University of Groningen
Characterization of fosfomycin heteroresistance among multidrug-resistant Escherichia coli
isolates from hospitalized patients in Rio de Janeiro, Brazil
Campos, Ana Carolina C; Andrade, Nathália L; Couto, Natacha; Mutters, Nico T; de Vos,
Marjon; Rosa, Ana Cláudia P; Damasco, Paulo V; Lo Ten Foe, Jerome R; Friedrich, Alex W;
Chlebowicz-Flissikowska, Monika A
Published in:
Journal of Global Antimicrobial Resistance
DOI:
10.1016/j.jgar.2020.04.026
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from
it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2020
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Campos, A. C. C., Andrade, N. L., Couto, N., Mutters, N. T., de Vos, M., Rosa, A. C. P., Damasco, P. V.,
Lo Ten Foe, J. R., Friedrich, A. W., Chlebowicz-Flissikowska, M. A., & Rossen, J. W. A. (2020).
Characterization of fosfomycin heteroresistance among multidrug-resistant Escherichia coli isolates from
hospitalized patients in Rio de Janeiro, Brazil. Journal of Global Antimicrobial Resistance, 22, 584-593.
https://doi.org/10.1016/j.jgar.2020.04.026
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Characterization
of
fosfomycin
heteroresistance
among
multidrug-resistant
Escherichia
coli
isolates
from
hospitalized
patients
in
Rio
de
Janeiro,
Brazil
Ana
Carolina
da
C.
Campos
a,b,
Nathália
L.
Andrade
a,
Natacha
Couto
b,
Nico
T.
Mutters
c,
Marjon
de
Vos
d,
Ana
Cláudia
de
P.
Rosa
a,
Paulo
V.
Damasco
e,f,
Jerome
R.
Lo
Ten
Foe
b,
Alex
W.
Friedrich
b,
Monika
A.
Chlebowicz-Flissikowska
b,
John
W.A.
Rossen
b,*
a
UniversidadedoEstadodoRiodeJaneiro,FaculdadedeCiênciasMédicas,DepartamentodeMicrobiologia,InmunologiaeParasitologia,Boulevard28de Setembro,77-VilaIsabel,RJ-20551-030,RiodeJaneiro,Brazil
bUniversityofGroningen,UniversityMedicalCenterGroningen,DepartmentofMedicalMicrobiologyandInfectionPrevention,Hanzeplein1,9713GZ,
Groningen,TheNetherlands
c
HeidelbergUniversityHospital,CenterforInfectiousDiseases,MedicalMicrobiologyandHygiene,ImNeuenheimerFeld672,69120,Heidelberg,Germany
d
UniversityofGroningen,InstituteforEvolutionaryLifeSciences,Linnaeusborg5thfloor,
Nijenborgh7,9747AG,Groningen,TheNetherlands
e
UniversidadedoEstadodoRiodeJaneiro,DepartamentodeDoençasInfecciosaseParasitárias,Boulevard28deSetembro,77-VilaIsabel,RJ-20551-030, RiodeJaneiro,Brazil
fUniversidadeFederaldoEstadodoRiodeJaneiro,DepartamentodeDoençasInfecciosaseParasitárias,R.VoluntáriosdaPátria,107-Botafogo,RJ-
22270-000,RiodeJaneiro,Brazil
A R T I C L E I N F O
Articlehistory:
Received4November2019
Receivedinrevisedform6March2020 Accepted17April2020
Availableonline7May2020 Keywords:
Heteroresistance Fosfomycin Escherichiacoli Urinarytractinfection Next-generationsequencing murA
A B S T R A C T
Objectives:Urinarytractinfections(UTIs)causedbymultidrug-resistantEscherichiacolihavebecomea majormedicalconcern.Oldantibioticssuchasfosfomycinhavebecomeanalternativetherapeuticoption duetotheireffectivenessand,asaresult,fosfomycinisnowusedasafirst-linedrugforthetreatmentof UTIs in many countries. Despite low resistance rates, fosfomycin heteroresistance, defined as a phenomenonwheresubpopulationsofbacteriaareresistanttohighantibioticconcentrationswhereas mostofthebacteriaaresusceptible,isanunderestimatedproblem.
Methods:ThefrequencyofheteroresistanceinE.coliisolatedfromhospitalizedpatientsinBrazilandits effectonsusceptibilityofE.coliinbiofilmswasstudiedandtheisolatesweremolecularlycharacterizedto revealthemechanismsbehindtheirfosfomycinheteroresistanceusingwhole-genomesequencing. Results:Ahigherfrequencyoffosfomycinheteroresistancecomparedwithotherstudieswasfound.In biofilms,mostheteroresistantisolateswerelesssensitivetofosfomycinthancontrolisolatesandshowed overexpressionofmetabolicgenestherebyincreasingtheirsurvivalrate.Molecularcharacterization showedthatsomeresistantsubpopulationsderivedfromheteroresistantisolateshadadefectintheir fosfomycinuptakesystemcausedbymutationsintransporterandregulatorygenes,whereasothers overexpressedthemurAgene.Nonetominoreffectsonbacterialfitnesswereobserved.Oxidativestress protection,virulenceandmetabolicgenesweredifferentiallyexpressedinresistantsubpopulationsand heteroresistantisolates.
Conclusion:FrequentdetectionofheteroresistanceinUTIsmayplayaroleinthefailureofantibiotic treatmentsandshouldthereforebemorecarefullydiagnosed.
©2020TheAuthor(s).PublishedbyElsevierLtdonbehalfofInternationalSocietyforAntimicrobial Chemotherapy.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons. org/licenses/by-nc-nd/4.0/).
1.Introduction
Uropathogenic Escherichia coli is themain aetiologicalagent causingcommunity-andhospital-acquiredurinarytractinfections (UTIs)[1].Treatmentoftheseinfectionsisachallenge,especially when caused bymultidrug-resistant (MDR) E.coli[2–4].In the past,olddrugs,includingfosfomycin,wereusedasanalternative
* Correspondingauthor.
E-mailaddress:j.w.a.rossen@rug.nl(J.W.A. Rossen). http://dx.doi.org/10.1016/j.jgar.2020.04.026
2213-7165/©2020TheAuthor(s).PublishedbyElsevierLtdonbehalfofInternationalSocietyforAntimicrobialChemotherapy.ThisisanopenaccessarticleundertheCC BY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).
