Virtual screening of secondary metabolites of the genus Solanum with potential antimicrobial activity

(1)

w ww.e l s e v i e r . c o m / l o c a t e / b j p

Original Article

Virtual screening of secondary metabolites of the genus Solanum with potential antimicrobial activity

Renata Priscila Costa Barros

^a

, Emidio Vasconcelos Leitão da Cunha

^b

, Raïssa Mayer Ramalho Catão

^b

, Luciana Scotti

^a

, Maria Sallett Rocha Souza

^a

, Amanda Amona Queiroz Brás

^a

, Marcus Tullius Scotti

^a,∗

aLaboratoriodeQuimioinformática,ProgramadePós-graduac¸ãoemProdutosNaturaiseSintéticosBioativos,UniversidadeFederaldaParaíba,JoãoPessoa,PB,Brazil

bDepartamentodeFarmácia,UniversidadeEstadualdaParaíba,CampinaGrande,PB,Brazil

a r t i c l e i n f o

Articlehistory:

Received20April2018 Accepted11August2018 Availableonline18September2018

Keywords:

Rutin

VirtualScreening RandomForest Antibacterialactivity

a b s t r a c t

Infectiousdiseasesareahealthproblemtodayandhavehighmortalityrateswithawidediversityof potentiallypathogenicmicroorganisms.Researchthatisbasedeitheronthesearchfornewdrugsfrom plantsorontheimprovementofphytotherapeuticsisprominentandcontinuestoplayanimportant rolenowadays.Fromthisperspective,useofinsilicostudiestocarryoutinvestigationsofnewmolecules potentiallyactiveformethicillin-resistantStaphylococcusaureusandEscherichiacoliusinganin-house databasewith421differentsecondarymetabolitesselectedfromtheliteraturefromSolanumgenus wasperformed.WealsorealizedaninvitrostudywithstrainsofS.aureusandE.coliandcompared theresults.TwodatabasesfromChEMBLwereselected,theﬁrstonewithactivityagainstmethicillin- resistantS.aureusandanotheragainstE.coli.ThecompoundswereclassiﬁedaccordingtothepIC50

valuestogenerateandvalidatethemodelusinga“RandomForest”.The“RandomForest”prediction modelformethicillin-resistantS.aureusobtainedanaccuracyof81%,areaundertheReceiverOperating Characteristiccurveof0.885,selectingeightmoleculeswithanactivepotentialabove60%.Theprediction modelforE.coliobtainedanaccuracyrateof88%,areaundertheReceiverOperatingCharacteristiccurve of0.932,selectingfourmoleculeswithpotentialprobabilityabove84%.Rutinprovedtobepotentially activeintheinsilicostudyforS.aureusandE.coli.Microbiologicaltestshaveshownthatrutinhasactivity onlyforE.coli.AninteractionstudywithstrainsofS.aureusATCC25923,astandardstrainsensitivetoall antibiotics,andSAM-01,amultidrug-resistantstrain,wasdesigned.Therewasinteractiononlybetween rutinandoxacillin,oneofthethreeantibioticsstudiedintheinteraction,forthestrainSAM-01,reducing theresistanceofthisstrain.

Introduction

Researchthatisbasedonthesearchfornewmedicinesfrom plantsorintheimprovementofalreadyexistingphytotherapyhas beenprominentandcontinuestoplayanimportantrolethesedays.

InaccordancewithresearchrealizedbyNewmanandCragg(2016) betweentheyears1981and2014,theareaofnaturalproductsstill producesorisinvolvedinroughly50.6%ofallnewdrugsapproved bytheFDAandsimilarorganizations.

SolanumisthebiggestgenusoftheSolanaceaefamily,with1500 speciesandroughly 5000epithetsdescribed (Meloet al.,2011;

VorontsovaandKnapp,2012).Thegenushasawidedistribution intheworld;Brazilisrepresentedby260species,ofwhich127are endemic(Agraetal.,2009).

∗ Correspondingauthor.

E-mail:mtscotti@ccae.ufpb.br(M.T.Scotti).

The Solanum genus presents great wealth and diversity of properties; between them, theyshowthe ability tobiosynthe- sizesteroidsand free orglycosylated alkaloids,ﬂavonoids, that are of interest therapeutically, which present a big variety of pharmacologicalactivities,likecytotoxicactivity,anticancer,anti- inﬂammatory,antiulcerogenicandantimicrobial(Pintoetal.,2011;

Ordazetal.,2011).

