College of Bioinformatics Science and Technology, Harbin Medical University, Harbin, China.
Department of Gynecology, The First Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, China.
J Biol Chem. 2024 Sep;300(9):107710. doi: 10.1016/j.jbc.2024.107710. Epub 2024 Aug 22.
Molecular subtypes play a pivotal role in guiding preclinical and clinical risk assessment and treatment strategies in cancer. In this study, we extracted whole-tissue transcriptomic data from 1987 ovarian cancer patients spanning 26 independent Gene Expression Omnibus cohorts. A total of four consensus subtypes (C1-C4) were identified, notably, subtype C1 samples exhibited a poor prognosis and higher M2 macrophages infiltration, whereas subtype C2 samples demonstrated the best prognosis and higher CD4 resting T cells infiltration. Additionally, we characterized cancer- and stromal-specific gene expression profiles, and conducted an analysis of ligand-receptor interactions within these compartments. Based on cancer compartment, subtype-specific interactions as well as gene signatures for each molecular subtype were identified. Leveraging single-cell transcriptomic data, we delineated malignant epithelial cells with four molecular subtypes and observed an increase in C1 cell proportions from primary to relapse to metastasis stages, with a corresponding decrease in C2 cell proportions. Furthermore, we investigated subtype-specific interaction with T cells through integrated analysis of bulk and single-cell datasets. Finally, we developed a robust ten-gene risk model based on subtype gene signatures for prognostic evaluation in ovarian cancer, demonstrating its efficacy across independent datasets. In summary, this study systematically explored ovarian cancer molecular subtypes and provided a framework for other cancer types.
分子亚型在指导癌症的临床前和临床风险评估以及治疗策略方面发挥着关键作用。在这项研究中，我们从跨越 26 个独立基因表达综合数据集的 1987 名卵巢癌患者中提取了全组织转录组数据。总共确定了四个共识亚型（C1-C4），值得注意的是，亚型 C1 样本表现出较差的预后和更高的 M2 巨噬细胞浸润，而亚型 C2 样本则表现出最佳的预后和更高的 CD4 静止 T 细胞浸润。此外，我们还对肿瘤和基质特异性基因表达谱进行了特征分析，并对这些部位的配体-受体相互作用进行了分析。基于肿瘤区室，确定了亚型特异性相互作用以及每个分子亚型的基因特征。利用单细胞转录组数据，我们描绘了具有四个分子亚型的恶性上皮细胞，并观察到从原发性到复发再到转移阶段 C1 细胞比例增加，C2 细胞比例相应降低。此外，我们通过对批量和单细胞数据集的综合分析研究了与 T 细胞的特定亚型相互作用。最后，我们基于亚型基因特征开发了一种稳健的十基因风险模型，用于卵巢癌的预后评估，在独立数据集上证明了其有效性。总之，本研究系统地探讨了卵巢癌的分子亚型，并为其他癌症类型提供了一个框架。
