Quantitative HPLC-based Metabolomics Approach for the Discrimination of Processed Rhizomes of Atractylodes Macrocephala

Background: The current study presents a method for the simultaneous quantification of atractylenolide I, II, and III, together with syringin, syringaresinol-4-O-beta-D-glucoside and caffeine in Atractylode macrocephala (AM) rhizomes. Contents of the metabolites, in combination with the metabolomics approach, were used to discriminate AM rhizomes, which were processed by different methods. Methods: An HPLC Agilent 1100 system with a Thermo Hypersil BDS C18 column (L x I.D. 250 mm x 4.6 mm, 5.0 mu m particle size) was used for the quantification of the compounds in the AM samples. The detection wavelengths were set up at 220 nm and 280 nm, respectively. A gradient of acetonitrile and water was utilized as the mobile phase. From the quantification results, the process AM rhizomes were discriminated using multivariate statistical methods, such as Principle component analysis and Hierarchical clustering analysis. Results: The contents of atractylenolide I, II, and III, syringin, syringaresinol-4-O-beta-D-glucoside, and caffeine in the AM samples were simultaneously quantified. The linear range of each reference compound was selected from 5 to 100 mu g/mL, the linearity with R-2 values varied from 0.9990 to 0.9997, the limits of quantification (LOD) ranged from 0.1 to 0.9 mu g/mL, LOQ ranged from 0.2 to 2.6 mu g/mL, while the intra- and inter-day recovery distributed between 96.0% and 104.8% indicated the precision and accuracy of the quantification method. These satisfied the criteria FDA standards for bioanalytical method validation. Multivariate statistical results revealed that atractylenolide I was the marker of the alcohol presoaking samples, syringaresinol-4-O-beta-D-glucoside, and atractylenolide III were representative compounds for the terra stirring AM rhizomes. Conclusion: For the first time, six investigated bioactive compounds in Atractylodes macrocephala were simultaneously quantified using the HPLC-DAD method. About 30 samples in four types of processed rhizomes of A. macrocephala were discriminated using the quantification results in combination with multivariate statistical methods. These results revealed a promising method for discrimination and quality assurance of products from processed AM rhizomes.