Journal of Fungi Review Vitamin Biosynthesis as an Antifungal Target ZoharMeirandNirOsherov* DepartmentofClinicalMicrobiologyandImmunology,SacklerSchoolofMedicine,Tel-AvivUniversity, Ramat-Aviv,Tel-Aviv69978,Israel;[email protected] * Correspondence:[email protected];Tel.:+972-3-640-9599;Fax:+972-3-640-9160 (cid:1)(cid:2)(cid:3)(cid:1)(cid:4)(cid:5)(cid:6)(cid:7)(cid:8)(cid:1) (cid:1)(cid:2)(cid:3)(cid:4)(cid:5)(cid:6)(cid:7) Received:29May2018;Accepted:13June2018;Published:17June2018 Abstract: The large increase in the population of immunosuppressed patients, coupled with the limited efficacy of existing antifungals and rising resistance toward them, have dramatically highlighted the need to develop novel drugs for the treatment of invasive fungal infections. Anattractivepossibilityistheidentificationofpossibledrugtargetswithinessentialfungalmetabolic pathwaysnotsharedwithhumans. Here,wereviewthevitaminbiosyntheticpathways(vitamins A–E,K)ascandidatesforthedevelopmentofantifungals. Wepresentasetofrankingcriteriathat identifythevitaminB2(riboflavin),B5(pantothenicacid),andB9(folate)biosynthesispathwaysas beingparticularlyrichinnewantifungaltargets. Weproposethatrecentscientificadvancesinthe fieldsofdrugdesignandfungalgenomicshavedevelopedsufficientlytomeritarenewedlookat thesepathwaysaspromisingsourcesforthedevelopmentofnovelclassesofantifungals. Keywords: antifungals;fungalvitaminmetabolism;drugtarget;essentialgenes 1. Introduction The number of life-threatening fungal infections has risen dramatically over the last twenty years. Recentestimateshaveidentifiedaglobalburdenofalmosttwomillionpatientswithsystemic andinvasivefungalinfections, including~700,000casesofinvasivecandidiasis, ~500,000casesof Pneumocystisjiroveciipneumonia,~250,000casesofinvasiveaspergillosis,~220,000casesofcryptococcal meningitis,and~100,000casesofdisseminatedhistoplasmosis[1,2]. Theprimaryreasonforthisisthe rapidriseinthenumberofchronicallyimmunosuppressedanddebilitatedpatients. Thisisdueto aggressivechemotherapytotreatleukemiaandotherhematologicalmalignancies,theriseinbone marrowtransplantations(BMTs),andAIDS. Treatments for invasive fungal infections remain unsatisfactory. There are only four classes of established antifungal drugs on the market—polyenes (e.g., amphotericin B formulations), triazoles(e.g.,voriconazole),thenewlyintroducedechinocandins(e.g.,caspofungin),andallylamines (e.g., terbinafine). Of these, only the first three classes are currently used to treat systemic fungal infections[3]. Nevertheless,despitetreatment,thereremainsanunacceptablyhighmortalityratein high-riskpatients. Inaddition,someofthecurrentantifungaltreatmentsinteractunfavorablywith other medications, have resistance problems, a low spectrum of activity, limited formulation, are fungistaticasopposedtofungicidal,andarefrequentlytoxic[3]. Thisisprimarilybecausefungiare eukaryotesandsharemanybiochemicalpathwaysandsubcellularstructureswithmammaliancells. Consequently,mostcurrentlyusedantifungalsarenottrulyfungal-specific. Onlytheechinocandins inhibit a specific target of the fungal cell-wall, and indeed exhibit an excellent safety profile and clinicalefficacy[4]. However,theyarenotorallyavailable,haveanarrowtherapeuticrange,andare fungistaticagainstmolds[4]. Becauseofdownsizing,consolidation,andlowprofitability,mostlargepharmaceuticalcompanies haveconsiderablyreducedorevenhaltedtheireffortstodevelopnovelantifungals,evenasresistance J.Fungi2018,4,72;doi:10.3390/jof4020072 www.mdpi.com/journal/jof J.Fungi2018,4,72 2of14 totheexistingdrugsrapidlyemergesamongstclinicalisolates[5]. Thus,thereisanurgentandunmet needtodevelopadditionalandnovelantifungaldrugsthatinhibitessentialfungal-specificcellular targetsandpathways[6]. During the last two decades, extensive molecular studies have helped identify several fungal-specificdrugtargetssharedbythemostimportanthumanpathogenicfungi,Candidaalbicans, Aspergillus fumigatus, and Cryptococcus neoformans, and not found in