molecules Article Molecular Informatics Studies of the Iron-Dependent Regulator (ideR) Reveal Potential Novel Anti-Mycobacterium ulcerans Natural Product-Derived Compounds Samuel K. Kwofie 1,2,3,*, Kweku S. Enninful 1,4, Jaleel A. Yussif 1, Lina A. Asante 1, Mavis Adjei 1, Kwabena Kan-Dapaah 1, Elvis K. Tiburu 1, Wilhelmina A. Mensah 2, Whelton A. Miller III 3,5,6, Lydia Mosi 2 and Michael D. Wilson 3,4 1 Department of Biomedical Engineering, School of Engineering Sciences, College of Basic and Applied Sciences, University of Ghana, Legon P.O. Box LG 54, Accra, Ghana; kenninful@noguchi.ug.edu.gh (K.S.E.); jaleela96@yahoo.com (J.A.Y.); linaagyekumwaaa@gmail.com (L.A.A.); madjei008@st.ug.edu.gh (M.A.); kkan-dapaah@ug.edu.gh (K.K.-D.); etiburu@ug.edu.gh (E.K.T.) 2 West African Centre for Cell Biology of Infectious Pathogens, Department of Biochemistry, Cell and Molecular Biology, University of Ghana, Legon P.O. Box LG 54, Accra, Ghana; wamensah@gmail.com (W.A.M.); lmosi@ug.edu.gh (L.M.) 3 Department of Medicine, Loyola University Medical Center, Maywood, IL 60153, USA; wheltonm@seas.upenn.edu (W.A.M.III); Mwilson@noguchi.ug.edu.gh (M.D.W.) 4 Department of Parasitology, Noguchi Memorial Institute for Medical Research (NMIMR), College of Health Sciences (CHS), University of Ghana, Legon P.O. Box LG 54, Accra, Ghana 5 Department of Chemical and Biomolecular Engineering, School of Engineering and Applied Science, University of Pennsylvania, Philadelphia, PA 19104, USA 6 Department of Chemistry & Physics, College of Science and Technology, Lincoln University, Philadelphia, PA 19104, USA * Correspondence: skkwofie@ug.edu.gh; Tel.: +233-20-3797922  Received: 1 April 2019; Accepted: 29 May 2019; Published: 21 June 2019  Abstract: Buruli ulcer is a neglected tropical disease caused by the bacterium Mycobacterium ulcerans. Its virulence is attributed to the dermo-necrotic polyketide toxin mycolactone, whose synthesis is regressed when its iron acquisition system regulated by the iron-dependent regulator (ideR) is deactivated. Interfering with the activation mechanism of ideR to inhibit the toxin’s synthesis could serve as a possible cure for Buruli ulcer. The three-dimensional structure of the ideR for Mycobacterium ulcerans was generated using homology modeling. A library of 832 African natural products (AfroDB), as well as five known anti-mycobacterial compounds were docked against the metal binding site of the ideR. The area under the curve (AUC) values greater than 0.7 were obtained for the computed Receiver Operating Characteristics (ROC) curves, validating the docking protocol. The identified top hits were pharmacologically profiled using Absorption, Distribution, Metabolism, Elimination and Toxicity (ADMET) predictions and their binding mechanisms were characterized. Four compounds with ZINC IDs ZINC000018185774, ZINC000095485921, ZINC000014417338 and ZINC000005357841 emerged as leads with binding energies of −7.7 kcal/mol, −7.6 kcal/mol, −8.0 kcal/mol and −7.4 kcal/mol, respectively. Induced Fit Docking (IFD) was also performed to account for the protein’s flexibility upon ligand binding and to estimate the best plausible conformation of the complexes. Results obtained from the IFD were consistent with that of the molecular docking with the lead compounds forming interactions with known essential residues and some novel critical residues Thr14, Arg33 and Asp17. A hundred nanoseconds molecular dynamic simulations of the unbound ideR and its complexes with the respective lead compounds revealed changes in the ideR’s conformations induced by ZINC000018185774. Comparison of the lead compounds to reported potent inhibitors by docking them against the DNA-binding domain of the protein also showed the lead compounds to have very Molecules 2019, 24, 2299; doi:10.3390/molecules24122299 www.mdpi.com/journal/molecules Molecules 2019, 24, 2299 2 of 21 close binding affinities to those of the potent inhibitors. Interestingly, structurally similar compounds to ZINC000018185774 and ZINC000014417338, as well as analogues of ZINC000095485921, including quercetin are reported to possess anti-mycobacterial activity. Also, ZINC000005357841 was predicted to possess anti-inflammatory and anti-oxidative activities, which are relevant in Buruli ulcer and iron acquisition mechanisms, respectively. The leads are molecular templates which may serve as essential scaffolds for the design of future anti-mycobacterium ulcerans agents. Keywords: buruli ulcer; Mycobacterium ulcerans; iron dependent regulator (ideR); metal binding site; DNA–binding site; natural product compounds; molecular docking; molecular dynamics simulation 1. Introduction Buruli ulcer is an infectious disease caused by Mycobacterium ulcerans [1]. It is a skin necrotizing disease that kills the cells of the skin and other soft tissues [2] and characterized by chronic ulceration of subcutaneous fat that leaves victims with unbearable deformity and disability when left untreated [3]. The pathogenesis of the disease starts as a painless nodule on the skin and may eventually grow into an extensive ulcer that can cover up to about 15% of an individual’s body. It is often referred to as the disease of the poor because most people stricken by the disease are inhabitants of poor rural communities with inadequate or no basic social amenities like potable water [4]. There are over 30 countries worldwide with reported cases of Buruli ulcer [5] and most of them are in Central and West Africa with few exceptions, including Australia. Cote d‘Ivoire, Ghana and Benin rank as the three countries with the highest prevalent rates [3]. About 1200 Buruli ulcer cases were reported in Ghana between 1993 and 1998 by a passive surveillance system established in the country. Between 2004 and 2014, reported cases exponentially increased to more than 9000 [6]. M. ulcerans is a slow growing bacterium doubling every 72 h [7] and like other slow-growing bacteria Francisella tularensis and Borrelia burgdorferi, it is classified as an environmental pathogen which implies that it survives in the external environment [8]. Its association with water and water bodies is well recognized and has generated hypotheses, such as bites of insects as the mode of transmission [2,9], which suggest mechanical transmission [10]. However, a recent study has demonstrated that acanthamoeba could be the host of the mycobacterium in the environment [11]. Despite the fact that the age or sex of the host have not been proven to be risk factors, statistics of recorded cases available suggest that the disease is more prevalent among women and children between the ages of 5 and 15 [3]. Virulence by M. ulcerans is attributed to the synthesis of a dermo-necrotic polyketide toxin called mycolactone [12]. The toxin is exported through the bacterial envelope and accumulates in an extracellular matrix [13]. It has also been shown to have immunosuppressive properties by inhibiting the phagocytic abilities of the phagocytic white blood cells and killing neutrophils dispatched to infected tissues [2,12]. Mycolactone also blocks exocytosis by blood platelets and mast cells, impairing wound healing processes [14]. Like all mycobacteria, M. ulcerans requires iron for growth [15]. Insufficient iron retards the growth of the bacterium and high intracellular level could cause irreparable oxidative damage [16]. The iron acquisition pathway of the mycobacterium ensures that an optimum amount of iron is taken in by the bacteria and this is regulated by the iron dependent regulator (ideR). Upon iron binding to ideR, it is activated and then binds to the iron boxes in the promoter regions of iron regulated genes, thereby deactivating iron acquisition (MbtB gene), activating iron storage (BfrB) and deactivating irtA (iron transport) and the reverse happens when iron levels are low. The binding of iron also induces structural changes in ideR, with the protein moving from an ‘open’ conformation in its inactive state to a ‘close’ conformation [15]. However, research has shown that a decrease in intracellular iron levels, which deactivates ideR reduces the synthesis of mycolactone [17]. This evidence led us to suggest that Molecules 2019, 24, 2299 3 of 21 any molecule that targets the ideR to either prevent iron binding or induce conformational changes is potentially a drug. Natural products are chemical compounds that are produced by a living organism from nature which has the bioactivity capable to be used as drugs [18]. They represent an enormous reservoir of diverse sources of bioactive chemicals and is very essential to drug discovery [18]. This study employed computer-aided drug design techniques to screen for potential inhibitory compoundMsolfercuolems 20a19n, 2A4, xfr FiOcRa nPEEnRa RtEuVrIaEWl p roduct database (AfroDB). Also, undertake molec3u olfa 2r1 modeling of the structure of ideR of M. ulcerans, as well as molecular dynamics simulations to identify compounds evidence led us to suggest that any molecule that targets the ideR to either prevent iron binding or with the pointdeunctei acol nofforimndatuiocnianl gchcaonngefso irsm poatteinotniaalllyc ah darnugg.e s in the ideR. Natural products are chemical compounds that are produced by a living organism from nature 2. ResultswahnidchD haiss cthues bsiiooanctivity capable to be used as drugs [18]. They represent an enormous reservoir of diverse sources of bioactive chemicals and is very essential to drug discovery [18]. 2.1. Three DimeTnhsisi osntuadly( 3emDp)lMoyeod ecloPmrpeudtiecrt-iaoidneadn ddruVga dliedsiagtnio tnechniques to screen for potential inhibitory compounds from an African natural product database (AfroDB). Also, undertake molecular Usingmtohdeelianmg oifn tohea sctriudctsuereq oufe idnecRe oof Mf t. huleceirdanesR, asf owrelMl as. muolcleecrualnasr d(yUnnamiPicrso stimIDulaAtio0nPs Tto6 i6d)enatsifya query to the Basic LcomcaploAunldigs nwmithe tnhte SpeoaterncthialT oof oinld(uBcLinAg ScoTn)fo[r1m9a],ti2on3ahl cihtsanwg esr ien ftoheu inddeRf. rom the Protein Data Bank (PDB) [20]. The best five of these 23 hits with a percentage sequence identity of 80% or more were 2. Results and Discussion selected (Table 1). Chain A of the crystal structure of the ideR from M. tuberculosis with PDB ID 1FX7 (1FX7_A) a2.n1.d Th2r.e0e DÅimrenssoiolnuatl i(o3Dn) wMaodselc Phroedsiectniona asntdh Vealtiedamtiopnl a te for modeling because it had the highest sequence coveUrsaingge thtoe atmhienot aacrigde steqaunedncea orf ethaes iodneRa bfolry Mg. ouolcedraEns- (vUanliuPeroto IfD8 A×0PT1606−) 1a5s4 a. quFeirvye tom thoed els were predicted wBaistihc LMocoadl eAllliegrnmvenrst iSoenar9ch.1 T7ouols i(nBLgAESaTs) y[1M9]o, d23e lhlietsr w4.e0re[ 2fo1u]nads faronmi nthter Pfaroctee.inT Dhaetam Boadnke l with the lowest Dis(cPrDeBte) [O20p].t iTmhei zbeesdt fPivoet eonf tthiaelseE 2n3e hrigtsy w(DithO aP pEer)csecnotargee wseaqusecnhceo sideenntaitsy tohfe 80b%e sotr mmoordee wl.eDre OPE is an selected (Table 1). Chain A of the crystal structure of the ideR from M. tuberculosis with PDB ID 1FX7 atomic dis(t1aFnXc7e_A-d) eapnde n2.d0 eÅn rtesstoaluttiisotnic waalsp cohtoesnent iaasl ,thwe hteimchpliastec faolrc muoladteelidngf rboecmauases iat mhapd lteheo hfinghaetsitv e protein structuressteoquaesnscees csotvheerageen etor gthiees taorfggete annedr aat eredaspornoabtelyi ngomodo dEe-vlasl.uMe oof d8e ×ls 1w0−1i5t4h. Ftihvee mloowdeelss twDerOe PE scores have the mporesdticstteadb wleithm Minodimelliezre vderesinone r9g.1y7 u[2si2n]g. ETahsyeMliosdteollefrg 4e.0n [e21r]a ates danm inotedrfealcse. wThieth mtohdeeil rwDithO thPeE scores is shown in Tloawbelset SD1is.crTethee Ospetilmecizteedd Pmoteondtieall E3nheragdy (tDhOePleEa) ssctoDreO wPasE chsocsoerne aos tfh−e 2be5s,t3 m55o.d7e7l.7 D3O4P, Ea nisd anc onsists of atomic distance-dependent statistical potential, which is calculated from a sample of native protein 230 residusetsrufcoturmresi ntog aessleesvse tnhea elnpehrgaiehs eolfi cgeensearantdeds pervoeteninb metoadeslhs.e Metosd(eFlsi gwuitrhe th1e) .lowest DOPE scores have the most stable minimized energy [22]. The list of generated models with their DOPE scores is Table s1h.oHwonm ino Tloagbulee Ss1o. fTMhey sceolbeactcetder miuomdeull c3e hrands t’hIer olena-sdt eDpOenPdE esncotrRe eogf u−l2a5t3o5r5.