Frequency
of
Mycobacterium
tuberculosis-specific
CD8+
T-cells
in
the
course
of
anti-tuberculosis
treatment
Rebecca
Axelsson-Robertson
a,
Martin
Rao
b,
Andre
G.
Loxton
c,
Gerhard
Walzl
c,
Matthew
Bates
d,e,
Alimuddin
Zumla
d,e,f,
Markus
Maeurer
a,b,*
a
CentreforAllogeneicStemCellTransplantation(CAST),KarolinskaUniversityHospital,Stockholm,Sweden
b
DivisionofTherapeuticImmunology(TIM),DepartmentofLaboratoryMedicine(LABMED),KarolinskaInstitutet,Ha¨lsova¨genF79,KarolinskaUniversity HospitalHuddingeCampus,SE14186,Stockholm,Sweden
cDST/NRFCentreofExcellenceforBiomedicalTuberculosisResearchandMRCCentreforMolecularandCellularBiology,DivisionofMolecularBiologyand
HumanGenetics,DepartmentofBiomedicalSciences,FacultyofHealthSciences,StellenboschUniversity,SouthAfrica
d
DivisionofInfectionandImmunity,UniversityCollegeLondon,London,UK
e
UNZA-UCLMsResearchandTrainingProject,UniversityTeachingHospital,Lusaka,Zambia
f
NIHRBiomaedicalResearchCentreatUniversityCollegeLondonHospitals,London,UK
1. Introduction
Cell-mediatedimmuneresponsesareessentialforprotection against tuberculosis (TB).1 CD8+ T-cells produce cytotoxic
moleculesaswellas cytokines.CytotoxicT-cellsarecapableof killingcellsinfectedwithMycobacteriumtuberculosis(MTB),2while
aidingtherecruitmentandactivationofotherimmunecelltypes. CD8+T-cellsrecognizeshortantigen-derivedpeptides(epitopes) presented on major histocompatibility complex (MHC) class I molecules on the surface of antigen-presenting cells (APCs).3
ClinicalandpreclinicalassessmentofCD8+T-cellsinTBshowthat
CD8+T-cellsplayanimportantroleinrecallresponses,aswellasin long-termprotectiontoMTBinfection.2,4
OnlyalimitednumberofT-cellepitopesof9–11aminoacidsin length havebeen identifiedtodate,5 usingtechnologiessuch as
ELISA,6 ELISPOT,7 the chromium release assay,8 the thymidine
incorporation assay,9 intracellular cytokine staining (ICS),10 and solubleMHCclassI/peptidemultimers.11Alargenumberofthese
epitopesoriginatefromonlyafewwell-characterized, immunologi-cally relevant MTB proteins, e.g., early-secreted antigenic target 6kDa(ESAT-6,Rv3875),12culturefiltrateprotein10kDa(CFP10,
Rv3874),7antigen85B(Ag85B,Rv1886c),11TB10.4(Rv0288),13anda
conservedtransmembraneprotein(Rv1733c).14This representsa
rather biased viewof T-cellresponses toonlya handful ofover 4000existingMTBproteinspotentiallyexpressedduringinfection.15
Fromanother60MTBproteins,oneoronlyseveralT-cellepitopes
ARTICLE INFO Articlehistory:
Received1December2014
Receivedinrevisedform16January2015 Accepted16January2015
CorrespondingEditor:EskildPetersen, Aarhus,Denmark Keywords: CD8+T-cells Mycobacteriumtuberculosis Tetramers Multimers MHC SUMMARY
Anti-tuberculosisdrugtreatmentisknowntoaffectthenumber,phenotype,andeffectorfunctionalityof antigen-specificT-cells.InordertoobjectivelygaugeMycobacteriumtuberculosis(MTB)-specificCD8+ T-cellsatthesingle-celllevel,we developedsolublemajorhistocompatibilitycomplex(MHC)classI multimers/peptide multimers, which allow analysis of antigen-specific T-cells without ex vivo manipulationorfunctionaltests.Weconstructed38MHCclassImultimerscoveringsomeofthemost frequentMHCclassIalleles(HLA-A*02:01,A*24:02,A*30:01,A*30:02,A*68:01,B*58:01,andC*07:01) pertinenttoaSouthAfricanorZambianpopulation,andpresentingthefollowingMTB-derivedpeptides: theearlyexpressedsecretedantigensTB10.4(Rv0288),Ag85B(Rv1886c),andESAT-6(Rv3875),aswell asintracellular enzymes, i.e., glycosyltransferase1(Rv2957), glycosyltransferase 2(Rv2958c), and cyclopropanefattyacidsynthase(Rv0447c).Anti-TBtreatmentappearedtoimpactonthefrequencyof multimer-positiveCD8+T-cells,withageneraldecreaseafter6monthsoftherapy.Also,areductionin thetotalcentralmemoryCD8+ T-cellfrequencies, aswellastheantigen-specific compartmentin CD45RA CCR7+T-cellswasobserved.Wediscussourfindingsonthebasisofdifferentialdynamicsof MTB-specificT-cellfrequencies,impactofMTBantigenloadonT-cellphenotype,andantigen-specific T-cellresponsesintuberculosis.
ß2015TheAuthors.PublishedbyElsevierLtdonbehalfofInternationalSocietyforInfectiousDiseases. ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(
http://creativecommons.org/licenses/by-nc-nd/4.0/).
* Correspondingauthor.Tel.:+46708627566. E-mailaddress:markus.maeurer@ki.se(M.Maeurer).
ContentslistsavailableatScienceDirect
International
Journal
of
Infectious
Diseases
j o urn a l hom e pa ge : ww w. e l s e v i e r. c om/ l o ca t e / i j i d
http://dx.doi.org/10.1016/j.ijid.2015.01.017
1201-9712/ß2015TheAuthors.PublishedbyElsevierLtdonbehalfofInternationalSocietyforInfectiousDiseases.ThisisanopenaccessarticleundertheCCBY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/).
havebeenidentified–leavingthemajorityoftheMTBproteometo beexploredforT-cellresponses.
