的列线图可以很好地预测 BRCA 患者的预后。敲低 PSCA 可抑制 BRCA 细胞的增殖、侵袭和迁移。
本研究建立了一个由五个基因组成的与糖酵解相关的特征，以区分高危和低危 BRCA 患者。基于 RiskScore 建立的列线图可可靠地预测 BRCA 的生存情况。我们的模型为 BRCA 患者的治疗提供了临床指导。
