Main Human Immunity ToT Cell Subsets

www.elsevierhealth.com/journals/tube

Tuberculosis

Human immunity to M. tuberculosis: T cell subsetsand antigen processing

W.H. Boom*, David H. Canaday, Scott A. Fulton, Adam J. Gehring,Roxana E. Rojas, Marta Torres

Tuberculosis Research Unit (TBRU), Case Western Reserve University and Division of Infectious Diseases,University Hospitals of Cleveland, OH, USA

Summary A hallmark of M. tuberculosis infection is the ability of most (9095%)healthy adults to control infection through acquired immunity, in which antigenspecific Tcells and macrophages arrest growth of M. tuberculosis bacilli and maintaincontrol over persistent bacilli. In addition to CD4+ Tcells, other Tcell subsets such as,gd, CD8+ and CD1-restricted T cells have roles in the immune response to M.tuberculosis. A diverse T cell response allows the host to recognize a wider range ofmycobacterial antigens presented by different families of antigen-presentingmolecules, and thus greater ability to detect the pathogen. Macrophages are keyantigen presenting cells for T cells, and M. tuberculosis survives and persists in thiscentral immune cell. This is likely an important factor in generating this T celldiversity. Furthermore, the slow growth and chronic nature of M. tuberculosisinfection results in prolonged exposure to antigens, and hence further T cellsensitization. The effector mechanisms used by T cells to control M. tuberculosis arepoorly understood. To survive in macrophages, M. tuberculosis has evolvedmechanisms to block immune responses. These include modulation of phagosomes,neutralization of macrophage effector molecules, stimulating the secretion ofinhibitory cytokines, and interfering with processing of antigens for T cells. Therelative importance of these blocking mechanisms likely depends on the stage ofM. tuberculosis infection: primary infection, persistence, reactivation or activetuberculosis. The balance of the hostpathogen interaction in M. tuberculosisinfection is determined by the interaction of T cells and infected macrophages. Theoutcome of this interaction results either in control of M. tuberculosis infection oractive disease. A better understanding of this interaction will result in improvedapproaches to treatment and prevention of tuberculosis.r 2003 Elsevier Science Ltd. All rights reserved.

Natural history of M: tuberculosisinfection

Aerosolized M. tuberculosis bacilli are efficientlytransmitted from person to person. Only small

numbers of bacilli need enter distal alveoli ofhuman lungs to establish infection. In most persons,local innate immunity, mediated primarily byalveolar macrophages, fails to control the slowlyreplicating bacilli. As a result, the immune systemis exposed to increasing amounts of mycobacterialantigen resulting in development of adaptiveimmunity. In most healthy adults, adaptive immu-nity mediated by T cells controls but does noteradicate M. tuberculosis infection. Thus, ongoing

*Corresponding author. Tuberculosis Research Unit, CaseWestern Reserve University, 10900 Euclid Ave., Cleveland, OH44106-4984, USA. Tel.: +1-216-368-4844; fax: +1-216-368-2034E-mail address: [email protected] (W.H. Boom).

1472-9792/03/$ - see front matter r 2003 Elsevier Science Ltd. All rights reserved.doi:10.1016/S1472-9792(02)00054-9

Tuberculosis (2003) 83, 98106

protective immunity is required to maintain controlover persistent bacilli, which are thought to bepresent in one-third of the worlds population.Adaptive immunity also protects against reinfec-tion. The latter is particularly important in areaswith high levels of M. tuberculosis transmission.During the last 10 years great progress has been

made in understanding the human immune responseto M. tuberculosis. The interaction of T cells andinfected macrophages is central to protectiveimmunity to M. tuberculosis. CD4+ T cells have anessential role but are supported by other T cellsubsets such as CD8+, gd TCR+ Tcells (gd Tcells), andCD1 restricted T cells. Antigens for many of these Tcells have been defined, and their function in theimmune response is starting to be elucidated. Insighthas been gained into mechanisms used by macro-phages to control M. tuberculosis. How T cells helpthem perform this task remains poorly understood.TNF-a, IL-12 and IFN-g are central cytokines inregulatory and effector phases of the immuneresponse to M. tuberculosis. Macrophages are notonly primary effector cells for control of M.tuberculosis but also essential for processing andpresentation of antigens to T cells. To survive (andthrive) in macrophages, M. tuberculosis has evolvedmechanisms to block immune responses. Theseinclude modulation of phagosomes, neutralizationof macrophage effector molecules, inducing secre-tion of inhibitory cytokines, and interference withprocessing of antigens for T cells.The relative importance of different T cell

subsets and mechanisms employed by M. tubercu-losis to interfere with macrophage and T cellfunction likely depends on the phase of theinfection. As outlined in Fig. 1, after an initialinnate phase, acute adaptive immunity develops tocontrol rapidly dividing bacilli. This is followed by achronic memory immune phase necessary forcontrol of persistent bacilli and surveillance for

possible re-infections. Immune failure during acuteor chronic adaptive immunity results in clinicaltuberculosis, resulting in M. tuberculosis spreadingto a new host. The balance of the hostpathogeninteraction in M. tuberculosis infection is deter-mined by the interaction of T cells and infectedmacrophages. The ability of M. tuberculosis tomodulate antigen presenting and microbicidalfunction of macrophages is essential for its survivalas one of the most successful human pathogens.Current knowledge about the role and function ofT cell subsets and the mechanisms used by M.tuberculosis to interfere with the ability of macro-phages to process T cell antigens and with T cellfunction, will be reviewed.

