Immunology and Cell Biology (1992) 70, 295-300
Antibodies to 65 kDa and 70 kDa heat shock rheumatoid arthritis and systemic lupus eryth JAI PANCHAPAKESAN, MALAMA DAGLIS, PAUL GATENBY Deparlment of Clinical Immunology, Royal Prince Alfred Hospital Camperdown, New South Wales, A ustralia Summary To test the potential role of autoimmunity to the highly conserved heat shock proteins (HSP) in immune arthritides, the sera from 99 patients with rheumatoid arthritis (RA), 48 patients with systemic lupus erythematosus (SLE) and 65 normal controls were examined by ELISA for IgG and IgM antibodies to the 65 kDa and 70 kDa heat shock proteins from Mycobacterium bovis (Bacille Calmette-Guerin; BCG). In RA sera there are significant numbers of individuals with increased IgM anti-65 kDa and anti-BCG reactivity as well as IgG anti-70 kDa when compared with controls. In SLE both IgM and IgG anti-BCG, together with IgM anti-65 kDa, differed significantly from controls. The results were compared with previous reports in simitar groups of patients, and it is clear that no consistent pattern of reactivity emerges. While further work may be justified looking carefully at the disease duration and other subsets of both RA and SLE, it is difficult at this stage to conclude that antibodies to autologous HSP that cross-react with mycobacterial HSP play a major role in disease pathogenesis.
Introduction Heat shock or stress proteins (HSP) are highly conserved throughout phylogeny, are induced by a wide range of noxious environmental insults and in microbial infections of a number of types, and serve as immunodominant antigens.' As both microbial and host HSP may be expressed during infection, HSP have emerged as prime candidates for the stimulation of autoreactive B and T cell responses because of their highly conserved structures.^""'* Indeed, a clear role for reactivity to one HSP family member, especially the 65 kDa HSP, has been demonstrated in a number of animal models of arthritis."* Furthermore, both cellular^"" and humoraP"^' responses to various HSP have been reported in a number of human arthritides, although other authors have been unable to reproduce these findings. ^^ This study was carried out to extend our knowledge about the occurrence
of antibodies to mycobacterial HSP from the 60-65 kDa and 70 kDa HSP families m rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE).
Materials and methods Subjects Sera were obtained from 99 patients with definite RA from the Combined Arthritis Unit of the Royal Prince Alfred Hospital and from 48 patients with definite SLE obtained from the Immunology Clinic at the Royal Prince Alfred Hospital. All patients met the appropriate American Rheumatism Association diagnostic criteria for their clinical diagnosis.'-''-^'* Normal sera age-matched to the RA patients were obtained as small aliquots from 65 normal blood donors at the Central Sydney Blood Bank by courtesy of the NSW Red Cross Blood Transfusion Service.
Correspondence: Associate Professor Paul Gatenby, Department of Clinical Immunology, Royal Prince Alfred Hospital, Camperdown, NSW 2050. Australia. Accepted for publication 28 May 1992.
296
J. Panchapakesan et al.
The study had appropriate institutional ethical approval.
with PBS. One hundred microlitres of 3% bovine serum albumin (BSA) in PBS were added to eacb well and incubated for 2 h at 37''C, after which wells were washed once in Serum preparation PBS. Fifty microlitres of sera diluted to 1 : 100 The serum was separated from clotted blood as in 1% BSA-PBS-Tween 0.05% was added to soon as it was received and stored in multiple duplicate wells and incubated for 2 h at 37'"C. aliquots at -2O''C. It was thawed only once Plates were tben washed twice in PBS-Tween prior to testing. and twice with PBS. Fifty microlitres of alkaline phosphatase-conjugated second antibody, Antigens either goat antihuman IgG or goat antibuman Mycobacterium bovis (Bacille Calmette-Guerin IgM (Sigma, Sydney, Australia) diluted 1 : 1000 in 1% BSA-PBS-Tween, was added [BCG]; Commonwealth Serum Laboratories, per well, and tbe plates were incubated a Melbourne, Australia) was suspended in sterile