JournalofGlobalAntimicrobialResistance22(2020)xxx–xxx
ContentslistsavailableatScienceDirect
Journal
of
Global
Antimicrobial
Resistance
totreatinfectionscausedbyMDRbacteria[5].Fosfomycinenters the bacterial cell through two sugar phosphorylated transport systems(GlpTandUhpT)[6],whereitinhibitscellwallsynthesis, interferes with the first cytoplasmic step of peptidoglycan synthesis and inhibits the formation of UDP-N-acetylmuramic acid,aprecursorofpeptidoglycan,byinhibitingthemurA-encoded enzymeUDP-N-acetylglucosamineenolpyruvyltransferase[5–7]. Fosfomycinisusedinseveralcountries,includingBrazil,forthe treatmentofuncomplicated UTIs and is currentlyconsidered a first-lineagentforthetreatmentofacuteuncomplicatedcystitisin womenindifferentcountries[8,9]owingtoitshigheffectiveness andlowtoxicity[10].However,previousstudieshaveindicatedan increaseinfosfomycinresistanceinthepastyears,particularlyin Asian countries [11,12]. Geneticalterations in bacteria, suchas mutationsintheglpTanduhpTtransportergenesandmurAgene, can cause resistance to fosfomycin, usually resulting in a high fitnesscost.Thisisthemainreasonwhyresistancetofosfomycin remains low in the population, since resistant strains are outcompeted by their non-resistant counterparts [6]. Indeed, whereas fosfomycinresistance is described in vitro, it is rarely observed during treatmentof UTIs caused by E. coli. Although heteroresistance is rarely observed during UTI treatment, an associationbetweenantibioticheteroresistance,treatmentfailure andrecurrentinfectionshasbeendescribedpreviously[9,13,14].
Heteroresistancecanbedefinedasthepresenceof subpopula-tions of bacteria able to grow in the presence of antibiotic concentrationshigherthantheminimuminhibitoryconcentration (MIC)whereasthemajorityofthepopulationconsistsofsusceptible cellsthatarekilled,differentfromafosfomycin-tolerantpopulation
where heteroresistance isnon-reversible anddifferent fromthe persistence phenomenon where bacteria only survive in the presenceofantibiotics,theheteroresistantisolatescanalsorapidly grow[14–16].Heteroresistancetofosfomycinhasbeendescribedas beingassociatedwithmutationsinthemurAgene[14,17].Clearly,it is clinically important and influences the treatment of UTIs. Nevertheless, information about fosfomycin heteroresistance is largelylacking.Inthisstudy,thefrequencyofheteroresistanceinE. coliisolatedfromhospitalizedpatientsinBrazilanditseffectonthe susceptibilityofE.coliinbiofilmswasstudied,andtheisolateswere molecularlycharacterizedtorevealthemechanismsbehindtheir fosfomycinheteroresistance.
2.Materialandmethods
2.1.Bacterialisolates,patients'characteristicsandantimicrobial susceptibilitytesting
Escherichia coli isolates (n=66) were obtained from urine samples of 66 patients hospitalized in four different hospitals located in Rio de Janeiro, Brazil,between November 2015and November2016.Intotal,70%ofthepatientswerewomenand90% weresufferingfromsymptomaticUTI,withameanageof76.05 years(range,26–97years).OfallUTIcases,57.6%wereclassifiedas community-acquiredinfections.Alloftheisolateswereprocessed by the clinical microbiology laboratory in Brazil and were discarded soon after. Escherichia coli K-12 MG1655 (GenBank CP025268) was used as a reference strain for bioinformatics analysesand asthecontrolstrainintheaccumulationofactive
Fig.1.Obtainingfosfomycin-heteroresistantisolatesandtheirresistantsubpopulations.Schematicrepresentationofisolationofheteroresistantisolates(populationA)and obtainingtheirresistantsubpopulations(populationB)andperformingtheantimicrobialsusceptibilitytests.PopulationAobtainedfromtheisolateswere fosfomycin-sensitivebyVITEK12,butwithcoloniesinsideofahalo.PopulationBwasobtainedusingresistantcoloniesB(inthelowpartofthehalo)andresistantcoloniesC(inthe
intracellular fosfomycinassay. MICs for fosfomycinwere deter-mined for each isolate in duplicate on separate days by Etest (bioMérieux,Marcy-l’Étoile,France)inordertorevealthestability oftheheteroresistantphenotype. Resistantcolonies were culti-vatedwithoutantibioticsfor1weekbyplatingthemevery24h ontofreshbloodagarplates.Subsequently,Etestwasperformedto assesstheirresistanceprofileaccordingtoEuropeanCommitteeon AntimicrobialSusceptibilityTesting(EUCAST)clinicalbreakpoints. 2.2.Obtainingfosfomycin-resistantsubpopulations
Thefosfomycin-resistantsubpopulationofeachheteroresistant isolatewasobtainedbyEtestusingselectingcoloniesinthelower, mediumandhigherregionsinsidethehaloformedfollowing24h ofincubation.Afterpickingthesecolonies,theywereincubatedin Mueller–Hintonbroth(MHB)withoutantibioticsandtheEtestwas repeated.FromthissecondEtest,resistantcoloniesinsidethehalo wereselectedandtheprocedurewasrepeateduntilcoloniesthat wereconsideredtobecompletelyresistant(coloniesthatdidnot formahalo)withanMIC>1024mg/Lwereobtained(Fig.1). 2.3.Populationanalysisprofiling(PAP)
PAP was performed to confirm the heteroresistant profile. Isolateswereculturedonbloodagarplatesandthecolonieswere transferredtotubeswitha0.9%salinesolutioninordertoattainan opticaldensityat600nmof0.7.Subsequently,serialdilutions(10–2 to10–8)weremade.A10-
m
Laliquotofeachbacterialdilutionwas plated onto Mueller–Hinton agar plates supplemented with glucose-6-phosphate(G6P)(25mg/mL),withfosfomycin (Media-products BV, Groningen, The Netherlands) at concentrations ranging from 0–400mg/L [14]. Following 24h and 48h of incubationat37C,colonieswerecountedbyeye.Inaddition,a modifiedPAPwasperformedbymeasuringbacterialgrowthafter 24hofincubationinMHBsupplementedwithG6Pbydetermining theopticaldensityat570nm[14].Isolateswereconsideredtobe heteroresistantifthedifferencebetweenthelowestconcentration of fosfomycin at which bacteria cannot grow and the highest fosfomycinconcentrationwherethebacteriastill cangrowwas 8-fold[18].2.4.Carbohydrateuse
Carbohydrateusewasevaluatedbythecapacityof heteroresist-antisolatestogrowonplateswithasinglecarbonsource.Bacteria wasseriallydilutedandwereculturedonM9(minimalmedium) agarplates supplemented with0.2%G6Por sn-glycerol-3-phosphate (G3P)(MediaproductsBV).Plateswereincubatedfor24–96hat 37Candthenumberofcolonieswascountedbyeye[19]. 2.5.Accumulationofactivefosfomycin
Accumulation assays were performed to determine the intracellular active fosfomycin concentration. Bacterial isolates weregrownin20mLofLuriabroth(LB)medium,werewashed twicewithfreshLBmedium,wereresuspendedin1mLofMHB supplementedwithG6Pandwereincubatedinthepresenceof 2mg/mL fosfomycin for 60min at 37C. Bacteria were then collected by centrifugation at 10 000rpm for 5min and were washed with hypertonic buffer (10mM Tris, 0.5mM MgCl2, 150mM NaCl,pH 7.3) to remove the antibiotic. Cells were re-suspendedin0.5mLofdistilledwaterandthen100
m
Lwasplated onLBagartodeterminethenumberofCFU/mL.Theother400m
L was sonicated for 30min to release the fosfomycin. Following centrifugation,theantibioticconcentrationinthesupernatantwas determinedbyadiskdiffusionassay.Sterilisedassaydisks(10mm)weresaturatedwith0.1mLofthesupernatantandweredeposited ontoLBagarplatesoverlaidwitha1:10dilutionofanovernight culture of E.coli MG1655. Fosfomycinconcentrations in super-natants were quantified by measuring the diameter of the inhibition zones on the LB agar culture and are presented in microgramsper107cells[20].
