Correlation analysis of physicochemical properties with anti-inflammatory activity of Andrographis paniculata (Burm.f.) Nees based on HPLC-DAD, colorimeter and multivariate statistics: A comprehensive quality evaluation strategy

Abstract

Andrographis paniculata (Burm.f.) Nees is an essential traditional herbal medicine with various bioactivities in Asian countries. The quality of herbal medicines directly affects its clinical efficacy, so a comprehensive strategy, including color assay, chemical profiles, quantitative analysis, and anti-inflammatory activity assay, is constructed to evaluate the quality of A. paniculata in this paper. Here, 16 batches of commercial A. paniculata were collected, and there were noticeable differences in appearance, chemical compounds, and bioactivity among batches. Principal component analysis (PCA) indicated the color and diterpenoid lactones could be used to classify A. paniculat. And then, correlation analysis found the batches with greener color have higher contents of diterpenoid lactones and more efficient anti-inflammatory activity. Based on the results above, the partial least squares (PLS) regression models were finally established to predict the contents of diterpenoid lactones using the color assay data. All diterpenoid lactones models showed good performance, especially neoandrographolide. Notably, PLS regression models were first used for the nondestructive quantitative analysis of A. paniculata. This comprehensive quality evaluation strategy can theoretically elucidate the relationship between good appearance, high contents of diterpenoid lactones, and superior bioactivity of A. paniculata, which is meaningful for its quality control and evaluation. It also laid a theoretical basis for appearance and bioactivity evaluations of high-quality traditional herbal medicine.