Single-cell and spatial sequencing identifies senescent and germinal tumor cells in adamantinomatous craniopharyngiomas

Xianlong Wang,Jincheng Lin,Hongxing Liu,Chuan Zhao,Zhiwei Tu,Dapeng Xu,En Zhang,Zhongqing Zhou,Xueling Qi,Xingfu Wang,Zhixiong Lin
DOI: https://doi.org/10.1186/s13578-024-01299-1
2024-09-02
Abstract:Adamantinomatous craniopharyngioma (ACP) is a clinically aggressive tumor without effective treatment method. Previous studies proposed a paracrine tumorigenesis model, in which oncogenic β-catenin induces senescence in pituitary stem cells and the senescent cells lead the formation of paracrine tumors through secretion of pro-tumorigenic factors. However, there lacks characterization on senescent cells in ACPs. Here, we profiled 12 ACPs with single-cell RNA and TCR-sequencing to elucidate the cellular atlas in ACPs and 3 of them were also subject to spatial sequencing to localize different subpopulations of the tumor cells. In total, we obtained the transcriptome profiles of 70,682 cells. Tumor cells, which were unambiguously identified through the cellular mutation status of the driver CTNNB1 mutations, were clustered into 6 subsets. The whorl-like cluster (WC) cells show distinct molecular features from the other tumor cells and the palisading epithelium (PE) cells consists of a proliferating subset. Other than typical PE and WC, we identified two novel subpopulations of the tumor cells. In one subpopulation, the cells express a high level of cytokines, e.g., FDCSP and S100A8/A9, and are enriched with the senescence-associated secretory phenotype (SASP) factors. Hematoxylin and eosin staining reveals that these SASP cells lack an ordered structures and their nuclei are elongated. In the other subpopulation, the cell sizes are small and they are tightly packed together with an unusual high density expressing a high level of mitochondrial genes (median 10.9%). These cells are the origin of the tumor developmental trajectories revealed by RNA velocity and pseudo-time analysis. Single-cell RNA and TCR analysis reveals that some ACPs are infiltrated with clonally expanded cytotoxic T cells. We propose a hypothesis that WC and PE are formed via different negative regulation mechanisms of the overactivated WNT/β-catenin signaling which provides a new understanding on the tumorigenesis of ACPs. The study lays a foundation for future studies on targeting senescent cells in ACPs with senolytic compounds or other therapeutic agents.
What problem does this paper attempt to address?