Infectious diseases are a world health problem today and have high mortality rates. There is a wide diversity of potentiallypathogenicmicroorganisms,includingStaphylococcusaureus, Escherichia coli, Pseudomonas aeruginosa, and their ability to becomeresistanttotheantibioticsavailableonthemarketreduces thechancesofinfectionpreventionandcontrol(Ventola,2015;Das etal.,2016).

Currently, there is a series of tools for studying the effects of diverse substances on an organism. The techniques of vir- tualscreeningrepresentamajoradvanceindrugplanningtoday, throughtheuseofinsilicomethods,largebanksofmoleculesare

https://doi.org/10.1016/j.bjp.2018.08.003

creativecommons.org/licenses/by-nc-nd/4.0/).

(2)

automaticallyanalyzed.Beyondidentiﬁcationofnewpotentially activemolecules,thevirtualscreeningalsoaims attheremoval of moleculesidentiﬁed astoxic or havingunfavorable pharma- cokinetic and pharmacodynamic properties (Song et al., 2009;

Prado-Pradoetal.,2007,2009;Speck-PlancheandCordeiro,2014;

Kleandrovaetal.,2016).

Inthisstudy,wasrealizedabibliographicsurveyofsecondary metabolitesisolatedfromtheSolanumgenus,creatingabankof molecules.Fromthemoleculebankcreated,twopredictionmod- elsofmoleculespotentiallyactiveagainstpathogenicbacterium methicillin-resistantS.aureus(MRSA)andE.coliweregenerated.

Inaddition,astudyinvitrowithstrainsofthesebacterialspecies wasrealizedandtheresultscompared.

Experimental Dataset

FromtheChEMBL database,we selectedtwo sets of chemi- calstructuresforconstructionoftwopredictivemodels.Theﬁrst sethad 1032diversechemicalstructuresthathad beenstudied (invitro)andhad inhibitedstrainsofS.aureus.Thecompounds wereclassiﬁedusingvaluesof−logIC₅₀(mol/l)=pIC₅₀,whichled ustoassign470actives(pIC₅₀≥5)and562inactives(pIC₅₀<5).

Inthiscase,IC50representedtheconcentrationrequiredfor50%

inhibitionofstrainsofS.aureus.

Thesecondsetofchemicalstructureswascomposedof1325 moleculesthathadinhibitedstrainsofE.coli.Thecompoundswere alsoclassiﬁedusingthevalueofpIC50,whichledustoassign777 actives(pIC₅₀≥4.6)and548inactives(pIC₅₀<4.6).Theclassiﬁca- tionofpIC₅₀wasdifferentfromthesetofS.aureusbecauseitwas attemptedtodividetheamountsofactiveandinactivemolecules closetohalf.

AnotherdatasetofmoleculesisolatedfromtheSolanumgenus wasbuiltfromabibliographicsurvey intheresearchbaseWeb ofScience,coveringatotalof550publishedpapersbetweenthe years2016and1991.Inthisbank,421chemicalstructuresofdif- ferentclassesofsecondarymetabolites(highlightingsteroidsand steroidalalkaloids)ofmany speciesoftheSolanumgenuswere cataloged(Scottietal.,2018).

Forallstructures,Smilescodeswereusedasinputdatain a Marvin 14.9.1.0,2014,ChemAxon (http://www.chemaxon.com).

WeusedStandardizersoftware[JChem14.9.1.0,2014;ChemAxon (http://www.chemaxon.com)]tocanonizestructures,addhydro- gens, perform aromatic form conversions, clean the molecular graphinthreedimensions,andsavecompoundsinsdfformat.

Three-dimensional (3-D) structureswere usedas inputdata inthesoftwareDragon7.0(Kode,2016)togeneratedescriptors, whereitispossibletopredictthebiologicalandphysicochemical propertiesofthemolecules.Thiscalculuswasrealizedforbothsets ofchemicalstructureswithknowledgeoftheactivityforMRSAand E.coli.

Predictionmodel

The Knime 3.4.0 software (Knime 3.4.0 the Konstanz Infor- mationMinerCopyright, 2003–2014,www.knime.org)wasused toperformallthefollowinganalyses.Thedescriptorsand class variables wereimportedfromthesoftware Dragon7.0,and for eachonethedataweredividedusingthe“Partitioning”nodewith the“stratiﬁedsample”optiontocreatea trainingsetand atest set,encompassing80%and20%ofthecompounds,respectively.

Althoughthecompoundswereselectedrandomly,thesamepro- portionofactiveandinactivesampleswasmaintainedinbothsets.

