Scholarlink

<p>Tumors are embedded within complex microenvironments composed of malignant, immune and stromal cells, yet how coordinated cellular organization shapes tumor biology across cancers remains incompletely understood. Here we integrate single-cell transcriptomic profiles from over two million cells spanning 19 human epithelial cancers to construct a comprehensive pan-cancer atlas of tumor microenvironment (TME) organization. We identify cell subsets selectively enriched in tumors compared with normal tissues and show that their relative abundance is associated with clinical outcomes. By clustering tumors according to their cellular composition, we define eight recurrent cellular ecosystem subtypes that capture coordinated variation in immune and stromal architectures across cancer types. Comparative analyses reveal that these ecosystem states are associated with distinct tumor-intrinsic molecular programs, including differences in signaling pathway activity, transcriptional regulation and copy number alterations. We further uncover subtype-specific patterns of ligand–receptor interactions linking malignant cells with immune and stromal compartments and provide independent spatial support for these interactions using spatial transcriptomic data. Together, this work provides a unified framework for tumor ecosystem stratification and highlights cellular organization as a key dimension of tumor heterogeneity with potential implications for precision cancer therapy.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Single-cell profiling defines cellular ecosystems and malignant cell state coupling across pan-cancer tumor microenvironments

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月25日

查看全文
  • Xiaoxiao Yang,
  • Yang Tong,
  • Yuting Wang,
  • Peng Xie,
  • Xin Li,
  • Jiapei Yuan,
  • Yang Yang

摘要

Tumors are embedded within complex microenvironments composed of malignant, immune and stromal cells, yet how coordinated cellular organization shapes tumor biology across cancers remains incompletely understood. Here we integrate single-cell transcriptomic profiles from over two million cells spanning 19 human epithelial cancers to construct a comprehensive pan-cancer atlas of tumor microenvironment (TME) organization. We identify cell subsets selectively enriched in tumors compared with normal tissues and show that their relative abundance is associated with clinical outcomes. By clustering tumors according to their cellular composition, we define eight recurrent cellular ecosystem subtypes that capture coordinated variation in immune and stromal architectures across cancer types. Comparative analyses reveal that these ecosystem states are associated with distinct tumor-intrinsic molecular programs, including differences in signaling pathway activity, transcriptional regulation and copy number alterations. We further uncover subtype-specific patterns of ligand–receptor interactions linking malignant cells with immune and stromal compartments and provide independent spatial support for these interactions using spatial transcriptomic data. Together, this work provides a unified framework for tumor ecosystem stratification and highlights cellular organization as a key dimension of tumor heterogeneity with potential implications for precision cancer therapy.