Identification of the lipid-lowering component of triterpenes from Alismatis rhizoma based on the MRM-based characteristic chemical profiles and support vector machine model
Identification of the lipid-lowering component of triterpenes from Alismatis rhizoma based on the MRM-based characteristic chemical profiles and support vector machine model
复制标题
基于MRM特征化学谱和支持向量机模型鉴定泽泻三萜类降脂成分
DOI：
10.1007/s00216-019-01818-x
复制
发表时间：
2019-06-01
影响因子：
4.3
通讯作者：
Jiang, Hongliang
中科院分区：
文献类型：
--
作者：
Li, Sen;Wang, Lu;Jiang, Hongliang
泽泻中的三萜类化合物对高脂饮食诱导的高脂血症具有降脂作用。泽泻醇B 23-乙酸酯是泽泻中含量丰富的三萜类化合物之一，《中国药典》将泽泻醇B 23-乙酸酯作为泽泻的质量控制指标，但由于泽泻中其他三萜类化合物也具有显著的药效，因此泽泻醇B 23-乙酸酯不能反映其降脂作用。为筛选ZX中具有显著生物活性的三萜类化合物，采用基于多反应监测（MRM）的特征化学谱（CCP）-支持向量机（SVM）模型研究ZX中三萜类化合物与降脂作用的关系。首先，采用超高效液相色谱-定量捕集-串联质谱（UHPLC-QTRAP-MS/MS）技术，建立了87种目标三萜类化合物的定量分析方法。第三，通过粒子群优化支持向量机模型，从87个具有高平均影响值的三萜中，识别出9个具有显著的降脂作用。由9种三萜类化合物构建的新SVM模型表现出良好的预测性能，总体预测准确率达到81.94%。最后，部分证实了这些三萜类化合物的真实的活性，并与支持向量机的预测结果相一致。这些结果表明，该方法用于发现ZX中具有显著降血脂活性的三萜化合物是可靠的。该方法有望为中药活性成分的筛选和药物发现提供一种高效、快速的方法。图形摘要。
It has been demonstrated that triterpenes in Alismatis rhizoma (Zexie in Chinese, ZX) contributed to the lipid-lowering effect on high-fat diet-induced hyperlipidemia. Alisol B 23-acetate, one of the abundant triterpenes in ZX, was used as the marker of quality control for ZX in Chinese Pharmacopoeia, while it could not reflect the lipid-lowering effect because other triterpenes in ZX also had prominent medicinal efficacy. To identify the significantly bioactive triterpenes in ZX, a multiple reaction monitoring (MRM)-based characteristic chemical profile (CCP)-support vector machine (SVM) model was used to explore the relationship between triterpenes and lipid-lowering effect of ZX. Firstly, the content of 87 targeted triterpenes was quantified by the MRM-based CCP using UHPLC-QTRAP-MS/MS. Secondly, the lipid-lowering effect of 30 ZX samples was assessed by 3T3-L1 preadipocytes. Thirdly, 9 of the 87 triterpenes possessing high mean impact value were identified to have significant lipid-lowering effect via the particle swarm-optimized SVM model. The new SVM model constructed by the 9 triterpenes showed good prediction performance and the overall prediction accuracy reached 81.94%. Finally, the real activity of these triterpenes was partly confirmed and was consistent with the prediction of SVM. These results showed that the method for discovery of triterpenes with prominent lipid-lowering activity in ZX was reliable. The proposed method is expected to provide an efficient and rapid approach for screening of active component and drug discovery in traditional herbs. Graphical abstract.