furtber 1 b at 37°C. Plates were then washed phosphate-buffered saline (PBS), sonicated for twice in PBS-Tween, once in triple distilled five 4 min bursts (Branson Sonifier 20, water and once in p-nitropbenyl phosphate Connecticut, USA) and then centrifuged for disodium buffer (NPP; Sigma, Sydney, 20 mins at 20 000^^. The presence of 65 kDa Australia) before addition of the chromogenic and 70 kDa HSP antigens was confirmed by agent. After further incubation the optical SDS-PAGE and subsequent Western Blot density (OD) was read at 405 nm with an analysis using antibodies L22 and L7, respecELISA plate reader equipped witb appropriate tively (a gift from Dr W. Britton, Department filters (Multiskan, Flow Laboratories). Each of Medicine, University of Sydney). The 65 kDa protein was purified from M. bovis BCG plate contained pooled cord blood-derived serum and pooled sera from leprosy patients as sonicate supernatant by affmity column negative and positive controls, respectively. chromatography using the L22 monoclonal These conditions were establisbed as the antibody (mAb) fixed to cyanogen bromideoptimum in a series of preliminary experiactivated sepharose 4B (Pharmacia, Uppsala, ments using a wide range of antigen concenSweden). Likewise the 70 kDa protein was trations, serum dilutions and second step purified on a similar affmity column containantibody diluted in a checkerboard manner. ing the L7 mAb. Both crude antigens were then further purified by loading tbe L22 and L7 eluates onto an L7 and L22 column, Statistical analysis respectively, and collecting the unbound The OD from all assays was converted to a fractions. In this way any contamination of ratio of the positive control in each plate to tbe 65 kDa antigen with 70 kDa antigen and allow interassay comparisons to be made. Tbe tbe converse was minimized. Purity at each positive control was arbitrarily designated as 1. stage of fractionation was monitored by SDSGroups were compared by the MannPAGE and Western Blot analysis.'^'^^ Whitney rank sum test using tbe STATVIEW software package for the Apple-Macintosh Measurement of antibodies computer. The P value obtained was subject to Enzyme-linked immunoassays (ELISA) similar the Bonferoni correction for multiple to those reported previously^'"* were set up comparisons.*^ for both proteins and for sonicated BCG. In brief, antigens were diluted in carbonate/ bicarbonate buffer pH 9.6 (65 kDa HSP and Results 70 kDa HSP) at 1 or 2 ^ig/mL for tbe IgG and IgM assays, respectively, BCG at 40 |ig/nil Tbe subjects witb RA ranged in age from 31 and 100 |J.L of antigen added to each well of a to 82 years (mean 63.4 years). Subjects witb 96-weIl flat-bottomed polystyrene plate SLE were younger witb an age range of 13 to (Flow Laboratories, VI, USA). Plates were 67 years (mean 37.4 years). The controls had a incubated overnight at 37 ° C and washed once range of 50-73 years (mean 57.4 years).
Antibodies to heat shock proteins in RA and SLE Tbe purity of botb tbe 65 kDa and 70 kDa antigens are demonstrated by SDS-PAGE and Western Blot in Figs 1 and 2, respectively. Two bands are seen in the 65 kDa lanes; tbis is a common finding witb the 65 kDa antigen and probably represents a minor degree of antigen breakdown. Tbe results of tbe median values of tbe OD ratio for the different antibodies are summarized in Table 1. When compared witb the controls with corrected P values in tbe RA sera tbere was a significant increase in tbe number of individuals with IgG anti-70 kDa, IgM anti-65 kDa and IgM anti-BCG antibodies. In tbe patients with SLE there was a significant increase in the number of individuals witb IgM anti-65 kDa and anti-BCG antibodies but no difference in either isotype with anti-70 kDa and a
MW(kDa}
116 84 58 48
36.526.6-
A
B
C
Fig. 1. A Coomassie Blue stained, SDS-PAGE separated gel showing: (a) 65 kDa HSP; (b) BCG sonicate; and (c) 70 kDa HSP. The 65 kDa antigen is seen as two bands, the one of lower molecular weight representing a minor degree of breakdown.
297
significant increase in tbe number of positives with IgG antibodies but not IgM to BCG. These statistical differences are summarized in Table 2.
