Metabonomics Analysis of Different Extracts from Chrysanthemum morifolium and Screening of Its Antioxidant Active Components

To explore differences in chemical components of Chrysanthemum morifolium extracted by four different methods, i.e. boiling water extract (BWE), reflux water extract (RWE), ultrasonic assisted water extract (UWE) and ultrasonic assisted ethanol extract (UEE) and to identify key compounds reflective of antioxidant activity of Chrysanthemum morifolium extracts, the high-throughput liquid chromatography-mass spectrometry metabolomics was applied. A total of 776 compounds were detected including alcohols and polyols, amino acids, carbohydrates, flavonoid glycosides, etc., contents of total phenols and flavonoids of Chrysanthemum extracts were determined by spectrophotometry. The antioxidant activity of different extracts was evaluated by ABTS⁺ and DPPH free radical scavenging rate. The key antioxidant compounds were identified using supervised machine learning. Results showed that chemical components of Chrysanthemum morifolium significantly differed between four extracts: Relative contents of 33 and 35 metabolites greatly discriminated BWE, RWE and UWE, or UWE and UEE, respectively (P<0.05 after Bonferroni correction). The contents of total phenols of BWE, RWE, UWE and UEE were 13.858%, 10.708%, 16.644%, 20.160%, respectively; the contents of total flavonoids were 26.401%, 15.984%, 27.299%, 40.769%, respectively. The EC₅₀ of ABTS free radical scavenging rate was respectively 0.048, 0.036, 0.055, 0.056; the EC₅₀ of DPPH free radical scavenging rate was respectively 0.048, 0.053, 0.059, 0.047. No correlation was found between contents of total phenols or total flavonoids with the antioxidant activity indicated by EC₅₀ of ABTS and DPPH free radical scavenging rate. Of note, glycitein, serine, 1, 3-dicaffeoylquinic acid as well as ergothioneine, apigenin, phosphatidylcholines were identified, which could optimally characterize the antioxidant activity of the water extracts of Chrysanthemum morifolium, respectively. Using metabolomics and machine learning, key compounds characterizing antioxidant activity of the Chrysanthemum morifolium water extracts were identified, providing novel insights and theoretical basis for optimizing extraction processes targeting functional components of Chrysanthemum morifolium.