(7id77e3R4), .aTnhde csoenasirsetsb oefs t 5 of 23 hits2w30i trhespideurecse nfotramgeinigd elnetvietyn afrlpohma 8h0e%liceasn adnda bsoeveno bbettai nsheedetfsr o(FmiguBraes i1c).L ocal Alignment Search Tool (BLAST). Table 1. Homologues of Mycobacterium ulcerans’ Iron-dependent Regulator (ideR). These are best 5 of 23 hits wiHthi tpserDceenstcargiep tiidoenntity from 80% and above obQtauineerdy fCroomv eBrasic LEocVala lAuleignmeIndt eSnetairtcyh PDB ID Tool (BLAST). Chain A, Crystal Structure of ideR from Mycobacterium tuberculosis Query 100% 8 × 10 −154 92% PDB 1FX7_A Hits Description E value Identity Chain A, Crystal Structure of a two-domain ideR-DNA Cover ID Chain Ac, oCmrypstlael xStrcurcytsutrae loff oidremR from Mycobacterium 64% 1 × 10 −96 91% 2ISZ_A 100% 8 × 10−154 92% 1FX7_A tuberculosis I CChh −95 aiani nA,A C,ryisdtaelR StfrruoctmureM of. at utwbeor-dcuomloasiins ideR-DNA 60% 2 × 10 96% 1B1B_A Chain A, Crystal Structure of the Nickel—activated 64% 1 × 10−96 91% 2ISZ_A complex crystal form I −95 two-domaiCnhIarion nA—, iddeeRp freonmd Men. tuRbee 64% 6 × 10 91% 2ISY_Arcguulolsaist or 60% 2 × 10−95 96% 1B1B_A Chain AC,hDainip Ah, tChreyrsitaal TStorxucRtuerep oref tshseo Nr i(cCke1l0 –2 adctiMvautetda tnwto)- 64% 6 × 10−95 −62 91% 2ISY_A complexed with Nickdeolmaanind IrDontx –r dceopnensdeennst uRsegbuilnatdoirn g sequence 52% 1 × 10 80% 1F5T_A Chain A, Diphtheria Tox Repressor (C102d Mutant) 52% 1 × 10−62 80% 1F5T_A complexed with Nickel and Dtxr consensus binding sequence Figure 1. AFigcuarret o1o. An rcearptoreonse rneptaretisoentaotifotnh oef 3thDe 3hDo mhoomlolgoygym moodell off IIddeReR ofo Mf .M ul.ceurlacnesr. aAnlsp.hAa hlpelhixaesh elixes are colored inarreed c,olboeretda isnh reeedt, sbeinta ysheelelotsw in ayneldlowlo aonpds loinopgs riene gnre.eRn.e Rgeigoinon‘ A‘A’’ cciirrcclleedd ini nblbuelu inediincadteisc athtee s the area area of the metal binding pockets whilst region ‘B’, circled in violet shows the DNA–binding domain of the metoafl tbhein pdroitnegin.p ockets whilst region ‘B’, circled in violet shows the DNA–binding domain of the pro tein. Molecules 2019, 24, 2299 4 of 21 Molecules 2019, 24, x FOR PEER REVIEW 4 of 21 InInv avlaidliadtaintigngth tehe3 D3Dm modoedle, lt,h tehRe aRmamacahcahnadnrdarnanp lpoltost hsohwowededth tehep eprecrecnetnatgaegeo forfe rseidsiudeuseisn inth tehe alalollwowededre rgeigoinosnsto tob ebe5 .05%.0%, r,e rseidsiudeusesin inth tehefa fvaovroerdedre rgeigoinonto tob ebe9 39.43%.4%a nadndre rseidsiudeuseisn inth teheo uotulitelirer rergeigoniotno tboe b1e.5 1%.5(%Fi g(Fuirgeu2r)e. 2A).R Aam RaacmhaanchdaranndrZa-nsc Zor-secoofre1 .4o8f 81w.48a8s owbatasi noebdtauinseindg uWsiHngA TWCHHAETCCKH[2E3C].K Th[2is3]in. Tdhiciast iensdtihcaatteosu trhpart eoduicr tpedremdioctdeedl mis oodf erel aiss oonf arbealysognoaobdlyq ugoaloitdy qsuinaclietyo vsienrc9e0 o%veorf 9re0s%id oufe rsewsideruees inwtheeref aivno trheed fraevgoiorend[ 2r4e]g.ioTnh e[2Z4-]s.c Tohree eZx-psrceosrsee sexhporweswseesl lhtohwe b wacekllb othnee bcaocnkfobromnea tcioonnsfoorfmaalltiroensisd oufe sall arreecsoidrureessp aorned cionrgretosptohnedkinnogw ton tahlleo kwnaobwlena arellaoswinabtlhee aRreaams aicnh tahned Rraanmpalcohta[n2d3]r.aAn ppleortf e[2ct3]Z. -Asc opreerfiesct exZp-escctoerde tios beexp1e.0c,teadn dtoZ b-sec o1.r0e,s aanbdo vZe-s4c.0oraensd abboelvoew 4.−04 a.0ndar ebevleorwy u−4n.c0o amrem voenr.yF uunrtchoemr mveornifi. cFautirothner wvitehriPfiRcaOtiCoHn EwCitKh sPhRoOwCeHd EthCeKr esshoolwuteiodn th(ne orremsoalluittyio)no f(ntohremstarluitcyt)u oref tthoeb setr1u.5ctÅurew thoi cbhe i1n.d5 iÅca tweshiach goinoddirceastoelsu ati ognoosdin rceesmolousttiohnig shin-rcees omluotsiot nhiXg-hr-aryessotrluuctitounre Xs -hraavye satrruecstoulruetsio hnawveit ha irnes1o.5luatniodn2 .w0 iÅth[i2n4 1]..5 and 2.0 Å [24]. Figure 2. Ramachandran plot of ideR. This plot gives a general overview of the quality of the protein strFuigctuurree .2. TRhaemsaeclehcatneddramno pdleolt iosfs iedeenR.t oThhiasv peloat rgeiavseosn aa bgleynegroaol dovqeuravliietwy soifn tcheem qousatliotyf oitfs trhees ipdruoetesin (>s9t0r%uc)tufarlel.w Tihthei nsetlheectfeadv omreodderel gisio sne.en to have a reasonably good quality since most of its residues 2.2. Bin(>d9in0%g )S fiatell Iwdeitnhtiinfi cthatei ofanvored region. 2.2.T Bhiendbiningd Sinitge Isdietnetsifoicbattaioinn e d from KVFinder, COACH and COFACTOR were compared with identified binding sites of co-crystallized structure, 1FX7_A, which was pre-processed and used as a tempTlahtee bfoinrdtihneg 3sDitems oodbetal i[n2e3d– 2f6r]om(T aKblVeF2in).deTrh, eCrOesAuCltHo batnadin eCdOfFroAmCTCOORF AwCerTeO cRomprpoavriedde dwaith coindfiendteinficeed sbcionrdeinofg 0s.i5te4s, boaf sceod-corynsatarlalinzgede bstertuwcetuenre0, 1aFnXd71_,Aw, withhivcahl uweassc plorsee-prrtooc0esisneddi caantdin ugsaedle asss a retleiamblpelaptree dfiocrt iothnea n3dD vmaloudeselc l[o2s3e–r2t6o]1 (iTnadbiclea t2in).g Tahme orreesruellti aobbletapirneeddi cftiroonm. FCroOmFAthCeTvOalRu eporbotvaiidneedd , a confidence score of 0.54, based on a range between 0 and 1, with values closer to 0 indicating a less reliable prediction and values closer to 1 indicating a more reliable prediction. From the value Molecules 2019, 24, 2299 5 of 21 Molecules 2019, 24, x FOR PEER REVIEW 5 of 21 it can be concluded that the binding site predicted is reasonably reliable. The modeled structure had a bionbdtianignesdit,e ist imcainla briet yco(BnSc-luscdoerde )tohfa1t .t9h,ea bBiSn-dscinogre s>ite1 prerfledecictstead siisg nreifiacsaontablolyca rlemliabtclhe. bTehtwe emenodtehleed prsetdruicctteudrea nhdadte am bpinladtienbgi snidtein sgimsiitlear[i2ty7] (.BTSh-secCorOe)F AofC 1T.9O, aR BsSer-svceorrael s>o 1p rreoflveicdtse da asigtenmifpiclantet lmocoadle mlinatgch (TbMet)wsceoerne tohfe0 .p9r9e9dfiocrtetdh eamndo dtelmedplparteo tebiindwinhgic hsisteh o[w27e]d. Tthaet aCOgoFoAdCtTemORpl asteervwear saulsoed p. rTohveidTeMd a sctoerme pralantge ems obdetewlinegen (T0Ma)n dsco1rwe oitfh 0a.9b99e tftoer ttheme mploadteelheadv pinrogteainv awluheicchl ossheorwtoedo tnhea[t 2a8 ]g.oTodh etermespullatte prwovaisd uesdedb.y TChOe ATMCH scsohroew raendgaesc obnetfiwdeennc e0 sacnodre 1o wf 0it.h0 1a wbehtitcehr itsemlepsslartel ihaabvlein. gIt aa vlsaoluper ecdloicseterd tot hoene bi[n2d8i]n. gThsiet ererseusildt upersovasidHed21 b9ya CndOAY2C2H3. sThwoowoefdt ha ecobninfiddienngcsei tsecsorep ofr t0e.0d1b wy hFiecehs eise lteasls. r[2el9i]a(bblein. dIti naglso sitpersed1iactned 3th) ew beirnedsinimg isliatre troestihdousees parse Hdi2c1t9e danbdy YC2O23A. CTHwoa nodf thCeO bFiAndCiTnOg Rsi.teAsl rletphoerbteidnd biyn gFeseitses et praeld. i[c2t9e]d (biyndthinegK sVitFeIsN 1D aEnRd m3) awtcehred sifmouilraor ftoth tohsoesep rperdeidcitcetdedin btyh eCtOemACplHat ae.ndB iCndOiFnAg CsiTteOsR1. aAnldl t2he wberinedthinegm soitsets pprreefdericreteddb binyd thineg KsViteFsINsiDncEeRt hmeaytcbhoethd pfolauyr eosf stehnotsiael prroeldesicitnedth ien atchteiv taetmiopnlaotfe.i dBeiRndaisng mseittaelsb 1i nadnidn 2g wsiteerse 1thaen md o2s, tr epsrpeefecrtirveedl ybi[n1d3,i1n4g] .siHteosw sienvceer ,thmeeyt ablobthin pdlianyg essitse n2t(iFali gruolres3 i)np trhoev aecdtitvoabtieon thoef midoesRt saus imtabetlael bbiinnddiinngg ssiitteesb e1c anudse 2o, freitsspreecltaivtievleyl y[1l3a,1rg4e]. vHooluwmeveeorf, m66e.9ta6l Åbi3nadnindgs usirtfea 2c e(Fairgeuaroef 3) 10p0r.8ovÅe2d, wtoh ibceh tmhea ymaocscto msumitaobdlaet ebirnedlaitnivge slyitela brgeecacuosme poofu intsd rse. latively large volume of 66.96 Å3 and surface area of 100.8 Å2, which may accommodate relatively large compounds. Table 2. Summary of predicted binding sites of the model and reported binding sites of the template (1FTXa7b,leid 2e.R SufrmommaMry. toufb perrceudloicstise)da bnidndthinogse spitreesd oifc ttehde fmorotdheel manodd reelpvoiartCedO bAinCdHinagn sditCesO oFfA tCheT OteRm.plate (1FX7, ideBRI fNroDmIN MG. tSuIbTerEcuPlRosEisD) aICndT ItOhoNse prediRcEteSdI DfoUr tEhSe AmTodTeHl EviBa ICNODAINCHG aSnITd ECOFACTOR. BINDCINOGFA SCITOE RPREDICTION HRiEsS79ID, GUlEuS8 3A,TH TisH98E, BGIlNu1D7I2N, Gl nS1I7T5E COCAOCFAHCTOR His79, Glu H83is, 2H19is, 9H8i,s G22l3u172, Gln175 REPORTECDOABICNHD ING SITES OF THE TEMHPisL2A19T, EH(i1sF2X237 ) [29] REPORTED BINDING SITES OF THE TEMPLATE (1FX7) [29] 1 (M 2 (M1 e t(aMl ebinding site 1) His79etal btianld biningdsinitge 2si)te 1) HiMs7e9 ,, GGlluu8833, His98,t10, Cys,1 H02is,9G8l, G Glluu117u105, 7 2, Gln175 H2i,s G10ln6175 2 (Meta3l binding site 2) Met10, CHyiss211092,, HGilsu212035, His106 4 3 HisH21i9s,2 H12is223 4 His212 FiFgiugruer3e. M3. etMaleBtainl dBiningdsiinteg2 osiftteh e2I doefR tmheo dIdele.RB inmdoindgelp. oBckinedt hinags bpeoencksehto whansi nbseuernfa csehorewpnre sienn tsautirofna.ce Threepimreasgeentwataisonge. nTehrea tiemdawgeit wh Pasy MgeOneLramteodle wcuiltahr PvyisMuaOliLz amtioonlectouolal.r visualization tool. 2.3. Anti-Mycobacterial Lead Discovery 2.3. Anti-Mycobacterial Lead Discovery To the best of our knowledge, no known inhibitors have been screened against the ideR of M. ulceTroan tsh.eH boewst eovfe or,ufir vkenpoowtleendtgien,h niboi tkonroswconm inphriisbiintogrsN hSaCv3e0 b36e0en0 (sIcCre50e:n5e.d4 8agµagi/nmstL t)h, eN iSdCeR12 o4f5 3M. (ICul5c0e:ra1nµs.g H/moLw)e, vNeSr,C f6iv5e7 4p8o(teICnt5 0i:nh23ib.9itoµrgs/ mcoLm),pNriSsiCn2g0 N17S7C33(0I3C65000: 1(I4Cµ50g: /5m.4L8 )μagn/dmNL)S, CN2S8C21629495(3I C(I5C0:50: 241.3 μµgg//mL)), hNavSeC6b5e7en48i d(eICnt5i0fi: e2d3.a9g μaign/smt ML)., tNubSeCrc2u0l1os7i7s3[ 3(0I]C. 5T0:h 1e4se μcgo/mmpLo)u anndds aNreSfCro28m26th99e N(IaCt5i0o: n2a4l.3 μg/mL) have been identified against M. tuberculosis [30]. These compounds are from the National Cancer Institute (NCI) database and were screened against the DNA-binding site of the ideR of M. tuberculosis, a close homologue to the ideR of M. ulcerans [30]. A library composed of the five potent Molecules 2019, 24, 2299 6 of 21 Cancer Institute (NCI) database and were screened against the DNA-binding site of the ideR of M. tuberculosis, a close homologue to the ideR of M. ulcerans [30]. A library composed of the five potent inhibitors and 832 compounds retrieved from AfroDB [31] were virtually screened against the metal binding site 2 of the modeled structure of the ideR, M. ulcerans. A total of 9272 different poses were generated from the compounds retrieved from AfroDB after the virtual screening. Hits were selected from the docking results by filtering with four criteria, which were drug likeness determined with Lipinski’s rule of five [32] and Veber’s rules [33], proper fit into the binding pocket, biomolecular interactions between ligands with binding site residues, and high binding affinity. Twenty compounds were shortlisted as top hits (Table 3). Figure 4 shows one of the selected hits firmly docked in the binding site. The biomolecular interactions between the five inhibitors and the binding site residues of the modeled ideR were also compared to those of the hits. The biomolecular interactions between the compounds and the metal binding site 2 of the modeled protein were generated with the processed_VinaResult.py script in the Autodock tools package, a python script useful for retrieving amino acid residues interacting with the docked ligand. We focused on residues located within the metal binding site of the ideR. The most common binding site residues with which the compounds interact with are Cys102 and Met10, an indication that these residues may perhaps play essential roles in the metal binding site. The five potent inhibitors, however, formed interactions with only His98, an amino acid residue of the metal binding site 1. All the Top 20 hits also formed interactions with His98, which indicates the amino acid could be very essential for ideR and may be involved in the activities of both metal binding sites. Table 3. Table of Top 20 hits and five potent inhibitors with their binding energies and interacting metal binding site-2 residues. Residues were determined using the processed_VinaResult.py script in the Autodock tools package. LIGAND ID BINDING ENERGY RESIDUES LIGANDINTERACTS WITH NSC12453 −7.5 His98 NSC201773 −7.5 His98 NSC282699 −7.5 His98 NSC303600 −7 His98 NSC65748 −7 His98 ZINC000005357841 −7.4 Cys102, His98, Met10 ZINC000013327497 −7.1 Met10, Cys102, His98 ZINC000013481884 −7.3 Glu172, His98, Cys102 ZINC000014417338 −8 Glu172, His98 ZINC000014811038 −7.7 Glu172, His98, Cys102 ZINC000014819573 −7.4 His98, Cys102, Met10 ZINC000018185774 −7.7 His98, Cys102, Met10 ZINC000033831303 −7.4 Met10, Cys102, His98 ZINC000095485893 −7.2 Met10, His98, Cys102 ZINC000095485918 −6.9 Glu172, Cys102, His98 ZINC000095485921 −7.6 Met10, His98, Glu172 ZINC000095486065 −7.5 Met10, His98, Cys102, Glu172 ZINC000095486093 −7.1 Glu172, Cys102, His98 ZINC000095486151 −7.3 Cys102, Met10, His98 ZINC000095486157 −7 His98, Met10, Cys102 ZINC000095486193 −7.2 Glu172, His98, Cys102 ZINC000095486235 −8.3 His98, Cys102 ZINC000095486265 −7.8 Met10, His98, Cys102 ZINC000095486301 −7.6 His98, Met10, Cys102 ZINC000095486336 −8.4 Glu172, His98 Molecules 2019, 24, x FOR PEER REVIEW 6 of 21 inhibitors and 832 compounds retrieved from AfroDB [31] were virtually screened against the metal binding site 2 of the modeled structure of the ideR, M. ulcerans. A total of 9272 different poses were generated from the compounds retrieved from AfroDB after the virtual screening. Hits were selected from the docking results by filtering with four criteria, which were drug likeness determined with Lipinski’s rule of five [32] and Veber’s rules [33], proper fit into the binding pocket, biomolecular interactions between ligands with binding site residues, and high binding affinity. Twenty compounds were shortlisted as top hits (Table 3). Figure 4 shows one of the selected hits firmly docked in the binding site. The biomolecular interactions between the five inhibitors and the binding site residues of the modeled ideR were also compared to those of the hits. The biomolecular interactions between the compounds and the metal binding site 2 of the modeled protein were generated with the processed_VinaResult.py script in the Autodock tools package, a python script useful for retrieving amino acid residues interacting with the docked ligand. We focused on residues located within the metal binding site of the ideR. The most common binding site residues with which the compounds interact with are Cys102 and Met10, an indication that these residues may perhaps play essential roles in the metal binding site. The five potent inhibitors, however, formed interactions with only His98, an amino acid residue of the metal binding site 1. All Molecules 2019, 24, 2299 7 of 21 the Top 20 hits also formed interactions with His98, which indicates the amino acid could be very essential for ideR and may be involved in the activities of both metal binding sites. Fiigurree 44.. ZIINC00001188118855777744 ffiirrmllyy docckeed iintto meettall biindiing ssiittee 2 off iideeR.. Thee biindiing ssiittee iiss sshoown aass aa ssoolliid ssurrffaaccee whiillsstt tthee lliiggaand iiss sshoown iin whiittee ssttiicckk moodeell.. Thee iimaaggee waass ggeeneerraatteed wiitthh PPyyMOLL.. 2.4. ETvaablulea t3i.o nTaobflAe uoft oTdoopck 2V0 ihniats’s aPnedr ffoirvme apnocteent inhibitors with their binding energies and interacting Tmheetaal binlidtyinog fsiAteu-2t oredsoidcukeVs.i Rneasitdouaecsc wuerraet edleyterramnikneddo ucskiendg tlhige apnrodcseswseads_eVvinaaluRaetseudlt.upysi sncgripdti ffine rent metritchse cAoumtopdroicskin tgootlhs epacrkeague.n der the curve (AUC) of the Receiver Operating Characteristic (ROC) curve, BoltzmannLI-GenAhNaDn IcDe d disBcIrNimDIiNnaGt iEoNnERoGf YR OCRE(SBIDEUDERSO LICG)AaNnDd INenTrEiRcAhmCTeSn Wt IfTaHct ors (EFs) [34]. The Receiver OperNaStiCn1g24C53h aracteristic (R−7O.5 C) curve is useful for meHasisu9r8i ng the ability of a docking or virtual screening NsoSfCt2w01a7r7e3 to distinguis−h7.5b etween active and inactHivise98c ompounds with respect to a receptor [32–35]. TNhSeC2p8e2r6f9o9 rmance of th−e7d.5o cking is measured by caHlcius9l8a ting the area under the curve NSC303600 −7 His98 (AUC) value. TheNcSlCo6s5e7r48th e AUC value−7i s to 1, the better the abiliHtyis9o8f the docking to discriminate between activeZIaNnCd00in00a0c5t3i5v7e84c1o mpound−s7..4A UC value less thaCnys01.052,i Hs icso98n, sMidete1r0e d poor discrimination ability, 0.5 to 0Z.I7NiCs00m00o1d33e2r7a4t9e7, 0.7 to 0.−87.i1s acceptable, 0.8 toM0et.190,i Cs yrse1a0s2,o Hnias9b8l y good and 0.9 to 1 is excellent [34,3Z6]IN. CT0h0e00A13U48C188v4a lues wer−e7.o3 btained by screenGilnug1732,4 Haisc9t8i,v Ceyss1a0n2d 1689 decoys against ZINC000014417338 −8 Glu172, His98 metal binding sZiItNeC200a0n0d14D81N10A38– binding s−i7t.e7 s of the modeled iGdeluR17o2f, HMis.9u8,l cCeyrsa1n0s2 and that of the ideR of M. tuberculosisZ. ITNhCe00A00U14C819o5f73R OC (ROC−7_.4A UC) values obtaiHneisd98w, Ceyrse10a2,b Movete100 .70 for all the binding sites, which falZlsINwCi0t0h0i0n18t1h8e57a7c4 ceptable d−7is.7c rimination abilityHraisn98g, eCy(Ts1a0b2l, eM4eta1n0 d Figure 5). Since ROCZINeCv0a0l0u0a3t3e8s31t3h03e overall p−7e.r4f ormance of the dMocekt1i0n, gCyms1e02t,h Hoids98to distinguish between ZINC000095485893 −7.2 Met10, His98, Cys102 actives and inaZcItNivCe00(o00v9e5r4a85ll91e8n richment−)6[.93 7–39], early enrichGmlue17n2t, wCyass10e2v, Halius9a8t ed using the BEDROC and EFs (TableZI4N).C0B0E00D95R48O59C21v alues ob−t7a.i6n ed were betweenM0e.t10a, nHdis908.,2 G,luw1h72i lst EF scores averaged a round 2.7 for 1% to 20% of the screened compounds. BEDROC values obtained were higher than the expected value for random selection, which is 0.05 [40], BEDROC values greater than 0.5 are considered as the best performance for early enrichment. Even though, ROC_AUC values indicated that the overall performance of docking was successful in distinguishing between the entire library of actives and decoys, early recognition was suboptimal, as shown by EFs and BEDROC values (Table 4). Therefore, methods used for selecting hits considered the whole screened library and then predicted leads were consolidated with prediction of anti-mycobacterial activity, which was reinforced with structural similarity analysis of known anti-mycobacterials. Molecules 2019, 24, 2299 8 of 21 Table 4. Respective values of ROC_AUC, BEDROC and Enrichment Factor for the metal binding site 2 and DNA- binding sites of ideR for M. ulcerans (BU_MBS and BU_DBS) and that of M. tuberculosis (TB_MBS and TB_DBS). BEDROC Enrichment Factor ROC_AUC (alpha = 20.0) 1% 10% 20% BU_MBS 0.702 0.137 2.979 2.355 2.208 BU_DBS 0.743 0.143 2.979 2.355 2.650 TB_MBS 0.727 0.174 0 2.355 2.797 Molecules 2019, 24, xT FBO_RD PBESER REVI0E.W70 3 0.175 5.95 2.650 2.355 8 of 21 FiguFrieg5u.reR e5c. eRiveceerivOepr eOrapteinragtiCnhga Crahcatrearcistteircis(tRicO (CR)OcCu)r vceurfvoer ifdore RidoefRM o.f uMlc.e uralcnesraannsd aMnd. tMub. etrucbuelrocsuisl.osis. MetaMl ebtianld biningdsinitge 2sitaen 2d aDnNd AD—NbAi n–d biinngdsinitge soifteid oefR idoefRM o.fu Mlc.e rualncesraarnes sahroe wshnoiwnnre idn arendd agnrdee gnr,ewenh,i lwsthilst thattohfatM o.f tMub.e rtcuubleorsciusloasries sahreo wshnoiwnnb lianc kblaanckd avniodl evti,orleest,p reecstipveecltyi.veTlyh.e Tchuer vceusrvweesr ewoebret aoinbetadinaefdte rafter dockdioncgk3in4ga c3t4i vaecstiavneds a1n6d89 1d68e9co dyescaogyasi angstaitnhset rtehsep reecstipveectbiivned biningdsiintegs s. ites. 2.5. In Silico ADMET Studies 2.5. In silico ADMET Studies The in silico Absorption, Distribution, Metabolism, Elimination and Toxicity (ADMET) test profiles the pharmThaceo kinin esitliicc,os tArubcstuorrpaltiaonnd, pDhiysstiroibchuetimonic,a lMperotapbeortliiessmo,f sEmlimallincoamtiopno uanndds [T4o1x].icTihtye m(AaDinMaiEmTo) ftest precplirnoifcialel sA tDheM pEhTartemstawcoaksintoeteilci,m stirnuactetuwraela aknddr upghycasinodcihdeamteiscainl pthreopeaerrltyiesst oagf essmoaflld croumg pdoevuenldops m[4e1n].t ,The whimchahinel pasimin opfl apcrinecglimniocrael eAmDpMhaEsiTs otenstp owteans ttiaol leylibmetitneartde rwugeacakn ddriduagt ecsan[4d2i]d. aInteds riung tdhees ieganr,lyli gsatangdess of are drercuogg dneizveedlobpymceenrtt,a winhipcrho pheelrptise sint hpalatcdinisgt imngoureis ehmdprhuags-ilisk oena pnodtennotnia-ldlyru bge-tltiekre dcroumg pcaonudniddsat[e4s3 ][.42]. TheIrnef dorreu,gs odmeseigonf,t lhigeapnrdops earrtei ersecthoagtnwizeerde bcoyn cseirdtearined pwroepreerhtiyeds rtohgaet ndibsotinndgsu, issohl udbriuligty-l,ibkieo aanvadi lnaobnil-idtyr,ug- molleickuel acromwepioguhtn,dpse r[m43e]a. bTihlietyr,etfooxriec,i tsyo,mpoel aorf stuhref apcreoapreerat,iems etthaabto wlisemre, acnodnsliipdoeprehdil iwcietyr.e Thhyedtroopg2e0n hbiotsnds, weresoelvuabluilaittye,d buiosianvgaFilraebeiAlitDyM, mEo-TleocxuFlialtre wrinegigThoto, lp(eFrAmFe-Dabruilgityse, rtvoexri)c[i4ty1,] .pTohlaerc osumrpfaocuen adrsewa,e mreestcaobroedlism, as “aacncde plitpedo”ph(liilgicaintyd.s Twhieth tnoop s2t0ru hctiutsr awl earleer tesvaanludastaetdis fuysiinnggt hFerepeh AysDicMocEh-eTmoixc aFliflitleterirn),g“ iTnoteorlm (FeAdiFa-teD”rug (ligasnerdvsewr)i t[h41lo].w T-rhiesk csotmrupcotuurnadl sa lewrtesrea nsdcofreewd pahsy “siaocccheepmteidca”l (plirgoapnedrtsi ews bitehl onwo thsteruthcrtuesrhalo ladl)eratnsd and “rejescatteidsf”y(ilnigga tnhdes pthhyatsidciodchneomt picaasls ftihlteerp)h, “yisnictoercmheemdiicaatel ”fi l(tleigrawnhdisc hwiinthcl ulodwe-arihskig sht-rruiscktusrtarul catluerrtasl aanledrt few andp/ohryesxicoecehdemthiecathl rpesrhoopledrtoief so cbcuelrorewn ctehoef ltohwre-srhisokldst)r uacntdur a“lraeljeercttse)d. ”T h(eligpahnydsisc otchhaetm dicidal fnilotet rpuassesd the waspDhryusgic-loikcheneemssic.aSlt rfuilctteurr awl haliecrht sianrcelumdoel eac uhlaigr hsu-rbisstkr uscttruurcetsuorralr eaalcetritv eagnrdo/uopr setxhcaeteadr etrheela ttehdretsohtohled of carcoincocugerrneicncaen domf ulotawge-rniisckp rsotrpuecrttiuersaol f tahleercths)e.m Tichael s panhdyspicoosechriesmksictaol cofimltpero