Discussion The results obtained in tbis study demonstrate an increase in binding of the IgM isotype to tbe 65 kDa HSP in both RA and SLE when compared witb normal controls. These results are quite different from previous descriptions of similar or even reduced binding by IgM anti-65 kDa HSP when compared with their controls.^'^ Tbe binding for anti-65 kDa HSP here was paralleled by tbe results obtained by IgM anti-BCG. As tbe 65 kDa is a dominant antigen in BCG and other mycobacteria,' tbese antibodies may reflect exposure to mycobacteria, most likely environmental as BCG vaccination is not widely used in Australia. A further contrast to tbe above studies was seen wjtb tbe results obtained for IgG anti-65 kDa HSP, where no differences from controls were demonstrated. Tbe only significant results witb tbe 70 kDa antigen were with IgG antibodies and RA. Here the results differ from Tsoulfa et a!., who observed no difference between RA, SLE or tuberculosis patients with regard to antibodies to tbis antigen, and concluded tbat there appeared to be something particular about the IgG response to 65 kDa antigen in RA when compared with controls or the 70 kDa responses.^ However, tbe data reported here clearly do not support tbat view. Indeed, other groups have been unable to confirm reactivity to the 65 kDa or tbe 70 kDa antigen in RA, although Jarjour et al. did demonstrate sucb reactivity in a wide range of other rheumatic and inflammatory diseases.'^ It is very unlikely that tbis is due to technical variation between laboratories as most groups have bad little trouble identifying anti-HSP antibodies. However, the antibodies are detected in different groups of patients. Otber Australian workers bave only found increased activity to 65 kDa antigen in individuals who bad experienced mycobacterial infections such as
J. Panchapakesan et al.
298
L22
A
B
C
L7
D
E
F
Fig. 2. Western blotting of separated 65 kDa, 70 kDa and BCG sonicate. Two blots are illustrated, stained with either L22 (anti-65 kDa monoclonal antibody) or L7. (a) 65 kDa HSP; (li) 70 kDa HSP; (c) BCG sonicate; (d) BCG sonicate; (e) 70 kDa HSP; (f) 65 kDa HSP. The 65 kDa HSF shows two bands, the one of lower molecular weigbt representing a minor degree of breakdown. Table 1. Median values of binding for antibodies to 65 kDa. 70 kDa, HSP and BCG in RA, SLE and control sera for both IgG and IgM isotypes. Anti-65 kDa
Anti-70 kDa
Anti-BCG
Controls
IgG IgM
0.20(0.19-0.30) 0.24 (0.25-0.32)
0.20(0.19-0.29) 0.25 (0.23-0.30)
0.29 (0.28-0.34) 0.26 (0.24-0.29)
RA
IgG IgM
0.15 (0.18-0.27) 0.39(0.41-0.55)
0.12(0.12-0.18) 0.28(0.31-0.50)
0.29 (0.28-0.34) 0.38 (0.37-0.45)
SLE
IgG IgM
0.24 (0.24-0.49) 0.51 (0.44-0.56)
0.17(0.16-0.24) 0.31 (0.28-0.41)
0.41 (0.37-0.52) 0.37(0.31-0.41)
Values are expressed as anopticaldensity ratio of percentage of the positive control (designated as 1). The 95% confidence limits are shown in brackets. tuberculosis or leprosy (M. Rowley, pers. comm.). Tsoulfa et al. '** and McLean et al. also reported that IgA anti-65 kDa HSP was elevated in RA.'*^ The present study could not
confirm these findings in a series of preliminary experiments (data not shown), The idea of immune responses to these highly conserved HSP participating in auto-
Antibodies to heat shock proteins in RA and SLE Table 2. Summary of differences (P values) becween RA and SLE binding of IGG and IGM antibodies to the three antigens.
3.
Anti-65 kDa Anti-70 kDa Anti-BCG 4. RA IpG 0.42 IgM 0.0003
0.0009 0.25
2.8 0.0003
SLE IgG 0.27 IgM 0.0003
1.15 0.11
0.037 0.018
5.
Corrected P values are shown for each of the three antigens for both IgG and IgM antibodies. 6. immune disease by molecular mimicry, including the demonstration of the autologous 65 kDa analogue in inflamed rheumatoid synovium^** and the binding of monoclonal antibodies to this protein in inflamed joints, is attractive. Despite this, the highly variable nature of the results, with different groups finding either no increased binding or binding with different isotypes, is consistent with the conclusion that humoral responses to 65 kDa HSP are extremely variable in both RA and SLE. The groups of patients studied by different workers appear, as best can be judged, to be relatively comparable, and this variability makes it unlikely that the antibodies play any specific role in the pathogenesis of disease. These results, together with negative results reported for T cell reactivity,*^'^' must lead to a re-evaluation of the whole idea of molecular mimicry to HSP and the development of autoimmune disease.
7.
8.
9.
10.
Acknovrledgements tl. We would like to thank Dr J. R. York, Director of the Royal Prince Alfred Hospital Combined Arthritis Unit. The study was supported by the NH&MRC. 12. References 1. Young, D. B., Ivanyi, J., Cox, J. H. and Lamb, J. R. 1987. The 65 kDa antigen of mycobacteria—A common bacterial protein?/mmMfio/. Today 8: 215-219. 2. Shoenfeld, Y. and Isenberg, D. A. 1988. Myco-
13.