The central role of CD4+ T cells inprotective immunity to M: tuberculosis

Studies in humans and animal models demonstratethat acquired immunity to M. tuberculosis requirescontributions by multiple T cell subsets, whichinclude a dominant role for CD4+ T cells andsignificant roles for CD8+ and gd Tcells.1 The reasonsfor involvement of these multiple T cell subsets arenot known. Diversity in T cells that differ in antigenprocessing mechanisms and molecules used forantigen presentation greatly expands the repertoireof mycobacterial antigens recognized. The ability ofM. tuberculosis to survive and persist in one of themajor antigen-presenting cells, the macrophage,contributes further to ready T cell activation. Theslow growth and chronic nature of M. tuberculosisinfection results in prolonged exposure to a largediversity of antigens. From the hosts perspective, adiverse T cell repertoire allows recognition of a widerange of mycobacterial antigens presented bydifferent families of antigen-presenting molecules,

Bact

eria

l Loa

d

Innate (TST -)Adaptive (TST +)

Latent Infection (90%)

ReactivationDisease (5-10%)

1 ProgressiveDisease (< 5%)

Acute ChronicImm

unity

Figure 1 Natural history of M. tuberculosis infection. TST is Tuberculin Skin Test.

Human immunity to M. tuberculosis 99

and thus greater ability to detect the pathogen. ForM. tuberculosis to be such a successful persistentpathogen requires that it evolve mechanisms tointerfere with T cell function.CD4+ abTCR+ T cells (CD4+ T cells) are central to

the human immune response to M. tuberculosis.The HIV pandemic provides direct evidence thatloss of CD4+ T cell number and function resulted inprogressive primary infection, reactivation of en-dogenous M. tuberculosis and enhanced suscept-ibility to re-infection.24 Early cell transfer studiesin mice established that CD4+ T cells transferedprotection against M. tuberculosis.5 In vivo deple-tion with anti-CD4 monoclonal antibodies con-firmed these findings.6,7 Mice with deleted genesfor CD4 or MHC class II molecules are markedlysusceptible to M. tuberculosis, firmly establishing acentral role for CD4+ T cells in protection.8,9

Human CD4+ T cells activated by M. tuberculosisantigens and specific CD4+T cells secrete themacrophage activating cytokines IFN-g and TNF-a/b, are cytotoxic (CTL) for M. tuberculosisinfected macrophages and help macrophages con-trol intracellular mycobacteria.1018 As CTL, afteractivation by M. tuberculosis, CD4+ T cells expressgranzymes, Fas-L (CD95L), granulysin, and perfor-in.19 However, its not clear if CTL function isassociated with increased control of M. tubercu-losis in macrophages. CD4+ T cells provide help forgd and CD8+T cells through IL-2 secretion. All thesefunctions are likely important for CD4 Tcells role inprotective immunity but may be expressed differ-entially during the different stages of the infection.CD4+ T cells however, clearly are critical at allstages of M. tuberculosis infection.The importance of IFN-g was demonstrated by

the increased susceptibility to mycobacterial in-fection of children with IFN-gR deficiency.20,21 IFN-gR knock-out mouse also are markedly susceptibleto M. tuberculosis.22,23 How IF-g mediates protec-tion is less clear. In humans, IFN-g alone isinefficient in enhancing the ability of humanmacrophages to control M. tuberculosis.24,25 Inmice, protective immunity by CD4+ T cells cannotbe explained by IFN-g alone, since MHC class II andCD4 knock-out mice produce large amounts of thiscytokine yet fail to control M. tuberculosis.8

Besides activating microbial killing mechanisms,IFN-g also is a critical regulator of antigen-present-ing cell function by increasing MHC and co-stimulatory molecule expression. This latter activ-ity is essential for optimal T cell responses to M.tuberculosis, and may represent the primary rolefor IFN-g in mycobacterial immunity. In addition toIFN-g, TNFa and IL-12 have been demonstrated inhuman and murine models to have critical roles in

protective immunity to M. tuberculosis.26,27 IL-12functions to increase IFN-g production. How TNF-gexerts its protective effect in humans is not defined.The antigen repertoire for human M. tuberculosis

specific CD4+ T cells is characterized by responsesto a large number of antigens.17,28,29 CD4+T cellsrecognize mycobacterial peptide fragments pre-sented to them by MHC class II molecules onantigen-presenting cells such as macrophages. Nosingle immunodominant antigen has emerged todate, but a number of antigens have beenidentified that are recognized by a majority ofhealthy tuberculin skin-test positive persons. Theseinclude the three 3032kDa 85 complex proteins,ESAT-6 and CFP-10, the 19 and 38 kDa lipoproteinsand two recently described proteins of 32 (serineprotease) and 39 kDa.3036 The 3032 kDa 85A,B,Ccomplex antigens are mycolyl transferases involvedin mycobacterial cell wall synthesis, and arerestricted to mycobacterial species.37 85B, themajor 30 kDa protein of M. tuberculosis, is recog-nized readily by T cells from M. tuberculosisinfected persons, and a number of MHC class IIrestricted epitopes have been mapped.38 Genes forESAT-6 and CFP-10 are present in M. tuberculosisbut deleted in all BCG substrains (region ofdifference 1, RD-1, between M. tuberculosis andM. bovis) and thus are of particular interest fordiagnostic tests aimed at distinguishing thesemycobacterial species. Although CD4+ T cell re-activity to the 19 and 38 kDa lipoproteins can befound, these proteins have been disappointing inanimal vaccine studies. They have, however, gainedpromise as stimuli of toll-like receptors (TLR) onmacrophages. Proteins of 32 and 39 kDa wereidentified by screening protein antigens of M.tuberculosis for reactive T cells from healthytuberculin skin-test positive persons. Vaccine stu-dies in animal models by combining ESAT-6 with 85Band 32 kDa with the 39 kDa protein have beenpromising, warranting their development for test-ing as subunit vaccines in humans. Access to the M.tuberculosis genome will greatly facilitate theidentification and expression of additional CD4+T cell antigens. Studies in human populations withdifferent susceptibilities and responses to M.tuberculosis will determine which antigens andfunctions are critical for CD4+ T cells during thedifferent phase of M. tuberculosis infection.

