UNIVERSITÄTSKLINIKUM HAMBURG-EPPENDORF
Institut für Medizinische Mikrobiologie, Virologie und Hygiene
Prof. Dr. med. Martin Aepfelbacher
Untersuchung zum Effekt von Tigecyclin auf die Biofilmbildung von
Staphylococcus epidermidis
Dissertation
zur Erlangung des Grades eines Doktors der Medizin an der Medizinischen Fakultät der Universität Hamburg.
vorgelegt von:
Dr. med. dent. Julian Otto Weiser aus Frankfurt am Main
Angenommen von der
Medizinischen Fakultät der Universität Hamburg am: 17.05.2017
Veröffentlicht mit Genehmigung der
Medizinischen Fakultät der Universität Hamburg.
Prüfungsausschuss, der/die Vorsitzende: Prof. Dr. Holger Rohde
Inhaltsverzeichnis
1. Originalarbeit
Sub-inhibitory tigecycline concentrations induce extracellular matrix binding protein Embp dependent Staphylococcus epidermidis biofilm formation and immune
evasion ... 4
2. Zusammenfassende Darstellung der Publikation ... 12
Einleitung ... 12
Ziel der Arbeit ... 14
Material und Methoden ... 14
Ergebnisse und Diskussion ... 16
Fazit ... 22
3. Literaturverzeichnis... 23
4. Zusammenfassung ... 26
5. Summary ... 27
6. Erklärung des Eigenanteils an der Publikation ... 28
7. Danksagung ... 29
8. Lebenslauf ... 30
InternationalJournalofMedicalMicrobiology306(2016)471–478
ContentslistsavailableatScienceDirect
International
Journal
of
Medical
Microbiology
jo u r n al hom e p a g e :w w w . e l s e v i e r . c o m / l o c a t e / i j m m
Sub-inhibitory
tigecycline
concentrations
induce
extracellular
matrix
binding
protein
Embp
dependent
Staphylococcus
epidermidis
biofilm
formation
and
immune
evasion
JulianWeisera,1,HanaeA.Henkea,1,NinaHectora,AnnaBotha,MartinChristnera,
HenningBüttnera,JefferyB.Kaplanb,HolgerRohdea,∗
aInstitutfürMedizinischeMikrobiologie,VirologieundHygiene,UniversitätsklinikumHamburg-Eppendorf,Martinistraße52,D-20246Hamburg,Germany bDepartmentofBiology,AmericanUniversity,Washington,DC20016,USA
a r t i c l e i n f o Articlehistory:
Received24February2016
Receivedinrevisedform19May2016 Accepted24May2016
Keywords:
Staphylococcusepidermidis Biofilmformation Antibioticresistance
Extracellularmatrixbindingprotein Persistence
Immuneevasion
a b s t r a c t
Biofilm-associatedStaphylococcusepidermidisimplantinfectionsarenotoriouslyreluctanttoantibiotic treatment.Herewestudiedtheeffectofsub-inhibitoryconcentrationsofpenicillin,oxacillin,vancomycin, daptomycin,linezolidandtigecyclineonS.epidermidis1585biofilmformation,expressionofextracellular matrixbindingprotein(Embp)andpotentialimplicationsforS.epidermidis−macrophageinteractions. Penicillin,vancomycin,daptomycin,andlinezolidhadnobiofilmaugmentingeffectatanyofthe concen-trationstested.Incontrast,atsub-inhibitoryconcentrationstigecyclineandoxacillinexhibitedsignificant biofilminducingactivity.InS.epidermidis1585,SarAisanegativeregulatorofgiant1MDaEmbp,and downregulationofsarAinducesEmbp-dependentassemblyofamulti-layeredbiofilmarchitecture.Dot blotimmuneassays,confocallaserscanningmicroscopy,andqPCRshowedthatunderbiofilminducing conditions,tigecyclineaugmentedembpexpressioncomparedtothecontrolgrownwithoutantibiotics. Conversely,expressionofregulatorsarAwassuppressed,suggestingthattigecyclineexertsitseffects onembpexpressionthroughSarA.Tigecyclinefailedtoinducebiofilmformationinembptransposon mutant1585-M135,provingthatundertheseconditionsEmbpup-regulationisnecessaryforbiofilm accumulation.Asafunctionalconsequence,tigecyclineinducedbiofilmformationsignificantlyimpaired theup-takeofS.epidermidisbymousemacrophage-likecelllineJ774A.1.Ourdataprovidenovelevidence forthemolecularbasisofantibioticinducedbiofilmformation,aphenotypeassociatedwithinherently increasedantimicrobialtolerance.Whilethiscouldexplainfailureofantimicrobialtherapies,persistence ofS.epidermidisinfectionsinthepresenceofsub-inhibitoryantimicrobialsisadditionallypropelledby biofilm-relatedimpairmentofmacrophage-mediatedpathogeneradication.
©2016ElsevierGmbH.Allrightsreserved.
1. Introduction
Theabilitytosubstituteorganfunctionsbytheuseofimplanted medical devices represents a remarkable advance in modern medicine.Indeed,implantslikecentralvenouscatheters,artificial lenses,heartvalvesorprostheticjointsarepartofestablished ther-apeuticstrategiesinalmosteverymedicalspeciality.Interestingly, however,implantationofadeviceputsthepatientatsignificant riskforcomplications,andamongthese,infectionsrankamongthe mostfearedmanifestations(Rosenthaletal.,2014;Chuetal.,2009;
∗ Correspondingauthor.
E-mailaddress:rohde@uke.uni-hamburg.de(H.Rohde).
1 Contributedequallytothiswork.
Rohacek andBaddour, 2015;Lentino, 2003;Wisplinghoff etal., 2004).Althoughusuallyrelativelylowinfectionratesbetween1 and5%areencountered,relatedtotheabsolutenumberofdevices used,itisestimatedthatintheUSAalone,over400,000 device-associatedinfectionsoccurperyear(Darouiche,2004;Makietal., 2006).Apartfromsignificantlycontributingtothemorbidityand mortalityofnosocomialinfections,device-relatedinfectionsalso leadtoasubstantialeconomicburden.Indeed,itisestimatedthat intheUSAannually250–500,000primarybloodstreaminfections fromintravasculardevicesoccur,causingattributableexcesscosts ofupto56,000US$perpatient(Makietal.,2006).Intriguingly,upto 80%ofdeviceassociatedinfectionsarecausedbyotherwise harm-lessskincommensalsbelongingtothegroupofcoagulase-negative staphylococci,mostnotablyStaphylococcusepidermidis(Darouiche, 2004;Macketal.,2006;Beckeretal.,2014).
http://dx.doi.org/10.1016/j.ijmm.2016.05.015 1438-4221/©2016ElsevierGmbH.Allrightsreserved.
Electronmicroscopicstudiesandevidencefromanimal mod-elsofdeviceassociated-infectionslinkS.epidermidispathogenicity withtheabilitytoformadherent,multi-layeredbiofilmson arti-ficialimplantsurfaces(Marrieetal.,1982;Ruppetal.,1999a,b; Schaefferetal.,2015;Stoodleyetal.,2010).Thestabilityand dura-bilityofthemulticellularbiofilmarchitectureessentiallydepends ontheorganizationofanextracellularmatrixthattightlytethers thebacterialcells(FlemmingandWingender,2010;Büttneretal., 2015). Thebiofilm matrix consistsof polysaccharides,proteins, DNA(i.e.extracellularDNA/eDNA),andteichoicacids.Specifically, polysaccharideintercellularadhesinPIA(Macketal.,1996;Gerke etal.,1998),accumulationassociatedproteinAap(Schaefferetal., 2015;Rohdeetal.,2005; Deckeretal.,2015),and extracellular matrixbindingproteinEmbp(Christneretal.,2010;Christneretal., 2012)havebeencharacterizedwithrespecttotheirfunctionduring biofilmformation.PIA,Aap,andEmbpallindependentlymediate S.epidermidisbiofilmformation,althoughthespecificmechanism underlyingtheintercellularadhesivepropertiesaremostlikely dif-ferent(Schommeretal.,2011).
Importantly,regardlessofthebasicmechanismleadingtocell aggregation,biofilmformationingeneralrendersS.epidermidisless susceptibletophagocytosisandeffectormechanismsofthehost’s innateimmuneresponse,i.e.antimicrobialpeptidesand comple-ment(Scherret al.,2014).In fact,comparedtobiofilmforming wildtypeS.epidermidisstrains,inanimalmodelsofimplant infec-tionsisogenicbiofilm-negativeS.epidermidismutantsweremore readilyeliminatedfromhosttissuesand failedtoestablish per-sistentinfections(Ruppetal.,1999a,b;Feyetal.,1999;Schaeffer etal.,2014).Thus,giventhegeneralchronicpersistentcourseof device-associatedinfections,itisreasonabletoassumethatbiofilm formation is directly associated with progression of disease in humans,too.
Asamajorconsequenceofthebiofilmmodeofgrowth, bac-teriadramaticallychangetheirmetabolicactivity(Lewis,2005). Inturn,evenifstandardsusceptibilitytestingindicatesclinically usefulactivity of a given antibiotic,in contrast totheir planc-toniccounterpartsbiofilmformingS.epidermidisareabletopersist in thepresence of antibiotics farabove theMIC (Lewis, 2010). Asaconsequence,treatmentofdevice-associatedinfections usu-allydemandsremovaloftheinfecteddeviceincombinationwith long-termantibiotictreatment(Darouiche,2004).Giventhehigh prevalenceofmethicillin-resistantstrainsinclinicalS.epidermidis populations,treatmentoptionsare oftenlimited, makingusage oflastresortantibiotics(e.g.vancomycin,linezolid,daptomycin, tigecycline)necessary.
TigecyclinehasbeenreportedtoinduceS.epidermidisbiofilm formationinsomestrains(Kaplanetal.,2011).Thisstudyanalysis thedeterminantsofS.epidermidisbiofilmformationinthe pres-enceofsub-inhibitorytigecyclineconcentrations.Geneticevidence isprovidedtodemonstratethattigecycline-inducedS.epidermidis biofilmformationisfunctionallylinkedtoEmbpover-expression, leadingtoproductionofanextracellularmatrixandcell aggrega-tion.Importantly,paralleltotheinductionofbiofilmformation, resistancetophagocytosisemerged. Ourresultsshedimportant mechanisticlightontoeffectsofalastresortantibioticthatunder certaincircumstancesobviouslycaninducebacterialphenotypes, i.e.biofilm formation, that are directlylinked topersistence of infection.Furthermore,theevolvingconceptofcombining tradi-tionalantimicrobialtherapywithdedicatedanti-biofilmstrategies isfurtherstrengthened.