2.6.RNAisolation,sequencingandgeneexpressionanalysis To testthehypothesisthat fosfomycin-heteroresistantE.coli andtheirresistantsubpopulationsdifferedintheirgene expres-sion,theirinvitrotranscriptionalprofileswerecomparedusing RNAsequencing(RNA-Seq).RNAwasisolatedusingamiRNeasyKit (QIAGEN, Hilden, Germany) according to the manufacturer’s instructions.PurifiedRNAwasquantifiedusingaQubit fluorome-ter(ThermoFisherScientific,Groningen,TheNetherlands)andthe qualitywasassessedusinganAgilentTapeStation4200(Agilent Technologies, SantaClara,CA,USA).rRNAwas removedusinga MICROBExpressTMKit(Ambion;ThermoFisherScientific). Bacte-rial mRNA was fragmented and double-stranded cDNA was generatedusingaKAPARNAHyperPrepKit(Roche,Almere,the Netherlands)accordingtothemanufacturer’sinstructions. Sam-plesweresequencedonaNextSeq500System(IlluminaInc.,San Diego,CA,USA)usingtheNextSeq500/550MidOutputv.2Kit(150 cycles,paired-end).Mappingthereferencegenomesand normal-isationofgeneexpressionwereperformedusingCLCGenomics Workbench v.11.0 (QIAGEN, Aarhus, Denmark). RNA-Seq reads werealignedtoE.colistrainK-12MG1655(RefSeqaccessionno. CP025268).Geneexpressionwasnormalisedbycalculatingreads perkilobasepermillionmappedreads(RPKM),givenbydividing themappedreadsbythetotalnumberofreads(inmillions)and thegenelengthinkilobases[21].RPKMcorrectsfordifferencesin bothsequencingdepthandlength.
2.7.Quantitativereal-timePCR(qPCR)
TotalRNAwasextractedusinganRNeasyMiniKit(QIAGEN). After quantification using a Qubit fluorometer (Thermo Fisher Scientific),cDNAwasproduced usingaFirst andSecond Strand cDNASynthesisKit(NewEnglandBiolabs,Ipswich,MA,USA)and was purified using a QIAquick PCR Purification Kit (QIAGEN). TaqManqPCRwasperformedusingprimersandprobesspecificfor therrsAand murAgenesdesigned usingPrimerExpressTMv.3.0 (Thermo Fisher Scientific) (Supplementary Table S1). Non-tem-platecontrolswereincludedforeachrun,andqPCRreactionswere performedintriplicate.RelativeexpressionlevelsofthemurAgene werecalculatedusingthecomparativeCtmethodwhererelative amountsofmurARNAwerenormalisedusingtherrsAgene(see SupplementaryTableS1forprimersandprobesusedinthisstudy). 2.8.Whole-genomesequencing(WGS)
Molecularcharacterizationofsixheteroresistantisolates,their sixresistantsubpopulationsandgeneticallycloselyrelatedisolates wasperformedusingWGS.DNAwasisolatedandlibrarieswere prepared using a Nextera XT Kit (Illumina Inc.) as previously described [22].LibrarieswererunonaMiSeqSystem(Illumina Inc.)togenerate250-bppaired-endreads.Denovoassemblywas performedusingthedefaultsettingsofCLCGenomicsWorkbench v.10.0.1 (QIAGEN)as described previously[23].Sequenceswere submitted to the European Nucleotide Archive (ENA) database (ProjectnumberPRJEB23420).Nucleotidemutationsand amino acid substitutionswere investigatedusing Artemis Comparison Tool (ACT), whilstPROVEAN v.1.1 [24] was used to predictthe functional effect of amino acid substitutions.Single nucleotide polymorphism (SNP) analyses were performed comparing the
genome of heteroresistant isolates with the reference E. coli MG1655,withtheirresistantsubpopulationsandwithgenetically closelyrelatedisolates(seeSupplementaryTableS2)byuploading thesequencesintoSeqSpherev.5.1.0(Ridom,Munster,Germany). 2.9.Individualfitnesstest
The fitness of heteroresistant isolates and their resistant subpopulations was analysed using an individual fitness assay as described previously [25] with modifications. In short, heteroresistantisolatesandtheirresistantsubpopulations were grown in tryptic soy broth (TSB) medium at 37C overnight.
Aliquotsof1%oftheovernightculturesweretransferredtofresh TSB medium and were incubated under agitation at 37C overnight. Cell densities were then determined using a BioTek spectrometer(BioTek Instruments Inc.,Winooski, VT, USA) at a wavelengthof620nm(OD620)twiceeveryhour(R1andR2).OD620 valuesbetween0.2to2.5,wheregrowthwasexponential,were usedtocalculatethemaximumgrowthrate.Fitnesswasmeasured asthemaximumgrowthratecalculatedas[(R2–R1)/
D
T]2.303, whereRequalsthelogoftheconvertedOD620readings(R1andR2) andTequalsthetimeinhours.2.10.Biofilmformation
Toquantitativelyassessbiofilmformation, 175
m
LofLBmedium wastransferred into96-well flat-bottompolystyrenemicrotitre plates(TPK)containing25m
L(105bacterialcells)ofanLBculture grownovernightat37Candtheplatewasincubatedfor24hat 37C.Thegrowthmediumwasdiscardedandreplacedafter12h. Eachwellwaswashedthreetimeswithphosphate-bufferedsaline (PBS)andthen200
m
LoffosfomycindilutedinMHBwasadded. Following24hofexposure tofosfomycin,planktoniccells were removedbyrinsingthreetimeswithwaterandthebiofilmwas stainedwith0.5%crystalvioletfor5min.Afterwashingwithwater, 200m
L of 95% ethanol was added and biofilm formation was quantified spectrophotometrically at 492nm. To assess the viability of the biofilm, mature biofilms were treated with fosfomycinfor24h.Subsequently,wellswerewashedthreetimes usingPBSandwerethenfilledwith200m
L ofPBS.Plateswere wrappedinplasticandwereplacedinasonicationbathfor30min. Then, 100m
L was used for preparing serial dilutions and subsequentplatingontoLBplatesforestimatingthenumberof viablecellsbycountingcolonies.Thetestwasperformedatleast threetimesinduplicate[26].2.11.Statisticalanalysis
Allphenotypictestswereperformedthreetimesinduplicate. Survival rates after the PAP test were analysed using a linear
regressionmodel,andbiofilmformationwasanalysedbyStudent’s t-test.qPCRresultswereanalysedbyanunpairedt-test,andfitness experimentresultswerecomparedbytwo-tailedStudent’st-test usingGraphPadPrismstatisticalsoftware(GraphPadSoftwareInc., LaJolla,CA,USA).Differentiallyexpressedgeneswereidentified using the generalised linear model test (negative binomial distribution), which involves determining whether there is evidence for a significant difference in expression of genes betweentheheteroresistantisolatesandtheirresistant subpopu-lation[27]withafalsediscoveryrate(FDR)correction[28].Genes with an adjusted P-value of <0.05 were considered to be differentiallyexpressed.
