INFECnION AND IMMUNIrY, JUlY 1994, p. 2963-2972 Vol. 62, No. 70019-9567/94/$04.00+0Copyright X) 1994, American Society for Microbiology

Epitope Specificity and Isoforms of the Mycobacterial19-Kilodalton Antigen

DAVID P. HARRIS,1 HANS-MARTIN VORDERMEIER,1 SARA J. BREIT,2 GEOFFREY PASVOL,3CARLOS MORENO,1 AND JURAJ IVANYIV*

MRC Clinical Sciences Centre, Tuberculosis and Related Infections Unit, Royal Postgraduate Medical School, HammersmithHospital, London W12 OHS,1 St. Mary's Hospital Medical School, Imperial College of Science, Technolog and

Medicine, Infection and Tropical Medicine, Northwick Park Hospital, Harrow, Middlesex, HAI 3UJ,and Department of Cell Biology, The Wellcome Research Laboratories,

Beckenham, Kent BR3 3BS,2 United Kingdom

Received 13 December 1993/Returned for modification 25 January 1994/Accepted 14 April 1994

The topography and specificity of B- and T-cell stimulatory epitopes from the 19-kDa protein ofMycobacterium tuberculosis were investigated by using overlapping synthetic peptides. Murine antiseraidentified two cryptic epitopes (residues 11 to 30 and 61 to 80) and one species-specific immunodominantepitope (residues 140 to 159). Immunoglobulins GI and G2a antibody isotypes varied for the respective peptideimmunogens but without relationship to the T-cell cytokine profiles which were characterized by high gammainterferon and low interleukin 5 levels. Antisera to recombinant M. tuberculosis 19-kDa protein (rGST-19)cross-reacted with homologous proteins of similar size from organisms of the Mycobacterium avium-intracellulare complex. Two-dimensional gel electrophoresis revealed differences in the number, relativemobility, and charge of isoforms of the 19-kDa protein, possibly reflecting posttranslational modifications. Theimmunodominant T-cell epitope from the M. tuberculosis 19-kDa protein (residues 61 to 80) and thecorresponding peptide sequence from Mycobacterium avium subsp. intracellare (residues 64 to 83), differing atfive residues, were both recognized in a genetically permissive manner. Peptides 61-80 and 64-83 stimulatedcross-reactive responses in BALB/c (H1-2d) mice, while in the C57BL/10 (H-2b) strain, responses to peptide61-80 were species specific. In purified protein derivative-positive healthy individuals, the M. avium subsp.intracellulare peptide stimulated stronger responses than did the M. tuberculosis peptide, whereas patients withactive tuberculosis had enhanced in vitro T-cell responses to both peptides.

Mycobacteria possess a number of highly conserved proteinfamilies, whose members characteristically share a high degreeof amino acid sequence homology. Conserved protein familiesof pronounced immunogenicity include the DnaK, GroEL,GroES, and low-molecular-weight (hsp-18) heat shock pro-teins (5, 19, 38, 41). The 19-kDa protein is not a member of anyknown family of stress proteins. Nevertheless, it is present inboth the Mycobacterium tuberculosis complex (3, 9) and inorganisms of the Mycobacterium avium-intracellulare complex(MAC) (5, 29). The sequences of the M. tuberculosis (3) andMycobacterium bovis (9) 19-kDa proteins are identical andshare approximately 75% amino acid identity with the homol-ogous antigens from Mycobacterium avium subsp. avium (5)and Mycobacterium avium subsp. intracellulare (5, 29). Further-more, 19-kDa protein gene probes have been shown to hybrid-ize with genomic DNA isolated from several other mycobac-terial species including Mycobacterium phlei, Mycobacteriumasiaticum (29), Mycobacterium scrofulaceum, Mycobacteriumkansasii, and Mycobacterium avium subsp. paratuberculosis(hla), suggesting that this antigen may be present in a range ofmycobacterial species.To date, no significant sequence homology has been de-

tected between the mycobacterial 19-kDa proteins and se-quences found in the protein data base, and the biologicalfunction of the molecule remains unknown. The presence ofhighly conserved six-residue consensus sequences for lipida-tion (M. tuberculosis, 19-LSGCSS-24; M. avium subsp. intracel-lulare, 19-ISGCSG-24) located immediately adjacent to the

* Corresponding author. Phone: 081-740 3161. Fax: 081-743 8536.

hydrophobic signal peptide, together with more direct bio-chemical evidence (29, 40), strongly suggests that the 19-kDaprotein undergoes posttranslational acylation and subsequentcleavage of the hydrophobic signal peptide prior to secretion ofthe mature lipoprotein across the cell membrane. The 19-kDaprotein also possesses concanavalin A binding properties ei-ther when produced in its natural host (15) or when expressedas a recombinant protein in Mycobacterium smegmatis (18),indicating that the molecule is probably glycosylated. Althoughthe complete 19-kDa gene product has been overexpressed asa recombinant fusion protein in Escherichia coli (21b) and alsoin the rapidly growing species M. smegmatis (18), it has provendifficult to purify the protein in quantities sufficient for immu-nological evaluation because of its hydrophobicity.The secreted nature of the 19-kDa antigen (15) may con-

tribute to its serological immunodominance by enhancing itsaccessibility in a native form for B-cell recognition. Themonoclonal antibodies (MAbs) TB-23 (IT-19) and F29-47(IT-10), originally used to identify the M. tuberculosis 19-kDaprotein (13, 22), recognize serologically distinct B-cell epitopes(4). Whereas TB-23 recognizes a conformational epitopewhich is dependent upon intramolecular disulfide bonding, theepitope recognized by F29-47 is thought to be linear in nature(4). Following either immunization or tuberculous infection ofdifferent inbred strains of mice, levels of antibody to the19-kDa protein are influenced by H-2 genes (7, 23, 39).Serological studies in humans indicated that antibodies tonon-TB-23 epitopes of the 19-kDa protein were produced inless severe forms of tuberculosis and that titers of antibody tothe 19-kDa protein were sufficiently discriminating, whensmear-negative pulmonary tuberculosis and healthy contacts

2963

on April 16, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

2964 HARRIS ET AL.

were compared, to permit successful diagnosis in about 40% ofcases (24). However, this is not applicable in countries wheretuberculosis is endemic, where titers of antibody to the 19-kDaantigen were found to be significantly elevated in healthysubjects, possibly reflecting increased exposure to environmen-tal mycobacteria or the higher incidence of self-healing tuber-culosis (6).

Cell-mediated immune responses to the M. tuberculosis19-kDa protein have been studied with mice (2, 21a, 21b),guinea pigs (26), and humans (14, 21a, 25, 32, 36, 37), and anumber of T-cell epitopes have been mapped in detail withoverlapping synthetic peptides. Analysis of peptide immuno-genicity in different inbred strains of mice has revealed at leastsix distinct T-cell epitopes, four of which are recognized in acryptic manner. In contrast, T-cell specificity following process-ing and presentation of the intact recombinant protein isfocused on two adjacent epitope cores spanning residues 54 to61 and 71 to 78 (21a, 21b). Significantly, these immunodomi-nant epitopes are recognized in a genetically permissive man-ner in both mice and humans. In human populations, aconsiderably larger number of potential epitopes have beenidentified with polyclonal T cells, with frequent and strongresponses to epitopes within the amino-terminal peptides (1-20and 11-30) being observed (14, 25, 32, 37).