Since the expression of MTB proteins is associated with different stages of MTB growth as well as TB disease,16 the
identificationofabroaderMTBepitopeprofileisessentialforthe discoveryofmarkersassociatedwithactiveorlatentTBinfection (LTBI).HalfoftheMTBepitopesassociatewithasingleMHCclassI allele (HLA-A*02:01) out of more than 3000 different alleles describedin humans thus far.17Based on this information,it is
clear that more CD8+ T-cell epitopes derived from novel MTB antigensassociatingwithclinicallyrelevantMHCclassIallelesneed tobedescribed,especiallysincetheregionaldistributionofalleles variesbetweencontinentsaswellasdifferentethnicgroups.18
One technique to identify epitope-specific (MTB antigen-specific)T-cellswithoutinvitromanipulationistheuseofMHC classImultimers.Wehavepreviouslydesignedabroadpanelof MHC class I multimers covering seven different MHC alleles presentingepitopesfromsixdifferentMTB-derivedproteins:19the early expressed secreted antigens TB10.4 (Rv0288), Ag85B (Rv1886c),andESAT-6(Rv3875),aswellasintracellularenzymes glycosyltransferase 1 (Rv2957), glycosyltransferase 2 (Rv2958c), and cyclopropane fatty acid synthase (CFA synthase; Rv0447c). Since anti-TB drug treatment is known to affect the number, phenotype, and effector functionality of antigen-specific CD8+ T-cells,20,21 we chose to develop and work with MHC class I
multimersinthisstudyasameanstodecipherthenatureof antigen-specificCD8+T-cellsbeforeandafteranti-TBdrugtherapy. 2. Materialsandmethods
2.1. Patientdata
Seven HIV patients with untreated, active pulmonary TB, diagnosedwithatleasttwopositivesputumsmearsforacid-fast bacilliorpositivesputumcultureforMTB,wereenrolled atthe University of Stellenbosch, South Africa, as well as from the University Hospital Lusaka, Zambia. Five were female and two weremale;theyrangedinagefrom20to55years.Sampleswere takenatthetimeof diagnosis(beforeinitiatingtreatment)and after6monthsoftreatment(endoftherapyfordrug-sensitiveTB). Peripheralbloodmononuclearcells(PBMCs)wereacquiredfrom the patient after receiving their informed consent, as well as endorsementfromtheinstitutionalreview boards(No.N05/11/ 187;HealthResearchandEthicsCommittee,Stellenbosch Univer-sity).FrozenPBMCs wereshipped to Swedenwhere theywere HLA-typedusingsequence-specificprimer(SSP)typingkits(One Lambda Inc., Canoga Park, CA, USA). Ethical consent was also obtainedfromtheethicscommitteeinStockholm(Ref. 2011/863-31/2,includingthesamplesfromZambia).
2.2. Cellularanalysiswithmultimers
Thirty-eight fluorescently labelled multimers (streptavidin– phycoerythrin (PE), streptavidin-allophycocyanin (APC), and fluorescein isothiocyanate (FITC)) covering HLA-A*02:01, A*24:02, A*30:01, A*30:02, A*68:01, B*58:01, and C*07:01 presenting peptides from the MTB-derived proteins Rv0288, Rv1886c, Rv3875, Rv2958c, Rv2957, and Rv0447c, were either constructedin-house(aspreviouslydescribed22)orcommercially
acquired (Beckman Coulter, San Diego, USA, and Immudex, Copenhagen, Denmark). In the CD3+CD8+CD4 compartment, multimer-positiveeventswererecordedusinganti-CD3-PE/Texas red (ECD) (Clone UCHT1; Beckman Coulter), anti-CD4-Pacific orange(CloneS3.5;Invitrogen,Carlsbad,CA,USA),and anti-CD8a-APC/Cy7(CloneSK1;BectonDickinson,FranklinLakes,NJ,USA). Cellsin theCD3+CD8 CD4+ compartment wereexcludedfrom
enumeration of multimer-positive events. All cellular analyses wereperformedusingaFACS GalliosFlow-cytometer(Beckman Coulter).Onlymultimerresponsesthatwereatleastthreetimes higherthanthenegativecontrolandforwhichwecoulddetect morethan50eventswereanalyzedfurther.
2.3. PhenotypicanalysisoftheT-cells
Analysisofthephenotype,degranulationmarker(CD107a),and survival marker (interleukin 7 receptor subunit alpha, IL7R
a
) expressionprofile ofthemultimer-specificcells wasperformed using anti-CD45RA-PerCP/Cy5.5 (Clone HI100; Biolegend, San Diego, CA, USA), anti-CCR7-PE/Cy7 (Clone 3D12; Becton Dick-inson),anti-CD107a-Pacificblue(PB)(CloneH4A3;Biolegend),and anti-CD127-APC/Alexa-700(CloneR34.34;BeckmanCoulter). 2.4. StatisticalanalysisTheStudent’st-testorpairedt-testwasappliedfortheanalysis of statistical significance between different T-cell populations usingGraphPadPrism4.0 software(GraphPadInc.,LaJolla,CA, USA).Ap-valueoflessthan0.05wasconsideredsignificant.
3. Results
3.1. MultimericanalysisofselectedTBepitopes
Wesuccessfullyconstructed38MHCclassIMTBpeptide-loaded multimerscoveringsomeofthemostfrequentlyoccurringMHCclass Ialleles(HLA-A*02:01,A*24:02,A*30:01,A*30:02,A*68:01,B*58:01, andC*07:01)amongSouthAfricanandZambianpopulations.The multimersweretestedonsamplesobtainedfromsevenHLA-typed individualsdiagnosedwithacutepulmonaryTB,priortoandpost anti-TBtreatment.Theantigen-specificanti-TBresponsesbyCD8+ T-cellsweregenerallylowbutdiversebeforeandaftertreatment. Furthermore,thefrequenciesofantigen-specificCD8+T-cellsranged between0and2.2%beforetreatment,whilethefrequencieswere foundtorangebetween0and1.8%postanti-TBtherapy(Table1).
