目的 建立牛蒡子-甘草药对的HPLC指纹图谱，研究探讨牛蒡子-甘草药对抗炎活性的谱效关系。方法 采用HPLC法建立牛蒡子-甘草药对的指纹图谱，结合聚类分析与主成分分析（principal component analysis，PCA）对收集到的18批牛蒡子-甘草药对进行质量评价；以脂多糖刺激RAW264.7细胞制备炎性细胞模型，以一氧化氮分泌量为指标考察18批牛蒡子-甘草药对的抗炎活性，采用灰色关联度分析（gray correlation analysis，GCA）和偏最小二乘法回归（partial least squares regression，PLSR）分析HPLC共有峰与炎性指标间的谱效相关性。结果 指纹图谱研究共标定38个共有峰，经混合对照品指认出13个成分，聚类分析与PCA结果一致，综合评价结果显示，样品S7质量最佳。GCA结果显示，牛蒡子-甘草药对HPLC指纹图谱各共有峰抗炎活性关联度为F5（甘草苷）＞F17（甘草酸）＞F1＞F3＞F4＞F9＞F11（异甘草苷）＞F13（牛蒡子苷）＞F19＞F23＞F21（牛蒡苷元）＞F14；通过PLSR分析发现，牛蒡子-甘草药对38个共有峰中17个特征峰（VIP＞1）与其抗炎作用显著相关，且F21（VIP＞1.5）贡献较大。结论 建立的牛蒡子-甘草药对HPLC指纹图谱可用于其质量评价，谱效相关性研究初步筛选出的17个成分可作为其抗炎作用有效成分群，为牛蒡子-甘草药对抗炎活性物质基础与作用机制的阐释提供研究基础和科学依据。

Objective To establish the HPLC fingerprints of Niubangzi (Arctii Fructus, AF)-Gancao (Glycyrrhizae Radix et Rhizoma, GRR) and to investigate the spectrum-effect relationship of the anti-inflammatory activity of AF-GRR. Methods To establish the fingerprint of AF-GRR and evaluate the quality of AF-GRR from different manufacturers by combining cluster analysis and principal component analysis (PCA); To prepare an inflammatory cell model using lipopolysaccharide-stimulated RAW264.7 cells, and to investigate the anti-inflammatory activity of AF-GRR from different manufacturers by using NO secretion as an index, and to apply gray correlation analysis (GCA) and partial least squares regression (PLSR) analysis. The spectroscopic correlations between the HPLC peaks and the inflammatory indexes were analyzed by GCA and PLSR analysis. Results A total of 38 shared peaks were identified in the fingerprint profile, and 13 components were recognized by the mixed control finger. The clustering analysis was consistent with the results of PCA, and the comprehensive evaluation results showed that sample S7 had the best quality. The results of GCA showed that the correlation degree of anti-inflammatory activity of each shared peak of AF-GRR fingerprint was F5 (liquiritin) > F17 (glycyrrhizic acid) > F1 > F3 > F4 > F9 > F11 (isoliquiritin) > F13 (arctigenin) > F19 > F23 > F21 (arctiin) > F14; By PLSR analysis, it was found that AF-GRR fingerprint had 38 shared peaks. Among the 38 peaks, 17 peaks (VIP > 1) were significantly correlated with their anti-inflammatory effects and F21 (VIP > 1.5) contributed more. Conclusion The HPLC fingerprints of AF-GRR established in this study can be used for their quality evaluation, and the 17 components initially screened in the spectroscopic correlation study can be regarded as the effective components of their anti-inflammatory effects. This study can provide a research basis and scientific basis for the elucidation of the material basis and mechanism of action of the anti-inflammatory activity of AF-GRR.

甘肃省中医药研究中心开放课题（zyzx-2023-18）；2022年甘肃省高等学校青年博士基金项目（2022QB-090）；甘肃省教育厅双一流重大科研项目（GSSYLXM-05）；中医药公共卫生服务补助专项子课题（2305191901）