higher eukaryotes including humans[7–10]. Theyincludeessentialgenesthatsynthesize,maintain,andcontrolthestructureofthe fungalcellwall(e.g.,FKS1glucansynthase,PKCproteinkinaseC,CHS1chitinsynthase)[11],unique pathways participating in the uptake of iron (e.g., SIT1 siderophore transporter, sidA siderophore biosynthesis, FTR1 iron permease) [12,13], zinc and copper (e.g., zrfA-C zinc transporters, crpA coppertransporter)[14],andthetransportandsynthesisofessentialaromaticaminoacids,metabolic precursors,andvitamins(e.g.,ARO1-9,HisB,SNO1/SNZ1,ABZ1-2,andPabaA,respectively)[15–18]. Importantly, deletion of genes participating in these pathways, in one or more of the main systemic human fungal pathogens (C. albicans, A. fumigatus, and C. neoformans), abolishes their virulence[9,19–23],suggestingthattheyarepromisingdrugtargets. This review focuses exclusively on the essential vitamin biosynthetic pathways (A, B1–12, C, D,E,andK)andtheirsuitabilityasantifungaldrugtargets. Amongthecriteriaqualifyingthemas validdrugtargetsaretheirabsenceinthehumanhost,conservationamongfungi,well-understood biochemistry/structuraldata,andessentialityforfungalgrowthintheinfectedhost. Animportant caveatisthatthehostnicheshouldnotsupplytheinfectingfunguswiththosenutrients,asitwould allowittobypasspathwayinhibition. Anotablelimitationinouranalysisisthatthevirulenceofthe fungalvitaminauxotrophsdescribedherewastestedindiverseinfectionmodels,usingdifferentmouse strains,immunosuppressiveregimens(neutropenic,non-neutropenic),infectionroutes(disseminated, lungs),andreadouts(mortality,fungalload)thatcanresultindifferentoutcomes. Thus,ourapproach hasbeentofocusonpathwaysinwhichauxotrophiesresultinavirulenceinthemaximalnumberof infectiousmodels,givingthemostrobustresults. Thereareseveraloverlookedadvantagesintargetingfungal-uniqueenzymesparticipatingin vitaminbiosynthesis: (i)Thesubstratesandproductsoftheseenzymesarewell-characterizedsmall molecules with high potential druggability. This has already been exploited in the generation of inhibitorsofvitaminB5(pantothenicacid)andB9(folate)biosynthesis(seerelevantsectionsbelow)[11]; (ii) competitive inhibitors of vitamin biosynthesis can be identified relatively easily because their activityisblockeduponadditionofexcessproduct. Thispropertycanbeexploitedtorapidlyidentify pathway-specific inhibitors from large compound libraries [24,25]. Note that this approach can effectively identify metabolic inhibitors in general (e.g., essential amino acids, nucleotides, metal uptake, etc.); (iii) several of the vitamin biosynthetic pathways, including vitamin B2 (riboflavin), vitaminB5(pantothenicacid),andvitaminB9(folate)biosynthesis,supplycofactorsforhundredsof enzymesinvolvedinessentialmetabolicprocesses. Therefore,inhibitionofthesevitaminbiosynthetic pathwayswillhaveaverylargeinhibitorycascadeeffectonmanymetabolicprocesses,potentially leadingtolethalcellulardamage;(iv)someofthevitaminpathwayssupplyprecursorsneededforthe biosynthesisofknownvirulencefactors. Asaresult,inhibitingthesepathwayswillalsoinhibitthe productionofthedependentvirulencefactors. Forexample,riboflavin/vitaminB2servesasacofactor forornithine-N5-monooxygenaseSidA,whichcatalysestheinitialstepinsiderophorebiosynthesisin A.fumigatus. Siderophoresareessentialvirulencefactors,allowingthefungustoovercomesevereiron limitationinthehost[26]. Inhibitionofriboflavinbiosynthesisstrongly(>80%)reducestheproduction ofsiderophores[27],deliveringanunexpectedadvantagebyinhibitingfungalironacquisitionand growthduringinfection. Despite their potential, to date there are few antifungals targeting the vitamin biosynthetic pathways. Pneumocystis pneumonia, caused by the yeast P. jirovecii, is treated with a synergistic combination of sulfamethoxazole and trimethoprim, which inhibit two key vitamin B9/folate J.Fungi2018,4,72 3of14 biosynthetic enzymes. A similar combinatorial approach is used to treat malaria caused by Plasmodiumfalciparum,aswellasmanytypesofbacterialinfections(seebelow). While we have limited our target search to include only fungal-specific enzymes not found in humans, there are many examples of existing antimicrobial drugs (e.g., azole and allylamine antifungals, antimicrobial DHFR inhibitors, etc.) and pipeline antifungals (AX001 inositol acyltransferaseinhibitor,F901318dihydroorotatedehydrogenaseinhibitor)thatinhibittargetsshared withhumans[6]. Theantifungalspecificityofthesedrugswasachievedbypainstakinglyoptimizing theirstructuretobindmoretightlyandselectivelytothemicrobialenzyme. 