299
bacteria and autoimmunity. Immunol Today 9: 178-181. Winfield, J. B. 1989. Stress proteins, arthritis and autoimmunity. Arthr. Rheum. 32: 14971504. Young, D. B. 1990. The immune response to mycobacterial heat shock proteins. Autoimmunity 7: 237-244. van Eden, W., Hogervosst, E. J. M., Hcnsen, E. J., van der Zee, R., van Embden J. D. A. and Cohen, I. R. 1989. A cartilage-mimicking T-cell epitope on a 65 K mycobacterial heatshock protein: Adjuvant arthritis as a model for human rheumatoid arthritis. Curr. Top. Microbiol. Immunol. 145: 27-43. Res, P. C. M., Schaar, C. G., Breedveld, F. C . van Bden, W., van Embden, J. D. A., Cohen, I. R. and De Vries, R. R. P. 1988. Synovial fluid T cell reactivity against 65 kDa heat shock protein of mycobacteria in early chronic arthritis. Lancet ii: 478-480. Holoshitz, J., Kosek, J., Sibley, R., Brown, D. A. and Strober, S. 1991. T lymphocytesynovial fibroblast interactions induced by mycobacterial proteins in rheumatoid arthritis. Arthr. Rheum. 34: 679-686. Pope, R. M., Lovis, R. M. and Gupta, R. S. 1992. Activation of synovial fluid T lymphocytes by 60 kDa heat shock proteins in patients with inflammatory synovitis. Arlhr. Rheum. 35: 43-48. Tsoulfa, G., Rook, G. A. W., Bahr. G. M. et al. 1989. Elevated IgG antibody levels to the mycobacterial 65 kDa heat shock protein are characteristic of patients with rheumatoid arthritis. Scand. J. Immunol. 30: 519-527. Tsoulfa, G., Rook, G. A. W., van Embden, J. D. A. et al. 1989. Raised serum IgG and IgA antibodies to mycobacterial antigens in rheumatoid arthritis. Ann. Rheum. Dis. 48: 118123. Minota, S., Cameron, B., Welch, W. J. and Winfield J. B. 1988. Autoantibodies to the constitutive 73 kDa member of the hsp70 family of heat shock proteins in systemic lupus erythematosus./ Exp. Med. 168: 1475-1480. Jarjour, W. N., Jeffries, B. D., Davis, J. S., Welch, W.J., Mimura, T. and Winfield, J. B. 1991. Autoantibodies to human stress proteins. A survey of various rheumatic and other inflammatory diseases. Arthr. Rheum. 34: 11331138. Arnett, F. C , Edworthy, S. M., Block, D. A. et al. 1988. The American Rheumatism Association 1987 revised criteria for the classification
300
14.
15. 16.
17. 18.
J. Panchapakesan et al. of rheumatoid arthritis. Arthr. Rheum. 31: 315-324. Tan,E.M.,Cohen,A.S.,Fries,J.F.erj/. 1982. The 1982 revised criteria for the classification of systemic lupus erythematosus. Arlhr. Rheum. 25:1271-1277. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685. Towbin, H., Staehelin T. and Gordon, J. 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc. Natl Acad. Sci. USA 76: 4350-4354. Glantz, S. 1981. Primer of Biostatistics. McGraw-Hill, New York. pp. 87-88. McLean, I. L,, Archer, J. R., Cawley, M. I. D., Pegley F. S., Kidd, B. L. and Thompson, P. W. 1990. Specific antibody response to the mycobacterial 65 kDa stress protein in ankylosing
spondylitis and rheumatoid arthritis. Br. J. Rheum. 29: 426-429. 19. Karlsson-Parra, A., Soderstrom, K., Perm. M., Ivanyi. J., Kiessling. R. and Klareskog, L. 1990. Presence of human 65 kDa heat shock protein (hsp) in inflamed joints and subcutaneous nodules of RA patients. Scand. J. Immunol. 31: 283-288. 20. de Graeff-Meeder, E. R.. Voorhorst, M.. van Eden, W. et al. 1990. Antibodies to the mycobacterial 65 kd heat shock protein are reactive with synovial tissue of adjuvant arthritic rats and patients with rheumatoid arthritis and osteoarthritis. Am.J. Path. 137: 1013-1017. 21. Crick, F. D. and Gatenby, P. A. 1992. Limiting dilution analysis of T cell reactivity to mycobacterial antigens in peripheral blood and synovium from rheumatoid arthritis patients. Clin. Exp. Immunol. 88: 424-429.
