Development of fluorescence expression tools to study host-mycoplasma interactions and validation in two distant mycoplasma clades

(1)

Contents lists available atScienceDirect

Journal

of

Biotechnology

j o u r n a l h o m 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 / j b i o t e c

Development

of

ﬂuorescence

expression

tools

to

study

host-mycoplasma

interactions

and

validation

in

two

distant

mycoplasma

clades

Tiffany

Bonnefois

a,b

_,

_{Marie-Stéphanie}

_Vernerey

c

_,

_Valérie

_Rodrigues

a,b

_,

_Philippe

_Totté

a,b

_,

Carinne

Puech

a,b

_,

_Chantal

_Ripoll

d

_,

_Franc¸

_ois

_Thiaucourt

a,b

_,

_Lucía

_{Manso-Silván}

a,b,∗

a_CIRAD,_UMR_CMAEE,_F-34398_Montpellier,_France

b_INRA,_UMR1309_CMAEE,_F-34398_Montpellier,_France

c_INRA,_Joint_Research_Unit₃₈₅_UMR_BGPI,_Campus_{International}_de_Baillarguet,_Montpellier,_France

d_INSERM_{U1051—Hôpital}_Saint_Eloi_INM._80,_rue_Augustin_Fliche,₃₄₀₉₁_Montpellier_cedex_5,_France

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received28June2016

Receivedinrevisedform27July2016

Accepted4August2016

Availableonline4August2016

Keywords: Mycoplasma Fluorescence

Wholecelllabelling

Host-pathogeninteractions

Flowcytometry

Confocalmicroscopy

a

b

s

t

r

a

c

t

Fluorescenceexpressiontoolsforstableandinnocuouswholemycoplasmacelllabellinghavebeen devel-oped.ATn4001-derivativemini-transposonaffordingunmarked,stablemutagenesisinmycoplasmas wasmodifiedtoallowtheconstitutive,high-levelexpressionofmCherry,mKO2andmNeonGreen. Thesetoolswereusedtointroducetherespectivefluorescentproteinsaschromosomaltagsinthe phylogeneticallydistantspeciesMycoplasmamycoidessubsp.mycoidesandMycoplasmabovis.The pro-duction,selectionandcharacterisationoffluorescentcloneswerestraightforwardandresultedinthe unprecedentedobservationofredandgreenfluorescentmycoplasmacoloniesinthetwospecies,with noapparentcytotoxicity.Equivalentfluorescenceexpressionlevelswerequantifiedbyflowcytometry inbothspecies,suggestingthatthesetoolscanbebroadlyappliedinmycoplasmas.Amacrophage infec-tionassaywasperformedtoassesstheusefulnessofmNeonGreen-expressingstrainsformonitoring mycoplasmainfections,andnotablycellinvasion.Thepresenceoffluorescentmycoplasmasinsidelive phagocyticcellswasdetectedandquantifiedbyflowcytometryandcorroboratedbyconfocalmicroscopy, whichallowedtheidentificationofindividualmycoplasmasinthecytoplasmofinfectedcells.The fluo-rescenceexpressiontoolsdevelopedinthisstudyaresuitableforhost-pathogeninteractionstudiesand offerinnumerableperspectivesforthefunctionalanalysisofmycoplasmasbothinvitroandinvivo.

1. Introduction

Theclass Mollicutesconstitutes a groupof wall-lessbacteria witha small genome size,commonly referred to as mycoplas-mas, which evolved from gram-positive bacteria with low GC

Abbreviations: cfu,colonyformingunits;GFP,greenﬂuorescentprotein;MFI,

medianﬂuorescenceintensity;mRFP1,monomericredﬂuorescentprotein;MOI,

multiplicityofinfection;Mmm,Mycoplasmamycoidessubsp.mycoides;PBS,

phos-phatebufferedsaline;PMT1,photomultipliertube1;PFA,paraformaldehyde;PI,

propidiumiodide;RT,roomtemperature;VSD,variablesecondarydichroic.

∗ Correspondingauthorat:CIRAD,UMRCMAEE,F-34398Montpellier,France.

E-mailaddresses:tiffany.bonnefois@gmail.com(T.Bonnefois),

marie-stephanie.vernerey@supagro.inra.fr(M.-S.Vernerey),

valerie.rodrigues@cirad.fr(V.Rodrigues),philippe.totte@cirad.fr

(P.Totté),carinne.puech@cirad.fr(C.Puech),chantal.ripoll@inserm.fr

(C.Ripoll),francois.thiaucourt@cirad.fr(F.Thiaucourt),lucia.manso-silvan@cirad.fr

(L.Manso-Silván).

contentthrough a processof massivegenomereduction (Razin etal.,1998).Sincetheyrepresentthesmallestorganismscapableof autonomousreplication,theyhaveservedasmodelsforthe deﬁni-tionoftheminimalgenesetrequiredforindependentlife,leading tothegenerationofa“minimal”artiﬁcialcellbysyntheticbiology (Hutchisonetal.,2016).Through thecourseofreductive evolu-tionmycoplasmaslostmanymetaboliccapabilitiesandadaptedto a commensalorparasiticmodeoflife.Asaconsequence,many speciesare knownas importantpathogensof humans,animals, andplants.However,despitetheapparentsimplicityoftheirsmall genomes,mycoplasmashavedevelopedsophisticatedmechanisms for colonisation and persistence in theirhost (Browninget al., 2014).Still,thesemechanismsremainvirtuallyunknownformost speciesand,althoughincreasingamountsofgenomesequencedata haveaccumulatedoverthelasttwodecades,thefunctionalanalysis ofmycoplasmashaslaggedfarbehindthatofothereubacteria.

http://dx.doi.org/10.1016/j.jbiotec.2016.08.006

(2)

Theanalysisofbacterialgeneexpressionandfunctionis depen-dentontheavailabilityof toolsforthegeneticmanipulation of thesebacteria,typicallyallowingtheproductionofmutantsand theexpressionofgeneproductssuchasselectionmarkers,reporter systemsorspecificgenesformutantcomplementation.Yet,this hasbeenhinderedbytheincompatibilityofvectorsengineered inconventionalbacteria,togetherwithdifficultiesencounteredin thedevelopmentofspecificgenetictoolsforthesepeculiar organ-isms(Renaudinet al.,2014).Significantlimiting factorsinclude thescarcityofnaturalplasmidsandselectiveantibioticresistance markersandtheslow,fastidiousgrowthofcultivablemycoplasmas. However,setting aside thelatestdevelopments in synthetic biology allowingtheengineering of bacterialgenomes in yeast and their transplantation in recipient cells (Hutchison et al., 2016), numeroustoolsarenow availableforthesimplegenetic manipulationof theselong consideredintractableorganisms. A comprehensivereview ofthesetools,and theircontributionsto thefunctionalgeneticanalysisofmycoplasmas,hasbeenrecently published(Renaudinetal.,2014).Notableexamplesinclude sophis-ticatedsystemsforrandominsertionalmutagenesisderivedfrom the transposon Tn4001 of Staphylococcus aureus. These mini-transposonshavebeenimprovedbyplacingthetransposasegene outside the transposable element, which prevents re-excision andsecondarytransposition,conferringstabilitytotheinsertions (ZimmermanandHerrmann,2005).Anotherinteresting improve-menthasbeenthecombineduseofthesemini-transposonswith the␥d TnpR/res recombinationsystem to excise theantibiotic resistancegenefromthemycoplasma chromosomeafter trans-positionin order toproduce unmarked mutations(Janis et al., 2008).Also, artificialplasmids usingtheoriginof replicationof the mycoplasma chromosome (oriC) have been used to allow genecloningandexpressioninavarietyofmycoplasmaspecies (Renaudin et al., 2014). However, the host range of oriC plas-midsiscircumscribedtocloselyrelatedspeciesandthereisnot asingleself-replicatingvectorforuniversaluseinmycoplasmas. The wide host range of Tn4001 transposition in mycoplasmas hasbeenexploited toovercomethe lackof plasmidvectors in somespeciesandmini-transposonshavebeenmodifiedtoallow genedeliveryandexpression(ZimmermanandHerrmann,2005). Thesesystemshavebeenusedtoexpressfluorescentproteinsin fusion withmycoplasma proteins, serving as reporters of gene expressionandallowingthesubcellularlocalisationofmycoplasma components(Balishetal.,2003;Kenrietal.,2004;Tulumetal., 2014;ZimmermanandHerrmann,2005).However,thesuccessful expressionoffluorescentproteinsinmycoplasmasremainsrare and,toourknowledge,therearenofluorescenceexpression sys-temsavailabletomonitorthedynamicsofmycoplasmainfections invivoorinvitro.