2. Materialandmethods 2.1. Bacterialstrains
S.epidermidis1585wasisolatedfromaportcatheterinfection. Thestrainisbiofilm-negativewhengrownintrypticasesoybroth
(TSB;BectonDickinson,Cockeysville,USA)(Rohdeetal.,2005).In thepresenceofgoatserum,S.epidermidis1585formsan Embp-dependentbiofilm(Christneretal.,2010).S.epidermidis12 was isolatedfromaprostheticjointinfection(Rohdeetal.,2007a).The strainisicaADBC-positive,butdoesnotformabiofilminTSB.S. epidermidis1457andisogenicicaA-transposoninsertion mutant 1457-M10servedasPIA-andbiofilm-positiveandbiofilm-negative controls,respectively(Macketal.,1994).Insomeexperiments,S. epidermidis1585transducedwithpCM29(Berendsetal.,2010), resultinginconstitutiveexpressionofgfp,wasused(Deckeretal., 2015).
2.2. Biofilmassay
Biofilmformingabilitywastestedusingaquantitativebiofilm assayaccordingtoapublishedprotocol(Macketal.,2001).Inbrief, bacteriawereinoculatedinto2mlofTSBandallowedtogrowfor 6hat37◦C.Theculturewasdiluted1:100intofreshTSBwithor withoutantibiotics,andwellsofa96-wellmicrotiterplatewere inoculatedwith200lofthebacterialsuspension.Aftergrowth for20h,theabsorbanceat600nmwasrecorded,andmediumwas removed.Non-adherentcellswereremovedfromthesurfaceby washingwith200lPBSthreetimes.Adherentcellsweredried, andstainedwithcrystalviolet.Biofilmformationwasquantified byassessingtheabsorbanceat570nmand405nmasareference wavelength.
2.3. Quantitativereal-timePCR
Forthegeneexpressionanalysisofembp,triplicatesofovernight culturesofstrains1585and12werediluted1:100andgrownfor 6hand 16hin 20ml of trypticsoy broth(TSB) orTSB supple-mentedwithtigecycline.Bacterialcellswereharvested,immersed in RNAprotect Bacteria reagent (Qiagen, Hilden, Germany) and lysedasdescribedpreviously(Frankeetal.,2007).RNAwasisolated usingtheRNeasyMiniKit(Qiagen,Hilden,Germany)accordingto themanufacturer’sinstructions.RNAyieldwasquantifiedwiththe QubitRNAHSAssayKitonaQubit2.0Fluorometer(bothThermo FisherScientific,Bremen,Germany). Followingtheextraction, a samplevolumecontaining1gofRNAwasdigestedtwicebyuseof DNA-freeDNase(Ambion,LifeTechnologies,Darmstadt,Germany) accordingtothemanufacturer’sinstructions.
Subsequently, 5l of DNase-digested RNA were reverse-transcribed,usingtheiScriptcDNASynthesisKit(BioRad,Hercules, CA,USA)accordingtothemanufacturer’sinstructions.Two tech-nicalreplicatesofthereversetranscriptionperRNAisolationwere made.TheresultingcDNAwasfrozenimmediatelyat−80◦Cuntil
furtheruse.
Relativegene expression quantification wasperformed ona LightCycler480instrument,runningwiththeLightCycler480 Soft-wareVersion1.5.1(RocheAppliedScience,Mannheim,Germany). Primers for target genes (embp, GenBank accession number SE1011;sarA,GenBankaccessionnumberSE0390;icaA,GenBank accession number SE2293) and reference genes (gyrB, Gen-BankaccessionnumberSE0004;rho,GenBankaccessionnumber SE1719;tpiAGenBankaccessionnumberSE0559)(Crawfordetal., 2014;Theisetal.,2007)weredesignedusingthePrimer3software (Untergasseretal.,2012;KoressaarandRemm,2007).Primersand probesaresummarizedinTable1.HydrolysisprobescontainFAM asfluorophoreandBHQ1asquencher,exceptforsarAtaqprobe, whichcontainsaminorgroovebinder(MGB).Primersandprobes weresynthesizedbyLifeTechnologies(sarAtaqprobe)orMWG Eurofins.TheTaqManFastAdvancedMasterMix2x(ThermoFisher Scientific,Bremen,Germany)wasusedasrecommended bythe manufacturer.Thereactionwassetupmanuallyinareaction vol-umeof15l.Cyclingconditionswereasfollows:10min95◦C,40
J.Weiseretal./InternationalJournalofMedicalMicrobiology306(2016)471–478 473 Table1
Primersusedinthisstudy.
Primername Sequence Ampliconlength PCREfficiencya
rhotaqfor 5‘-TCGATGAACCACCAGAACACCATG-3 231bp 98%
rho taq rev 5‘-CTTCAATGTTACGTGCTGCACCGAA-3
rhotaqprobe 5‘-ACGTACATTGTCAGGTGGGCTTGATCCCGC-3
tpiAfortaq 5‘-GTGTGCTCACGTACGTCAAACA-3 194bp 97%
tpiArevtaq 5‘-TGCACCTTCTAACAATTGTACCAAATC-3
tpiAprobetaq 5‘-CGATGGCGCTCTTGTAGGTGGCGC-3
gyrBfortaq 5‘-TGGTCTGCGTTCATTTCACCAAGAC-3 246bp 94%
gyrBrevtaq 5‘-CTTGCCGATGTTGATGGTGCACA-3
gyrB probe taq 5‘-GGCGGCTGAGCAATATAAACGTAGCCCGC-3
embptaqfor 5‘-ACCTGGTGCTGTGCGTAATATAC-3 189bp 100%
embptaqrev2 5‘-GCAGTTCCGTTATTTGTTGGTCCG-3
embpprobe2 5‘-ATGGTCGTTGGACTGTTGAAACTGGGTC-3
icaAfor2 5‘-TGCCTTATTTATTGACAGTCGCTACG-3 187bp 92%
icaAfor2 5‘-CGTTGGATATTGCCTCTGTCTGG-3
icaAprobetaq 5‘-ATACTGGGTTATCAATGCCGCAGTTGTCA-3
sarAtaqfor 5‘-TGCTTCTTTGATACGGTTGTTTACTCG-3 316bp 88%
sarAtaqrev 5‘-AGCTATGGTCACTTATGCTGACAGATT-3
sarA taq probe 5‘-ACAGTTCTCTCATCGTGTTCATTACGT-3 aasdeterminedusingtheLightCycler480SoftwareVersion1.5.1.
cyclesof95◦Cfor15s,60◦Cfor60s.Twotechnicalreplicatesof eachcDNA-sampleandthecorrespondingRNAsamples,as con-trolforcontaminatingDNA,weremeasured.AnalysisoftheqPCR datawascarriedoutinGenExTMsoftware(multiD,Gothenburg,
Sweden).Theexpressionofembp,icaA,andsarAwasnormalized tothegeometricmeanofallthreereferencegenes,Cq-valueswere subjectedtoanefficiencycorrectioninGenExtoaccountfor differ-encesinqPCRassayefficiency.Technicalreplicateswereaveraged. Differencesingeneexpressionlevelswereassessedwiththe two-tailedStudent’st-test.Thesignificancelevelwassetto0.05.
2.4. Embpdetectionbydotimmuneassay(DIA)
Forsemi-quantitative detectionofEmbp,S.epidermidiswere grownin50mlTSBunderstaticgrowthconditionsat37◦C.After 20h,themediumwasremoved,adherentcellswerescrapedoff andre-suspendedinLDS-buffer(Invitrogen,Heidelberg,Germany). Afterheatingfor 10min at70◦C,suspensions werecentrifuged and supernatantswere transferredtoa fresh tube.Cell surface extractsweredilutedgeometrically,and5lofeachdilutionwas spottedontoaPVDF-membrane.Membraneswereblockedat4◦C overnight,andrabbitanti-rEmbp6599antiserumatadilutionof 1:10,000 and peroxidase-coupled anti rabbit IgG were used to detectEmbpbychemiluminescence(Christneretal.,2010).
2.5. Confocallaserscanningmicroscopyandquantificationof phagocyticuptake
For analysis of S. epidermidis biofilm formation by confocal laserscanningmicroscopy,strain1585xpCM29wasgrowninTSB withoutantibioticsforsixhours.Culturesweresubsequently sus-pended1:100inplainTSBorTSBcontainingantibioticstigecycline (0.6mg/L).200lof thesuspensionwereusedtoinoculatethe wellofa-slide8-wellcellculturedish(ibidi,Munich,Germany) andbacteriawereallowedtogrowfor20h.Afterremovalofthe medium,wells werewashed once with200l PBS buffer, bac-teriawerefixedusingformaldehyde (3.7%[vol/vol]),andfinally blockedovernightusingfetalcalfserum(FCS;20%[vol/vol]).Embp wasdetectedusingrabbitanti-rEmbp6599antiserumandan anti-rabbitIgGcoupledtoAlexa568.Imageswerethenacquiredusing a Leica TCS SP2 confocal microscope (Leica, Solms, Germany) andprocessedusingtheVolocitysoftwarepackage(Perkin-Elmer, Waltham,USA).
Phagocytic uptake was quantified by differential detection of extra- and intracellular bacteria essentially as described
(Schommeretal.,2011),withtheexception thatgfp-expressing S.epidermidis1585xpCM29wereused,makingtheimmunologic detectionof intracellularbacteria unnecessary.In brief,J774A.1 murine macrophages were exposed for 2h to S. epidermidis 1585xpCM29growninthepresenceorabsenceoftigecycline,and thenfixedin2%(vol/vol)paraformaldehydefor10minatroom temperature.AfterbeingwashedwithPBS,outsidebacteriawere detectedusingarabbit anti-S.epidermidis antiserum(1:1000in PBScontaining10%[vol/vol]FCS)andanti-rabbitIgGcoupledwith Cy5(Invitrogen,Karlsruhe,Germany).Stackimageswerecollected withanAxiovert200Mconfocallaserscanningmicroscopeat 100-and60-foldmagnifications,usingtheVolocity4.3Improvision soft-ware,andmeannumbersofintracellularbacteriawerecalculated fromthreeindependentexperimentsinwhichintracellular bacte-riawerecountedin15macrophages.