uunsdesdw wheans uDserduign-ldikruengess. discSotvreurcytu[4ra4l] .aAlefrttesr athree mevoalleucautiloanr ,stuwbostcroumctpuoreusn odrs rweearcetisvceo rgerdou“pacsc tehpatet da”r,e trweolawteedr eto“ itnhtee rcmarecdiniaotge”enic (Tabalend5) manudta1g6enwiecr per“orpejeercttieeds ”o.f Sthome cehoefmthiceasltsr aunctdu praolsael erristsksth taot cwomerepoiduenndtisf iwedheinn musoesdt oinf tdhreu‘gre djeicstceodv’ery [44]. After the evaluation, two compounds were scored “accepted”, two were “intermediate” (Table 5) and 16 were “rejected”. Some of the structural alerts that were identified in most of the ‘rejected’ ligands included phenols, alpha (α) and beta (β)-unsaturated carboxylic acids, ketones, quinones and amides. The compounds that were considered “accepted” and “intermediate” were shortlisted as lead compounds (Table 5). Derek Nexus [41,42] provided an overall conclusion about the likelihood of lead compound toxicity by applying expert knowledge-based rules in toxicology (Table 6). Derek Nexus is a knowledge-based software that provides toxicity predictions in silico by identifying potentially toxic chemicals, which aids in the rejection of unsuitable drug candidates [45,46]. The pharmacological profiles of the lead compounds were then compared with that of five known Buruli ulcer drugs, namely Rifampicin, Streptomycin, Clarithromycin, Moxifloxacin, and Amikacin (Table 7). Lipinski’s violations, solubility and bioavailability of the compounds were computed. The leads violated none of the Lipinski’s rule of 5, whereas all the known drugs except Molecules 2019, 24, 2299 9 of 21 ligands included phenols, alpha (α) and beta (β)-unsaturated carboxylic acids, ketones, quinones and amides. The compounds that were considered “accepted” and “intermediate” were shortlisted as lead compounds (Table 5). Derek Nexus [41,42] provided an overall conclusion about the likelihood of lead compound toxicity by applying expert knowledge-based rules in toxicology (Table 6). Derek Nexus is a knowledge-based software that provides toxicity predictions in silico by identifying potentially toxic chemicals, which aids in the rejection of unsuitable drug candidates [45,46]. The pharmacological profiles of the lead compounds were then compared with that of five known Buruli ulcer drugs, namely Rifampicin, Streptomycin, Clarithromycin, Moxifloxacin, and Amikacin (Table 7). Lipinski’s violations, solubility and bioavailability of the compounds were computed. The leads violated none of the Lipinski’s rule of 5, whereas all the known drugs except moxifloxacin did not fully comply. The known drugs, however, had good solubility whilst the lead compounds had reduced solubility except ZINC000018185774. All the lead compounds, as well as rifampicin, moxifloxacin, and clarithromycin, had good bioavailability whilst the rest had low bioavailability. These results show some considerable difference between the properties of the known drugs and the lead compounds especially in the Lipinski’s rule of five. Table 5. Free ADME-Tox Filtering results showing ligands that passed the physicochemical filtering rules with no structural alerts. Ligands Status ZINC000005357841 Accepted ZINC000014417338 Accepted ZINC000018185774 Intermediate ZINC000095485921 Intermediate Table 6. Results obtained from Derek Nexus Software. The reasoning of outcome is indicated based on the confidence level of the predicted toxicological endpoint. The most common end point among the compounds was Skin sensitization which had a plausible outcome indicating a low confidence level. Ligands. End Point Species Reasoning Negative Strongest Ec3Outcome Outcome Prediction Hepatotoxicity Mammal Plausible - - Mutagenicity In Vitro Bacterium - Inactive - ZINC000005357841 Carcinogenicity Mammal Plausible - - Skin Sensitization Mammal Plausible - 2.9% ModerateSensitizer Teratogenicity Mammal Equivocal - - Mutagenicity In Vitro Bacterium - Inactive ZINC000018185774 Skin Sensitization Mammal Plausible - 0.15% StrongSensitizer Photoallergenicity Mammal Plausible - - Teratogenicity Mammal Equivocal - - ZINC000095485921 Mutagenicity In Vitro Bacterium - Inactive Skin Sensitization Mammal Plausible - 0.15% StrongSensitizer Mutagenicity In Vitro Bacterium - Inactive ZINC000014417338 Skin Sensitization Mammal Plausible - 0.16% StrongSensitizer Molecules 2019, 24, 2299 10 of 21 Table 7. Pharmacological profiles of the lead compounds and five known drugs for Buruli ulcer. The known drugs are Rifampicin, Streptomycin, Clarithromycin, Moxifloxacin, and Amikacin. Ligand ID Lipinski’s Solubility (mg L) Solubility Oral BioavailabilityViolation / Forecast Index (Veber) ZINC000014417338 0 3279.99 Reduced Solubility Good ZINC000005357841 0 4526.38 Reduced Solubility Good ZINC000018185774 0 8434.39 Good Solubility Good ZINC000095485921 0 2928.86 Reduced Solubility Good Moxifloxacin 0 30151.64 Good Solubility Good Amikacin 3 5077659.17 Good Solubility Low Streptomycin 3 3508974.65 Good Solubility Low Clarithromycin 2 1491.87 Good Solubility Good Rifampicin 4 246.01 Good Solubility Good 2.6. Molecular Dynamics (MD) Simulations A 100 ns MD simulation using GROMACS [47] was performed for the complexes of ideR and each lead compound and the results were compared to that of the unbound protein. This was done to investigate any influence on the structural conformation of the protein, due to the binding of the predicted lead compounds. Root Mean Square Deviation (RMSD) and RMS–Fluctuation (RMSF) graphs were generated after the simulations and the results of the complexes and the unbound ideR were compared (Figure 6). The RMSD graph accounts for the deviation of the atoms of the protein from the backbone of the protein and RMSF shows the movement of the protein residues during the simulation [48]. From the RMSD graph (Figure 6a), it was observed that most of the complexes experienced little fluctuations throughout the simulation with RMSDs close to that of the unbound protein. The RMSDs of ideR_l21, ideR_l38, ideR_l41 and unbound ideR structure fell between 0.3 nm and 0.4 nm from 20 ns, with ideR_l21 showing the closest RMSD fluctuations to that of the unbound protein. However, ideR_l74 endured very huge fluctuations in the first 60 ns of the simulation. The RMSD of ideR_l74 rose to 0.5 nm from 0–15 ns and again to 0.7 nm within the 55th–60th ns, where it remains stable till the end of the simulation. This indicates that the protein structure may have experienced conformational changes induced by the binding of ZINC000018185774. The RMSF graph (Figure 6b) affirms this observation, as huge fluctuations can be observed in the residues’ positions of the ideR_l74 complex. The fluctuations of the ideR_l74 peak the highest for residues within the DNA–binding domain of ideR (residues 1 to 75); the overall fluctuations with respect to that of the unbound ideR indicate high instability in the protein’s structure, due to its binding to ZINC000018185774. Some fluctuations were observed in the other complexes though not as high as that of ideR_l74. The complex of ideR_l38 showed some high fluctuations for residues 30–60 whilst ideR_l41 showed high fluctuations for residues between 100 and 150. On the other hand, ideR_l21 showed very high stability since it experienced minimal fluctuations as compared to the others and was close to that of the unbound ideR. Induced changes in the protein’s conformation by a ligand, particularly ZINC000018185774 can disrupt the protein’s iron acquisition system causing difficulty in the mycobacterium’s survival within its host. Molecules 2019, 24, 2299 11 of 21 Molecules 2019, 24, x FOR PEER REVIEW 11 of 21 FigFuirgeu6r.e R6o. oRtoMote ManeaSnq uSaqruea–rDee–vDieavtiioantio(Rn M(RSMDS) Dan) dan–dF lu–Fcltuucattuioantio(Rn M(RSMF)SgFr)a gprhaspohfs tohfe trhees preescptievcetive comcopmlepxelesxeasn adndg egneenreartaetdeda a mmooddeell ooff idideReR; (;a)( ar)eprreepsreenstesn tthse tRhMe SRDM gSrDaphg rvaeprhsusv etirmsues antidm (eb)a snhdows (b) tshheo RwMs SthFe pRerM reSsFidpueer grerasipdhuse. Ggrraapphhss .arGer raepphrsesaerneteredp irne sdeifnfteerdenitn cdoilffouerres natncdo ilnoduircsaatendd inin tdhiec aletegdend; in tbhlaeckle–gpernedd;ictbeldac km–opdreedl ioctfe didemRo, dreeldo–ifdiedre_Rl2,1 re(ZdI–NidCe0r_0l020195(4Z8I5N9C210-0id00e9R5 4c8o5m92p1le-ixd),e Rgrceoemn–pidleexr)_,l38 gre(eZnI–NidCe0r_0l03081(4Z4I1N73C3080-0id0e1R44 c1o7m33p8l-eidxe),R bcluome–pidleexr)_,l4b1lu (eZ–IiNdeCr_0l04010(0Z5I3N57C804010-0id0e5R35 c7o8m41p-ildexeR), caonmd pyleelxlo),w– andidyeerl_lol7w4– (iZdIeNr_Cl70400(Z01IN81C850707040-1i8d1e8R5 7c7o4m-ipdleeRx)c. omplex). 2.7. Induced Fit Docking 2.7. Induced Fit Docking Docking has become a widely accepted and standard method in computational drug discovery. It is, howDeovcekri,nligm hiatesd bbeycodmiffie cau wltiiedselliyk eacmceopdteeldin agnflde sxtiabnilditayrodf mtheethporodt einin cuopmopnultiagtainodnabli nddruingg d[i4s9c,o5v0e].ry. AsIstu icsh, ,hIonwduevceedr, Fliimt Ditoecdk binyg d(IiFffDic)uwltiaess elmikpe lmoyoeddeflionrgt hfelelxeiabdilictoym opf othuen dpsroctoeminp luepxoend wligitahndid ebRinodfing M.[u4l9c,e5r0a]n. sA, ss isnuccehI,F IDndmucoedde lFsitp Drooteckininflge (xIiFbDil)it wy ausp eomn plilgoaynedd bfoinr dthineg le[a5d1, 5co2]m. pIFoDunsdcos rceosmapnldexGeldid weith scoirdeesRw oefr eMo.b utalcinereadnsf,o sritnhcee cIoFmDp mleoxdese.lsT phreoItFeDins fcloerxeibs ielisttyim uaptoenth leigbaensdt pblianudsiinbgle [c5o1n,5f2o]r.m IFaDtio sncoorfetsh eand ligaGnldidceo mscpolreexs anwderteh eoGbltiadineesdco rfeosr gtihvee s caommepalseuxrees.o fTthhee bIiFnDd insgcoaresn ietystibmetawtee etnhea libgeasnt dpalnaudsaible ffi recceopntoforr[m53a,t5i4o]n. Aof mthoer elingeagnadt ivcoemvaplluexe iannbdo tthhet hGeliIdFeD sacnodreGs lgidiveessc oar emseraespurrees eonft tahem boirnedpinlagu saifbfilneity conbfeotrwmeaetnio an laignadndbe attnedr ba inredcienpgtoorf [t5h3e,5p4r]o. tAei nm–olirgea nndegcaotimvep lveaxlurees ipne cbtoivthe lyth. eZ IIFNDC a0n0d00 G95li4d8e5 9s2c1o,res ZINreCp0r0es0e0n1t8 1a8 57m7o4r,eZ IpNlaCu0s0ib0l0e1 4c4o1n7f3o3r8maantidonZ IaNnCd 00b0e0tt0e5r3 5b7i8n4d1incgo mopf lethxee s porbottaeiinne–dligGalnidde csocmorpelex of r−e7s.p6e0ctkivcaellym. oZlI,N−C70.00000k9c5a4l8m59/ / o2l1,, −Z6I.N75C0k0c0a0l 1818577/mol and4, −Z5I.N82C0k0c0a0l1m44o1l,73a3s8 / waenldl aZsINIFCD00s0c0o0r5e3s5o7f841 −47c1o.m96plkecxaelsm obotl,ai−n4e7d1 .G27lidkec aslcomreo lo, f− −477.16.00 5kckacla/ml molo, l−7a.n0d0 k−c4a6l4/m.79olk, c−a6l.7m5 oklc,arle/mspoelc atinvde l−y5. .T82h ekcLal/mol, / / / / ead comasp owuenlld assf oIrFmDe sdcoinrteesr aocft i−o4n7s1w.96it hkcraels/imduoel,s ,−s4u7c1h.2a7s kMcaelt/1m0,oGl, l−u417712.,0C5 yksc1a0l2/m, aonld aHndis 9−84,6w4.h79ic hkchaal/vme ol, respectively. The Lead compounds formed interactions with residues, such as Met10, Glu172, Cys102, and His98, which have already been predicted as essential residues of the metal binding site after the Molecules 2019, 24, 2299 12 of 21 Molecules 2019, 24, x FOR PEER REVIEW 12 of 21 Molecules 2019, 24, x FOR PEER REVIEW 12 of 21 already been predicted as essential residues of the metal binding site after the virtual screening. Tvhivreitruytauala lsl cosrcefroeerneminniegnd.g T.i nhTtheyer ayac latsilos onfo sfrowmrmiethde dTin hitnret1rea4rc,atAciotrinogsn3 3sw waitnhitd hT AhTrsh1pr411,47 A,. ArTgrh3ge33s 3ea nardnes dAi dAsupse1ps71.f7 oT. rhTmehsehesy erd ersreoisdgiudeeunse bsf ofrnomdrm