molecules Article Characterization of α-Glucosidase Inhibitors from Clinacanthus nutans Lindau Leaves by Gas Chromatography-Mass Spectrometry-Based Metabolomics and Molecular Docking Simulation SuganyaMurugesu1,ZalikhaIbrahim1 ,Qamar-UddinAhmed1,Nik-IdrisNikYusoff1, Bisha-FathamahUzir1,VikneswariPerumal2,FaridahAbas3 ,KhozirahSaari3, HeshamEl-Seedi4,5andAlfiKhatib1,3,* 1 DepartmentofPharmaceuticalChemistry,KulliyyahofPharmacy,InternationalIslamicUniversityMalaysia, Kuantan25200,PahangDarulMakmur,Malaysia;[email protected](S.M.); [email protected](Z.I.);[email protected](Q.-U.A.);[email protected](N.-I.N.Y.); [email protected](B.-F.U.) 2 FacultyPharmacy&HealthSciences,UniversitiKualaLumpurRoyalCollegeofMedicinePerak, Ipoh30450,PerakDarulRidzuan,Malaysia;[email protected] 3 LaboratoryofNaturalProducts,InstituteofBioscience,UniversitiPutraMalaysia,Serdang43300, SelangorDarulEhsan,Malaysia;[email protected](F.A.);[email protected](K.S.) 4 DivisionofPharmacognosy,DepartmentofMedicinalChemistry,BiomedicalCentre,UppsalaUniversity, Box574,SE-75123Uppsala,Sweden;[email protected] 5 H.E.J.ResearchInstituteofChemistry,InternationalCentreforChemicalandBiologicalSciences, UniversityofKarachi,Karachi75270,Pakistan * Correspondence:alfi[email protected];Tel.:+60-9-570-4940 (cid:1)(cid:2)(cid:3)(cid:1)(cid:4)(cid:5)(cid:6)(cid:7)(cid:8)(cid:1) (cid:1)(cid:2)(cid:3)(cid:4)(cid:5)(cid:6)(cid:7) Received:24August2018;Accepted:18September2018;Published:19September2018 Abstract: Background: Clinacanthusnutans(C.nutans)isanAcanthaceaeherbalshrubtraditionally consumed to treat various diseases including diabetes in Malaysia. This study was designed to evaluate the α-glucosidase inhibitory activity of C. nutans leaves extracts, and to identify the metabolites responsible for the bioactivity. Methods: Crude extract obtained from the dried leaves using 80% methanolic solution was further partitioned using different polarity solvents. The resultant extracts were investigated for their α-glucosidase inhibitory potential followed by metabolites profiling using the gas chromatography tandem with mass spectrometry (GC-MS). Results: Multivariate data analysis was developed by correlating the bioactivity, and GC-MS data generated a suitable partial least square (PLS) model resulting in 11 bioactive compounds, namely, palmitic acid, phytol, hexadecanoic acid (methyl ester), 1-monopalmitin, stigmast-5-ene, pentadecanoicacid,heptadecanoicacid,1-linolenoylglycerol,glycerolmonostearate,alpha-tocospiro B, and stigmasterol. In-silico study via molecular docking was carried out using the crystal structureSaccharomycescerevisiaeisomaltase(PDBcode: 3A4A).Interactionsbetweentheinhibitors and the protein were predicted involving residues, namely LYS156, THR310, PRO312, LEU313, GLU411,andASN415withhydrogenbond,whilePHE314andARG315withhydrophobicbonding. Conclusion: Thestudyprovidesinformativedataonthepotentialα-glucosidaseinhibitorsidentified inC.nutansleaves,indicatingtheplant’stherapeuticeffecttomanagehyperglycemia. Keywords: α-glucosidaseinhibitors;Clinacanthusnutans;diabetes;GC-MS;partialleastsquare Molecules2018,23,2402;doi:10.3390/molecules23092402 