Antibodies to 65 kDa and 70 kDa heat shock rheumatoid arthritis and systemic lupus eryth JAI PANCHAPAKESAN, MALAMA DAGLIS, PAUL GATENBY Deparlment of Clinical Immunology, Royal Prince Alfred Hospital Camperdown, New South Wales, A ustralia Summary To test the potential role of autoimmunity to the highly conserved heat shock proteins (HSP) in immune arthritides, the sera from 99 patients with rheumatoid arthritis (RA), 48 patients with systemic lupus erythematosus (SLE) and 65 normal controls were examined by ELISA for IgG and IgM antibodies to the 65 kDa and 70 kDa heat shock proteins from Mycobacterium bovis (Bacille Calmette-Guerin; BCG). In RA sera there are significant numbers of individuals with increased IgM anti-65 kDa and anti-BCG reactivity as well as IgG anti-70 kDa when compared with controls. In SLE both IgM and IgG anti-BCG, together with IgM anti-65 kDa, differed significantly from controls. The results were compared with previous reports in simitar groups of patients, and it is clear that no consistent pattern of reactivity emerges. While further work may be justified looking carefully at the disease duration and other subsets of both RA and SLE, it is difficult at this stage to conclude that antibodies to autologous HSP that cross-react with mycobacterial HSP play a major role in disease pathogenesis.
Introduction Heat shock or stress proteins (HSP) are highly conserved throughout phylogeny, are induced by a wide range of noxious environmental insults and in microbial infections of a number of types, and serve as immunodominant antigens.' As both microbial and host HSP may be expressed during infection, HSP have emerged as prime candidates for the stimulation of autoreactive B and T cell responses because of their highly conserved structures.^""'* Indeed, a clear role for reactivity to one HSP family member, especially the 65 kDa HSP, has been demonstrated in a number of animal models of arthritis."* Furthermore, both cellular^"" and humoraP"^' responses to various HSP have been reported in a number of human arthritides, although other authors have been unable to reproduce these findings. ^^ This study was carried out to extend our knowledge about the occurrence
of antibodies to mycobacterial HSP from the 60-65 kDa and 70 kDa HSP families m rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE).
Materials and methods Subjects Sera were obtained from 99 patients with definite RA from the Combined Arthritis Unit of the Royal Prince Alfred Hospital and from 48 patients with definite SLE obtained from the Immunology Clinic at the Royal Prince Alfred Hospital. All patients met the appropriate American Rheumatism Association diagnostic criteria for their clinical diagnosis.'-''-^'* Normal sera age-matched to the RA patients were obtained as small aliquots from 65 normal blood donors at the Central Sydney Blood Bank by courtesy of the NSW Red Cross Blood Transfusion Service.
Correspondence: Associate Professor Paul Gatenby, Department of Clinical Immunology, Royal Prince Alfred Hospital, Camperdown, NSW 2050. Australia. Accepted for publication 28 May 1992.
296
J. Panchapakesan et al.
The study had appropriate institutional ethical approval.
with PBS. One hundred microlitres of 3% bovine serum albumin (BSA) in PBS were added to eacb well and incubated for 2 h at 37''C, after which wells were washed once in Serum preparation PBS. Fifty microlitres of sera diluted to 1 : 100 The serum was separated from clotted blood as in 1% BSA-PBS-Tween 0.05% was added to soon as it was received and stored in multiple duplicate wells and incubated for 2 h at 37'"C. aliquots at -2O''C. It was thawed only once Plates were tben washed twice in PBS-Tween prior to testing. and twice with PBS. Fifty microlitres of alkaline phosphatase-conjugated second antibody, Antigens either goat antihuman IgG or goat antibuman Mycobacterium bovis (Bacille Calmette-Guerin IgM (Sigma, Sydney, Australia) diluted 1 : 1000 in 1% BSA-PBS-Tween, was added [BCG]; Commonwealth Serum Laboratories, per well, and tbe plates were incubated a Melbourne, Australia) was suspended in sterile furtber 1 b at 37°C. Plates