3.Resultsanddiscussion
Despitefosfomycinbeingagoodtreatmentoptionforinfections caused by MDR E. coli [10,29–31], the increase in fosfomycin-heteroresistant bacteria is an underestimated problem. In this study, the frequency of fosfomycin heteroresistanceamong 66 MDRE.coliisolatesobtainedfromurineofpatientsadmittedto hospitalsinRiodeJaneirowasinvestigated.Initially,13isolates were identified as heteroresistant as they had small colonies growing within the inhibition zone. Heteroresistance was con-firmedbyPAP(Fig.2A,B)foronlysixisolates;theothersevenwere therefore considered non-heteroresistant (see Supplementary Fig. S1 for individual growth curves). Despite low MICs to fosfomycin(Table1), thereweremoreresistantsubpopulations present that could grow in the presence of up to 300mg/L fosfomycin (Fig. 2A). In this study, a frequency of 9% (6/66) heteroresistant E. coli was found, which is higher than the frequency of 3% reported by others [14,32,33]. However, comparingstudiesisdifficultasthereisnostandardmeasureto definea heteroresistantpopulation, which mayleadtoa false-positiveor false-negativeidentificationofheteroresistance[17]. Nevertheless, we hypothesise that the frequency and clinical relevanceofheteroresistanceiscurrentlyunderestimated. Resis-tantsubpopulationsremainedresistantafter1weekofculturing without antibiotics, indicating that the phenotype is stable (Fig. 2C). In addition, extended-spectrum β-lactamase (ESBL) productionhasbeendescribed asa riskfactor forresistance to fosfomycininEnterobacteriaceae[5]andindeed67%(4/6)ofthe heteroresistant isolates in this study were ESBL-producers (Table1).Furthermore,they originatedfromdifferenthospitals andbelongedtodifferentclonalgroups(Table1),indicatingthat heteroresistance was not associated with a specific clone or hospitalandmayratherevolvespontaneously.
Resistance to fosfomycin hasbeen associated with reduced permeabilityofthebacterium,inmostcasesduetomutationsin genes encoding the sn-glycerol-3-phosphate (G3P) transporter (GlpT) and the glucose-6-phosphate (G6P) transporter (UhpT)
Fig.2. HeterogeneousresponseofEscherichiacoliisolatestofosfomycin.(A)Populationanalysisprofiling(PAP)ofE.coliisolates;inredthenegativecontrol(E.coliATCC 25922)andingreenthepositivecontrol(fosfomycin-resistantisolate708).(B,C)ExampleofanEtest(isolate3218)showingcoloniesinsidethehalo(B)andEtestforthesame resistantsubpopulation(3218UP)after1weekwithoutanyantibioticshowingthepersistenceofresistance(C).
proteins[7].Toinvestigatewhethertheseproteinswereinvolved in the fosfomycin heteroresistant profile, bacterialgrowth was assessedonaminimalmediumagarsupplementedwithG3Por G6P, the substrates for GlpT or UhpT, respectively. The results showedthatallisolateswereabletogrowusingbothcarbohydrate sources (Table 1). However, growth rates were decreased in isolates666and1469andcolonieswereonlyvisibleafter96hand 72h,respectively.ThismayresultinapossiblelossinGlpTand UhpTactivityandcouldresultinareductionoffosfomycinuptake asshowninpreviousstudies[9,34,35].
Therefore,wedecidedtoinvestigatetheintracellular concen-trationoffosfomycinintheheteroresistantisolatesasameasureof fosfomycinuptake.Resistantsubpopulationsofisolates666and 1469presentedwithastatisticallysignificantlylower concentra-tionofintracellularfosfomycincomparedwiththeir heteroresist-ant populations. Surprisingly, this was also observed for the
resistantsubpopulationsof isolates2877and7198(Fig.3).The resistant subpopulations of isolates 666 and 1469 may be less effectiveintheirfosfomycinuptakeowingtothelossoffunctionof their GlpT and UhpT transporter proteins as they showed a decreased growthrateonminimalmedium agarsupplemented with G3P and G6P. However, the mechanism attributed tothe decreaseintheactiveintracellularfosfomycinconcentrationinthe 2877 and 7198 resistant subpopulations is not clear. We hypothesisethatin the7198and2877resistantsubpopulation, mutationsingenesencodingregulatoryproteinsandthatcontrol theexpressionoffosfomycintransportersandcAMPlevels,both involved in fosfomycin uptake, are involved in the observed decreaseinintracellularfosfomycin.
We subsequently investigated whether genes known to be associated with fosfomycin resistance were associated with heteroresistance. The nucleotide sequences and corresponding
Table1
Characteristicsoffosfomycin-heteroresistantEscherichiacoliisolatesandreferenceisolates.
Isolate ESBL MLST Heteroresistantprofile GrowthonM9plates FosfomycinMIC(mg/L) Hospital WithG3P WithG6P
5770d Yes ST131 No Yes Yes 0.38 HRL
3218 Yes ST131 Yes Yes Yes 0.38 HUPE
1710D Yes ST131 No Yes Yes 0.5 HRL
9893 Yes ST131 No Yes Yes 0.25 HRL
9581A Yes ST131 Yes Yes Yes 0.25 HRL
666 No ST69 Yes Yesa Yesa 0.25 HUPE
7719 No ST69 No Yes Yes 0.25 HUPE
6050 Yes ST405 No Yes Yes 0.064 HRL
2877 Yes ST405 Yes Yes Yes 0.64 HUPE
7198 Yes ST1703 Yes Yes Yes 0.75 HRL
421 Yes ST1703 No Yes Yes 0.5 HRL
1469 No ST617 Yes Yesb
Yesb
1 HRL
ESBL, extended-spectrum β-lactamase; MLST, multilocus sequence typing; G3P, sn-glycerol-3-phosphate; G6P, glucose-6-phosphate; MIC, minimum inhibitory concentration;HRL,HospitalRioLaranjeiras;HUPE,UniversityHospitalPedroErnesto.
a
Isolatesgrewonlyafter72h.
b
Isolatesgrewonlyafter96h.