2. TheVitaminA,C,D,E,andKPathwaysAreNotSuitableasAntifungalTargets Thefirstpartofthisreviewbrieflydescribestheessentialvitaminbiosyntheticpathwaysthatare unsuitable,inouropinion,forthedevelopmentofnewantifungals. TheyincludethevitaminA,C,D, E,andKpathways. Vitamin A compounds (retinol, retinal, retinoic acid, and their precursors, the carotenoids) areimportantforgrowthanddevelopment, forthemaintenanceoftheimmunesystemandgood vision[28]. AnimalslackthevitaminAbiosyntheticpathwayandrelyonexogenoussourcessuch as plants. Carotenoids are organic pigments that are found in the chloroplasts and chromoplasts ofplantsandsomeotherphotosyntheticorganisms, includingsomebacteriaandfungi. Themost common carotenoids include lycopene and the vitamin A precursor β-carotene. β-carotene is an intensered-orangepigmentabundantinplantsandfruits. Fungiproducecarotenoidsfordifferent non-essentialfunctions,includingstresstoleranceandsynthesisofphysiologicallyactiveby-products. However,inhumanpathogenicfungithatproducecarotenoids(RhizopusandAspergillusspp.),mutants unable to produce them do not display phenotypic alterations in the laboratory, apart from lack of pigmentation [29]. Many fungi do not produce carotenoids, including pathogenic species of CandidaandCryptococcus[29]. Thus,thecarotenoidpathwayisnotagoodtargetforthedevelopment ofantifungals. VitaminCorascorbicacidisacofactorforanumberofenzymesandanimportantantioxidant. Itisproducedinallhigherplantsandmostanimals. Humansandanthropoidapescannotsynthesize ascorbate,becauseofmutationsintheL-gulono-γ-lactoneoxidase(GLO)genecatalyzingthelaststep initsbiosynthesis[30].InthepathogenicfungusC.albicans,deletionoftheGLOhomologalo1increases sensitivitytooxidativestressinvitroandattenuatesvirulenceininfectedimmunocompetentmice[31]. Nevertheless,deletionoftheA.fumigatusGLOhomologAfu1g14950doesnotaffectoxidativestress sensitivityorvirulence(ourunpublishedresults). Collectively, theseresultssuggestthateventhe mostpotentinhibitorofAlo1wouldonlypartiallyreduceC.albicansvirulenceandwouldprobablybe ineffectiveagainstA.fumigatus. Finally, vitamin D (vitamin D /ergocalciferol) is produced by fungi from ergosterol in a 2 non-enzymaticphotochemicalreactioncatalyzedbyUV-Brays[32]. Itthuslacksadruggabletarget, whilevitaminsEandKareonlysynthesizedbyplantsandnotfungi. 3. TheBvitaminsasAntifungalTargets B vitamins are a chemically diverse group of water-soluble compounds that are important cofactors in cell metabolism. They include vitamins B1 (thiamine), B2 (riboflavin), B3 (niacin), B5(pantothenicacid),B6(pyridoxine),B7(biotin),B9(folate),andB12(cobalamin). Wewillbriefly describetheirbiosyntheticpathwaysandfocusonthosecontainingtargetssuitableforthedevelopment ofnewantifungals. AsummaryisprovidedinTable1. J.Fungi2018,4,72 4of14 Table1.EvaluationofthesuitabilityoffungalvitaminBbiosyntheticgenesforantifungaldevelopment. Essentialfor FUNGALCRYSTAL FungalInhibitors Pathway Gene* FungalVirulence STRUCTURE Developed THI4 No[33] Yes[34] No THI5 ND† Yes[35] No VitaminB1thiamine THI6 No[27] Yes[36] No THI20 ND Yes[37] No RIB1 Yes[27,33] No No RIB2 Yes[9,38] No No VitaminB2riboflavin RIB3 ND Yes[39] No RIB4 ND Yes[40] Yes[42]‡ RIB5 ND Yes[41] No ECM31 Yes[9] No No PAN2 ND Yes[43] No VitaminB5pantothenate PAN5 ND No No PAN6 Yes[27,38] No No VitaminB6pyridoxine SNZ1 Partially[27,33] Yes[44] No VitaminB7biotin BIO2 No[33,38,45] No No ABZ1 Yes[19,46] No No VitaminB9folate ABZ2 ND Yes[47] No FOL1 Essentialgene Yes[48] Yes[49]‡ *Saccharomycescerevisiae;†ND=notdetermined;‡Onlyactiveagainstrecombinantfungalenzyme. 