were then washed phosphate-buffered saline (PBS), sonicated for twice in PBS-Tween, once in triple distilled five 4 min bursts (Branson Sonifier 20, water and once in p-nitropbenyl phosphate Connecticut, USA) and then centrifuged for disodium buffer (NPP; Sigma, Sydney, 20 mins at 20 000^^. The presence of 65 kDa Australia) before addition of the chromogenic and 70 kDa HSP antigens was confirmed by agent. After further incubation the optical SDS-PAGE and subsequent Western Blot density (OD) was read at 405 nm with an analysis using antibodies L22 and L7, respecELISA plate reader equipped witb appropriate tively (a gift from Dr W. Britton, Department filters (Multiskan, Flow Laboratories). Each of Medicine, University of Sydney). The 65 kDa protein was purified from M. bovis BCG plate contained pooled cord blood-derived serum and pooled sera from leprosy patients as sonicate supernatant by affmity column negative and positive controls, respectively. chromatography using the L22 monoclonal These conditions were establisbed as the antibody (mAb) fixed to cyanogen bromideoptimum in a series of preliminary experiactivated sepharose 4B (Pharmacia, Uppsala, ments using a wide range of antigen concenSweden). Likewise the 70 kDa protein was trations, serum dilutions and second step purified on a similar affmity column containantibody diluted in a checkerboard manner. ing the L7 mAb. Both crude antigens were then further purified by loading tbe L22 and L7 eluates onto an L7 and L22 column, Statistical analysis respectively, and collecting the unbound The OD from all assays was converted to a fractions. In this way any contamination of ratio of the positive control in each plate to tbe 65 kDa antigen with 70 kDa antigen and allow interassay comparisons to be made. Tbe tbe converse was minimized. Purity at each positive control was arbitrarily designated as 1. stage of fractionation was monitored by SDSGroups were compared by the MannPAGE and Western Blot analysis.'^'^^ Whitney rank sum test using tbe STATVIEW software package for the Apple-Macintosh Measurement of antibodies computer. The P value obtained was subject to Enzyme-linked immunoassays (ELISA) similar the Bonferoni correction for multiple to those reported previously^'"* were set up comparisons.*^ for both proteins and for sonicated BCG. In brief, antigens were diluted in carbonate/ bicarbonate buffer pH 9.6 (65 kDa HSP and Results 70 kDa HSP) at 1 or 2 ^ig/mL for tbe IgG and IgM assays, respectively, BCG at 40 |ig/nil Tbe subjects witb RA ranged in age from 31 and 100 |J.L of antigen added to each well of a to 82 years (mean 63.4 years). Subjects witb 96-weIl flat-bottomed polystyrene plate SLE were younger witb an age range of 13 to (Flow Laboratories, VI, USA). Plates were 67 years (mean 37.4 years). The controls had a incubated overnight at 37 ° C and washed once range of 50-73 years (mean 57.4 years).
Antibodies to heat shock proteins in RA and SLE Tbe purity of botb tbe 65 kDa and 70 kDa antigens are demonstrated by SDS-PAGE and Western Blot in Figs 1 and 2, respectively. Two bands are seen in the 65 kDa lanes; tbis is a common finding witb the 65 kDa antigen and probably represents a minor degree of antigen breakdown. Tbe results of tbe median values of tbe OD ratio for the different antibodies are summarized in Table 1. When compared witb the controls with corrected P values in tbe RA sera tbere was a significant increase in tbe number of individuals with IgG anti-70 kDa, IgM anti-65 kDa and IgM anti-BCG antibodies. In tbe patients with SLE there was a significant increase in the number of individuals witb IgM anti-65 kDa and anti-BCG antibodies but no difference in either isotype with anti-70 kDa and a
MW(kDa}
116 84 58 48
36.526.6-
A
B
C
Fig. 1. A Coomassie Blue stained, SDS-PAGE separated gel showing: (a) 65 kDa HSP; (b) BCG sonicate; and (c) 70 kDa HSP. The 65 kDa antigen is seen as two bands, the one of lower molecular weight representing a minor degree of breakdown.
297
significant increase in tbe number of positives with IgG antibodies but not IgM to BCG. These statistical differences are summarized in Table 2.