Theaim of this workwas toextendthe array ofmolecular toolsfor usein Mollicutes, notablythrough thedevelopmentof universalfluorescenceexpressionsystemsforwholemycoplasma celllabellingallowingthemonitoringofmycoplasmasduringthe course of infection. The level of fluorescence conferred to the mycoplasmasshould be sufficient toenable their direct detec-tioninsidehostcellsusingaffordabletechniquessuchasconfocal microscopyandflowcytometry,thusallowingtheanalysisofthe mechanismsinvolvedincolonisationand persistenceinsidethe host(e.g.,adhesion,invasion,survival and multiplicationinside hostcells).Forthispurpose,aTn4001-derivativemini-transposon affordingstable,randommutagenesisinavarietyofmycoplasma specieswasmodifiedtoallowtheconstitutive,high-level expres-sion of green and red fluorescent proteins in mycoplasmas. The species Mycoplasma mycoides subsp. mycoides (Mmm) and Mycoplasma bovis, known as important cattle pathogens but belongingto distant phylogenetic groups (Sirand-Pugnet et al., 2007),wereusedtodemonstratethebroadspectrumofthe

ﬂuores-cenceexpressiontoolsforuseinmycoplasmasandamacrophage infectionassaywasperformedtoshowthatﬂuorescent mycoplas-mascouldbedirectlydetectedinsidehostcells,thusdemonstrating theusefulnessofthesenewtoolsforthestudyofhost-mycoplasma interactions.

2. Materialandmethods

2.1. Bacterialstrainsandcultureconditions

CompetentDH10␤-derivativeEscherichiacolistrainNEB 10-beta(New EnglandBiolabs, USA) wasusedfor plasmidcloning and propagation according to the supplier’s instructions, with 100␮g/mLkanamycinaddedtotheculturemediumforselection oftransformants.

MmmtypestrainPG1T_(CIRAD,_France)_was_used_for_PCR ampli-fication of the intergenic region containing the tufA promoter. Thetwopathogenicmycoplasmastrainsusedinthisstudywere bovinelungisolatesfromcasesofpneumoniaandsubjectedtofew invitropassages.Mmmstrain8740-Ritawasisolatedfroma con-tagiousbovinepleuropneumoniacaseinCameroon,in1987(Dr. Aboubakar,LaboratoireNationalVétérinaire,Garoua,Cameroon). M.bovisstrainOger2wasisolatedfromacaseofcalfpneumonia inArdennes,France,in1975(CIRAD,France).Allstrainswere cul-turedat37◦C,5%CO2inmodifiedHayflick’smedium:2.1%PPLO brothwithout crystal violet (Difco, USA), 15% horseserum de-complementedfor1hat56◦C,5%freshbaker’syeastextract,0.2% sodiumpyruvate,0.1%glucose,with1%NobleAgar(Difco,USA) addedforplating.

2.2. Plasmidconstructions 2.2.1. DNAtechniques

PlasmidDNA wasextractedusingeither theWizardPlusSV MiniprepsDNAPurificationSystem(Promega,USA)orthe End-oFree Plasmid Maxi Kit (Qiagen, USA) depending on required DNAyield.Standardmoleculartechniqueswereusedforplasmid constructions,usingthefollowingreagentsaccordingtothe man-ufacturer’sinstructions.Restrictionenzymeswereobtainedfrom NewEnglandBiolabs,USA.CalfIntestinalAlkalinePhosphataseand T4DNAligasewerepurchasedfromInvitrogen,LifeTechnologies, USA.OnemicrogramofplasmidDNAwasdephosphorylatedwith1 unitofCIAPduring5minat37◦C.Ligationwasperformed accord-ing to the manufacturer’s recommendations for blunt cloning. GenomicDNAwasextractedfrombacterialculturesattheendof theexponentialphaseusingtheDNeasyBlood&TissueKit(Qiagen, USA).Alloligonucleotidesusedinthisstudy(Table1)wereobtained fromSigma-Aldrich,France.TominimizemutationscausedbyPCR amplification,highfidelityPhusionHotStartIIDNAPolymerase (ThermoScientific,USA)wasusedforDNAcloningexperiments. Otherwise,TopTaqDNAPolymerase(Qiagen,USA) wasusedfor PCRverificationsand priortosequencing.DNAsequencing was performedbyBeckmanCoulterGenomics,Takeley,UK.

2.2.2. Constructionofﬂuorescenceexpressiontools

Thestrategy forthedevelopmentoffluorescenceexpression toolsispresentedintheresultssection.PlasmidspMT85/2resand pPS3.1werekindlyprovidedbyPascalSirand-Pugnet(INRA, Bor-deaux, France) and all primer sequences are listed in Table 1. pMT85/2res was modified by cloning a cassette constructed usingtheoverlapextensionPCRtechnique.Theentireintergenic sequenceprecedingtheelongationfactorTugene(tufA)followed bythefirst33nucleotidesinits5end,wasamplifiedbyPCRfrom thegenomicDNAofMmmstrainPG1T_using_primers_SpeI_Ptuf_F and NcoIPtuf+R.The transcription terminator sequence ofthe

(3)

Table1

Oligonucleotidesusedinthisstudy.

Name Nucleotidesequence(5-3)

SpeIPtufF TTTACTAGTTATTTTTTGAATTAAGTATTAAATAAGTG

NcoI Ptuf+R TTTCCATGGAGGTAAACTACGGTCAAATTG

MCSPtuf+TﬁbF GTAGTTTACCTCCATGGAAAGTCGACCCGCGGCTTAAGAAATTAAAGTTGGTTCATTC TﬁbXbaIR GGGTCTAGATTTAAAAAATAAAAAACAACCTTTTGG TnpA-F GTTCTTATGGACCTACATGATG pMT85-SpeIcloningF AACCCATAGCTTTGGACACAC Adapt1BclI CTCGTACGTGGGATCGTAACGTT Adapt2BclI GATCAACGTTACGATCCC BclI-Primer CGTGGGATCGTAACGTTGATCA MT85-R CCGCCTTTGAGTGAGCTGATACC

fibrilproteingenefibofSpiroplasmacitriwasamplifiedfrom

plas-midpPS3.1usingprimersMCSPtuf+ TﬁbFandTﬁbXbaIR.The

resultingproductswerehybridisedtoeachotherandextendedby

primer-freeampliﬁcation.AﬁnalPCRwasperformedtoamplify

thislastproductusingprimersSpeIPtufFandTﬁbXbaIR.AllPCR

reactionsconsistedin initialdenaturation at98◦C for 30s,

fol-lowedby35cyclesofdenaturationat98◦Cfor7s,annealingat

57◦Cfor30sandelongationat72◦Cfor30s,ﬁnishedbyaﬁnal

extensionat72◦Cfor5min.Theonlyexceptionwasthe

primer-freeampliﬁcation,forwhichonly12PCRcycleswereperformed.

Theresultingcassettewasdouble-digestedwithSpeIandXbaIand

clonedintheSpeIsiteofpMT85/2res,resultinginplasmidpMT/exp.

TheentireclonedsequencewasveriﬁedbyDNAsequencingwith

primerTnpA-F. ThepMT/expplasmidmap,annotated sequence

andFASTAsequencearepresentedinSuppl.Fig.1.

The gene sequences of mCherry, mKO2 and mNeonGreen

were retrieved from NCBI GenBank database (Table 2). These

nucleotidesequences wereoptimised takinginto consideration the codon usage bias in mycoplasmas (Suppl. Table 1). An in-house bioinformatic tool was developed at CIRAD toautomate codonoptimisation.Thistoolisfreelyavailablefromtheauthors on request. The Codon Usage Database (http://www.kazusa.or. jp/codon/; last accessed 24/06/2016) (Nakamura et al., 2000) was used to analyse the frequency of codon use in the query organism “Mycoplasma mycoides subsp. mycoides SC str. PG1 [gbbct]:1016CDS’s(330,592codons)”.Themostfrequentlyused codon for each amino acid (Suppl. Table 1) was selected for implementation in the bioinformatic tool. Synthetic constructs correspondingtomycoplasma-optimisedgenesequencesofeach ofthethreefluorescentproteinswereproducedbyProteoGenix (Schiltigheim, France).The NcoIandAflII recognitionsiteswere included at the extremities of the different fluorescent genes to allow directional cloning in pMT/exp, resulting in plasmids pMT/mNeonGreen, pMT/mCherry and pMT/mKO2. The cloned insertswereentirelyvalidatedbysequencingwithprimer pMT85-SpeIcloningF(Table1).