3.1. VitaminB1(Thiamine) Thiamine is synthesized by bacteria, plants, and fungi but not by animals. It is a cofactor in carbohydrateandaminoacidmetabolism,glycolysis,TCAcycle,andthepentoseshunt[50]. Thiamine is generated by the Thi6p-dependant coupling of thiazole and pyrimidine, each synthesized by a separate pathway (Figure 1A). Fungi can also obtain thiamine from their environment, using a dedicatedtransporter,Thi7p. Theimportanceoffungalthiaminebiosynthesisduringanimalinfection hasonlybeendeterminedintheAspergilli. InAspergillusnidulans,athiamineauxotrophgeneratedby pointmutationofthi4wasfullyvirulentinamurinemodelofsystemicinfection[33]. InA.fumigatus, a thiamine auxotroph generated by thi6/thiB gene deletion showed attenuated virulence in both pulmonaryandsystemicmodelsofinfection. Therewasnodifferencebetweenwild-typeandmutant infungalloadandhistopathology,suggestingthatthereissufficientthiamineintheinfectedhostto allowstronggrowthofthethi6/thiBmutant[27]. Therefore,inhibitingthiaminebiosynthesismaynot beagoodstrategyforantifungaldevelopment,unlessperhapsifcombinedwiththedevelopment ofinhibitorsofthiamineuptake. Interestingly,thiaminebiosynthesisisessentialforvirulenceinthe plant pathogenic fungus Verticillium dahlia, suggesting that this pathway could be suitable for the developmentofagriculturalfungicides[51]. 3.2. VitaminB2(Riboflavin) Riboflavin in its active forms, flavin adenine dinucleotide (FAD) and flavin mononucleotide (FMN),functionsasacofactorfornumerousflavocoenzyme-catalyzedreactions,includingelectron transport, fatty acid oxidation, and vitamin B3/B6/B9 synthesis. Riboflavin is synthesized from onemoleculeofGTPandtwomoleculesofribulose5phosphatebysixenzymes(Rib1–Rib5pand Rib7p)[52](Figure1B).InS.cerevisiae,deletionofRIB1-RIB5resultsinriboflavinauxotrophy. ThevirulenceofriboflavinauxotrophshasbeenassessedinthepathogenicfungiC.albicans[9], Histoplasmacapsulatum[38],A.nidulans[33],andA.fumigatus[27]. InC.albicans,conditionalrepression ofRib2resultsinavirulenceinsystemicallyinfectedimmunocompetentmice[9]. InH.capsulatum, disruptionofRib2preventsfungalproliferationinsidemacrophagesinvitro. Virulenceinintranasally infected immunocompetent mice is severely attenuated as evidenced by an inability to replicate in the lungs or disseminate [38]. Rib2 may be unsuitable as a drug target because it has high J.Fungi2018,4,72 5of14 homologytoanuncharacterizedpseudouridylatesynthasedomain-containingprotein(NP689473)in humans. MutationalinactivationofRib1inA.nidulansattenuatesvirulenceinsystemicallyinfected immunocompetentmice[33]. DeletionofRib1(riboB)inA.fumigatusabolishesvirulenceinintranasally infected cortisone-compromised mice and systemically infected neutropenic mice, and attenuates virulenceinintranasallyneutropenicmice[27]. Underriboflavinlimitation,the∆riboBstrainismore sensitive to nitric oxide and produces less siderophores, further reducing its ability to survive in thehost[27]. Together, thesefindingsindicatethatriboflavinbiosynthesisoffersattractivetargets for the development of novel antifungals. The crystal structure of the last two enzymes in the pathway, Rib4 lumazine synthase and Rib5 riboflavin synthase, has been elucidated in the yeast Schizosaccharomycespombe,C.albicans,andCandidaglabrata[40]. Small-moleculescreenshaveidentified severalcompoundsthatinhibittheactivityofpurifiedS.pombelumazinesynthaseRib4p,although they lack antifungal activity because of poor cell penetration [42]. To identify compounds that inhibitriboflavinbiosynthesisandcanenterthefungus,wescreenedasmall-moleculelibraryagainst A.fumigatusandidentifiedtwocompoundsthatlosetheirantifungalactivitywhenexcessriboflavinis addedtothegrowthmedium,bypassingtheneedfordenovosynthesis[24]. Wepresumethatthese compoundsinhibitakeystepinfungalriboflavinbiosynthesis. Ongoingworkwilldeterminetheir precisetargets. Interestingly, many species of bacteria, including Mycobacterium tuberculosis, lack exogenous riboflavinuptakesystemsandarecompletelyreliantonendogenousbiosynthesis. Severalsubstrate analogs or antimetabolites with moderate activity against purified Rib4p or Rib5p and weak activity against M. tuberculosis bacteria in culture have been described. Taken together, these findingsdemonstratethefeasibilityofdevelopingriboflavinbiosynthesisinhibitorsasantimicrobial drugs[53,54]. 