MHC class I restricted CD8+ T cells

CD8+abTCR+ T cells (CD8+T cells) are activated byM. tuberculosis and BCG.3941 M. tuberculosis

100 W.H. Boom et al.

reactive CD8+ T cells are found among alveolar andperipheral T cells of healthy tuberculin skin-testpositive persons.39 M. tuberculosis activated CD8+ Tcells secrete IFN-g, but less than CD4+ T cells.40 Theyexpress granzymes, Fas-L (CD95L), granulysin, andperforin, which enables them to lyse infected macro-phages.19 CD8+ T cells can help macrophages controlintracellular mycobacteria. Murine studies establisheda role for MHC-I restricted CD8+ T cells in protectiveimmunity particularly during the later stages ofinfection in the lung. To what extent protection byCD8+ T cells depends on CTL activity is unclear. Forhumans, it is not known at what stages of M.tuberculosis infection CD8+ T cells are most critical.The majority of M. tuberculosis reactive CD8+

T cells recognize mycobacterial peptides in thecontext of MHC class I molecules. Epitope mappingstudies, usually guided by epitope predictionalgorithms, have defined HLA-A2 restricted epi-topes on already defined M. tuberculosis antigenssuch as ESAT-6, 85B, 85A, the 19k Da lipoprotein andCFP-10.4245 Others have used a proteinomicapproach by eluting peptides from human MHCclass I molecules (HLA-A2).46 This approach canidentify novel M. tuberculosis antigens by deter-mining the epitopes presented by MHC-I moleculeson infected cells. Whether the repertoire ofantigens recognized by CD8+ and CD4+ T cellsoverlap or are distinct will be determined as moreantigens for both T cell populations are defined byproteinomic approaches. Class I and class II MHCmolecules have distinct intracellular traffickingpathways and mechanisms for loading of microbialpeptides. These different pathways have a key rolein determining the repertoire of antigens recog-nized by CD4+ and CD8+ T cells.M. tuberculosis antigens can enter the MHC class

I antigen processing pathway in human macro-phages through an alternate pathway, that does notrequire traffic through the classic ER-Golgi path-way.40 Traditionally, class I MHC molecules presentmicrobial peptides derived from antigens present incytosol either through de novo synthesis (viralantigens) or carried by microbial agents capableof penetrating into the cytoplasm (e.g. Listeriamonocytogenes, Leishmania etc.). Once in thecytosol, antigens are cleaved by proteasomes andpeptide fragments delivered to the endoplasmicreticulum, where they bind to newly synthesizedclass I MHC molecules and are transported toplasma membrane for recognition by CD8+ T cells.Particulate bacterial antigens, such as M. tubercu-losis, can be processed for class I MHC presentationby an alternate pathway that does not requirepenetration into cytosol and thus providing anadditional mechanism for activation of CD8+ Tcells.

cd T cells and other T cell populations

gd TCR expressing T cells (gd T cells) are character-ized by a unique T cell antigen receptor (TCR)comprised of g and d chains. The majority ofcirculating gd Tcells in adults express Vd9 (aka Vg2)and Vd2 elements (Vd2+ T cells). M. tuberculosisbacilli readily activate Vd2+ Tcells.47 Similar to CD4and CD8+ Tcells, Vd2+Tcells secrete IFN-g, can lyseinfected macrophages and can help contain myco-bacterial growth.17,48 Individuals sensitized tomycobacterial antigens have a greater ability toactivate gd-T cells in response to M. tuberculosisthan tuberculin negative persons.49,50 Vaccinationof adults with BCG increases in vitro expansion ofVg9 Vd2 T cells after stimulation with M. tubercu-losis antigens.51 Tuberculosis patients have adiminished ability to activate gd T cells in responseto M. tuberculosis.50 gd T cells are found in thelungs of tuberculosis patients, and there is acorrelation between the absence or loss of Vg9Vd2 Tcells in blood and lung, and extent of disease.52 In aprimate model gd T cell number and reactivity tophosphate antigens increases during primary my-cobacterial infection and during challenge afterBCG vaccination, suggesting a role in protectiveimmunity.53

Vd2+ T cells react to small phosphate containingmolecules that can be divided into two groups.There are nucleotide-conjugated phosphatemolecules, such as TUBAg3-4, isopentenyl-ATP,and pyrophosphate molecules such as TUBag1-2,isopentenyl pyrophosphate [IPP], mono-ethylpyrophosphate (MEPP) and others (reviewed inRef.5456). Recognition of phospho-molecules isTCR-dependent but not restricted or dependenton any known MHC or MHC-like molecules. Speci-ficity of prenylphosphate recognition depends onparticular CDR3 regions in the Vg9 chain and theVd2 chain. The phosphoantigens are intracellularand not secreted by M. tuberculosis. Prenylpyrophosphates are ubiquitous precursors for cho-lesterol and its derivatives, and for terpenoids.Discrimination between infected and non-infectedcells may involve recognition of metabolic inter-mediates produced by bacteria.57 Thus, despitesimilar functions, Vd2+ T cells recognize an entirelydifferent range of mycobacterial molecules thanCD4+ and CD8+ Tcells. How these phospho-antigensare processed and presented to gd T cells (if at all)by infected macrophages is unknown.58,59 Even ifthere is no antigen processing, macrophages areimportant for activating gd T cells because wheninfected they represent a concentrated source ofphospho-antigen and because they enhance gdT cell responses by providing co-stimulation.