Discussion The results obtained in tbis study demonstrate an increase in binding of the IgM isotype to tbe 65 kDa HSP in both RA and SLE when compared witb normal controls. These results are quite different from previous descriptions of similar or even reduced binding by IgM anti-65 kDa HSP when compared with their controls.^'^ Tbe binding for anti-65 kDa HSP here was paralleled by tbe results obtained by IgM anti-BCG. As tbe 65 kDa is a dominant antigen in BCG and other mycobacteria,' tbese antibodies may reflect exposure to mycobacteria, most likely environmental as BCG vaccination is not widely used in Australia. A further contrast to tbe above studies was seen wjtb tbe results obtained for IgG anti-65 kDa HSP, where no differences from controls were demonstrated. Tbe only significant results witb tbe 70 kDa antigen were with IgG antibodies and RA. Here the results differ from Tsoulfa et a!., who observed no difference between RA, SLE or tuberculosis patients with regard to antibodies to tbis antigen, and concluded tbat there appeared to be something particular about the IgG response to 65 kDa antigen in RA when compared with controls or the 70 kDa responses.^ However, tbe data reported here clearly do not support tbat view. Indeed, other groups have been unable to confirm reactivity to the 65 kDa or tbe 70 kDa antigen in RA, although Jarjour et al. did demonstrate sucb reactivity in a wide range of other rheumatic and inflammatory diseases.'^ It is very unlikely that tbis is due to technical variation between laboratories as most groups have bad little trouble identifying anti-HSP antibodies. However, the antibodies are detected in different groups of patients. Otber Australian workers bave only found increased activity to 65 kDa antigen in individuals who bad experienced mycobacterial infections such as
J. Panchapakesan et al.
298
L22
A
B
C
L7
D
E
F
Fig. 2. Western blotting of separated 65 kDa, 70 kDa and BCG sonicate. Two blots are illustrated, stained with either L22 (anti-65 kDa monoclonal antibody) or L7. (a) 65 kDa HSP; (li) 70 kDa HSP; (c) BCG sonicate; (d) BCG sonicate; (e) 70 kDa HSP; (f) 65 kDa HSP. The 65 kDa HSF shows two bands, the one of lower molecular weigbt representing a minor degree of breakdown. Table 1. Median values of binding for antibodies to 65 kDa. 70 kDa, HSP and BCG in RA, SLE and control sera for both IgG and IgM isotypes. Anti-65 kDa
Anti-70 kDa
Anti-BCG
Controls
IgG IgM
0.20(0.19-0.30) 0.24 (0.25-0.32)
0.20(0.19-0.29) 0.25 (0.23-0.30)
0.29 (0.28-0.34) 0.26 (0.24-0.29)
RA
IgG IgM
0.15 (0.18-0.27) 0.39(0.41-0.55)
0.12(0.12-0.18) 0.28(0.31-0.50)
0.29 (0.28-0.34) 0.38 (0.37-0.45)
SLE
IgG IgM
0.24 (0.24-0.49) 0.51 (0.44-0.56)
0.17(0.16-0.24) 0.31 (0.28-0.41)
0.41 (0.37-0.52) 0.37(0.31-0.41)
Values are expressed as anopticaldensity ratio of percentage of the positive control (designated as 1). The 95% confidence limits are shown in brackets. tuberculosis or leprosy (M. Rowley, pers. comm.). Tsoulfa et al. '** and McLean et al. also reported that IgA anti-65 kDa HSP was elevated in RA.'*^ The present study could not
confirm these findings in a series of preliminary experiments (data not shown), The idea of immune responses to these highly conserved HSP participating in auto-
Antibodies to heat shock proteins in RA and SLE Table 2. Summary of differences (P values) becween RA and SLE binding of IGG and IGM antibodies to the three antigens.
3.
Anti-65 kDa Anti-70 kDa Anti-BCG 4. RA IpG 0.42 IgM 0.0003
0.0009 0.25
2.8 0.0003
SLE IgG 0.27 IgM 0.0003
1.15 0.11
0.037 0.018
5.
Corrected P values are shown for each of the three antigens for both IgG and IgM antibodies. 6. immune disease by molecular mimicry, including the demonstration of the autologous 65 kDa analogue in inflamed rheumatoid synovium^** and the binding of monoclonal antibodies to this protein in inflamed joints, is attractive. Despite this, the highly variable nature of the results, with different groups finding either no increased binding or binding with different isotypes, is consistent with the conclusion that humoral responses to 65 kDa HSP are extremely variable in both RA and SLE. The groups of patients studied by different workers appear, as best can be judged, to be relatively comparable, and this variability makes it unlikely that the antibodies play any specific role in the pathogenesis of disease. These results, together with negative results reported for T cell reactivity,*^'^' must lead to a re-evaluation of the whole idea of molecular mimicry to HSP and the development of autoimmune disease.
7.
8.
9.
10.
Acknovrledgements tl. We would like to thank Dr J. R. York, Director of the Royal Prince Alfred Hospital Combined Arthritis Unit. The study was supported by the NH&MRC. 12. References 1. Young, D. B., Ivanyi, J., Cox, J. H. and Lamb, J. R. 1987. The 65 kDa antigen of mycobacteria—A common bacterial protein?/mmMfio/. Today 8: 215-219. 2. Shoenfeld, Y. and Isenberg, D. A. 1988. Myco-
13.