ihnyhtedyrrdaorcgotiegonen nsb wobnoitdnh dit nhitenetlreeaarcadtcicotoinomsn psw owiuthint hdt hstaeht elme laeedata dcl obcmionmdpiopnuognusdnitsde sa2 ta( Ttm ambetleaetla8 )lb ,ibpnirdnoidvniignd gisn isgteiti en2 s 2i(g T(haTtbaslbienl et8 o)8,o )p,t hrpeorrvonivdoiidvneignl g rineisniisdgiuhgethst siwn ithnoit cooh tohctaehnre rnb onevofeuvlre trlh ereseirsdieudxeupsel osw iwtheidhc.ihcT hch acenab nbi neb defu ifnrugtrhtpehore sree xeapxnlpodliotieintdtee.d rT.a hTctehi oebn ibnmidnaidnpigno gpf opZsoIeNs eaC na0dn0 d0in0 i9tne5tr4ea8rca5tc9iot2ino1 n amrmeapash po owfo nf ZinZINIFNCigC0u00r00e00s907594a58n458d95829,12r 1e aspraer cet sihvseohlwoy,wnw nh inili snt FtihFgoiugsrueerose fsZ 7 I7N aCnad0n0 d0 80,18 8, r1e8rse5ps7ep7ce4tc,ivtZievIlNeylC,y ,0 w0w0h0ih1lsi4lt4s t1 t7h3toh3so8esa en odfo f ZZININCC0000000100851138851578785747417,4 aZ,r ZIeNIsNChCo00w000n0010i4n144F14i71g37u33r83e 8as naSdn3 dZa nZINdINCSC040.000T00h00e5035li53g75p87l48o14t 1ao raferZ esI hsNohCwow0n0 ni0n 0i n9F 5iFg4iug8r5ue9rs2e sS1 3Ss 3ha noadwn dSin 4Sg.4 T.t hTeh e ilnigltipegrlpoalcto toti ofo nZf sZINoINfCtC0h0e000l00ig90a59n458d458i95n291t2h 1se hsMohwoewtinailgnb gtih ntehd eiin itngetrseairtcaetci2otinoosfn stoh foe tfh mteho eldi gleiaglneadn diidn ie ntRh tehb eMe fMoerteaetlaI bFl iDbnidnisidnaignls gso istseiht e2o 2wo fon tfh itneh e Fmimgoudoredelee9lde. did iedReR b ebfeofroer eIF IDFD is i sa laslos os hsohwown nin i nF iFgiugruer e9 .9 . Figure 7. The induced fit pose of ZINC000095485921 (grey) in metal binding site 2 of the ideR of M. Figure 7. TThhee inindduucceedd fifit tppooses eofo fZZININCC000009059458458952912 1(g(rgerye)y i)ni mn emtaelt ablinbdinindgin sgites i2te o2f tohfet hideeiRd eoRf Mof. ulcerans model. Mulc.eurlacnesr amnsomdeold. el. Figure 8. Two–dimensional interaction map of ZINC000095485921–ideR complex obtained af te r IFD. HFyidgruorgee 8n. bTownod–sdairmeesnhsoiwonnailn invtieorlaecttliionne smwapit hofa rZrIoNwCh0e0a0d0s9.5485921–ideR complex obtained after IFD. Figure 8. Two–dimensional interaction map of ZINC000095485921–ideR complex obtained after IFD. Hydrogen bonds are shown in violet lines with arrow heads. Hydrogen bonds are shown in violet lines with arrow heads. Molecules 2019, 24, 2299 13 of 21 Molecules 2019, 24, x FOR PEER REVIEW 13 of 21 FFiigguurree 99.. LLiiggpplloott ooff ZZIINNCC000000009955448855992211s shhoowwininggt htheel igliagnadnsdsin itnetrearcatciotinosnps rper–eIF–IDF.DL.i gLaignadnids sish oswhonwinn yine lyloewlloawn danhdy dhryodgreongebno nbdosnadrse asrheo swhnowinng irne egnreberno kberonkleinne lsi.nRese.f ReretfoerF tiog uFrigeuSr1ef So1r tfhoer tlihgep lliogtpslooftst hoef oththe eorthleeard lecaodm cpoomupnodus.nds. 3. Screening of Lead Compounds and Known Inhibitors against the DNA-Binding Site 3. Screening of Lead Compounds and Known Inhibitors against the DNA-Binding Site Selected lead compounds comprising ZINC000018185774, ZINC000095485921, ZINC000014417338 Selected lead compounds comprising ZINC000018185774, ZINC000095485921, and ZINC000005357841 were screened against the DNA–binding site of the modeled ideR alongside ZINC000014417338 and ZINC000005357841 were screened against the DNA–binding site of the the five known potent inhibitors (Table 8). This screening was done to investigate the activity of the modeled ideR alongside the five known potent inhibitors (Table 8). This screening was done to chosen leads against the DNA–binding site of the modeled ideR in comparison to the potent inhibitors investigate the activity of the chosen leads against the DNA–binding site of the modeled ideR in discovered for the M. tuberculosis ideR’s DNA–binding site. This was possible since both proteins comparison to the potent inhibitors discovered for the M. tuberculosis ideR’s DNA–binding site. This are close homologues and showed high conservation at the DNA–binding domain. Binding energies was possible since both proteins are close hom−ologue−s and showed high conservation at the DNA–obtained for the lead compounds ranged from 5.7 to 5.9 kcal/mol, which is very close to those of the binding domain. Bind−ing en−ergies obtained for the lead compounds ranged from −5.7 to −5.9 five potent inhibitors ( 5.5 to 6.0 kcal/mol) (Table 8). The hydrogen bond interactions between the kcal/mol, which is very close to those of the five potent inh+ibitors (−5.5 to −6.0 kcal/mol) (Table 8). compounds and the protein were also analysed using LigPlot [55] to investigate common interacting The hydrogen bond interactions between the compounds and the protein were also analysed using residues the lead compounds might share with the five potent inhibitors. It was observed that the lead LigPlot+ [55] to investigate common interacting residues the lead compounds might share with the compounds and the potent inhibitors shared common hydrogen bond interactions with Arg47, Arg27, five potent inhibitors. It was observed that the lead compounds and the potent inhibitors shared Thr44 and Thr7. Ser37, Pro39, and Gln43, are essential residues within the DNA–binding site necessary common hydrogen bond interactions with Arg47, Arg27, Thr44 and Thr7. Ser37, Pro39, and Gln43, for binding to DNA [30]. Two of the potent inhibitors, namely NSC12453 and NSC65748 formed are essential residues within the DNA–binding site necessary for binding to DNA [30]. Two of the hydrogen bond interactions with Gln43 and Ser37, respectively. None of the lead compounds formed potent inhibitors, namely NSC12453 and NSC65748 formed hydrogen bond interactions with Gln43 hydrogen bond interactions with any of the three essential residues; however, ZINC000014417338 and and Ser37, respectively. None of the lead compounds formed hydrogen bond interactions with any ZINC000095485921 formed hydrogen bond interactions with Ser42, which has been shown to be a of the three essential residues; however, ZINC000014417338 and ZINC000095485921 formed novel critical residue which could be exploited for discovery of inhibitors against ideR [30]. hydrogen bond interactions with Ser42, which has been shown to be a novel critical residue which Induced Fit Docking was also performed for the binding of the lead compounds to the DNA–binding could be exploited for discovery of inhibitors against ideR [30]. site to account for protein flexibility. ZINC000018185774, ZINC000014417338, ZINC000095485921 and Induced Fit Docking was also performed for the −binding of the lead compounds to the DNA–ZINC000005357841 complexes obtained Glide score of 5.27 kcal/mol, −5.04 kcal/mol, −4.79 kcal/mol bindi−ng site to account for prote−in flexibility. Z−INC000018185774, − ZINC000014417338, and 4.99 kcal/mol with IFD scores of 468.50 kcal/mol, 467.32 kcal/mol, 466.87 kcal/mol and −ZINC000095485921 and ZINC000005357841 complexes obtained Glide score of −5.27 kcal/mol, −5.04 461.93 kcal/mol, respectively. The compounds formed interactions with common residues, such as kcal/mol, −4.79 kcal/mol and −4.99 kcal/mol with IFD scores of −468.50 kcal/mol, −467.32 kcal/mol, Ala28, Arg60, and Ser42. −466.87 kcal/mol and −461.93 kcal/mol, respectively. The compounds formed interactions with common residues, such as Ala28, Arg60, and Ser42. Molecules 2019, 24, 2299 14 of 21 Table 8. Binding energies and hydrogen bond interactions of selected lead compounds and the five potent inhibitors screened against metal binding site 2 and the DNA–binding site respectively. A hydrogen bond interaction was generated with LigPlot+ software. Ligand ID Metal Binding Site 2 DNA-Binding Site Binding Energy (KCAL MOL) Hydrogen Bonds Binding Energy (KCAL MOL) Hydrogen Bonds/ / NSC12453 −7.5 Gly176, Arg13, His98 −5.9 Gln43, Arg47, Thr7 NSC201773 −7.5 Gly176, His173, His98 −6 Arg27 NSC282699 −7.5 - −5.9 Arg47, Thr44 NSC303600 −7 Thr14, Arg13, Arg33 −5.9 Thr8, Thr7, Asn2 NSC65748 −7 Arg33, Asp17, His173, Arg13 −5.5 Ser37, Thr40, Gln36, Glu35 ZINC000014417338 −8 Arg33, Asp17, His98 −5.9 Ser42 ZINC000018185774 −7.7 Asp3, Arg103, Arg33, Asp17,Glu172 −5.8 Arg47, Thr44 ZINC000095485921 −7.6 Thr14, His98, Glu172 −5.7 Arg60, Ser42, Arg27, Ala28 ZINC000005357841 −7.4 His98 −5.9 Thr7, Asn2, Thr44 4. Exploring the Anti-Mycobacterial Activity of the Predicted Leads Due to the limited financial resources for Buruli ulcer drug discovery, repurposing of antimycobacterials by screening against M. ulcerans is gaining attention [56,57]. A library of compounds from the tuberculosis lead generation and optimization programs was screened in a whole-cell assay against M. ulcerans, where five compounds were discovered to be potent inhibitors with high activity (IC90 ≤ 1 µM) [56]. Therefore, exploring the plethora of antitubercular structures to unravel potential anti-M. ulcerans scaffolds was adopted. Analogues, derivatives or structurally similar compounds to the leads were investigated for possible anti-mycobacterial related activities. ZINC000018185774, also popularly known as luteolin, has been shown to exhibit anti-mycobacterial activity against M. tuberculosis via fractionations from crude samples of Annona sylvatic and Ficus chlamydocarpa with MIC values of 236.8 µg/mL and 78.12 µg/mL, respectively [58,59]. However, there was no report of the compound being tested against M. ulcerans. A structural similarity search performed in Drugbank revealed luteolin to be highly structurally similar to a flavonol compound, quercetin with a similarity score of 0.884. An analogue of quercetin, quercetin–3–O–β–d–glucoside has been reported to inhibit glutamine synthetase enzyme in M. tuberculosis (IC50 = 0.048 µM) [60]. Quercetin–3–O–β–d–glucoside (ZINC4096845) was casually docked against M. ulcerans of ideR and a high binding energy of −8.2 kcal/mol was obtained. The quercetin–3–O–β–d–glucoside docked firmly within the active site pocket of the M. ulcerans’ ideR. We, therefore, suggest that both quercetin–3–O–β–d–glucoside and luteolin could be investigated as potential novel anti-Buruli ulcer leads. ZINC000014417338, also popularly known as Alpinumisoflavone, has been reported to exhibit antibacterial and anti-mycobacterial activity with MIC value of 19.53 µg/mL against M. smegmatis [59]. Therefore, it is plausible to explore repurposing ZINC000014417338 as an anti-mycobacterial ulcerans. However, similarity search via Drugbank did not yield any structurally similar compound which exhibits anti-mycobacterial activity. This may be due to the stringent similarity threshold adopted for the query. Similarly, similarity searches for ZINC000095485921 (1,4,8-trihydroxy-5-(3-methylbut-2-enyl)xanthen-9-one) and ZINC000005357841 ((6-methoxybenzo [1,3]dioxol-5-yl)BLAHone) yielded no structure which has shown anti-mycobacterial activity. Even though, we could not find any report describing the anti-mycobacterial activity of ZINC000095485921, some novel 1,2,3-triazolyl xanthenones were reported to have shown good to excellent antimicrobial and anti-tubercular activity with MIC values from 3.12–6.25 µg/mL [61]. It is worth exploring ZINC000095485921 as a potential anti-mycobacterial lead, since it is also a xanthenone analogue. Due to the fact that no similar compounds or analogues were found for ZINC000005357841, the possible biological activity of the lead compound was predicted with Prediction of Activity Spectra for Substances (PASS) [62,63] and their Probable activity (Pa), as well as Probable inactivity (Pi) values were obtained. Among the results retrieved from PASS, the ones