Spectrum-effect relationships between high performance liquid chromatography fingerprint and analgesic property of Anisodus tanguticus   (Maxim) Pascher (Solanaceae) roots

Purpose: To investigate the spectrum-effect relationships between high performance liquid chromatography with photodiode array detection (HPLC-DAD) fingerprint and analgesic activity of Anisodus tanguticus (Maxim.) Pascher (Solanaceae) (AT) roots.
Methods: Analgesic activity of AT roots was evaluated by acetic acid-induced writhing test in mice. Fingerprint of AT roots was established by HPLC-DAD. After oral administration of AT roots extract, intra-gastric contents of caffeoylputrescine, anisodine, fabiatrin, scopolin, scopolamine, anisodamine and atropine in mice were determined by HPLC-DAD. Spectrum-effect relationships between HPLC-DAD fingerprint and analgesic activity were investigated using bivariate correlation analysis.
Results: Following treatment with different batches of AT roots extract, acetic acid-induced writhing responses in mice were inhibited significantly (p < 0.05 or 0.01), with inhibitions of 26.62 - 55.13 %, relative to the control group. Sixteen common peaks were obtained by fingerprint analysis. Peaks 1, 2, 6, 7, 8, 9 and 12 were identified as caffeoylputrescine, anisodine, fabiatrin, scopolin, scopolamine, anisodamine and atropine, respectively. Bivariate correlation analysis between analgesic activity of AT roots and 16 common peaks areas indicated the contributions of 16 common peaks to analgesic activity of AT roots. Surprisingly, bivariate correlation analysis between analgesic activity of AT roots and intragastric contents of above-named 7 constituents revealed that the contributions of the 7 constituents to analgesic activity of AT roots were different from those based on their peak areas.
Conclusion: This study provides scientific justification for the investigation of the active constituents of AT root with a view to its standardization.

Keywords
Anisodus tanguticus root; Analgesic activity; HPLC-DAD fingerprint; Bivariate correlation analysis