Fig.3.Intracellularaccumulationoffosfomycininheteroresistantisolatesandtheirresistantsubpopulations(indicatedwiththesuffixUP).Accumulationamongthese isolatesisgivenastheamountoffosfomycin(mg)per107
cells.Dataplottedarethemeanofthreeindependenttests.*Statisticallysignificantdifference(P<0.05)between theheteroresistantisolatecomparedwithitsresistantcounterpart.EscherichiacolistrainsATCC25922and042knowntobesusceptibletofosfomycinwereusedascontrols. A.C.C.Camposetal./JournalofGlobalAntimicrobialResistance22(2020)xxx–xxx 5
aminoacidsequencesofthetwotransporter(uhpTandglpT),the fosfomycinintracellulartarget(murA)andregulatorygenesthat controltheexpressionofthetransportersandcAMPlevels(ptsI, cyaA,uhpA,uhpCanduhpB)oftheheteroresistantpopulationand resistant subpopulations of all isolates were compared with geneticallycloselyrelated non-heteroresistantisolates collected in the same period. PROVEAN analyses of WGS data showed mutationsinglpTanduhpCgenesleadingtodeleteriousaminoacid substitutions(Table2).In theresistantsubpopulationof isolate 7198,thesesubstitutionsareinconservedregionsoftheUhpCand GlpTproteins potentiallyaffecting their function. The resistant subpopulationofisolate666hadaframeshiftmutationintheglpT geneandadeleteriousdeletionoffiveaminoacidsintheUhpA proteinsequence(P58del,D59del,I60del,S61delandG62del),and the1469 resistantsubpopulationhadaninsertion ofsix amino acids in the UhpT protein (Table 2).Mutations in this protein related to fosfomycin resistance were previously reported for enterohaemorrhagicE.coli [36]. Theresistantsubpopulation of 9581AhadasubstitutionintheGlpTproteinandamutationinthe uhpBgene,andtheresistantpopulationof3218hadamutationin
theuhpCgeneresultingina prematurestopcodon.Incontrast, none of the tested isolates had mutations in the murA gene (Table2; SupplementaryFig.S2).Theseresultsstronglysuggest thatthereisnocommongeneticbasisforheteroresistancesinceit wasassociatedwithdifferentmolecularmechanismsindifferent isolates.Fluctuationingeneexpressionbetweentheisolatesmay alsobe explainedby thepresence of isolateswitha persistent profileamongtheheteroresistantisolates.
RNA-Seq revealeddifferences in expressionlevelsof several genesintheheteroresistantisolatesandtheirresistant subpopu-lations(Fig.4).Overall,overexpressionofthemurAgenewasfound intheresistantsubpopulationcomparedwiththeheteroresistant isolate(SupplementaryTableS3).ThiswasconfirmedbyqPCRthat showed an 80.85195.34-fold (P=0.0001), 8.9754.95-fold (P=0.0001) and 9.7913.77-fold(P=0.0001) overexpressionof themurA genein theresistantsubpopulationsof isolates3218, 2877and 7198, respectively. However,nostatistical differences werefoundforthe666,9581Aand1469resistantsubpopulations (Fig. 5). In contrast, other genes known to be associated with fosfomycinuptakeandregulationshowednosignificantdifference
Table2
Aminoacidsubstitutionsinthetarget,transporterandregulatoryproteinsoffosfomycin-heteroresistantEscherichiacoliisolates.a
Resistantsubpopulation MurA GlpT UhpT UhpA UhpB UhpC PtsI
666UP None Q134# None P58del,D59del,I60del,
S61delandG62del
None None None
1469UP None None G177L(insertionof
6aminoacids)
None None None None
7198UP None L255Q None None None S332L None
3218UP None None None None L362* M346# None
2877UP None None None None None None None
9581AUP None D88E None None E409H* None None
a
Aminoacidssubstitutionspresentintheresistantsubpopulationsaftercomparationwithheteroresistantisolates,theirgeneticcloselyisolatesandcontrolstrain042.The#
indicatesframeshiftmutationsandthe*indicatesastopcodon.Underlinedaminoacidsubstitutionsindicateadeleteriouseffectontheproteinstructure.
Fig.4.GeneexpressionlevelsbyRNAsequencinginfosfomycin-heteroresistantEscherichiacoliisolatesandtheirresistantsubpopulations(indicatewiththesuffixUP).The redtobluegradientindicatesahighertolowergeneexpression.Theyellowlineindicatesthegroupformedbytheheteroresistantisolates,andthepurplelineindicatesthe groupoftheircorrespondingresistantsubpopulations.
intheirexpressionlevels.In addition,severalothergeneswere statisticallysignificantlyunderexpressedoroverexpressedinthe resistant subpopulations compared with the heteroresistant isolates (Supplementary Table S3). Among the overexpressed genes, many encode uncharacterized or hypothetical proteins, genes activated during environmental stress including the
multidrugtransporterABCgene,oxidativestressprotectivegenes (lexAand rexB), and majorfacilitatory superfamily (MFS) trans-porterandporingenes,e.g.mdtA.Interestingly,itisknownthatthe ABC transporter proteinin E.coli K-12 strain confersincreased resistancetofosfomycinandthatmdtAupregulationwasdescribed toresultinmultidrugresistance[37,38].TheilvBandadeQgenes
Fig.5.RelativemurAgeneexpressioninfosfomycin-heteroresistantEscherichiacoliisolatesandtheirresistantsubpopulations(indicatedwiththesuffixUP).Geneexpression wascalculatedusingCtvalueswith95%ofconfidence.*Statisticallysignificantdifference(P<0.05)comparingtheheteroresistantisolatesandtheirresistantsubpopulations.
Fig.6. Individualfitnesstest.Testcomparingthefitnessbetweentheheteroresistantisolatesandtheirresistantsubpopulations(indicatedwiththesuffixUP).Fitnesswas measuredasthemaximumgrowthrateattwodifferentreadingtimepoints,withthegrowthratecalculatedas[(R2–R1)/DT]2.303,whereRequalsthelogoftheconverted opticaldensityreadingatawavelengthof620nmatthetworeadingtimepoints(R1andR2)andTequalsthetimeinhours.Alltestswereperformedthreetimesinduplicate. *Statisticallysignificantdifference(P<0.05).
wereunderexpressedintheresistantsubpopulations. Interesting-ly,apreviousstudyshoweddeletionofthesegenesin fosfomycin-resistantisolates[32],alsoindicatingthatgenesnotinvolvedin fosfomycinuptakemaybeinvolvedintheheteroresistantprofile. Inaddition,somevirulencegeneswereunderexpressedamongthe resistantsubpopulationcompared withtheheteroresistant iso-lates,suchashlyE(haemolysinE)andfepE(enterobactin),while othervirulencegenessuchasnlpD(associatedwithinvasion),clsB (involved in biofilm formation) [39] and prfB (adhesin) were overexpressed(SupplementaryTableS3).Wehypothesisethatthe differentialexpressionofvirulencegenesbetweenheteroresistant isolates and their resistant subpopulations may be a survival mechanism for bacteria in the presence of high fosfomycin concentrations.
Mutationsthoughttobeinvolvedinresistancetofosfomycin arereportedtobepresentatlowfrequenciesastheyimposeahigh fitness cost [7,19,40,41]. Hence, we determined the individual fitness of the heteroresistant isolates and their resistant sub-populations. No significant reduction of fitness was observed betweentheheteroresistantisolatesandtheirresistant subpopu-lation,except for isolate2877(Fig. 6).Aspreviously described, overexpressionofthemurAgeneimposesalowfitnesscost[40].In addition,isolatesthatdidnotoverexpressthemurAgenebuthad mutationsintheglpTanduhpTgenesalsodidnotappeartohavea
decreasedfitness,similartopreviousstudieswithPseudomonas aeruginosa[42,43](seeSupplementaryFig.S3forgrowthratesof all isolates). Overexpression of the murA gene leading to fosfomycinheteroresistanceis particularly ofrisk in O25-ST131 isolates belonging to a highly virulent and MDR lineage and frequently causing UTIs and bloodstream infections [44]. If fosfomycinisconsideredanalternativedrugtotreatUTIscaused by such isolates, this may eventually lead to an increase in resistancetofosfomycin[44].