However,fortheE.colimodel,thedataweredividedinto70%for

thetrainingsetand30%forthetestset;duetothegreaternumberof moleculesintheE.colibankitwaspossibletoputmoremolecules intothetrainingset.

Forinternalvalidation,weemployedcross-validationusingten randomlyselected,stratiﬁedgroups,andthedistributionsaccord- ingtoactivityclassvariableswerefoundtobemaintainedinall validationgroupsandinthetrainingset.Descriptorswereselected, andamodelwasgeneratedusingthetrainingsetandtheRandom Forestalgorithm(RF)(Salzberg,1994),usingtheWEKAnodes(Hall etal.,2009).TheparametersselectedforRFincludedthefollowing settings:numberoftreestobuild=100,seedforrandomnumber generator=1,fortheS.aureusmodel,and50thenumberoftrees tobuildandtwoseedsfortheE.colimodel.

Theinternalandexternalperformancesoftheselectedmod- els were analyzed for sensitivity (true positive rate, i.e., active rate),speciﬁcity(truenegativerate,i.e.,inactiverate),andaccuracy (overallpredictability).Inaddition,thesensitivityandspeciﬁcity oftheReceiverOperatingCharacteristic(ROC)curvewerefoundto describetrueperformancewithmoreclaritythanaccuracy.

Biologicalactivity

Fortheantimicrobialactivityscreening,oneclinicalstrainof S.aureuswasused,SAM-01(MRSA),belongingtotheLaborato- riodeMicrobiologiadaUniversidadeEstadualdaParaíba,andtwo referencestrains, S.aureusATCC25923andE.coliATCC25922.

For theinoculum preparation,isolated coloniesofnewcultures (24h)wereselected,andwiththeaidofaninoculationloop,they weretransferred toa tubecontaining5mlof 0.85%NaCl. They werehomogenized,comparingtheirturbiditywitha0.5tubeof theMcFarlandscale(1.5×10⁸UFC/ml).

Antibiograms were carried out by disk diffusion in a solid mediumaccordingtoCLSI,2010,forthedeterminationofthesen- sitivityproﬁleofstrainSAM-01.

Fordeterminationoftheantimicrobialactivityandthemini- muminhibitoryconcentration(MIC),sterilemicroplateswereused containing96wellswithﬂatbottoms,where100␮lofBrainHeart Infusion(BHI)brothwaspouredintoeachwell.Rutinwasusedina concentrationof6674␮Mand100␮ltransferredtotheﬁrstwell.

Dilutionswerethenperformedtoobtainconcentrationsbetween 10and6.55␮M.BHIbrothwithinoculumwasusedasthepositive controlandjusttheBHIbrothwasusedasthenegativecontrol.In addition,thesolventcontrolusedforthedissolutionoftheprod- ucts,DMSO,wasinserted.

Theexperimentwasrealizedwithrutindilutedindistilledwater and 5% v/v aqueous solution of DMSO, which in the ﬁrst well became2.5%.10␮loftheinoculuminthatconcentrationwasalso dispensed1.5×10⁸CFU/ml.Theplateswereincubatedat35–37^◦C for24h,andtheexperimentswereperformedintriplicate.

Bacterialviabilitywasdetectedby adding20␮lof resazurin (0.01%)inaqueoussolution.Theplateswerereincubatedat37^◦C for2h,andin thosewellswhere bacterialgrowthoccurredthe resazurinchangedtopink.MICwasdeﬁnedasthelowest con- centrationofantibacterialagentsthatinhibitedvisiblegrowth,as indicatedbyresazurinstaining.

Theanalysisofrutininterferenceovertheeffectivenessofcon- ventionalantibiotics,oxacillin,penicillin,andamoxicillin+Acwas performed.Clavulanic,forthestrainsofS.aureus,wasperformed bydiskdiffusion.Fortheinteractiontest,50␮lofrutinataconcen- trationof6674␮M,i.e.,2048␮g/ml,wasaddedtoeachantibiotic usedaswellastosterilediskstoobservecomparativelywhether theadditionoftheproductcausedsomechangeinthesizeofthe inhibitionhalos.Thedisksofeachantibioticwerealsoinsertedin theplateforvisualizationofthesensitivityproﬁleofthestrainsand theoccurrenceofsynergismorantagonismwithrutinuse.Sterile

(3)

100 93 80 60 40 20

0 A B

B A

C

C D

D

E

F

F G

G

H

I

I J

J

K

L

54 49

40

21 19 17

11 10 9 7

91

Classes of secondary metabolites

Amount of mlecules

Steroidal glycoalkaloid Steroidal alkaloid Saponin Alkaloid Lignanamide Others Flavone

Amide Couarin Steroid

Fatty acid Alkaloid tropanic

Fig.1. ClassesandsubclassesofmetabolitesisolatedfromtheSolanumgenus.

diskswerefurtherﬁlledwiththediluentusedtodiluterutin,2.5%

DMSOandsteriledistilledwater.