In view of the prevalence of nontuberculous mycobacterialinfections, the requirement for molecularly defined alterna-tives to purified protein derivative (PPD) for the diagnosis ofmycobacterial infections has become a priority. One of themain aspects in producing improved reagents for diagnosis isthe identification of species-specific B- and T-cell epitopeswhich are immunogenic in a high percentage of the population.Although homologous 19-kDa proteins have been identified invarious mycobacterial species, delayed-type hypersensitivitystudies with guinea pigs indicate the existence of species-specific epitopes (26).

In this study, we have undertaken a comprehensive analysisof linear B-cell epitopes of the M. tuberculosis 19-kDa proteinwith synthetic peptides and also studied the isoforms ofhomologous 19-kDa proteins from different mycobacterialspecies by two-dimensional gel electrophoresis. Finally, theT-cell immune recognition of an immunodominant peptideepitope from the M. tuberculosis 19-kDa protein (residues 61 to80) has been compared with recognition of the homologoussequence from M. avium subsp. intracellulare in mice andhumans.

MATERIALS AND METHODS

Mice. The following recombinant inbred strains of micewere obtained from Olac Harlem (Shaws - Farm, Bicester,Oxfordshire, United Kingdom): C57BL/10 (H-2b), B10.BR(H-2k), B10.D2 (H-2d), BALB/c (H-2d), and BALB.K (H-2k).Mice were age and sex matched and used when 6 to 8 weeksold.Human donors. Blood samples were obtained on a voluntary

basis from donors who were healthy, were PPD positive (n =10) or PPD negative (n = 5) (as determined by in vitroresponse of peripheral blood mononuclear cells to 100 U ofPPD per ml), and had no previous history of clinical tubercu-losis. Blood specimens were also obtained with consent from14 patients with tuberculosis (8 pulmonary, 6 extrapulmonary)admitted to the Lister Unit, Northwick Park Hospital, Harrow,Middlesex, United Kingdom. Blood samples were obtainedfrom patients who were confirmed by smear, culture, orhistology.

Mycobacterial strains and antigens. We have adopted a

nomenclature system (and abbreviations) for species belongingto the MAC as described in a previous publication (12). M.tuberculosis H37Ra, M. bovis BCG, Mycobacterium africanum,Mycobactenum malmoense, M. kansasii, Mycobactenium vaccae,M. smegmatis, Mycobacterium maninum, M. scrofulaceum, My-cobactenum leprae, Mycobactenium duvalii, M. avium subsp.avium RFLP-A6, M. avium subsp. avium serovar 8 strainSM351, M. avium subsp. avium SM483, M. avium subsp.intracellulare serovar 8 strain B12A, M. avium subsp. paratuber-culosis III plus V-R250, and M. avium subsp. intracellulareBattey were produced as soluble extracts following gammairradiation and homogenization with glass beads as describedelsewhere (24). M. tuberculosis (strain H37Rv) culture filtratewas prepared as previously described (1). Heat-killed M.tuberculosis organisms (H37Ra) were obtained from DifcoLaboratories (Detroit, Mich.). Recombinant M. tuberculosis19-kDa fusion protein (rGST-19) was prepared from celllysates of overexpressing E. coli MC1061 (21b). Purified re-combinant M. tuberculosis 19-kDa protein (19-kDa-PC-2) waskindly provided by Bill Russell and Glyn Hewinson, CentralVeterinary Laboratory (Weybridge, Surrey, United Kingdom).PPD was obtained from Evans Medicals (Langhurst, UnitedKingdom).

Synthetic peptides. Peptides were synthesized by a simulta-neous (solid-phase) multiple peptide synthesis procedure asdescribed in detail elsewhere (21b). Peptides were purified bySephadex G-15 gel filtration and preparative high-perfor-mance liquid chromatography. Amino acid sequence compo-sition was verified by Edman degradation (Biopolymer Synthe-sis and Analysis Unit, University of Nottingham, Nottingham,United Kingdom). As previously described (21b), peptidescomprised various residues as follows: pl, 1 to 20; p2, 11 to 30;p3, 21 to 40; p4, 31 to 50; p5, 41 to 60; p6, 51 to 70; p7, 61 to80; p8, 71 to 91; p9, 81 to 100; plO, 91 to 110; pll, 101 to 120;p12, 111 to 130; p13, 121 to 140; p14, 131 to 150; p15, 140 to159. The amino acid sequence of peptide p64-83 from the M.avium subsp. intracellulare 19-kDa protein is VSGSVVCTNAGGTINIAIGG.PAGE and Western blotting (immunoblotting). Antigens

were separated under reducing conditions by sodium dodecylsulfate-polyacrylamide gel electrophoresis (SDS-PAGE) at 17mA per gel for 60 min and then transferred onto nitrocellulosemembranes (Amersham, Little Chalfont, United Kingdom) bybeing blotted overnight at 10 V. Nitrocellulose membraneswere blocked with phosphate-buffered saline (PBS) containing5% skimmed milk powder (PBS-M)-0.05% Tween 20 for 1 hand then incubated for a further 2 h with the appropriateprimary antibody. Membranes were then washed and incu-bated for 1 h with goat anti-mouse immunoglobulin G (IgG)-horseradish peroxidase (HRP) conjugate (Bio-Rad Laborato-ries, Richmond, Calif.), and bound peroxidase activity wasdetected with diaminobenzidine dihydrochloride (0.1 mg/ml;Sigma, Poole, United Kingdom) and 0.01% H202 in 0.5 Mcitrate buffer (pH 5.0). For two-dimensional PAGE, mycobac-terial extracts were treated with sample buffer containing 9.5 Murea and 10% Nonidet P-40 and separated by isoelectricfocusing (IEF) in tube gels as described by O'Farrell (30) with4% Ampholytes (pH 5 to 7) and 1% Ampholytes (pH 3.5 to 10)(Pharmacia LKB, Uppsala, Sweden). SDS-PAGE in the sec-ond dimension and Western blotting were then carried out asdescribed above.

Enzyme-linked immunosorbent assay (ELISA). Syntheticpeptides (5 ,ug per well) or recombinant antigen (0.5 ,ug perwell) diluted in 0.05 M carbonate-bicarbonate buffer (pH 9.6)was dispensed into polyvinyl microtiter plates (Nunc, Roskilde,Denmark) and incubated at 4°C for 48 h. After blocking with

INFECT. IMMUN.

on April 16, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

IMMUNE RESPONSES TO THE MYCOBACTERIAL 19-kDa PROTEIN 2965

PBS-M, primary antibodies were added and plates were incu-bated for a further 2 h. Plates were washed with PBS-T andthen incubated with either goat anti-mouse IgG-HRP, goatanti-mouse IgGl-specific-HRP, or IgG2a-specific-HRP (South-ern Biotechnology Associates Inc., Birmingham, Ala.). Afterwashing, bound peroxidase activity was detected with tetram-ethyl benzidine tetrahydrochloride (0.1 mg/ml; Aldrich, Gill-ingham, United Kingdom) in 0.1 M citrate buffer (pH 5.0)containing 0.01% H202. The reaction was stopped with 0.5 MH2SO4, and the A450 was read in a multiscan spectrophotom-eter (Titertek, MCC/340). In all experiments, the binding oftest sera to wells containing carbonate buffer alone served as anegative control. Antibody titers (ABT30 X 10-3) are ex-pressed as the dilution of test serum giving 30% of the plateaubinding of the MAb TB-23 to M. tuberculosis-soluble extract.

Immunization for antibody responses. Mice (groups ofthree) were injected intraperitoneally with 100,ug of syntheticpeptide, 50 pg of rGST-19, or 50 ,ug of heat-killed M.tuberculosis (H37Ra) organisms emulsified in incompleteFreund's adjuvant. Mice were boosted intraperitoneally atweekly intervals with 50 to 100,ug of antigen suspended inPBS, and pooled serum was collected after the fourth boost.Immunization procedure and T-cell prolifetraion assay.

