RESUMO
Objective: Totally 12 batches of Salvia miltiorrhiza stem and leaves were taken as the research object,in order to investigate the correlation between the chromatographic peaks and the blood circulation activity in the HPLC fingerprint,and establish and verify the activity spectrum-effect model based on enzyme activity. Method: HPLC method and fibrinogen plate method were used to determine the fingerprint and in vitro activity values of stems and leaves of S. miltiorrhiza in 6 batches. SPSS was used to establish its spectrum-effect model by the principal component analysis-correlation analysis-multiple linear regression analysis method,and six batches of S. miltiorrhiza stems and leaves were used to verify the model. Result: Totally 7 characteristic peaks were obtained in the fingerprints of S. miltiorrhiza leaves. Kendalls tau-b correlation analysis showed the correlations between the blood circulating activity of S. miltiorrhiza stems and leaves and the 7 characteristic peak areas. The steep slope map analysis showed five optimal principal components of the model, namely protocatechualdehyde,caffeic acid,common peak 5 (to be identified), rosmarinic acid and salvianolic acid B. Among them,protocatechualdehyde,common peak 5 and salvianolic acid B showed a positive correlation with the activity of blood circulation,while caffeic acid and rosmarinic acid showed a negative correlation with the activity of blood circulation,with low mean relative errors between the predicted value and the measured value. Conclusion: The blood circulation activity spectrum-effect model based on enzyme activity can predict the blood circulation activity of stems and leaves of S. miltiorrhiza.