目的 超高效液相色谱（UPLC）法结合模式识别方法分析贯叶连翘不同生长部位成分和抗炎活性的差异，建立同时定量测定贯叶连翘5个不同生长部位中6种黄酮类活性成分的方法。方法 采用ACQUITY UPLC^® BEH C₁₈（50 mm×2.1 mm，1.7 μm）色谱柱，0.2%甲酸-水-乙腈作为梯度洗脱系统。采集贯叶连翘花、果、叶、茎、根5个不同部位样品的色谱信息，对积分数据进行中心化、归一化等预处理，采用偏最小二乘判别分析（partial least squares discriminant analysis，PLS-DA）及PLS-tree聚类分析方法，以NO活性的半数抑制浓度（IC₅₀）作为抗炎活性的监督因子，评价贯叶连翘不同部位在黄酮成分、抗炎活性方面的相似性和差异性。结果 方法学验证结果显示，5种化合物线性关系良好，r>0.999，精密度、重复性和稳定性良好，平均加样回收率为97.28%~102.84%。根据不同部位的黄酮类成分含量和抗炎活性PLS分析情况，显示贯叶连翘不同部位的质量为叶 > 花 > 茎 > 果 > 根。结论 说明贯叶连翘采收部位应以地面上茎干部分开始采收为宜。

Objective To provide a chemometric analytical approach for different parts classification from Hypericum perforatum using ultra performance liquid chromatography (UPLC) combined with chemometrics methods. Methods The ACQUITY UPLC^® BEH C₁₈ (50 mm×2.1 mm, 1.7 μm) column was used, and 0.2% formic acid aqueous solution-acetonitrile as gradient elution system. The chromatograms information of different parts that including flower, fruit, leaf, stem and root from Hypericum perforatum L. was collected. The original data were pretreated by centralization and normalization, analyzed by partial least squares discriminant analysis (PLS-DA) and PLS-tree cluster analysis, and monitored by the half inhibition concentratiom of nitric oxide (NO) production activity as an anti-inflammatory factor, in order to evaluate the similarities and differences in flavonoids cotents and nitric oxide production inhibition in different parts from H. perforatum.Results All the calibration curves of 6 flavonoids showed good linearity in each range with correlation coefficients greater than 0.999 that had good precision, repeatability and stability, and the average recovery ranged from 97.28% to 102.84%. By PLS-DA and PLS-tree, according to the data of flavonoids contents and NO product inhibition activities, it showed the quality of leaf > flower > stem > fruit > root for H. perforatum. Conclusion The established method suggested that an appropriate harvest part of H. perforatum is the stem part on the ground.

国家自然科学基金资助项目（81501229）