Mice (groups of three) were immunized subcutaneously inboth hind footpads with 80 ,ug of synthetic peptide or PBSemulsified in incomplete Freund's adjuvant, and 8 to 10 dayslater, draining popliteal lymph nodes were removed andpooled and single cell suspensions were prepared in eitherRPMI 1640 medium supplemented with 10% fetal calf serum(Gibco, Paisley, Scotland), 5 x 10-5 M,B-mercaptoethanol, 2mM L-glutamine, 100 U of penicillin per ml, and 100 ,g ofstreptomycin sulfate per ml or HL-1 serum-free medium(Ventrex Laboratories, Portland, Oreg.) supplemented withL-glutamine, ,B-mercaptoethanol, and penicillin-streptomycinas indicated above. Cells (4 x 105 cells per well) were thenadded in triplicate to 96-well flat-bottom microtiter plates(Nunc) containing antigen at different concentrations rangingfrom 0 to 50 j±g/ml. After 3 days of incubation at 37°C in anatmosphere of 5% CO2, cells were radiolabelled with 37 kBq of[3H]thymidine per well (Amersham International, Amersham,United Kingdom). After a further 6 to 8 h, each well washarvested onto glass-fiber filter paper and radioactive incorpo-ration was determined by liquid scintillation counting. Stan-dard error of the mean for triplicate microcultures was usually<10%.

Nineteen-kilodalton protein-specific T-cell line. The gener-ation and specificity of the CD4+ murine T-cell line specific forrecombinant M. tuberculosis 19-kDa antigen (rGST-19) havebeen described in detail elsewhere (18).

Cytokine assays. For analysis of gamma interferon (IFN--y)and interleukin 5 (IL-5) production, peptide- or PBS-primedlymph node cells were cultured in 24-well plates (4 x 106 cellsper well) containing antigen (0 to 50 ,ug/ml) and supernatantsamples were removed after 24 or 48 h. IFN-y was measuredby sandwich ELISA, by using the anti-murine IFN-y MAbR46A2 to coat ELISA plates and a rabbit polyclonal anti-murine IFN-y to detect (28). The assay was developed with agoat anti-rabbit peroxidase conjugate (Sigma) followed by aTMB (tetramethyl benzidine tetrahydrochloride) substrate.Amounts of IFN--y in test supernatants were quantified byreference to a recombinant IFN--y standard (Genzyme) andare expressed in units per milliliter. IL-5 was measured bysandwich ELISA with a pair of affinity-purified rat anti-murineIL-5 MAbs (TRFK4 and TRKF5) (33). Amounts of IL-5 in testsupernatants were quantified by reference to a standard curve

derived from recombinant murine IL-5 and are expressed inunits per milliliter.Human T-lymphocyte proliferation assay. Blood samples

were obtained by venipuncture, collected in 10-ml Vacutainerscontaining 1.0 ml of CDP Adenine-1 (Baxter, Thelford, Nor-folk, United Kingdom), and then mixed with an equal volumeof PBS. Peripheral blood mononuclear cells were isolated andcultured (2 x 105 cells per well) for 6 days in quadruplicate in96-well round-bottom microtiter plates containing antigen ormitogen (0 to 50,g/ml) as previously described (21a). Stan-dard error of the mean of quadruplicate microcultures wasusually <10%.

RESULTS

Antibody responses following immunization with syntheticpeptides. Groups of C57BL/10 mice (n = 3) were hyperimmu-nized with each of the 15 peptides (pl to p15) spanning thecomplete amino acid sequence of the 19-kDa protein of M.tuberculosis, and pooled sera were assayed by ELISA forbinding to the homologous immunogen. A total of 3 of the 15peptides (p2, p7, and p15) induced strong anti-peptide anti-body responses (Fig. 1). The highest antibody titer was inducedby p15 (ABT30 x 10-3 = 62), followed by p7 (27.8) and p2(6.8). Titers of antibody to all other peptides were <0.1, andnone of the 15 peptides was recognized by normal mouseserum. To define the B-cell epitopes within p2, p7, and p15more precisely, antisera were tested for their ability to recog-nize adjacent peptides. Anti-p2 serum failed to recognizeeither pl (residues 1 to 20) or p3 (residues 21 to 40);consequently, the core B-cell epitope within p2 could belocated between residues 15 and 25. However, anti-p7 andanti-p15 serum did react weakly (ABT30, <103) with theadjacent peptides p8 (residues 71 to 91) and p14 (residues 131to 150), indicating that the core B-cell epitopes within p7 andp15 were probably localized between residues 70 and 80 and140 and 150, respectively.Immunodominance and mycobacterial specificity of the p15

epitope. The MAb F29-47 recognized a single band (molecularmass, -45 kDa) in the recombinant GST-19 fusion protein(Fig. 2A, lane 2) and a doublet (molecular mass, -19 to 21kDa) in the native protein (Fig. 2A, lane 1). An identicalpattern of antibody recognition was also seen with the anti-p1Sserum (Fig. 2B, lanes 1 and 2). With the exception of anti-p2serum, which was weakly positive, all other antipeptide sera,including anti-p7, were consistently negative for binding to thenative or recombinant 19-kDa protein by Western blotting(Fig. 2B, lane 3). High-titer antibody binding to purifiedrecombinant 19-kDa protein (19-kDa-PC-2) by ELISA wasalso demonstrated only with anti-p15 serum (Fig. 2C). Giventhe high degree of sequence homology among the mycobacte-rial 19-kDa proteins (see Fig. 5), it was of interest to determinewhether the epitope recognized by anti-p1S serum was M.tuberculosis specific. Western blotting analysis of soluble ex-tracts prepared from 14 different mycobacterial species indi-cated that only M. tuberculosis, M. bovis, and M. africanumwere recognized, while all non M. tuberculosis-complex speciestested were negative (see Fig. 2B, lane 4, for a representativeexample). Significantly, despite sharing extensive sequencehomology at the p15 epitope with M. avium subsp. avium andM. avium subsp. intracellulare (15 identical residues out of 20),cross-reactivity was observed neither with the MAC nor withM. kansasii, M. marinum, or M. duvalii, all of which arerecognized by the anti-19-kDa protein MAb TB23 (22).Antibody responses following immunization with rGST-19

and H37Ra. Antisera obtained from rGST-19-immunized

VOL. 62, 1994

on April 16, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

2966 HARRIS ET AL.

105-

0 4

E-~~~~~~

r-' 104-:

_S

- ~~3" 1 --b

0~op-*-O

110

[2 Anti-peptWe serum

[1 Normal mouse serum

/ ,.. i... R;,......

Ri ..* .;.. ::::i:'.:::::i,i, :::i.:-.-i,:::.- >:-i: :i:-:.:: :::i:i.:i::Ri,: *i,: i.: :-.

.s a- _P3 ( ),i., ...............i:: :i.::i.::::::::

R-. ..........R>. i.ji.4-.-....RR Rj,, .,,, ,i,

i. :R: ::::-:.:R ,.R j.-i;..RRw j--RR

:::i. -:i.::: '::::'ri. .: i,: .. :j:

R- sRRi.*RR:R: *::::: :j*: ::-:::::-:

*R 'R R '-R* :-RR " 0':* X,:: -: ,::: * ,¢:U:

..,i,. .*ii. *.-.-4. >,:,.-.. '..::tePS

p6 (-)No p8 (+) pl4 (+)

pl p2 p3 p4 p5 p6 p7 p8 p9 plO pll p12 p13 p14 p15

Immunogen and test peptideFIG. 1. Antibody responses following immunization with synthetic peptides. ELISA of C57BL/10 antisera for IgG antibodies specific for the

homologous immunizing peptide (5 jtg per well). Positive antisera were also assayed for recognition of adjacent peptides, and results are shownas positive (+) or negative (-).