Itwasregardedasaninteractiveeffectwhentherewasachange in diameter of the inhibition zones (halo) of microbial growth afterthisprocess,asynergisticinteractiveeffectiftheantibacterial showedanincreaseof≥2mmwhencomparedwiththeinhibition zonesformedbythetestedalone.Ifitshowedasmallerdiame- terthantheoneformedbyisolatedactivity,anantagonisticeffect wasconsidered.Thesetestswereperformedintriplicateandthe resultswereobtainedbythemeanoftheinhibitionzonesformed (Oliveira,2006).

Resultsanddiscussion

The secondary metabolites’ dataset was composed of 734 molecules,with421differentchemicalstructures,from110species oftheSolanumgenus.These421structureswereclassiﬁedinto54 classesofsecondarymetabolites,highlighting:steroidalglycoalka- loid(93),steroidalalkaloid(54),saponin(49),steroid(40),ﬂavones (21),amongothersrepresentedinFig.1.Thisdatasetisavailablein SistematX(Scottietal.,2018).

Inaccordwiththeseresults,itispossibletoobservethatthe mainclassesare thesteroidal glycoalkaloids,alkaloids,saponin, steroids, and ﬂavones, which are considered chemotaxonomic markersofthisgenus(Pereiraetal.,2016).

TheDragon7.0(Kode,2016)softwaregenerated1232descrip- torsfor 1033moleculesofknownactivityagainst MRSA.These descriptorswere usedas inputdata totheKnime software for generationofthepredictivemodel.

TheDragon 7.0 (Kode, 2016)calculation gave roughly 5270 moleculardescriptorscoveringmostofthetheoreticalapproaches.

These descriptors are organized into thirty logical blocks. The descriptors’listincludesthesimplesttypesofatoms,functional groupsandfragmentcounting,topologicalandgeometricdescrip- tors,three-dimensionaldescriptors,butalsoseveralestimatesof propertiessuchaslogPandLipinski(Kode,2016).

The molecular descriptors with the information of qualita- tivebiologicalactivitywereusedformodelgenerationwiththe

RFmachine learning. TheS. aureusmodel withmolecules with pIC50≥5wereconsideredasactive,withatotalof470molecules, and the molecules with pIC₅₀<5 were considered as inactive, totaling562molecules.FortheE.colimodel,themoleculeswith pIC50≥4.6wereconsideredasactive,withatotalof777molecules, and themoleculeswithpIC₅₀<4.6were consideredasinactive, totaling548molecules.

Analyzing the S. aureus model, you can see that the cross-validationandthetestdemonstratedsimilarstatisticalper- formance,withhitrateshigherthan74%,whiletheE.colimodel hadhitrateshigherthan83%.Thetraininghadanalmostperfect performance,onbothmodels,withhitratesof99%,ascanbeseen inTables1and2,whichsummarizethestatisticalratesoftheRF model.

Forthetrainingandtest,theS.aureusRFmodelhadsimilarrates fortheactiveandinactivemolecules(99%and81%,respectively), butforthecross-validation,therewasahigherhitratefortheinac- tivecompounds,78%,whilethehitrateforactivecompoundswas 74%.In addition,for thecross-validationandtest,theE.coliRF modelhadsimilarratesfortheactivemoleculesgreaterthan88%, butfortheinactive,theywereslightlysmaller,74%cross-validation and80%inthetest.

Withtheseresults,itwaspossibletocalculatetheMatthewscor- relationcoefﬁcient(MCC)forgeneralevaluationofthetwomodels.

TheMCCcorrelatestheobservedandpredictedbinaryclassiﬁca- tions,resultinginavaluebetween−1and+1,where+1isaperfect predictionand−1indicatesacompletedisagreementbetweenpre- dictionandobservation.Forthetest,thevalueobtainedfortheS.

aureusmodelwas0.68and0.62inthecross-validation,andforthe E.colimodelwas0.71inthetestsetand0.70inthecross-validation, revealingthegoodpredictionofthetwomodels.