One additional T cell subset that respondsto M. tuberculosis is the CD1 restricted abTCR+ T cell.60 It is the least common T cellsubset in human peripheral blood and lung. Inhumans, most of these T cells express neitherCD4 or CD8 and are referred to as doublenegative (DN) cells. A minority expresses CD8.CD1 is an MHC-like molecule that associateswith b2-microglobulin that can be induced bycytokines or constitutively expressed on antigen-presenting cells such as dendritic cells. CD1molecules have very little polymorphism andhave the unique ability to bind polar non-peptide lipid antigens of M. tuberculosis suchas mycolic acids and phosphatidyl-inositol-mannosides (PIMs). CD1 restricted T cells secreteIFN-g, are cytotoxic for infected macrophages andcan help macrophages control intracellular myco-bacteria.

Thus, the human T cell response to M. tubercu-losis is characterized by the participation of multi-ple T cell subsets with similar functions (Fig. 2).They secrete IFN-g and TNF-a, can lyse infectedcells as CTL and can help macrophages control M.tuberculosis growth. They differ markedly, how-ever in the range of mycobacterial antigens theyrecognize and the antigen-processing mecha-nisms used to process and present these antigensto their TCRs. T cell subsets likely also differ instage of infection when they are most active oressential, and in ability to enter sites ofactive infection (granuloma, cavity, lymph node).Some have suggested that gd T cells and CD1restricted T cells may provide a link between theinnate and adaptive phases of the immuneresponse to M. tuberculosis. CD4+ T cells areknown to have a key role through all stages of M.tuberculosis infection but may differ as to whencytokine secretion vs. CTL function is the mostimportant.

T cell inhibition by M: tuberculosis

The complexity and diversity of the adaptiveimmune response to M. tuberculosis require devel-opment of mechanisms to interfere and inhibit hostimmunity for pathogen survival. M. tuberculosisevolved to survive and persist in key immuneeffector cells, the macrophage. Macrophages read-ily take up microbial pathogens, have effectivemicrobicidal mechanisms, process and presentantigens for T cell recognition, and express thenecessary costimulatory molecules to activateT cells.M. tuberculosis has evolved at least two mechan-

isms to interfere with innate immune defenses ofmacrophages. M. tuberculosis bacilli are taken upby receptor mediated phagocytosis using a varietyof macrophage receptors including CR3, CR4 andmannose receptor. Mycobacteria remain withinphagosomes and do not penetrate into cytoplasm.M. tuberculosis modulates its phagosomal compart-ment by preventing fusion with acidic lysosomalcompartments and actively excludes vesicularproton ATP-ases, resulting in an elevated pH of6.36.5 (compared to the normal lysosomal pH of4.5).61 IFNg can partially overcome M. tubercu-losis mediated inhibition of phagosomal acidifica-tion. In addition to modulation of the phagosome toprevent critical proteases from attacking it, M.tuberculosis resists killing by oxygen radical inter-mediates, through superoxide dismutases and otherenzymes. M. tuberculosis is sensitive to nitric oxide(NO) made by inducible nitric oxide synthase(iNOS), but whether enough iNOS can be inducedin human macrophages to produce sufficient levelsof NO remains controversial.M. tuberculosis also has a variety of mechanisms

to interfere with adaptive immune function. First,molecules of M. tuberculosis readily induce macro-phages to produce cytokines that inhibit T cell

Phos. Ag

-TCRCD 4T cell

class II MHCclass IMHC

-TCRCD 8T cell

CD1 restricted T cellsDN TCR

M

T cell

-TCR

Figure 2 Many T cell subsets respond to M. tuberculosis infected macrophages.


function. These include inhibitory cytokines such asIL-10 and TGFb, and are produced by the samecells producing pro-inflammatory and Th-1 promot-ing cytokines such as IL-12, IL-15, TNFa, and IL-1.Excess production of IL-10 and TGFb, as seen inactive tuberculosis, inhibits the effects of the pro-inflammatory cytokines and directly inhibits T cellfunction.Recent studies suggest that recognition of patho-

gen-associated molecular patterns (PAMPs) by Toll-like receptors (TLR) may be a primary signal forcytokine release by macrophages. There are 10known TLRs of which macrophages express many.62

TLR-4 is the primary receptor for LPS from Gram-negative bacteria, and TLR-2 recognizes Gram-positive bacteria.6365 TLR-2 and TLR-4 recognizeM. tuberculosis associated PAMPs, with TLR-2recognizing mycobacterial lipoproteins such as19 kDa.65,66 Signalling through TLR results in macro-phage cytokine secretion, which includes both pro-inflammatory and inhibitory cytokine secretion.Whether specific M. tuberculosis PAMPs differ inthe range or balance of pro-inflammatory andinhibitory cytokines they stimulate, is not known.Second, active M. tuberculosis infection is

associated with increased apoptosis of mycobac-terial antigen-specific T cells.67 M. tuberculosis-specific T cell apoptosis has a prolonged effect on Tcell responses to mycobacterial antigens. Long (612 months) after initiation of TB treatment andcontrol of bacterial replication, CD4+ T cellresponses to proteins of M. tuberculosis remainmarkedly diminished compared to responses tocontrol antigens such as tetanus toxoid or themitogen PHA.68 This suggests that T cell apoptosisduring acute infection can result in prolongeddefects in T cell repertoire and function. Themolecules of M. tuberculosis and mechanismresponsible for T cell apoptosis have not beendefined.