Onaverage,thefrequenciesdecreasedsignificantlyfrom0.25% beforetreatmentto0.20%aftertreatment(p=0.02)(Figure1A), but thetrend varied considerably between individual antigen-specificT-cellpopulations(Figure1B).Anti-TBtreatmentresulted inonlyalimiteddecreaseinfrequenciesofantigen-specificCD8+ T-cellsformostmultimer-specificpopulations.However,forafew multimer-specificpopulations,adramaticdecreaseinfrequencies could be observed, e.g., A2-TB10.4IMYNYPAML (
D
0.77%),A24-ESAT6ELNNALQNL (
D
0.54%), and B58-Ag85BQTYKWETFL (D
0.34%).Despite anobserved decreasein frequencies in themajority of antigen-specificCD8+T-cellpopulations,anincreasein antigen-specificT-cellfrequenciescouldbeobservedposttreatmentwith regard to some populations, e.g., T-cells directed against A2-Ag85BFIYAGSLSA, A24-Ag85BIYAGSLSAL, and A68-TB10.4ANTMAMMAR
(Table 1). The average decrease in frequencies after anti-TB
treatment was independent of allele-restricted and peptide-derivedTBproteins.However,someT-cellpopulationsrestricted bycertainHLAallelesexhibitedthetendencyformoredramatic decreases in T-cell frequencies post anti-TB treatment, e.g., B*58:01(
D
0.34%)andA*24:02(D
0.12%).Adecreaseinfrequency could also be observed with respect to antigen-specific T-cell populationsrecognizingpeptidesderivedfromtheearlyexpressed TBantigens(TB10.4,Ag85B,andESAT-6)(D
0.11%)(Figure1C)as comparedtotheglycosyltransferasesandCFAsynthase(D
0.01%) (Figure 1D). The decrease in frequencies for most multimer-specific populations was also true (Supplementary Material, FigureS1),despitethefactthatsomeindividuals(PAT1andPAT4)remained culture-positive for MTB after 6 months of anti-TB treatment(informationnotshown).
3.2. Phenotypicanalysisofantigen-specificT-cells
PhenotypicanalysesoftotalCD8+T-cellsaswellasthe antigen-specificT-cellcompartmentbasedonCD45RAandCCR7expression were performed to gauge differences before and after anti-TB treatment. In general, themajority of total as well as antigen-specific CD8+ T-cells belonged to the precursor compartment (CD45RA+CCR7+), followed by terminally differentiated (CD45RA+CCR7 ) and effector memory cells (CD45RA CCR7 ). Importantly,thefrequenciesof these populations didnot differ considerably prior to and post anti-TB treatment. The central memoryT-cellcompartment(CD45RA CCR7+)decreased signifi-cantly post anti-TB treatment, both among total CD8+ T-cells (p=0.04) and antigen-specific T-cell populations (p=0.008) (Figure 2). The frequencies of antigen-specific CD8+ T-cells belonging to a certain phenotype (based on CD45RA/CCR7 expression)were independent ofthe presentingHLAallele and theantigen-derivedMTBepitope(datanotshown).
3.3. Analysisofdegranulationandsurvivalmarkers
IL7R
a
is an important marker for T-cell survival, and this receptor can normally be found on most mature T-cells.23 Nodifference couldbe detected interms of IL7R
a
(also known as CD127)expressionamongtotalCD8+T-cellspriorto(80%) and posttreatment(78%).However,asignificantdecreaseinIL7Ra
was detected among antigen-specific T-cells(88% vs. 83%;p=0.04) (Figure3AandB).CD8+T-cellcytotoxicitycanbecorrelatedwiththeexpression ofthedegranulationmarkerCD107a(LAMP-1).24Theexpressionof
CD107a was generally upregulated on antigen-specific T-cells comparedtothetotalCD8+T-cellpopulation.Atrendofdecreased frequenciesofCD107aexpressionpost-treatmentcouldbeseen among total CD8+T-cells, whilethe opposite was truefor the antigen-specificCD8+T-cellpopulations(Figure3CandD).