TheROCgraphthatanalyzesthemodelperformancewasgener- atedforthetestsetforthetwomodelsandtheareaunderthecurve obtainedfortheS.aureusmodelwas0.885,Fig.2.Theareaunder thecurveobtainedfortheE.colimodelwas0.9329,Fig.3.Knowing thataperfectmodelhasanareaunderthecurveequalto1,itis possibletostatethatthemodelsabovearecapableofperforming ahighclassiﬁcationrateforthisRFmethod.

Thetwo modelswereusedtotriage thesecondary metabolites’bankoftheSolanumgenusforinvestigatingpossiblebioactive moleculesagainst MRSAand E.coli. Moleculeswithprobability greaterthan50%,pIC50≥5fortheS.aureusmodelandpIC50≥4.5 fortheE.colimodel,wereconsideredactive,totaling30molecules selectedbytheS.aureusmodeland221moleculesselectedbythe E.colimodel.

OntheS.aureusmodel,moleculeswithprobabilitygreaterthan 60%wereselectedforthisstudy,toincreasetherestriction,total- ingeightpotentiallyactivemoleculesagainstMRSA.Table3shows themoleculesselectedandtheirrespectiveclassesofsecondary metabolitesandspeciesfromwhichtheywereisolated.

Of the 221 active molecules on the E. coli model, 26 are likely to be active between 80 and 88%, 77 with potential activity between 70 and 79% probability, 64 molecule with a probability of activity between 60 and 69%, and ﬁnally, 54 molecules between 50 and 59% of active potential. The moleculeswiththehighestactivepotential,of84–88%probability,

Table1

Summaryofcross-validationandtestforthechemicalcompoundsoftheactivityknownasmethicillin-resistantStaphylococcusaureus(MRSA)usingtheRandomForest model.

Crossvalidation Test

Sample Predict %Hit Sample Predict %Hitrate

Active 376 280 74 94 77 81

Inactive 410 322 78 102 82 80

General 786 602 76 196 159 81

(4)

Table2

Summaryofcross-validationandtestforchemicalcompoundsofknownactivityformultidrug-resistantEscherichiacoliusingtheRandomForestmodel.

Crossvalidation Test

Sample Predict %Hit Sample Predict %Hitrate

Active 544 486 89 233 206 88

Inactive 383 285 74 165 127 80

General 927 771 83 398 398 83

1.000 0.899 0.899 0.899 0.600 0.500 0.400 0.300 0.200

0.100 0

0 0.100

0.200 0.300

0.400 0.500

0.600 0.700

0.799 0.899 1.000 Area under the curve = 0.8852

1-Specificity

Sensibility

Fig.2. ROCplotwiththeareaunderacurvefortheMRSAmodeltestsetobtainedwithRandomForest.

1.000 0.950 0.950 0.950 0.800 0.750 0.700 0.650 0.600 0.549 0.499 0.449 0.399 0.350 0.300 0.250 0.200 0.150 0.100 0.050 0

0 0.050 0.1000.1500.200 0.2500.300 0.3500.399 0.4490.4990.549 0.600 0.6500.700 0.7500.8000.8500.900 0.9501.000

1-Specificity

Sensibility

Area under the curve = 0.9329

Fig.3.ROCplotwiththeareaunderthecurveforthetestsetoftheE.colimodelobtainedwithRandomForest.

(5)

Table3

Moleculesselectedwithhigheractivepotential,ontheStaphylococcusaureusmodel, andtheirrespectiveclassesofsecondarymetabolitesandspeciesfromwhichthey wereisolated.

Chemicalstructures Class Species

MelongenamideD Lignanamide Solanummelongena

Grossamide Lignanamide Solanummelongena

N-cis-Grossamide Lignanamide Solanumtuberosum N-trans-Grossamide Lignanamide Solanumtuberosum MelongenamideB Lignanamide Solanummelongena

Tiliroside Flavone Solanumcrinitum

MelongenamideA Lignanamide Solanummelongena

CannamisinA Lignanamide Solanummelongena

Rutin Flavone Solanumlycopersicum

Table4

Moleculesselectedwithhigheractivepotential,ontheEscherichiacolimodel,and theirrespectiveclassesofsecondarymetabolitesandspeciesfromwhichtheywere isolated.

Chemicalstructures Class Species

AbutilosideJ Glycosteroid Solanumabutiloside AbutilosideA Glycosteroid Solanumabutiloside SolasodosideE Glycosteroid SolanumsodomaeumL.