3. Resultsanddiscussion
Inclinicalpractice,giventhehighprevalenceofPBP2a-mediated -lactamresistancein clinicalS. epidermidispopulations (Mack et al., 2009), device-associated S. epidermidis infections almost always requirethe administration of vancomycin. Even if van-comycinsusceptibilityofaclinicalS.epidermidisisproven,though, cureratescanbelow(Marculescuetal.,2006).Intriguingly,thereis evidencedemonstratingabiofilm-inducingeffectofsub-inhibitory vancomycinconcentrationsonS.epidermidis(Kaplanetal.,2011; Olsen,2015),potentiallycontributingtothefailureofantimicrobial therapiesindevice-associatedinfections.Undercircumstancesin whichvancomycinisprohibitedduetonephrotoxicsideeffectsor hypersensitivity,linezolid,daptomycinortigecyclinearepotential therapeuticalternatives(Johnetal.,2009;Leiteetal.,2011;Aybar etal.,2012).Sofar,however,potentialadverseeffectsrelatedto biofilminducingpropertiesofthesecompoundshaveonlyrarely beenanalysedsofar(Kaplanetal.,2011).Tothisend,we here analysedhowlinezolid,daptomycinandtigecyclineaffectS. epi-dermidis1585invitrobiofilmformation.S.epidermidis1585was recovered from a persistent port catheter associated infection (Rohdeetal.,2005;Christneretal.,2010;Christneretal.,2012). Despiteitsclinicalsignificance,S.epidermidis1585doesnotform a biofilm understandard laboratoryconditions whengrown in TSB(Christneretal.,2010;Christneretal.,2012).Indeed,there isclearindicationfromepidemiologicalstudiesthatthefrequency ofclinicallysignificantbutbiofilm-negativeS.epidermidisstrains, especiallyinbacterialpopulationscausingprostheticjoint infec-tions,canbeashighas60%(Rohdeetal.,2007b).Themajorityof
thesebiofilm-negativeS.epidermidisisolatesdonotcarryicaADBC, an operonencoding for the synthesis machinery necessary for PIA-mediatedbiofilmformation(Heilmannetal.,1996).Instead, theyare able to establish surface adherent cell aggregates via PIA-independentmechanisms(Rohdeetal.,2005;Christneretal., 2010).ThereismountingevidencethatPIA-independent mecha-nismsaresubjecttorigorousregulation.Whilebeingsuppressed understandardlaboratoryconditions(Christneretal.,2012),they areonlyexpressedinthepresenceofspecificenvironmental stim-uli,e.g.antibiotics(Kaplanetal.,2011)orthepresenceofserum (Christneretal.,2010).
Resistance testing using Etest strips showed that, accord-ing to the EUCAST clinical breakpoints, S. epidermidis 1585 is resistanttooxacillin (MIC1mg/L), but susceptible topenicillin (MIC 0.125mg/L), vancomycin (MIC 1mg/L), daptomycin (MIC 0.125mg/L),linezolid(MIC1mg/L),andtigecycline(MIC0.5mg/L). Neither growth in the presence of varying concentrations of penicillin (0.005–5mg/L), vancomycin (0.005–5mg/L), linezolid (0.005–5mg/L)ordaptomycin(0.005–5mg/L)hadanyeffectonthe biofilmphenotypeofS.epidermidis1585(Fig.1).Thisisincontrast topreviousreportsdemonstrating thatvancomycin hadbiofilm inducingpropertiesindifferentS.epidermidisstrains,including ref-erencestrains9142,1457andRP62A(Kaplanetal.,2011;Claessens etal.,2015;CargillandUpton,2009).Apparently,theimpactofa givenantibioticonbiofilmformationsignificantlydependsonthe geneticbackgroundoftheS.epidermidisstrainsunder investiga-tion(Kaplanetal.,2011).Indeed,herewefoundthatat1.25g/ml, vancomycininducedbiofilmformationinclonallyindependentS. epidermidis12(Supplementary Fig.S1Ain theonline versionat DOI:10.1016/j.ijmm.2016.05.015),isolatedfromaPJI(Rohdeetal., 2007b).
Importantly,withinanarrowrangeoftigecycline concentra-tionsbetween0.156and0.625mg/L,aclear2–3-foldinductionof S.epidermidis1585biofilmformationbecameevident(Fig.1). Anal-ysisofbacterialgrowthdemonstratedthatthephenotypicswitch consistently became apparent at thetransition from inhibitory tosub-inhibitory tigecyclineconcentrations and waslost when the antibiotic concentration was further lowered (Fig. 1). In addition,a2-foldbiofilm-inducingeffectonS.epidermidis1585 wasalsorecordedfor oxacillinatsub-inhibitory concentrations (0.02–0.35mg/L).
Tigecyclineisabacteriostaticcompoundbelongingtothegroup ofglycylcyclines.Beingaderivativeofminocycline,thesubstance exertsabacteriostaticeffectbyinterferingwiththeribosomal pro-tein synthesismachinery, but is lesssusceptible totetracycline resistance determinants, e.g. efflux mechanisms. Interestingly, tetracyclinedidnotinduceS.epidermidis1585biofilmformation atanyoftheconcentrationstested(datanotshown),indicating thattheeffectobservedwastigecycline-specific.Atsub-inhibitory concentrations,tigecyclinealsoinducedbiofilmformationinS. epi-dermidis12(SupplementaryFig.S1AintheonlineversionatDOI:
10.1016/j.ijmm.2016.05.015).
Augmentation of S. epidermidis biofilm formation by sub-inhibitoryantibioticconcentrationsissofarmainlyattributed totranscriptionalup-regulationoficaADBC,leadingtoincreased PIA synthesis (Chang et al., 2010; Gomes et al., 2011; Rachid etal.,2000).However,inthesestudiesusuallyicaADBC-positive S. epidermidis strains were investigated, thus neglecting the possibleinvolvement of PIA-independent mechanismsof inter-cellularadhesion.Infact,astudyanalysingthemolecularbasisof antibiotic-inducedbiofilmformationinicaADBC-positiveS. epider-midisRP62Afoundthatincreasedbiofilmaccumulationrelatedto exposuretovancomycinmostlikelywaslinkedtoaugmentedPIA production,whileincontrast,tigecyclineboostedbiofilm forma-tioninaPIA-independent,mostlikelyeDNA-dependentmanner (Kaplanetal.,2011;CargillandUpton,2009).Previousworkfrom
our group has demonstrated that S. epidermidis 1585 can use extracellularmatrixbindingproteinEmbpforcellaggregationand assemblyofamulti-layeredcellconsortium(Christneretal.,2010; Christneretal.,2012).Embpisagiantproteinwithanestimated molecularweight of 1 MDa that forms an extracellular matrix embeddingthebacterialcells(Schommeretal.,2011).Theembp geneiswidespreadinclinicalS.epidermidisisolates,beingpresent inalmosteverystrain(Rohdeetal.,2007b; Rohdeetal.,2004). UnderstandardcultureconditionsinTSBembpisexpressedonly atverylowamounts(Fig.2A)(Christneretal.,2010;Linnesetal., 2013).Presenceofserum(Christneretal.,2010)orhighosmotic pressure(Linnesetal.,2013),however,induceEmbpproduction andbiofilmformation.TherepressionofembpinTSBisassociated withproductionofS.epidermidismasterregulatorSarA(Christner et al., 2012). Indeed, high-level expression of sarA reasonably explains why icaADBC-negative, embp-positive S. epidermidis isolatesregularlyfail toproducea biofilminTSB (Rohde etal., 2007a).GiventhefindingthatinactivationofsarAingenetically independent S. epidermidis isolates induces Embp-dependent biofilm formation (Christner et al., 2012), it appears plausible toassumethat Embp-mediated biofilmformationis a common phenomenon in clinical S.epidermidis isolates. Here, Embp can functionallysubstituteoractinconcertwithPIA(Christneretal., 2010).
Buildingonthesepreviousfindings,wehypothesizedthatinS. epidermidis1585tigecycline-inducedbiofilmformationcouldbe functionallylinkedtoEmbp.Totest thishypothesis, firstEmbp productionin thepresenceor absenceofsub-inhibitory tigecy-clineconcentrationswasanalysed.Dotblotimmuneassays(DIA) usingarabbitantiserumraisedagainstarecombinantsub-domain ofEmbp(rEmbp6599)(Christneretal.,2010)foundnooronlyweak EmbpsynthesisinS.epidermidis1585after24hofgrowthinTSB (Fig.2A).Insharpcontrast,thepresenceoftigecyclineata concen-trationof0.625mg/Lresultedinan8-foldEmbpinduction(Fig.2A). Intriguingly,Embp-inductionadditionallybecameapparentinthe presenceofoxacillinandlinezolid,whileallotherantibioticstested failedaugmentEmbpproduction(Fig.2A).
S. epidermidis 1585 biofilm formation and Embp-expression wasalsoanalysedbyconfocallaserscanningmicroscopy(CLSM). CLSMnotonlycorroboratedthebiofilminducingtigecycline-effect observed inthe microtiterplate assay,but alsoshowedthat S. epidermidis1585formedanEmbp-containingextracellularmatrix (Fig.2B).
Embp-induction by tigecycline as demonstrated by DIA and CLSM was also confirmed by quantitative real-time RT-PCR (Christneretal.,2010).ComparedtoS.epidermidis1585grownin TSB,presenceoftigecyclineinducedasignificant3.7(p=0.04;CI 1.1–13.1)-and3.5(p=0.01;CI2–6.2)-foldembpup-regulation after6and16hofgrowth,respectively.ParallelanalysisofsarA expressionatthesetimepointsshowedasignificant3.9-(p=0.003; CI 2.1–7.1) and 3.1- (p=0.05; CI -1.0–10.2) down-regulation, respectively. Thus, accumulation of Embp in the presence of sub-inhibitory tigecycline apparently relates to transcriptional changes.Inlinewithpublishedevidence(Christneretal.,2012), embpup-regulationis likelya directorindirectconsequenceof sarAdown-regulationinthepresenceofsub-inhibitorytigecycline. Futurestudieswillhavetoaddresssignallingeventsand regula-torynetworksinvolvedinS.epidermidisadaptationtoantibiotic exposure.
When grown in the presence of tigecycline, in S. epider-midis12a3.7-foldembpup-regulation(p<0.005;CI3.0–4.5)and parallelinduction of Embpproduction (Supplementary Fig.S1B in the online version at DOI: 10.1016/j.ijmm.2016.05.015) was observed.Ontheotherhand,undertheseconditionsnorelevant icaA-regulationbecameapparent(1.3-foldregulation,CI-9.2–5.1),
J.Weiseretal./InternationalJournalofMedicalMicrobiology306(2016)471–478 475
Fig.1.BiofilminductioninthepresenceofantibioticsinS.epidermidis1585.ThestrainwasgrowninTSBinthepresenceofvaryingconcentrationsofantimicrobialsas indicated.After20hofgrowthcelldensitieswererecordedat600nm(redgraph,openboxes),whilebiofilmformationwasquantifiedafterstainingofadherentcellswith crystalvioletandabsorbancemeasurementat570nm(bluegraph,closedcircles).ThedottedlineindicatesquantitativebiofilmformationbyS.epidermidis1585grownin TSBwithoutantibiotics.Eachdatapointrepresentsthemeanof12valuesfromthreeindependentexperiments.(Forinterpretationofthereferencestocolourinthisfigure legend,thereaderisreferredtothewebversionofthisarticle.)
indicating that in S. epidermidis 12 PIA is not in involved in tigecycline-inducedbiofilmformation.