Modulation of macrophage antigen-presenting cell function byM: tuberculosis

Recent studies suggest an important third way forM. tuberculosis to evade adaptive immune re-sponses, namely interfering with antigen processingand presentation by macrophages. Processing ofantigens is essential for T cells to recognize cellsinfected by microbes and is regulated by complexcellular processes resulting in peptide presentationby two distinct sets of molecules: MHC class II forCD4+ and MHC class I for CD8+ T cells. Little is

known about inhibitors of antigen processing forMHC class I and class I-like molecules. However,recent studies are starting to shed light onmechanisms for inhibition of MHC class II antigenprocessing and the molecule(s) of M. tuberculosisresponsible for this inhibition.For class II MHC processing, antigens normally

are internalized by endocytosis or phagocytosis,and concentrated within endosomes. As endosomesfuse with lysosomes, proteases break downprotein into peptides. Class II MHC molecules areconcentrated in a late endocytic compartment, theclass II MHC compartment (MIIC). Class II MHCpeptide complexes are formed in these compart-ments with peptides binding to class II MHCpromoted by the acidic pH. Class II MHC moleculesare targeted for endosomes by specific sequenceson the invariant chain associated with the a and bchains of class II MHC as it emerges from endoplas-mic reticulum (ER). Invariant chain contains a CLIPregion which blocks the class II MHC peptide bindingsite. CLIP is exchanged for antigen peptide withhelp from HLA-DM molecules. In murine macro-phages M. tuberculosis phagosomes contain MHC IIcapable of presenting mycobacterial peptide di-rectly to T cells.69

M. tuberculosis infected monocytes do notpresent tetanus toxoid as well as uninfected cells.70

In murine macrophages, IL-6 secretion has beenshown under some circumstances to inhibit T cellfunction, and M. tuberculosis has been shown tointerfere with maturation of MHC class II and withIFN-g signaling in THP-1 cells.7174 In murinemacrophages, mycobacterial infection was asso-ciated with decreased CIITA (class II transactivator)expression, resulting in decreased MHC class IIlevels.75 Recent studies indicate that the 19 kDalipoprotein of M. tuberculosis through TLR-2inhibits MHC class II expression and antigenprocessing for CD4+ T cells in murine macro-phages.76 These studies have been extended tohuman macrophages, and indicate that M. tuber-culosis and the 19 kDa lipoprotein inhibit IFN-gmediated regulation of human HLA-DR.77 Thisresults in decreased presentation and thus activa-tion of M. tuberculosis specific CD4+ T cellresponses. The ability to interfere with CD4+ T cellactivation by hiding from the immune responsesis likely to be a major mechanism used by M.tuberculosis to avoid detection and eliminationduring the persistent phase of infection. Usingcarefully designed assays to measure specific stagesof antigen processing for MHC class II and Imolecules, it is likely that additional mycobacterialmolecules will be identified that can block antigenprocessing.


Thus, M. tuberculosis has evolved a number ofmechanisms to interfere with activation of bothinnate and adaptive phases of the immune re-sponse. It is likely that the importance of thesedifferent mechanisms will differ for each stage ofM. tuberculosis infection. For example, suppressivecytokines and apoptosis may be more importantduring phases when M. tuberculosis is replicatingrapidly during primary and reactivation phases,whereas inhibition of antigen processing may bethe most important defense mechanisms during thepersistent phase of infection. The balance betweenactivation of multiple T cell subsets capable ofrecognizing a wide range of mycobacterial mole-cules in the context of a wide range of antigen-presenting molecules and the ability of M. tuber-culosis to interfere with immune recognition and toblock the effector phase of adaptive immunitylikely determines the outcome of M. tuberculosisinfection, which ranges from active disease toindefinite persistence. Access to genomic andproteonomic information of M. tuberculosiscoupled with increased knowledge of the cellbiology of antigen processing and T cell activationwill undoubtedly characterize further the molecu-lar mechanims of the critical host pathogeninteraction in M. tuberculosis infection: the inter-action of T cell subsets with infected macrophages.

Acknowledgements

This work was supported by National Institutes ofHealth grants A127243 and HL55967, and contractA195383 to the Tuberculosis Research Unit.

References

1. Boom WH. The role of T cell subsets in M. tuberculosisinfection. Infectious Agents Dis 1996;5:7381.

2. Raviglione MC, Snider Jr. DE, Kochi A. Global epidemiologyof tuberculosis. Morbidity and mortality of a worldwideepidemic [see comments]. JAMA 1995;273:2206.

3. Barnes PF, Bloch AB, Davidson P, Snider D. Tuberculosis inpatients with human immunodeficiency virus infection. NEngl J Med 1991;324:164450.

4. Hopewell PC. Impact of human immunodeficiency virusinfection on the epidemiology, clinical features, manage-ment, and control of tuberculosis. Clin Infect Dis 1992;15:5407.

5. Orme IM, Collins FM. Protection against Mycobacteriumtuberculosis infection by adoptive transfer. J Exp Med1983;158:7483.

6. Muller I, Cobbold S, Waldmann H, Kaufmann SH. Impairedresistance to Mycobacterium tuberculosis infection afterselective in vivo depletion of L3T4+ and Lyt-2+ Tcells. InfectImmun 1987;55(9):203741.

7. Pedrazzini T, Hug K, Louis JA. Importance of L3T4+ and Lyt-2+ cells in the immunologic control of infection withMycobacterium bovis strain bacillus Calmette-Guerin inmice. Assessment by elimination of T cell subsets in vivo.J Immunol 1987;139(6):20327.

8. Caruso AMN, Serbina N, et al. Mice deficient in CD4 Tcells have only transiently diminished levels of IFN-gamma,yet succumb to tuberculosis. J Immunol 1999;162(9):540716.

9. Ladel CH, Daugelat S, et al. Immune response to Mycobac-terium bovis bacille Calmette Guerin infection in majorhistocompatability complex I- and II-deficient knock-outmice: contribution of CD4+ and CD8+ T cells to acquiredresistance. Eur J Immunol 1995;25:37784.