3.3. VitaminB3(Niacin/nicotinicAcid) Vitamin B3 or niacin is a precursor of NAD and NADP, which are coenzymes of many dehydrogenases. Three lines of evidence suggest that the niacin biosynthetic pathway is not a good target for antifungal development: (i) Niacin is not strictly an essential vitamin, as it can be synthesizeddenovofromtryptophan(throughtheconservedkynureninepathway)bymostorganisms, includingplants,bacteria,fungi,andanimals[55];(ii)niacinauxotrophydoesnotaffectvirulencein A.nidulans[33];and(iii)C.glabrata,whichisanaturalniacinauxotroph,isneverthelessaneffective pathogen[56]. 3.4. VitaminB5(PantothenicAcid) PantothenicacidisaprecursoroftheessentialcofactorCoA,whichfunctionsinmanycentral cellularmetabolicpathways,includingfattyacidandcarbohydratemetabolism,aswellaspolyketide and nonribosomal peptide biosynthesis [57]. Four enzymes, encoded by the genes panB-E, are responsibleforthebiosynthesisofpantothenicacidfromaspartate/spermineandketoisovalerateina pathwayfoundinmostbacteria,fungi,andplants,butnotinmammals(Figure1C).Inallorganisms, exogenouslyavailablepantothenicacidcanbeimportedintothecellbyapantothenatetransporter. In S. cerevisiae, deletion of ECM31 and PAN6, homologous to bacterial panB and panC, results in pantothenicacidauxotrophy. InC.albicans,downregulationofECM31expressionstronglyreduces fungalburden[9]. InthepathogenicyeastH.capsulatum,auxotropicpan6RNAiknockdownsshow reducedlungandspleenfungalloadsininfectedmice,indicatingreducedvirulence[38].InA.nidulans, deletionofpantoA/pan6andpantoB/ECM31resultsinpantothenicacidauxotrophy[58]. InA.fumigatus, panA/pan6deletionalsoresultsinpantothenicacidauxotrophy. Invitro,growthofthismutantunder limitingpantothenicacidandironisstronglyreduced,probablybecausepantothenicacidisneeded for siderophore biosynthesis. The A. fumigatus ∆pan6 mutant is avirulent in systemically infected neutropenicmiceandstronglyreducedinvirulenceinlung-infectedneutropenicmice[27].Importantly, pantothenicacidbiosynthesishasbeenstudiedasapossibleantibacterialtargetsincethe1940s. The J.Fungi2018,4,72 6of14 enzymesencodingpanBwerepurifiedandcrystalizedinE.coliandM.tuberculosis,enablingaccurate determinationoftheirstructureandthedevelopmentofspecificinhibitors[57].Amajorefforthasbeen madetodeveloppantothenatesynthase(panC)inhibitorsactiveagainstM.tuberculosis,asthisenzyme isessentialforvirulence[59]. Whiletheseapproachesyieldedeffectiveinhibitors(submicromolar range)ofthepurifiedenzyme,theiractivitywasstronglyreduced(~twoordersofmagnitude)against theintactorganismandnonewastestedinvivo. Itwouldbeofgreatinteresttoelucidatethecrystal structureofafungalpantothenatesynthaseandtotesttheactivityofavailablebacterialinhibitors andspecificallydesignedinhibitorsagainstthepurifiedfungalenzymeandtheintactfungusinthe presence/absenceofexogenouspantothenicacid. 3.5. VitaminB6(Pyridoxine) Vitamin B6, in the form of pyridoxal 5(cid:48)-phosphate (PLP), is a cofactor for a large number of essentialenzymestakingpartincarbohydrate,aminoacid,andfattyacidmetabolism. PLPcanalso actdirectlyasaprotectiveagentagainstreactiveoxygenspecies[60]. Bacteria,fungi,andplantsare abletosynthesizevitaminB6,whereasmostanimals,includinghumans,lackthisabilityandacquire itfromtheirfood. Fungiandplantsusethedeoxy-xylose5(cid:48)-phosphate(DXP)-independentvitamin B6biosynthesispathwaytogeneratePLPdenovo. Twosynthaseproteins,encodedbytheSNZ1and SNO1genesinS.cerevisiae,directlysynthesizePLPfromribose5(cid:48)-phosphateorribulose5(cid:48)-phosphate, in