作者
Fan Zhang,Yaqin Xu,Xueying Bu,Zitong Wang,Shilong Qi,Dalong Li,Libo Wang,Jingwen Bai,Yu Chul Yang
摘要
High pressure assisted aqueous two-phase system for simultaneous extraction of polysaccharides and polyphenols from Schisandra chinensis was modeled by response surface methodology (RSM) and artificial neural network (ANN). ANN displayed higher coefficient of determination (0.9725), and lower Pearson’s Chi-square (0.2950), absolute average deviation (1.0693) and root means square error (0.6479) than RSM model (0.9440, 0.6231, 1.6725, 0.9666), indicating ANN had better prediction capability than RSM. Under suitable aqueous two-phase system (ethanol/ammonium sulfate, 29.37%/19.83%) combining with optimal high pressure conditions (liquid-to-material ratio 87.10 mL/g, pressure 4.78 MPa, temperature 60.92 °C, and time 16.01 min) predicted by ANN, the yields of polyphenols, polysaccharides, and DPPH radical scavenging activities were separately 1.87 ± 0.06 mg/g, 16.76% ± 0.52%, 95.04% ± 0.97% and 84.84% ± 1.48%, which were higher than the response values optimized by RSM (1.80 ± 0.60 mg/g, 16.19%±0.37%, 91.28% ± 1.92%, 83.19% ± 0.97%). After purification by column chromatography, the polyphenols in top phase (SCTP) and polysaccharides in bottom phase (SCBP) were obtained. Rutin, hyperoside, quercetin and veronicastroside were the main phenolic compounds in SCTP, which displayed strong antibacterial effects against foodborne pathogens. SCBP was an acidic heteropolysaccharide with a molecular weight of 6.46 × 106 g/mol, and showed potent protective effects against erythrocyte hemolysis.