2.3. Mycoplasmatransformationandidentiﬁcationoftransposon insertionsites

PEG-mediatedtransformationofmycoplasmastrainswas per-formedasdescribed(Janisetal.,2008)withafewmodifications. TwentymicrogramsofpMT85/2res-derivedplasmidswereusedto transformapproximately109_mycoplasma_cells._Transformed_cells werere-suspendedin500␮Lbrothandincubatedforeither2h(M. bovis)or3h(Mmm)at37◦C,thenplated(100␮L)onsolidHayflick’s mediumsupplementedwithkanamycin(100and200␮g/mLfor MmmandM.bovisrespectively).Ten-folddilutionswerealsoplated (20␮Ldrops)forculturetitrationswithandwithoutantibioticsin ordertocalculatethetransformationefficiency,expressedasthe numberoftransformantcolonyformingunits(cfu)pertotalcfu. Selectedfluorescentcolonieswerepickedandgrowninselective

broth,ﬁlteredthroughaMillexHV0.45␮mDuraporePVDF mem-brane(MerkMillipore,Ireland)and platedonsolid mediumfor cloning.Filter-clonedmutantcultureswerestoredat−80◦_C_for fur-theranalysis.Subsequentcultureswereperformedinnon-selective medium.

TosequencethesiteofthetransposoninsertionintheMmm genome,totalDNAfromselectedcloneswasdigestedwithBclI. Anadaptorwasgeneratedbyhybridisationof100␮Mof oligonu-cleotides“Adapt1BclI”and“Adapt2BclI”(Table1)at100◦Cduring 5min,then ligated toBclI-digestedmycoplasma DNA.The liga-tion productwas usedas template for PCR ampliﬁcationusing primersBclI-PrimerandMT85-R(Table1).Standardreactionswere usedduring35 PCRcycleswithhybridisation at52◦C. ThePCR product wasthen sequencedusing primer MT85-R.The result-ingsequenceswereanalysedusingGeneiousR66.0.1(Biomatters Ltd,NewZealand).Aftereliminationofﬂankingmini-transposon sequences,theinsertionsitesequencewasobtainedbymapping withtheavailablegenomesequenceofMmmPG1T_(GenBank acces-sionno.BX293980.2).

2.4. Macrophageisolationandinfection

BloodsampleswerecollectedfromhealthyJerseycattlekept inananimalhousingfacility(CIRAD,Montpellier,France).Whole blood was collected from the jugular vein in heparinised BD Vacutainer tubes (Beckton Dickinson, USA) according to the manufacturer’s instructions. Experimental procedures for ani-mal maintenance and blood sampling were approved by the Languedoc-Roussillon regional ethics committee (French CE-LR #36) intheAuthorised Projectusing animalsfor scientific pur-poses#12ANI01.Monocyteisolation wasperformedbypositive selectionofCD14+cellsaspreviouslydescribed(Hopeetal.,2003). CD14+cellswerethenseededin6-wellplates,withorwithoutglass coverslips,at3×106_cells/well_and_cultured_in_Iscove’s_Modified Dulbecco’sMedium(LifeTechnologies,France)supplementedwith 2mMl-glutamine,50␮M2-mercaptoethanol,50␮g/mL gentam-icin(LifeTechnologies,France)and10%heatinactivatedfoetalcalf serum(EurobioAbCys,France)for6daysat37◦Cand5%CO2.Half ofthemediumwasreplacedafter3days.After6daysof differentia-tion,thecellswerereferredtoasmonocyte-derivedmacrophages. Aftertwowasheswithphosphatebufferedsaline(PBS)thecells wereinfectedwithwashedmycoplasmasattheendofthe expo-nentialphaseofgrowthusingamultiplicityofinfection(MOI)of around200.Twentypercentv/vofthecorrespondingMmmorM. bovisspecificantiserumwasaddedrespectivelyandthecellswere incubatedfor30minat37◦C.ThecellswerewashedtwicewithPBS beforefurtheranalysis.Thespecificantiserausedforopsonisation consistedeachinpooledserafromtenconvalescentanimals.The M.bovisantisera,originatingfromFrance,wereagiftfromFlorence Tardy(ANSESLyon,France).TheMmmantiserawereoriginated fromCameroon(LaboratoireNationalVétérinaire,Garoua).

(4)

Table2

Characteristicsofselectedﬂuorescentproteins.

Characteristics mCherry mKO2 mNeonGreen

Numberofaminoacids 236 218 236

Quaternarystructure Monomer Monomer Monomer

Excitation/Emissionmaxima(nm) 587/610 551/565 506/517

Brightnessa _15.8 _36.4 _92.8

pHsensitivity(pKa)b _<4.5 _5.5 _5.7

GenBankaccessionno. AY678264.1 AB370332.1 KC295282.1

Reference Shaneretal.(2004) Sakaue-Sawanoetal.(2008) Shaneretal.(2013)

a_Brightness:_Extinction_Coefﬁcient_(mM−1_cm−1₎_×_Fluorescence_Quantum_Yield.

b _pH_at_which_{ﬂuorescence}_intensity_is_50%_of_its_maximum_value.

2.5. Fluorescenceandconfocalmicroscopy

Direct observation of mycoplasma colonies was conducted after 3days incubation on PPLO agar using an AxioVert A1 invertedmicroscopeequippedwithCY3andEGFPfilterunits(Carl Zeiss, Germany) for observation of red and green fluorescence respectively.Theexposureconditionsweresetsothat no auto-fluorescencewasobservedonwild-typecultures.Theseconditions wereidenticalforallimagestakenusingCY3filters(0.3msat12dB) andEGFPfilters(0.1msat6dB)respectively.Imagesweretaken witha 10Xobjectiveusinga QImagingCDD digitalcamera and analysedusingtheArchimedsoftware(MicrovisionInstruments, France).

Forconfocalmicroscopy analysis,threedayoldmycoplasma brothcultureswerepelleted,re-suspended in4% paraformalde-hyde(PFA,Sigma-Aldrich,France)inPBSandﬁxedduring15min atroomtemperature (RT).Mycoplasmaswerewashed twicein PBS, re-suspended in 20␮L ProLong Diamond Antifade Moun-tant(Fisherscientiﬁc,USA)andtransferredtoastandardslidefor microscopyobservation.

Macrophagestobeanalysedbyconfocalmicroscopywere cul-turedonglasscoverslips.Threewashesin PBSwereperformed betweeneach of the following steps: (i) staining using a 1.5X workingsolutionofCellMaskDeepRedPlasmaMembraneStain (MolecularProbes,LifeTechnologies,USA)during10minat37◦C; (ii)ﬁxingwith4%PFAinPBSfor20minatRT;(iii)nucleistaining with2␮g/mLHoechst33258pentahydratedyesolution(Molecular Probes,LifeTechnologies,USA)for15minatRT.Finally,coverslips weremountedwithProLongDiamondAntifadeMountant.

Observationswere carriedout ona ZEISS LSM 700confocal laserscanningmicroscope.Allimageswereacquiredwitha Plan-Apochromat63Xoilobjectiveataresolutionof620×620pixels formycoplasmaobservationand512×512pixelsformacrophage imagingandwithapinholeapertureof1airyunit.mNeonGreen wasexcitedwitha488nmlaserandthevariablesecondarydichroic (VSD)beamsplitterwassetsoastorecoveralltheﬂuorescence emittedupto530nm.Twootherdistincttrackswerecreatedto recovertheemissionsignals of Hoechst(405nmlaser andVSD beamsplittersetat490nmonthephotomultipliertube1,PMT1) andCellMaskDeepRedPlasmaMembraneStain(555nmlaserand VSDbeamsplittersetat630nmonthePMT1).Allimageswere cap-turedusingthesameparameterstoallowcomparisonsandwere analysedusingtheZENsoftware(CarlZeiss,Germany).