combination with glyceraldehyde 3(cid:48)-phosphate and glutamine (Figure 1D) [61]. Baker’s yeast also use the high-affinity transporter Tpn1p to uptake pyridoxine from their environment [62]. Amongthepathogenicfungi,pyridoxineauxotrophshaveonlybeengeneratedinA.nidulansand A. fumigatus. In A. nidulans, mutational inactivation of pyroA, the homolog of S. cerevisiae SNZ1, resultsinauxotrophyandstronglyattenuatesvirulenceinsystemicallyinfectedmice[33]. Similarly, deletionofA.fumigatuspyroAresultsinauxotrophyandstronglyattenuatesvirulence(70%survival)in neutropeniclung-infectedmice,althoughvirulenceisonlypartiallyattenuatedinsystemicallyinfected neutropenicmice(100%mortality10dayspost-infection)[27]. Therefore,inhibitingPLPbiosynthesis maynotbeanoptimalantifungalstrategyunlesspyridoxineuptakeisalsoblocked. Interestingly, deletionofPDX1, thehomologofS.cerevisiaeSNZ1inthepathogenicbacteriaM.tuberculosisand Helicobacter pylori, results in pyridoxine auxotrophy and severely attenuates colonization in mice, suggestingthatthispathwaycontainspotentantibacterialdrugtargets[63,64]. Thepdx1geneofthe malariaparasiteP.falciparumisalsoessentialforvirulence,andinhibitorswithactivityagainstintact plasmodiumhavebeendescribed[65,66]. 3.6. VitaminB7(Biotin) Biotin,aprostheticgroupincarboxylases/decarboxylases,isproducedbyplants,bacteria,and mostfungi. Infungi, biotinissequentiallysynthesizedfrompimeloyl-CoAbythreeBioenzymes: Bio6p,aKAPAsynthase;Bio3/4p,achimericproteincomposedofDTBandDAPAsynthases;and Bio2p,abiotinsynthase(Figure1E)[67]. Severallinesofevidencesuggestthatthebiotinbiosynthetic pathwayisnotsuitableasanantifungaltarget: C.albicanslacksBIO6andisanaturalbiotinauxotroph invitro. Despite this, it is a successful pathogen and grows well invivo [45]. Likewise, RNAi knockdown/mutation/gene deletion of bio2/bioB in H. capsulatum, A. nidulans, and A. fumigatus, respectively,resultsinbiotinauxotrophyinvitro,butdoesnotaffectvirulenceininfectedanimals, suggestingthereissufficientbiotininthehosttoallownormalfungalgrowthandvirulenceper[33,38] andourunpublishedfindings. 3.7. VitaminB9(Folate) Folatesareessentialcofactorsforthebiosynthesisofpurinesandthymidylate,andhenceforDNA replication. Bacteria,plants,andfungicansynthesizefolatedenovo,generatingdihydrofolate(DHF) frompara-aminobenzoicacid(PABA;derivedfromchorismateandglutamine)anddihydropteridine diphosphate (DHPP; derived from GTP). This reaction is sequentially catalyzed by the enzymes J.Fungi2018,4,72 7of14 dihydropteroatesynthase(DHPS)encodedbyFOL1anddihydrofolatesynthase(DHFS)encodedby FOL3(Figure1F).HumanslackthoseenzymesandmustobtainfolateintheformofDHFfromtheir food,subsequentlyconvertingitintotetrahydrofolate(THF)inareactioncatalyzedbydihydrofolate reductase (DHFR). DHFR is highly conserved among all animals, plants, and microorganisms. In S. cerevisiae, deletion of ABZ1, encoding PABA synthase, and FOL1/DHPS result in PABA and folate auxotrophy, respectively [68]. The importance of the pathway in fungal virulence has only beenstudiedinA.nidulansandA.fumigatus. Inbothfungi,deletionofpabA/ABZ1,encodingPABA synthase,resultsinPABAauxotrophyandtotalavirulence[19,46]. VirulenceofthepabAnullmutant wasrestoredwhenthemicereceivedPABAintheirdrinkingwater. Importantly,withdrawalofPABA supplementation during late infection blocked further lethality, suggesting that inhibitors of pabA willlikelybeactiveagainstestablishedinfection. InhibitorsofbacterialPABAsynthasewererecently identifiedbuthavenotundergoneinvivoorantifungaltesting[69]. FOL1/DHPSisalsoaviabletarget forantifungaldevelopment: DeletionofA.fumigatusFOL1/DHPS(Afu2g09840)islethalandcannotbe complementedbyfolateorfolinicacidsupplementation,suggestingthatthefungusreliesexclusively onendogenousproduction(ourunpublishedfindings). InhibitorsofbacterialDHPS,betterknownas sulfadrugs,havebeeninclinicalusesincethe1930s[70]. TheyaresimilartothesubstratePABAand competitivelyinhibititsbindingtoDHPS.Theyalsoreactwiththesecondsubstrate,DHPP,depleting