[摘 要] 目的：基于公开的卵巢癌（OC）单细胞测序数据并利用生物信息学方法构建OC上皮细胞亚型，探究上皮细胞异质性 及其对OC肿瘤微环境（TME）的影响。方法：使用GSE118828卵巢癌单细胞数据集的分析得到的8种细胞类型及基因表达矩阵 结果，使用反卷积方法估计整体转录组测序（bulk RNA-seq）数据中的细胞组成，依据上皮细胞丰度将OC分为高上皮细胞亚型 （C1型）和低上皮细胞亚型（C2型）。分析C1型与C2型患者的生存时间及免疫微环境差异。基于C1、C2型及OC早期（Ⅰ~Ⅱ 期）、晚期（Ⅲ~Ⅳ期）组间的差异表达基因（DEG），筛选与OC进展相关的上皮细胞相关基因并分析其对TME的影响。通过人类 蛋白质图谱（HPA）数据库数据验证OC进展中上皮细胞相关基因的蛋白表达水平。结果：C1型患者的OS明显优于C2型患者 （P < 0.05）。C1、C2型具有完全不同的免疫微环境浸润特性，C1型中辅助性T（Th）细胞、M1型巨噬细胞和激活的树突状细胞浸 润程度显著高于C2型（均P < 0.05）。晚期高上皮细胞OC患者HAS1、DAPL1、ADH1B基因表达增加（均P < 0.05），同时，DAPL1 基因在OC中的表达与巨噬细胞、Treg细胞的浸润程度呈正相关（P < 0.05）。HPA数据库数据证实，DAPL1蛋白在OC组织中表 达显著增加。结论：卵巢癌高上皮细胞亚型可能由于Th细胞、M1型巨噬细胞和激活的树突状细胞的高度浸润发挥抗肿瘤作用； 上皮细胞相关基因DAPL1的高表达可能会诱导巨噬细胞、Treg细胞的浸润程度，从而促进OC的进展。

[Abstract] Objective: To investigate epithelial cell heterogeneity in ovarian cancer (OC) and its impact on the tumor microenvironment (TME) by constructing OC epithelial subtypes using publicly available single-cell RNA sequencing (scRNA-seq) data and bioinformatics approaches. Methods: Eight cell types and gene expression matrices were identified from the GSE118828 OC scRNA-seq dataset, and a deconvolution approach was used to estimate cellular composition in bulk RNA sequencing (bulk RNA-seq) data. OC samples were stratified into high-epithelial (C1) and low-epithelial (C2) subtypes based on epithelial cell abundance. Differences in overall survival (OS) and immune microenvironment features between patients with C1 and C2 subtypes were analyzed. Differentially expressed genes (DEGs) between C1 and C2 subtypes, as well as between early-stage (stageⅠ~Ⅱ) and late-stage (stage Ⅲ-Ⅳ) OC groups were identified to screen epithelial-related genes associated with OC progression, and their influence on the TME was analyzed. Protein expression of the epithelial-related genes in OC progression was validated using data from the Human Protein Atlas (HPA) database. Results: Patients with the C1 subtype exhibited significantly better OS compared to those with C2 subtype (P < 0.05). The C1 and C2 subtypes demonstrated distinct immune microenvironment infiltration profiles. Infiltration levels of T helper (Th) cells, M1 macrophages, and activated dendritic cells (DCs) were significantly higher in the C1 subtype (all P < 0.05). In late-stage OC patients with C1 subtype, the expression of HAS1, DAPL1, and ADH1B was significantly upregulated (all P < 0.05). Furthermore, DAPL1 expression was positively correlated with the degree of macrophage and regulatory T cells (Treg) infiltration in OC (P < 0.05). HPA database analysis confirmed that DAPL1 protein expression was significantly upregulated in OC tissues. Conclusion: The high epithelial subtype of ovarian cancer may exert anti-tumor effects through extensive infiltration of T helper (Th) cells, M1 macrophages, and activated dendritic cells within the tumor microenvironment. Conversely, upregulation of the epithelial-related gene DAPL1 may induce the infiltration of macrophages and Treg, thereby promoting OC progression.

[基金项目] 云南省“万人计划”名医专项（No. YNWR-MY-2019-037）；国家自然科学地区科学基金（No. 82360579）；昆明医 科大学第二附属医院对外合作项目（No. 2022dwhz06）；云南省科技厅-昆医联合专项重点项目（NO：202401AY07 0001-053）