目的 建立藏药铁棒锤Aconitum pendulum的HPLC指纹图谱及多成分定量分析方法，用于评价不同批次铁棒锤的质量。方法 建立15批铁棒锤的HPLC指纹图谱，结合化学模式识别方法分析共有峰，并对苯甲酰乌头原碱、乌头碱、3-脱氧乌头碱3种成分进行定量测定。结果 HPLC指纹图谱共标定10个共有峰，成功指认出3个特征色谱峰，分别为苯甲酰乌头原碱（峰4）、乌头碱（峰9）、3-脱氧乌头碱（峰10）;15批供试品指纹图谱与对照指纹图谱的相似度均在0.9以上，表明不同批次样品整体化学组成较为一致。层次聚类分析（HCA）与主成分分析（PCA）均将15批铁棒锤分为2大类；正交偏最小二乘法-判别分析（OPLS-DA）中，以变量投影重要性（VIP）值>1为阈值，获得4个主要差异成分，分别为9号峰（乌头碱）、1号峰、7号峰及8号峰，其中乌头碱是区分两类药材质量的关键指标。15批铁棒锤样品中，苯甲酰乌头原碱、乌头碱、3-脱氧乌头碱的质量分数分别为0.024 2～0.057 7、0.329 8～1.257 0、0.218 0～0.377 1 mg·g^-1。结论 建立的HPLC指纹图谱及多成分定量分析方法操作简便、结果可靠，结合化学模式识别方法，可为藏药铁棒锤的质量评价、质量标准提升及临床合理应用提供科学的数据支撑。

Objective To establish methods for HPLC fingerprint, and multi-component quantitative analysis of Aconitum pendulum, a Tibetan medicine, for quality evaluation of different batches of the herb. Methods HPLC fingerprints of 15 batches A. pendulum were established, and common peaks were analyzed combined with chemical pattern recognition. Quantitative determination was carried out for three components: benzoylaconine, aconitine, and 3-deoxyaconitine. Results A total of 10 common peaks were calibrated in the HPLC fingerprint, among which three characteristic peaks were identified: benzoylaconine(peak 4), aconitine(peak 9), and 3-deoxyaconitine(peak 10). The similarity degree between the fingerprint patterns of the 15 test samples and the control fingerprint pattern was all above 0.9, indicating that the overall chemical composition of the different batches of samples was relatively consistent. Both hierarchical cluster analysis(HCA) and principal component analysis(PCA) divided the 15 batches into two categories. In orthogonal partial least squares-discriminant analysis(OPLS-DA), with variable importance in projection(VIP) > 1 as the threshold, four main differential components were obtained: peak 9(aconitine), peak 1, peak 7, and peak 8. Aconitine was the key index distinguishing the two categories. The mass fractions of benzoylaconine, aconitine, and 3-deoxyaconitine in 15 batches were 0.024 2—0.057 7, 0.329 8—1.257 0, 0.218 0—0.377 1 mg·g?1, respectively. Conclusion The established HPLC fingerprint, and multi-component quantitative methods are simple and reliable. Combined with chemical pattern recognition, they can provide scientific data support for quality evaluation, improvement of quality standards, and rational clinical application of A. pendulum.

R286.2

西藏自治区科技计划项目(XZ202401ZY0015)