2.6. Flowcytometry

Mycoplasmabrothculturesattheendoftheexponentialphase ofgrowthwerewashedtwiceandcarefullyre-suspendedinPBS beforeflowcytometryanalysis.Dataacquisitionandanalysiswere performedonaFACSCantoIIflowcytometer(BDBiosciences,USA), withanexcitation wavelength of 488nm and detectionwith a 530/30-nmband-passfilter.Aminimumof 10,000mycoplasma cellswereanalysedandtheirglobalmedianfluorescence

intensi-ties(MFI)werecalculatedforcomparison.TheFACSDiva6software (BDBiosciences,USA)wasusedfordataanalysis.

Infectedandnoninfectedmacrophageswerealsoanalysedby flowcytometry.Cellviabilitywasmeasuredusingpropidiumiodide (PI).CellswerewashedtwiceinPBSbeforere-suspensioninPIat 2␮g/mLandthepercentageofdeadcellsmarkedwithPIwas mea-suredbyflowcytometry.SincemycoplasmasarepermeabletoPI, thismethodalsoalloweddiscriminatingintracellular(PI-negative) andextracellular(PI-stained)mycoplasmas.Dataacquisitionand analysiswereperformedasdescribedaboveforgreenfluorescence, anda585/42-nmband-passfilterwasusedforredfluorescence.A minimumof20,000cellswithinagatedregionexcludingcelldebris wereanalysed.

3. Results

3.1. Constructionofﬂuorescenceexpressiontoolsforwhole mycoplasmacelllabelling

Newgenetictoolsforconstitutive,highlevelexpressionof fluo-rescentproteinsinmycoplasmasweredevelopedassummarisedin Fig.1.Themini-transposonpMT85/2res(Fig.1C)wasusedforgene delivery.pMT85/2resisagenetictoolforstable,random mutagen-esisthatallowstheeliminationoftheantibioticresistancemarker fromselectedmycoplasmamutantsbytreatmentwitharesolvase enzyme(Janisetal.,2008).Thisvectorwasmodifiedtoallowthe expressionofclonedgenesinmycoplasmasunderthecontrolofthe strongpromoteroftheelongationfactorTugeneofMmm(tufA). The rho-independenttranscription terminator of thefibril pro-teingeneofthephytopathogenicmycoplasmaS.citri(fib)wasalso includedtoavoidundesirableread-throughofstopcodons(Duret etal.,2005).TheexpressionunitwasinsertedintheSpeIsiteof pMT85/2res,whichallowsitspreservationintheeventofresolvase treatment.TheresultingplasmidwasnamedpMT/exp(Fig.1B).

The red fluorescentproteins mCherry and mKO2 and green fluorescent protein mNeonGreen were selected for expression inmycoplasmas owingtotheirinteresting properties(Table2). Theirgenesequenceswereoptimisedtakingintoconsiderationthe codonusagebiasinmycoplasmas(Suppl.Table1).Alloptimised sequenceshadaGCcontentexceeding29%.Syntheticconstructs correspondingtomycoplasma-optimisedgenesequencesofeach ofthethreefluorescentproteinswereclonedinpMT/exp, result-inginplasmidspMT/mCherry,pMT/mKO2,andpMT/mNeonGreen (Fig.1A).

3.2. Expressionofbrightredandgreenﬂuorescentproteinsin twodistantmycoplasmaspecies

The performance of ﬂuorescence expression tools pMT/mCherry,pMT/mKO2,andpMT/mNeonGreeninthedistant species Mmm and M. bovis was assessed in order to demon-stratetheirbroadspectrumforuseinmycoplasmas.Bothstrains were successfully transformed with the three vectors and the

(5)

C

B

A

D

SpeI pMT85/2res 5.2 kpb res res tnp A Kan a R pMT/exp 5.5 kpb pMT/fluo 6.2 kpb pMT/fluo transposition MCS Tfib NcoI SalISacII AflII XbaI SpeI PtufA

fluorescent marker

NcoI AflII

co

lE1

res KanaR _res

Tfib

SpeI

PtufA fluorescent marker

NcoI AflII

colE1

Fig.1.Modiﬁcationofthemini-transposonpMT85/2resforexpressionofﬂuorescentproteinsinmycoplasmas.A:Cloningofmycoplasma-optimisedgenesequences

ofﬂuorescentproteinsmNeonGreen,mCherryandmKO2(ﬂuorescentmarker)inpMT/exp(B),resultinginplasmidspMT/mNeonGreen,pMT/mCherryandpMT/mKO2

respectively(pMT/ﬂuo).B:CloningofacassettecomprisingthetufApromoterofMycoplasmamycoidessubsp.mycoides(PtufA),amultiplecloningsite(MCS),andtheﬁb

transcriptionterminatorofSpiroplasmacitri(Tﬁb)inpMT85/2res(C),resultinginvectorpMT85/exp.C:pMT85/2resplasmidmap,displayingthetransposasegene(tnpA),

thecolE1originofreplication,andtheaacA-aphDgene(KanaR_)._This_selection_marker_is_located_between_the_two_inverted_repeats_(IR,_black_arrowheads)_that_deﬁne_the

extremitiesofthetransposedfragment(D),whereasthetransposaseislocatedoutside,forincreasedmutantstability.Theressequencesarethetargetsofthe␥␦resolvase,

whichinducestheexcisionoftheentiresequencecomprisedbetweenthem.D:SchematicrepresentationofapMT/ﬂuotranspositionevent.ThetwoIR(blackarrowheads)

deﬁnetheextremitiesofthetransposoninsertioninthemycoplasmagenome(discontinuousline).Thetransposedfragmentcontainstheﬂuorescenceexpressionsystem

(PTufA-fluorescentmarker-Tfib),aswellastheKanaR_gene_and_colE1_origin,_flanked_by_the_two_res_sequences.

transformationefﬁcienciesrangedbetween10−5and10−6 trans-formant cfu/total cfu regardless the strain and construct used. Eachtransformationtubeyieldedfromseveralhundredstoover onethousandtransformantclones.

Five toseven days after transformation fluorescent colonies were directly observed on selective plates by fluorescence microscopyforallmycoplasmastrainsandconstructsused. Fur-thermore,all thecoloniescorrespondingto thesameconstruct presentedequivalentlevelsoffluorescence,onlydependingonthe sizeanddensityofthecolonies.Fluorescentcloneswereselected fromthefastestgrowingmutants.However,thesiteofthe transpo-soninsertionshouldalsobeconsideredinordertolimittheeffect ofgene disruptioninsubsequentfunctional studies.Asa mode ofexample,thetransposoninsertionsite ofMmmmutantswas sequencedandthosepresentinginsertionsinintergenicsequences orirrelevantsitessuchaspseudogeneswerechosen(Suppl.Table 2).Sincestrainsmaybeattenuatedthroughconsecutiveinvitro passagesandthismayalsointerferewithfunctionalstudies,the numberofsubcultureswaskepttoaminimum.Thewholeprocess fromwild-typestrainseedingfortransformationuntilselectionof fluorescentclonesrequiredfivepassagesandtwocloning proce-dures.

The brightness of selected M. bovis clones expressing each of the three fluorescent proteins wasassessed by fluorescence microscopy(Fig.2).ExaminationofallsolidculturesusingCY3 fil-tersunderthesameexposureconditionsresultedinobservation ofred fluorescentcolonies onplates correspondingtomCherry andmKO2-expressingmycoplasmas,whereasnofluorescencewas detectedinwild-typeandmNeonGreen-expressingcultures. On theotherhand,identicalexposureconditionsusingEGFPfilters resultedinobservationofbrightgreenfluorescentcoloniesonlyin mNeonGreen-expressingcultures.Theintensityofredflorescence

washigherinmKO2thaninmCherry-expressingmycoplasmas,as expectedaccordingtothereportedbrightness oftherespective fluorescentproteins(Table2).Theintensityofgreenfluorescent colonieswasalsoremarkable,consideringthehigherlevelof auto-fluorescenceevidencedonmycoplasmaculturesdirectlyobserved usingEGFPfilters,whichimposedlowerexposureconditions. Sim-ilarresultswereobtainedbyobservationofthree-day-old plate culturesfromselectedMmmclonesexpressingeachofthethree fluorescentproteins,withapparentlylowerfluorescence intensi-tiesattributedtothesmallsizeofMmmcolonies,comparedtothose ofM.bovis(Suppl.Fig.2).However,nodifferenceincolonysizewas evidencedbetweenanyofthetransformantclonesandthe corre-spondingwild-typestrain.Furthermore,fluorescenceexpression wasconservedaftermorethan10invitropassagesintheabsence ofselectivepressure.

