颅咽管瘤免疫微环境的单细胞转录组学分析及PE_EGR2亚群生长信号通路研究

Single-Cell Transcriptomic Analysis of the Immune Microenvironment in Adamantinomatous Craniopharyngioma and Growth Signaling Pathways of the PE_EGR2 Subpopulation

摘要: 目的：系统解析成釉细胞型颅咽管瘤(Adamantinomatous Craniopharyngioma, ACP)肿瘤免疫微环境的细胞组成与功能状态，并重点探讨高干性肿瘤上皮亚群PE_EGR2的生长信号通路通讯网络，为ACP靶向治疗提供新的理论依据。方法：整合5个公开数据集共81,050个单细胞转录组数据，涵盖21例样本。采用Harmony算法进行批次效应校正，重建UMAP低维嵌入。通过Wilcoxon秩和检验鉴定免疫细胞标志基因，利用CytoTRACE 2深度学习框架评估肿瘤上皮细胞分化潜能，并以LIANA rank_aggregate共识方法推断细胞间配体–受体通讯。结果：共鉴定27种细胞类型，免疫细胞占44.1%，肿瘤上皮细胞占37.8%。髓系细胞(以M0样驻留巨噬细胞为主，占免疫细胞41.4%)构成免疫微环境的核心成分，呈现显著的M2样免疫抑制极化特征。T细胞亚群中检测到明显的耗竭信号(PDCD1、HAVCR2、TIGIT高表达)。CytoTRACE 2分析显示PE_EGR2亚群具有最高干性潜能(中位评分0.435)。细胞通讯分析识别出172对显著配体–受体互作，其中33对涉及生长通路，PE_EGR2通过GSTP1/ANXA1 → EGFR轴向其他肿瘤上皮亚群发送强烈的EGF信号，同时接收来自神经胶质细胞的FGF14 → FGFR2信号。结论：ACP肿瘤微环境呈现高度免疫抑制状态，PE_EGR2亚群兼具高干性和活跃的生长信号发送能力，是ACP肿瘤维持与扩张的潜在关键驱动亚群，EGFR抑制剂与FGFR抑制剂联合免疫检查点阻断可能是ACP综合治疗的潜在策略。

Abstract: Object: To systematically characterize the cellular composition and functional states of the tumor immune microenvironment (TME) in adamantinomatous craniopharyngioma (ACP), with a focus on elucidating the growth signaling communication network of the high-stemness tumor epithelial subpopulation PE_EGR2, thereby providing a theoretical basis for targeted therapy of ACP. Methods: Single-cell transcriptomic data from 81,050 cells across 21 samples were integrated from five public datasets. Batch effects were corrected using the Harmony algorithm, followed by UMAP dimensionality reduction. Wilcoxon rank-sum tests were applied to identify immune cell marker genes. The CytoTRACE 2 deep learning framework was used to assess the differentiation potential of tumor epithelial cells, and the LIANA rank_aggregate consensus method was employed to infer ligand-receptor-mediated intercellular communication. Results: A total of 27 cell types were identified, with immune cells comprising 44.1% and tumor epithelial cells 37.8% of all cells. Myeloid cells, predominantly M0-like resident macrophages (41.4% of immune cells), constituted the core of the immune microenvironment and exhibited prominent M2-like immunosuppressive polarization. T cell subpopulations displayed marked exhaustion signatures, characterized by high expression of PDCD1, HAVCR2, and TIGIT. CytoTRACE 2 analysis revealed that the PE_EGR2 subpopulation possessed the highest stemness potential (median score 0.435). Cell communication analysis identified 172 significant ligand-receptor interactions, of which 33 involved growth-related pathways. PE_EGR2 transmitted potent EGF signals to other tumor epithelial subpopulations via the GSTP1/ANXA1 → EGFR axis, while receiving FGF14 → FGFR2 signals from glial cells. Conclusion: The ACP tumor microenvironment is characterized by profound immunosuppression. The PE_EGR2 subpopulation combines high stemness with active growth signal transmission, representing a potential key driver of ACP maintenance and expansion. Combined targeting of EGFR/FGFR signaling with immune checkpoint blockade may constitute a promising therapeutic strategy for ACP.