Fosfomycinusedaloneorcombinedwithotherdrugsisknown tobeefficientagainstbiofilms[45,46].Inthecurrentstudy, we investigated whether a heteroresistant profile can affect the susceptibilityofbiofilmstofosfomycin.Theresultsshowedthat fosfomycinwasabletoreducebiofilmproductionforallisolates tested.Nevertheless,thereductioninbiomasswasonly statisti-cally significant for three of the six heteroresistant isolates investigated and their resistantsubpopulations (666, 7198and 3218)(Fig.7A).Inaddition,fosfomycinappearstobeabletoreduce theviabilityofbiofilmsinmostisolates(5/6)sincethenumberof viablecellsaftertreatmentwassignificantlyreducedby fosfomy-cinformostisolates(Fig.7B).Nevertheless,for3ofthe6isolates the biofilm viabilityshowed no significant differencefollowing fosfomycintreatment(isolates9581A,3218 and7198) (Fig.7B). Theseresultsindicatethataheteroresistantprofilecanaffectthe
Fig.7.SusceptibilitytofosfomycinofbiofilmsformedbyheteroresistantEscherichiacoliisolatesandtheirresistantpopulations(indicatedwiththesuffixUP).(A)Volumeof biofilmbiomassafter24hofexposureto2000mg/Lfosfomycinforsixheteroresistantisolatesandtheirresistantsubpopulations.(B)Viabilityofthebiofilm.*Statistically significantdifference(P<0.05).EscherichiacolistrainsATCC25922and042shownbeforethedottedlinewereusedascontrols.
efficacyoffosfomycinagainstbiofilmformation.Thisisofclinical relevanceandunderlinestheimportanceofbiofilmsinUTI,since isolatesformingbiofilmonurothelialcellsandonthesurfaceof medicaldevices[47]areassociatedwithrecurringinfectionsand catheter-associatedUTIs[26].Inaddition,theviabilityofbiofilmof heteroresistantisolates7198and9581Awashigherthanforthe controlstrains(042andATCC25922), showingtheircapacityto surviveinsidebiofilmsinthepresenceofhighconcentrationsof fosfomycin. The capacity of heteroresistant isolates and their resistantsubpopulationtoform biofilmsalsoindicatesthat the bacterialfitnesscostisnothigh.Thisstudyhassomelimitations, includingtherelativelysmallsamplesize.Also,itwasnotpossible to assess a possible reduction in fosfomycin efficacy among patientssufferingwithUTIscausedbyfosfomycin-heteroresistant E.coli.
Insummary,fosfomycinisbeingusedasafirst-linedrugtotreat uncomplicatedUTIsinseveralcountries,asanalternativetreatment optionorasacombinationdrugtotreatnumerousinfectionscaused by MDR bacteria. The emergence of antimicrobial resistance to clinically important antibiotics together with thedecline in the number of newlydeveloped antibiotics may increase the needforuse ofthisantibiotic.Inordertosafeguardthefuture,acquisitionof fosfomycinresistanceandheteroresistanceshouldbeinvestigated. Inthisstudy,differentmolecularmechanismswereidentifiedthat canleadtofosfomycinheteroresistance,indicatingthatthereare bacteria capable of growth in the presence of high fosfomycin concentrationintheseBrazilianhospitalswithoutbeingdetectedby conventional antimicrobialscreening. In addition, theimpact of heteroresistanceinrelationtopatientoutcomeandtreatmentneeds to be further assessed as thepresence of these heteroresistant isolatesisalarming.Furtherknowledgeofthemolecular mecha-nisms causing heteroresistance couldimproverecommendations for the futureuse of fosfomycin in bacterial infections.Until then, heteroresistantprofilesmaypersistinthepopulation,beingeithera step towardstheevolutionofresistanceor analternativemechanism tosurviveinthepresenceofantibioticwithouttheneedtoacquire geneticresistance,whichcomeswithahighfitnesscost.Considering theabilityoftheisolatesstudiedinthisinvestigationto express resistance/heteroresistancewhilecontinuingtoproducebiofilms, heteroresistance combined with biofilm formation might be sufficient for strains to survive treatment with fosfomycin and causerecurrentinfections,especiallycatheter-associatedUTIsbut alsootherMDRinfectionsincluding foreignbodyinfectionsand medicaldevice-relatedinfections(e.g.prostheses),whichis very problematic.Thelowornofitnesscostoffosfomycin heteroresist-ance combined with underexpression of virulence factors and continuous biofilm formation might result in chronic device-associatedinfections.
Funding
ACCC has received financial support from the Abel Tasman TalentProgramfor biomedicalresearchtalentof theUniversity MedicalCenterGroningen (Groningen,TheNetherlands), which aims to support international high-quality research at the University of Groningen/University Medical Center Groningen. This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Brazil [Finance Code 001]. NTM was supported by a grant from the MinistryofScience,ResearchandtheArtsofBaden-Württemberg (MWK),Germanywithintheproject‘Surveillancevon Mehrfach-Antibiotika-Resistenzen’.
Competinginterests Nonedeclared.
Ethicalapproval
Thisstudywas approvedbytheEthical Committeeof Pedro Ernesto University Hospital (Rio de Janeiro, Brazil) and in accordance with Brazilian legislation and received registration no.CAAE:45780215.8.0000.5259.Patientssigneda consentform foruseoftheirsamplesinthisresearch;non-personaldatafrom patientsnortheiridentificationwasusedinthispublication. AppendixA.Supplementarydata
Supplementarymaterialrelatedtothisarticlecanbefound,in the online version, at doi:https://doi.org/10.1016/j. jgar.2020.04.026.
References
[1]DoumithM,DayM,CiesielczukH,HopeR,UnderwoodA,ReynoldsR,etal. RapididentificationofmajorEscherichiacolisequencetypescausingurinary tractandbloodstreaminfections.JClinMicrobiol2015;53:160–6,doi:http:// dx.doi.org/10.1128/JCM.02562-14.
[2]SastryS,ClarkeLG,AlrowaisH,QuerryAM,ShuttKA,DoiY.Clinicalappraisalof fosfomycinintheeraofantimicrobialresistance.AntimicrobAgents Chemo-ther2015;59:7355–61,doi:http://dx.doi.org/10.1128/AAC.01071-15. [3]MoreiradaSilvaRCR,deOliveiraMartinsJúniorP,GonçalvesLF,dePaulo
MartinsV,deMeloABF,Pitondo-SilvaA,etal.Ciprofloxacinresistancein uropathogenicEscherichiacoliisolatescausingcommunity-acquiredurinary infectionsinBrasília,Brazil.JGlobAntimicrobResist2017;9:61–7,doi:http:// dx.doi.org/10.1016/j.jgar.2017.01.009.
[4]Yeganeh-SefidanF,GhotaslouR,AkhiMT,SadeghiMR,Mohammadzadeh-Asl Y,BaghiHB.Fosfomycin,interestingalternativedrugfortreatmentofurinary tractinfectionscreatedbymultipledrugresistantandextendedspectrum β-lactamaseproducingstrains.IranJMicrobiol2016;8:125–31.
