目的 评价市售吴茱萸Euodiae Fructus的质量并研究不同商品规格吴茱萸中化学成分与肝细胞毒性的相关性，为吴茱萸全面质量控制提供依据。方法 按吴茱萸药材商品规格收集小花吴茱萸7批，中花吴茱萸22批，大花吴茱萸10批，建立了UPLC法同时测定吴茱萸中柠檬苦素（limonin，LIM）、吴茱萸碱（evodiamine，EVD）、吴茱萸次碱（rutaecarpine，RUT）、1-甲基-2-壬基-4-（1H）-喹诺酮[1-methyl-2-nonyl-4（1H）-quinolone，MNQ]、1-甲基-2-十一烷基-4（1H）-喹诺酮[1-methyl-2-undecyl-4（1H）-quinolone，MUQ]和吴茱萸新碱（evocarpine，EVC）的含量；采用CCK-8法检测吴茱萸对L02肝细胞的生长抑制作用。运用SIMCA-P 14.1统计软件对吴茱萸的细胞存活率进行主成分分析（principal component analysis，PCA）。结果 有19批商品符合《中国药典》2020年版规定，不符合药典要求的20批样品中有10批为药材基原不符，有1批EVD和RUT的总量不符合药典规定，9批LIM含量不符合药典规定。7批小花吴茱萸中1批不符合药典要求，22批中花吴茱萸中10批不符合药典要求，10批大花吴茱萸中9批不符合药典要求。同时发现小花吴茱萸中LIM含量高于中花和大花吴茱萸，而生物碱含量中花和大花吴茱萸中较高。其中S33～S39样品按果实大小属于中花吴茱萸但不含LIM，将其称为中花样吴茱萸劣品。基于建立的UPLC方法同时测定吴茱萸中LIM和5种生物碱含量；PCA分析模型将小花吴茱萸、中花吴茱萸、大花吴茱萸和中花样吴茱萸劣品的毒性很好地区分为4个区域，对L02肝细胞的生长抑制作用为中花样吴茱萸劣品>大花吴茱萸>中花吴茱萸>小花吴茱萸，且LIM与5种生物碱的比值和吴茱萸对L02肝细胞的半数抑制率（IC₅₀）总体上呈正相关。结论 吴茱萸的毒性与LIM和5种生物碱比例有关，LIM与5种生物碱含量和的比值越大，吴茱萸样品毒性越小。提示大花吴茱萸不适合作为商品流通，同时吴茱萸质量控制应考虑药材中LIM和生物碱类成分含量的比例。

Objective To evaluate the quality of Wuzhuyu (Evodiae Fructus, EF) in the market, and study the correlation between components and hepatotoxicity of different commercial grades of EF, and provide the foundation for the quality control of EF. Methods Seven batches of small flower EF (SEF), twenty-two batches of medium flower EF (MEF) and ten batches of big flower EF (BEF) were collected according to the commercial specifications of EF. The contents of limonin (LIM), evodiamine (EVD) and rutaecarpine (RUT) in 39 batches of EF samples were determined according to 2020 Edition of Chinese Pharmacopoeia. A UPLC method was established for the simultaneous determination of LIM, EVD, RUT, 1-methyl-2-nonyl-4(1H)-quinolone (MNQ), 1-methyl-2-undecyl-4(1H)-quinolone (MUQ), and evocarpine (EVC) in EF. The inhibitory effects of EF samples on L02 cells were detected by CCK-8 method. The principal component analysis (PCA) was performed on the basis of cell viability of 39 batches of EF samples to evaluate the differences toxicity of different categories of EF samples using SIMCA-P 14.1 statistical software. Results The results showed that there were 19 batches of EF meeting the requirements of Pharmacopoeia. Of the 20 batches of samples that do not meet the requirements of Pharmacopoeia, 10 batches were inconsistent with the original medicinal materials, EVD and RUT contents of one batch, and LIM contents of nine batches did not meet the requirements. One of seven batches SEF, ten of twenty-two batches MEF, and nine of ten batches BEF did not meet the requirements of Pharmacopoeia. At the same time, it was found that the LIM contents in SEF were higher than that in MEF and BEF, while the alkaloid contents were higher in MEF and BEF. Samples S33~S39 belonged to MEF but do not contain LIM according to the fruit size, which was called inferior products of MEF(IMEF). The method based on UPLC could be applied to determine LIM and five alkaloids in EF. The results of PCA suggested that 39 batches of samples were divided into four groups:SEF, MEF, BEF, IMEF. The inhibition effects of EF samples on L02 cell were IMEF>BEF>MEF>SEF. The result also demonstrated that the ratio of contents of LIM to five alkaloids was positively correlated with the IC₅₀ of EF samples. Conclusion The toxicity of EF was related to the ratio of LIM to 5 alkaloids. The greater the ratio of LIM to five alkaloids, the less toxic the EF was. It was suggested that BEF was not suitable for circulation as a commodity, and the proportion of LIM and alkaloids in EF should be considered in quality control.

R286

国家自然科学基金资助项目（81773879）；辽宁省高等学校国（境）外培养项目（2018LNGXGJWPY-YB024）