Interestingly,induction of Embp onlybecame evident when S.epidermidis1585wasexposed totigecyclineintheearlyand mid-exponentialgrowthphase,whileexposureaftersixhoursof growthin TSB didnot increase Embp production (Supplemen-tary Fig.S2 in the online version at DOI: 10.1016/j.ijmm.2016. 05.015).Likewise,biofilminduction in S.epidermidis 1585only becameapparentifexposuretotigecyclineoccurredintheearly
or mid-exponential growth phase, suggesting a functional link betweenEmbpexpressionandtigecycline-inducedS.epidermidis biofilmformation.Totestthishypothesis,wemadeuseofisogenic embpmutant1585-M135,carryingaTn917insertionwithinembp (Christneretal.,2010).Whenmutant1585-M135wasgrowninTSB containingsub-inhibitorytigecyclineconcentrations,asignificant reductioninbiofilmformationbecameevident,providinggenetic evidencethatinS.epidermidis1585,tigecycline-inducedbiofilm formationisatleastpartiallyEmbp-dependent(Fig.2C).Likewise,
Fig.2.DetectionofEmbpproductioninthepresenceofantimicrobials.(A)CellsurfaceadherentproteinswerepreparedfromS.epidermidis1585growninthepresenceof antimicrobialsatconcentrationsasindicated.SerialdilutionsofproteinpreparationswerespottedontoPVDFmembranes,andEmbpwasdetectedbychemiluminescence usingarabbitanti-rEmbp6599antiserumandaperoxidase-coupledanti-rabbitIgG.ProteinsfromS.epidermidis1585growninTSBwithoutantibioticsservedasacontrol. UD,undiluted.(B)AnalysisofS.epidermidis1585biofilmformationandEmbpproductioninthepresenceoftigecyclinebyconfocallaserscanningmicroscopy.Gfp-expressing S.epidermidis1585xpCM29wasgrowninTSBorTSB+tigecycline(0.6mg/L)overnight.Embpwasdetectedbyrabbitanti-rEmbp6599antiserumandanAlexa568-coupled anti-rabbitIgGantibody.Whitebars=13m.(C)RelevanceofEmbpfortigecycline-andoxacillin-inducedbiofilmformation.S.epidermidis1585andcorrespondingisogenic embptransposonmutant1585-M135weregrowninthepresenceof0.6mg/Ltigecyclineoroxacillinandbiofilmformationwasquantifiedafter20hofgrowth.Barsrepresent meanof12datapointsrecordedinthreeindependentexperiments.Thedifferencebetweenmeansaresignificantlydifferent(Wilcoxonmatchedpairtest;**p<0.01;* p<0.05).
inactivationofembpalsoabolishedoxacillin-inducedbiofilm accu-mulation.Theresidualbiofilmformingcapacityofembpmutant 1585-M135grownunderinducing conditions(Fig.2C)suggests thathere,additionalintercellularadhesivemechanisms,e.g.eDNA
release (Qin et al., 2007), are contributing toadhesive growth. ThereleaseofeDNAcouldbeofsignificantimportancefor induc-tion of an antibiotic-tolerant state, as evidence indicates that eDNAcanbindantibiotics,soakthemwithinthebiofilmmatrix
J.Weiseretal./InternationalJournalofMedicalMicrobiology306(2016)471–478 477
Fig.3. UptakeofS.epidermidis1585bymousemacrophage-likeJ774A.1cells. GfpproducingS.epidermidis1585xpCM29wasgrowninplainTSBorTSB sup-plementedwithtigecycline(0.6mg/L)overnight.Mousemacrophage-likeJ774A.1 cellswereadded.After2hofincubationandfixationwithformaldehyde(3.7% [vol/vol]),outsidecellswerestainedusingarabbitanti-S.epidermidisantiserum andanAlexa568-coupledanti-rabbitIgG.Phagocytosedbacteriawereenumerated for15macrophages.Thegraphrepresentsdataacquiredduringthree indepen-dentexperiments.Themeannumberofphagocytosedbacteriawassignificantly (p<0.001;MannWhitneytest)higherintheabsenceoftigecycline.
(Mulcahyetal.,2008),andrenderbiofilmformingS.epidermidis less susceptible to e.g. vancomycin (Doroshenko et al., 2014). Moreover,inhibitorsofproteinbiosynthesisusedabovetheirMIC havebeenshown tosupporttheemergenceofpersistercellsin E.coli(Kwanetal.,2013).Thus,it istemptingtospeculatethat tigecycline-induced biofilm formation coulddrive the develop-mentofpersistercellsinS.epidermidis.
As a key consequence, assembly of a multicellular biofilm architectureprotectsS.epidermidisnotonlyfromantibiotics,but alsofromeffector mechanismsofthe hostinnateimmune sys-tem(Schommeretal., 2011;Vuong et al.,2004a;Vuong etal., 2004b).Therefore, we thoughtto investigatethe consequences of tigecycline-induced biofilm formation on interactions with mousemacrophage-likecellsJ774A.1.J774A.1cellswereaddedto gfp-expressingS.epidermidis1585xpCM29grownovernightinthe presenceorabsenceof0.6mg/Ltigecycline.After2h,intracellular (i.e.phagocytosed)bacteriawereenumeratedusingfluorescence microscopy.Indeed,comparedtobacteriathatweregrownunder tigecycline-freeconditions,tigecycline-inducedS.epidermidis1585 weresignificantly(p<0.0001;MannWhitneytest)lessefficiently taken-up by mouse macrophages (Fig. 3). Given that tigecy-cline evenat highconcentrations onlyhasa limitedimpact on bacterial uptake by phagocytes (Naess et al., 2011), it appears thatindeed,tigecycline-inducedbiofilmformationinterfereswith phagocytosis.Thisisinlinewithpreviousresults,showingthat Embp-mediated biofilm formation protects S. epidermidis 1585 fromuptakeintomousemacrophages.Thus,thebiofilminducing effectsofsub-inhibitoryantimicrobialspotentiallypromotes bac-terialpersistencebyaugmentingbacterialmechanismsfostering immuneevasion.
In conclusion, our resultsnot only provide additional infor-mationonmolecularmechanismsrelatedtosub-MIC-inducedS. epidermidisbiofilmformation,butalsounderscorethenotionthat thisphenomenoncouldhavesignificantdetrimentaleffectsonthe courseofadeviceassociated-infection.Giventheproblematic pen-etrationofantibioticsintoinfectedtissues,e.g.jointsandbones (Garazzinoetal.,2008;Rodvoldetal.,2006),itappearsplausible toassumethatindeed,duringtherapyS.epidermidisis exposed againstonlysub-inhibitoryantibioticconcentrations,drivingthe
developmentofbiofilmformation.Thus,aphenotypeispromoted whichitselfisassociatedwithaninherentlyincreasedresistance toantimicrobials,andeffectorsofthehostimmunesystem.In par-ticular,thequestionifspecificantimicrobialshaveunfavourable effectsinhost-pathogeninteractionsissofarunderappreciatedand demandsfurtherinvestigation.
Funding
ThisworkwasfundedbyanunrestrictedgrantfromPfizer,given toH.R.
Acknowledgements
WethankGescheKrollforexcellenttechnicalassistance.
References
Aybar,Y.,Ozaras,R.,Besirli,K.,Engin,E.,Karabulut,E.,Salihoglu,T.,etal.,2012. Efficacyoftigecyclineandvancomycininexperimentalcatheter-related Staphylococcusepidermidisinfection:microbiologicalandelectronmicroscopic analysisofbiofilm.Int.J.Antimicrob.Agents39(April(4)),338–342. Büttner,H.,Mack,D.,Rohde,H.,2015.StructuralbasisofStaphylococcus
epidermidisbiofilmformation:mechanismsandmolecularinteractions.Front. Cell.Infect.Microbiol.5,14.
Becker,K.,Heilmann,C.,Peters,G.,2014.Coagulase-negativestaphylococci.Clin. Microbiol.Rev.27(October(4)),870–926.
Berends,E.T.,Horswill,A.R.,Haste,N.M.,Monestier,M.,Nizet,V.,von
Kockritz-Blickwede,M.,2010.NucleaseexpressionbyStaphylococcusaureus facilitatesescapefromneutrophilextracellulartraps.J.InnateImmun.2(6), 576–586.
Cargill,J.S.,Upton,M.,2009.Lowconcentrationsofvancomycinstimulatebiofilm formationinsomeclinicalisolatesofStaphylococcusepidermidis.J.Clin.Pathol. 62(December(12)),1112–1116.
Chang,Y.M.,Jeng,W.Y.,Ko,T.P.,Yeh,Y.J.,Chen,C.K.,Wang,A.H.,2010.Structural studyofTcaRanditscomplexeswithmultipleantibioticsfromStaphylococcus epidermidis.Proc.Natl.Acad.Sci.U.S.A.107(May(19)),8617–8622. Christner,M.,Franke,G.C.,Schommer,N.N.,Wendt,U.,Wegert,K.,Pehle,P.,etal.,
2010.Thegiantextracellularmatrix-bindingproteinofStaphylococcus epidermidismediatesbiofilmaccumulationandattachmenttofibronectin.Mol. Microbiol.75(January(1)),187–207.
Christner,M.,Heinze,C.,Busch,M.,Franke,G.,Hentschke,M.,Bayard,D.S.,etal., 2012.sarAnegativelyregulatesStaphylococcusepidermidisbiofilmformation bymodulatingexpressionof1MDaextracellularmatrixbindingproteinand autolysis-dependentreleaseofeDNA.Mol.Microbiol.86(October(2)), 394–410.
Chu,V.H.,Miro,J.M.,Hoen,B.,Cabell,C.H.,Pappas,P.A.,Jones,P.,etal.,2009. Coagulase-negativestaphylococcalprostheticvalveendocarditis–a contemporaryupdatebasedontheInternationalCollaborationon Endocarditis:prospectivecohortstudy.Heart95(April(7)),570–576. Claessens,J.,Roriz,M.,Merckx,R.,Baatsen,P.,VanMellaert,L.,VanEldere,J.,2015.
