2010-08-012024-05-13https://scholars.lib.ntu.edu.tw/handle/123456789/650306摘要：許多毒理學家對於發展一個全面、公正、快速且準確的毒物篩檢方法極感興趣。篩檢的結果將幫助化學物風險評估及風險管理。毒理基因體學方法提供有利的篩檢工具，其不僅高通量且全面性，更可了解外源物(xenobiotics)的分子毒理機制。在各個不同的毒理基因體學技術中，代謝體學提供最有效測量細胞功能及狀態的方法。所以，我們將利用這個技術進行毒物篩檢及分類。我們假設化學物質造成的代謝干擾可被代謝體學方法偵測，且引起類似生物效應的化合物將會引起類似的代謝體改變。所以，代謝體學可以依據分子作用機制不同來分類化學物質，並且預測其他化學物質的可能毒性反應。此研究將以農藥為例，說明利用代謝影響的不同來篩檢及分類化學物質。大鼠腹腔注射不同劑量的農藥(例如：dicofol - 有機氯殺蟲劑，ethion - 有機磷殺蟲劑，及bifenox - 聯苯醚除草劑)並取得其組織。萃取並分析肝及睪丸組織樣本中的水溶性及疏水性代謝物。我們將開發極致效能液相層析串聯式質譜儀(UPLC-MS2)的脂質分析方法，並應用在分析組織樣本。核磁共振技術(NMR)包括1H NMR 及2D J-resolvedNMR 也將用於組織中水溶性及疏水性代謝物的分析。接著，應用多變相統計分析例如主成份分析及偏最小平方-判別分析，來分析並分類農藥造成的代謝影響。這將產生一個代謝profiling 合併數學規則，並進一步利用此特殊模式預測其它農藥的分組狀況及毒性反應。在這次研究中，選定測試的農藥(例如：clofentezine - 有機氯殺蟲劑，phorate -有機磷殺蟲劑，及chlomethoxyfen - 聯苯醚除草劑)造成之大鼠代謝影響將被檢驗並套入已建立好的運算法，來確定分組預測模型的效力。這個計畫為概念證明(proof-of-concept)之研究，欲建立以代謝體學做為化學危害物質篩檢及分類的可行技術。未來，代謝體篩檢的技術也可應用於疾病診斷及藥物開發。<br> Abstract: A comprehensive, unbiased, rapid, and accurate chemical hazard screening method is ofinterests of many toxicologists. The screening results will assist chemical risk assessmentand hazard management. Toxiocogenomic approaches provide valuable screening toolswhich are not only high-through and comprehensive, but also able to understand insight toxicmechanisms of xenobiotics. Among different toxicogenomic techniques, metabolomicsprovides the most functional measure of cellular status and can help to describe an organism’stoxic responses. Therefore, we will apply this technique to chemical screening andclassification. We hypothesize that metabolic turbulences of chemicals can be recorded byusing metabolomic approach. Compounds elicit similar biological effects will elicit similarmetabolic effects. Therefore, metabolic profiling can be used to classify chemicals based onmodes of action. Further, metabolomics is capable to predict toxic actions of an unknown.In this study, we will use pesticides to demonstrate chemical screening and classificationbased on metabolic effects.Tissues from rats treated with series doses of various pesticides (ie. dicofol-- organochlorineinsecticide, ethion-- organophosphate insecticide, and bifenox-- diphenyl-ether herbicide) viaip will be obtained. Both aqueous and hydrophobic metabolites from the liver and testistissue samples will be extracted and analyzed. A validated lipid analysis method using ultraperformance liquid chromatography combined with tandem mass spectrometry (UPLC-MS2)will be developed and applied to analyze the tissue samples. Both aqueous and hydrophobicmetabolites from the samples will be examined by 1H nuclear magnetic resonance (NMR) and2D J-resolved NMR. Then, multivariate statistically analysis, such as principal componentsanalysis and partial least squares-discriminant analysis, will be applied to analyze and classifymetabolic effects of pesticides. A particular pattern of metabolic profiling combinedmathematical rules will be generated, further used to predict class memberships of otherpesticides. In this study, metabolic effects of selected testing pesticides includingclofentezine-- organochlorine insecticide, phorate-- organophosphate insecticide, andchlomethoxyfen-- diphenyl-ether herbicide will be examined and encoded in the establishedalgorithm to check the effectiveness of the class prediction model.This project intends to do a proof-of-concept study to establish metabolomics as a viabletechnique for chemical hazard screening and classification. The results from themetabolomic screening can also apply to disease diagnosis and drug development in thefuture.代謝體學毒理學農藥核磁共振多變量分析toxicogenomicsmetabolomicsnuclear magnetic resonancemultivariate analysespesticides