人的肿瘤组织是一个高度动态的异质性组织,由多个不同遗传背景的克隆组成,不同克隆之间相互竞争促进肿瘤的发展、进化。虽然现代生物医学的研究已经和正在揭示肿瘤进化的遗传学和表观遗传学分子机制，但是克隆间竞争直接的细胞学机制并不清楚。Entosis是最近报道的一种细胞生物学行为，这个过程中一个细胞能被相邻的细胞（宿主）卷入形成"cell-in-cell"结构，并最终在宿主中死亡，该结构在人各种肿瘤组织中存在，提示该机制可能介导肿瘤克隆间的竞争并促进肿瘤进化。与此相一致，我们发现将不同细胞混合时，一些肿瘤细胞总是被另外一些肿瘤细胞"吞入"并"杀死"，最终导致宿主肿瘤细胞获得生长优势。本课题将一方面利用体、内外模型进一步确认该机制在克隆选择、肿瘤进化中的作用；同时，以此为基础深入研究决定肿瘤细胞"宿主身份"的分子机制，鉴定通过此机制促进肿瘤克隆选择的致癌基因突变，以期为肿瘤临床诊断和治疗提供新思路。

Human tumor is a type of highly dynamic and heterogeneous tissue composed of genetically different clones, inter-clonal competition is believed to be able to promote tumor evolution. Although continuing efforts in biology and medicine have made great progress in deciphering the molecular mechanisms genetically and epigenetically, the cellular mechanisms underlying direct inter-clonal competition remain unclear. Entosis was a recently reported non-apoptotic cell death process, in which one cell gets internalized into and finally dead inside of its neighbouring cell. The structure of so called "cell-in-cell" formed by entosis resembled those documented in a wide range of human tumors for decades, where "cell cannibalism" was used to describe this phenomenon. We hypothesized entosis a mechanism mediating inter-clonal competition and tumor evolution. In support of this, we found that some tumor cells can outcompete others, when mixed together, by serving as "engulfer" host cells and gain an advantage of outgrowth over the loser cells. The project proposed here aimed to first set up a model of direct competition among intra-tumoral clones base on entosis in vitro as well as in vivo, and then explore the molecular determinants which dictate the identity of "winner/engulfer"of tumor cells, finally identify the oncogenic mutations promoting clonal selection within tumor through this mechanism. Thus, we hope to provide a new vision and clue on clinical diagnosis and therapy of human tumors by the study of this project.