复发性体细胞变异对肿瘤基因表达的决定作用及其分子机制

Tumour cells acquire frequent somatic changes in the genome including somatic copy number aberrations, mutations and epigenetic changes such as DNA methylation. These somatic changes are major causes of genome instability and account for various abnormalities of the tumour cells by altering the expression levels of the affected genes, which are associated with the clinical phenotypes. Identification of the recurrent somatic changes in the tumour is the key to understand the biology of tumourigenesis, genetics and to discover new therapy. Previous studies are limited by the data availability to reveal the global image of the somatic features in tumours. The latest cancer genome atlas (TCGA) and ENCODE database with large matched multilevel somatic data provide ideal platforms for systematic study of the somatic changes in various tumour types. In the study proposed, we use various statistical methods and complex network model to understand the somatic changes from three aspects: 1) identification of the recurrent somatic aberrations in the tumours; 2) the genetic determinants of the somatic aberrations and 3) identification of the somatic determinants of the tumour gene-expression and the corresponding biological mechanism. The associations among the different types of somatic features and tumour gene expression are likely to indicate important regulatory mechanisms of the tumourigenesis and differentiation; moreover, the unique pathways of which the expression levels are associated with the somatic changes will help to explain the behaviours of the tumour cells and serve as potential targets for the future personalised therapy. On the other hand, the microenvironment of tumour cells can cause substantial randomness and complexity in the somatic features. In coping with this, our previous studies have established solid analytical pipelines to integrate multilevel high throughput data sets and to adjust for possible confounding factors. In summary, the proposed study will lead us to a better understanding of the biology underlying the somatic aberrations of the tumour cells and reveal the important functional clinical impacts.

体细胞水平上的分子异常是肿瘤基因组的重要特征。体细胞变异直接影响肿瘤细胞的基因表达并决定若干重要的临床表型。由于受到样本和技术的限制，我们对肿瘤中不同类型体细胞变异的遗传背景和对基因表达的影响尚缺乏系统性认识。大型肿瘤数据库如TCGA，ENCODE的建立，使我们可以通过多层次的匹配信息认识肿瘤中的各种体细胞变异。据此，我们提出通过统计学模型和复杂网络模型从三个层次上对体细胞分子异常进行研究：首先发现肿瘤细胞中复发性的体细胞突变，拷贝数变化，甲基化异常；其次，寻找体细胞变异的遗传学决定因素；最后结合表观遗传学数据认识体细胞变异的对肿瘤细胞基因表达的决定性作用及其调控机制。由于肿瘤细胞所处的组织微环境各有不同，其体细胞变异和基因表达具有高度复杂性。为此，我们建立了整合多层次组学数据库的分析方法。上述研究将帮助我们构建完整的肿瘤体细胞变异图谱并深入认识体细胞变异对基因表达和细胞功能的决定性作用。

本课题经过为期三年的工作，按计划建立了三个系统行的研究平台并形成了相关成果。首先，我们建立了泛肿瘤的拷贝数改变，等位基因不平衡的贝叶斯模型，并根据拷贝数改变和等位基因的分布特征建立基于关联网络的谱聚类分析。我们系统地评价了等位基因不平衡性和体细胞拷贝数改变，基因表达之间的关联，描述了等位基因在肿瘤群体中的偏好性及其生物学内涵。其次，我们进行了泛肿瘤eQTL分析，并整合表观遗传学信息提出了优选致病位点的方法学，并以此解释了卵巢癌风险位点的若干顺式作用基因。最后我们针对食道鳞癌中的突变过程，整合胚系遗传负荷，体细胞突变和拷贝数改变，解释了肿瘤体细胞进化过程的遗传决定因素，并据此报告了两个潜在的食道鳞癌的驱动基因：ZNF750和CDC27。综上，我们初步建立了肿瘤基因组改变的系统生物学模型，并得到了一系列有生物学意义的结果。