Human Stem Cells for Toxicity Screening(RMI)

点击翻译按钮获取中文摘要

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.