10. Boom WH, Wallis RS, Chervenak KA. Human Mycobacteriumtuberculosis-reactive CD4+ T-cell clones: heterogeneity inantigen recognition, cytokine production, and cytotoxicityfor mononuclear phagocytes. Infect Immun 1991;59(8):273743.

11. Haanen JB, de-Waal-Malefijt R, Res PC, Kraakman EM,Ottenhoff TH, de-Vries RR, Spits H. Selection of a human Thelper type 1-like T cell subset by mycobacteria. J Exp Med1991;174(3):58392.

12. Hansen PW, Petersen CM, Povlsen JV, Kristensen T. Cytotoxichuman HLA class II restricted purified protein derivativereactive T lymphocyte clones. Scand J Immunol1987;25:295303.

13. Mustafa AS, Godal T. BCG-induced suppressor T cells optimalconditions for in vitro induction and mode of action. Clin ExpImmunol 1985;62(3):47481.

14. Kumararatne DS, Pithie AS, Drysdale P, Gaston JS, KiesslingR, Iles PB, Ellis CJ, Innes J, Wise R. Specific lysis ofmycobacterial antigen-bearing macrophages by class II MHC-restricted polyclonal T cell lines in healthy donors orpatients with tuberculosis. Clin Exp Immunol 1990;80(3):31423.

15. Ottenhoff TH, Ab BK, Van-Embden JD, Thole JE, Kiessling R.The recombinant 65-kD heat shock protein of Mycobacter-ium bovis Bacillus Calmette-Guerin/M. tuberculosis is atarget molecule for CD4+ cytotoxic T lymphocytes that lysehuman monocytes. J Exp Med 1988;168(5):194752.

16. Hancock GE, Cohn ZA, Kaplan G. The generation of antigen-specific, major histocompatability complex-restricted cyto-toxic T lymphocytes of the CD4+ phenotype. J Exp Med1989;169:90919.

17. Tsukaguchi K, Balaji KN, et al. CD4+ alpha-beta T cell andgamma delta T cell responses to Mycobacterium tubercu-losis: similarities and differences in antigen recognitioncytotoxic effector function and cytokine production.J Immunol 1995;154:178696.

18. Tsukaguchi K, de Lange B, et al. Differential regulation ofIFN-gamma, TNF-alpha, and IL-10 production by CD4(+)alphabetaTCR+ T cells and vdelta2(+) gammadelta T cells inresponse to monocytes infected with Mycobacterium tuber-culosis-H37Ra. Cell Immunol 1999;194(1):1220.

19. Canaday DH, Wilkinson RJ, Li Q, Harding CV, Silver RF, BoomWH. CD4+ and CD8+ Tcells kill intracellular M. tuberculosisby a perforin and FAS/FASL independent mechanism.J Immunol 2001;167:273442.

20. Jouanguy E, Altare F, et al. Interferon-gamma-receptordeficiency in an infant with fatal bacille Calmette-Guerininfection. N Engl J Med 1996;335(26):195661.

21. Jouanguy E, Lamhamedi-Cherradi S, et al. Partial interferon-gamma receptor 1 deficiency in a child with tuber-culoid bacillus Calmette-Guerin infection and a sibling withclinical tuberculosis. J Clin Invest 1997;100(11):265864.


22. Flynn JL, Chan J, et al. An essential role for interferongamma in resistance to Mycobacterium tuberculosis infec-tion. J Exp Med 1993;178(6):224954.

23. Cooper AM, Dalton DK, et al. Disseminated tuberculosis ininterferon-gamma gene-disrupted mice. J Exp Med 1993;178:22437.

24. Douvas GS, Looker DL, et al. Gamma interferon activateshuman macrophages to 4become tumoricidal and leishma-nicidal but enhances replication of macrophage-associatedmycobacteria. Infect Immun 1985;50(1):18.

25. Silver RF, Li Q, et al. Lymphocyte-dependent inhibition ofgrowth of virulent Mycobacterium tuberculosis H37Rv withinhuman monocytes: requirement for CD4 + T cells in purifiedprotein derivative-positive, but not in purified proteinderivative-negative subjects. J Immunol 1998;160(5):240817.

26. de Jong R, Altare F, Haagen IA, Elferink DG, Boer T, vanBreda Vriesman PJ, Kabel PJ, Draaisma JM, van Dissel JT,Kroon FP, Casanova JL, Ottenhoff TH. Severe mycobacterialand Salmonella infections in interleukin-12 receptor-defi-cient patients. Science 1998;280:14358.

27. Keane J, Gershon S, Wise RP, Mirabile-Levens E, Kasznica J,Schwieterman WD, Siegel JN, Braun MM. Tuberculosisassociated with infliximab, a tumor necrosis factor alpha-neutralizing agent. N Engl J Med 2001;345:1098104.

28. Havlir DV, Wallis RS, Boom WH, Daniel TM, Chervenak K,Ellner JJ. Human immune response to Mycobacteriumtuberculosis antigens. Infect Immun 1991;59(2):66570.

29. Schoel B, Gulle H, Kaufmann SH. Heterogeneity of therepertoire of T cells of tuberculosis patients and healthycontacts to Mycobacterium tuberculosis antigens separatedby high-resolution techniques. Infect Immun 1992;60(4):171720.

30. Wiker HG, Harboe M. The antigen 85 complex: a majorsecretion product of Mycobacterium tuberculosis. MicrobiolRev 1992;56(4):64861.

31. Sorensen AL, Nagai S, Houen G, Andersen P, Andersen A.Purification and characterization of a low-molecular-mass Tcell antigen secreted by Mycobacterium tuberculosis. InfectImmun 1995;63:17107.