AbutilosideR Glycosteroid Solanumabutiloside

Rutin Flavone Solanumlycopersicum

are available in Table 4, as well as the class of secondary metabolitesto which they belongand thespecies from which thesemolecules have been isolatedand reportedin the literature.

TherutinmoleculeisalsopresentinTables3and4;ithas53%

and56%probabilityofbeingpotentiallyactiveontheS.aureusand E.colimodels,respectively,anditwaschosenfortheinvitrostudies becauseofitsavailability.

Rutindilutedinsteriledistilledwaterorin2.5%DMSOshowed noactivityagainstthetwostrainsofS.aureus,ATCC25923and SAM-01.However,fortheE.colistandardstrain,ATCC25922,rutin exhibitedactivitywhendilutedin2.5%DMSOwithMICvalueof 455␮M(256␮g/ml).TheseresultscanbeseeninFig.4.

Souza(2009)andOliveira(2014)foundsimilarresultswhen evaluating rutin’s antimicrobial effect (Sigma-Aldrich) against Salmonella enterica, E. coli, S. aureus, and P. aeruginosa. In thesestudies,differentconcentrationsbetween15.625␮g/mland

10000␮g/ml and different ways of solubilizing rutin, e.g., in methanol,distilledwater,3%DMSO,wereused.

Thoseresultsdifferfromthatofthisresearchregardingrutin activityforE.coli.Thesedifferencesinresultscanbeattributedto thewayitwasdiluted,aswellastheproductbrand(rutin).

Rutinisaglycosidicﬂavonolandhasgreattherapeuticimpor- tanceforimprovingtheresistanceand permeabilityofcapillary vessels,withantioxidantactivities,anti-inﬂammatory,anticarcino- genicproperties,amongothers(Brechoetal.,2009).Accordingto Martinietal.(2009),rutinhasanactivityforsomeGram-positive andGram-negativebacteria.

Rutinshowed activityagainstE. coli,corroboratingwiththe resultobtainedinthepredictivemodelofE.coliwhererutinhad 56% probability of active potential. However, in the predictive modelofS.aureus,rutinhadanactivepotentialof53%,whichdis- agreeswiththeinvitroresultsbecauserutinshowednoactivityfor anyoftheS.aureus,ATCC25923andSAM-01strains.

Afacttobeobservedisthatforthegenerationofthemodels,the IC₅₀wasconsidered,andintheinvitrotests,MICwasconsidered.

Astudywasalsocarriedoutoftheinteractionbetweenrutinand threeoftheantibioticsthattheMRSAstrain,SAM-01,isresistant to,namely,oxacillin(whichisequivalenttomethicillin),amoxicillin+clavulanicacid,and penicillin,toobserveifrutin hasthe capacitytointeractwithsomeoftheseantimicrobialsandalterthe ﬁnalresponseinthebacterium.

Onlyrutindilutedinsteriledistilledwaterpresentedaninterac- tiveeffectforonetheMRSAstrains,SAM-01.Thisstrainisresistant tooxacillin,thereisnoinhibitionhaloformation;whenevaluating theeffectofrutinonit,a14mminhibitionhalowasformed(Fig.5).

Thepresenceofthishalo,accordingtothesensitivityvaluesinthe literature,isnotsufﬁcienttomakethestrainsensitivetotheaction ofthisinteraction,butitwasenoughtodecreasetheresistanceof thisstraintooxacillin.

Conclusion

Through the in silico tools used in this work,it was possi- ble to generate models to trace virtually the database of the Solanumgenus.TheS.aureusmodelselectedthirtymoleculeswith potentialeffectagainstMRSA,whereeightmoleculeshaveaprob- abilitygreaterthan60%.WiththeE.colimodel,itwaspossibleto

1 2 3 4 5 6 7 8 9 10 + -

Rutin diluted in sterile distilled water

Rutin diluted in DMSO 2.5%

DMSO 2,5%

control

DMSO 2,5%

control Control sterility

of rutin

Control sterility of rutin

S.aureus ATCC 25923

P. aeruginosa ATCC 27853 - rutin DMSO

E. coli ATCC 25922 - rutin DMSO 2,5%

DMSO 2,5% control in s. aureus - Rutin DMSO 2,5%

Staphilococcus aureus − SAM-01 Staphilococcus aureus − ATCC 25923

Control sterility of rutin

DMSO 2,5% control in p. aeruginosa

DMSO 2,5% control in E. coli

Legends: 1- [3,337 µM], 2- [1,678 µM], 3- [830 µM], 4- [410 µM], 5- [201 µM], 6- [105 µM], 7- [52.14 µM], 8- [26.2 µM], 9- [13.1 µM], 10- [6.55 µM], (+): positive control of

Fig.4. Analysisoftheantimicrobialactivityofrutinsolution(dilutedin2.5%DMSOandanothersolutiondilutedinsteriledistilledwater)forS.aureusstrainATCC25923 andSAM-01andE.coliATCC25922.

