[Quality evaluation of Aralia taibaiensis based on spectrum-activity relationship].

A spectrum-activity relationship is established with high performance liquid chromatography(HPLC) fingerprints and the in vitro antioxidant activity to improve the quality evaluation system of Aralia taibaiensis. The HPLC profiles of 12 batches of samples were collected, and the similarity evaluation, heat map analysis and principal component analysis were conducted for the chemometric study of the fingerprint data. Combined with grey correlation analysis, the contributions of the common peaks in the fingerprints to the antioxidant activity were clarified, and the important peaks reflecting the efficacy were identified. The results showed that 17 common peaks were found in 12 batches of A. taibaiensis samples, and 6 of them were identified as saponins. Similarity evaluation, heat map analysis and principal component analysis roughly classified the A. taibaiensis herbs into two categories, i.e.,(1) S1-S10, S12 and(2) S11. Twelve batches of samples showed different antioxidant activities in a dose-dependent manner. In particular, S9 had the strongest antioxidant activity, while S11 was the weakest in antioxidant capacity, which was basically consistent with the overall score results. The results of grey correlation analysis demonstrated that the 17 common peaks scavenged DPPH radicals in the following order: X_3>X_(17)>X_4>X_8>X_7>X_(13)>X_2>X_6>X_(11)>X_(10)>X_(16)>X_(12)>X_9>X_5>X_(14)>X_1>X_(15), and scavenged ABTS radicals in the order of X_4>X_3>X_7>X_8>X_2>X_(17)>X_(13)>X_6>X_(16)>X_(11)>X_5>X_(12)>X_(10)>X_9>X_(14)>X_1>X_(15). Among them, X_3, X_4, X_7(araloside C), X_8 and X_(17) were the important peaks reflecting the efficacy of A. taibaiensis, which were basically consistent with those contained in the principal component 1. In this study, the correlation between the HPLC fingerprints of 12 batches of A. taibaiensis and its antioxidant activity provides a reference for the Q-marker screening and quality control of A. taibaiensis.