32. Dillon DC, Alderson MR, Day CH, Bement T, Campos-Neto A,Skeiky YA, Vedvick T, Badaro R, Reed SG, Houghton R.Molecular and immunological characterization of Mycobac-terium tuberculosis CFP-10, an immunodiagnostic antigenmissing in Mycobacterium bovis BCG. J Clin Microbiol2000;38:328590.

33. Skeiky YA, Lodes MJ, Guderian JA, Mohamath R, Bement T,Alderson MR, Reed SG. Cloning, expression, and immunolo-gical evaluation of two putative secreted serine proteaseantigens of Mycobacterium tuberculosis. Infect Immun1999;67:39984007.

34. Dillon DC, Alderson MR, Day CH, Lewinsohn DM, Coler R,Bement T, Campos-Neto A, Skeiky YA, Orme IM, Roberts A,Steen S, Dalemans W, Badaro R, Reed SG. Molecularcharacterization and human T-cell responses to a memberof a novel Mycobacterium tuberculosis mtb39 gene family.Infect Immun 1999;67:294150.

35. Surcel HM, Troye-Blomberg M, Paulie S, Andersson G, MorenoC, Pasvol G, Ivanyi J. Th1/Th2 profiles in tuberculosis, basedon the proliferation and cytokine response of bloodlymphocytes to mycobacterial antigens. Immunology1994;81:1716.

36. Harris DP, Vordermeier HM, Friscia G, Roman E, Surcel HM,Pasvol G, Moreno C, Ivanyi J. Genetically permissiverecognition of adjacent epitopes from the 19-kDa antigenof Mycobacterium tuberculosis by human and murineT cells. J Immunol 1993;150(11):504150.

37. Belisle JT, Vissa VD, et al. Role of the major antigen ofMycobacterium tuberculosis in cell wall biogenesis. Science1997;276(5317):14202.

38. Silver RF, Wallis RS, Ellner JJ. Mapping of T cell epitopesof the 30-kDa alpha antigen of Mycobacterium bovisBCG in PPD-positive individuals. J Immunol 1994;154:4665.

39. Tan JS, Canaday DH, Boom WH, Balaji KN, Schwander SK,Rich EA. Human alveolar T lymphocyte responses toMycobacterium tuberculosis antigens: role for CD4+ andCD8+ cytotoxic T cells and relative resistance of alveolarmacrophages to lysis. J Immunol 1997;159:2907.

40. Canaday DH, Ziebold C, Noss EH, Chervenak KA, Harding CV,Boom WH. Activation of human CD8+ ab TCR+ cells byMycobacterium tuberculosis via an alternate class IMHC antigen-processing pathway. J Immunol 1999;162:3729.

41. Turner J, Dockrell H. Stimulation of human peripheral bloodmononuclear cells with live Mycobacterium bovis BCGactivates cytolytic CD8+ T cells in vitro. Immunology1996;87(3):33942.

42. Mohagheghpour N, Gammon D, et al. CTL response toMycobacterium tuberculosis: identification of an immuno-genic epitope in the 19-kDa lipoprotein. J Immunol1998;161(5):24006.

43. Lalvani A, Brookes R, et al. Human cytolytic and interferongamma-secreting CD8+ T lymphocytes specific for Mycobac-terium tuberculosis. Proc Natl Acad Sci USA 1998;95(1):2705.

44. Smith SM, Brookes R, et al. Human CD8+ CTL specific for themycobacterial major secreted antigen 85A. J Immunol2000;165(12):708895.

45. Geluk A, van Meijgaarden KE, et al. Identification of majorepitopes of Mycobacterium tuberculosis AG85B that arerecognized by HLA-A*0201-restricted CD8+ T cells in HLA-transgenic mice and humans. J Immunol 2000;165(11):646371.

46. Flyer DC, Ramakrishna V, Miller C, et al., Identification bymass spectrometry of CD8(+)-T-cell Mycobacterium tuber-culosis epitopes within the Rv0341 gene product. InfectImmun 2002;70:292632.

47. Havlir DV, Ellner JJ, Chervenak KA, Boom WH. Selectiveexpansion of human gamma delta T cells by monocytesinfected with live Mycobacterium tuberculosis. J Clin Invest1991;87(2):72933.

48. Dieli F, Troye-Blomberg M, Ivanyi J, Fournie JJ, Krensky AM,Bonneville M, Peyrat MA, Caccamo N, Sireci G, Salerno A.Granulysin-dependent killing of intracellular and extracel-lular Mycobacterium tuberculosis by Vgamma9/Vdelta2 Tlymphocytes. J Infect Dis 2001;184:10825.

49. Kabelitz D, Bender A, Schondelmaier S, Schoel B, KaufmannSH. A large fraction of human peripheral blood gamma/delta+ T cells is activated by Mycobacterium tuberculosisbut not by its 65-kD heat shock protein. J Exp Med1990;171(3):66779.

50. Barnes PF, Grisso CL, Abrams JS, Band H, Rea TH, Modlin RL.Gamma delta T lymphocytes in human tuberculosis. J InfectDis 1992;165(3):50612.

51. Hoft DF, Brown RM, et al. Bacille Calmette-Guerin vaccina-tion enhances human gamma delta T cell responsiveness tomycobacteria suggestive of a memory-like phenotype.J Immunol 1998;161(2):104554.

52. Li B, Rossman MD, et al. Disease-specific changes ingammadelta T cell repertoire and function in patientswith pulmonary tuberculosis. J Immunol 1996;157(9):42229.


53. Shen Y, Zhou D, Qiu L, Lai X, Simon M, Shen L, Kou Z, WangQ, Jiang L, Estep J, Hunt R, Clagett M, Sehgal PK, Li Y, ZengX, Morita CT, Brenner MB, Letvin NL, Chen ZW. Adaptiveimmune response of Vgamma2Vdelta2+ T cells duringmycobacterial infections. Science 2002;295:22558.