4. Discussion
Theuseofabroadpanelof38differentmultimerspresenting peptidesfromdifferentMTB-derivedproteinsinthisstudyallowed
Table1
Prevalenceofepitope-specificT-cellsidentifiedbymultimerstaininga
Tetramer/patient 1 2 3 4 5 6 7
Bef Aft Bef Aft Bef Aft Bef Aft Bef Aft Bef Aft Bef Aft
NegPE 0.1 0.0 0.1 0.1 0.0 0.1 0.0 0.0 0.1 0.1 0.1 0.1 0.0 0.1 NegAPC 0.0 0.1 0.0 0.0 0.0 0.0 0.1 0.1 0.0 0.1 0.0 0.1 A2-TB10.4IMYNYPAML 0.8 0.1 1.0 0.2 A2-TB10.4AMLGHAGDM 0.2 0.1 0.1 0.2 A2-TB10.4MLGHAGDMA 0.3 0.0 0.1 0.1 A2-Ag85BYLLDGLRAQ 0.0 0.1 0.3 0.5 A2-Ag85BKLVANNTRL 0.2 0.2 1.4 0.9 A2-Ag85BFIYAGSLSA 0.0 0.4 0.0 0.1 A2-ESAT6AMASTEGNV 0.1 0.1 0.4 0.4 A2-ESAT6LLDEGKQSL 0.0 0.1 0.0 0.1 A2-Rv2958ALADLPVTV 0.0 0.0 0.0 ND A2-Rv2957SIIIPTLNV 0.2 0.1 0.0 0.1 A2-Rv0447VLAGSVDEL 0.0 0.1 0.0 ND A24-TB10.4IMYNYPAML 0.4 0.0 0.3 0.1 A24-Ag85BWYYQSGLSI 0.2 0.2 0.5 0.3 A24-Ag85BFLTSELPQW 0.9 0.0 0.1 0.1 A24-Ag85BIYAGSLSAL 0.0 0.4 0.0 0.0 A24-ESAT6AYQGVQQKW 0.0 0.0 0.1 0.1 A24-ESAT6ELNNALQNL 1.9 1.2 2.2 1.8 A24-Rv2958KYIAADRKI 0.0 0.0 0.0 0.1 A24-Rv2957PYNLRYRVL 0.3 0.1 0.1 0.4 A24-Rv0447KYIFPGGLL 0.0 0.0 0.0 0.1 A3001-TB10.4QIMYNYPAM 0.3 0.2 0.2 0.1 A3001-TB10.4LVRAYHAMS 0.2 0.2 0.5 0.3 A3001-Rv2957IVLVRRWPK 0.1 0.2 0.3 0.3 A3002-TB10.4QIMYNYPAM 0.3 0.2 0.2 0.4 A3002-TB10.4IMYNYPAML 0.7 0.2 0.2 0.2 A3002-TB10.4AMEDLVRAY 0.2 0.2 0.3 0.1 A3002-ESAT6AMASTEGNV 0.1 0.1 A3002-Rv2958SARLAGIPY 0.5 0.2 0.8 0.6 A3002-Rv0447RMWELYLAY 0.1 0.2 0.3 0.2 A68-TB10.4HAMSSTHEA 0.3 0.2 A68-TB10.4ANTMAMMAR 0.1 0.2 A68-Ag85BLPQWLSANR 0.2 0.1 A68-Ag85BWGAQLNAMK 0.2 0.1 A68-Rv2958AAPEPVARR 0.2 0.1 A68-Rv2957LVYGDVIMR 0.0 0.2 A68-Rv0447AASAAIANR 0.0 0.0 B58-Ag85BQTYKWETFL 0.0 0.0 0.0 0.0 1.1 0.1 Cw07-Ag85BANNTRLWVY 0.0 0.0 0.1 0.0
Bef,beforetherapy;Aft,aftertherapy;PE,streptavidin–phycoerythrin;APC,streptavidin–allophycocyanin;ND,notdetermined;PBMC,peripheralbloodmononuclearcells; TB,tuberculosis;MHC,majorhistocompatibilitycomplex.
a PBMCsfromindividualswithTBwereincubatedwithMHC-matchedMHCclassITBmultimersandstainedforT-cellmarkers.Resultsarereportedasthepercentage
multimer-positiveeventsintheCD3+CD8+T-cellpopulation;negativegatingwasperformedtoexcludeCD4+T-cells.Negativemultimerswereincludedtodecipherthe backgroundstaining.ForshadedpointsitwaspossibletoretrievethephenotypeaswellasdataregardingthedegranulationmarkerCCR7andthesurvivalmarkerCD127of theantigen-specificT-cells.
ustocharacterizetheTB-specificCD8+T-cellrepertoirebeforeand after (6 months) anti-TB treatment in patients from countries endemicforTB.25SincetheimportanceofCD8+T-cellshasbeen increasinglyacknowledged,e.g.,incombinationwithanti-tumour necrosisfactoralpha(anti-TNF-
a
)treatment,26itisimportanttofurthercharacterizetheCD8+T-cell-mediatedimmuneresponse inpatientswithactiveTB,andidentifythehithertounknowneffect ofanti-TBtreatmentonantigen-specificT-cellpopulationsthereof.
The effect of anti-TB treatment concerning the frequency, phenotype,andeffectorfunctionsofantigen-specificCD8+T-cells hasbeenstudiedpreviouslyindifferentsettings,involvingadults and children with active TB.2,10,21,27,28 However, therespective
outcomesofthesestudies havebeenconflicting. In thecurrent report, the decreased frequencies of the antigen-specific CD8+ T-cells detected after TB treatment were contradictory to data reportedbyothergroupsevaluatingTB-specificmultimerstaining
Figure1.Frequencypercentagesofmultimer-positiveCD8+T-cells.(A)Averagedetectionofantigen-specificCD8+T-cellsinbloodfrompatientsbeforeanti-TBtreatment andafter6monthsofcompletedtreatment.(B)Thetrendofeachofthe38antigen-specificT-cellpopulationsdetectedbydifferentmultimersbeforeandafteranti-TB treatmentintheCD8+T-cellcompartment.Theaveragepercentageofmultimer-specificCD8+T-cellsbeforeandafteranti-TBtreatmentrecognizingepitopesderivedfrom (C)thepreviouslywell-studiedantigensTB10.4(Rv0288),Ag85B(Rv1886c),andESAT-6(Rv3875),andfrom(D)thenewlydiscoveredTBantigensglycosyltransferase 1(Rv2958c),glycosyltransferase2(Rv2957),andCFAsynthase(Rv0447c).Thetwo-sidedStudent’st-testwasperformedandsignificantvalueswerecalculatedbasedon p-values;*p<0.05.(Bef,beforetherapy;Aft,aftertherapy.)
