| In this study,the homology of medicine and food plant pueraria and dandelion,Chinese herbal medicine circium japonicum and periploca forrestii schltr,edible plants artichokes,and celery seed were prepared for test raw materials.The composition,antioxidant activity and xanthine oxidase inhibitory activity of water and ethanol extracts from test raw materials were investigated.In addition,an efficient hyperuricemic cell model was established for the anti-hyperuricemic functional test.Also,molecular docking and molecular dynamics simulations on compounds from six samples with target proteins was performed,followed by the structure-docking energy relationship analysis.The content of flavonoids,polyphenols,polysaccharides and proteins of all extracts were determined.The results shown that the content of active components is closely related to raw materials,extraction technology and other parameters.And experiment of measuring DPPH·scavenging activity by microplate method was optimized.The absorption wavelength is 517nm,concentration of DPPH·is 0.2 mg/m L,the temperature is 37°C and reaction time is 20min.The water extract and ethanol extract of dandelion showed high DPPH·scavenging activity with the IC50 value 0.45 mg/m L and 0.46 mg/m L respectively.Also extract of pueraria had the best oxygen radical absorption capacity（ORAC）.The trolox equivalent of ethanol extract of pueraria was 55.06 mol/L,and the water extract of pueraria was 31.69mol/L.In addition,the results from the in vitro XO inhibitory activity shown that ethanol extract of pueraria with the IC50 value 6.14 mg/mL was significantly higher than that of others.There was no significant correlation between DPPH·scavenging activity and XO inhibitory activity,but the oxygen radical absorption capacity was significantly related to xanthine oxidase inhibitory activity,and the Pearson coefficient was 0.697.Hyperuricemic cell model was established for screening the anti-hyperuricemic substances via adenosine stimulus in human proximal tubul cell.When the concentration of water extract of celery seed was 4.0 mg/m L,the yield of uric acid was 65.70%of the model group.And when the concentration of ethanol extract of pueraria was 0.5 mg/m L and 2.0mg/m L,the yield of uric acid was 60.28%and 13.14%of the model group respectively.When the concentration of water extract of periploca forrestii schltr was 0.125 mg/m L,the yield of uric acid was 49.05%of the model group.Flexible docking simulations were carried out on 311 small molecular compounds derived from six samples as adenine nucleoside deaminase（ADA）,purine nucleoside phosphorylase（PNP）and xanthine oxidase（XO）inhibitors.And 112,46 and 74 small molecules have quite low docked enery with ADA,PNP,XO comparing with FRK,IM5 and TEI respectively.Among them,corylin has the highest docking enery with XO and PNP,and Apigenin-7-O-caffeoylglucoside has the highest docking enery with ADA.Among the ADA inhibitors,26.13%were from pueraria,22.52%from dandelion,18.02%from celery seeds and remarkably 68.48%were flavonoids and aglycones.Among the XO inhibitors,36.11%were from pueraria,16.67%were from dandelion,16.67%were from cirsium japonicum,and also 83.05%were flavonoids and aglycones.Among the 46 PNP inhibitors,25.58%were from celery seed,20.93%were from dandelion,20.93%were from pueraria,and 80%were flavonoids and aglycones.The molecular docking results showed that the small molecules in pueraria had better inhibitory effects on ADA,PNP and XO,which was consistent with the results of double-enzyme-coupling method and hyperuricemic cell model method.Also it was found that the main amino acids residues involved in hydrogen-bond formation of ADA are ASP19,ASP185,ASP295,HIS238,HIS17 and HIS157.The main amino acids residues involved in hydrogen-bond formation of XO are ARG880,THR1010,SER876,MOS3004,ALA1079 and ASP872.The main amino acids residues involved in hydrogen-bond formation of PNP are PO4291,TYR88,ASN243,GLU201,HIS257,ALA116,THR242,MET219.Relationships between molecular structure and binding energy with ADA,PNP,and XO were carried out to develop a structure-docking energy relationship（SDER）model based on multiple linear regressions（MLR）,principle component regression analysis（PCR）,and artificial neural network（ANN）.The results showed that ANN model with two hidden layers was the best model to predict the binding energy of XO with the coefficient of determination（R2）and mean squared error（MSE）were 0.8746 and 0.9414,respectively.In addition,MLR is the best way for the prediction of the binding energy for PNP（R2=0.8762,MSE=0.6205）and ADA（R2=0.8362,MSE=0.6620）... |