most relevant to anti-buruli Molecules 2019, 24, x FOR PEER REVIEW 15 of 21 Molecules 2019, 24, x FOR PEER REVIEW 15 of 21 aMnotleic-utulebs 2e0r1c9u, l2a4r, x aFOctRiv PiEtyER wREiVthIE WM IC values from 3.12–6.25 μg/mL [61]. It is worth exp1l5o roifn 2g1 ZananIN ti-tub ti-Ctu0b00 e er0 rc9ular activity with MIC values from 3.12–6.25 μg/mL cu5l4a8r5 9a2c1t iavsi tay pwotietnht iaMl aICn ti-vmalyuceosb afcrtoemri al3 l.e1a2d–6, .s2in5 ceμ igt /ims Lal so [61]. [6 a1 ]x. an It Itt is hiesn o wnowoe rathrthn a leoxexg puleo.r i ng ZINCD0u0e0 0to9 5t4h8e5 f9a2c1t tahsa at pnoot seinmtiialal ra ncotim-mpyocuonbdasc toerr iaanl aleloagdu, essin wcee rite ifso aulnsod afo xra ZnItNheCn0o0n0e analog puloe.r i ng ZINCD0u0e0 0t9o5 t4h8e5 f9a2c1t aths aat pnoot esinmtiialla ra ncotim-mpyocuonbdasc toerr iaanl aleloagdu, esisn wcee rite ifso aulnsod af oxra ZntIhNeCn0o0n0e 0 a0n53a5lo7g8u41e,. t005357841, t he possiDbluee b tioo logic he possible bioltohgei fa acl activity of the lead compound was predicted with Prediction of Activity Spectra for Substances (cPaAl ta tchtaivt intyo osifm thilea rle caodm cpoomupnodusn odr awnaasl opgruedesic wteedr ew fiothu nPdr efdoric ZtiIoNnC o0f 0A00ct0i5v3it5y7 8S4p1e,c tthrae pfoors sSiubbles tbainocloegs i(cPaAl a ScSt)i v[i6ty2, 6o3f ]t haen dle athde ciro mPrpoobuanbdle w aacstivity (Pa), as well as Probable inactivity (Pi) values were obtaineSdS. )A [m62o,6n3g] tand predicted with Pr he r etshuelitrs Prertorbieavbelde farcotmiv iPtyA (SPSa, )t,h aes o wneesl l eadsi cPtiroonb aobf lAe citnivaicttyiv Sitpye c(tPria) fvMoaorll eucSeuulseb sws2t0ea1rn9e,c 2oe4sb, 2t(a2P9in9AeSdS.) A [m62o,63] and their Probable activity (Pa), as well m aos sPt rroelbeavbalnet itnoa acntitvii-tbyu r(uPlii) uvlacleure sa cwtievriety o batnadin ierdon. A amcqou nigsi ttihoen r mesults retng the reseuclhtsa rneits rmiesv ewde from PASS, the ones most relevant to anti-b15uorfu2l1i ulcer activity and iron acquisition mechanisrmiesv ewde f r rr e eo an man tPi-Ainfti-inSfS la l,a mt mamhme oan toersy (Pa =tory m (Posat =r e0le.4v5a8n, tP tio = 0 0.458, Pi =a n0t .0i-7b)u arnudli aunlctei-ro xidative (P .07) and anti-oaxcidtiavtiitvye a(nd a =ir 0o.n30 a5c, qPui i=si 0ti.022) actPa = 0.305, Pi = 0.o0n2 2m) aeccth ivaintiiesms. sT hwee Prea avnatliu-iensf wlamerme gartoearyte r(P tha a=n 0t.h4o5s8e, oPfi P=i 0o.0b7ta) inanedd fauonlrct ei-rbooaxtcihdti avtaiitfvyoer ae(nPmdae =nir to0io.n3n0ae5cd,q Puibi si=oi t0lio.o0gnicmale cah icavities. The Pa values were grea22) activtniivtiiisetmsie. ssT,w hepe orPeiana tvninatlgiu- ienosfl uwatm ertmhe ega troneraey teeted( r than rP tahtoa= nf0 uth t.h4r o ot5h s s8e e,r Po ife Pi obtained for both aforementioned bi e of =Pxip 0ol.o0br7tea) iantnhedde pfaonhrta ir-bomoxatidhcoa ltaoivfgoeirc(eaPmla ae=cnt0tiiv.o3i0nty5e ,doP fi Zb=iI oN0l.o0Cg202i0c)0aa0lc 0ta5ivc3it5tii7ve8ist4.i1eT s[h,6 e2pP].oa invtailnuge s owuetr etghree anteeretdh atnot hfousrethoefrP ieoxbptaloinree dtfhoer pharm ological bothTahfaoecr oeremloseugnilcttsaio lf nraoecmdtivb tiihtoyel o osgtfri cZuaIclNtauCcrta0ilv0 si0ti0ime0 a5c3t5iv78it4ie1s [,6 2p]o. inting out the need to further explore the pharmacological activity of ZINC00000s5i,l3ap5ro7iti8ny4 ts1ine [ag6r2oc]hu. et sth aennde tehde teonfruicrthheedr ienxfpolromreatthioenp ohbatraminaecdol aobgoicuatl tahceti lveTTiat h hyd e eso re rwfeZa s sI ults fr urNrlatCsn 0ftr 0e o0xm0p0 et5rh3iem5 7se8tnr4ut1acl[t 6ue2vr]aa.ll usiamtioilna roitfy t hseeiarr achnetis- Baunrdu tlhi ue lecnerri acchteivdi tiny.f oTrhmisa stitound yo bctoaminpelde mabout the leads warrant eoxmp etrhiem setnrutaclt uevraall usiamtiiolnar oitfy t hseeairr cahnetis- Banudru tlhi eu lecnerri cahcteidv iitnyf. oTrhmisa tsitound yo bctoaminpelde mab eonuts entst cthuer r currl eeTnahtd eesfrfeosrutslt gsefent e fwfoarrtrant ar eorxempde ttrhoiemwsaetrnrdutascl tu uenvraraalluvsieamltliiionlnagr o ittfhy tehs meeairer ccahhnaetnis-iBasnumdr uotlfhi aeucletcnieorrni acshc oteifdv piitonytf.e oTnrhtmiiasal st Bitouundruyol bci touamlicnpeerlde dmarbueongutsst. Scthuuermrleemnaatd resyfwf ooafr rtlr sea agndet aceroxempde ptrooimwuneadnrdtsa sil sue snvhraoaluvwaentlli ioinnng To tafhbtelh eme 9ire. cahnatin-iBsumr uolfi aucltcieornasc otifv pitoyt.ential Buruli ulcer drugs. Summary of lse agdea croemd ptoowunarddss i su nshravelling the mechanism of actions of pote Tnhtiiasls Btuudruylci oumplements current own in Table 9. lcer drugs. Summareyff oofr tlseagdea croemd ptoowunarddss isu snhraovwenll iinng Ttahbelem 9e. chanism of actions of potential Buruli ulcer drugs. SummTaabrlye o9f. lLeiastd ocfo pmrepdoicutnedd sleiasds hcoomwpnoiunndTas,b tlhee9ir. common names and two–dimensional structures MolecuoTTlbe astb a2li0e 9ablen1 e99d, .2 fL4r,i osxmt F oO ZfR ipn PrcEe dEdRaict RtaebEdaV slIeeE.aW d compounds, their common names and two–dimensional structures obtained. Lfriostm o Zf ipnrce ddiacttaebda sleea. d compounds, their common names and two–dimensional structur 1e5s of 21 TobabtaLlieing9ead.n Ldfri osIDtmo Zf ipnrce ddiacttaebdasleea. d cCoommpmouonnd Ns,amtheesir common names aTnwdot-wDoim–deinmsieonnsaiol nSatrlusctrtuucretu res anti-toubbtaeLiringceuadnlafdrr o ImDa cZtiivncitdya twabiatshe . MCICo mvmaolune Ns amfroesm 3.12–6.25 μg/TmwLo -D[6i1m].e nIsti oinsa l wStorurtchtu reex ploring ZINC00L00ig9a5n4d8 5ID92 1 as a potential Canommon Alpinumisoftlia-voynec;o Nbame 5-ahcytde sr roiaxly -l7e-a(d4-, since it Tisw aol-sDo iam xeannsitohneanlo Snteru acntuarleo gue. DueL tiog atnhde IfDact that nAol psiinmuimla Cohyirsd ocrfool mamvmopononeu; Nxyphenn5d- ahmesyl)sy- 2od,r2r o-axnya-7lo-(g4u- es were T fwouo-nDdi mfoern ZsiIoNnCal0S0t0ru00ct5u3r5e7841, the possZibINleC b0i0o0l0o1g4i4c1a7l3 a38c tivitAyd lipomifn etuthhmehy iyllspedoyarfroldaax vncyooopn[mh3ee,;2p n5-oyg-hul])cy-nh2ddr,r2o o-wmxyae-ns7 --p6(-4r-edicted with Prediction of Activity Spectra Alpinumisoflavone; for SZuINC0ZINbsCt0a 0n0c0e1s4 417338 hydroxyphenyl)-2,2-000144(1P7A33S8S ) [62d,6im3]e tahnydlp ytrhaenior [n3Pe,r 2o-gb]achbrleo maecnti-v6i-ty (Pa), as well as Probable inactivity (Pi) valueZsI NwCe0re00014417338 5-hydroxy-7-(4-hydroxyphenyl)- obtained. Amdoinmge tthhyel preysruano[3,2-g]chromen-6-2,2-dimeltthosyn rleep tyrriaenvoe[d3, one 2 fr-om PASS, the ones most relevant to anti-bur uli ulcer activity and iron acquisitiong ]mchercohmaennis-6m-osn ewere anti-inflammatory (Pa = 0.458, Pi = 0.07) a nd anti-oxidative (Pa = 0.305, Pi = 0.022) activities. The Pa values were greater than those of Pi obtain ed for both aforementioned biological activities, pointing out the need to further explore the (6-methoxybenzo[1,3]dioxol-5- pharZmINaCco0l0o0g00ic5a3l5 a78c4ti1v ity of (Z6-ImNeCth0o0yx0ly0)bB0e5n3z5o7[814,31] d[6io2x]. ZZININCC000000000055335578 LAH ne 784411 (6(-6m-metehtohxoyxbyebneznoz[o1[,13,]3d]idoixo oxlo-5l-- ZTINheC r0e0s0u00lt5s3 5fr7o8m41 the structural 5ys-ily)mBl)LBilALaArHitHoyno sene earchole-s5 -and the enriched information obtained about the leads warrant experimental evaluyla)BtiLoAn Hofo nthee ir anti-Buruli ulcer activity. This study complements current efforts geared towards unravelling the mechanism of actions of potential Buruli ulcer dru gs. Summary of lead compounds is shown in Table 9. Table 9. List of predicted lead compLouutneodlsi,n t;heir common names and two–dimensional structures obtained from Zinc datLaubtaesolin; 2-(3,4-Dihydroxy-phenyl)-ZZININCC00000118811885577744 Lut5e,7o e2l. -di-n(i3;h ,2y4--d(D3ro,i4h-yDdirhoyxdyr-opxhye-npyhle)-nyl)-ZINC000018185774 Lute5o,7li-nd;i h2y-(d3r,4o xxyy--Di- chromen-4-one Ligand ID h chyrdormoxeyn--p4h-oeneyl)- ZINC000018185 774 5,7-dihCyodmromxyo-nc hNroameens -4-one Two-Dimensional Structure 5,7-dihydroxy-chromen-4-one Alpinumisoflavone; 5-hydroxy-7-(4- hydroxyphenyl)-2,2- ZINC000014417338 ZINC000095485921 di1m,4e,8th-tyrliphyyrdar noox[y3-,52--(g3]-cmherothmyelbnu-6t-- 2-enyl)xaonnteh en-9-one 1,4,8-trihydroxy-5-(3-methylbut-2- ZINC000095485921 1,4,8-triehnyydlr)xoaxnyt-h5-e(n3--9m-oenthey lbut-2- 5. MZZa INC ItNeCri 0a0ls00a900009n 5 54d 48M59218592e1th ods1,4,8-triehnyydlr)oxxayn-t5h-e(n3--m9-oetnhey lbut-2- enyl)xanthen-9-one 5.1. ZHIoNmCo0l0o0g0y0M53o5d78el4i1n g of M(y6c-ombeatchteorxiyubmenuzloc[e1r,a3n]ds iiodxeoRl-S5-tructure yl)BLAHone The ideR of M. ulcerans protein had no experimentally solved structure in any of the prot ein databases, including the Protein Data Bank (PDB). Therefore, the 3D structure was generated using in5 5si . .l M Mico atheoriaterim alos and Methods alsl oagnyd mMoedtheloidngs . The amino acid sequence of M. ulcerans (strain Agy99) with UniProt ID 5A5..1 0M.P HTaot6em6roiwalolagss yau nMsdeo ddMetleointhagco oqdfu sMi ryechoboamctoerloiugmu euslcaesratnems ipdelRat eSstrfuocrtumreo deling via the Basic Local Alignment 5S.e1a. rHchomTooloogl y( BMLoAdSelTin)g[ 1o9f ]M. yAcobsaucittearbiulemt uelmcepralantse idweRit hStarugcotuorde E-value and high sequence identity w5.1a.s HsTeohmle ociltdoegedyR f Mrofom dMeal.i mnuglo conefrg LMaunthytse ceopolbrienoatc;t rte2iei-ern(vi3 u,eh4md-aD duhi hloncymeodr aroeonlxxospy gie-dupreehiRmse nSaeytnsrlt)u-aclttleuymr es pollvateed fostrrumcotudreel iinng atnhye opfr othteei nptraortgeient. dTahtaZbTIaNC0000e strhsueec sti,ud ireneRc1l 8uo1f8 5M77. 4 w ulcerans protein had no experimentally solved structure in any of the protein databTahsee si,d ienRcl uo adf siMnmg. o utdhlceel erPadrnou5st ,se7pii-nrdog iDhtMeyaidntoar do hBxeayaldl-nec rkhn rvo(oPe merDsxeiBpno)-en4.r -Ti9om.hn1ee7 refmorbee,d tdhed 3Din sEtarusycMtuored ewllaesr g4e.0n[e2r1a]t.ed using ding the Protein Data Bank (PDB). The nretafollrye ,s tohlev e3dD s sttrruuccttuurree iwn aasn gye onfe rtahtee dp ruostiening d 5.a2t.aSbtarsuecst,u irnecVluadlidinagti otnhe Protein Data Bank (PDB). Therefore, the 3D structure was generated using The quality of the generated model was assessed using a Ramachandran plot and further valida ted with WHATCHECK and PROCHECK [21,22]. 1,4,8-trihydroxy-5-(3-methylbut-2- ZINC000095485921 enyl)xanthen-9-one 5. Materials and Methods 5.1. Homology Modeling of Mycobacterium ulcerans ideR Structure The ideR of M. ulcerans protein had no experimentally solved structure in any of the protein databases, including the Protein Data Bank (PDB). Therefore, the 3D structure was generated using Molecules 2019, 24, 2299 16 of 21 5.3. Binding Site (Pocket) Identification After the quality of the model has been assessed to be reasonably accurate, putative binding sites were identified using KVFINDER, COFACTOR and COACH [23–25]. Predicted binding sites were also compared with the binding sites of the co-crystallised template [29] since proteins with similar folds are normally found to have similar binding sites. 5.4. Virtual Screening Two stages of virtual screening were carried out for lead discovery. In the first stage, a library composed of 832 compounds retrieved from AfroDb [31] and five potent inhibitors of ideR of M. tuberculosis recently discovered [30], retrieved from the database of the National Cancer Institute (NCI), (Bethesda, MD, USA) were screened against the predicted binding site of the ideR. The compounds were energy minimized with OpenBabel in PyRx using the universal force field (uff) and conjugate gradients as the optimization algorithm with a total number of steps of 200. Virtual screening against the ideR model was done using AutoDock Vina embedded in PyRx version 0.8 [64] using grid box size of 24.3, 27.7, and 20.9 Å; as well as center dimensions of 4.5, 47.2, and −2.8 Å in the X, Y and Z coordinate axes, respectively. The second stage of virtual screening involved a library of identified potential leads and the five potent inhibitors. They were screened against the DNA–binding site of the modeled ideR with a grid box size of 25.0, 25.0, and 25.0 Å, as well as center dimensions of 13.11, 64.95, and −3.09 Å. An exhaustiveness of 8 was used for both screenings. 