Figure2.Frequenciesofthe(A)totalCD8+T-cells,and(B)antigen-specificCD8+T-cellsexpressingdifferentphenotypicmarkersbefore(naı¨ve (CD45RA+CCR7+),dotted; centralmemory(CD45RA CCR7+),horizontallines;effectormemory(CD45RA CCR7 ),diagonallines;terminallydifferentiatedcells(CD45RA+CCR7 ),white)andafter anti-TBtreatment(naı¨ve(CD45RA+CCR7+),chequered;centralmemory(CD45RA CCR7+),verticallines;effectormemory(CD45RA CCR7 ),grey;terminallydifferentiated cells(CD45RA+CCR7 ),black).Thetwo-sidedStudent’st-testwasperformedandsignificantvalueswerecalculatedbasedonp-values;*p<0.05.(Bef,beforetherapy;Aft, aftertherapy.)
priortoandpostanti-TBtreatment.Onereasonforthismightbe thelowpatientnumberincludedinourstudy(n=7),butalsoother pointsareworthdiscussing.Onestudydescribedanincreaseinthe frequencyofAg85A-specificCD8+T-cells(0.4%)at4monthspost treatmentin childrenusingMHC classImultimers.20Thesame
groupalso described a similar trend in antigen-specific T-cells (specificforthe16kDaantigen,Rv1490andESAT-6)derivedfrom MTB-infectedadults.However,theresultsarenotconclusivesince the frequencies of antigen-specific CD8+ T-cells recognizing peptides derived from other TB antigens (Hsp65, Ag85B, and Rv1614)showsimilarorreducedfrequenciesofantigen-specific T-cells post anti-TB treatment.10 Also, a decrease in
antigen-specificCD4+T-cellscouldbeseenpostanti-TBtreatmentusing HLA-DR8multimerspresentingESAT-6-derivedpeptides.29Since
ourstudyshowsadecreaseinfrequencyofantigen-specificCD8+ T-cells particularly directed against the highly-expressed early antigensTB10.4,Ag85B,andESAT-6,thedecreasemightbearesult of reduced antigenic load during/after treatment (discussed below). However, this and other studies implementing MTB-specificMHCclassImultimers,suggestthatthereappearsbea broad anti-TB CD8+ T-cell response involving many different epitopesderivedfromdifferentgroupsofTBproteins,albeitata verylowfrequency.10,19
Thephenotype of thetotal CD8+T-celland antigen-specific populationsthereinshowedasimilarpatterninourstudy,witha significantdecrease in the central memory population post TB treatment.Furthermore,thesizeoftheotherphenotypicCD8+ T-cellcompartmentswasnotaffectedbyanti-TBtreatment.Inline withour findings, a reportpublished last year attested to the informativerole of CD8+ T-cellsin sensing bacterialburden in patients who hadundergone standard anti-TBtreatment. Since CD8+ T-cells recognize and respond to intrinsic (intracellular)
antigens,theauthorsclaimedthatadecliningCD8+T-cellresponse maybeindicativeofdiminishingMTBpopulationsduetoanti-TB drugtherapy.21Nonetheless,ourcurrentreportisthefirststudyto shedlightonCD8+T-cellsspecificforseveralTBepitopesbasedon multimerstaining.
Anotherimportantpointiswhethermycobacterialburdenin thepatientaffects themaintenanceofantigen-specificmemory CD8+T-cellnumbers.InanevaluationbyTheronandcolleaguesof thelongitudinalT-cellreactivityofSouthAfricanTBpatientsusing interferon gamma release assays (IGRAs) such as TSPOT.TB, QuantiFERONGoldIn-tube,and PPD-ELISpotin thebackdrop of anti-TBtreatment,mycobacterialloadinsputumwasfoundnotto correlatewiththeamountofinterferongamma(IFN-
g
)secretedby thepatients’PBMCs.30AlthoughIGRAslargelyaccountforMTB-specificCD4+T-cellresponsesintheperiphery,31CD8+T-cellsalso
contributewithIFN-
g
secretioninresponsetostimulationwith epitopesfromthesame antigens.32 PatientswithTB whowereculture-positiveforMTBatthetimeofdiagnosisandconverted after6monthsofanti-TBtreatmentdidnotexhibitdifferencesin IFN-
g
levels. The authors concluded that cellular immune responses,atleastinpart,arenotdirectlyaffectedbyMTBload inpatientsinhigh-burdencountriesforTB.Amorerecentstudyby Rozotetal.evaluatedthecytokineprofileofCD4+andCD8+T-cell response inpatients withactiveTB as wellasthose withLTBI, originatingfromvarioushigh-burdencountriesforTB.Similarto thestudybyTheronandcolleagues, MTBculturepositivitywas shown not toaffect theantigen-specific T-cell response of the patients, while polyfunctionality was postulated as a reliable diagnostic measure of the disease state.27 Nevertheless, TB serodiagnostictests donot accountfor qualitativevariations in TCR(T-cellreceptor)specificitiesarisingfromdifferentclustersof MTBantigensoccurringinthepatientfollowinganti-TBtreatment.Figure3.Frequenciesof(A)totalCD8+T-cells,and(B)antigen-specificCD8+T-cellsexpressingtheCD127(IL7Ra)cell-surfacemarkerbeforeandafteranti-TBtherapy. Frequenciesof(C)totalCD8+T-cells,and(D)antigen-specificCD8+T-cellsexpressingtheCD107a(LAMP-1)degranulationmarkerbeforeandafteranti-TBtherapy.The two-sidedStudent’st-testwasperformedandsignificantvalueswerecalculatedbasedonp-values;*p<0.05.(Bef,beforetherapy;Aft,aftertherapy.)
