Quality evaluation of Salvia miltiorrhiza Bge. by ultra high performance liquid chromatography with photodiode array detection and chemical fingerprinting coupled with chemometric analysis.

Abstract

An ultra high performance liquid chromatography with photodiode array detection method is developed for the simultaneous quantitative determination of five water-soluble compounds including danshensu, protocatechualdehyde, rosmarinic acid, salvianolic acid B, and salvianolic acid A in Salvia miltiorrhiza Bge. Through method optimization, the five compounds all expressed good linearity (R(2) > 0.9990) in a wide concentration range together with satisfactory accuracy, precision, and stability. Moreover, through qualitative analysis of the chemical fingerprint combined with similarity analysis, hierarchical cluster analysis, principle component analysis, and partial least-squares discriminate analysis, we determined that the 13 batches of Salvia miltiorrhiza Bge. were similar in internal quality and the differences resulted from various cultivation environments, recovery elements, and others. Seen from the results of hierarchical cluster analysis and principle component analysis, the classification of 13 batches was in accordance, and partial least-squares discriminate analysis technique was more suitable than the principle component analysis model to provide a distinct classification of test samples on the basis of their different components. Moreover, a permutation test verified the rationality of partial least-squares discriminate analysis and variable importance plot showed that peaks 37 and 38 were the most significant variables in distinguishing the Salvia miltiorrhiza Bge. The idea of the quantitative and qualitative analysis of Salvia miltiorrhiza Bge. was convenient, sensitive, and comprehensive, which could be applied to evaluate the quality of more traditional Chinese medicines.