C57BL/10, B1O.BR, and BlO.D2 strains of mice showed signif-icant binding to peptides p5 (residues 41 to 60) and p15(residues 140 to 159). Titers of antibody to p5 and p15 were notsignificantly different in the three strains of mice, and resultsfor C57BL/10 mice only are shown in Table 1. Neither peptide

p2 nor p7, both of which induced antibody responses (Fig. 1),was recognized by anti-rGST-19 sera. Moreover, antisera fromC57BL/10 mice immunized with killed M. tuberculosis H37Rareacted only against the p15 epitope (Table 1). Thus, immu-nization with synthetic peptide, recombinant 19-kDa antigen,

0.8 -

- 0.6-

0.4 -

0.2 -

o.o -o'

I 'o ".2 II n II -o1 I .6'''9lo -l lo -2 10 -3 10 -4 to0-5 lo 6

Dilution of antisera

FIG. 2. Mycobacterial specificity of anti-peptide antibodies. (A and B) Staining of M. tuberculosis culture filtrate (lanes 1) and recombinant M.tuberculosis 19-kDa fusion protein (rGST-19) (lanes 2) with F29-47 MAb (A) and anti-p15 serum (B). M. tuberculosis culture filtrate was developedwith a negative antipeptide serum (p7) (lane 3). A representative example of a non-M. tuberculosis complex species (M. avium subsp. intracellulare)stained with anti-p15 serum (1/200) is shown in lane 4. (C) ELISA of antipeptide sera with purified 19-kDa protein (19-kDa-PC-2) (0.5 ,ug perwell). MW, molecular mass.

MW AkDa

80 -

49.5 -

323 -

273 -

185 -

1 2

C * anti-plSanti-p2

'\ fa A anti-p7

MW BkDa

80-

49.5 - -_

323 -

2735 -

183--

1 2 3 4

L-r--LL

INFECT. IMMUN.

/-V-V-V -V -V -V -V -V

II

/

on April 16, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

IMMUNE RESPONSES TO THE MYCOBACTERIAL 19-kDa PROTEIN 2967

TABLE 1. Epitope and isotype specificity of antibodies elicited by synthetic peptides and intact 19-kDa protein

Test Antibody titer" Immunoglobulin isotypec IFN--(d IL-5Immunogen' peptide ([ABT301 X 10-3) IgGl/IgG IgG2a/IgG (U/ml) (U/ml)

p2 p2 6.7 0.87 0.01 230 3p7 p7 27.8 0.08 0.96 360 22p15 p15 62.2 1.00 0.10 1,064 3rGST-19 p5 10.4 1.00 1.00 NDe ND

p15 17.3 1.00 0.77 ND NDKilled H37Ra p15 15.2 0.47 0.81 ND ND

a C57BL/10 mice (n = 3) were hyperimmunized with peptides (p2, p7, or p15), rGST-19, or killed H37Ra in incomplete Freund's adjuvant.b Serum dilutions giving 30% of the plateau binding.c Ratios of optical density values obtained with antisera at a dilution of 1/100.d Cytokine-specific ELISA of supematants from primed lymph node cells cultured in vitro with homologous peptide for 24 h.e ND, not determined.

or tubercle bacilli revealed the immunogenicity of the linearepitope contained within p15.Antibody isotype and T-cell cytokine production. Antibodies

produced in response to immunization wfth peptide p2, p7, orp15 showed striking differences in isotype: results from differ-ent experiments consistently showed that while anti-p7 serumwas predominantly IgG2a, anti-p2 and anti-p1S sera werealmost exclusively the IgGl isotype (Table 1). In contrast,anti-p1S and anti-p5 antibodies raised by immunization withrGST-19 or killed M. tuberculosis contained similar propor-tions of both IgGl and IgG2a isotypes, suggesting that T-cellhelp in support of a single antibody specificity, when presentedin the context of the whole molecule, may have been stimu-lated by more than one T-cell epitope. Differences in antibodyisotype could not, however, be directly attributed to theselective secretion of either IFN--y or IL-5 since all threepeptides elicited strong in vitro proliferative and IFN--y re-sponses while levels of IL-5 were either only modest (p7) ornot significantly elevated (p2 and p15) (Table 1). AlthoughIFN--y responses were generally reduced, no concomitantincrease in IL-5 secretion was detected in any peptide-primedT-cell cultures following multiple peptide immunizations (re-sults not shown).

Species specificity of anti-rGST-19 serum. To evaluateexpression of the 19-kDa protein in different mycobacterialspecies, the specificity of anti-rGST-19 serum was analyzed byWestern blotting (Fig. 3). The native 19-kDa protein present inM. tuberculosis extracts was identified in the form of a doublet,with molecular masses of - 19 and 21 kDa. Antibody binding toM. avium subsp. intracellulare, M. avium subsp. paratuberculo-sis, and M. avium subsp. avium extracts was also observed,indicating the likely presence of homologous 19-kDa proteins.The protein identified in M. avium subsp. intracellulare extractsby anti-rGST-19 serum was also present as a doublet, whereasthe proteins recognized in M. avium subsp. paratuberculosisand M. avium subsp. avium extracts appeared to migrate assingle bands in SDS gels, although differences in molecularmass were observed among the three strains ofM. avium subsp.avium tested, ranging from -21 to 25 kDa. The anti-rGST-19serum failed to react specifically with the 19-kDa protein whenextracts prepared from a number of other mycobacterialspecies including M. leprae, M. scrofulaceum, M. smegmatis,and M. vaccae were assayed by Western blotting (results notshown). The simplest interpretation of this result is that theantiserum used (anti-rGST-19) does not cross-react with anyputative 19-kDa protein homolog present in these species (hla,29); however, the lack of gene expression cannot be excluded.

Isoforms of the 19-kDa protein detected by two-dimensionalgel electrophoresis. Different mycobacterial extracts were sep-

arated by two-dimensional gel electrophoresis, and Westernblots were subsequently stained with anti-rGST-19 serum. Toaid comparison between MAC subspecies, extracts were runboth individually (Fig. 4A) and as mixtures with M. tuberculosisculture filtrate (Fig. 4B). The 19-kDa protein in M. tuberculosisculture filtrate resolved into at least four distinct protein spots(indicated with arrows in top panel), each of which migratedinto the acidic region of the IEF gel. The same isoforms werealso recognized by the anti-p1S serum (results not shown). Asall isoforms of theM tuberculosis 19-kDa protein migrated toapproximately the same molecular weight position, it was notpossible to determine conclusively whether particular isoformswere derived from the upper or from the lower band of thedoublet (seen in Fig. 3). An identical staining pattern wasobserved with the 19-kDa protein present in the M tuberculo-sis-soluble extract (results not shown). Analysis of M bovis-BCG extracts on two-dimensional gels revealed the presenceof two isoforms of the 19-kDa protein (Fig. 4A) which wereshown by mixing experiments to be completely overlapping

FIG. 3. Mycobacterial specificity of anti-rGST-19 serum. Westernblotting of different mycobacterial extracts with anti-rGST-19 serum.MW, molecular mass; Maa, M. avium subsp. avium; Mai, M aviumsubsp. intracellulare; Map, M avium subsp. paratuberculosis; Mtb, M.tuberculosis.

