Human Stem Cells for Toxicity Screening

描述（由申请人提供）： 人类基因组测序的完成，阐明疾病分子机制的最新进展，以及组合化学、高通量筛选和“组学”技术的出现，为发现新的治疗药物提供了前所未有的机会。然而，许多候选药物由于在人体中的疗效差、不利的药效学特性、不可接受的不良反应和主要的不可预测的毒性而失败。目前的非临床安全性评估是昂贵的，耗时的，使用大量的代理，并涉及大量的动物，使得它不切实际的评价大量的新化学实体（NCE）的毒性在药物开发管道的早期。此外，典型的毒性研究无法可靠地预测人体潜在的安全问题，这是昂贵的临床试验损耗的主要原因。因此，需要新的经济方法来有效地确定吸收，分布，代谢，排泄和毒性（ADMET），这些方法可以更早地集成到药物开发管道中，以确定在临床试验中成功的可能性更大的NCE。这一高风险和高影响的建议将建立成人肝和肾干细胞作为受体介导毒性的体外高通量模型的效用。将研究雌激素、过氧化物酶体增殖物激活、芳烃和甾烷X受体介导的活性，以比较评估成人肝和肾干细胞对体内引发的啮齿动物效应的反应。使用正交人、小鼠和大鼠cDNA微阵列进行的剂量和时间依赖性基因表达谱分析将与传统（例如组织病理学、细胞增殖、临床化学）、分子（例如染色质免疫沉淀）和计算方法（例如反应元件搜索）结合使用，以阐明保守和不同的反应。所有数据将在dbZach（http：//www.example.com）中进行管理，这是一个符合MIAME-compliant的模块化毒理基因组学支持性关系数据库。dbzach.fst.msu.edu将通过使用先进的统计方法（例如遗传算法/偏最小二乘框架）计算整合这种不同的数据来重建调控途径和网络，并进行比较检查，以鉴定推定的基于机械的跨物种保守生物标志物，这些生物标志物可用于高通量筛选，并预测体内人类毒性。一旦得到验证，这些成体干细胞模型和基于机械的生物标志物可以用于优化，排名和优先考虑新的先导化合物。此外，该提案还将开发计算工具，以改善药物发现和开发期间的知识管理和决策。

DESCRIPTION (provided by applicant): Completion of the human genome sequence, recent advances in elucidating the molecular mechanisms of disease, and the emergence of combinatorial chemistry, high-throughput screening and "omic" technologies have provided unprecedented opportunities to discover new therapeutic agents. However, many candidate agents are failing due to poor efficacy in humans, unfavorable pharmacodynamic properties, unacceptable adverse effects, and major unpredicted toxicities. Current nonclinical safety assessments are costly, time consuming, utilize large amounts of agent and involve significant numbers of animals, making it impractical to evaluate the toxicity of large numbers of new chemical entities (NCEs) early in the drug development pipeline. Moreover, typical toxicity studies fail to reliably predict potential safety problems in humans, a leading cause of costly clinical trial attrition. Consequently, new economical approaches to efficiently determine the absorption, distribution, metabolism, excretion and toxicity (ADMET) are required that can be integrated earlier in the drug development pipeline in order to identify NCEs that have a greater likelihood of success in clinical trials. This high-risk and high-impact proposal will establish the utility of adult human liver and kidney stem cells as in vitro high-throughput models for receptor-mediated toxicity. Estrogen, peroxisome proliferator-activated, aryl hydrocarbon and pregnane X receptor mediated activities will be investigated to comparatively assess adult human liver and kidney stem cell responses to in vivo elicited rodent effects. Dose- and time-dependent gene expression profiling using orthologous human, mouse and rat cDNA microarrays will be used in conjunction with traditional (e.g. histopathology, cell proliferation, clinical chemistry), molecular (e.g. chromatin immunoprecipitation) and computational approaches (e.g. response element searches) to elucidate conserved and divergent responses. All data will be managed within dbZach (http://dbzach.fst.msu.edu) a MIAME-compliant modular toxicogenomic supportive relational database. Regulatory pathways and networks will be reconstructed by computationally integrating this disparate data using advanced statistical approaches (e.g. genetic algorithm/partial least squares framework) and comparatively examined in order to identify putative mechanistically-based cross-species conserved biomarkers that are amendable to high throughput screening, and predictive of in vivo human toxicity. Once validated, these adult stem cell models and mechanistically-based biomarkers could be employed in optimizing, ranking and prioritizing new lead compounds. Moreover, this proposal will also develop computational tools that will improve knowledge management and decision-making during drug discovery and development.