利用Toxtree平台预测中草药重要成分的毒理学关注阈值

Prediction of threshold of toxicological concern of important components in Chinese herbal medicine by toxtree

目的通过对中草药重要成分进行毒理学关注阈值(Threshold of toxicological concern,TTC)预测,以建立针对中草药成分毒性筛选和危险评定方法。方法利用本研究已建立的中草药重要成分的QSAR预测毒性数据库,Toxtree平台中的Cramer决策树,扩展的Cramer决策树和Krose TTC决策树3个软件预测中草药重要成分的TTC。结果本研究无法获得暴露数据,只能得到中草药重要成分的TTC值,因此简化了Krose TTC决策树。对984个中草药重要成分进行TTC预测,发现共有255个中草药成分具有遗传毒性结构预警,预测的TTC为每人0.15μg/day;共有729个中草药成分无遗传毒性结构预警,预测的TTC分为I、II、III三类,每类分别有141、9、579个,对应的TTC分别为每人1 800、540和90μg/day。对具有代表性的黄酮类化合物的无遗传毒性的III类各化学物进行列表分析。结论 TTC方法可对暴露量低或者缺乏毒性资料的中草药重要成分进行毒性预测,为危险评定提供有价值的初步信息,并可评价进一步试验的优先性。

Objective To establish toxicity screening and risk assessment methods for important components in Chinese herbal medicine. Methods TTCs of important components in Chinese herbal medicine were predicted by Cramer decision tree,Cramer decision tree,with extensions and Krose TTC decision tree in Toxtree,with an established toxicity database of important ingredients in Chinese herbal medicine predicted by QSAR. Results Unable to obtain exposure data,the Krose TTC decision tree is simplified with only TTCs. Among984 important components of Chinese herbal medicine,there are 255 ones with genotoxic structure warning,and the TTC level was per person0. 15 μg / day; there are 729 ones with non-genotoxic structure warning,and the TTC was divided into Class I（ 141）,Class II（ 9）,Class III（ 579） of which TTC levels were 1800,540,90 μg/persor/day. Flavonoids could be responsible for Chinese herbal medicine,so flavonoids which were non-genotoxic and in Class III were analyzed in detail. Conclusions TTC method could predict the toxicity of substances with insufficient toxicological data or lower exposure level. Application of the TTC method could not only be used to provide valuable preliminary information for the risk assessment,but could also be used for priority setting for further toxicity testing.