目的 采用HPLC法对珠子参Panax japonicus中多指标成分进行定量检测，建立主成分分析（principal component analysis，PCA）、正交偏最小二乘判别分析（orthogonal partial least-squares discrimination analysis，OPLS-DA）及加权逼近理想解排序-灰色关联度（technique for order preference by similarity to an ideal solution-grey relation analysis，TOPSIS-GRA）融合模型对不同产地珠子参进行质量评价，以提高珠子参药材整体质量控制水平。方法 采用外标法测定珠子参中三七皂苷R₁、人参皂苷Rg₁、人参皂苷Re、人参皂苷Rb₂、人参皂苷Rd₂、竹节参皂苷V、楤木皂苷A、竹节参皂苷IVa、越南参皂苷R₄、齐墩果酸、镰叶芹醇、豆甾醇和β-谷甾醇含量，定量检测条件为Agilent Eclipse Plus C₁₈色谱柱，流动相乙腈-0.2%磷酸梯度洗脱，检测波长205 nm，柱温30 ℃。以珠子参中13种成分含量为变量，通过PCA降维获得主成分并实现对样品分组，在样品分组基础上采用OPLS-DA挖掘影响珠子参质量的差异性标志物。建立加权TOPSIS-GRA融合模型对不同产地珠子参综合质量进行评价。结果 珠子参中13种成分在各自范围内线性关系良好（r＞0.999），平均加样回收率为96.76%～100.06%（RSD＜2.0%），精密度、稳定性和重复性试验结果符合《中国药典》2020年版规定。PCA结果显示前2个主成分特征值分别为9.757和1.976，累积方差贡献率为90.255%，15批珠子参聚为3组，呈现一定产地差异；OPLS-DA结果显示竹节参皂苷IVa、竹节参皂苷V、人参皂苷Rd₂、β-谷甾醇、楤木皂苷A和人参皂苷Re是影响珠子参质量的差异性标志物。加权TOPSIS-GRA融合模型结果显示15批珠子参样品的平均相对贴近度（γ_i）为0.261 3～0.743 2，珠子参产品质量存在一定产地差异，同一产地差异较小，不同产地产品质量差异较大，同时聚类结果与PCA结果一致。结论 建立的HPLC多指标成分定量控制方法操作便捷、结果准确，可用于珠子参中13种成分的同步检测；PCA、OPLS-DA及加权TOPSIS-GRA融合模型合理可靠，可用于不同产地珠子参质量评价。

Objective To quantitatively detect the multi-index components of Zhuzishen (Panacis Majoris Rhizoma) by HPLC, and establish a fusion model of principal component analysis (PCA), orthogonal partial least-squares discrimination analysis (OPLS-DA) and weighted technique for order preference by similarity to an ideal solution-grey relation analysis (TOPSIS-GRA) to evaluate the quality of Panacis Majoris Rhizoma from different places, so as to improve the overall quality control level of Panacis Majoris Rhizoma. Method Quantitative detection conditions were determined by HPLC on Agilent Eclipse Plus C₁₈ column with gradient elution of mobile phase acetonitrile-0.2% phosphoric acid at 205 nm and column temperature at 30℃. The contents of notoginsenoside R₁, ginsenoside Rg₁, ginsenoside Re, ginsenoside Rb₂, ginsenoside Rd₂, chikusetsusaponin V, araloside A, chikusetsusaponin IVa, vina ginsenoside R₄, oleanolic acid, falcarinol, stigmasterol and β-sitosterol were determined by external standard method. Based on the content of 13 components in Panacis Majoris Rhizoma, the principal components were obtained by PCA and the samples were grouped. On the basis of the sample grouping, OPLS-DA was used to find the difference markers affecting the quality of Panacis Majoris Rhizoma. A weighted TOPSIS-GRA fusion model was established to evaluate the comprehensive quality of Panacis Majoris Rhizoma from different origin. Results The 13 components had good linear relationships within their respective ranges (r > 0.999), and the average recovery rate (n=9) were 96.76%-100.06% (RSD<2.0%), the precision, stability and repeatability test results were in accordance with the provisions of Chinese Pharmacopoeia. The results of PCA showed that the eigenvalues of the first two principal components were 9.757 and 1.976, respectively, and the cumulative variance contribution rate was 90.255%. The 15 batches of Panacis Majoris Rhizoma were clustered into three groups, showing certain differences in origin. The results of OPLS-DA showed that chikusetsusaponin IVa, chikusetsusaponin V, ginsenoside Rd₂, β-sitosterol, araloside A and ginsenoside Re were the different markers affecting the quality of Panacis Majoris Rhizoma. The results of the weighted TOPSIS-GRA fusion model showed that the average relative proximity (γ_i) of 15 batches of Panacis Majoris Rhizoma samples ranged from 0.261 3 to 0.743 2, and there were some differences in the quality of Panacis Majoris Rhizoma products from different regions, with small differences in the same region and large differences in the quality of products from different regions. Meanwhile, the clustering results were consistent with the PCA results. Conclusion The HPLC method can simultaneously determine the 13 components in Panacis Majoris Rhizoma, which is simple and practical, OPLS-DA with weighted TOPSIS-GRA can be used to evaluate the quality of Panacis Majoris Rhizoma.

R286.2

黑龙江省卫生健康委科研课题（2019-403