[5]FalagasME, VouloumanouEK,SamonisG, VardakasKZ.Fosfomycin. Clin MicrobiolRev2016;29:321–47,doi:http://dx.doi.org/10.1128/CMR.00068-15. [6]KurabayashiK,TanimotoK,FuekiS,TomitaH,HirakawaH.Elevatedexpression ofGlpTandUhpTviaFNRactivationcontributestoincreasedfosfomycin susceptibility in Escherichia coli under anaerobic conditions. Antimicrob Agents Chemother 2015;59:6352–60, doi:http://dx.doi.org/10.1128/ AAC.01176-15.
[7]Castañeda-GarcíaA,BlázquezJ,Rodríguez-RojasA.Molecularmechanisms andclinicalimpactofacquiredandintrinsicfosfomycinresistance.Antibiotics (Basel)2013;2:217–36,doi:http://dx.doi.org/10.3390/antibiotics2020217. [8]Ballestero-TéllezM, Docobo-PérezF,Rodríguez-Martínez JM,ConejoMC,
Ramos-GuelfoMS,BlázquezJ,etal.Roleofinoculumandmutantfrequencyon fosfomycin MIC discrepancies by agar dilution and broth microdilution methodsinEnterobacteriaceae.ClinMicrobiolInfect2017;23:325–31,doi: http://dx.doi.org/10.1016/j.cmi.2016.12.022.
[9]SilverLL.Fosfomycin:mechanismandresistance.ColdSpringHarbPerspect Med2017;7:a025262,doi:http://dx.doi.org/10.1101/cshperspect.a025262. [10]Giancola SE, Mahoney MV, Hogan MD, Raux BR, McCoy C, Hirsch EB.
Assessmentoffosfomycinforcomplicatedormultidrug-resistanturinarytract infections:patientcharacteristicsandoutcomes.Chemotherapy2017;62:100– 4,doi:http://dx.doi.org/10.1159/000449422.
[11]AghamaliM,SedighiM,BialvaeiAZ,MohammadzadehN,AbbasianS,Ghafouri Z,etal.Fosfomycin:mechanismsandtheincreasingprevalenceofresistance.J MedMicrobiol2019;68:11–25,doi:http://dx.doi.org/10.1099/jmm.0.000874. [12]Peretz A, NaamnehB, TkhawkhoL, Nitzan O. Highrates of fosfomycin
resistanceinGram-negativeurinaryisolatesfromIsrael.MicrobDrugResist 2019;25:408–12,doi:http://dx.doi.org/10.1089/mdr.2018.0393.
[13]BandVI,SatolaSW,BurdEM,FarleyMM,JacobJT,WeissDS. Carbapenem-resistant Klebsiella pneumoniae exhibiting clinically undetected colistin heteroresistanceleadstotreatmentfailureinamurinemodelofinfection. mBio2018;9:,doi:http://dx.doi.org/10.1128/mBio.02448-17e02448-17. [14]EngelH,Gutiérrez-FernándezJ,FlückigerC,Martínez-RipollM,MühlemannK,
Hermoso JA, et al. Heteroresistance to fosfomycin is predominant in Streptococcus pneumoniae and depends on the murA1 gene. Antimicrob Agents Chemother 2013;57:2801–8, doi:http://dx.doi.org/10.1128/ AAC.00223-13.
[15]BalabanNQ,HelaineS,LewisK,AckermannM,AldridgeB,AnderssonDI,etal. Definitionsandguidelinesforresearchonantibioticpersistence.NatRev Microbiol2019;17:441–8,doi:http://dx.doi.org/10.1038/s41579-019-0196-3. [16]FalagasME,MakrisGC,DimopoulosG,MatthaiouDK.Heteroresistance:a
concernofincreasingclinicalsignificance?ClinMicrobiolInfect2008;14:101– 4,doi:http://dx.doi.org/10.1111/j.1469-0691.2007.01912.x.
[17]El-HalfawyOM,ValvanoMA.Antimicrobialheteroresistance:anemerging fieldinneedofclarity.ClinMicrobiolRev2015;28:191–207,doi:http://dx.doi. org/10.1128/CMR.00058-14.
[18]El-HalfawyOM,ValvanoMA.Chemicalcommunicationofantibioticresistance byahighlyresistantsubpopulationofbacterialcells.PLoSOne2013;8:e68874, doi:http://dx.doi.org/10.1371/journal.pone.0068874.
[19]NilssonAI,BergOG,AspevallO,KahlmeterG,AnderssonDI.Biologicalcosts andmechanismsoffosfomycinresistanceinEscherichiacoli.AntimicrobAgent Chemother 2003;47:2850–8, doi:http://dx.doi.org/10.1128/AAC.47.9.2850-2858.2003.
[20]KurabayashiK,HirakawaY,TanimotoK,TomitaH,HirakawaH.Roleofthe CpxARtwo-componentsignaltransductionsystemincontroloffosfomycin resistanceandcarbonsubstrateuptake.JBacteriol2014;196:248–56,doi: http://dx.doi.org/10.1128/JB.01151-13.
[21]Mortazavi A, WilliamsBA, McCue K,Schaeffer L, Wold B. Mappingand quantifying mammalian transcriptomes by RNA-Seq. Nat Methods 2008;5:621–8,doi:http://dx.doi.org/10.1038/nmeth.1226.
[22]CamposACC,AndradeNL,FerdousM,ChlebowiczMA,SantosCC,CorrealJCD, etal.ComprehensivemolecularcharacterizationofEscherichiacoliisolates fromurinesamplesofhospitalizedpatientsinRiodeJaneiro,Brazil.Front Microbiol2018;9:243,doi:http://dx.doi.org/10.3389/fmicb.2018.00243. [23]FerdousM,ZhouK,MellmannA,MorabitoS,CroughsPD,deBoerRF,etal.Is
Shiga toxin-negativeEscherichia coli O157:H7enteropathogenicor enter-ohemorrhagic Escherichia coli? Comprehensive molecular analysis using whole-genomesequencing.JClinMicrobiol2015;53:3530–8,doi:http://dx. doi.org/10.1128/JCM.01899-15.
[24]ChoiY,SimsGE,MurphyS,MillerJR,ChanAP.Predictingthefunctionaleffect ofaminoacidsubstitutionsandindels.PLoSOne2012;7:e46688,doi:http://dx. doi.org/10.1371/journal.pone.0046688.
[25]DursoLM,SmithD,HutkinsRW.Measurementsoffitnessandcompetitionin commensalEscherichiacoliandE.coliO157:H7strains.ApplEnvironMicrobiol 2004;70:6466–72,doi:http://dx.doi.org/10.1128/AEM.70.11.6466-6472.2004. [26]MarcheseA,BozzolascoM,GualcoL,DebbiaEA,SchitoGC,SchitoAM.Effectof
fosfomycinaloneandincombinationwithN-acetylcysteineonE.colibiofilms. Int J Antimicrob Agents 2003;22:95–100, doi:http://dx.doi.org/10.1016/ S0924-8579(03)00232-2.