(6)

Fig.5.Interactiveeffectofrutinsolution,dilutedinsteriledistilledwater,with theantibioticsoxacillin,amoxicillin+ac.clavulanicacid,andpenicillin,againstthe strainSAM-01bymeansofthediskdiffusiontechnique.Firstline–steriledisk imbibedwithrutinsolution;fromlefttorightwehave:secondline–oxacillindisk (OXI),OXIdisksoakedwithrutinsolution,OXIdisksoakedwithsteriledistilled water;thirdline–penicillindisk(PEN),PENdiskimbibedwithrutinsolution,PEN disksoakedwithsteriledistilledwater;fourthline–amoxicillin+ac.clavulanicacid (AMC),AMCdisksoakedwithrutinsolution,AMCdisksoakedinsteriledistilled water.

identify221moleculespotentiallyactiveforthisbacteriumfrom thedatabaseof theSolanumgenus.Amongthesemolecules,26 moleculeshadanactivepotentialbetween80and88%probability.Theinvitrotestsperformedrevealedthatrutinhadnoactivity forS.aureusstrains,beingactivejustfortheE.colistrain.Rutinwas abletointeractwiththeoxacillinantibioticintheSAM-01(MRSA) strain,beingabletoreducetheresistanceofthisbacteriumtothis antibiotic.

Conﬂictsofinterest

Theauthorsdeclarenoconﬂictsofinterest.

Acknowledgements

TheauthorswouldliketothanktheCNPqforﬁnancialsupport.

References

Agra,M.F.,Nurit-Silva,K.,Berger,L.R.,2009.FloradaParaíba,Brasil:SolanumL.

(Solanaceae).ActaBot.Bras.23,826–842.

Brecho,J.R.M.,Machado,H.,Guerra,M.O.,2009.Rutina–estrutura,metabolismoe potencialfarmacológico.Rev.Int.Est.Exp.1,21–25.

CLSI,2010.PerformanceStandardsforAntimicrobialSusceptibilityTesting:Twen- tiethInformationalSupplement,ClinicalAndLaboratoryStandardsInstitute, DocumentM100-S20.CLSI,Wayne,PA,USA.

Das,S.,Lindemann,C.,Young,B.C.,Osterreich,B.,Ternette,N.,Winkler,A.C.,Paprotka, K.,Reinhardt,R.,Förstner,K.U.,Allen,E.,Flaxman,A.,Yamaguchi,Y.,Rollier,C.S.,

VanDiemen,P.,Blättner,S.,Remmele,C.W.,Selle,M.,Dittrich,M.,Müller,T., Vogel,J.,Ohlsen,K.,Crook,D.W.,Massey,R.,Wilson,D.J.,Rudel,T.,Wyllie,D.H., Fraunholz,M.J.,2016.NaturalmutationsinaStaphylococcusaureusvirulence regulatorattenuatecytotoxicitybutpermitbacteremiaandabscessformation.

Proc.Natl.Acad.Sci.U.S.A.113,3101–3110.

Hall,M.,Frank,E.,Holmes,G.,Pfahringer,B.,Reutemann,P.,Witten,I.H.,2009.The WEKAdataminingsoftware:anupdate.SIGKDDExplor.11,10–18.

Kleandrova,V.V.,Ruso,J.M.,Speck-Planche,A.,DiasSoeiroCordeiro,M.N.,2016.

Enablingthediscoveryandvirtualscreeningofpotentandsafeantimicrobial peptides.Simultaneouspredictionofantibacterialactivityandcytotoxicity.ACS Comb.Sci.18,490–498.

Kode srl, 2016. Kode Chemoinformatics. DRAGON 7.0,

https://chm.kode-solutions.net/index.php(accessed09.12.16).

Martini,S.,D’Addario,C.,Colacevich,A.,Focardi,S.,Borghini,F.,Santucci,A.,Figura, N.,Rossi,C.,2009.AntimicrobialactivityagainstHelicobacterpyloristrainsand antioxidantpropertiesofblackberryleaves(Rubusulmifolius)andisolatedcom- pounds.Int.J.Antimicrob.Agents34,50–59.