Several reasonsmayexplainthedifferentresultsconcerning thenumberofantigen-specificT-cellsinthecourseofTBtherapy. The patient’s genetic background, which reflects the predomi-nantlyexpressedHLA allelesin a given population,determines whichMTBepitopestheCD8+T-cellsareabletoencounterand respond to. Other factors are the molecular modifications harbouredbytheMTBstrain(s)infectingtheindividual, bioavail-ability of MTB-derived factors, and previous bacille Calmette– Gue´rin(BCG)vaccination,aswellascontactwithenvironmental mycobacteria that play a fundamental role in shaping theTCR antigen specificity repertoire.33,34 These phenomena impose
continuousexposureoftheinfectedindividualtoantigenicstress, collectively contributing to counter-modulation of T-cell func-tionality–andqualityofresponsivenessofindividualT-cells.High ratesofMTBtransmissioninhigh-burdencountriesforTBsuchas SouthAfricaandZambiaisanotherinevitablefactorthatinfluences thediversityofCD8+T-cell-dependentantigenrecognitionandthe ensuing effector response.25 It is important to appreciate that geographicalvariationanddiseaseendemicityarecritical param-etersthataffectanti-TBCD4+andCD8+T-cellresponsesobserved inthepopulation,aphenomenonthathasbeentermed‘antigenic editingoftheT-cellrepertoire’inthecancerfield.35
Highantigenconcentrationinanindividualpriortoinitiationof therapymayinducethegenerationofpotentiallyhighnumbersof antigen-specificT-cells,giventhatMTBbacillaryloadinapatient withpulmonaryTBcanreach109viablebacteriabasedon
sputum-smear microscopy and culture.36 As mentioned earlier, the
majorityof drug-sensitive MTB bacilliarekilled in theprocess ofanti-TBtreatment,sparingonlythosethatundergometabolic andphysiological adaptation andbecomerefractory to antibio-tics.37 This serves as an indication of antigen turnover in the patient, shaping the repertoire while influencing shifts in bioavailability of MTB epitopes presented in association with theirMHCclassI/IIbackgroundtogetherwithpro-inflammatory cytokines.Furthermore,thereisalackofknowledgeconcerning whetheranti-TBtreatmentdrivessurvivingMTBbacillitoalatent (dormant)stateinhumans,suchasthatobservedintheCornell mouse model of TB.37,38 Thus, the significant reduction in
mycobacterial load due to either bacillary death or dormancy followinganti-TBtherapyislikelytoaltertheantigenrecognition repertoire of available T-cells. The remarkable reduction in frequencies of IL7R
a
-expressing antigen-specific CD8+ T-cells after anti-TB therapy supports this notion. Therefore, some populationsofantigen-specificT-cellsthatwereinitiallyprimed by productive infection are possibly deprived of their cognate antigenictarget(s)afteranti-TBtherapy,resultinginT-celldeathin thelongrun.39A noteworthy point is the fitness andfunctionality of the T-cells themselves. The T-cell response to stimuli can be impaired in the event of excessive inflammation or anergy, duetothedownregulationofCD3zetachainexpression,which directly affects TCR activation.40 A similar phenomenon has
alreadybeenobservedinpulmonaryTBandleprosywhereT-cell activity could be restored with the addition of exogenous interleukin 2 (IL-2).41 Since pulmonary TB is a disease
characterized by a largely TNF-
a
-driven, tissue-destructive, hyper-inflammatoryenvironment inthe lung,16 thequalityof antigen-specific T-cell responses is inevitably perturbed. This hasveryrecently beenreconfirmedinpatientswithactiveTB whopresentwith a large proportionof TNF-a
single-positive CD4+T-cellsin peripheralblood.27Besidesdownregulation ofexpression,delayedrecruitmentoftheCD3zetachaintotheTCR complexowingtobindingoflow-affinityantigenalso contrib-utestoreducedCD8+T-cellfunctionality.42
The factthatthecentralmemory CD8+T-cellfrequenciesin blood significantly declined following anti-TB therapy could
imply that these cells home totissue or organ compartments. Okhrimenko and colleagues recently showed that the bone marrowharboursrestingmemory CD4+andCD8+T-cellsalike, thusestablishinganicheforlong-termmemoryT-cells.43Homingof
antigen-experienced memoryCD8+ T-cellstothe lung following anti-TBtherapyisstillunderdebate,consideringthatactivehuman TB granulomas do not harbour functional CD8+ T-cell popula-tions.44,45Indeed,terminally differentiatedCD45RA+CCR7 CD8+
T-cells have been associated with host-protective immune responsesinLTBI.10Thesefindingssuggestthatthelackoffunctional
CD8+T-cellsinactiveTBgranulomasmay,inpart,beresponsiblefor decreased immune surveillance. If anti-TB therapy prompts enhancedtraffickingofMTB-specificmemoryCD8+T-cellstotissue compartments,itmayjustifywhytheirnumbersinperipheralblood decreasedwithtime.
Inconclusion,ourfindingssuggestthatanti-TBtherapymight alterthequalityofCD8+T-cellresponses,basedonthechangesin antigen recognition. This is a crucial observation – anti-TB treatmentmayindirectlytailorthecellularimmunecompartment to better manage the disease. Our report provides a view of antigen-specificCD8+T-cellpopulationsthatcaninlargerstudies beusedtoevaluateandlongitudinallymonitortheeffectivenessof anti-TBtreatmentandvaccineefficacytrials.Thisalsoappliesto theassessmentofnovelhost-directedtherapiesforTB,wherethe assessment of the quality of targeted and protective anti-MTB immuneresponsesisdesirable.
Acknowledgements
ThestudywasfundedinpartbyEDCTPTBNEAT(RAR,MB,MM, and AZ), and grants from VR, HLF, and Vinnova to MM. RAR receivedagrantfromKarolinskaInstitutet(KID);AGLreceivedan InnovationPostdoctoralfellowshipfromtheDST/NRF.
Ethical approval: Peripheral blood mononuclear cells were acquiredfromthepatientafterreceivingtheirinformedconsent,as wellasendorsementfromtheinstitutionalreviewboards(number N05/11/187;HealthResearchandEthicsCommittee,Stellenbosch University).Ethicalconsentfromtheethicscommitteein Stock-holm was also obtained (Ref. 2011/863-31/2, including the samplesfromZambia).