Inefficacyofvancomycinandteicoplaninineradicatingandkilling
Staphylococcusepidermidisbiofilmsinvitro.Int.J.Antimicrob.Agents45(April (4)),368–375.
Crawford,E.C.,Singh,A.,Metcalf,D.,Gibson,T.W.,Weese,S.J.,2014.Identification ofappropriatereferencegenesforqPCRstudiesinStaphylococcus
pseudintermediusandpreliminaryassessmentoficaAgeneexpressionin biofilm-embeddedbacteria.BMCRes.Notes7(July),451–457,http://dx.doi. org/10.1186/1756-0500-7-451.
Darouiche,R.O.,2004.Treatmentofinfectionsassociatedwithsurgicalimplants.N. Engl.J.Med.350(April(14)),1422–1429.
Decker,R.,Burdelski,C.,Zobiak,M.,Büttner,H.,Franke,G.,Christner,M.,etal., 2015.An18kDascaffoldproteiniscriticalforStaphylococcusepidermidis biofilmformation.PLoSPathog.11(March(3)),e1004735.
Doroshenko,N.,Tseng,B.S.,Howlin,R.P.,Deacon,J.,Wharton,J.A.,Thurner,P.J., etal.,2014.ExtracellularDNAimpedesthetransportofvancomycinin Staphylococcusepidermidisbiofilmspreexposedtosubinhibitory
concentrationsofvancomycin.Antimicrob.AgentsChemother.58(December (12)),7273–7282.
Fey,P.D.,Ulphani,J.S.,Götz,F.,Heilmann,C.,Mack,D.,Rupp,M.E.,1999. Characterizationoftherelationshipbetweenpolysaccharideintercellular adhesinandhemagglutinationinStaphylococcusepidermidis.J.Infect.Dis.179 (June(6)),1561–1564.
Flemming,H.C.,Wingender,J.,2010.Thebiofilmmatrix.Nat.Rev.Microbiol.8 (September(9)),623–633.
Franke,G.C.,Dobinsky,S.,Mack,D.,Wang,C.J.,Sobottka,I.,Christner,M.,etal., 2007.Expressionandfunctionalcharacterizationofgfpmut3.1anditsunstable variantsinStaphylococcusepidermidis.J.Microbiol.Methods71(November (2)),123–132.
Garazzino,S.,Aprato,A.,Baietto,L.,D’Avolio,A.,Maiello,A.,DeRosa,F.G.,etal., 2008.Glycopeptidebonepenetrationinpatientswithsepticpseudoarthrosis ofthetibia.Clin.Pharmacokinet.47(12),793–805.
Gerke,C.,Kraft,A.,Süssmuth,R.,Schweitzer,O.,Götz,F.,1998.Characterizationof theN-acetylglucosaminyltransferaseactivityinvolvedinthebiosynthesisof theStaphylococcusepidermidispolysaccharideintercellularadhesin.J.Biol. Chem.273(July(29)),18586–18593.
Gomes,F.,Teixeira,P.,Cerca,N.,Ceri,H.,Oliveira,R.,2011.Virulencegene expressionbyStaphylococcusepidermidisbiofilmcellsexposedtoantibiotics. Microb.DrugResist.17(June(2)),191–196.
Heilmann,C.,Schweitzer,O.,Gerke,C.,Vanittanakom,N.,Mack,D.,Götz,F.,1996. Molecularbasisofintercellularadhesioninthebiofilm-formingStaphylococcus epidermidis.Mol.Microbiol.20(June(5)),1083–1091.
John,A.K.,Baldoni,D.,Haschke,M.,Rentsch,K.,Schaerli,P.,Zimmerli,W.,etal., 2009.Efficacyofdaptomycininimplant-associatedinfectiondueto methicillin-resistantStaphylococcusaureus:importanceofcombinationwith rifampin.Antimicrob.AgentsChemother.53(July(7)),2719–2724.
Kaplan,J.B.,Jabbouri,S.,Sadovskaya,I.,2011.ExtracellularDNA-dependentbiofilm formationbyStaphylococcusepidermidisRP62Ainresponsetosubminimal inhibitoryconcentrationsofantibiotics.Res.Microbiol.162(June(5)), 535–541.
Koressaar,T.,Remm,M.,2007.Enhancementsandmodificationsofprimerdesign programPrimer3.Bioinformatics23(May(10)),1289–1291.
Kwan,B.W.,Valenta,J.A.,Benedik,M.J.,Wood,T.K.,2013.Arrestedprotein synthesisincreasespersister-likecellformation.Antimicrob.Agents Chemother.57(March(3)),1468–1473.
Leite,B.,Gomes,F.,Teixeira,P.,Souza,C.,Pizzolitto,E.,Oliveira,R.,2011.Invitro activityofdaptomycin,linezolidandrifampicinonStaphylococcusepidermidis biofilms.Curr.Microbiol.63(September(3)),313–317.
Lentino,J.R.,2003.Prostheticjointinfections:baneoforthopedists,challengefor infectiousdiseasespecialists.Clin.Infect.Dis.36(9),1157–1161.
Lewis,K.,2005.Persistercellsandtheriddleofbiofilmsurvival.Biochemistry (Mosc.)70(February(2)),267–274.
Lewis,K.,2010.Persistercells.Annu.Rev.Microbiol.64,357–372.
Linnes,J.C.,Ma,H.,Bryers,J.D.,2013.Giantextracellularmatrixbindingprotein expressioninStaphylococcusepidermidisisregulatedbybiofilmformationand osmoticpressure.Curr.Microbiol.66(June(6)),627–633.
Mack,D.,Nedelmann,M.,Krokotsch,A.,Schwarzkopf,A.,Heesemann,J.,Laufs,R., 1994.Characterizationoftransposonmutantsofbiofilm-producing Staphylococcusepidermidisimpairedintheaccumulativephaseofbiofilm production:geneticidentificationofahexosamine-containingpolysaccharide intercellularadhesin.Infect.Immun.62(August(8)),3244–3253.
Mack,D.,Fischer,W.,Krokotsch,A.,Leopold,K.,Hartmann,R.,Egge,H.,etal.,1996. TheintercellularadhesininvolvedinbiofilmaccumulationofStaphylococcus epidermidisisalinearbeta-1,6-linkedglucosaminoglycan:purificationand structuralanalysis.J.Bacteriol.178(January(1)),175–183.
Mack,D.,Bartscht,K.,Fischer,C.,Rohde,H.,deGrahl,C.,Dobinsky,S.,etal.,2001. GeneticandbiochemicalanalysisofStaphylococcusepidermidisbiofilm accumulation.MethEnzymol336,215–239.
Mack,D.,Horstkotte,M.A.,Rohde,H.,Knobloch,J.K.M.,2006.Coagulase-negative Staphylococci.In:Pace,J.L.,Rupp,M.E.,Finch,R.G.(Eds.),Biofilms,Infection,and AntimicrobialTherapy.CRCPress,BocaRaton,pp.109–153.
Mack,D.,Davies,A.P.,Harris,L.G.,Knobloch,J.K.,Rohde,H.,2009.Staphylococcus epidermidisbiofilms:functionalmolecules,relationtovirulence,andvaccine potential.Top.Curr.Chem.288,157–182.
Maki,D.G.,Kluger,D.M.,Crnich,C.J.,2006.Theriskofbloodstreaminfectionin adultswithdifferentintravasculardevices:asystematicreviewof200 publishedprospectivestudies.MayoClin.Proc.81(September(9)),1159–1171. Marculescu,C.E.,Berbari,E.F.,Hanssen,A.D.,Steckelberg,J.M.,Harmsen,S.W.,
Mandrekar,J.N.,etal.,2006.Outcomeofprostheticjointinfectionstreated withdebridementandretentionofcomponents.Clin.Infect.Dis.42(February (4)),471–478.
Marrie,T.J.,Nelligan,J.,Costerton,J.W.,1982.Ascanningandtransmissionelectron microscopicstudyofaninfectedendocardialpacemakerlead.Circulation66 (December(6)),1339–1341.
Mulcahy,H.,Charron-Mazenod,L.,Lewenza,S.,2008.ExtracellularDNAchelates cationsandinducesantibioticresistanceinPseudomonasaeruginosabiofilms. PLoSPathog.4(November(1)),e1000213.
Naess,A.,Andreeva,H.,Sornes,S.,2011.Tigecyclineattenuates
polymorphonuclearleukocyte(PMN)receptorsbutnotfunctions.ActaPharm. 61(September(3)),297–302.
Olsen,I.,2015.Biofilm-specificantibiotictoleranceandresistance.Eur.J.Clin. Microbiol.Infect.Dis.34(May(5)),877–886.
Qin,Z.,Yang,X.,Yang,L.,Jiang,J.,Ou,Y.,Molin,S.,etal.,2007.Formationand propertiesofinvitrobiofilmsofica-negativeStaphylococcusepidermidis clinicalisolates.J.Med.Microbiol.56(January(Pt1)),83–93. Rachid,S.,Ohlsen,K.,Witte,W.,Hacker,J.,Ziebuhr,W.,2000.Effectof
subinhibitoryantibioticconcentrationsonpolysaccharideintercellular adhesinexpressioninbiofilm-formingStaphylococcusepidermidis.Antimicrob. AgentsChemother.44(December(12)),3357–3363.
Rodvold,K.A.,Gotfried,M.H.,Cwik,M.,Korth-Bradley,J.M.,Dukart,G.,Ellis-Grosse, E.J.,2006.Serum,tissueandbodyfluidconcentrationsoftigecyclineaftera single100mgdose.J.Antimicrob.Chemother.58(December(6)),1221–1229. Rohacek,M.,Baddour,L.M.,2015.Cardiovascularimplantableelectronicdevice
infections:associatedriskfactorsandprevention.SwissMed.Wkly.145(July), w14157,http://dx.doi.org/10.4414/smw.2015.14157.eCollection@2015. Rohde,H.,Kalitzky,M.,Kroger,N.,Scherpe,S.,Horstkotte,M.A.,Knobloch,J.K.,
etal.,2004.Detectionofvirulence-associatedgenesnotusefulfor discriminatingbetweeninvasiveandcommensalStaphylococcusepidermidis strainsfromabonemarrowtransplantunit.J.Clin.Microbiol.42(December (12)),5614–5619.
Rohde,H.,Burdelski,C.,Bartscht,K.,Hussain,M.,Buck,F.,Horstkotte,M.A.,etal., 2005.InductionofStaphylococcusepidermidisbiofilmformationviaproteolytic processingoftheaccumulation-associatedproteinbystaphylococcalandhost proteases.Mol.Microbiol.55(March(6)),1883–1895.
