Relationship between cardiotonic activity of Fuzi (Radix Aconiti Lateralis Preparata) and its fingerprint determined by liquid chromatography-mass spectrometry.

OBJECTIVE To investigate the relationship between the cardiotonic activity of Fuzi (Radix Aconiti Lateralis Preparata, RALP) and its fingerprint determined by liquid chromatography-mass spectrometry (LC-MS). METHODS First, the fingerprints of six processed products of RALP were established by high performance liquid chromatography quadrupole time-of-flight mass spectrometry (HPLC-Q-TOF-MS) followed by analysis of the principal component of the relative peak area of its common peaks. Next, the scores of the first five principal components were used as input for an artificial neural network (ANN). Additionally, the therapeutic effect of RALP was assessed by measuring the hemodynamic indexes of heart failure model rats. Subsequently, fluorescence semi-quantitative polymerase chain reaction and an enzyme-linked immunosorbent assay kit were used to determine the effects of RALP-processed products on the serum levels of noradrenaline (NA), angiotensin-Ⅰ (Ang-Ⅰ), and the expression of β-norepinephrine receptor mRNA (β-NRm) in the rat cardiac tissues. P < 0.05 was used as the output of the ANN. Finally, a network was constructed to display the relationship between the LC-MS fingerprints and the cardiotonic activity of the RALP-processed products. RESULTS Several types of RALPs can improve diastolic function, systolic function and heart rate. On the basis of the findings from the principal component analysis (PCA) of 16 common peaks of fingerprints of six RALP-processed products, it was revealed that the first five principal components may include 100% of the information of the original data. As observed from the multilayer perceptron neural network analysis, principal component 4 presented with the strongest effects on serum levels of NA and Ang-Ⅰ in rats, while principal component 1 exerted the greatest effect on β-NRm expression in cardiac tissue. CONCLUSION The key findings obtained from this study indicated that the network constructed by the PCA-ANN may predict pharmacodynamic effects of the main ingredients of Traditional Chinese Medicine (TCM). This method may serve as a new approach to identify the relationship between LC-MS fingerprints and the pharmacodynamic effects of TCM ingredients.