Conflictofinterest:Theauthorsdeclarenoconflictofinterest.
AppendixA.Supplementarydata
Supplementarydataassociatedwiththisarticlecanbefound,in theonlineversion,athttp://dx.doi.org/10.1016/j.ijid.2015.01.017.
References
1.O’GarraA,RedfordPS,McNabFW,BloomCI,WilkinsonRJ,BerryMP.The immuneresponseintuberculosis.AnnuRevImmunol2013;31:475–527.
2.PrezzemoloT,GugginoG,LaMannaMP,DiLibertoD,DieliF,CaccamoN. FunctionalsignaturesofhumanCD4andCD8TcellresponsestoMycobacterium tuberculosis.FrontImmunol2014;5:180.
3.AbbasA,LichtmanA,PilaiS.Basicimmunology:functionsanddisordersofthe immunesystem.Philadelphia:ElsevierSaunders;2014.
4.BeharSM.Antigen-specificCD8(+)Tcellsandprotectiveimmunityto tubercu-losis.AdvExpMedBiol2013;783:141–63.
5.VitaR,ZarebskiL,GreenbaumJA,EmamiH,HoofI,SalimiN,etal.Theimmune epitopedatabase2.0.NucleicAcidsRes2010;38:D854–62.
6.HammondAS,KleinMR,CorrahT,FoxA,JayeA,McAdamKP,etal. Mycobacte-riumtuberculosisgenome-widescreenexposesmultipleCD8Tcellepitopes.Clin ExpImmunol2005;140:109–16.
7.LewinsohnDA,WinataE,SwarbrickGM,TannerKE,CookMS,NullMD,etal. ImmunodominanttuberculosisCD8antigenspreferentiallyrestrictedby HLA-B.PLoSPathog2007;3:1240–9.
8.ShamsH,BarnesPF,WeisSE,KlucarP,WizelB.HumanCD8+Tcellsrecognize epitopesofthe28-kDahemolysinandthe38-kDaantigenofMycobacterium tuberculosis.JLeukocBiol2003;74:1008–14.
9.ChaitraMG,ShailaMS,NayakR.Detectionofinterferongamma-secretingCD8+ TlymphocytesinhumansspecificforthreePE/PPEproteinsofMycobacterium tuberculosis.MicrobesInfect2008;10:858–67.
10.CaccamoN,GugginoG,MeravigliaS,GelsominoG,DiCarloP,TitoneL,etal. Analysis of Mycobacterium tuberculosis-specific CD8 T-cells in patients withactivetuberculosisandinindividualswithlatentinfection.PLoSOne 2009;4:e5528.
11.WeicholdFF,MuellerS,KortsikC,HitzlerWE,WulfMJ,HoneDM,etal.Impact ofMHCclassI allelesonthe M.tuberculosis antigen-specificCD8+ T-cell responsein patients with pulmonary tuberculosis. Genes Immun 2007;8: 334–43.
12.LalvaniA,BrookesR,WilkinsonRJ,MalinAS,PathanAA,AndersenP,etal. Humancytolyticandinterferongamma-secretingCD8+Tlymphocytesspecific forMycobacteriumtuberculosis.ProcNatlAcadSciUSA1998;95:270–5.
13.Axelsson-RobertsonR,WeicholdF,SizemoreD,WulfM,SkeikyYA,SadoffJ, etal.ExtensivemajorhistocompatibilitycomplexclassIbindingpromiscuity forMycobacteriumtuberculosisTB10.4 peptidesandimmunedominanceof humanleucocyteantigen(HLA)-B*0702andHLA-B*0801allelesinTB10.4CD8 T-cellresponses.Immunology2010;129:496–505.
14.CommandeurS,LinMY,vanMeijgaardenKE,FriggenAH,FrankenKL,Drijfhout JW,etal.Double-andmonofunctionalCD4(+)andCD8(+)T-cellresponsesto MycobacteriumtuberculosisDosRantigensandpeptidesinlong-termlatently infectedindividuals.EurJImmunol2011;41:2925–36.
15.SchubertOT,MouritsenJ,LudwigC,RostHL,RosenbergerG,ArthurPK,etal. TheMtbproteome library:aresource ofassaystoquantifythecomplete proteomeofMycobacteriumtuberculosis.CellHostMicrobe2013;13:602–12.
16.ZumlaA,RaoM,ParidaSK,KeshavjeeS,CassellG,WallisR,etal.Inflammation andtuberculosis:host-directedtherapies.JInternMed2014.http://dx.doi.org/ 10.1111/joim.12256[Epubaheadofprint].
17.RobinsonJ,HalliwellJA,McWilliamH,LopezR,ParhamP,MarshSG.TheIMGT/ HLAdatabase.NucleicAcidsRes2013;41:D1222–7.
18.MiddletonD,MenchacaL,RoodH,KomerofskyR.Newallelefrequency data-base:http://www.allelefrequencies.net.TissueAntigens2003;61:403–7.
19.Axelsson-RobertsonR,LoxtonAG,WalzlG,EhlersMM,KockMM,ZumlaA,etal. Abroadprofileofco-dominantepitopesshapestheperipheralMycobacterium tuberculosisspecificCD8+T-cellimmuneresponseinSouthAfricanpatients withactivetuberculosis.PLoSOne2013;8:e58309.
20.CaccamoN,MeravigliaS,LaMendolaC,GugginoG,DieliF,SalernoA. Pheno-typicalandfunctionalanalysisofmemoryandeffectorhumanCD8Tcells specificformycobacterialantigens.JImmunol2006;177:1780–5.
21.NyendakMR,ParkB,NullMD,BasekeJ,SwarbrickG,Mayanja-KizzaH,etal. TuberculosisResearchUnitandtheTuberculosisTrialsConsortium. Mycobac-teriumtuberculosisspecificCD8(+)Tcellsrapidlydeclinewithantituberculosis treatment.PLoSOne2013;8:e81564.