Broth cultures of fluorescent mycoplasma clones were then analysed by confocal microscopy in order to visualize individ-ual mycoplasma cells. An example is presented in Suppl. Fig. 3. Individual mycoplasmas expressing mNeonGreen could also be detected by flow cytometry (Suppl. Fig 4). Flow cytometry was therefore used to quantify the relative green fluorescence intensity of mNeonGreen-expressing and wild-type mycoplas-mas at the end of the exponential phase of growth in order to confirm that the expression system developed here could induce comparable fluorescence levels in M. bovis and Mmm (Table 3). The MFI values registered in M. bovis strains were higherthanthoserecordedinMmm,indicatingthatthegreen flu-orescencebackgroundwashigher inthis species. However,the MFIratiobetweenmNeonGreen-expressingandwild-typeclones was equivalentin both species (corresponding to a 10–17-fold increaseinMFIvalues,asrecordedinthreeindependent experi-ments).

(6)

Fig.2.Micrographsofthree-day-oldMycoplasmaboviscoloniesonsolidmediumobservedunderaﬂuorescencemicroscope.Fromlefttoright:M.boviswild-typeand

mCherry,mKO2andmNeonGreen-expressingstrains.Fromtoptobottom:imagingusingphasecontrast(PC)andepifluorescencewithCY3filters(red)andEGFPfilters

(green).Themycoplasmacoloniespresentatypical“fried-egg”appearance,withadensecentreembeddedintheagarshowingmoreintenseﬂuorescencesignals.(For

interpretationofthereferencestocolourinthisﬁgurelegend,thereaderisreferredtothewebversionofthisarticle.)

Table3

Fold-changeincreaseingreenﬂuorescenceintensityassociatedtomNeonGreen

expressioninthespeciesMycoplasmamycoidessubsp.mycoidesandMycoplasma

bovis.

Experiment M.mycoidessubsp.mycoides M.bovis

WTa _Neona _FCb _WTa _Neona _FCb

1 89 1015 11.4 225 3769 16.8

2 72 1140 15.8 693 7222 10.4

3 65 929 14.3 578 7240 12.5

a_Median_{ﬂuorescence}_intensity_values.

b _Fold-change_increase_in_median_{ﬂuorescence}_intensity_in_the

mNeonGreen-expressingstrain(Neon)comparedtothewild-typestrain(WT).

3.3. Detectionofﬂuorescentmycoplasmasinsidehostcellsby ﬂowcytometryandconfocalmicroscopy

Amacrophageinfectionassaywasusedtoassessifthe bright-ness of fluorescently-labelled mycoplasmas was sufficient to allowtheirdirectdetection insidehostcells.The mNeonGreen-expressing mycoplasma clones were selected for these assays owingtopracticalreasons,sincetheyweresuitedforflow cytom-etryanalysis.

Bovinemacrophageswereinfectedwithspecifically-opsonised M. bovis strains at an MOI of around 200 and the cells were analysedbyflowcytometry30minpost-inoculation(Fig.3).The forwardversusside-scatterplotanalysesallowedthedefinition ofagateexcludingcelldebrisandaggregates.M.bovis-infected macrophages presentedhigher levels of debristhanuninfected cells,suggesting anincrease inmortalitydue toM.bovis infec-tion,independentlyofthecloneused.Themacrophagesinfected withwild-type M. bovis presented a higher green fluorescence backgroundthantheuninfectedcells(representingapproximately a 4-fold increase in MFI). However, the cells infected with mNeonGreen-expressing M.bovis weresignificantly more fluo-rescent than those infected with the wild-type strain (10-fold increasein MFI).Moreover,thesetwo populations wereclearly discriminatedaccordingtogreen fluorescence,since96% ofthe cellsinfectedwiththemNeonGreen-expressingstrain(versus2% of those infected with the wild-type strain) fluoresced above backgroundlevels(Fig.3).Finally,toconfirm thatthedetected fluorescence resulted from intracellular bacteria, and not from

mycoplasmasattachedtothecellsurface, aPIprotectionassay was performed. The percentage of PI-positive cells among the mNeonGreen-positivecellsneverexceeded2%(resultsnotshown). Thisimpliedthatthegreatmajorityofthemycoplasmas,which arepermeabletoPI,wereprotectedfromstaining,indicatingthat theywerelocatedinsidethecells.Thesameexperimentperformed usingMmmwild-typeandgreenﬂuorescentstrainsprovidedvery similarresults(Suppl.Fig.5).

Theintracellularlocationoffluorescentmycoplasmaswas fur-therdemonstratedbyconfocalmicroscopyanalysis.Fig.4shows confocalmicrographsofmacrophagesinfectedwithM.bovis30min post-inoculation.ManygreenfluorescentM.boviswereobserved insidethemacrophagesinfectedwiththemNeonGreen-expressing strain(Fig.4C),bothintightclustersandasindividualparticles dis-persedinthecytoplasm.Thesizeoftheparticleswasestimatedat nearly0.5␮m,whichcorrespondstothesizeofamycoplasmacell (CittiandBlanchard,2013),and mostofthemwereobservedin themiddlefocalplanewiththecellnucleus,demonstratingthat theywerelocatedinsidethemacrophage.Nogreenfluorescence wasobservedeitherinuninfectedmacrophagesorinmacrophages infectedwithwild-typemycoplasmas(Fig.4AandBrespectively) andnomorphologicalorfunctionaldifferencesweredetectedin macrophagesinfectedwiththemNeonGreen-expressingM.bovis comparedtothoseinfectedwiththewild-typestrain(Fig.4Cand Brespectively).Verysimilarresultswereobtainedbyinfectionof macrophages withMmmstrains(Suppl. Fig.6),confirmingthat individualmNeonGreen-expressingmycoplasmascouldbedirectly detectedinsidephagocyticcells.

4. Discussionandconclusions

4.1. Developmentofﬂuorescenceexpressiontoolsforwhole mycoplasmacelllabelling

Theaim ofthis work wastodevelop universalfluorescence expressiontoolsforwholemycoplasmacelllabellingtomonitor thedynamicsofmycoplasma infections. Thefirstobjective was thereforetoselecttheappropriatevector,expressionsystemand fluorescentproteinsallowingtheproductionofstableand innocu-ousbrightfluorescenttagstolabelmycoplasmacellsfromdifferent species.

(7)

Fig.3.Flowcytometryanalysisofbovinemacrophagesinfectedwithspeciﬁcally-opsonisedMycoplasmabovisstrainsatamultiplicityofinfectionofaround200during

30min.A:uninfectedcells;B:cellsinfectedwithwild-typeM.bovis;C:cellsinfectedwithmNeonGreen-expressingM.bovis.Forwardversussidescatter(FSCandSSC,

respectively)contourplotsshowthegatedregion(P1).Histogramplotsdisplaythemedianﬂuorescenceintensity(MFI)andthepercentageofpositivecells(P2interval)

foreachcondition.P2wassetbasedonmacrophagesinfectedwithwild-typeM.bovistoexcludebackgroundduetoautoﬂuorescence.Arepresentativeresultfromtwo

experimentsisshown.

