Optimization of extraction of flavonoids from shallot skin using response surface methodology based on multiple linear regression and artificial neural network and evaluation of its xanthine oxidase inhibitory activity

In this study, extracts of shallot skin, flesh and whole bulbs were evaluated their inhibitory effect on xanthine oxidase activity in vitro. The results showed that the effect of shallot skin extract was superior to the extracts from flesh and whole bulb. The extraction of flavonoid from shallot skin was comprehensively investigate using one-factor-at-a-time and response surface methodology based on multivariable linear regression and artificial neural network. Four established models accurately illustrated the quantitative relationship between temperature, time of extraction, solvent and total flavonoids content, yield of extraction. The optimal conditions were determined as: temperature of 85 °C, time of 120 min and ethanol concentration of 50%. The IC50 values on xanthine oxidase of the extracts in optimal and sub-optimal conditions were 3.23 and 21.09 µg/mL while the total flavonoids contents were 20.26% and 13.89%, respectively. The docking results suggested that isorhamnetin-3-glucoside and other flavonoids in shallot skin might play a key role in xanthine oxidase activity of shallot extract.