22.AltmanJD,MossPA,GoulderPJ,BarouchDH,McHeyzer-WilliamsMG,BellJI, etal.Phenotypicanalysisofantigen-specificTlymphocytes.Science1996;274: 94–6.
23.MazzucchelliR,DurumSK.Interleukin-7receptorexpression:intelligent de-sign.NatRevImmunol2007;7:144–54.
24.deSaintBasileG,MenascheG,FischerA.Molecularmechanismsofbiogenesis andexocytosisofcytotoxicgranules.NatRevImmunol2010;10:568–79.
25.World Health Organization. Global tuberculosis report 2014. Geneva, Switzerland:WHO;2014.
26.BrunsH,MeinkenC,SchauenbergP,HarterG,KernP,ModlinRL,etal.Anti-TNF immunotherapyreducesCD8+Tcell-mediatedantimicrobialactivityagainst Mycobacteriumtuberculosisinhumans.JClinInvest2009;119:1167–77.
27.RozotV,PatriziaA,ViganoS,Mazza-StalderJ,IdriziE,DayCL,etal.Combined useofMycobacteriumtuberculosis-specificCD4andCD8T-cellresponsesisa powerfuldiagnostictoolofactive tuberculosis.ClinInfect Dis2015;60(3): 432–7.http://dx.doi.org/10.1093/cid/ciu795.Epub2014Oct31.
28.Tena-CokiNG,ScribaTJ,PeteniN,EleyB,WilkinsonRJ,AndersenP,etal.CD4 andCD8T-cellresponsestomycobacterialantigensinAfricanchildren.AmJ RespirCritCareMed2010;182:120–9.
29.LiY,ZhuY,ZhouL,FangY,HuangL,RenL,etal.UseofHLA-DR*08032/E7and HLA-DR*0818/E7tetramersintrackingofepitope-specificCD4+Tcellsinactive andconvalescenttuberculosispatientscomparedwithcontroldonors. Immu-nobiology2011;216:947–60.
30.TheronG,PeterJ,LendersL,vanZyl-SmitR,MeldauR,GovenderU,etal. CorrelationofMycobacteriumtuberculosisspecificandnon-specificquantitative Th1T-cellresponseswithbacillaryloadinahighburdensetting.PLoSOne 2012;7:e37436.
31.GarciaJacoboRE,SerranoCJ,EncisoMorenoJA,GasparRamirezO,Trujillo OchoaJL,UrestiRiveraEE,etal.AnalysisofTh1,Th17andregulatoryTcellsin tuberculosiscasecontacts.CellImmunol2014;289:167–73.
32.LinPL,FlynnJL.Understandinglatenttuberculosis:amovingtarget.JImmunol 2010;185:15–22.
33.AndersenP,DohertyTM.ThesuccessandfailureofBCG—implicationsfora noveltuberculosisvaccine.NatRevMicrobiol2005;3:656–62.
34.CohenT,vanHeldenPD,WilsonD,ColijnC,McLaughlinMM,AbubakarI,etal. Mixed-strainMycobacteriumtuberculosisinfectionsandtheimplicationsfor tuberculosistreatmentandcontrol.ClinMicrobiolRev2012;25:708–19.
35.AlpizarYA,ChainB,CollinsMK,GreenwoodJ,KatzD,StaussHJ,etal.Tenyearsof progressinvaccinationagainstcancer:theneedtocounteractcancerevasionby dualtargetinginfuturetherapies.CancerImmunolImmunother2011;60:1127–35.
36.CanettiG.Presentaspectsofbacterialresistanceintuberculosis.AmRevRespir Dis1965;92:687–703.
37.GengenbacherM,KaufmannSH.Mycobacteriumtuberculosis:successthrough dormancy.FEMSMicrobiolRev2012;36:514–32.
38.McCuneRM,FeldmannFM,LambertHP,McDermottW.Microbialpersistence. I.Thecapacityoftuberclebacillitosurvivesterilizationinmousetissues.JExp Med1966;123:445–68.
39.MarkiewiczMA,BrownI,GajewskiTF.DeathofperipheralCD8+Tcellsinthe absenceofMHCclassIisFas-dependentandnotblockedbyBcl-xL.EurJ Immunol2003;33:2917–26.
40.BaniyashM.TCRzeta-chaindownregulation:curtailinganexcessive inflam-matoryimmuneresponse.NatRevImmunol2004;4:675–87.
41.SeitzerU,KayserK,HohnH,EntzianP,WackerHH,PloetzS,etal.ReducedT-cell receptorCD3zeta-chainproteinandsustainedCD3epsilonexpressionatthe siteofmycobacterialinfection.Immunology2001;104:269–77.
42.YachiPP, AmpudiaJ,ZalT,GascoigneNR.Alteredpeptideligandsinduce delayed CD8-Tcell receptor interaction—arole forCD8 in distinguishing antigenquality.Immunity2006;25:203–11.
43.OkhrimenkoA,GrunJR,WestendorfK,FangZ,ReinkeS,vonRothP,etal. HumanmemoryTcellsfromthebonemarrowarerestingandmaintain long-lastingsystemicmemory.ProcNatlAcadSciUSA2014;111:9229–34.
44.RahmanS,GudettaB,FinkJ,GranathA,AshenafiS,AseffaA,etal. Compart-mentalizationofimmuneresponsesinhumantuberculosis:fewCD8+effector TcellsbutelevatedlevelsofFoxP3+regulatorytcellsinthegranulomatous lesions.AmJPathol2009;174:2211–24.
45.AnderssonJ,SamarinaA,FinkJ,RahmanS,GrundstromS.Impairedexpression ofperforinandgranulysininCD8+Tcellsatthesiteofinfectioninhuman chronicpulmonarytuberculosis.InfectImmun2007;75:5210–22.