Themini-transposonpMT85/2res,agenetictoolforstable, ran-dommutagenesisallowingtheproductionofunmarkedmutations (Janisetal.,2008;ZimmermanandHerrmann,2005),wasselected as vector for gene delivery in mycoplasmas. pMT85/2res was derivedfromthetransposonTn4001ofS.aureus,whichpresents abroadhostspectrumandhasbeensuccessfullyusedina vari-etyofmycoplasmaspecies(HalbedelandStülke,2007).Thisvector provedtobeverypracticalandconvenient.TransformationofMmm andM.bovisstrainswithpMT85/2res-basedconstructsallowedthe directisolationoftransformantcoloniesonagar,limitingthe num-berofinvitropassagesrequiredforcloneselection.Clonescould thenbestablypropagatedinnon-selectivemedium,whichmaybe criticalforcertainfunctionalstudies.Althoughtheuseofa trans-posondeliveryvectorratherthanareplicatingplasmidforgene deliveryimpliedtheintegrationinthemycoplasmachromosome andthesubsequentriskoffunctionalalterations,the transforma-tionefficiencywassufficienttoallowtheselectionoflargenumbers ofmutantsforwhichtheinsertionsitecouldeasilybecharacterised. Furthermore,transposonmutagenesismayactuallybeextremely useful for the construction of mutant libraries of fluorescently labelledmycoplasmas.

pMT85/2res wasmodiﬁed to expressheterologous genes in mycoplasmas, resulting in the expression vector pMT/exp. The strongpromoteroftheelongationfactorTugeneofMmm(tufA) wasusedtodrivetheexpressionofclonedgenesinthissystem. Theexpressionunitsofconstitutively,highlyexpressedgenesare thecandidatesofchoicetomaximisegeneexpressionandthetuf promoterhasalreadybeenusedtodrivetheexpressionof heterol-ogousgenesinseveralmycoplasmaspecies(Renaudinetal.,2015; Tulumetal.,2014;ZimmermanandHerrmann,2005),including M.bovis(Sharmaetal.,2014)andM.mycoides(Karasetal.,2014). However,inthis studythetufApromoterofMmmwas success-fullyusedforgeneexpressionintheheterologousspeciesM.bovis. TheequivalentexpressionofhighlevelsofﬂuorescenceinMmm

andM.bovis,twomycoplasmaspeciesbelongingtodistant phy-logenetic groups(Sirand-Pugnetet al.,2007), indicatedthat the expressionsystemmaybebroadlyappliedinmycoplasmas.This wasnotwithoutprecedent,sincethepromoterofthespiralingene, themostabundantproteinofS.citri,provedtobeefﬁcientin driv-inggeneexpressioninavarietyofmycoplasmaspecies(Renaudin etal.,2014).

Theefﬁciencyoftheexpressionsystemindistantmycoplasma species, added to the broad host range of Tn4001 transposi-tion makes pMT/exp a versatile tool for gene expression in mycoplasmas.OnelimitationtotheuseofTn4001-derivative mini-transposonsisthattheaacA-aphDgeneisnotfunctionalincertain mycoplasma species includingM. penetrans and M. fermentans, which are naturally resistant to kanamycin,and M.genitalium, where theexpression of theaacA-aphD gene resultsin growth impairment.AlsoinM.arthritidisandM.pulmonisTn4001 inser-tionwasnotpossible,presumablyduetofunctionalfailureofthe aacA-aphDgene(Renaudinetal.,2014).However,thislimitation mayeasilybesolvedbyusingeitherthechloramphenicol acetyl-transferasegene(cat)orthetetracyclineresistancemarker(tetM), which havebeenshown tofunction inthesespecies(Renaudin etal.,2014),fortheselectionoftransposonmutants.

pMT/expwasusedtoexpressfluorescentproteinsinMmmand M.bovis.Sincetheexpressionoffluorescentmarkersin mycoplas-masremainsrareandhasmetwithdifficulties(Duretetal.,2003), severalunrelatedproteinswereusedtoincreasethechancesof success.ThegreenfluorescentproteinmNeonGreen(Shaneretal., 2013)andredfluorescentproteinsmCherry(Shaneretal.,2004) andmKO2(Sakaue-Sawanoetal.,2008)wereselectedowingto theirinterestingproperties(Table2).Theyareallmonomeric pro-teins,whicharegenerallynontoxic(Shaneretal.,2005),andare characterisedbyhighintrinsicbrightness,particularlymKO2and mNeonGreen.Ontheotherhand,mCherryisparticularlyresistant

(8)

Fig.4.ConfocalmicrographsofbovinemacrophagesinfectedwithMycoplasma

bovisasinFig.3.A:uninfectedmacrophages;B:macrophagesinfectedwith

wild-typeM.bovis;C:macrophagesinfectedwithmNeonGreen-expressingM.bovis.

Theypresenttheoverlayofthreeﬁeldsshowinginredthemacrophage’s

mem-braneandcytosol,inbluethecellnucleusandmycoplasmaDNA,andingreen

themNeonGreen-expressingM.bovis.Thewhitearrowheadsindicateﬂuorescently

labelledmycoplasmacells.(Forinterpretationofthereferencestocolourinthis

ﬁgurelegend,thereaderisreferredtothewebversionofthisarticle.)

tolowpH,whichmaybeusefulforobservationofmycoplasmas insideacidiccompartmentssuchaslysosomes.

Allthree fluorescentproteinsweresuccessfullyexpressedin bothspecies,tosuchextentthatbrightfluorescentcoloniescould bedirectly observed on conventional culture medium by fluo-rescencemicroscopy.Furthermore,nosignsoftoxicity couldbe attributedtoanyofthem.Nogrowthimpairmentorother phe-notypic changes were observed onfluorescent mutant strains, indicatingthattheywereneitheraffectedbythetransposon inser-tion,norbytheproductionofhighlevelsoffluorescentproteins, andthestrainsremainedfluorescentaftermorethan10invitro

passagesintheabsenceofselectivepressure.Thiswasan impor-tantachievement,sinceattemptstoexpressfluorescentproteins in mycoplasmas were not alwayssuccessful. Expression of the greenfluorescentprotein(GFP)infusionwithspiralin,themajor surfacelipoproteinofS.citri,resultedinmorphologicchangesin cellsandinreducedcolonysize(Duretetal.,2003).Moreover,the mutantswereunstableandresultedintheproductionofsubclones thatdidnotexpressGFP,leadingtheauthorstohypothesisethat thefluorescentproteinwastoxicforthisspecies.Thereareonly afewstudiesreportingtheexpressionofGFPderivatives,which wereusedasfusiontagstolocalisemycoplasmacomponentsinM. pneumoniae(Balishetal.,2003;Kenrietal.,2004)andM.mobile (Tulumetal.,2014).Alsoin M.pneumoniae,themonomericred fluorescentprotein(mRFP1)wasusedasfluorescentreporter sys-tem(ZimmermanandHerrmann,2005).However,untilnow,there havebeennoreportsoffluorescenceexpressiontoolsforwhole mycoplasmacelllabellingallowingthedirectobservationof flu-orescentcolonies on agarand providing sufficient fluorescence levelsforuseinhost-mycoplasma interactionstudies.Sincethe strategyusedhereforfluorescentproteinexpressionwasvery sim-ilartothatpreviouslyusedforGFPderivativesandmRFP1(withthe exceptionthatthemRFP1genesequencewasnotcodon-optimised) thepresent successmustbemainlyattributed tothechoiceof fluorescentproteins.

4.2. Useofmycoplasmaﬂuorescenceexpressionsystemsfor host-pathogeninteractionstudies

Fluorescence expressionsystems forwholemycoplasma cell labelling were developed in this study as tools for monitoring the dynamics of mycoplasma infections. Although pathogenic mycoplasmascanbeindirectlydetectedusingfluorescentantibody labelling,traditionalimmunofluorescencedetectiontechniquesare time-consuming,tedious and expensive. Alternatively,theycan bedirectlycoatedwithfluorochromes,butthefluorescencesignal mayfadeovertimeandisdilutedduringreplication.Consequently, theseapproachesareoflimitedvaluefortracinginvasionpathways inlivingcells(Lingetal.,2000).Undeniably,theuseof chromo-somallytaggedmycoplasmaswouldfacilitatetheanalysisofthe mechanismsinvolvedincolonisationandpersistencebyallowing thedirectmonitoringofmycoplasmaadhesion,invasion,survival andmultiplicationinsidehostcells.Therefore,thenextobjectiveof thisstudywastodemonstratethatthebrightnessof fluorescently-labelledmycoplasmaswassufficienttoallowtheirdirectdetection insidehostcells.ThemNeonGreen-expressingmycoplasmaswere selectedforthis“proofofconcept”,sincetheycouldbeanalysed byflowcytometry,andamacrophageinfectionassaywasusedin ordertowarranttheintracellularlocationofmycoplasmas.The presenceoffluorescentmycoplasmasinsidelivephagocyticcells wasdetectedandquantifiedbyflowcytometryandcorroborated byconfocalmicroscopyimaging,whichallowedthedirect observa-tionofindividualmycoplasmasinthecytoplasmofinfectedcells. Furthermore,nosignsofcytotoxicitycouldbeidentifiedandtheuse offluorescentlylabelledmycoplasmasdidnotinterferewiththe cellinfectionassay.Theseresultsindicatedthatthefluorescence expressionsystemsforwholemycoplasmacelllabellingwerewell suitedfortheanalysisofhost-pathogeninteractions.Thesenew toolsofferveryinterestingperspectivesformycoplasmaresearch byallowingthedirectmonitoringofmycoplasmainfectionsboth invitroandinvivo.