299
bacteria and autoimmunity. Immunol Today 9: 178-181. Winfield, J. B. 1989. Stress proteins, arthritis and autoimmunity. Arthr. Rheum. 32: 14971504. Young, D. B. 1990. The immune response to mycobacterial heat shock proteins. Autoimmunity 7: 237-244. van Eden, W., Hogervosst, E. J. M., Hcnsen, E. J., van der Zee, R., van Embden J. D. A. and Cohen, I. R. 1989. A cartilage-mimicking T-cell epitope on a 65 K mycobacterial heatshock protein: Adjuvant arthritis as a model for human rheumatoid arthritis. Curr. Top. Microbiol. Immunol. 145: 27-43. Res, P. C. M., Schaar, C. G., Breedveld, F. C . van Bden, W., van Embden, J. D. A., Cohen, I. R. and De Vries, R. R. P. 1988. Synovial fluid T cell reactivity against 65 kDa heat shock protein of mycobacteria in early chronic arthritis. Lancet ii: 478-480. Holoshitz, J., Kosek, J., Sibley, R., Brown, D. A. and Strober, S. 1991. T lymphocytesynovial fibroblast interactions induced by mycobacterial proteins in rheumatoid arthritis. Arthr. Rheum. 34: 679-686. Pope, R. M., Lovis, R. M. and Gupta, R. S. 1992. Activation of synovial fluid T lymphocytes by 60 kDa heat shock proteins in patients with inflammatory synovitis. Arlhr. Rheum. 35: 43-48. Tsoulfa, G., Rook, G. A. W., Bahr. G. M. et al. 1989. Elevated IgG antibody levels to the mycobacterial 65 kDa heat shock protein are characteristic of patients with rheumatoid arthritis. Scand. J. Immunol. 30: 519-527. Tsoulfa, G., Rook, G. A. W., van Embden, J. D. A. et al. 1989. Raised serum IgG and IgA antibodies to mycobacterial antigens in rheumatoid arthritis. Ann. Rheum. Dis. 48: 118123. Minota, S., Cameron, B., Welch, W. J. and Winfield J. B. 1988. Autoantibodies to the constitutive 73 kDa member of the hsp70 family of heat shock proteins in systemic lupus erythematosus./ Exp. Med. 168: 1475-1480. Jarjour, W. N., Jeffries, B. D., Davis, J. S., Welch, W.J., Mimura, T. and Winfield, J. B. 1991. Autoantibodies to human stress proteins. A survey of various rheumatic and other inflammatory diseases. Arthr. Rheum. 34: 11331138. Arnett, F. C , Edworthy, S. M., Block, D. A. et al. 1988. The American Rheumatism Association 1987 revised criteria for the classification
300
14.
15. 16.
17. 18.
J. Panchapakesan et al. of rheumatoid arthritis. Arthr. Rheum. 31: 315-324. Tan,E.M.,Cohen,A.S.,Fries,J.F.erj/. 1982. The 1982 revised criteria for the classification of systemic lupus erythematosus. Arlhr. Rheum. 25:1271-1277. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685. Towbin, H., Staehelin T. and Gordon, J. 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc. Natl Acad. Sci. USA 76: 4350-4354. Glantz, S. 1981. Primer of Biostatistics. McGraw-Hill, New York. pp. 87-88. McLean, I. L,, Archer, J. R., Cawley, M. I. D., Pegley F. S., Kidd, B. L. and Thompson, P. W. 1990. Specific antibody response to the mycobacterial 65 kDa stress protein in ankylosing
spondylitis and rheumatoid arthritis. Br. J. Rheum. 29: 426-429. 19. Karlsson-Parra, A., Soderstrom, K., Perm. M., Ivanyi. J., Kiessling. R. and Klareskog, L. 1990. Presence of human 65 kDa heat shock protein (hsp) in inflamed joints and subcutaneous nodules of RA patients. Scand. J. Immunol. 31: 283-288. 20. de Graeff-Meeder, E. R.. Voorhorst, M.. van Eden, W. et al. 1990. Antibodies to the mycobacterial 65 kd heat shock protein are reactive with synovial tissue of adjuvant arthritic rats and patients with rheumatoid arthritis and osteoarthritis. Am.J. Path. 137: 1013-1017. 21. Crick, F. D. and Gatenby, P. A. 1992. Limiting dilution analysis of T cell reactivity to mycobacterial antigens in peripheral blood and synovium from rheumatoid arthritis patients. Clin. Exp. Immunol. 88: 424-429.