[27]RobinsonMD,McCarthyDJ,SmythGK.edgeR:aBioconductorpackagefor differentialexpressionanalysisofdigitalgeneexpressiondata.Bioinformatics 2009;26:139–40,doi:http://dx.doi.org/10.1093/bioinformatics/btp616. [28]PawitanY,MichielsS,KoscielnyS,GusnantoA,PlonerA.Falsediscoveryrate,
sensitivity and sample size for microarray studies. Bioinformatics 2005;21:3017–24,doi:http://dx.doi.org/10.1093/bioinformatics/bti448. [29]AbbottIJ,MeletiadisJ,BelghanchI,WijmaRA,KaniouraL,RobertsJA,etal.
FosfomycinefficacyandemergenceofresistanceamongEnterobacteriaceaein an in vitro dynamic bladder infection model. J Antimicrob Chemother 2018;73:709–19,doi:http://dx.doi.org/10.1093/jac/dkx441.
[30]SabharwalER,SharmaR.Fosfomycin:analternativetherapyforthetreatment ofUTIamidstescalatingantimicrobialresistance.JClinDiagnRes2015;9: DC06–09,doi:http://dx.doi.org/10.7860/JCDR/2015/15227.6951.
[31]FalagasME,KastorisAC,KapaskelisAM,KarageorgopoulosDE.Fosfomycinfor the treatment of multidrug-resistant, including extended-spectrum β-lactamase producing, Enterobacteriaceae infections:a systematic review. Lancet Infect Dis2010;10:43–50,doi:http://dx.doi.org/10.1016/S1473-3099 (09)70325-1.
[32]LucasAE,ItoR,MustaphaMM,McElhenyCL,MettusRT,BowlerSL,etal. Frequency and mechanisms of spontaneous fosfomycin nonsusceptibility observed upondiskdiffusiontesting of Escherichia coli.J ClinMicrobiol 2018;56:e01368–17,doi:http://dx.doi.org/10.1128/JCM.01368-17.
[33]KarageorgopoulosDE,WangR,YuXH,FalagasME.Fosfomycin:evaluationof thepublishedevidenceontheemergenceofantimicrobialresistancein Gram-negativepathogens.JAntimicrobChemother2012;67:255–68,doi:http://dx. doi.org/10.1093/jac/dkr466.
[34]TakahataS,IdaT,HiraishiT,SakakibaraS,MaebashiK,TeradaS,etal.Molecular mechanismsoffosfomycinresistanceinclinicalisolatesofEscherichiacoli.IntJ Antimicrob Agents 2010;35:333–7, doi:http://dx.doi.org/10.1016/j.ijantimi-cag.2009.11.011.
[35]TsengSP,WangSF,KuoCY,HuangJW,HungWC,KeGM,etal.Characterization offosfomycinresistantextended-spectrumβ-lactamase-producingEscherichia coliisolatesfromhumanandpiginTaiwan.PLoSOne2015;10:e0135864,doi: http://dx.doi.org/10.1371/journal.pone.0135864.
[36]KurabayashiK,TanimotoK,TomitaH,HirakawaH.Cooperativeactionsof CRP-cAMPandFNRincreasethefosfomycinsusceptibilityofenterohaemorrhagic Escherichiacoli(EHEC)byelevatingtheexpressionofglpTanduhpTunder anaerobic conditions. Front Microbiol 2017;8:426, doi:http://dx.doi.org/ 10.3389/fmicb.2017.00426.
[37]NishinoK,Yamaguchi A.Analysisofacompletelibraryofputativedrug transportergenesinEscherichiacoli.JBacteriol2001;183:5803–12,doi:http:// dx.doi.org/10.1128/JB.183.20.5803-5812.2001.
[38]HirakawaH,NishinoK,YamadaJ,HirataT,YamaguchiA.β-Lactamresistance modulatedbytheoverexpressionofresponseregulatorsoftwo-component signal transduction systems in Escherichia coli. J Antimicrob Chemother 2003;52:576–82,doi:http://dx.doi.org/10.1093/jac/dkg406.
[39]LorenzC,DoughertyTJ,LoryS.TranscriptionalresponsesofEscherichiacolitoa small-moleculeinhibitorofLolCDE,anessentialcomponentofthelipoprotein transport pathway. J Bacteriol 2016;198:3162–75, doi:http://dx.doi.org/ 10.1128/JB.00502-16.
[40]CouceA, Briales A, Rodríguez-Rojas A, Costas C, Pascual Á, Blázquez J. GenomewideoverexpressionscreenforfosfomycinresistanceinEscherichia coli:MurAconfersclinicalresistanceatlowfitnesscost.AntimicrobAgents Chemother2012;56:2767–9,doi:http://dx.doi.org/10.1128/AAC.06122-11. [41]MarcheseA,GualcoL,DebbiaEA,SchitoGC,SchitoAM.Invitroactivityof
fosfomycinagainstGram-negativeurinarypathogensandthebiologicalcostof fosfomycinresistance.IntJAntimicrobAgents2003;22:53–9,doi:http://dx. doi.org/10.1016/S0924-8579(03)00230-9.
[42]Rodríguez-RojasA,MaciáMD,CouceA,GómezC,Castañeda-GarcíaA,OliverA, etal.Assessingtheemergenceofresistance:theabsenceofbiologicalcostin vivomaycompromisefosfomycintreatmentsforP.aeruginosainfections.PLoS One2010;5:e10193,doi:http://dx.doi.org/10.1371/journal.pone.0010193. [43]Castañeda-GarcíaA,Rodríguez-RojasA,GuelfoJR,BlázquezJ.The
glycerol-3-phosphatepermeaseGlpTistheonlyfosfomycintransporterinPseudomonas aeruginosa. J Bacteriol 2009;191:6968–74, doi:http://dx.doi.org/10.1128/ JB.00748-09.
[44]DowningT.Tacklingdrugresistantinfectionoutbreaksofglobalpandemic Escherichia coli ST131 using evolutionary and epidemiological genomics. Microorganisms 2015;3:236–67, doi:http://dx.doi.org/10.3390/microorgan-isms3020236.
[45]TangHJ,ChenCC,ChengKC,TohHS,SuBA,ChiangSR,etal.Invitroefficacyof fosfomycin-containingregimensagainstmethicillin-resistantStaphylococcus aureusinbiofilms.JAntimicrobChemother2012;67:944–50,doi:http://dx.doi. org/10.1093/jac/dkr535.
[46]ShiJ,MaoNF,WangL,ZhangHB,ChenQ,LiuH,etal.Efficacyofcombined vancomycin and fosfomycin against methicillin-resistant Staphylococcus aureusinbiofilmsinvivo.PLoSOne2014;9:e113133,doi:http://dx.doi.org/ 10.1371/journal.pone.0113133.
[47]RijavecM,Müller-PremruM,ZakotnikB,Žgur-BertokD.Virulencefactorsand biofilm productionamong Escherichiacoli strainscausingbacteraemia of urinarytractorigin.JMedMicrobiol2008;57:1329–34,doi:http://dx.doi.org/ 10.1099/jmm.0.2008/002543-0.