Cell invasion can be a major contributor to the establish-mentofpersistentinfectionsandchronicdiseasebymycoplasmas (Browningetal.,2014),butthisphenomenonhasonlybeenstudied inafewmycoplasmaspeciesandthemechanismsinvolvedremain totallyunknown.Sofar,evidenceoftheintracellularlocalisationof mycoplasmashasrequiredacombinationofmicroscopic

(9)

observa-tionswithimmunochemistry,differentialstainingorfluorescence labellingtechniques,oftensubstantiatedbyinvitroinvasionassays suchasthe“gentamicinsurvivalassay”(Cittietal.,2005).Theuseof chromosomallytaggedmycoplasmasoffersinterestingadvantages, sincetheycanbedirectlyvisualizedwithnodilutionofthe fluores-cencesignalbyconfocalmicroscopy,allowingtheaccurateanalysis ofmycoplasmainvasionpathways.Furthermore,time-lapse imag-ingmaybeperformedformonitoringcellinfectionsinreal-time (O’Neilletal.,2016).Flowcytometryhasalsobeenveryusefulfor evaluatingmycoplasmaattachmenttoeukaryoticcells, whichis consideredapre-requisiteforcolonisationandinvasion.Invitro adherenceassaysbasedonflowcytometryhavebeenusedina vari-etyofmycoplasmamodelsofferingdiverseapplications(asamode ofexample,Ayeetal.,2015;García-Moralesetal.,2014;Leighand Wise,2002;Schurwanzetal.,2009).However,theseassaysused either fluorescent antibodiesor fluorochromes for mycoplasma labelling.Theuseofchromosomallytaggedmycoplasma strains canprovidea quick,quantitativeand objectivemethodto com-pareadherencetoandinvasionofdifferenthostcells,andoffers the possibility of evaluating intracellular survival and replica-tion.Furthermore,thismethodmaybeextremelyusefulforhigh throughputscreeningofcelladherenceandinvasionamong dif-ferentmycoplasmastrainsormutants.Indeed,fluorescentmutant libraries,whichcanbeeasilygeneratedbyrandommutagenesis usingthefluorescenceexpressiontoolspresented here,maybe directlyscreenedbyflowcytometrytoidentifydeficientmutants. Althoughinvitrocellinfectionmodelscanprovideinvaluable informationonthemechanismsofmycoplasmapathogenicity,the studyoftheinvasionpathwaysinthenaturalhostremains essen-tialfor afullunderstandingofthepathogenesis ofmycoplasma infections.Thetoolsdevelopedinthisstudyareperfectlysuited fortheanalysisofinvivoinfections,sincefluorescenceexpression bymycoplasmasisstablymaintainedintheabsenceofselective pressureandtheantibioticmarkercanbeeliminatedbyresolvase treatment.Moreover,chromosomallytaggedmycoplasmasmaybe directlymonitoredinhosttissues,allowinginfectionkinetics stud-iesbothexvivo,byhistologicalexamination,andinvivo,bydirect imagingofsmallanimalhostsormodels,asreportedinother bac-teria(Lingetal.,2000;Zelmeretal.,2012).Finally,theavailability ofmultiplefluorescentmarkersmaybeexploitedtolabel differ-entmycoplasma speciesor strainsinorder toanalysemultiple infections,aspreviouslyreported(Lingetal.,2000).

Inconclusion,newgenetictoolshavebeendevelopedforthe stableintroductionofgreenandredfluorescentproteinsas chro-mosomaltagsinmycoplasmas.Theprocessofproduction,selection andcharacterisationoffluorescentcloneswasstraightforwardand resultedintheunprecedentedobservationofredandgreen fluores-centmycoplasmacoloniesonstandardculturemedium,whereas nosigns of cytotoxicity could beidentified. The green fluores-centmycoplasmascouldbedirectlydetectedinsidehostcellsby flowcytometryandconfocalmicroscopy,thusdemonstratingtheir usefulnessforhost-mycoplasmainteractionstudies.Furthermore, thesenewtoolsforstable,wholemycoplasmacelllabellingoffer innumerableperspectivesforthefunctionalanalysisof mycoplas-masbothinvitroandinvivo.

Authors’contributions

LMS, VR and FT conceived and designed the study. TB and LMScarriedouttheexperimentsandtheanalysisanddraftedthe manuscript.MSVcarriedouttheconfocalmicroscopyimagingand participatedindraftingthemanuscript.PTandVRcontributedto theflowcytometryanalyses,CPparticipatedincellculture experi-mentsandCRcontributedtocellfixationandmountingprotocols. Alltheauthorsread,correctedandapprovedthefinalmanuscript.

Acknowledgments

WearegratefultoMakotoMiyataandIsilTulum(OsakaCity University,Japan) fortheiradviceregardingtheconstructionof ﬂuorescenceexpressiontoolsandtoJulienCau(IGH-CNRS, Mont-pellier,France) for hisexcellent adviceregarding the choiceof ﬂuorescentproteinmarkers,forpreliminaryanalysisbyconfocal microscopy andfor constructivediscussions. Wewish tothank Axel Verdierforthedevelopmentof theautomaticcodon opti-mizer,FlorenceTardy(Anses,Lyon,France)forsupplyofM.bovis convalescentsera,PascalSirand-Pugnet(INRA,Bordeaux,France) forplasmidspMT85/2resandpPS3.1,andVictoriaChalker(Public HealthEngland,London,UK)forrevisionoftheEnglishlanguage.

AppendixA. Supplementarydata

Supplementarydataassociatedwiththisarticlecanbefound,in theonlineversion,athttp://dx.doi.org/10.1016/j.jbiotec.2016.08. 006.

References

Aye,R.,Mwirigi,M.K.,Frey,J.,Pilo,P.,Jores,J.,Naessens,J.,2015.Cyto-adherenceof

Mycoplasmamycoidessubsp.mycoidestobovinelungepithelialcells.BMCVet.

Res.11,27,http://dx.doi.org/10.1186/s12917-015-0347-3.

Balish,M.F.,Santurri,R.T.,Ricci,A.M.,Lee,K.K.,Krause,D.C.,2003.Localizationof

Mycoplasmapneumoniaecytadherence-associatedproteinHMW2byfusion withgreenﬂuorescentprotein:implicationsforattachmentorganelle structure.Mol.Microbiol.47,49–60.

Browning,G.F.,Noormohammadi,A.H.,Markham,P.F.,2014.Identiﬁcationand

characterizationofvirulencegenesinmycoplasmas.In:Browning,G.F.,Citti,C. (Eds.),Mollicutes;MolecularBiologyandPathogenesis.CaisterAcademic Press,Norfolk,UK,pp.77–90.

Citti,C.,Blanchard,A.,2013.Mycoplasmasandtheirhost:emergingand

re-emergingminimalpathogens.TrendsMicrobiol.21,196–203,http://dx.doi.

org/10.1016/j.tim.2013.01.003,S0966-842X(13)00014-0.

Citti,C.,Browning,G.F.,Rosengarten,R.,2005.Phenotypicdiversityandcell

invasioninhostsubversionbypathogenicmycoplasmas.In:Blanchard,A., Browning,G.F.(Eds.),Mycoplasmas;MolecularPathogenicityBiology StrategiesforControl.Horizonbioscience,Norfolk,UK,pp.439–483.

Duret,S.,Berho,N.,Danet,J.-L.,Garnier,M.,Renaudin,J.,2003.Spiralinisnot

essentialforhelicity,motility,orpathogenicitybutisrequiredforefﬁcient

transmissionofSpiroplasmacitribyitsleafhoppervectorCirculifer

haematoceps.Appl.Environ.Microbiol.69,6225–6234,http://dx.doi.org/10.

1128/AEM.69.10.6225-6234.2003.