www.mdpi.com/journal/molecules Molecules2018,23,2402 2of21 1. Introduction Diabetesmellitus(DM)isoneofthemostchronicmetabolicdisordersthatoccursduetoabnormal carbohydratemetabolismthatismainlylinkedwithinsensitivityofthetargetorganstoinsulin[1,2]. Commonly, it is characterized by hyperglycemia in which the level of blood sugar is significantly elevated,whichinturndisturbsthenormalbodymetabolismrelatedtocarbohydrate,fat,andprotein breakdown and eventually proves fatal if not treated or managed properly [3]. Cases of DM are increasingrapidlyworldwideandaffectingallpartsoftheworld. Highincreasingnumbersofcases have been reported in most developing countries, including Malaysia; caused mainly by inactive lifestyleandunhealthyfoodconsumption[4]. Around170millionpeopleworldwidehaddiabetesin theyearof2000anditisprojectedtoincreaseto366millionby2030[5]. Additionally,theprevalence of DM is also rapidly rising among the middle-class and low-income communities in developing countries[6]. Theuseofseveraltraditionalmedicinalplantsinthetreatmentofdiabeteshasbeenreported[7,8]. Likewise,Clinacanthusnutans(Burm. F)Lindau;(C.nutans)orSabahsnakegrass(locallyknownas ‘belalaigajah’),isaherbalshrubfoundabundantlyinsomeAsiancountriesincludingThailandand Malaysia.Thisherbhasbeentraditionallyusedtotreatvariousdiseasesincludinggout,hyperuricemia, inflammation, fever, skin rashes, and diabetes [9]. Commonly, the plant leaves are used for the treatmentofburns,allergicreactions,mucositis,skinrashes,andlesionsofvaricella-zostervirus(VZV) andherpessimplexvirus(HSV)[10,11]. InMalaysiaandThailand,ithasbeenpracticedasaremedy for scorpion and insect stings as well as venomous snake bites [12,13]. In recent times, the leaves ofC.nutansarewidelyconsumedbycancerpatientsthroughoutMalaysia[14]whileinIndonesia, theleavesareusedtotreatdiabetes,hyperuricemia,anddysentery[15,16]. C.nutansleavesarereportedtocontainseveralcompoundsthatpossesshypoglycemicproperties suchascaffeicacid,chlorogenicacid,quercetin,lupeol,betulin,stigmasterol,andβ-sitosterol[17–20]. Mostoftheexistingstudieswerecarriedoutseparatelytoevaluatethebiologicalactivitiesofsome individualcompounds. However,thebioactivityoftheindividualcompoundsmaybedifferentwhen thesecompoundsarewithinthecomplexmatrixoftheleaves,whichcouldbeduetothesynergistic effect with other components [21]. Thus, the reported antidiabetic compounds may not be solely responsible for the actual antidiabetic activity of the plant. Therefore, one way of detecting the compoundsresponsibleforagivenactivityisbyusingthemetabolomicsapproach. Itisaholistic approachthatincludesthedetectionofallmetabolitesineachsamplethatcanbecorrelatedtothe tested bioactivity using multivariate data analysis (MVDA) [22]. Metabolomics techniques can be achievedwiththeaidofchromatographyorspectroscopyinstrumentstoprofileallthemetabolites present in the analyzed sample. MVDA is a suitable statistical tool for managing large data sets obtainedusingspectroscopictoolsandisemployedinclassifyingsamplesbasedontheirchemical constituents[23]. Severalcommonanalyticalinstrumentshavebeenusedasananalyticalplatforminmetabolomics, includingnuclearmagneticresonancespectroscopy(NMR),gaschromatography-massspectrometry (GC-MS),andliquidchromatography-massspectrometry(LC-MS)[24]. Inthisstudy, GC-MSwas utilizedduetothenon-polarnatureofthemostactivehexaneextract. Moreover,GC-MSprovides highresolutionofseparation,highsensitivity,andthereproducibilityneededtoanalyzeacomplex biologicalmixture[25]. Hence,theaimsofthisstudyweretoevaluatetheα-glucosidaseinhibitory activityofdifferentextractsofC.nutansleaves,andtoidentifytherelatedbioactivecompoundsinthe extractusingGC-MSbasedmetabolomics. Thestudyhasalsodepictedthemolecularinteractionbetweentheenzymesandtheinhibitors (ligands) identified in the plant extracts. With molecular docking, the conformations and binding affinitiesofthepotentialphytoconstituentscouldbepredicted. Inaddition,thisin-silicotechniqueaids inthevisualizationoftheligand-proteincomplexformedwhileportrayingthecharacteristicactivity andbindingsitesofthecompounds[26]. Molecules2018,23,2402 3of21 2. Results 2.1. α-GlucosidaseInhibitoryActivity The α-glucosidase inhibitory activity of the C. nutans leaves extracts obtained from different solventextractionaredisplayedinTable1asthehalf-maximalconcentrationvalues(IC ,µg/mL). 50 AlowerIC valueisdesirableforhigherα-glucosidaseinhibitionactivity[27]. TheIC valuesranged 50 50 from3.07to133.57µg/mL.Thehighestα-glucosidaseinhibitoryactivitywasobservedforhexane fraction,withanIC valueof3.05µg/mL,whichwasfoundcomparableorevenbettertotheIC 50 50 value obtained for quercetin, the positive control. Meanwhile, the methanol extract exhibited the lowestactivitywiththehighestIC viz.,133.57µg/mL. 50 Table 1. The half maximal inhibitory concentration (IC ) of α-glucosidase inhibitory activity of 50 C.nutansleavesextracts. Samples IC (µg/mL) 50 Hexane(H) 3.05±0.05d Hexane:Ethylacetate(HE) 5.54±0.09d Ethylacetate(EA) 8.42±0.17c Ethylacetate:Methanol(EAM) 37.45±0.90b Methanol(M) 133.57±0.30a 5.77±1.01 Quercetin (19.09µM) Dataexpressedasmean±SD(n=6).SD=StandardDeviation.Meansthatdonotsharealetteraresignificantly differentwithpvalue<0.05. 2.2. GC-MSChromatogramofC.nutansLeavesExtracts Allidentifiedmetaboliteswereconfirmedbyscrutinizingthespectralpatternandcomparison with GC-MS NIST14 database library. The metabolites identified were primarily plant sterols, fatty acids, organic acids, and others, including palmitic acid, phytol, hexadecanoic acid, 1-monopalmitin,stigmast-5-ene,pentadecanoicacid,heptadecanoicacid,1-linolenoylglycerol,glycerol monostearate,alpha-tocospiroBandstigmasterol.Theremainingcompoundsincludesucrose,maltose, D-gluconic acid and D-glucose, which were abundant in the non-active extract (methanol extract). Alltheabove-mentionedmetabolitesarelabelledandshowninthechromatogramofeachoftheextract analyzed,asdisplayedinFigure1. SeveralcompoundsarefoundabundantlyinH,HEandEextracts especially palmitic acid, phytol, stigmast-5-ene and stigmasterol. The intensity of the metabolites reducedapparentlyasthepolarityoftheextractsolventincreased. Meanwhile,sucroseandD-glucose werefoundmoreprominentinthemorepolarextract(EMandM). Molecules2018,23,2402 4of21 Figure 1. GC-MS chromatogram of metabolites in H, HE, E, EA and M extracts from