5.5. Validation of Docking Protocol To validate the docking protocol used, a receiver operating characteristic (ROC) curve was generated by screening 34 actives [30] and their respective decoys against the Metal binding site 2 and DNA–binding sites of the modeled ideR of the M. ulcerans and the ideR of the M. tuberculosis (PDB ID: 1FX7) (Supplementary file S1). The actives are composed of the five potent inhibitors, as well as 12 other compounds and analogs of some of the highly performing compounds screened against the ideR of M. tuberculosis [30]. The decoys were generated with the Directory of Useful Decoys–Enhanced (DUD-E) [65]. A total of 1689 decoys were used together with the 34 actives for the screening. The docking results were used to generate the ROC curves and AUC values utilizing ROCKER [66] and screen explorer [67]. BEDROC values with an alpha of 20.0 and EFs at 1%, 10% and 20% were also evaluated. 5.6. In Silico ADMET Studies FAF-Drug [41] and DEREK NEXUS version 2.1 [41,42] were used for ADMET analysis and ADME/tox elimination. The ligands were uploaded as Simplified Molecular Input Line Entry System (SMILES) or Structure Data File (SDF) and were scored as “accepted”, “intermediate” and “rejected”. The physicochemical filter used was Drug-likeness. DEREK NEXUS was used to further evaluate the toxicity profiles of the “accepted” and “intermediate” ligands obtained from FAF-Drug. The pharmacological profiles of the chosen leads were also compared to that of five known drugs. 5.7. Molecular Dynamic Simulations GROningen MAchine for Chemical Simulation (GROMACS) version 5.1.1 [47] was used to perform the molecular dynamics simulations using the GROMOS96 43A1 force field. The ligands’ topology was, however, generated using PRODRG [68] since its topology could not be built in GROMACS. In running the simulation, the complex was first solvated in a 1 nm dodecahedron water box. The system was neutralized by adding 10 positive ions to balance the net charge of the complexes. The complex was then relaxed through energy minimization to remove any steric clashes or bad geometry. Thereafter, the system was equilibrated to the required temperature (300 K) and density (1020 kg/m3). After the system was equilibrated and set in the desired temperature and density, a 100 ns production run Molecules 2019, 24, 2299 17 of 21 was then performed, and the results of the simulation were analysed using Xmgrace version 5.1.25. The unbound protein was also subjected to molecular dynamics using the same simulation parameters as those of the complexes with the OPLS force field. 5.8. Induced Fit Docking Induced Fit Docking (IFD) of the lead compounds was done using the Schrödinger software suite and the GlideScores and IFD scores were generated for all the plausible poses. This was done to understand the flexibility of the modeled ideR since most docking programs dock a flexible ligand to rigid receptors [49]. The lead compounds were docked against the metal binding site 2 and the DNA–binding domain respectively. 5.9. Lead Structural Similarity Searches and Antimycobacterial Association The popular names of predicted leads were retrieved from the ZINC database to investigate whether the compounds have been reported elsewhere to have shown anti-mycobacterial activity. This was undertaken to find reports of a possible relationship between the leads and mycobacteria. Also, SMILES files of the leads were used to retrieve structurally similar compounds present in the Drugbank [69]. The search was done with a similarity threshold of 0.7 and molecular size ranged between 100 Da and 500 Da. This was done to identify compounds with relatively high structural similarity to the queried leads, which may have known anti-mycobacterial activity and will further support the possibility of the predicted leads to exhibiting anti-mycobacterial activity. For further investigation of ZINC000005357841 since no report or analogue with anti-mycobacterial activity was found for it, ZINC000005357841 SMILES files were uploaded to the Prediction of Activity Spectra for Substances (PASS) online tool [63] to predict its biological activity. 6. Conclusions African natural compounds ZINC000018185774, ZINC000095485921, ZINC000014417338 and ZINC000005357841 were identified as potential novel leads against the modeled structure of ideR of M. ulcerans by docking, which was validated with an AUC value of the ROC curve above 0.7. Five potent inhibitors of ideR of M. tuberculosis had similar binding energies as the leads when screened against the DNA–binding domain of the ideR of M. ulcerans. Novel critical residues of metal binding site 2 comprising Thr14, Arg33 and Asp17 were predicted. A hundred nanoseconds molecular dynamics simulations showed conformational changes in the ideR-ZINC000018185774 complex with implications in iron acquisition. Interestingly, quercetin, which is structurally similar to ZINC000018185774 was previously shown to exhibit antimycobacterial activity. Similarly, ZINC000014417338 (Alpinumisoflavone) was reported to exhibit anti-mycobacterial activity, whilst analogues of ZINC000095485921 have shown antimicrobial and antitubercular activity. ZINC000005357841 was predicted to possess anti-oxidant and anti-inflammatory activities. Since this work is largely computational, experimental confirmation of the anti-Buruli ulcer activity of the leads is critical. Furthermore, the scaffolds of the leads could be used for designing novel inhibitors. Supplementary Materials: The following are available online at http://www.mdpi.com/1420-3049/24/12/2299/s1, Figure S1: Ligplots showing the residue interactions of the lead compounds docked in the metal binding site 2 of the modeled ideR for M. ulcerans. Figure S2: Induced fit docking poses of the ligands docked in the metal binding site 2 of the modeled ideR for M. ulcerans. Figure S3: Interaction maps (I–map) of lead compounds docked in the metal binding site 2 of the modeled ideR for M. ulcerans obtained after IFD. Table S1: Summary of successfully generated models with DOPE scores. Supplementary File S1: List of actives and decoys used for validation of docking method and screening results. Author Contributions: S.K.K., L.M., W.M. and M.D.W. conceptualized the research project. Computational and data analysis was predominantly undertaken by S.K.K., M.A., K.S.E., J.A.Y., L.A.A. and M.D.W.; with inputs from K.D., E.K.T. and W.A.M. W.A.M.III contributed to the structural modeling and molecular dynamics simulations. S.K.K. and K.S.E. co-wrote the first draft. All authors contributed to the revision of the drafts and agreed on the final version of the manuscript prior to submission. Molecules 2019, 24, 2299 18 of 21 Funding: This work was supported by funds from the World Bank African Centres of Excellence grant (ACE02-WACCBIP: Awandare). Acknowledgments: The authors express their profound gratitude to the West African Center for Cell Biology of Infectious Pathogens, University of Ghana, for the use of Zuputo, the high performance computing systems. Also, we thank the entire faculty members of the Department of Biomedical Engineering, University of Ghana, for all the useful advice on the research. Conflicts of Interest: The authors declare no conflict of interest. References 1. Nyabadza, F.; Bonyah, E. On the transmission dynamics of Buruli ulcer in Ghana: Insights through a mathematical model. BMC Res. Notes 2015, 8, 656. [CrossRef] [PubMed] 2. Hall, B.S.; Hill, K.; McKenna, M.; Ogbechi, J.; High, S.; Willis, A.E.; Simmonds, R.E. The Pathogenic Mechanism of the Mycobacterium ulcerans Virulence Factor, Mycolactone, Depends on Blockade of Protein Translocation into the ER. PLoS Pathog. 2014, 10, e1004061. [CrossRef] [PubMed] 3. Williamson, H.R.; Benbow, M.E.; Nguyen, K.D.; Beachboard, D.C.; Kimbirauskas, R.K.; McIntosh, M.D.; Quaye, C.; Ampadu, E.O.; Boakye, D.; Merritt, R.W.; et al. Distribution of Mycobacterium ulcerans in Buruli Ulcer Endemic and Non-Endemic Aquatic Sites in Ghana. PLoS Negl. Trop. Dis. 2008, 2, e205. [CrossRef] [PubMed] 4. CDC. CDC—Neglected Tropical Diseases—Diseases. Available online: http://www.cdc.gov/globalhealth/ ntd/diseases/index.html (accessed on 18 January 2018). 5. WHO. WHO|Buruli Ulcer. Available online: http://www.who.int/buruli/en/ (accessed on 17 February 2017). 6. Ampah, K.A.; Asare, P.; Binnah, D.D.-G.; Maccaulley, S.; Opare, W.; Röltgen, K.; Pluschke, G.; Yeboah-Manu, D. Burden and Historical Trend of Buruli Ulcer Prevalence in Selected Communities along the Offin River of Ghana. PLoS Negl. Trop. Dis. 2016, 10, e0004603. [CrossRef] [PubMed] 7. Anand, U.; Sinisi, M.; Fox, M.; MacQuillan, A.; Quick, T.; Korchev, Y.; Bountra, C.; McCarthy, T.; Anand, P. Mycolactone-mediated neurite degeneration and functional effects in cultured human and rat DRG neurons: Mechanisms underlying hypoalgesia in Buruli ulcer. Mol. Pain 2016, 12, 174480691665414. [CrossRef] [PubMed] 8. Garchitorena, A.; Guégan, J.-F.; Léger, L.; Eyangoh, S.; Marsollier, L.; Roche, B. Mycobacterium ulcerans dynamics in aquatic ecosystems are driven by a complex interplay of abiotic and biotic factors. eLife 2015, 4, e07616. [CrossRef] 9. Amofah, G.; Bonsu, F.; Tetteh, C.; Okrah, J.; Asamoa, K.; Asiedu, K.; Addy, J. Buruli ulcer in Ghana: Results of a national case search. Emerg. Infect. Dis. 2002, 8, 167–170. [CrossRef] 10. Mosi, L.; Williamson, H.; Wallace, J.R.; Merritt, R.W.; Small, P.L.C. Persistent association of Mycobacterium ulcerans with West African predaceous insects of the family belostomatidae. Appl. Environ. Microbiol. 2008, 74, 7036–7042. [CrossRef] 11. Azumah, B.K.; Addo, P.G.; Dodoo, A.; Awandare, G.; Mosi, L.; Boakye, D.A.; Wilson, M.D. Experimental demonstration of the possible role of Acanthamoeba polyphaga in the infection and disease progression in Buruli Ulcer (BU) using ICR mice. PLoS ONE 2017, 12, e0172843. [CrossRef] 12. Bieri, R.; Pluschke, G.; Huber, S.; Li, J.; Scherr, N.; Bomio, C.; Hug, M.N.; Gersbach, P.; Altmann, K.-H.; Dangy, J.-P. Antibody-Mediated Neutralization of the Exotoxin Mycolactone, the Main Virulence Factor Produced by Mycobacterium ulcerans. PLoS Negl. Trop. Dis. 2016, 10, e0004808. 13. Adusumilli, S. Understanding Immune Response in Mycobacterium ulcerans Infection. Ph.D. Thesis, University of Tennessee, Knoxville, TN, USA, 2005. Available online: https://trace.tennessee.edu/utk_ graddiss/656 (accessed on 1 April 2019). 14. Kwofie, S.; Dankwa, B.; Enninful, K.; Adobor, C.; Broni, E.; Ntiamoah, A.; Wilson, M. Molecular Docking and Dynamics Simulation Studies Predict Munc18b as a Target of Mycolactone: A Plausible Mechanism for Granule Exocytosis Impairment in Buruli Ulcer Pathogenesis. Toxins 2019, 11, 181. [CrossRef] [PubMed] 15. Ghosh, S.; Chandra, N.; Vishveshwara, S. Mechanism of Iron-Dependent Repressor (IdeR) Activation and DNA Binding: A Molecular Dynamics and Protein Structure Network Study. PLoS Comput. Biol. 2015, 11, e1004500. [CrossRef] [PubMed] Molecules 2019, 24, 2299 19 of 21 16. Ranjan, S.; Yellaboina, S.; Ranjan, A. IdeR in mycobacteria: From target recognition to physiological function. Crit. Rev. Microbiol. 