Rohde,H.,Burandt,E.C.,Siemssen,N.,Frommelt,L.,Burdelski,C.,Wurster,S.,etal., 2007a.Polysaccharideintercellularadhesinorproteinfactorsinbiofilm accumulationofStaphylococcusepidermidisandStaphylococcusaureusisolated fromprosthetichipandkneejointinfections.Biomaterials28(March(9)), 1711–1720.
Rohde,H.,Burandt,E.C.,Siemssen,N.,Frommelt,L.,Burdelski,C.,Wurster,S.,etal., 2007b.Polysaccharideintercellularadhesinorproteinfactorsinbiofilm accumulationofStaphylococcusepidermidisandStaphylococcusaureusisolated fromprosthetichipandkneejointinfections.Biomaterials28(March(9)), 1711–1720.
Rosenthal,V.D.,Maki,D.G.,Mehta,Y.,Leblebicioglu,H.,Memish,Z.A.,Al-Mousa, H.H.,etal.,2014.InternationalNosocomialInfectionControlConsortium (INICC)report,datasummaryof43countriesfor2007–2012.
Device-associatedmodule.Am.J.Infect.Control42(9),942–956.
Rupp,M.E.,Ulphani,J.S.,Fey,P.D.,Bartscht,K.,Mack,D.,1999a.Characterizationof theimportanceofpolysaccharideintercellularadhesin/hemagglutininof Staphylococcusepidermidisinthepathogenesisofbiomaterial-basedinfection inamouseforeignbodyinfectionmodel.Infect.Immun.67(May(5)), 2627–2632.
Rupp,M.E.,Ulphani,J.S.,Fey,P.D.,Mack,D.,1999b.Characterizationof
Staphylococcusepidermidispolysaccharideintercellularadhesin/hemagglutinin inthepathogenesisofintravascularcatheter-associatedinfectioninarat model.Infect.Immun.67(May(5)),2656–2659.
Schaeffer,C.R.,Woods,K.M.,Longo,G.M.,Kiedrowski,M.R.,Paharik,A.E.,Büttner, H.,etal.,2014.Accumulation-associatedprotein(Aap)enhances
Staphylococcusepidermidisbiofilmformationunderdynamicconditionsandis requiredforinfectioninaratcathetermodel.Infect.Immun.,IAI–14. Schaeffer,C.R.,Woods,K.M.,Longo,G.M.,Kiedrowski,M.R.,Paharik,A.E.,Buttner,
H.,etal.,2015.Accumulation-associatedproteinenhancesStaphylococcus epidermidisbiofilmformationunderdynamicconditionsandisrequiredfor infectioninaratcathetermodel.Infect.Immun.83(January(1)),214–226. Scherr,T.D.,Heim,C.E.,Morrison,J.M.,Kielian,T.,2014.Hidinginplainsight:
interplaybetweenStaphylococcalbiofilmsandhostimmunity.Front. Immunol.5(February),37,http://dx.doi.org/10.3389/fimmu.2014.00037. eCollection@2014.
Schommer,N.N.,Christner,M.,Hentschke,M.,Ruckdeschel,K.,Aepfelbacher,M., Rohde,H.,2011.Staphylococcusepidermidisusesdistinctmechanismsof biofilmformationtointerferewithphagocytosisandactivationofmouse macrophage-likecells774A.1.Infect.Immun.(March).
Stoodley,P.,BraxtonJr.,E.E.,Nistico,L.,Hall-Stoodley,L.,Johnson,S.,Quigley,M., etal.,2010.DirectdemonstrationofStaphylococcusbiofilminanexternal ventriculardraininapatientwithahistoryofrecurrentventriculoperitoneal shuntfailure.Pediatr.Neurosurg.46(August(2)),127–132.
Theis,T.,Skurray,R.A.,Brown,M.H.,2007.Identificationofsuitableinternal controlstostudyexpressionofaStaphylococcusaureusmultidrugresistance systembyquantitativereal-timePCR.J.Microbiol.Methods70(August(2)), 355–362.
Untergasser,A.,Cutcutache,I.,Koressaar,T.,Ye,J.,Faircloth,B.C.,Remm,M.,etal., 2012.Primer3–newcapabilitiesandinterfaces.NucleicAcidsRes.40(August (15)),e115.
Vuong,C.,Voyich,J.M.,Fischer,E.R.,Braughton,K.R.,Whitney,A.R.,DeLeo,F.R., etal.,2004a.Polysaccharideintercellularadhesin(PIA)protectsStaphylococcus epidermidisagainstmajorcomponentsofthehumaninnateimmunesystem. Cell.Microbiol.6(March(3)),269–275.
Vuong,C.,Kocianova,S.,Voyich,J.M.,Yao,Y.,Fischer,E.R.,DeLeo,F.R.,etal.,2004b. Acrucialroleforexopolysaccharidemodificationinbacterialbiofilm formation,immuneevasion,andvirulence.J.Biol.Chem.279(December(52)), 54881–54886.
Wisplinghoff,H.,Bischoff,T.,Tallent,S.M.,Seifert,H.,Wenzel,R.P.,Edmond,M.B., 2004.NosocomialbloodstreaminfectionsinUShospitals:analysisof24,179 casesfromaprospectivenationwidesurveillancestudy.Clin.Infect.Dis.39(3), 309–317.
12 2. Zusammenfassende Darstellung der Publikation
Einleitung
Staphylococcus epidermidis stellt einen großen Anteil der residenten Hautflora und ist
unter physiologischen Bedingungen harmlos (Kloos and Musselwhite, 1975). Allerdings ist S. epidermidis, gerade durch seine weite Verbreitung und den häufigen Einsatz von Fremdkörpern (zentrale Venenkatheter oder Totalendoprothesen) in der Medizin, als häufiger Erreger von Fremdkörper-assoziierten Infektionen von großer klinischer Bedeutung (Otto, 2009). So ist S. epidermidis die häufigste Ursache für Katheterinfektionen (Rogers et al., 2009), nach S. aureus die zweithäufigste Ursache für Herzklappenprothesen-Infektionen (Chu et al., 2009) und maßgeblich an Liquorshunt-, Schrittmacher-, Gefäßprothesen- sowie Gelenkprothesen-Infektionen beteiligt (Rogers et al., 2009). Diese sind oft auch mit Reserveantibiotika (wie Vancomycin, Linezolid, Daptomycin und Tigecyclin) schwer zu therapieren und erfordern nicht selten eine Explantation des infizierten Fremdkörpers. Diesem Umstand liegt die Fähigkeit von S. epidermidis zu Grunde feste Oberflächen als Biofilm zu kolonisieren, d.h. eine organisierte Struktur von mehreren Schichten fest haftender Bakterien zu bilden, die in einer extrazellulären Matrix eingebettet sind (Stoodley et al., 2010, Buttner et al., 2015). Für die meisten Bakterien stellt der bakterielle Biofilm in ihrem natürlichen Habitat die bevorzugte Lebensform dar (Costerton et al., 1995). Funktionell können 5 Phasen der Biofilmbildung unterschieden werden (Abbildung 1). Der planktonischen, frei schwebenden Lebensform folgt die primäre Adhäsion vereinzelter Bakterienzellen an Oberflächen, hauptsächlich aufgrund ihrer Hydrophobizität oder in vivo durch Bindung an wirtseigene Matrixproteine, welche Fremdkörperoberflächen nach kürzester Zeit überziehen (Otto, 2009). Im weiteren Verlauf kommt es durch Zellteilung zur Ausbildung von Mikrokolonien sowie durch Expression extrazellulärer Matrixproteine, welche Zell-Zell Interaktionen ermöglichen, zur Reifung eines stabilen Biofilmes (Flemming and Wingender, 2010, Buttner et al., 2015). In einer solchen Lebensgemeinschaft hat der Großteil der Bakterienzellen den Kontakt zur Oberfläche verloren und ist mit veränderten Lebensbedingungen, zum Beispiel im Hinblick auf das Nährstoffangebot konfrontiert. Die letzte Phase beschreibt die Möglichkeit vereinzelter Bakterien, sich aus dem Verband zu lösen und über die planktonische Form Absiedlungen zu bilden, wodurch das Modell zu einem Kreislauf geschlossen wird (Monds and O'Toole, 2009).
13
Abbildung 1 Phasen der bakteriellen Biofilmbildung.
Planktonische Bakterienzellen (1) haften an einer durch wirtseigene Matrixproteine konditionierten Fremdkörperoberfläche und vollziehen somit zunächst die primäre Adhäsion (2). Im weiteren Verlauf entstehen durch Entwicklung einer extrazellulären Matrix und Zellteilung Mikrokolonien (3) und nach Ausbildung einer dreidimensionalen Biofilmarchitektur Makrokolonien (4). Durch die Möglichkeit von Bakterienzellen sich aus dem reifen Biofilm abzulösen, wird das Modell zu einem Kreislauf geschlossen.
Die Architektur des reifen Biofilmes ist von der Bakterienspezies und den Lebensbedingungen abhängig, S. epidermidis bildet auf festen Oberflächen entweder pilzförmige Türme oder Rasen-ähnliche Strukturen (Rani et al., 2005, Okajima et al., 2006).
Über ein Zusammenwirken verschiedener Mechanismen wie niedrige Teilungsraten, Veränderungen des Metabolismus, die Etablierung von Persister-Zellen und extrazellulärer Biofilmmatrix sowie der Unfähigkeit von antimikrobiellen Substanzen in Biofilme einzudringen, erhöht dieser Wachstumsmodus die Antibiotikatoleranz (Romling and Balsalobre, 2012).
Die extrazelluläre Matrix ist aus Polysacchariden, Proteinen, eDNA und Teichonsäuren aufgebaut und ist wichtige Voraussetzung für die Ausbildung eines stabilen Biofilmes. Als beteiligte Faktoren wurden bisher vor allem das polysaccharide intercellular
adhesin (PIA) (Mack et al., 1996), das accumulation associated protein (Aap) (Rohde
et al., 2005) und das extracellular matrix binding protein (Embp) (Christner et al., 2010) identifiziert, welche voneinander unabhängige Faktoren der Biofilmbildung darstellen.
14 Ziel der Arbeit
Der S. epidermidis Stamm 1585 ist ein klinisches Isolat einer besiedelten Portkatheterspitze und bildet trotz seiner klinischen Relevanz unter Standard- Laborbedingungen keinen Biofilm aus. Ziel der vorliegenden Arbeit war die Identifikation biofilminduzierender Effekte von Antibiotika auf S. epidermidis 1585 sowie der dafür erforderlichen Bedingungen. Um den genauen Mechanismus zu untersuchen sollte der resultierende Biofilm insbesondere im Hinblick auf das beteiligte Adhäsionsprotein molekularbiologisch analysiert werden. Eine dreidimensionale Darstellung dieser organisierten Bakterienzelladhäsion mitsamt seiner extrazellulären Matrix sowie die Untersuchung dieses Gefüges auf eine entstandene Phagozytoseresistenz als beispielhafter erworbener Resistenzmechanismus von Biofilmen gegen das wirtseigene Immunsystem sollte die Arbeit komplettieren.