C. nutans. Eachlabelledpeakrefersto;1—palmiticacid,2—phytol,3—sucrose,4—hexadecanoicacid,5—maltose, 6—1-monopalmitin, 7—D-glucose, 8—stigmast-5-ene, 9—D-gluconic acid, 10—pentadecanoic acid, 11—heptadecanoicacid,12—1-Linolenoylglycerol,13—Glycerolmonostearate,14—Alpha-TocospiroB and15—Stigmasterol. Molecules2018,23,2402 5of21 2.3. MultivariateDataAnalysis ThebioactivecompoundscanbeidentifiedthroughGC-MSbasedmetabolomicsutilizingMVDA wherebyapartialleastsquare(PLS)modelwasdeveloped. Automatedfittingresultsintwoprincipal componentsnamelyPLScomponent1and2. Themasstochargeratio(m/z)ofacertainretention timewasextractedandtheIC (µg/mL)valuesofeachofthesampleswereconsideredasxandy 50 variables,respectively. PLSwasperformedtodiscriminatetheconstituentsamongtheextractsandto pinpointthem/zvaluescontributingtothebioactivity[1]. Briefly,themodeldiagnosticofthePLS modeldevelopedcanbeevaluatedusingseveralparameterssuchasgoodnessoffit,permutationtest, andcapabilityofthemodeltopredicttheyvalueusingactualtopredictedplot. Thegoodnessoffit isexplainedbythecumulativevaluesofR2Yindicatingthepercentageofvariationoftheresponse explainedbythemodel,andthecumulativevalueofQ2Yrepresentingthepercentageofthevariation of the response predicted by the model according to cross validation. The fitness of model and predictiveabilityareconsideredgoodifthecumulativevaluesofbothR2YandQ2Yaregreaterthan 0.5[28]. Inthisstudy,theR2YandQ2Yvalueswere0.95and0.87,respectively. Otherimportantparametersthatareconsideredinmeasuringtheaccuracyandtheperformance ofthemodelareroot-mean-squareerrorsofestimation(RMSEE)andtherootmeansquareerrorof cross-validation(RMSECV).RMSEEisameasureofaveragedeviationofthemodelfromthedata, whileRMSECVisameasureofthequalityofthemodelinpredictingnewsamples. Thelowerthe RMSECVvalue,thehigherthepredictiveaccuracyfornewsamples. Forthismodel,theRMSEEand theRMSECVvaluesare0.0286046and0.0441324,respectively. Figure2showsthescorescatterplotofthePLSmodelwithgoodseparationofthesamplesbased onthechemicalprofileandbioactivity(IC value)ofthesamples. Thehighactivesamples(H,HE, 50 and EA) with the IC value less than 10 µg/mL were observed at the positive PLS component 1, 50 whilethelessactivesamples(EAMandM)weredistributedatthenegativePLScomponent1. Figure2.Partialleastsquare(PLS)scorescatterplotofdifferentsolventextractsofC.nutansleaves. Figure 3 displays the permutation test plot, which depicts the validation of the PLS model developed. Theplotdisplaystheinterceptsforthefractionofthetotalsumofthesquaresintercept Y-valueandthepredictiveabilityofthemodelinterceptY-value. TheR2YandQ2Yvaluesinthis studyare0.407and −0.288, respectively, whichareacceptablesincetheR2 andQ2 isequalorless than0.4and−0.05,respectively[29]. Thesupplementaryplot,FigureS1displaystheobservedversus predictedplotthatmeasurestheaccuracyofthePLSmodelwiththeregressionequationthatmeasures itsvalidationwithR2valueof0.9497,whichisvalidonlyiftheR2valueis≥0.9. Molecules2018,23,2402 6of21 Figure3.ThepermutationplotwiththeR2Y=0.407andQ2Y=−0.288. Figure4showstheloadingcolumnplotofthedifferentextractsofC.nutansleaves. Eachpeak indicatestheidentifiedmetabolitefromtheC.nutansleavesextractsandthecompoundsdetailshave