Melo,C.A.F.,Martins,M.I.G.,Oliveira,M.B.M.,Benko-Iseppon,A.M.,Carvalho,R., 2011.KaryotypeanalysisfordiploidandpolyploidspeciesoftheSolanumL.

PlantSyst.Evol.293,227–235.

Newman,D.J.,Cragg,G.M.,2016.Naturalproductsassourcesofnewdrugsfrom 1981to2014.J.Nat.Prod.79,629–661.

Oliveira,L.M.N.,M.Sc.Dissertation2014.Quantiﬁcac¸ãoderutina,atividadesantiox- idanteeantimicrobianadeextratosdepolpasesubprodutosdefrutastropicais.

UniversidadeFederaldoCeara,Ceará,pp.84.

Oliveira,R.A.G.,Lima,E.O.,Vieira,W.L.,Freire,K.R.L.,Trajano,V.N.,Lima,I.O.,Souza, E.L.,Toledo,M.S.,Silva-Filho,R.N.,2006.Estudodainterferênciadeóleosessen- ciaissobre aatividadedealguns antibióticosusadosnaclínica.Rev.Bras.

Farmacogn.16,77–82.

Ordaz,G.,D’Armas, H., Yá ˜nez,D.,Moreno, S.,2011. Composición químicade losaceitesesencialesdelashojasdeHelicteresguazumifolia(Sterculiaceae), Pipertuberculatum(Piperaceae),Scopariadulcis(Arecaceae)ySolanumsubin- erme (Solanaceae), recolectadas en Sucre, Venezuela. Rev. Biol. Trop. 59, 585–595.

Pereira,I.S.P.,Rodrigues,V.F.,Vega,M.R.G.,2016.FlavonoidesdogêneroSolanum.

Rev.Virt.Quim.8,4–26.

Pinto,F.C.L.,Uchoa,D.E.A.,Silveira,E.R.,Pessoa,O.D.L.,Braz-Filho,R.,Silva,F.M., 2011.Glicoalcaloidesantifúngicos,ﬂavonoideseoutrosconstituintesquímicos deSolanumasperum.Quim.Nova34,284–288.

Prado-Prado,F.J.,Uriarte,E.,Borges,F.,Gonzalez-Diaz,H.,2009.Multi-targetspectral momentsforQSARandComplexNetworksstudyofantibacterialdrugs.Eur.J.

Med.Chem.44,4516–4521.

Prado-Prado,F.J.,Gonzalez-Diaz,H.,Santana,L.,Uriarte,E.,2007.UniﬁedQSAR approachtoantimicrobials.Part2:predictingactivityagainstmorethan90dif- ferentspeciesinordertohaltantibacterialresistance.Bioorg.Med.Chem.15, 897–902.

Salzberg,S.L.,1994.BookReview:C4.5:ProgramsforMachineLearningMorgan KaufmannPublishersbyJ.RossQuinlan.Inc.,1993.MachineLearning,©Kluwer AcademicPublishers,Boston.ManufacturedInTheNetherlands,vol.16.,pp.

235–240.

Scotti, M.T.,Herrera-Acevedo, C., Oliveira, T.B., Costa, R.P.O., Santos, S.Y.K.O., Rodrigues,R.P.,Scotti,L.,Da-Costa,F.B.,2018.SistematX,anonlineweb-based cheminformaticstoolfordatamanagementofsecondarymetabolites.Molecules 23,103–113.

Song,C.M.,Lim,S.J.,Tong,J.C.,2009.Recentadvancesincomputer-aideddrugdesign.

Brief.Bioinform.10,579–591.

Souza,A.J.F.,M.Sc.Dissertation2009.Avaliac¸ãodosefeitosantimicrobianosderutina equercetinainvitro.InstitutodeBiologia,UniversidadeEstadualdeCampinas, Campinas,pp.62.

Speck-Planche,A.,Cordeiro,M.N.D.S.,2014.Simultaneousvirtualpredictionof anti-EscherichiacoliactivitiesandADMETproﬁles:achemoinformaticcomple- mentaryapproachforhigh-throughputscreening.ACSComb.Sci.16,78–84.

Ventola,L.C.,2015.Theantibioticresistancecrisis,Part1:causesandthreats.P.T.

40,277–283.

Vorontsova,M.S.,Knapp,S.,2012.AnewspeciesofSolanum(Solanaceae)fromSouth Africarelatedtothecultivatedeggplant.PhytoKeys8,1–11.