Application of Boolean Networks to discover stem and progenitor cells

描述(由申请人提供)：我们已经开发了几种新的方法来分析大规模平行的基因表达数据集。首先是StepMiner，这是一个识别时间进程微阵列数据集中逐步转变的工具。第二，BoolanNet，这是一种使用大量基因表达数据集发现基因之间的布尔暗示的方法。最近，我们发表了一种名为MiDReG(挖掘发育调节基因)的新方法，它使用布尔蕴涵成功地预测了发育途径中的基因。通过最初将这种方法应用于淋巴细胞分化，我们发现了以前未知的B细胞分化标志物，以及一个新的B细胞和T细胞发育的分支点。 拟议的项目将建立在我们使用MiDReG成功预测人类B细胞发育基因的基础上，开发一种发现癌症干细胞和祖细胞的通用方法。我计划在人类膀胱癌(移行细胞癌)中验证这种方法，首先是因为它是一个简单的癌症模型来测试，其次是因为我们的实验室拥有分离和测试细胞群启动肿瘤潜力的专业知识。这种方法将在发现膀胱癌干细胞和祖细胞方面得到优化，并有望成为其他类型癌症的类似研究的起点。超过90%的人类膀胱癌起源于一种称为尿路上皮的简单上皮组织，通常被称为移行细胞癌(TCC)。此外，人类膀胱癌是美国第五大常见癌症。本实验室建立了小鼠人膀胱癌异种移植模型，并于近期在膀胱癌中发现了一个肿瘤始发群体。最近的研究表明，癌症是异质性的，并形成了原始肿瘤细胞群体的层次结构。然而，详细的膀胱癌发展等级仍不清楚。我在这项提案中的主要近期目标是使用系统生物学方法来预测(诊断/预后)标记特定人群的基因，以了解膀胱癌细胞的这种层次结构。我将与Matt van de Rijn博士合作使用人类癌症组织微阵列，并与Robert Chin博士和Jens-Peter Volkmer博士合作使用异种移植来验证这种方法。从长远来看，在我的独立研究阶段，我将把这种方法推广到识别其他类型癌症的干细胞和祖细胞。

DESCRIPTION (provided by applicant): We have developed several novel approaches for analysis of massively parallel gene expression datasets. First, StepMiner, a tool that identifies step-wise transitions in the time course microarray datasets. Second, BooleanNet, a method of discovering Boolean implications between genes using these large numbers of gene expression datasets. Recently, we published a new method called MiDReG (Mining Developmentally Regulated Genes) that uses Boolean implications to successfully predict genes in developmental pathways. By initially applying this approach to lymphocyte differentiation, we discovered previously unrecognized markers for B cell differentiation, as well as a novel branchpoint of B cell and T cell development. The proposed project will build on our successful prediction of human B cell developmental genes using MiDReG to develop a general method for discovering cancer stem and progenitor cells. I am planning to validate this approach in human bladder cancer (Transitional Cell Carcinoma), first, because it is a simple model cancer to test, and, second, because our laboratory has the expertise to isolate and test cell populations for tumor-initiating potential. This method will be optimized for the discovery of stem and progenitor cells in bladder cancer and hopefully it will serve as a starting point for similar studies in other types of cancers. More than 90% of human bladder cancers arise from a simple epithelial tissue called urothelium, and are commonly called transitional cell carcinomas (TCC). Furthermore, human bladder cancer is fifth most common cancer in the United States. Our laboratory has established a working model for the xenotransplantation of human bladder cancer in mice and has recently discovered a tumor-initiating population in bladder cancer. Recent studies show that cancer is heterogeneous and forms a hierarchy of original tumor cell populations. However, a detailed bladder cancer developmental hierarchy remains unknown. My primary near-term goal in this proposal is to understand this hierarchy of bladder cancer cells using a systems biology approach to predict (diagnostic/prognostic) genes that mark specific populations. I will validate this approach using human cancer tissue microarrays in collaboration with Dr. Matt van de Rijn, and using xenotransplantation in collaboration with Drs. Robert Chin and Jens-Peter Volkmer. In the longer term, I will extend the method to identify stem and progenitor cells in other types of cancers during my independent investigator phase.