itspoolandforminganinhibitorydead-endproduct. Currentsulfadrugsexhibitonlyweakantifungal activity. DerivativeswithhigheractivityagainstpurifiedC.albicansDHPSwereidentifiedbuttheyare ineffectiveagainstthewholeorganism,possiblybecauseofefflux[49]. TheS.cerevisiaeDHPScrystal structureisavailableandcouldbeusedtofacilitatethediscoveryofimprovedfungalDHPSinhibitors. AnattractiveoptionistodevelopbisubstrateinhibitorsmimickingbothPABAandDHPP.Thiscan increasetheiraffinityandspecificitycomparedwiththeexistingsulfamonosubstrateinhibitors[48]. TheenzymeDHFRishighlyconservedbetweenhumansandmicroorganisms, andseemsan unlikelycandidateforthedevelopmentofantimicrobials. Nevertheless,drugsthatselectivelyinhibit humanDHFR(e.g.,thechemotherapyagentmethotrexate)andmicrobialDHFR(e.g.,trimethoprim) weredevelopedandareinwidespreadclinicaluse[16]. Theyarecompetitiveinhibitorsthatmimicthe naturalsubstrate,DHF.However,forbothantimicrobialDHPSandDHFRdrugs,resistancerapidly emergesbecauseofpointmutationsintheactivesiteoftheenzyme. Thisproblemhasbeenaddressed by combination therapy. The simultaneous inhibition of DHPS and DHFR is both synergistic and highlyeffectiveincombattingresistance.Itisusedtotreatbacterialinfections(e.g.,resprim,containing sulfamethoxazole and trimethoprim), malaria (e.g., Lapdab, a combination of chlorproguanil and dapsone),andfungalpneumoniacausedbyP.jorovecii(resprim)[71]. J. Fungi 2018, 4, x FOR PEER REVIEW 7 of 13 and cannot be complemented by folate or folinic acid supplementation, suggesting that the fungus relies exclusively on endogenous production (our unpublished findings). Inhibitors of bacterial DHPS, better known as sulfa drugs, have been in clinical use since the 1930s [70]. They are similar to the substrate PABA and competitively inhibit its binding to DHPS. They also react with the second substrate, DHPP, depleting its pool and forming an inhibitory dead-end product. Current sulfa drugs exhibit only weak antifungal activity. Derivatives with higher activity against purified C. albicans DHPS were identified but they are ineffective against the whole organism, possibly because of efflux [49]. The S. cerevisiae DHPS crystal structure is available and could be used to facilitate the discovery of improved fungal DHPS inhibitors. An attractive option is to develop bisubstrate inhibitors mimicking both PABA and DHPP. This can increase their affinity and specificity compared with the existing sulfa monosubstrate inhibitors [48]. The enzyme DHFR is highly conserved between humans and microorganisms, and seems an unlikely candidate for the development of antimicrobials. Nevertheless, drugs that selectively inhibit human DHFR (e.g., the chemotherapy agent methotrexate) and microbial DHFR (e.g., trimethoprim) were developed and are in widespread clinical use [16]. They are competitive inhibitors that mimic the natural substrate, DHF. However, for both antimicrobial DHPS and DHFR drugs, resistance rapidly emerges because of point mutations in the active site of the enzyme. This problem has been addressed by combination therapy. The simultaneous inhibition of DHPS and DHFR is both synergistic and highly effective in combatting resistance. It is used to treat bacterial infections (e.g., resprim, containing sulfamethoxazole and trimethoprim), malaria (e.g., Lapdab, a combination of J.Fungi2018,4,72 8of14 chlorproguanil and dapsone), and fungal pneumonia caused by P. jorovecii (resprim) [71]. (a) (b) Figure1.Cont. J. JF.uFnugnig 2i021081,8 4,,4 x, 7F2OR PEER REVIEW 8 9ofo 1f31 4 (c) (d) (e) Figure1.Cont. J. FJu.nFguin 2g0i1280,1 48,, x4 ,F7O2R PEER REVIEW 9 1o0f 1o3f 14 (f) Figure 1. Overview of the S. cerevisiae pathways responsible for the biosynthesis of (A) vitamin Figure 1. Overview of the S. cerevisiae pathways responsible for the biosynthesis of (A) vitamin B1/thiamine;(B)vitaminB2/riboflavin;(C)vitaminB5/pantothenicacid;(D)vitaminB6/pyridoxine; B1/thiamine; (B) vitamin B2/riboflavin; (C) vitamin B5/pantothenic acid; (D) vitamin B6/pyridoxine; (E) vitamin B7/biotin; and (F) vitamin B9/folate. Red box indicates the vitamin product. Genes (E) vitamin B7/biotin; and (F) vitamin B9/folate. Red box indicates the vitamin product. Genes denoted denotedwith*havenohumanhomologs. with * have no human homologs. 