2006, 32, 69–75. [CrossRef] [PubMed] 17. Deshayes, C.; Angala, S.K.; Marion, E.; Brandli, I.; Babonneau, J.; Preisser, L.; Eyangoh, S.; Delneste, Y.; Legras, P.; De Chastellier, C.; et al. Regulation of Mycolactone, the Mycobacterium ulcerans Toxin, Depends on Nutrient Source. PLoS Negl. Trop. Dis. 2013, 7, e2502. [CrossRef] [PubMed] 18. Siddiqui, A.A.; Iram, F.; Siddiqui, S.; Sahu, K. Role of natural products in drug discovery process. Int. J. Drug Dev. Res. 2014, 6, 172–204. 19. Altschul, S. Basic Local Alignment Search Tool. J. Mol. Biol. 1990, 215, 403–410. [CrossRef] 20. Moorhouse, M.; Barry, P. The Protein Databank. In Bioinformatics Biocomputing and Perl; John Wiley & Sons, Ltd.: Chichester, UK, 2005; pp. 173–210. 21. Kuntal, B.K.; Aparoy, P.; Reddanna, P. EasyModeller: A graphical interface to MODELLER. BMC Res. Notes 2010, 3, 226. [CrossRef] 22. Shen, M.; Sali, A. Statistical potential for assessment and prediction of protein structures. Protein Sci. 2006, 15, 2507–2524. [CrossRef] 23. Vriend, G. WHAT IF: A molecular modeling and drug design program. J. Mol. Graph. 1990, 8, 52–56. [CrossRef] 24. Laskowski, R.A.; MacArthur, M.W.; Moss, D.S.; Thornton, J.M. PROCHECK: A program to check the stereochemical quality of protein structures. J. Appl. Crystallogr. 2002, 26, 283–291. [CrossRef] 25. Yang, J.; Roy, A.; Zhang, Y. Protein-ligand binding site recognition using complementary binding-specific substructure comparison and sequence profile alignment. Bioinformatics 2013, 29, 2588–2595. [CrossRef] [PubMed] 26. Roy, A.; Yang, J.; Zhang, Y. COFACTOR: An accurate comparative algorithm for structure-based protein function annotation. Nucleic Acids Res. 2012, 40, W471–W477. [CrossRef] 27. Yang, J.; Roy, A.; Zhang, Y. BioLiP: A semi-manually curated database for biologically relevant ligand-protein interactions. Nucleic Acids Res. 2013, 41, D1096–D1103. [CrossRef] [PubMed] 28. Zhang, Y.; Skolnick, J. Scoring function for automated assessment of protein structure template quality. Proteins Struct. Funct. Genet. 2004, 57, 702–710. [CrossRef] [PubMed] 29. Goranson-Siekierke, J.; Ingason, B.P.; Holmes, R.K.; Hol, W.G.J.; Feese, M.D. Crystal Structure of the Iron-dependent Regulator from Mycobacterium tuberculosis at 2.0-Å Resolution Reveals the Src Homology Domain 3-like Fold and Metal Binding Function of the Third Domain. J. Biol. Chem. 2002, 276, 5959–5966. 30. Rohilla, A.; Khare, G.; Tyagi, A.K. Virtual Screening, pharmacophore development and structure based similarity search to identify inhibitors against IdeR, a transcription factor of Mycobacterium tuberculosis. Sci. Rep. 2017, 7, 4653. [CrossRef] 31. Ntie-Kang, F.; Zofou, D.; Babiaka, S.B.; Meudom, R.; Scharfe, M.; Lifongo, L.L.; Mbah, J.A.; Mbaze, L.M.A.; Sippl, W.; Efange, S.M.N. AfroDb: A select highly potent and diverse natural product library from African medicinal plants. PLoS ONE 2013, 8, e78085. [CrossRef] 32. Lipinski, C.A.; Lombardo, F.; Dominy, B.W.; Feeney, P.J. Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Adv. Drug Deliv. Rev. 2012, 64, 4–17. [CrossRef] 33. Veber, D.F.; Johnson, S.R.; Cheng, H.Y.; Smith, B.R.; Ward, K.W.; Kopple, K.D. Molecular properties that influence the oral bioavailability of drug candidates. J. Med. Chem. 2002, 45, 2615–2623. [CrossRef] 34. Truchon, J.-F.; Bayly, C.I. Evaluating Virtual Screening Methods: Good and Bad Metrics for the “Early Recognition” Problem. J. Chem. Inf. Model. 2007, 47, 488–508. [CrossRef] 35. Goksuluk, D.; Korkmaz, S.; Zararsiz, G.; Karaagaoglu, A.E. EasyROC: An interactive web-tool for roc curve analysis using r language environment. R J. 2016, 8, 213–230. [CrossRef] 36. Triballeau, N.; Acher, F.; Brabet, I.; Pin, J.P.; Bertrand, H.O. Virtual screening workflow development guided by the “receiver operating characteristic” curve approach. Application to high-throughput docking on metabotropic glutamate receptor subtype 4. J. Med. Chem. 2005, 48, 2534–2547. [CrossRef] [PubMed] 37. Mandrekar, J.N. Receiver operating characteristic curve in diagnostic test assessment. J. Thorac. Oncol. 2010, 5, 1315–1316. [CrossRef] 38. Kwofie, S.; Dankwa, B.; Odame, E.; Agamah, F.; Doe, L.; Teye, J.; Agyapong, O.; Miller, W.; Mosi, L.; Wilson, M. In Silico Screening of Isocitrate Lyase for Novel Anti-Buruli Ulcer Natural Products Originating from Africa. Molecules 2018, 23, 1550. [CrossRef] [PubMed] Molecules 2019, 24, 2299 20 of 21 39. Shamsara, J. Correlation between Virtual Screening Performance and Binding Site Descriptors of Protein Targets. Int. J. Med. Chem. 2018, 2018, 1–10. [CrossRef] [PubMed] 40. Quiroga, R.; Villarreal, M.A. Vinardo: A scoring function based on autodock vina improves scoring, docking, and virtual screening. PLoS ONE 2016, 11, e0155183. [CrossRef] [PubMed] 41. Lagorce, D.; Sperandio, O.; Galons, H.; Miteva, M.A.; Villoutreix, B.O. FAF-Drugs2: Free ADME/tox filtering tool to assist drug discovery and chemical biology projects. BMC Bioinform. 2008, 9, 396. [CrossRef] 42. Sliwoski, G.; Kothiwale, S.; Meiler, J.; Lowe, E.W. Computational Methods in Drug Discovery. Pharmacol. Rev. 2014, 66, 334–395. [CrossRef] [PubMed] 43. Kubinyi, H. Hydrogen Bonding: The Last Mystery in Drug Design? In Pharmacokinetic Optimization in Drug Research; Verlag Helvetica Chimica Acta: Zürich, Switzerland, 2007; pp. 513–524. 44. Benigni, R.; Bossa, C. Structural Alerts of Mutagens and Carcinogens. Curr. Comput. Aided-Drug Des. 2006, 2, 169–176. [CrossRef] 45. Greene, N.; Judson, P.N.; Langowski, J.J.; Marchant, C.A. Knowledge-Based Expert Systems for Toxicity and Metabolism Prediction: DEREK, StAR and METEOR. Sar Qsar Environ. Res. 1999, 10, 299–314. [CrossRef] 46. Dobo, K.L.; Greene, N.; Fred, C.; Glowienke, S.; Harvey, J.S.; Hasselgren, C.; Jolly, R.; Kenyon, M.O.; Munzner, J.B.; Muster, W.; et al. In silico methods combined with expert knowledge rule out mutagenic potential of pharmaceutical impurities: An industry survey. Regul. Toxicol. Pharmacol. 2012, 62, 449–455. [CrossRef] [PubMed] 47. Berendsen, H.J.C.; Hess, B.; Lindahl, E.; Van Der Spoel, D.; Mark, A.E.; Groenhof, G. GROMACS: Fast, flexible, and free. J. Comput. Chem. 2005, 26, 1701–1718. 48. Tiburu, E.K.; Issah, I.; Darko, M.; Armah-Sekum, R.E.; Gyampo, S.O.A.; Amoateng, N.K.; Kwofie, S.K.; Awandare, G. Investigating the Conformation of S100β Protein Under Physiological Parameters Using Computational Modeling: A Clue for Rational Drug Design. Open Biomed. Eng. J. 2018, 12, 36–50. [CrossRef] [PubMed] 49. Xu, M.; Lill, M.A. Induced fit docking, and the use of QM/MM methods in docking. Drug Discov. Today Technol. 2013, 10, e411–e418. [CrossRef] [PubMed] 50. Farid, R.; Day, T.; Friesner, R.A.; Pearlstein, R.A. New insights about HERG blockade obtained from protein modeling, potential energy mapping, and docking studies. Bioorganic Med. Chem. 2006, 14, 3160–3173. [CrossRef] [PubMed] 51. Sherman, W.; Day, T.; Jacobson, M.P.; Friesner, R.A.; Farid, R. Novel procedure for modeling ligand/receptor induced fit effects. J. Med. Chem. 2006, 49, 534–553. [CrossRef] [PubMed] 52. Sherman, W.; Beard, H.S.; Farid, R. Use of an induced fit receptor structure in virtual screening. Chem. Biol. Drug Des. 2006, 67, 83–84. [CrossRef] 53. Friesner, R.A.; Murphy, R.B.; Repasky, M.P.; Frye, L.L.; Greenwood, J.R.; Halgren, T.A.; Sanschagrin, P.C.; Mainz, D.T. Extra precision glide: Docking and scoring incorporating a model of hydrophobic enclosure for protein− ligand complexes. J. Med. Chem. 2006, 49, 6177–6196. [CrossRef] 54. Medina-Franco, J.L.; Méndez-Lucio, O.; Yoo, J. Rationalization of activity cliffs of a sulfonamide inhibitor of DNA methyltransferases with induced-fit docking. Int. J. Mol. Sci. 2014, 15, 3253–3261. [CrossRef] 55. Laskowski, R.A.; Swindells, M.B. LigPlot+: Multiple ligand-protein interaction diagrams for drug discovery. J. Chem. Inf. Model. 2011, 51, 2778–2786. [CrossRef] 56. Scherr, N.; Pluschke, G.; Panda, M. Comparative Study of Activities of a Diverse Set of Antimycobacterial Agents against Mycobacterium tuberculosis and Mycobacterium ulcerans. Antimicrob. Agents Chemother. 2016, 60, 3132–3137. [CrossRef] [PubMed] 57. Scherr, N.; Bieri, R.; Thomas, S.S.; Chauffour, A.; Kalia, N.P.; Schneide, P.; Ruf, M.-T.; Lamelas, A.; Manimekalai, M.S.S.; Grüber, G.; et al. Targeting the Mycobacterium ulcerans cytochrome bc1:aa3 for the treatment of Buruli ulcer. Nat. Commun. 2018, 9, 5370. [CrossRef] [PubMed] 58. Araujo, R.C.P.; Neves, F.A.R.; Formagio, A.S.N.; Kassuya, C.A.L.; Stefanello, M.E.A.; Souza, V.V.; Pavan, F.R.; Croda, J. Evaluation of the anti-mycobacterium tuberculosis activity and in vivo acute toxicity of Annona sylvatic. BMC Complement. Altern. Med. 2014, 14, 209. [CrossRef] [PubMed] 59. Kuete, V.; Ngameni, B.; Simo, C.C.F.; Tankeu, R.K.; Ngadjui, B.T.; Meyer, J.J.M.; Lall, N.; Kuiate, J.R. Antimicrobial activity of the crude extracts and compounds from Ficus chlamydocarpa and Ficus cordata (Moraceae). J. Ethnopharmacol. 2008, 120, 17–24. [CrossRef] Molecules 2019, 24, 2299 21 of 21 60. Safwat, N.A.; Kashef, M.T.; Aziz, R.K.; Amer, K.F.; Ramadan, M.A. Quercetin 3-O-glucoside recovered from the wild Egyptian Sahara plant, Euphorbia paralias L., inhibits glutamine synthetase and has antimycobacterial activity. Tuberculosis 2018, 108, 106–113. [CrossRef] [PubMed] 61. Goud, G.L.; Ramesh, S.; Ashok, D.; Reddy, V.P.; Yogeeswari, P.; Sriram, D.; Saikrishna, B.; Manga, V. Design, synthesis, molecular-docking and antimycobacterial evaluation of some novel 1,2,3-triazolyl xanthenones. MedChemComm 2017, 8, 559–570. [CrossRef] [PubMed] 62. Jamkhande, P.G.; Pathan, S.K.; Wadher, S.J. In silico PASS analysis and determination of antimycobacterial, antifungal, and antioxidant efficacies of maslinic acid in an extract rich in pentacyclic triterpenoids. Int. J. Mycobacteriol. 2016, 5, 417–425. [CrossRef] 63. Lagunin, A.; Stepanchikova, A.; Filimonov, D.; Poroikov, V. PASS: Prediction of activity spectra for biologically active substances. Bioinformatics 2000, 16, 747–748. [CrossRef] 64. Dallakyan, S.; Olson, A.J. Small-molecule library screening by docking with PyRx. Methods Mol. Biol. 2015, 1263, 243–250. 65. Mysinger, M.M.; Carchia, M.; Irwin, J.J.; Shoichet, B.K. Directory of Useful Decoys, Enhanced (DUD-E): Better Ligands and Decoys for Better Benchmarking. J. Med. Chem. 2012, 55, 6582–6594. [CrossRef] 66. Lätti, S.; Niinivehmas, S.; Pentikäinen, O.T. Rocker: Open source, easy-to-use tool for AUC and enrichment calculations and ROC visualization. J. Cheminform. 2016, 8, 45. [CrossRef] 67. Empereur-Mot, C.; Zagury, J.F.; Montes, M. Screening Explorer-An Interactive Tool for the Analysis of Screening Results. J. Chem. Inf. Model. 2016, 56, 2281–2286. [CrossRef] [PubMed] 68. Schüttelkopf, A.W.; Van Aalten, D.M.F. PRODRG: A tool for high-throughput crystallography of protein-ligand complexes. Acta Crystallogr. Sect. D Biol. Crystallogr. 2004, 60, 1355–1363. 69. Wishart, D.S.; Feunang, Y.D.; Guo, A.C.; Lo, E.J.; Marcu, A.; Grant, J.R.; Sajed, T.; Johnson, D.; Li, C.; Sayeeda, Z.; et al. DrugBank 5.0: A major update to the DrugBank database for 2018. Nucleic Acids Res. 2018, 46, D1074–D1082. [CrossRef] [PubMed] Sample Availability: Samples of the compounds are not available from the authors. © 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).