Duret,S.,André,A.,Renaudin,J.,2005.SpeciﬁcgenetargetinginSpiroplasmacitri:

improvedvectorsandproductionofunmarkedmutationsusingsite-speciﬁc

recombination.Microbiol.Read.Engl.151,2793–2803,http://dx.doi.org/10.

1099/mic.0.28123-0.

García-Morales,L.,González-González,L.,Costa,M.,Querol,E.,Pi ˜nol,J.,2014.

QuantitativeassessmentofMycoplasmahemadsorptionactivitybyﬂow

cytometry.PLoSOne9,e87500,http://dx.doi.org/10.1371/journal.pone.

0087500.

Halbedel,S.,Stülke,J.,2007.ToolsforthegeneticanalysisofMycoplasma.Int.J.

Med.Microbiol.297,37–44,http://dx.doi.org/10.1016/j.ijmm.2006.11.001.

Hope,J.C.,Whelan,A.O.,Hewinson,R.G.,Vordermeier,M.,Howard,C.J.,2003.

Maturationofbovinedendriticcellsbylipopeptides.Vet.Immunol. Immunopathol.95,21–31(S0165242703001041).

Hutchison,C.A.,Chuang,R.-Y.,Noskov,V.N.,Assad-Garcia,N.,Deerinck,T.J.,

Ellisman,M.H.,Gill,J.,Kannan,K.,Karas,B.J.,Ma,L.,Pelletier,J.F.,Qi,Z.-Q.,

Richter,R.A.,Strychalski,E.A.,Sun,L.,Suzuki,Y.,Tsvetanova,B.,Wise,K.S.,

Smith,H.O.,Glass,J.I.,Merryman,C.,Gibson,D.G.,Venter,J.C.,2016.Designand

synthesisofaminimalbacterialgenome.Science351,http://dx.doi.org/10.

1126/science.aad6253,aad6253.

Janis,C.,Bischof,D.,Gourgues,G.,Frey,J.,Blanchard,A.,Sirand-Pugnet,P.,2008.

UnmarkedinsertionalmutagenesisinthebovinepathogenMycoplasma

mycoidessubsp.mycoidesSC:characterizationofalppQmutant.Microbiol.

Read.Engl.154,2427–2436,http://dx.doi.org/10.1099/mic.0.2008/017640-0.

Karas,B.J.,Wise,K.S.,Sun,L.,Venter,J.C.,Glass,J.I.,Hutchison,C.A.,Smith,H.O.,

Suzuki,Y.,2014.Rescueofmutantﬁtnessdefectsusinginvitroreconstituted

designertransposonsinMycoplasmamycoides.Front.Microbiol.5,369,http://

dx.doi.org/10.3389/fmicb.2014.00369.

Kenri,T.,Seto,S.,Horino,A.,Sasaki,Y.,Sasaki,T.,Miyata,M.,2004.Useof

ﬂuorescent-proteintaggingtodeterminethesubcellularlocalizationof

Mycoplasmapneumoniaeproteinsencodedbythecytadherenceregulatory

locus.J.Bacteriol.186,6944–6955,

http://dx.doi.org/10.1128/JB.186.20.6944-6955.2004.

Leigh,S.A.,Wise,K.S.,2002.Identiﬁcationandfunctionalmappingofthe

(10)

Ling,S.H.,Wang,X.H.,Xie,L.,Lim,T.M.,Leung,K.Y.,2000.Useofgreenﬂuorescent protein(GFP)tostudytheinvasionpathwaysofEdwardsiellatardaininvivo andinvitroﬁshmodels.Microbiology146(Pt1),7–19.

Nakamura,Y.,Gojobori,T.,Ikemura,T.,2000.Codonusagetabulatedfrom

internationalDNAsequencedatabases:statusfortheyear2000.NucleicAcids Res.28,292.

O’Neill,A.M.,Thurston,T.L.,Holden,D.W.,2016.Cytosolicreplicationofgroupa

Streptococcusinhumanmacrophages.MBio7.

Razin,S.,Yogev,D.,Naot,Y.,1998.Molecularbiologyandpathogenicityof

mycoplasmas.Microbiol.Mol.Biol.Rev.62,1094–1156.

Renaudin,J.,Breton,M.,Citti,C.,2014.MoleculargenetictoolsforMollicutes.In:

Browning,G.F.,Citti,C.(Eds.),Mollicutes;MolecularBiologyandPathogenesis. CaisterAcademicPress,Norfolk,UK,pp.55–76.

Renaudin,J.,Béven,L.,Batailler,B.,Duret,S.,Desqué,D.,Arricau-Bouvery,N.,

Malembic-Maher,S.,Foissac,X.,2015.Heterologousexpressionandprocessing

oftheﬂavescencedoréephytoplasmavariablemembraneproteinVmpAin Spiroplasmacitri.BMCMicrobiol.15,82,10.1186/s12866-015-0417-5.

Sakaue-Sawano,A.,Kurokawa,H.,Morimura,T.,Hanyu,A.,Hama,H.,Osawa,H.,

Kashiwagi,S.,Fukami,K.,Miyata,T.,Miyoshi,H.,Imamura,T.,Ogawa,M.,

Masai,H.,Miyawaki,A.,2008.Visualizingspatiotemporaldynamicsof

multicellularcell-cycleprogression.Cell132,487–498.

Schurwanz,N.,Jacobs,E.,Dumke,R.,2009.Strategytocreatechimericproteins

derivedfromfunctionaladhesinregionsofMycoplasmapneumoniaefor

vaccinedevelopment.Infect.Immun.77,5007–5015,http://dx.doi.org/10.

1128/IAI.00268-09.

Shaner,N.C.,Campbell,R.E.,Steinbach,P.A.,Giepmans,B.N.G.,Palmer,A.E.,Tsien,

R.Y.,2004.Improvedmonomericred,orangeandyellowﬂuorescentproteins

derivedfromDiscosomasp.redﬂuorescentprotein.Nat.Biotechnol.22,

1567–1572,http://dx.doi.org/10.1038/nbt1037.

Shaner,N.C.,Steinbach,P.A.,Tsien,R.Y.,2005.Aguidetochoosingﬂuorescent

proteins.Nat.Methods2,905–909.

Shaner,N.C.,Lambert,G.G.,Chammas,A.,Ni,Y.,Cranﬁll,P.J.,Baird,M.A.,Sell,B.R.,

Allen,J.R.,Day,R.N.,Israelsson,M.,Davidson,M.W.,Wang,J.,2013.Abright

monomericgreenﬂuorescentproteinderivedfromBranchiostomalanceolatum.

Nat.Methods10,407–409,http://dx.doi.org/10.1038/nmeth.2413.

Sharma,S.,Markham,P.F.,Browning,G.F.,2014.Genesfoundessentialinother

mycoplasmasaredispensableinMycoplasmabovis.PLoSOne9,e97100,http://

dx.doi.org/10.1371/journal.pone.0097100.

Sirand-Pugnet,P.,Citti,C.,Barré,A.,Blanchard,A.,2007.Evolutionofmollicutes:

downabumpyroadwithtwistsandturns.Res.Microbiol.158,754–766,

http://dx.doi.org/10.1016/j.resmic.2007.09.007.

Tulum,I.,Yabe,M.,Uenoyama,A.,Miyata,M.,2014.LocalizationofP42and

F(1)-ATPase␣-subunithomologoftheglidingmachineryinMycoplasmamobile

revealedbynewlydevelopedgenemanipulationandﬂuorescentprotein

tagging.J.Bacteriol.196,1815–1824,http://dx.doi.org/10.1128/JB.01418-13.

Zelmer,A.,Carroll,P.,Andreu,N.,Hagens,K.,Mahlo,J.,Redinger,N.,Robertson,B.D.,

Wiles,S.,Ward,T.H.,Parish,T.,Ripoll,J.,Bancroft,G.J.,Schaible,U.E.,2012.A

newinvivomodeltotestanti-tuberculosisdrugsusingﬂuorescenceimaging.J.

Antimicrob.Chemother.67,1948–1960,http://dx.doi.org/10.1093/jac/dks161.

Zimmerman,C.-U.,Herrmann,R.,2005.Synthesisofasmall,cysteine-rich,29

aminoacidslongpeptideinMycoplasmapneumoniae.FEMSMicrobiol.Lett.