54. Rojas R, Boom WH. Gamma delta T cells and HIV-1 infection.In: Cellular aspects of HIV infection. Cytometric cellularanalysis series, New York, NY: p. 147205. 2001. Wiley-LissPress.

55. Constant P, Davodeau F, Peyrat MA, et al. Stimulation ofhuman gamma delta T cells by nonpeptidic mycobacterialligands. Science 1994;264(5156):26770.

56. Tanaka Y, Morita C, Tanaka Y, Nieves E, Brenner MB, BloomBM. Natural and synthetic non-peptide antigens recognizedby human gamma delta T cells. Nature 1995;375:1558.

57. Sicard H, Fournie JJ. Metabolic routes as targets forimmunological discrimination of host and parasite. InfectImmun 2000;68(8):43757.

58. Morita CT, Beckman EM, Bukowski JF, et al. Directpresentation of nonpeptide prenyl pyrophosphate antigensto human gamma delta T cells. Immunity 1995;3(4):495507.

59. Lang F, Peyrat MA, Constant P, et al. Early activation ofhuman V gamma 9V delta 2 Tcell broad cytotoxicity and TNFproduction by nonpeptidic mycobacterial ligands. J Immunol1995;154(11):598694.

60. Ulrichs T, Porcelli SA. CD1 proteins: targets of T cellrecognition in innate and adaptive immunity. Rev Immuno-genet 2000;2:41632.

61. Sturgill-Koszycki S, Sclesinger P, Chakraborty P, et al. Lack ofacidification in Mycobacterium phagosomes produced byexclusion of the vesicular proton-ATPase. Science1994;263:678.

62. Medzhitov R, Janeway Jr. CA. Innate immunity: the virtuesof a nonclonal system of recognition. Cell 1997;91:2958.

63. Yoshimura A, Lien E, et al. Recognition of Gram-positivebacterial cell wall components by the innate immunesystem occurs via Toll-like receptor. J Immunol 1999;1633(1):15.

64. Takeuchi O, Hoshino K, Kawai T, Sanjo H, Takada H, Akira S.Differential roles of TLR2 and TLR4 in recognition of Gram-negative and Gram-positive bacterial cell wall components.Immunity 1999;11:44351.

65. Means TK, Wang S, Lien E, Yoshimura A, Golenbock D, FentonM. Human toll-like receptors mediate cellular activationby Mycobacterium tuberculosis. J Immunol 1999;163:39207.

66. Brightbill HD, Libraty DH, Krutzik SR. Host Defense Mechan-isms Triggered by Microbial Lipoproteins through the Toll-like Receptors. Science 1999;285:7326.

67. Hirsch CS, Toossi Z, Johnson JL, et al. Augmentation ofapoptosis and interferon-gamma production at sites ofactive Mycobacterium tuberculosis infection in humantuberculosis. J Infect Dis 2001;183:77988.

68. Hirsch CS, Toossi Z, Vanham G, et al. Apoptosis and T cellhyporesponsiveness in pulmonary tuberculosis. J Infect Dis1999;179:94553.

69. Ramachandra L, Noss EN, Boom WH, Harding CV. Processingof M. tuberculosis antigen 85B involves intra-phagosomalformation of peptide: MHC-II complexes and is inhibited bylive bacilli that decrease phagosome maturation. J Exp Med2001;194:142132.

70. Gercken J, Pryjma J, et al. Defective antigen presentationby Mycobacterium tuberculosis-infected monocytes. InfectImmun 1994;62(8):34728.

71. VanHeyningen TK, Collins HL, et al. IL-6 produced bymacrophages infected with Mycobacterium species sup-presses T cell responses. J Immunol 1997;158(1):3307.

72. Hmama Z, Gabathuler R, Jefferies WA, de Jong G, Reiner NE.Attenuation of HLA-DR expression by mononuclear phago-cytes infected withMycobacterium tuberculosis is related tointracellular sequestration of immature class II heterodi-mers. J Immunol 1998;161:488293.

73. Hussain S, Zwilling BS, Lafuse WP. Mycobacterium aviuminfection of mouse macrophages inhibits IFN-gamma Januskinase-STAT signaling and gene induction by down-regulationof the IFN-gamma receptor. J Immunol 1999;163:20418.

74. Ting LM, Kim AC, Cattamanchi A, Ernst JD. Mycobacteriumtuberculosis inhibits IFN-g transcriptional responseswithout inhibiting activation of STAT1. J Immunol 1999;163:3898906.

75. Wojciechowski W, DeSanctis J, et al. Attenuation of MHCclass II expression in macrophages infected with Mycobac-terium bovis bacillus Calmette-Guerin involves class IItransactivator and depends on the Nramp1 gene. J Immunol1999;163(5):268896.

76. Noss EH, Sellati TJ, Radolf JD, Belisle J, Golenbock DT, BoomWH, Harding CV. Toll-like receptor 2-dependent inhibition ofmacrophage class II MHC expression and antigen processingby 19 kD lipoprotein of M. tuberculosis. J Immunol2001;167:9108.

77. Gehring A, Rojas R, Canaday D, et al. Inhibition of humanMHC-II antigen processing by 19 kD lipoprotein ofM. tuberculosis. Submitted.


Human immunity to M. tuberculosis: T cell subsets and antigen processingNatural history of M.tuberculosis infectionThe central role of CD4+ T cells in protective immunity to M.tuberculosisMHC class I restricted CD8+ T cellsgammadelta T cells and other T cell populationsT cell inhibition by M.tuberculosisModulation of macrophage antigen-presenting cell function by M.tuberculosisAcknowledgementsReferences

Documents

Main Human Immunity ToT Cell Subsets