目的 利用UPLC法指纹图谱结合化学计量学，探索不同产地甘草Glycyrrhiza uralensis的化学轮廓变化，并筛选能够区分产地的特征标志物，为具有统一属性的甘草质量评价提供科学依据和参考。方法 建立5个产地30批甘草药材的UPLC指纹图谱，并进行方法学验证。通过相似度评价初步分析质量一致性，进一步使用SIMCA 14.0软件进行主成分分析（principal component analysis，PCA）、聚类分析（hierarchical cluster analysis，HCA）和正交偏最小二乘法-判别分析（orthogonal partial least squares-discriminant analysis，OPLS-DA）等化学计量学方法，对不同产地甘草样品进行模式识别和差异区分。进一步结合变量权重值（variable importance projection，VIP），筛选关键差异标志物。最后，对筛选出的5种特征标志物成分进行含量测定。结果 成功建立了30批甘草样品的UPLC指纹图谱，各批次样品与对照图谱的相似度各不相同，但均大于0.849，表明所用样品质量稳定的同时又存在差异性。化学计量学分析显示，不同产地甘草样品存在明显聚类趋势，且PCA和HCA结果一致，证实了产地间的化学成分存在差异。OPLS-DA模型有效区分了不同产地，置换检验证实模型可靠。通过VIP＞1筛选出区分产地的5种关键特征标志物。含量测定结果表明，5个特征标志物成分在不同产地甘草中的含量范围差异显著，进一步验证了产地间化学成分的差异性。结论 建立的UPLC指纹图谱结合化学计量学（PCA、HCA、OPLS-DA）的分析方法，能有效区分不同产地甘草，并且甘草苷、甘草酸、芹糖甘草苷、甘草查耳酮B和三色堇黄苷可作为关键的产地特征标志物。该方法体系为客观评价不同产地甘草的质量差异、鉴别其产地来源及后续质量标准提升提供了可靠的分析策略和数据支持。

Objective To explore the chemical profile variations of Glycyrrhiza uralensis from different origins and screen origin-specific characteristic markers by using UPLC fingerprinting integrated with chemometrics. This approach provides a scientific basis and reference for the quality evaluation of G. uralensis with unified attributes. Methods This study established UPLC fingerprints for 30 batches of G. uralensis collected from five origins and validated the analytical method. The quality consistency of the samples was preliminarily analyzed through similarity evaluation. Subsequently, chemometric methods such as principal component analysis (PCA), hierarchical cluster analysis (HCA), and orthogonal partial least squares discriminant analysis (OPLS-DA) were used to identify the patterns and differences of G. uralensis samples from different origins by SIMCA 14.0 software. Further, the key discriminant markers were identified based on the variable importance projection (VIP) values. Finally, the contents of the five characteristic markers were determined. Results UPLC fingerprints of 30 batches of G. uralensis samples were established. The similarities between each sample batch and the reference fingerprint varied but all exceeded 0.849. This indicates that while the overall chemical profiles of the samples were generally consistent, quantitative differences existed. Chemometric analysis showed clear clustering trends for samples from different origins, and the consistent results of PCA and HCA confirmed the presence of chemical differences among the origins. The OPLS-DA model effectively distinguished different origins, and the permutation test confirmed the reliability of the model. Five key characteristic markers were selected based on VIP > 1. The content determination results indicated that the contents of the five characteristic markers in G. uralensis from different origins varied significantly, further confirming the differences in chemical components among the origins. Conclusion The UPLC fingerprint combined with chemometrics (PCA, HCA, OPLS-DA) analysis method established in this study can effectively distinguish G. uralensis from different origins, and glycyrrhizin, glycyrrhizic acid, liquiritin apioside, glycyrrhizin B, and violanthin can be used as key characteristic markers for origin discrimination. This analytical system provides a reliable strategy and data support for objectively evaluating the quality differences of G. uralensis from different origins, identifying their origin sources, and improving future quality standards.

R282.6

国家自然科学基金面上项目（82374117）