Material und Methoden
Die verwendeten Bakterienstämme und entsprechende relevante Eigenschaften sind in Tabelle 1 dargestellt. Eine Stammhaltung erfolgte auf Columbia Blutagarplatten.
Tabelle 1 Verwendete Bakterienstämme.
S. epidermidis Stamm Beschreibung Referenz
1585 Klinisches Isolat einer Port-Katheterinfektion. Biofilm-negativer Phänotyp in TSB. embp-positiver, icaADBC-negativer Genotyp.
Rohde 2005 Christner 2010 1585v Biofilm-positive Subpopulation von S. epidermidis 1585,
Biofilmbildung durch Embp-Expression.
Christner 2010 1585-M135 Biofilm-negative Mutante von S. epidermidis 1585v, trägt
ein Tn917 in Embp
Christner 2010 12 Klinisches Wildtyp-Isolat aus einer
Endoprothesen-infektion. Biofilm-negativer Phänotyp in TSB. icaADBC und embp-positiver Genotyp.
Rohde 2007
Biofilmassay. Zunächst wurde der Einfluss verschiedener Antibiotika auf das Bakterienwachstum und die Biofilmbildung untersucht indem die Stämme S.
epidermidis 1585 und M135 mit Verdünnungsreihen eines Antibiotikums (0–10 mg/L
Tigecyclin, 0–10 mg/L Vancomycin oder 0–1 mg/L Penicillin) für 22 Stunden bei 37°C in Trypticase Sojabrühe (TSB) inkubiert wurden. Anschließend erfolgte die Bestimmung des Zellwachstums durch Messung der optischen Dichte bei 600nm
15 (OD600) sowie der Dicke des Biofilmes durch Messung bei 570nm (OD570). Um den
Einfluss der Bakterienwachtumsphase auf die Biofilminduktion zu ermitteln, wurde in weiteren Experimenten das Antibiotikum erst nach 12 Stunden, also in der stationären Phase zu den Bakterien gegeben.
Dot-blot Immunoassay. Um eine Veränderung der Embp-Expression im Biofilm nachzuweisen wurden die Stämme S. epidermidis 1585 und M135 für 22 Stunden bei 37°C in Zellkulturschalen mit TSB sowie Tigecyclin (0,0, 0,625 und 1,25 mg/L) inkubiert. Anschließend wurden die Proteine der adhärenten Bakterien präpariert und Dot-blot Immunoassays unter Verwendung von Embp-Antikörper angefertigt.
Quantitative real-time PCR Analyse (qPCR). Für eine quantitative Untersuchung von Veränderungen der Embp-Expression auf Transkriptionsebene wurde eine qPCR der Stämme S. epidermidis 1585, 1585v (Positivkontrolle) und M135 (Negativkontrolle) nach Wachstum in TSB/Tigecyclin durchgeführt und das relative Expressionslevel berechnet.
Konfokale Lasermikroskopie (CLSM). Zur 3D Darstellung des S. epidermidis Biofilms samt seiner Embp Verteilung wurde der Stamm 1585xpCM29 für 20 Stunden in Zellkulturschalen mit TSB oder TSB und Tigecyclin (0,6 mg/L) inkubiert. Nach Entfernen des Mediums, Waschen und Blocken wurde Embp durch Embp-Antikörper und Alexa568 gekoppelte IgG dargestellt. Mit einem konfokalen Lasermikroskop wurden dreidimensionale Bilder der Biofilme aufgenommen, wobei sich die Bakterienzellen des verwendeten Stammes durch das Exprimieren von Gfp darstellten und Embp, wie oben beschrieben, durch Alexa568 dargestellt wurde.
Makrophagen-Interaktion. Die Untersuchung der Bakterienzellaufnahme durch Makrophagen wurde ähnlich wie zuvor durch (Schommer et al., 2011) beschrieben durchgeführt mit der Ausnahme, dass durch die Verwendung des Gfp-exprimierenden Stammes 1585xpCM29 eine immunologische Detektion intrazellulärer Bakterien unnötig war. Murine Makrophagen vom Typ J774A.1 wurden für 2 Stunden mit S.
epidermidis 1585xpCM29, welcher zuvor mit oder ohne Tigecyclin kultiviert wurde,
exponiert. Nach Fixieren und Waschen wurden extrazelluläre Bakterien durch anti-S.
epidermidis Antikörper und IgG gekoppeltes Cy5 markiert. Mit einem konfokalen
Lasermikroskop konnten nun Bilder generiert werden, in welchen die intrazellulären Bakterienzellen ausgezählt wurden.
16 Ergebnisse und Diskussion
Fremdkörperassoziierte Infektionen mit S. epidermidis werden im klinischen Alltag aufgrund der häufigen Resistenzen gegen β-lactam Antibiotika (Mack et al., 2009) meist mit Vancomycin therapiert. Ist aufgrund von Nephrotoxizität oder Überempfindlichkeitsreaktionen Vancomycin kontraindiziert, wird auf Wirkstoffe wie Linezolid, Daptomycin oder Tigecyclin zurückgegriffen (John et al., 2009, Aybar et al.). Bisher wurden Nebenwirkungen im Zusammenhang mit Biofilm induzierenden Eigenschaften dieser Wirkstoffe wenig untersucht, es gibt aber Hinweise auf einen solchen Effekt durch subinhibitorische Vancomycin Konzentrationen (Kaplan et al., 2011). Im Rahmen der genaueren Untersuchung biofilminduzierender Effekte von Reserveantibiotika wurde in der vorliegenden Arbeit speziell die Wirkung von Linezolid, Daptomycin und Tigecyclin auf ein klinisches Isolat von S. epidermidis 1585 aus einer Port-Katheterinfektion untersucht (Rohde et al., 2005, Christner et al., 2010, Christner et al., 2012). Trotz seiner klinischen Relevanz bildet S. epidermidis unter Standard-Laborbedingungen keinen Biofilm aus (Christner et al., 2010, Christner et al., 2012). Die meisten der unter Laborbedingungen Biofilm-negativen Stämme sind trotz des Fehlens des icaADBC Operons als Vorraussetzung für eine PIA vermittelte Biofilmbildung dennoch in der Lage, als Reaktion auf externe Stimuli (z.B. Antibiotika) Biofilme über PIA unabhängige Mechanismen auszubilden (Rohde et al., 2005, Christner et al., 2010, Kaplan et al., 2011). Dieser Vorgang unterliegt der Regulation des SarA Gens (Staphylococcal accessory regulator).
Die Resistenztestung mittels E-Test ergab, dass S. epidermidis 1585 resistent gegen Oxacillin, jedoch sensibel für Penicillin, Vancomycin, Daptomycin, Linezolid und Tigecyclin ist. Verschiedene Konzentrationen (0.005 – 5 mg/L) von Penicillin, Vancomycin, Linezolid oder Daptomycin hatten keinen Einfluss auf den Biofilmphänotyp (Abbildung 2). Dies steht im Gegensatz zu früheren Studien die einen biofilminduzierenden Effekt von Vancomycin auf S. epidermidis 9142, 1457 und RP62A zeigen konnten (Kaplan et al., 2011, Cargill and Upton, 2009, Claessens et al., 2015), was die Vermutung zulässt, dass der Einfluss von Antibiotika auf die Biofilmbildung vom Genotyp des jeweiligen Stammes abhängt (Kaplan et al., 2011). Damit in Übereinstimmung konnten wir einen biofilminduzierenden Effekt von 1,25 μg/ml Vancomycin auf den genetisch unabhängigen Stamm S. epidermidis 12 (einem klinischen Isolat einer TEP-Infektion) zeigen (Rohde et al., 2007).
17 Insbesondere konnten wir auch Tigecyclin in einem schmalen Konzentrationsbereich zwischen 0,156 und 0,625 mg/L in S. epidermidis 1585 eine 2-3 fache Biofilmbildung nachweisen (Abbildung 2). Bei genauerer Betrachtung des Bakterienwachstums zeigte sich, dass der Phänotyp-Switch insbesondere am Übergang von inhibitorischen zu sub-inhibitorischen Tigecyclin-Konzentrationen auftrat und sich bei weiterer Verringerung der Konzentration wieder verlor (Abbildung 2).
Abbildung 2 Biofilminduktion in S. epidermidis 1585 unter Anwesenheit von Antibiotika.
Nach 20 Stunden Wachstum in TSB unter Anwesenheit verschiedener Antibiotikakonzentrationen wurde die Zelldichte durch Messung der Absorption bei 600nm (rote Kästchen) und die Biofilmbildung nach Färbung adhärenter Zellen mit Gentianaviolett bei 570nm (blaue Punkte) gemessen. Die gepunktete Linie zeigt die Biofilmbildung von S. epidermidis 1585 in TSB ohne Antibiotika an. Jeder Datenpunkt repräsentiert den Mittelwert von 12 Werten aus drei unabhängigen Experimenten.
18 Tigecyclin gehört zur Gruppe der Gycylcycline und wirkt ebenso wie die verwandten
(herkömmlichen) Tetrazykline über eine Hemmung der ribosomalen
Proteinbiosynthese bakteriostatisch bei jedoch geringerer Anfälligkeit für bakterielle Resistenzmechanismen. Erstaunlicherweise konnten wir für Tetrazyclin keinen Einfluss auf die Biofilmbildung in S. epidermidis 1585 nachweisen, was auf einen Tigecyclin-spezifischen Effekt schließen lässt. Durch subinhibitorische Tigecyclin- Konzentrationen konnte auch in S. epidermidis 12 die Biofilmbildung induziert werden.
Bisher wurde die Biofilminduktion von subinhibitorischen Antibiotikakonzentrationen der Hochregulation von icaADBC und der darauffolgenden PIA-Synthese zugeschrieben (Gomes et al., 2011, Chang et al., 2010, Rachid et al., 2000), wobei mit der Betrachtung von überwiegend icaADBC-positiven Stämmen mögliche PIA unabhängige Mechanismen der Zellaggregation weitestgehend vernachlässigt worden sind. Es gibt bereits Hinweise, dass der biofilminduzierende Effekt von Vancomycin PIA-vermittelt wirkt, während Tigecyclin diesen Effekt höchstwahrscheinlich über einen PIA unabhängigen, eDNA-abhängigen, Mechanismus hervorruft (Kaplan et al., 2011, Cargill and Upton, 2009). Vorangegangene Studien der hiesigen Arbeitsgruppe zeigten, dass S. epidermidis 1585 in der Lage ist einen mehrschichtigen Zellverband anhand von Embp auszubilden (Christner et al., 2010, Christner et al., 2012).
