Chang Hua, Zhu Yuyan, Zheng Jiahui, Chen Lian, Lin Jiaxing, Yao Jihang
Department of Gynaecology, The First Hospital of China Medical University, Shenyang, China.
Department of Urology, The First Hospital of China Medical University, Shenyang, China.
J Oncol. 2021 Nov 24;2021:1331031. doi: 10.1155/2021/1331031. eCollection 2021.
High-grade serous ovarian cancer (HGSOC) carries the highest mortality in the gynecological cancers; however, therapeutic outcomes have not significantly improved in recent decades. Macrophages play an essential role in the occurrence and development of ovarian cancer, so the mechanisms of macrophage infiltration should be elucidated.
We downloaded transcriptome data of ovarian cancers from the Gene Expression Omnibus and The Cancer Genome Atlas. After rigorous screening, 1566 HGSOC were used for data analysis. CIBERSORT was used to estimate the level of macrophage infiltration and WGCNA was used to identify macrophage-related modules. We constructed a macrophage-related prognostic model using machine learning LASSO algorithm and verified it using multiple HGSOC cohorts.
In the GPL570-OV cohort, high infiltration level of M1 macrophages was associated with a good outcome, while high infiltration level of M2 macrophages was associated with poor outcomes. We used WGCNA to select genes correlated with macrophage infiltration. These genes were used to construct protein-protein interaction maps of macrophage infiltration. IFL44L, RSAD2, IFIT3, MX1, IFIH1, IFI44, and ISG15 were the hub genes in the network. We then constructed a macrophage-related prognostic model composed of CD38, ACE2, BATF2, HLA-DOB, and WARS. The model had the ability to predict the overall survival rate of HGSOC patients in GPL570-OV, GPL6480-OV, TCGA-OV, GSE50088, and GSE26712. In exploring the immune microenvironment, we found that CD4 memory T cells and activated mast cells showed that the degree of infiltration was higher in the high-risk group, while M1 macrophages were the opposite, and HLA molecules were overexpressed in the high-risk group.
We constructed a macrophage infiltration-related protein interaction network that provides a basis for studying macrophages in HGSOC. Our macrophage-related prognostic model is robust and widely applicable. It predicts overall survival in HGSOC patients and may improve HGSOC treatment.
高级别浆液性卵巢癌（HGSOC）在妇科癌症中死亡率最高；然而，近几十年来治疗效果并未显著改善。巨噬细胞在卵巢癌的发生和发展中起重要作用，因此应阐明巨噬细胞浸润的机制。
我们从基因表达综合数据库（Gene Expression Omnibus）和癌症基因组图谱（The Cancer Genome Atlas）下载了卵巢癌的转录组数据。经过严格筛选，1566例HGSOC用于数据分析。使用CIBERSORT估计巨噬细胞浸润水平，使用加权基因共表达网络分析（WGCNA）识别巨噬细胞相关模块。我们使用机器学习LASSO算法构建了巨噬细胞相关的预后模型，并使用多个HGSOC队列进行验证。
在GPL570 - OV队列中，M1巨噬细胞高浸润水平与良好预后相关，而M2巨噬细胞高浸润水平与不良预后相关。我们使用WGCNA选择与巨噬细胞浸润相关的基因。这些基因用于构建巨噬细胞浸润的蛋白质 - 蛋白质相互作用图谱。IFL44L、RSAD2、IFIT3、MX1、IFI44、IFI44和ISG15是网络中的枢纽基因。然后我们构建了一个由CD38、ACE2、BATF2、HLA - DOB和WARS组成的巨噬细胞相关预后模型。该模型能够预测GPL570 - OV、GPL6480 - OV、TCGA - OV、GSE50088和GSE26712中HGSOC患者的总生存率。在探索免疫微环境时，我们发现CD4记忆T细胞和活化肥大细胞在高危组中的浸润程度较高，而M1巨噬细胞则相反，并且HLA分子在高危组中过表达。
我们构建了一个巨噬细胞浸润相关的蛋白质相互作用网络，为研究HGSOC中的巨噬细胞提供了基础。我们的巨噬细胞相关预后模型稳健且广泛适用。它可预测HGSOC患者的总生存率，并可能改善HGSOC的治疗。