MWkDa

49.5 -

32.5 -27.5-- *_ _

18.5-

CIOuCu, Cu

cn En c< O mccm

VOL. 62, 1994

on April 16, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

2968 HARRIS ET AL.

Basic Acidic

Mtb MwkDa1

A

I I09

. .

I...'1

W.. -:

a =a 0 X

I~~~4 0

B

r~ A

1~~~~.SS i X

K ----I~~06-A"I

, _ l

*4t-' I9IMM

FIG. 4. Two-dimensional Western blot analysis of mycobacterial 19-kDa proteins stained with mouse anti-rGST-19 serum. The top panel (M.tuberculosis [Mtb]) indicates the positions of the 27.5- and 18.5-kDa molecular mass (Mw) markers and the acidic and basic poles of the IEF gel.(A) Individual mycobacterial extracts. (B) Mycobacterial extract mixed with an approximately equal concentration of M. tuberculosis culturefiltrate. Mb, M. bovis; Maa, M. avium subsp. avium; Mai, M. avium subsp. intracellulare; Map, M. avium subsp. paratuberculosis.

with the corresponding spots of M. tuberculosis (Fig. 4B). Thethree strains of M. avium subsp. avium tested (A6, SM483, andSM351) produced similar staining patterns. This was con-firmed in mixing experiments which showed that spots from M.avium subsp. avium A6 were overlapping with spots derivedfrom M. avium subsp. avium SM351 (results not shown).Mixing experiments with M. tuberculosis (panel B) clearlyrevealed that isoforms derived from all strains ofM. avium subsp.avium possessed higher molecular weights (as seen on one-dimensional gels in Fig. 3) and had different isoelectric points.Isoforms ofM avium subsp. paratuberculosis appeared similar to

those seen in M. avium subsp. avium. In contrast, isoforms of thetwoM avium subsp. intracellulare strains tested (B2A and B) weredifferent from M avium subsp. avium. Unlike M avium subsp.avium, and M tuberculosis both M avium subsp. intracellularestrains contained at least one isoform which migrated to the basicend of the IEF gel, and mixing experiments confirmed that theywere distinct from those ofM tuberculosis.

T-cell cross-reactivity of peptide p61-80 (p7) homologs. Wehave previously identified an immunodominant T-cell epitopelocated within peptide p7 (residues 61 to 80) of the Mtuberculosis 19-kDa protein which is permissively recognized in

1 20 40 60 80

iTVAVAGAAII GLSGS NKST ETTT DGKDQN SVVC - IAIGG Mtb

VKRQLTVAVAGAAILAAGMSGCS1NK SIS 3TSAS GAAGTK IDGKDQNVSGSVVCTNAGG IAIGG Mai

GCSSG GGAAGTK IDGKDQNVSGSVVCTNAGGTVNIAIGG Maa

81 100 120 140 159

AATGIAAV DGNPP dKSVGLGNVNGVTLGYTSGTGQGNASA JDG H KITGTATGVDMANP

AATGIAAVLSDGNPPQVKSVGLGNVNGVTLGYTSGTGQGNAmASKDGNSYKIIGTATGVDMANPMQPVNKPFEINVTCNSAATGIAAVLSDGNPPQVKSVGLGNVNGVTLGYTSGTGQGNAS NSYKITGTATGVD PVNKPFEI0VTCNSFIG. 5. Amino acid sequences of the M. tuberculosis, M. avium subsp. intracellulare, and M. avium subsp. avium 19-kDa proteins. Amino acid

sequences shown in boldface correspond to the immunodominant p7 T-cell epitope (M. tuberculosis [Mtbl residues 61 to 80) and theimmunodominant p15 B-cell epitope (M. tuberculosis residues 140 to 159). Boxed residues indicate amino acid differences among M. tuberculosis,M. avium subsp. intracellulare (Mai), andM avium subsp. avium (Maa) sequences. Amino acid sequences were obtained from references 3, 5, and29.

Mb-BCG

Maa-A6

Maa-SM483

Maa-SM351

Mai-B2A

Mai-B

Map-250

INFEcr. IMMUN.

X7 C

o 11 0t t t

on April 16, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

IMMUNE RESPONSES TO THE MYCOBACTERIAL 19-kDa PROTEIN 2969

TABLE 2. Immunogenicity and T-cell cross-reactivity of peptide homologs from the mycobacterial 19-kDa protein

M. tuberculosis p6l80c M. avium subsp. intracellularePeptide immunogena Mouse strain p64p83c

cpm SI cpm SI

M. tuberculosis p61-80 C57BL/10 (b) 24,236 6.8 (2.1) 12,632 3.5 (3.7)BALB/c (d) 95,054 13.7 (1.2) 63,390 9.2 (2.7)BALB.K (k) 72,275 8.9 (1.2) 26,698 3.2 (1.6)

M. avium subsp. intracellulare p64-83 C57BL/10 10,299 2.5 (2.1) 72,679 17.9 (3.7)BALB/c 78,680 9.4 (1.8) 105,994 12.6 (3.3)BALB.K 50,484 8.6 (1.2) 97,653 16.6 (1.6)

aMice were immunized in the footpad with 80 p,g of peptide.b Peptide added to culture of 7-day immune lymph node cells.c Mean counts per minute of triplicate wells and SIs for peptide-immune and PBS-immune (in parentheses) lymph node cells (mean counts per minute +

peptide/background counts per minute without peptide). Background counts per minute (cells without antigen) for mice primed with M. tuberculosis p61-80 and M.avium subsp. intracellulare p64-83 were as follows: C57BIJ10, 3581 and 4046; BALB/c, 6870 and 8411; and BALB/K, 8157 and 5872, respectively. Standard error of themean of triplicates was <10%.

both mice and humans (21a, 21b). Comparison of the respec-tive amino acid sequences (Fig. 5) reveals that the homologousM. avium subsp. intracellulare peptide (p64-83) contains con-servative substitutions at residues 62 (T--S), 71 (A-->G), and74 (V->I) and two nonconservative changes at positions 69(T->N) and 73 (N-4T). Peptides p61-80 (M. tuberculosis) andp64-83 (M. avium subsp. intracellulare) were both found to beimmunogenic, eliciting vigorous proliferative responses to thehomologous peptide in three different strains of mice (b, d, andk) (Table 2). In contrast, when p61-80- or p64-83-immune Tcells were cultured in vitro with the heterologous peptide, thepattern of immune recognition was strongly influenced by thegenetic haplotype. Thus, no cross-reactivity was observed witheither peptide in the H-2b strain, whereas in H-2d mice, bothpeptides were reciprocally cross-reactive. Finally, in the H-2kstrain, p64-83-immune T cells responded when challenged invitro with p61-80 but not vice versa. Cross-reactivity of themycobacterial 19-kDa protein was also analyzed with a T-cellline (H-2b) specific for the recombinant (rGST-19) M. tuber-

200AA p61-80 (Mtb)|

10p64-83(Mai)i

3-0

is

00

so

culosis 19-kDa protein (Fig. 6). Although the line responded toboth peptides (Fig. 6A), p61-80 was recognized i103 timesmore efficiently than p64-83, confirming that T cells specific forp61-80 do not cross-react strongly with p64-83 in the context ofthe H-2b haplotype. Significantly, this lack of cross-recognitionin the H-2b haplotype was also reflected in the response to theintact 19-kDa protein; hence, strong responses to M. tubercu-losis and M. bovis extracts were observed while responses to M.avium subsp. intracellulare or M. avium subsp. avium extractswere not more than background (Fig. 6B).

Recognition of peptide p61-80 (p7) homologs by human Tcells. Human T-cell recognition of the homologous peptidesp61-80 and p64-83 was analyzed in three groups of donors,represented by PPD-positive healthy individuals (n = 10),PPD-positive tuberculosis patients (n = 14), and PPD-negativehealthy individuals (n = 5) (Fig. 7). Responses to PPD orpeptides, but not concanavalin A (results not shown) were notsignificantly elevated in the PPD-negative group. In contrast,PPD-positive healthy individuals responded more strongly

.001 .01 .1 1 10 100 /0.1 .1 1 10 100

Antigen (gg/ml)FIG. 6. T-cell recognition of the mycobacterial 19-kDa protein. A rGST-19-specific T-cell line was assayed for recognition of peptides p61-80

and p64-83 (A) and various mycobacterial extracts (B). T cells (104 per well) were cultured for 72 h with antigen (0 to 100 ,ug/ml) in the presenceof syngeneic (H-2b) irradiated antigen-presenting cells (3 x 105 per well), and radioactive thymidine incorporation was determined during the final6 h of culture. Mtb, M. tuberculosis; Mai, M. avium subsp. intracellulare; Mb, M bovis; Maa, M. avium subsp. avium.

VOL. 62, 1994

on April 16, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

2970 HARRIS ET AL.

* p61-80 (Mtb)% p64-83 (Mai)0 PPD

100-

- 10- * -w00

.:3 :

pO.'

p0.031

% Responders: 60 90 100

PPD-positive

86 93 100 0 0 0

Tuberculosis PPD-negativHealthy [N=10] Patients [N=14] Healthy [N=5]

FIG. 7. The recognition of peptide homologs by human T cells. Peripheral blood mononuclear cells (2 x 105 per well) were cultured in vitroin quadruplicate with optimal concentrations (0 to 100 ,ug/ml) of peptide p61-80, p64-83, or PPD (10 to 100 pAl/ml). After 6 days, incorporationof radioactive thymidine was determined. Individual values are expressed as the maximum SI (counts per minute with antigen/counts per minutewithout antigen), and an SI value of 3 was used as a cutoff for positive responses (horizontal dotted line). p, unpaired Student's t test; Mtb, M.tuberculosis; Mai, M. avium subsp. intracellulare.

(mean stimulation index [SI] = 10.2) and more frequently(90% responders) to p64-83 than to p61-80 (P = 0.009; meanSI = 4.0; 60% responders). The response of patients withtuberculosis to p64-83 was very similar to that seen in PPD-positive healthy individuals (mean SI = 15.9 and 93% re-

sponder status; P = 0.382). Responses to p61-80 were, how-ever, significantly stronger (mean SI = 7.8; P = 0.031) andmore frequent (86% responders) in patients with tuberculosis.

DISCUSSION

One of the main aspects examined in this paper concerns thelocalization of linear epitopes recognized by B cells. Antiseradirected against synthetic peptides identified the most immu-nogenic linear epitope of the M. tuberculosis 19-kDa protein atthe carboxy-terminal end of the molecule (p15 residues 140 to159). This epitope is not, however, recognized by the previ-ously reported MAbs TB-23 and F29-47 (20). Furthermore,antisera from C57BL/10 (H-2b) and BALB/c (H-2d) miceinfected with virulent M. tuberculosis (strain H37Rv), as well asa limited number of human antisera which reacted stronglyagainst the whole 19-kDa antigen, failed to show significantbinding to the p15 epitope (results not shown), suggesting that

the specificity of antibodies produced following tuberculousinfection is strongly biased toward conformational epitopes.The antibody binding site, shared between peptides p14

(residues 131 to 150) and p15 (residues 140 to 159), probablyresides within the sequence 140 to 150, but immunogenicitywas restricted to p15, indicating that helper T cells were

adequately stimulated by the carboxy-terminal residues of p15,but not by those contained in p14. This finding is surprising inview of our previous demonstration that in C57BL/10 mice(H-2b), both p14 and p15 stimulated primed lymph node T-cellproliferation (21b). These results corroborate with previousfindings obtained with hybrid synthetic peptides demonstratingthat T-cell help for antibody production was effective onlywhen the adjacent B epitope was linked in the N-terminalposition relative to the T epitope (10).The strict species specificity of the p15 epitope was unex-

pected given that the carboxy-terminal region of the 19-kDaprotein is highly conserved between M. tuberculosis and MACand suggests a critical role for the nonconserved serine resi-dues at positions 146 and 151 (Fig. 5) in determining epitopespecificity. Alternatively, posttranslational modifications suchas glycosylation might also influence epitope cross-reactivity.

p

in

*I14

IPo

0

14c

INFECT. IMMUN.

CDF -re on A

pril 16, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

IMMUNE RESPONSES TO THE MYCOBACTERIAL 19-kDa PROTEIN 2971

An immunodominant epitope was also identified withinpeptide p5 (residues 41 to 60) with antisera raised against therGST-19 fusion protein. It is possible that T-cell help for theantibody response to this epitope has originated from theglutathione-S-transferase fusion partner, since anti-pS antibod-ies were absent from antisera to killed M. tuberculosis (presentresults and reference 2). Furthermore, the lack of a suitableT-helper stimulatory epitope within p5 was suggested by thefailure to raise antibodies by immunization with this peptide.

Antisera produced by immunization of mice with syntheticpeptides also identified epitopes within p2 (residues 11 to 20)and p7 (residues 61 to 80), in addition to p15. Anti-p2 reactedonly weakly with the recombinant 19-kDa antigen, and p7antibodies reacted not at all, indicating that these epitopesprobably adopt a different conformation within the context ofthe whole molecule. The cryptic nature of these epitopes wasalso confirmed with antisera raised against either rGST-19 orkilled M. tuberculosis. In contrast, the latter antisera have beenreported to bind to amino acid sequences 1 to 20 (pl) and 61to 80 (p7) in another study (2).We explored whether the immunogenic peptides (p2, p7,

and p15), each representing a distinct B- and T-cell epitopepair, induced the same or different T-cell cytokines andwhether the cytokine profile correlated with antibody isotypes.Indeed, a striking segregation of antibody isotypes was consis-tently observed, with p2 and p15 eliciting an IgGl responsewhereas antibodies specific for p7 were predominantly IgG2a.However, analysis of the cytokines secreted by peptide-im-mune T cells following both short- and long-term immuniza-tion protocols did not reveal a clear correlation betweenantibody isotype and cytokine induction as has previously beensuggested (11, 16, 27). Although our data do not support thenotion of a stringent relationship between antibody isotypesand cytokine profiles, the conclusions are limited on groundsthat the low levels of IL-5 measured in vitro may not reflectpossible differences between cell populations in vivo. For thisreason, more sensitive methods such as mRNA analysis may berequired. However, our results are consistent with other stud-ies which have shown that multiple factors including the type ofadjuvant used (17, 35), the major histocompatibility complex(MHC) class II genotype (31), the dose of antigen used (31),the method of immunization (8), and the nature of the immunogen(34) can influence the phenotype of responding T cells.

Analysis by two-dimensional Western blotting revealed anew charge-based heterogeneity among antigenic isoforms ofthe 19-kDa protein. Isoforms of the 19-kDa protein detected inM. tuberculosis were clearly different in both size and chargefrom those found in the MAC. While isoforms were verysimilar within the M. tuberculosis complex, considerable heter-ogeneity was apparent between M. avium subsp. avium and M.avium subsp. intracellulare isolates. It is unlikely that thisheterogeneity is due entirely to charged residue differences inthe primary amino acid sequences. Rather, it may reflectposttranslational modifications of protein structure such as thepresence or absence of the hydrophobic signal peptide and thecovalent attachment of fatty acids or carbohydrate moieties(15, 18, 29, 40), capable of affecting the overall charge of themolecule. Whilst we cannot exclude that the mycobacterialculture conditions may influence the nature of the isoformsproduced, it will be important to evaluate whether this tech-nique could contribute to the taxonomic classification of theMAC and to a better understanding of the structural basis ofthe isoform differences that we have observed in the mycobac-terial 19-kDa protein.

In the light of our previous work (21a, 21b) and recentdelayed-type hypersensitivity studies with guinea pigs (26), we

analyzed the immunogenicity and T-cell cross-reactivity ofpeptide p61-80, an immunodominant and genetically permis-sive T-cell epitope from the M. tuberculosis 19-kDa protein,and the corresponding peptide p64-83, from M. avium subsp.intracellulare. While both peptides were immunogenic in thethree haplotypes tested (d, b, and k), the pattern of cross-reactivity was highly dependent on the nature of the MHCclass II haplotype, and only in BALB/c (H-2d) mice were bothpeptides fully cross-reactive. A possible explanation may bethat these peptides adopt different conformations when boundto the MHC molecule, thereby interfering with the T-cellreceptor-MHC interaction. Since the epitope core withinp61-80 is located between residues 71 and 78 (21a), it is likelythat the three amino acid substitutions within this region, atpositions 71, 73, and 74, strongly influence epitope-specificcross-reactivity. As all three strains of mice are highly sensitiveto substitution with alanine at position Asn-73 (21), theselective cross-reactivity with p64-83 in H-2d but not in H-2b orH-2k mice may be influenced by more than one residue. It is ofinterest that the M. avium subsp. intracellulare peptide (resi-dues 65 to 78) was found to be cross-reactive, eliciting delayed-type hypersensitivity responses in outbred guinea pigs sensi-tized with either M. avium subsp. intracellulare or M.tuberculosis (26).

Blood lymphocytes from human donors possessing a varietyof different HLA genotypes responded to both p61-80 andp64-83. Although responses to the M. avium subsp. intracellu-lare peptide were generally stronger and more frequent in bothPPD-positive healthy individuals and tuberculosis patients,both peptides elicited positive responses in more than 50% ofthe individuals tested. As neither peptide was recognized byPPD-negative individuals, reactivity to the M. avium subsp.intracellulare peptide is probably due to cross-reactive primingby BCG vaccination, either on its own or in combination withnatural immunity to environmental mycobacteria.

In conclusion, while neither of the T-cell epitopes analyzedin this report appears to be sufficiently species specific fordiagnostic purposes, the strict species specificity of the immu-nodominant B-cell epitope within peptide p15 for the M.tuberculosis complex may be of further interest for the sero-logical diagnosis of tuberculosis.

ACKNOWLEDGMENTS

We thank Bill Russell and Glyn Hewinson (Weybridge) for thepreparation of purified recombinant 19-kDa protein and Arend Kolkfor the F29-47 MAb. We are also grateful to A. Elsaghier for providingvarious strains of mycobacteria and E. Roman, A. Hills, and M. Hill fortechnical assistance.

REFERENCES1. Andersen, A. B., P. Andersen, and L. Ljungqvist. 1992. Structure

and function of a 40,000-molecular-weight protein antigen ofMycobacterium tuberculosis. Infect. Immun. 60:2317-2323.

2. Ashbridge, K. R., B. T. Backstrom, H. X. Liu, T. Vikerfors, D.Englebretsen, D. R. K. Harding, and J. D. Watson. 1992. Mappingof T helper cell epitopes by using peptides spanning the 19kDaprotein of Mycobacterium tuberculosis. Evidence for unique andshared epitopes in the stimulation of antibody and delayed-typehypersensitivity responses. J. Immunol. 148:2248-2255.

3. Ashbridge, K. R., R. J. Booth, J. D. Watson, and R. B. Lathigra.1989. Nucleotide sequence of the 19kDa antigen from M. tuber-culosis. Nucleic Acids Res. 17:1249.

4. Ashbridge, K. R., R. L. Prestidge, R. J. Booth, and J. D. Watson.1990. The mapping of an antibody-binding region on the Myco-bacterium tuberculosis 19kDa antigen. J. Immunol. 144:3137-3142.

5. Booth, R. J., D. L. Williams, K. D. Moudgil, L. C. Noonan, P. M.Grandison, J. J. McKee, R. L. Prestidge, and J. D. Watson. 1993.Homologs of Mycobacterium leprae 18-kilodalton and 19-kilodal-

VOL. 62, 1994

on April 16, 2020 by guest

http://iai.asm.org/

Dow

nloaded from

2972 HARRIS ET AL.

ton antigens in other mycobacteria. Infect. Immun. 61:1509-1515.6. Bothamley, G., H. Batra, V. Ramesh, A. Chandramuki, and J.

Ivanyi. 1992. Serodiagnostic value of the 19 kilodalton antigen ofMycobacterium tuberculosis in Indian patients. Eur. J. Clin. Micro-biol. Infect. Dis. 11:912-915.

7. Brett, S. J., and J. Ivanyi. 1990. Genetic influences on the immunerepertoire following tuberculous infection in mice. Immunology71:113-119.

8. Caulada-Benedetti, Z. F., F. Al-Zamel, A. Sher, and S. James. 1991.Comparison of Thl and Th2-associated immune reactivities stimu-lated by single versus multiple vaccination of mice with irradiatedSchistosoma mansoni cecariae. J. Immunol. 146:1655-1660.

9. Collins, M., A. Paiki, S. Wall, A. Nolan, J. Goodger, M. Woodward,and J. Dale. 1990. Cloning and characterization of the gene for the19kDa antigen of Mycobacterium bovis. J. Gen. Microbiol. 136:1429-1436.

10. Cox, J. H., J. Ivanyi, D. B. Young, J. R. Lamb, A. D. Syred, andM. J. Francis. 1988. Orientation of epitopes influences the immuno-genicity of synthetic peptide dimers. Eur. J. Immunol. 18:2015-2019.

11. DeKruyff, R. H., T. R. Mosmann, and D. T. B. Umetsu. 1990.Induction of antibody synthesis by CD4+ T cells: IL-5 is essentialfor induction of antigen-specific antibody responses by Th-2 butnot Th-1 clones. Eur. J. Immunol. 20:2219-2227.

11a.De Smet, K. 1993. Personal communication.12. Elsaghier, E., A. Nolan, B. Allen, and J. Ivanyi. 1992. Distinctive

Western blot antibody patterns induced by infection of mice withindividual strains of the Mycobactenium avium complex. Immunol-ogy 76:355-361.

13. Engers, H. D., V. Houba, J. Bennedsen, T. M. Buchanan, S. D.Chaparas, G. Kadival, 0. Closs, J. R. David, J. D. A. van Embden,T. Godal, S. A. Mustafa, J. Ivanyi, D. B. Young, S. H. E. Kaufmnann,A. G. Khomenko, A. H. J. Kolk, M. Kubin, J. A. Louis, P. Minden,T. M. Shinnick, L. Trnka, and R. A. Young. 1986. Results of aWorld Health Organization-sponsored workshop to characterizeantigens recognized by mycobacterium-specific monoclonal anti-bodies. Infect. Immun. 51:718-720. (Letter to the editor.)

14. Faith, A., C. Moreno, R. Lathigra, E. Roman, M. Fernandez, S. J.Brett, D. J. Mitchell, J. Ivanyi, and A. D. M. Rees. 1991. Analysisof human T cell epitopes in the 19kDa antigen of Mycobacteriumtuberculosis: influence of HLA-DR. Immunology 74:1-7.

15. Fifis, T., C. Costopoulos, A. J. Radford, A. Bacic, and P. R. Wood.1991. Purification and characterization of major antigens from aMycobacterium bovis culture filtrate. Infect. Immun. 59:800-807.

16. Finkelman, F. D., I. M. Katona, T. R. Mosmann, and R. L.Coffman. 1988. IFN-y regulates the isotypes of Ig secreted duringin vivo humoral immune responses. J. Immunol. 140:1022-1027.

17. Fox, B. S. 1992. Antibody responses to a cytochrome c peptide donot correlate with lymphokine production patterns from helperT-cell subsets. Immunology 75:164-169.

18. Garbe, T., D. Harris, M. Vordermeier, R. Lathigra, J. Ivanyi, andD. Young. 1993. Expression of the Mycobacterium tuberculosis19-kilodalton antigen in Mycobacterium smegmatis: immunologicalanalysis and evidence of glycosylation. Infect. Immun. 61:260-267.

19. Garsia, R. J., L. Hellqvist, R. J. Booth, A. J. Radford, W. J. Britton,L. Astbury, R. J. Trent, and A. Basten. 1989. Homology of the70-kilodalton antigens from Mycobacterium leprae and Mycobacte-rium bovis with the Mycobacterium tuberculosis 71-kilodalton anti-gen and with the conserved heat shock protein 70 of eucaryotes.Infect. Immun. 57:204-212.

20. Harris, D. P. Unpublished data.21. Harris, D. P. Unpublished data.21a.Harris, D. P., H. M. Vordermeier, G. Friscia, H.-M. Surcel, G.

Pasvol, C. Moreno, and J. Ivanyi. 1993. Genetically permissiverecognition of adjacent epitopes from the 19kDa antigen ofMycobacterium tuberculosis by human and murine T cells. J.Immunol. 150:5041-5050.

21b.Harris, D. P., H. M. Vordermeier, E. Roman, R. Lathigra, S. J.Brett, C. Moreno, and J. Ivanyi. 1991. Murine T cell stimulatorypeptides from the 19kDa antigen of Mycobacterium tuberculosis:epitope-restricted homology with the 28kDa protein of Mycobac-terium leprae. J. Immunol. 147:2706-2712.

22. Ivanyi, J., A. Morris, and M. Keen. 1985. Studies with monoclonal

antibodies to mycobacteria, p. 59-90. In A. J. L. Macario (ed.),Monoclonal antibodies against bacteria, vol. 1. Academic Press,Inc., New York.

23. Ivanyi, J., and K. Sharp. 1986. Control by H-2 genes of murineantibody responses to protein antigens of M. tuberculosis. Immu-nology 59:329-332.

24. Jackett, P. S., G. Bothamley, H. V. Batra, A. Mistry, D. B. Young,and J. Ivanyi. 1988. Specificity of antibodies to immunodominantmycobacterial antigens in pulmonary tuberculosis. J. Clin. Micro-biol. 26:2313-2318.

25. Lamb, J. R., A. D. M. Rees, V. Bal, H. Ikeda, D. Wilkinson, R. R. P.De Vries, and J. B. Rothbard. 1988. Prediction and identificationof an HLA-DR-restricted T cell determinant in the 19-kD proteinof Mycobacterium tuberculosis. Eur. J. Immunol. 18:973-976.

26. Mackall, J. C., G. H. Bai, D. A. Rouse, G. R. G. Armoa, F.Chuidian, J. Nair, and S. L. Morris. 1993. A comparison of thedelayed-type hypersensitivity epitopes of the 19kD antigens fromMycobactenium tuberculosis and Myco. intracellulare using overlap-ping peptides. Clin. Exp. Immunol. 93:172-177.

27. Mosmann, T. R., and R. L. Coffman. 1989. TH-1 and TH-2 cells:different patterns of lymphokine secretion lead to different func-tional properties. Annu. Rev. Immunol. 7:145-173.

28. Mosmann, T. R., and T. A. Fong. 1989. Specific assays for cytokineproduction by T cells. J. Immunol. Methods 116:151-158.

29. Nair, J., D. A. Rouse, and S. L. Morris. 1992. Nucleotide sequenceanalysis and serologic characterization of the Mycobacterium in-tracellulare homologue of the Mycobactenium tuberculosis 19kDantigen. Mol. Microbiol. 6:1431-1439.

30. O'Farrell, P. H. 1975. High resolution two-dimensional electro-phoresis of proteins. J. Biol. Chem. 250:4007-4021.

31. Pfeiffer, C., J. Murray, J. Madri, and K. Bottomly. 1991. Selectiveactivation of Th-1 and Th-2-like cells in the in vivo-response tohuman collagen IV. Immunol. Rev. 123:65-84.

32. Rees, A. D. M., A. Faith, E. Roman, J. Ivanyi, K.-H. Wiesmuller,and C. Moreno. 1993. The effect of lipoylation on CD4 T cellrecognition of the 19,000 MW Mycobacterium tuberculosis antigen.Immunology 80:407-414.

33. Schumacher, J. H., A. O'Garoa, B. Schrader, A. Kimmenade,M. W. Bond, T. R. Mosmann, and R. L. Coffmnan. 1988. Thecharacterization of four monoclonal antibodies specific for mouseIL-5 and development of mouse and human IL-5 enzyme-linkedimmunosorbent. J. Immunol. 141:1576-1581.

34. Scott, P., P. Natovitz, R. L. Coffman, E. Pierce, and A. Sher. 1988.Immunoregulation of cutaneous leishmaniasis: T cell lines thattransfer protective immunity or exacerbation belong to different Thelper subsets and respond to distinct parasite antigens. J. Exp.Med. 168:1675-1684.

35. Street, N. E., J. H. Schumacher, T. A. T. Fong, H. Bass, D. R.Fiorentino, J. A. Leverah, and T. R. Mosmann. 1990. Heteroge-neity of mouse helper T cells. Evidence from bulk cultures andlimiting dilution cloning for precursors of Thl and Th2 cells. J.Immunol. 144:1629-1639.

36. Surcel, H.-M., M. Troye-Blomberg, S. Paulie, G. Andersson, C. Moreno,G. Pasvol, andJ. Ivanyi. 1994. Thl/fh2 profiles in tuberculosis, based onthe proliferation and cytokine response of blood lymphocytes to myco-bacterial antigens. Immunology 81:171-176.

37. Tan, P. L J., S. Farmiloe, J. Young, J. D. Watson, and M. A. Skinner.1992. Lymphocyte responses to DR4/1-restricted peptides in rheumatoidarthritis: the immunodominant T cell epitope on the 19kd Mycobacte-nwn tuberculosis protein. Arthritis Rheum. 35:1419-1426.

38. Thole, J. E. R., and J. Van der Zee. 1990. The 65kD antigen:molecular studies on a ubiquitous antigen, p. 37-67. In J. McFad-den (ed.), Molecular biology of the mycobacteria. Surrey Univer-sity Press, London.

39. Verbon, A., S. Kulper, H. M. Jansen, P. Speelman, and A. H. J.KolL 1992. Antibodies against secreted and non-secreted antigensin mice after infection with live Mycobacterium tuberculosis. Scand.J. Immunol. 36:371-384.

40. Young, D. B., and T. R. Garbe. 1991. Lipoprotein antigens ofMycobacterium tuberculosis. Res. Microbiol. 142:55-65.

41. Young, D. B., and T. R. Garbe. 1991. Heat shock proteins andantigens ofMycobacterium tuberculosis. Infect. Immun. 59:3086-3093.

INFECT. IMMUN.

on April 16, 2020 by guest

http://iai.asm.org/

Dow

nloaded from