PO.IM03.01 · 免疫学
HPV18 阳性头颈癌细胞的单细胞 ATAC 测序分析揭示 HPV18 基因组可及性的异质性及其与癌症进展的关联
Single-cell ATAC-sequencing profiling of HPV18 positive head and neck cancer cells reveals heterogeneity of HPV18 genome accessibility and its association with cancer progression
作者与单位 Authors & Affiliations
摘要 Abstract
中文摘要
背景：HPV 是头颈癌（HNC）的主要致癌驱动因素。然而，HPV 基因组的表观基因组异质性——尤其是单细胞染色质可及性模式——及其与癌症进展的相关性仍不明确。本研究刻画了患者来源的 HPV 阳性 HNC 细胞的染色质可及性图谱，以鉴定与肿瘤异质性及潜在侵袭性亚群相关的调控特征。
方法：采用条件重编程细胞方法从患者肿瘤活检组织中生成原代 HPV18⁺ HNC 细胞。使用人-HPV18 联合参考基因组，通过单细胞 ATAC 测序评估染色质可及性。HPV18 基因组上的可及性模式由二值化的峰矩阵编码。我们对全基因组可及性谱应用 UMAP，并使用核密度估计来检查具有不同 HPV18 峰的细胞分布。在各 HPV18 可及性组间进行了差异可及性、峰-基因连锁和通路富集分析。
结果：我们在 HPV18 基因组内鉴定出三个主要可及峰。Peak 1 或 Peak 3 阳性的细胞在 UMAP 上呈弥散、无偏倚的分布，而 Peak 2 阳性细胞则强烈富集于离散的簇中，提示该峰的可及性代表肿瘤内异质性的关键标志。我们还观察到异质性的 HPV18 整合，将不同的 HPV18 峰与来自同一患者细胞中不同的人类基因组位点相连接，并与位点特异性的可及性变化相关。测序深度分析证实，HPV18 峰模式的差异是生物学性的，而非技术性伪影。峰-基因连锁分析显示，UGGT1 和 COL5A1 的转录起始位点在 Peak 2 阳性亚群中更易可及。基于 TCGA 数据，这两个基因均与宫颈癌和 HNC 的不良预后相关。通路富集显示了与神经元迁移、上皮形态发生和组织形态发生相关程序的激活，表明该亚群可能具有增强的侵袭性和干性特征。
结论：本研究提供了首个患者来源 HPV18⁺ HNC 细胞的单细胞染色质可及性图谱，并揭示了 HPV18 病毒基因组内部的显著异质性。一个特定的 HPV18 可及峰（Peak 2）标记着一个独特的、潜在具侵袭性的细胞亚群，其特征为不良预后相关基因的可及性和干性相关通路。这些发现凸显了一个可能促进 HNC 进展的新型病毒-表观基因组标志物，可进一步探索用于预后或治疗应用。
查看英文原文 English abstract
Background: HPV is a major oncogenic driver of head and neck cancers (HNC). However, the epigenomic heterogeneity of the HPV genome-particularly single-cell chromatin accessibility patterns-and its relevance to cancer progression remain unclear. This study characterizes the chromatin-accessibility landscape of patient-derived HPV positive HNC cells to identify regulatory signatures associated with tumor heterogeneity and potential aggressive subpopulations.
Methods: Primary HPV18⁺ HNC cells were generated using the conditional reprogramming cell method from patient tumor biopsies. Single-cell ATAC-sequencing was performed to assess chromatin accessibility using a combined human-HPV18 reference genome. Accessibility patterns across the HPV18 genome were encoded by a binary peak matrix. We applied UMAP on genome-wide accessibility profiles and used kernel density estimation to examine the distribution of cells with distinct HPV18 peaks. Differential accessibility, peak-to-gene linkage, and pathway-enrichment were performed across HPV18 accessibility groups.
Results: We identified three major accessible peaks within the HPV18 genome. Cells positive for Peak 1 or Peak 3 showed a dispersed, unbiased distribution on UMAP, whereas Peak 2-positive cells were strongly enriched in discrete clusters, suggesting that accessibility of this peak represents a key marker of intratumoral heterogeneity. We also observed heterogeneous HPV18 integration linking different HPV18 peaks to distinct human genomic loci across cells from the same patient, associated with locus-specific accessibility changes. Sequencing-depth analysis confirmed that differences in HPV18 peak patterns were biological rather than technical artifacts. Peak-to-gene linkage analysis revealed that UGGT1 and COL5A1 transcription start sites were more accessible in the Peak-2-positive subpopulation. Both genes correlate with poor prognosis in cervical cancer and HNC based on TCGA data. Pathway enrichment demonstrated activation of programs associated with neuron migration, epithelial morphogenesis, and tissue morphogenesis, indicating that this subpopulation may possess enhanced invasiveness and stemness features.
Conclusions: This study provides the first single-cell chromatin accessibility map of patient-derived HPV18⁺ HNC cells and reveals substantial heterogeneity within the HPV18 viral genome. A specific HPV18-accessible peak (Peak 2) marks a distinct, potentially aggressive cellular subpopulation characterized by accessibility of poor-prognosis-associated genes and stemness-related pathways. These findings highlight a novel viral-epigenomic marker that may contribute to HNC progression and could be explored for prognostic or therapeutic applications.
利益披露 Disclosure
C. Cao, None..
B. Zhao, None.