目的 优化奶制手参Gymnadenia conopsea的炮制工艺并采用HPLC法建立其指纹图谱及指标成分含量测定方法。方法 基于单因素实验结合Box-Behnken设计-响应面法（Box-Behnken design-response surface methodology，BBD-RSM），以料液比、炮制温度和炮制时间为考察因素，结合层次分析法（analytic hierarchy process，AHP）和基于指标相关性的权重分配法（criteria importance through intercriteria correlation，CRITIC）综合赋权，以综合评分作为工艺优化评价标准。采用HPLC法建立指纹图谱，应用中药色谱指纹图谱相似度评价系统（2012.130723版本）、IBM SPSS Statistics 27.0、SIMCA 14.1软件，结合层次聚类分析（hierarchical cluster analysis，HCA）、主成分分析（principal component analysis，PCA）和正交偏最小二乘法-判别分析（orthogonal partial least squares-discrimination analysis，OPLS-DA）对不同产地的奶制手参质量进行评价。结果 确定最佳炮制工艺参数为料液比1∶2、温度85 ℃、时间15 min，经3批验证试验综合评分为83.446（RSD为0.544%）。奶制手参中天麻素、对羟基苯甲醇、dactylorhin B、loroglossin、dactylorhin A和militarine的平均质量分数分别为0.143%、0.300%、0.600%、0.490%、0.870%和0.880%。指纹图谱确定了8个共有峰，13批样品相似度为0.897～0.996，HCA将13批奶制手参分为3类；OPLS-DA筛选出militarine和对羟基苯甲醇为关键差异性成分。结论 建立的炮制工艺稳定可靠，所构建的HPLC指纹图谱结合多成分含量测定的质量控制方法科学合理，可为奶制手参质量控制提供依据。

Objective To optimize the processing technology of milk-processed Shoushen (Gymnadenia conopsea, GC) and establish its fingerprint and determination method for index components by HPLC. Methods Based on single-factor experiments combined with Box-Behnken design-response surface methodology (BBD-RSM), the solid-liquid ratio, processing temperature, and processing time were investigated. The comprehensive weighting method combining analytic hierarchy process (AHP) and criteria importance through intercriteria correlation (CRITIC) was used, with the comprehensive score as the evaluation standard for process optimization. The HPLC method was applied to establish the fingerprint, and the traditional Chinese medicine chromatographic fingerprint similarity evaluation system (Version 2012.130723), IBM SPSS Statistics 27.0, and SIMCA 14.1 software were used. Quality evaluation of milk-processed GC from different producing areas was conducted by combining hierarchical cluster analysis (HCA) and orthogonal partial least squares-discrimination analysis (OPLS-DA). Results The optimal processing parameters were determined as solid-liquid ratio 1:2, temperature 85 ℃, and time 15 min. The comprehensive score of three batches of verification tests was 83.446 (RSD is 0.542%). The average contents of gastrodin, p-hydroxybenzyl alcohol, dactylorhin B, loroglossin, dactylorhin A, and militarine in milk-processed GC were 0.143%, 0.300%, 0.600%, 0.490%, 0.870%, and 0.880%, respectively. Eight common peaks were identified in the fingerprint, and the similarity of 13 batches of samples was 0.897—0.996. HCA classified the 13 batches of milk-processed GC into three categories; OPLS-DA screened out militarine and p-hydroxybenzyl alcohol as key differential components. Conclusion The established processing technology is stable and reliable, and the quality control method combining HPLC fingerprint and multi-component content determination is scientific and reasonable, which can provide a basis for the quality control of milk-processed GC.

R283.6

国家重点研发计划资助项目（2023YFC3504403）