Embp ist ein riesiges, etwa 1 Megadalton (MDa) großes Protein, welches eine extrazelluläre Matrix für Aufnahme und Zusammenhalt von Bakterienzellen bildet (Schommer et al., 2011). In klinischen S. epidermidis Isolaten ist das Embp Gen weit verbreitet und kann in fast jedem Stamm nachgewiesen werden (Rohde et al., 2007, Rohde et al., 2004). Unter Standard-Laborbedingungen wird embp nur sehr gering exprimiert (Abbildung 3A) (Christner et al., 2010), da diese bei Wachstum in TSB durch
SarA supprimiert wird (Christner et al., 2012) wodurch sich erklären lässt warum icaADBC-negative, Embp-positive S. epidermidis Isolate unter
Standard-Labor-bedingungen keinen Biofilm ausbilden (Rohde et al., 2007).
Aufbauend auf diesen Ergebnissen stellten wir die Hypothese auf, dass die Tigecyclin induzierte Biofilmbildung funktionell mit Embp zusammenhängt. Um diese Hypothese zu untersuchen wurde die Embp-Expression zunächst in An- oder Abwesenheit von Tigecyclin in Dot Blot Immunassays (DIA) mit einem Antiserum gegen eine Embp Subdomäne untersucht (Christner et al., 2010). Hierbei ließ sich nach 24 Stunden
19 Wachstum in TSB keine oder nur geringe Embp-Synthese nachweisen (Abbildung 3A), während in scharfem Kontrast dazu die Anwesenheit von 0,625 mg/L Tigecyclin zu einer 8-fachen Erhöhung der Embp-Synthese führte (Abbildung 3A).
Auch die Untersuchung der Biofilmbildung und Embp-Expression von S. epidermidis 1585 mittels konfokaler Lasermikroskopie bestätigte diese Ergebnisse nicht nur durch einen visuellen Nachweis der Biofilmbildung unter Tigecyclin Einfluss, sondern konnte sogar eine extrazelluläre Matrix aus Embp darstellen (Abbildung 3B).
Darüber hinaus konnte die Embp-Induktion auch durch quantitative real-time PCR (Christner et al., 2010) nachgewiesen werden, wobei die Anwesenheit von Tigecyclin 3,7-fach erhöhte Expressionslevel hervorrief, während die sarA Expression zugleich 3,9-fach herabreguliert war. Die Anhäufung von Embp ist somit durch Veränderungen auf Transkriptionsebene zu erklären und nicht durch langsamen Zellumsatz z.B. infolge Herunterregulation von extrazellulären Proteasen. Die Embp Hochregulation ist am ehesten als eine Konsequenz der sarA Herabregulation durch subinhibitorische Tigecyclin-Konzentrationen zu sehen (Christner et al., 2012).
Auch S. epidermidis 12 konnte unter subinhibitorischen Tigecyclinkonzentrationen eine 3,7-fache Heraufregulation von embp und eine damit einhergehende Induktion der Embp-Synthese nachgewiesen werden. Zugleich war in diesem Rahmen keine relevante icaA-Regulation zu beobachten, was darauf hindeutet, dass in S. epidermidis 12 PIA nicht an der Tigecyclin induzierten Biofilmbildung beteiligt ist.
Erstaunlicherweise konnte Embp nur in der frühen bis mittleren exponentiellen Wachstumsphase von S. epidermidis 1585 induziert werden, während eine Exposition mit Tigecyclin nach 20 Stunden Wachstum in TSB keinen Einfluss hatte. Da auch die Biofilmbildung ausschließlich in der frühen bis mittleren exponentiellen Wachstumsphase induziert werden konnte, stellt sich der Verdacht auf einen funktionellen Zusammenhang zwischen Tigecyclin induzierter embp-Expression und Biofilmbildung. Zur Überprüfung dieser Hypothese wurde die isogene mutante 1585-M135, welche eine Tn917 Insertion in embp trägt, verwendet (Christner et al., 2010). Die signifikant herabgesetzte Biofilmbildung der Mutante 1585-M135 unter subinhibitorischen Tigecyclin-Konzentrationen zeigt, dass die Tigecyclin induzierte Biofilmbildung durch S. epidermidis 1585 zumindest teilweise Embp abhängig ist Abbildung 3C).
20
Abbildung 3 Detektion der Embp Produktion unter Anwesenheit von Antibiotika.
(A) Zelloberflächen-adhärente Proteine wurden aus S. epidermidis 1585 nach Wachstum unter den
angegebenen Antibiotikakonzentrationen präpariert. Verdünnungseihen dieser Präparationen wurden auf vorhandenes Embp mittels Dotblot Immunoassays unter Verwendung von Anti-Embp Antikörper untersucht. UD= unverdünnt. (B) Untersuchung der S. epidermidis 1585 Biofilmbildung mittels Konfokaler Lasermikroskopie. Gfp-exprimierender S. epidermidis 1585xpCM29 (grün) wurde in TSB versus TSB + Tigecyclin kultiviert. Embp wurde Antikörper vermittelt markiert (rot). (C) Relevanz von Embp für die Tigecyclin und Oxacillin induzierte Biofilmbildung. S. epidermidis 1585 und die entsprechende isogene Transposon-Mutante 1585-M135 wurden unter Tigecyclin oder Oxacillin kultiviert. Im Anschluss wurde die Biofilmbildung quantifiziert. Jeder Balken repräsentiert 12 Werte aus drei unabhängigen Experimenten. Die Mittelwerte unterscheiden sich signifikant (** p < 0,01; * p < 0,05).
21 Die verbliebene Biofilmbildung unter diesen Bedingungen (Abbildung 3C) deutet auf weitere interzelluläre Mechanismen hin, die adhäsives Wachstum ermöglichen, wie zum Beispiel die Freisetzung von eDNA (Qin et al., 2007). eDNA könnte eine wichtige Rolle im Zusammenhang mit gesteigerter Antibiotikatoleranz darstellen, da bereits gezeigt werden konnte, dass eDNA Antibiotika in der Biofilmmatrix binden kann (Mulcahy et al., 2008) und die Sensibilität von biofilmbildenden S. epidermidis gegenüber Vancomycin verringert (Doroshenko et al., 2014). Da Hemmstoffe der Proteinbiosynthese in E. coli Persister-Zellen hervorrufen können (Kwan et al., 2013), ist die Vermutung naheliegend dass Tigecyclin induzierte Biofilmbildung einen ähnlichen Effekt auf S. epidermidis hat.
Das Wachstum in Biofilmen schützt S. epidermidis nicht nur vor Antibiotika, sondern auch vor Effektormechanismen des unspezifischen Immunsystems (Schommer et al., 2011, Vuong et al., 2004b, Vuong et al., 2004a). Diesen erworbenen Resistenzmechanismus wollten wir auch für den Tigecyclin-induzierten Biofilm untersuchen, in dem dieser murinen Makrophagen vom Typ J774A.1 ausgesetzt wurde. Nach zwei Stunden konnte durch den fluoreszenzmikroskopischen Vergleich der intrazellulären Bakterienzellen nachgewiesen werden, dass diese nach Wachstum in subinhibitorischer Tigecyclinkonzentration signifikant schlechter phagozytiert wurden als Bakterienzellen ohne vorherigen Tigecyclin-Kontakt (Abbildung 4).
Da Tigecyclin selbst in hohen Konzentrationen nur einen geringen Effekt auf die Aufnahme von Bakterien durch Phagozyten zu haben scheint (Naess et al., 2011), deutet das Ergebnis darauf hin, dass die Phagozytose tatsächlich durch den Tigecyclin-induzierten Biofilm beeinträchtigt wird. Somit wird durch diesen Wachstumsmodus über die Ausbildung von Mechanismen zur Immunevasion die Persistenz der Bakterien gesteigert. Aufgrund der erschwerten Verteilung von Antibiotika in infizierte Gewebe (Garazzino et al., 2008, Rodvold et al., 2006), kann davon ausgegangen werden, dass S. epidermidis im klinischen Alltag häufig sub-inhibitorischen Konzentrationen ausgesetzt ist, wodurch Biofilmbildung und letztendlich eine erhöhte Resistenz gegen antimikrobielle Wirkstoffe und das wirtseigene Immunsystem ausgebildet wird.
22
Abbildung 4 Phagozytose von S. epidermidis 1585 durch murine Makrophagen vom Typ J774A.1.
Gfp synthetisierender S. epidermidis 1585xpCM29 wurde in TSB versus TSB + Tigecyclin kultiviert. 2h nach Inkubation mit murinen Makrophagen vom Typ J774A.1 wurden die extrazellulären Zellen Antikörper vermittelt markiert und phagozytierte Bakterienzellen von 15 Makrophagen ausgezählt. Der Mittelwert phagozytierter Bakterien war signifikant (p < 0,001) höher in Abwesenheit von Tigecyclin.
Fazit
In der vorliegenden Studie konnten wir eine paradoxe Wirkung von Tigecyclin auf S.
epidermidis zeigen, welche bei subinhibitorischen Konzentrationen die Biofilmbildung
induzieren konnte. In molekularbiologischen Verfahren konnte ein Zusammenhang mit der Hochregulation der Embp-Synthese festgestellt werden. Damit einher ging auch eine gesteigerte Phagozytoseresistenz, woraus zu schließen ist, dass das untersuchte Phänomen einen abträglichen Effekt auf den Verlauf von Fremdkörper assoziierten Infektionen haben kann. Die Ergebnisse der Studie verdeutlichen die Wichtigkeit des genauen Verständnisses der an der Biofilmbildung beteiligten Komponenten für die zukünftige Entwicklung von Präventions- und Therapieregimes im Umgang mit
Fremdkörper-assoziierten Infektionen, an deren Ende möglicherweise
biofilmauflösende Enzyme stehen könnten.
Bisher wurde der Einfluss von antimikrobiellen Wirkstoffen auf die Interaktion zwischen Wirt und Pathogen von der Wissenschaft nicht ausreichend untersucht. Weitere Studien sind notwendig um auch die Signaltransduktionswege und regulatorische Netzwerke aufzuklären, die der Expression von extrazellulären Proteinen im Rahmen der Biofilmbildung zugrunde liegen. Des Weiteren ist immer noch nicht vollständig geklärt, wie das 1MDa große Protein Embp in den Extrazellularraum gelangen kann.