beentabulatedinTable2. Thepeakslabelledas3,5,7,and9representsucrose,maltose,D-gluconic acid,andD-glucose,respectively,thatarelocatedoppositetotheperIC50 valueassumedtoinduce α-glucosidaseactivity. Thesecompoundswerefoundabundantinthemethanolextract,withitsIC 50 valuemorethan100µg/mL,indicatingthatthecompoundswerelessactive.Meanwhile,therestofthe peaksthatareinthesamepositionastheperIC valueareeffectivetoinhibitα-glucosidaseactivity 50 whichareprimarilyplantsterols,fattyacids,organicacids,andothercompounds,namely;palmitic acid,phytol,hexadecanoicacid,1-monopalmitin,stigmast-5-ene,pentadecanoicacid,heptadecanoic acid,1-linolenoylglycerol,glycerolmonostearate,alpha-tocospiroB,andstigmasterol(Figure5). Figure 4. PLS loading column plot of different solvent extracts of C. nutans leaves. Assignments: 1—Palmitic Acid, 2—Phytol, 3—Sucrose, 4—Hexadecanoic acid, 5—Maltose, 6—1-Monopalmitin, 7—D-Glucose,8—Stigmast-5-ene,9—D-Gluconicacid,10—Pentadecanoicacid,11—Heptadecanoicacid, 12—1-Linolenoylglycerol,13—Glycerolmonostearate,14—alpha-TocospiroB,and15—Stigmasterol. Molecules2018,23,2402 7of21 Table2.CompoundsidentifiedintheC.nutansleavesextractsthroughGC-MSanalysis. Molecular No. RT(min) %ofArea Probability M+ TentativeMetabolites Formula 1 17.24 10.50 99 256.24 C H O Palmiticacid 16 32 2 2 26.64 1.42 95 296.31 C H O Phytol 20 40 3 45.47 36.18 93 342.12 C H O Sucrose 12 22 11 4 11.63 0.91 99 270.26 C H O Hexadecanoicacid 17 34 2 5 49.48 0.28 93 342.12 C H O Maltose 12 22 11 6 44.43 1.03 95 330.51 C H O 1-Monopalmitin 19 38 4 7 10.31 19.04 91 180.06 C6H12O6 D-Glucose 8 56.83 4.50 99 398.39 C H Stigmast-5-ene 29 50 9 14.30 3.42 93 196.16 C6H12O7 D-Gluconicacid 10 12.34 0.24 98 242.22 C H O Pentadecanoicacid 15 30 2 11 23.95 0.38 93 270.26 C H O Heptadecanoicacid 17 34 2 12 47.59 0.46 99 352.52 C H O 1-Linolenoylglycerol 17 36 13 47.99 0.40 91 358.31 C H O Glycerolmonostearate 21 42 4 14 49.24 0.07 96 462.37 C H O Alpha-TocospiroB 29 50 4 15 55.59 2.70 99 412.37 C H O Stigmasterol 29 48 RT=RetentionTimeC=Carbon,H=Hydrogen,O=Oxygen,N=Nitrogen. Figure5.Thebioactivecompoundsthatinhibitα-glucosidaseenzymeactivity. Molecules2018,23,2402 8of21 2.4. BioactivityConfirmationandQuantificationofSomeBioactiveCompounds A further investigation was required to confirm the bioactivity of afore-mentioned identified compounds through testing of the individual compounds on α-glucosidase inhibitory activity. However,notallthecompoundscouldbetestedandquantifiedduetoalimitationoftheavailability of some of these compounds. Therefore, three bioactive compounds were procured and tested. These include palmitic acid, heptadecanoic acid, and stigmasterol. The α-glucosidase inhibitory activityofthesecompoundswasanalyzed,andtheIC valuesweredisplayedinTable3. Inmany 50 cases,thepurifiedcompoundswereobservedtohavediminishedtheirbioactivitiesafterseparation or purification from the biologically active extract [30]. However, from the results obtained, thesecompoundsshowedfairlygoodα-glucosidaseinhibitionactivityasanindividualcompoundin theirpurifiedstate. Table 3. The half maximal inhibitory concentration (IC ) of α-glucosidase inhibitory activity of 50 quantifiedreferencecompounds. IC Value 50 Samples µg/mL µM Palmiticacid 24.09±0.44c 93.95 Heptadecanoicacid 26.10±0.77b 96.51 Stigmasterol 65.31±0.37a 158.25 Quercetin 5.77±1.01 19.09 Dataexpressedasmean±SD(n=6).SD=StandardDeviation.Meansthatdonotsharealetteraresignificantly differentwithpvalue<0.05. QuantificationofeachidentifiedcompoundthroughGC-MSwascarriedouttodeterminethe amountoftheidentifiedα-glucosidaseinhibitorsthatispresentintheactiveextract(H)oftheC.nutans leavesextract. Aninternalstandard,methylnonadecanoatewasusedtocalculatetherecoveryduring GC-MSanalysis. Theparametersmeasuredintermsofthelinearity,limitofquantification(LOQ), andlimitofdetection(LOD)aredisplayedinTable4. Besidesthat,thecalibrationcurvesestablished werefoundtobevalidwiththecoefficientoftheregression>0.99forallthecurvesdevelopedforeach ofthestandardreferences. Therecoveryofthesecompoundswascalculatedtobe93.07%,basedon thestandardcurvedevelopedfortheinternalstandard, MNDthatrepresentsthethreequantified compounds. Overall,therangeofLODvaluesobtainedwasfrom0.31to5.00µg/mg,andtherangeof LOQwasfrom0.63to10.00µg/mgforallthethreecompounds. Table4.Calibrationcurves,detectionlimitsandquantificationlimitsmeasuredbyGC-MS. Concentration LOD LOQ Analytes R21 (µg/mgSample)2 (µg/mgSample) (µg/mgSample) Palmiticacid 0.99 433.25±22.96a 2.50 5.00 Heptadecanoicacid 0.99 20.27±1.20c 5.00 10.00 Stigmasterol 0.99 386.48±21.03b 0.31 0.63 LOD=LimitofDetection;LOQ=LimitofQuantification.1R2isthecorrelationcoefficientoftheequationwith n=3. 2 Dataisexpressedasmean±SD(n=6),calculatedconsidering%recoveryofmethylnonadecanoate (≥98.0%).Meansthatdonotsharealetteraresignificantlydifferentwithpvalue<0.05. 2.5. MolecularDocking Molecular docking was carried out in order to understand the protein-ligand interaction at a molecular level. Generally, protein is comprised of amino acids linked in a sequence to form the complex. These amino acids have different properties depending on their side chains, which can behydrophobic(aliphaticoraromatic),polar(neutral)orevenelectricallycharged(acidicorbasic). Thishasbeenthekeytotheactualpropertiesandbehaviorsofaprotein[31]. Allthecompounds Molecules2018,23,2402 9of21 identifiedtobeactivelyresponsiblefortheα-glucosidaseinhibitionweredockedtotheSaccharomyces cerevisiaeisomaltase(SCI)crystalstructure(PDBID:3A4A).Theconformationsdisplayingthelowest bindingenergyforthecompoundswiththeinteractingresiduesaresummarizedinTables5and6. Table 5. Molecular interaction results of α-glucosidase enzyme protein with the known inhibitor (Quercetin)andtheactivecompoundsquantifiedusingGC-MS. BindingEnergy Other Compound H-bondInteractingResidues (kcal/mol) InteractingResidues ASH69,HIE112,GLN182, Controlligand −6.00 ARG213,ASH215,GLH277, TYR72 HIE351,ASP352,ARG442 LYS156,THR310,PRO312, Quercetin −8.15 PHE314,ARG315 LEU313,GLU411,ASN415 Palmiticacid −3.75 LYS156,LEU313,ARG315 HIE280,PHE314 Heptadecanoicacid −3.80 LYS156,SER236,GLN239,SER240 ARG315 TRP15,ILE262,ARG263, Stigmasterol −8.66 THR290 ILE272,VAL266,HID295 Table6.Molecularinteractionresultsofothertentativemetabolitesdockedontoα-glucosidaseenzyme. BindingEnergy H-bond Compound OtherInteractingResidues (kcal/mol) InteractingResidues 