3.8. VitaminB12(Cobalamins) 3.8. Vitamin B12 (Cobalamins) Cobalaminsareorganometalliccofactorsusedbyalimitednumberofessentialenzymes,notably Cobalamins are organometallic cofactors used by a limited number of essential enzymes, notably methionine synthase and methylmalonyl-CoA mutase [72]. Only bacteria and archaea have the methionine synthase and methylmalonyl-CoA mutase [72]. Only bacteria and archaea have the enzymes required for vitamin B synthesis. Fungi, plants, and animals (including humans) are enzymes required for vitamin B12 1s2ynthesis. Fungi, plants, and animals (including humans) are incapableofproducingvitaminB andthusthispathwaydoesnotconstituteaviableantifungaltarget. incapable of producing vitamin B1122 and thus this pathway does not constitute a viable antifungal target. 4. MetabolicPathwayInhibitorsandtheGenerationofResistance 4. MetaEbxopleicr iPenatchewhaays tIanuhgibhittuosrst haantda tnhteim Gicernoebriaatlisotna rogfe Rtiensgismtaentcaeb olicpathwaysarehighlyproneto thedevelopmentofresistance(e.g., antifolates, azoles)becauseofmutationsinthetargetenzyme. Experience has taught us that antimicrobials targeting metabolic pathways are highly prone to Nonetheless, many strategies can be used to slow down the development of drug resistance. the development of resistance (e.g., antifolates, azoles) because of mutations in the target enzyme. They include: (i) The use of drug combinations (e.g., inhibiting both DHPS and DHFR activity Nonetheless, many strategies can be used to slow down the development of drug resistance. They infolatebiosynthesis–seeabove);(ii)generatingdrugswithmultipletargets;(iii)designingdrugsthat include: (i) The use of drug combinations (e.g., inhibiting both DHPS and DHFR activity in folate producetoxicendproductsormetabolites(asoccurswithazoles)andwhicharefungicidalrather biosynthesis–see above); (ii) generating drugs with multiple targets; (iii) designing drugs that thanfungistatic,givingnotimeforresistancetodevelop(e.g.,amphotericinB);(iv)developingdrugs produce toxic end products or metabolites (as occurs with azoles) and which are fungicidal rather targetingconservedsubstratesorproductsratherthaneasilymutableenzymes(likevancomycinand than fungistatic, giving no time for resistance to develop (e.g., amphotericin B); (iv) developing drugs amphotericinB);(v)moreprecisetargetingofthedrug(e.g.,localizingazolestotheirtargetenzymein targeting conserved substrates or products rather than easily mutable enzymes (like vancomycin and theER[73]);and(vi)judicioususeofnewdrugs,includinglimitingtheirusetohumans. amphotericin B); (v) more precise targeting of the drug (e.g., localizing azoles to their target enzyme in 5t.hCe oEnRc [l7u3s]i)o; nasnd (vi) judicious use of new drugs, including limiting their use to humans. Insummary,thevitaminB2(riboflavin),B3(pantothenicacid),andB9(folate)pathwaysappear 5. Conclusions toofferthemostattractiveantifungaldrugtargetsamongtheessentialvitaminbiosyntheticpathways. TheIny scuonmtaminarwye, ltlh-ceh vairtaacmteirni zBe2d (ernibzoyfmlaevsinth),a Bt3a r(epeasnsteontthieanlfioc rafcuidn)g, aalnvdir Bu9le (nfcoela,itne)c lpuadtihnwgaoynse appapthewara y, to foolfafteer btihosey nmthoesst isa,twtrhacictihvhe asanatlirfeuandgyapl rodvruedg ittsarwgoetrsth aamsoannga ntthime icersosbeniatliaaln dviatnamtifiunn gbaiolsdyrnutghteatrigc et. paSthciwenatyisfi.c Tkhneyo wcolendtagien iwnetlhl-echfiaerladctoerfizderdu gendzeysmigens tahnadt afuren egsaslegnetinaol mforic fsuhnagsald veivruelleonpceed, isnucflfiucdiienngt ly one pathway, folate biosynthesis, which has already proved its worth as an antimicrobial and
Description: