Molecular Immunology, Vol. 29, No. 9, pp. 1153-l 156, 1992 Printed

0161.5890,‘92 $5.00 + 0.00 Pergamon Press Ltd

in Great Britain

SHORT COMMUNICATION Localisation of linear epitopes at the carboxy-terminal mycobacterial 71 kDa heat shock protein Ashraf Elsaghier, MRC Tuberculosis

Raju Lathigra and Juraj Ivanyi’

and Related Infections UnitRoyal

(Received

3 May

end of the

Postgraduate

1992; accepted

Medical School, London W12 OHS, UK.

18 May

1992)

Summary Four distinct linear epitopes localized within species-specific sequences at the carboxy-terminal end of the 71 kDa heat shock protein of M.tuberculosis have been identified by scanning 94 overlapping peptides with 13 human sera. One epitope (“C”) of entirely M.tuberculosis-specific core sequence (GEAGPG) has been found immunogenic in smear-negative tuberculosis, but not in non-tuberculous mycobacterial diseases. This peptide appears to be a valuable candidate for further serodiagnostic evaluation. Introduction The specificity of the immune repertoire in mycobacterial diseases has been of interest for the study of pathogenesis, the development of immuno-diagnostic tests and vaccines. Serological studies suggested that the antibody response in respect of species-specific protein antigens is restricted to relatively few constituents represented by the 38kDa, 19kDa and 14 kDa antigens in tuberculosis and by the 35 kDa antigen in leprosy (Ivanyi et al., 1991). Antibody levels to these antigens have been found pronounced only in multibacillary forms of the disease. However, it was recently reported that antibody levels to the mycobacterial hsp71 stress protein antigen are elevated to the same extent in patients with either smear-positive, or smear negative tuberculosis (Elsaghier et a/., 1991). Hence, it was of interest to identify the corresponding immunodominant and M.tubercu/osis -specific hsp71 epitopes. We avoided the bulk of the hsp71 sequences which are structurally conserved between microbial and eukaryotic cells and have chosen for epitope scanning the carboxy-terminal end of the molecule which has been found most polymorphic in the alignment of hsp71 sequences from mycobacteria and other species (McKenzie era/., 1991, Lathigra et al., 1992). Materials and Methods The general net synthesis of the 94 peptide pins purchased from Cambridge Research Biochemicals (Cambridge, England) was based on 101 residues derived from the sequence of hsp71 of M. tuberculosis residues 509-621 (Lathigra et al., 1992). Non-cleavable octapeptides, overlapping by one residue starting at residue 509 (VRNQAETL) and ending at residue 621 (GRCPPRLG) were synthesised in solid-phase using the Geysen technique on specially moulded, high-density polyethylene pins which are assembled into a plastic holder designed to hold 96 pins in the format and spacing of a microtitre plate. Unrelated positive and negative control peptide sequences were synthesised on the same block as the test peptides to give an indication of the sensitivity of the ELISA assay. Altogether 13 sera were collected from 2 cases of smear-positive pulmonary tuberculosis, 6 cases of smear-negative tuberculosis, and 1 case each of lepromatous leprosy, non-tuberculous pulmonary disease of proven infection with M.avium, lung fibrosis, Crohn’s disease and one healthy subject. Sera had previously been tested and all found as of high titre (l/2000 - l/4000) against the recombinant mycobacterial hsp71 (Elsaghier et al. 1991) Reactions of the tips of the pins were carried out in 96 well polypropylene microtitre plates following the producers recommended procedure. Briefly, the pins were incubated with 175 @/well of 11400 diluted human serum, washed, reacted with peroxidase labelled goat anti-human IgG and subsequently with substrate solution, 0.5 mg/ml of 2,2’-azino-di [3-ethyl-benthiazoline sulfonate] in citrate buffer pH 4.0 and 0.01% w/v hydrogen peroxide. The reaction was stopped by removing the pins and optical density was measured at 405 nm in the Titertek “Multiscan MCC II” spectrophotometer. After each test, the block was incubated for 10 minutes in a sonication

bath containing

v/v 2-mercaptoethanol,

O.lM phosphate

rinsed,

immersed

buffer pH 7.2 at 60°C

with 1% w/v sodium dodecyl sulphate and 0.1%

in boiling methanol for 15 seconds,

air dried and stored at 4OC.

Results Representative optical density values from the ELISA assay of serum from a patient with smear-negative pulmonary tuberculosis showed elevated antibody binding to several discretely separated clusters of adjacent peptides (figure 1). Four areas of peptides, with consistent localisation of increased binding had been designated as epitopes A (peptides 33-41) ,B (peptides 44-52) C (peptides 56-67) and D (peptides 7485). However, elevated

1153

1154

A. ELSAGHIER et al.

binding values to peptides conserved sequence, located further towards the amino-terminus were omitted from further evaluation. One peptide No.77 within epitope D showed low binding with all tested sera which we considered to be due to technically defective peptide structure or presentation in view of the consistent performance of adjacent peptides on both sides from peptide 77.

Fioure 1, Antibody binding to 94 octamer peptides using serum from a patient with sputum srnearnegative pulmonary tuberculosis.

In order to evaluate the contribution of individual amino acids to the epitope structure, we have calculated the mean absorbance values for each residue from all 6 overlapping octapeptides containing the respective amino acid. This analysis was applicable to sequences, starting from the leucine at position 516, which represented the last residue in peptide number 1. The results shown in figure 2 demonstrate the values obtained from testing one serum each from patients with pulmonary tuberculosis (calculated from values shown in figure l), lung disease caused by M.avium and with lepromatous leprosy. Antibody binding activities towards individual residues closely overlapped in their localisation and were clearly pronounced for epitopes B and D with each of the three tested sera. In contrast, binding to epitopes A and C identified with the tuberculous serum was not significant with the two non-tuberculous control sera. This individual variation in the binding patterns between the three sera shown in figure 2 and several other tested sera (results not shown) suggested that the identified epitopes are of distinct immunogenicity.

&t&2. Locallsatlon of epitopes wthm the carboxy-terminal hsp71 sequence. Mean absorbance values from eight adjacent octamer the individual amino acid restdues llsted on the horizonta scale. Sera from patients with smear-negative peptides each containin pulmonary tuberculosis T, see bindin to individual peptldes in figure 1). lung disease caused by M. avium (AM) and lepromatous leprosy (LL). A-D = localisation ofB epftope core 3, examers.

(

The core structures for each of the four identified epitopes have been arbitrarily allocated to six residues graded with the highest mean OD values. They are represented by sequences DAAVAE for epitope A, WRIGYF for epitope B, GEAGPG for epitope C and CVTGHW for epitope D (Figure 3). Alignment of these hexamers within the sequence of the hsp71 carboxy terminus from M.tuberculosis, with corresponding sequences of M.leprae hsp70, EcolidnaK, and of the human HSP70-1 gene revealed, that the core of epitope A is overlapping in 5 out of the 6 core residues with the M./eprae sequence. In contrast, epitopes D, B and C shared only 2, 1 or no residue

Epitopes of mycobacterial

hsp7l antigen

respectively with M. leprae. None of the four epitopes had significant homology with either dnaK or human HSP70-1 structure. Predictive epi!ope evaluation of the carboxy-terminal sequence by the “Protean” programme (Proteus Biotechnology Ltd., Manchester, UK; kindly performed by .f.Edwards at Peptide Technology ~td.Sydney~Aust~l~a) allocated the highest score to a single broad stretch of residues 565 - 591 which contains

GLK GDX **T a

k

a530

xsrp GEXTA E

E.

-554

LQV

nix*

QAL E

A-

,RCNPIISGLYQGAGG--PGPGGFGA

e

k?S

&gg#_& Alrgnmenror carbox~lermina~ amino-acid sequences. fju: _F$man __hsp70-f , ,Ec=, Escherkbia -.cd anaK, t.p= ~y,~oacrer~um reprae ftsprt, hJcwba,crertum tubylosls Hsp71 c enz~e er al., 1991, Lathi ra 81 a/., 1992). l = twquencesidentical with ycobacterium luberculosis. - = gaps.

Ip=

QQQEAQQQTAGADA#ANNAKo--‘QAIYEATQAASX’G*LASA*GGS 83Ea9P~BSSSE~YrEH1~~f---S02 14n_C_

GSGSGPTIEEVDDDVVDAEFLEVKD-*K *WSTTNGSPI NSTDDVLT’ ___-__ E&&aacai;~~a~s621 QGPKG-

_ __

A sem~~uantitati~e evatuatitbn of antibodies in the 13 tested sera against i~d~vidual egitopes has been done by calcufation of mean UD vatues from the 7 - If peptides, ~orres~nding to each of the four epitopes. The results represented in Figure 4, demonstrate significanfty higher antibody binding to each of the four epitopes in stara from tuberculous patients when compared with non-tuberculous controls.This difference was more significant for epitopes C and D (PC 0.001) than for epitopes A and B (pcO.01). However, it should be remembered, that the serum from the healthy subject and the other 3 controls were purposely selected to contain high antibody binding to fuifil the topographic aims of the study. f-fence, much better discrimination would be expected between TB patients and randomly selected non-tubercufous patients and healthy controls.

Hguq_& Semi-ran!itativs evaluation of antIbodIes la four (Aepltopes. Sera diluted 1:400 were collected from patients with smear-positive full circle) or smear-negative (shaded circle) tu b erculosis and from non-

)

were pre-selected to contain high &body levels to the whole hsp71 antigen. Absotbances (vertical scale) are mean values from adjacent peptldes correspcmding to each of the four epitopes (bottom horizontal scale).

Discussian Antibody formation is largely based on the recognition of epitopes displayed on native molecules by the immunoglobulin receptors of B cells. This is corroborated by the fact that the bulk of antibodies in patients with multibacillary tuberculosis are directed against conformational epitopes of the secreted lipoprotein 38kDa and 19kDa antigens (Ivanyi 7 991 f_ In the pau~~bac~~~a~ stage of infection however, the most prominent serological response was found to be directed against the 71 kDa stress protein (Efsaghisr ef al, tg91). The pronounced immunogenicify of hsp71 during paucibaciliary infection couid be attributed to enhanced stress-induced

A. ELSAGHIERet al.

1156

expression of hsp71 during intracellular replication (Buchemeier et al., 1990), or to surface expression (Vanburskirk. et a/., 1989) enabling recognition by the lg receptors of B cells or merely to earlier priming by multiple cross-reacting commensal microbial organisms and autologous hsp70. Although the hsp71 protein is a predominantly intracellular constituent (Mehlert and Young 1989) its presence in media from short-term mycobacterial cultures (Abou-Zeid et al, 1988) indicates some form of secretion from live bacteria which could be contributory to its immunogenicity for B cells. Elevated serum antibody levels to both microbial and human hsp71 have been found in a variety of infectious, chronic inflammatory and autoimmune diseases (Britton eta/., 1986, Minota et al., 1988, Tsoulfa et al, 1989, Mattei er al., 1989, Elsaghier et a/., 1991, 1992). However, it is notable, that antibodies produced following infection with Schistosoma mansoni were found to be predominantly directed against species-specific epitopes of hsp70 (Newport et a/., 1988). The present peptide-scan analysis of the most polymorphic carboxyterminal part of the hsp71 sequence allowed the localisation of four distinct epitopes. The C epitope composed of the GEAGPG core and adjacent residues, unique for the M. fuberculosis antigen reacted positively with sera from the majority (5 out of 6 tested) of smear-negative patients but with none of the non-tuberculous controls. The corresponding sequence also had the highest score in the prediction evaluation using the “Protean” programme. It is possible that the M.tubercu/osis epitope aligned sequences from other microbial and eukaryotic species may also carry immunogenic epitopes. Their potential as species-specific probes for the involvement of hsp 71 could be valuable in the study of immunopathogenesis of various infectious diseases or autoimmunity and possibly for immunodiagnostic applications. References Abou-Zeid C., Smith I., Grange J.M., Ratliff T.L., Steele J., and Rook G.A.W. (1988) The secreted antigens of M. tuberculosis and their relationship to those recognised by the available antibodies. J. Gen Microbial. 134: 531-538. Britton, W.J., Hellqvist, L., Basten, A. and lnglis A.S., (1986) lmmunoreactivity of a 70 kDa protein purified from Mycobacterium bovis Bacillus Calmette-Guerin by monoclonal antibody affinity chromatograph. J. Exp. Med. 16 4: 695-708. Buchemeier N.A., and Heffron F., (1990) Induction of Salmonella stress proteins upon infection of macrophages. Science. 24 8 : 730-732. Elsaghier A., Prantera C., Bothamley G., Wilkins E., Jindal S. and lvanyi J. (1991) Disease association of antibodies to human and mycobacterial hsp70 and hsp60 stress proteins. Clin Exp Immunol. In press. (1991) Elevated Elsaghier A.A.F., Wilkins E.G.L., Mehrotra P.K., Jindal S. and lvanyi J. antibody levels to stress protein hsp70 in smear-negative tuberculosis.lmmunol & Inf. Diseases. 1 :323. lvanyi J., Serological tests for the diagnosis of tuberculosis and leprosy. Rapid Methods and Automation in Microbiology and immunology (Edited by Vaheri A.), Springer-Verlag, Berlin. 1991. 267-277. Lathigra R., Alexander W., Stover K., Coadwell J., Young R., and Young D. (1992) Genetic characterization of the dnaK-grpE-dnaJ heat shock gene cluster in Mycobacterium tuberculosis. J Biol Chem. In press. L.P. (1989) A heat shock-like protein from the Mattei D., Scherf A., Bensaude O., and da Silva human malaria parasite Plasmodium falciparum induces autoantibodies. Eur. J. Immunol. 1 9 : 1823-l 828. Garsia R.J. and Basten A. (1991) Sequence and McKenzie K.R., Adams E., Britton W.J., immunogenicity of the 70-kDa heat shock protein of Mycobacterium leprae. J. Immunol. 147: 312-319. D.B. (1989) Biochemical and antigenic characterization of the Mycobacterium Mehlert A., and Young tuberculosis 77 kD antigen, a member of the 70kD heat-shock protein family. Mol Microbial. 3 :(2): 125130. Minota S., Cameron B., Welch W.J., and Winfield J.B. (1988) Autoantibodies to the constitutive 73kD member of the hsp70 family of heat shock proteins in systemic lupus erythematosus. J. Exp. Med. 168: 1475-l 880. Kallestad J.,Tarr P., Klebanoff S., Agabian N. (1988) Newport G-R., Hedstrom R.C., Identification, molecular cloning and expression of a schistosome antigen displaying diagnostic potential. Am. J. Trop.Med.Hyg. 5 4 0 :540-546. Tsoulfa G., Rook G.A.W., Bahr G.M., Sattar M.A., Behbehani K., Young D.B., Mehlert A., van-Embden J.D.A., Hay F.C., lsenberg D.A., and Lydyard, P.M. (1989) Elevated IgG antbody levels to the mycobacterial 65kDa heat shock protein are characteristic of patients with rheumatoid arthritis. Stand. J. Immunol. 30 : 519-527. (1989) A peptide binding protein Vanbuskirk A., Crump B.L., Margoliash E., and Pierce S. J. having a role in antigen presentation is a member of the Hsp70 heat shock family. J. Exp. Med. 170: 1799-l 809. ‘Corresponding

author

(FAX:

081. 743 3987, TEL: 081.740

3161)

Localisation of linear epitopes at the carboxy-terminal end of the mycobacterial 71 kDa heat shock protein.

Four distinct linear epitopes localized within species-specific sequences at the carboxy-terminal end of the 71 kDa heat shock protein of M. tuberculo...
433KB Sizes 0 Downloads 0 Views