1-linolenoylglycerol −4.76 LYS156,SER241 TYR158,SER240,PHE314 1-monoplamitin −4.01 TYR158,SER240,ASP242 PHE314,ILE419 LYS156,TYR158,PHE159, Alpha-tocospiroB −7.99 GLN279,GLU411 PHE178,HIE280,PHE314 LYS156,TYR158,SER240,ASP242, Glycerolmonostearate −3.95 SER241 PHE314,ARG315,TYR316 Hexadecenoicacid −4.00 ALA292,ASN259,HID295 VAL266 (methylester) Pentadecanoicacid −4.42 HID295 VAL266 LYS156,TYR158,VAL232, Phytol −5.51 GLN239,SER240 LEU313,PHE314 TYR158,LYS156,HIE280,PRO312, Stigmast-5-ene −9.09 - PHE314,ARG315,TYR316 AccordingtoAutoDock1.5.6simulationresultshowninTable5, theα-glucosidase-quercetin (positivecontrol)inhibitorcomplexshowed−8.15kcal/molbindingenergycontainingfivehydrogen bondswiththeinteractingresiduesviz.,ASH69,HIE112,GLN182,ARG213,ASH215,GLH277,HIE351, ASP352,ARG442,andhydrophobicinteractionwithTYR72.Meanwhile,thebindingaffinityofpalmitic acid in complex with α-glucosidase was −3.75 kcal/mol. A total of three hydrogen bonds were observedinthecomplexinvolvingLYS156,LEU313,andARG315alongwithhydrophobicinteractions involving HIE280 and PHE314. On the other hand, α-glucosidase-heptadecanoic acid complex showedabindingaffinityvalueof−3.80kcal/mol. FourresiduesincludeLYS156,SER236,GLN239, andSER240wereobservedtointeractwithheptadecanoicacidviahydrogenbond,whileARG315 wasinteractingthroughπ-bond. Despite,interactionwithonlyoneresiduebyhydrogenbondwith itshydroxylgroup,THR290,stigmasterolhascomparablyshownabetterbindingenergyof−8.66 kcal/mol than the positive control quercetin. The hydrophobic contacts seemed to be dominant in α-glucosidase-stigmasterol complex due to its cyclic skeleton and alkyl groups that preferably bindstoTRP15,ILE262,ARG263,ILE272,VAL266,andHID295. Thecrystallizedligand,α-D-glucose (ADG)exhibitedslightlylowerbindingenergy;i.e.,−6.00kcal/mol,comparedtothepositivecontrol, quercetin,andstigmasterol. However,thehydrogenbondingbetweenADGandtheproteinindicates Molecules2018,23,2402 10of21 astableinteractionwithatotalofninebonds,involvingASH69,HIE112,GLN182,ARG213,ASH215, GLH277,HIE351,ASP352,andARG442withahydrophobicinteractionwithTYR72.Unlikethedocked ADG, the docking results showed that the binding of the tested compounds are majorly assisted throughhydrophobiccontactswithα-glucosidase. Theresultsarecomparabletothosereportedby Seongetal.,[32]whotestedtheisoflavoneswheredaidzein,genisteinandcalycosinshowedmore hydrophobiccontactswithsimilarresidues,suchasILE262,ARG263,ILE272,andVAL266. The2D diagram showing the interaction between the compounds and the protein residues are shown in Figure6whilethe3Dsuperimposeddiagramofthequercetin,ADG,andthreereferencecompounds (stigmasterol,palmiticandheptadecanoicacid)isdisplayedinFigure7. Figure6.The2Ddiagramshowingtheinteractionbetweentheproteinresiduesandtheinhibitors.(A) Positivecontrol(Quercetin),(B)Palmiticacid,(C)Heptadecanoicacidand(D)Stigmasterol. Figure7. Thesuperimposed3DdiagramshowingthebindingsiteoftheADG,Quercetin,Palmitic acid,HeptadecanoicacidandStigmasterol. The binding energy and residues involved in the interaction of other tentative metabolites identifiedaredisplayedinTable6. Asfortheotherα-glucosidaseinhibitorsidentifiedinthisstudy,
Description: