目的 建立杜仲Eucommia ulmoides叶高效液相色谱(HPLC)指纹图谱,结合化学模式识别评价其质量。方法 采用Ultimate XB-C₁₈色谱柱(250 mm×4.6 mm,5 μm),以乙腈-0.1%磷酸溶液为流动相梯度洗脱,体积流量1.0 mL/min,柱温30℃,进样量20 μL,波长为207、240、277、350 nm,检测杜仲叶中12种活性成分含量并建立指纹图谱,应用SPSS 25.0和SIMCA 14.1软件进行聚类分析(clustering analysis,CA)、主成分分析(principal component analysis,PCA)和偏最小二乘法-判别分析(partial least squares discrimination analysis,PLS-DA),对不同产地杜仲叶进行质量评价。结果 不同产地杜仲叶桃叶珊瑚苷、京尼平苷酸、原儿茶素、绿原酸、儿茶素、车叶草苷、咖啡酸、京尼平苷、松脂醇二葡萄糖苷、芦丁、异槲皮素、紫云英苷差异显著,质量分数范围分别为0.053%~4.563%、0.078%~0.977%、0.018%~0.084%、2.634%~6.183%、0.010%~0.088%、0.124%~0.582%、0.001%~0.242%、0.033%~0.163%、0.065%~0.215%、0.045%~0.865%、0.199%~0.910%、0.020%~0.221%;20批杜仲叶HPLC指纹图谱确定了25个共有峰,经对照品指认11个色谱峰,相似度0.905~0.996;通过聚类分析将样品按不同产地分为2大类,最后筛选出绿原酸、桃叶珊瑚苷、车叶草苷、芦丁、松脂醇二葡萄糖苷是导致不同产地样品差异的标志性化合物。结论 建立的指纹图谱方法稳定、可靠、重复性好,结合化学模式识别可用于杜仲叶质量评价。

Objective To establish the HPLC fingerprint of Eucommia ulmoides leaves, and to evaluate its quality by chemical pattern recognition. Methods The chromatographic column was an Ultimate XB-C₁₈ column (250 mm × 4.6 mm, 5 μm) with gradient elution using acetonitrile-0.1% phosphate solution as the mobile phase at a flow rate of 0.1 mL/min and a column temperature of 30℃ with an injection volume of 20 μL. The sample was detected at 207 nm, 240 nm, 277 nm and 350 nm, respectively. The contents of 12 active ingredients in E. ulmoides leaves were detected and fingerprints were established. SPSS25.0 and SIMCA 14.1 software were applied to perform clustering analysis, principal component analysis and partial least squares-discriminant analysis to evaluate the quality of Eucommia leaves from different origins. ResultsSignificant differences were found between different origins of E. ulmoides leaves aucubin, geniposidic acid, protocatechin, chlorogenic acid, catechin, asperuloside, caffeic acid, geniposide, pinoresinol diglucoside, rutin, isoquercetin and astragalin, with contents ranging from 0.053% to 4.563%, 0.078% to 0.977%, 0.018% to 0.084%, 2.634% to 6.183%, 0.010% to 0.088%, 0.124% to 0.582%, 0.001% to 0.242%, 0.033% to 0.163%, 0.065% to 0.215%, 0.045% to 0.865%, 0.199% to 0.910%, and 0.020% to 0.221%; The HPLC fingerprinting of 20 batches of E. ulmoides leaves identified 25 common peaks and 11 peaks identified by the control, with similarities ranging from 0.905 to 0.996; the samples were divided into two major categories according to different origins by cluster analysis. Chlorogenic acid, aucubin, asperuloside, rutin and pinoresinol diglucoside being the signature compounds that caused the differences. Conclusion The established fingerprinting method is reliable and reproducible and shows a potential improvement in the quality evaluation of E. ulmoides leaves in combination with chemical pattern recognition.

R286.2

国家自然科学基金资助项目（32160388）；吉首大学校级科研项目（Jdy20050）