1. ¹Department of Microbiology and Immunology, Medical University of South Carolina, Charleston, SC, USA
2. ²Third Department of Internal Medicine, Semmelweis University, Budapest, Hungary
3. ³Department of Microbiology and Immunology, University of Tampere, Tampere University Hospital, Tampere, Finland

Correspondence: Dr JP Pandey, Department of Microbiology and Immunology, Medical University of South Carolina, Charleston, SC 29425-2230, USA. E-mail: pandeyj@musc.edu

Received 12 May 2003; Revised 27 August 2003; Accepted 29 August 2003.

Abstract

Immunoglobulin GM and KM genes have been associated with antibody responses to a variety of antigens. A promoter-region polymorphism of interleukin-6 (IL-6) gene (-174 G/C) has been shown to be associated with antibody responses to heat-shock proteins (hsp) 60 and hsp65. To examine the possible epistatic effects of these unlinked genetic systems on the autoimmune responses to hsp60 and hsp65, 176 healthy Caucasian subjects from Finland were genotyped for several allelic determinants of GM, KM, and IL-6 genes by PCR-RFLP methods. IgG antibodies to hsp60 and hsp65 were measured by an ELISA. Significant interactive effects of GM f,z and IL-6-174 genotypes were noted for both anti-hsp60 (P=0.002) and anti-hsp65 (P=0.038) antibody levels. Since these autoantibodies have been implicated in susceptibility to coronary heart disease and carotid atherosclerosis, the associations reported here might be relevant to the etiology of these diseases.

Immunoglobulin (Ig) GM and KM genes—located on human chromosomes 14 and 2, respectively—encode for antigenic determinants, called allotypes, on Ig polypeptide chains. Allelic variation at the GM locus results in amino-acid substitutions in the constant regions of 1, 2, and 3 chains; KM alleles correspond to amino-acid substitutions at three sites on the constant region of chains. Frequencies of GM and, to a lesser extent, KM allotypes vary widely among racial groups.^1,2 Linkage disequilibrium between particular alleles of the GM system is almost absolute. Although the mechanisms underlying the evolution of GM and KM polymorphisms are not understood, the marked differences in gene frequencies among races and strong linkage disequilibrium within a race, both suggest that differential selection over many generations may have contributed to the maintenance of polymorphism at these loci. One likely selective force may be association between GM and KM allotypes and specific antibody responses to autoantigens or pathogens, resulting in differential immunity to autoimmune or infectious diseases. Results from several studies support this contention.^3,4,5,6

The aim of the present investigation was to determine whether allelic variation at the GM and KM loci is associated with autoantibody responsiveness to hsp60 and hsp65 epitopes. In addition, we wished to determine whether a promoter-region polymorphism of the IL-6 gene, which has been shown to be associated with the height of antibody responses to hsp60 and hsp65,⁷ epistatically interacts with particular GM or KM alleles in inducing autoantibodies to these proteins. The results reported in this article show that particular GM genotypes have main as well as interactive (with the IL-6 gene) effects on the production of autoantibodies to hsp60 and hsp65 epitopes.

Blood samples of 176 healthy adult males were obtained from the Finnish Red Cross Blood Transfusion Center, Tampere. The mean age of the subjects was 44 (range 21–63) years. This study was approved by the Ethical Committee of the Finnish Red Cross. DNA samples were typed for G3 M b/5 and g/21 alleles by a previously described PCR-RFLP method.⁸ For the determination of G1 M f/3 and z/17 alleles, the CH1 region of 1 chain was amplified by PCR, using the primers described previously,⁹ and the purified double-stranded PCR product was subjected to an automatic DNA sequencing on an ABI PRISM 377. chain determinants KM 1 and 3 were characterized by a PCR-RFLP technique using the following primers: 5' TAG GGG GAA GCT AGG AAG AAA 3' and 5' AAA AAG GGT CAG AGG CCA AA 3'. After digestion of the amplified product with the restriction enzyme Acc1, the following products corresponding to the two alleles were detected: KM 1 (a 538 bp fragment), KM 3 (a 390 and a 148 bp fragment). A previously described PCR-RFLP method was used for the analysis of -174 G/C polymorphism of the IL-6 gene.¹⁰ Briefly, oligonucleotides 5'TGACTTCAGCTTTACTCTTGT3' and 5'CTGATTGGAAACCTTATTAAG3' were used as primers in the PCR. Conditions used were: five cycles of 96°C for 9 min, 55°C for 1 min and 72°C for 3 min, followed by 30 cycles of 95°C for 1 min, 55°C for 1 min and 72°C for 1 min, with a final incubation at 72°C for 10 min. The PCR-products were digested with 5 U of NlaIII at 37°C for 24 h and analyzed on a 9% polyacrylamide gel stained with ethidium bromide. This generated fragments of 119 and 49 bp for allele 2 and a single fragment of 168 bp for allele 1.

IgG antibodies to hsp60 and hsp65 epitopes [recombinant human hsp60, recombinant M. bovis hsp65 (batch MA14) Lionex GmbH, Braunschweig, Germany)] were assessed by an ELISA described elsewhere.¹¹ In brief, plates were coated with 0.05 g/well human hsp60 or M. bovis hsp65. After washing and blocking (PBS, 0.5% gelatin) the wells were incubated with 50 l of serum samples diluted 1 : 100 in PBS containing 0.5% gelatin and 0.05% Tween 20. Binding of anti-hsp antibodies was determined using -chain specific anti-human IgG peroxidase labelled antibodies (Sigma, St Louis, USA) and o-phenylene-diamine (Sigma) detection system. The optical density was measured at 490 nm (reference at 620 nm) and the means of duplicate wells were calculated. A serial dilution of a high-positive human serum was used as standard. Data obtained as optical density values were calculated to arbitrary unit/ml values related to this standard. Anti-hsp65 and anti-hsp60 antibodies were determined in two different assays, each with its own high-positive human serum standard. (All subjects were characterized for the presence of anti-hsp60 and anti–hsp65 antibodies. However, for technical reasons, six individuals could not be typed for certain alleles. Thus, both genotype and antibody data were available for 170 subjects for statistical analyses.)

Since the anti-hsp antibody levels were not normally distributed (Kolmogorov–Smirnov test, P<0.0001), nonparametric tests were used to determine the influence of genotypes on the antibody levels: the Mann–Whitney test for determining the influence of GM and KM genotypes (Table 1), the Kruskal–Wallis analysis of variance for measuring the global interactive effect of GM and IL-6 genotypes, and Dunn's post-test for comparing the influence of specific GM and IL-6 genotype combinations (Table 2).

Table 1 - Anti-Hsp antibody levels according to the GM z or g allele carrier status.

Full table

Table 2 - Anti-Hsp antibody levels according to IL-6 -174 C and GM z carrier status.

Full table

There was a significant correlation between the levels of autoantibodies to hsp60 and hsp65 (Spearman correlation coefficient=0.66, P<0.0001). Table 1 presents the anti-hsp60 and anti-hsp65 antibody levels stratified according to the GM z or g allele carrier status. The carriers of z or g allele had lower levels of both antibodies than noncarriers. This difference, however, was statistically significant only for the anti-hsp60 antibody levels and the z carrier status (P=0.023). In Caucasians, there is significant linkage disequilibrium between GM z and g alleles, which may, at least in part, be responsible for the similarity in antibody levels for the carriers of these alleles. No significant associations were found between KM genotypes and anti-hsp antibody levels (data not shown). Table 2 presents anti-hsp antibody levels stratified according to the GM z and IL-6 -174 C carrier status. Significant interactive effects of GM and IL-6 genotypes were noted for both anti-hsp60 (P=0.002) and anti-hsp65 (P=0.038) antibody levels. Subjects positive for IL-6 -174 C and GM z alleles appeared to have the lowest levels of both antibodies. Compared to these subjects, those who lacked the IL-6 -174 C allele, with or without the GM z allele, had significantly higher levels of anti-hsp60 antibodies (P<0.05). The levels of anti-hsp65 antibodies corresponding to these genotypes were not significantly different. No other significant associations were found. Also, there was no significant population association between IL-6 and GM or KM alleles (data not shown).

The most significant finding of the present investigation is that allelic determinants at IL-6 -174 and GM f,z loci epistatically interact to have a pronounced effect on the levels of anti-hsp60 and (to a lesser extent) anti-hsp65 antibodies. We have reported earlier⁷ that noncarriers of the C allele (subjects homozygous for the G allele) at the IL-6 -174 locus had almost two-fold higher anti-hsp60 antibody levels than those who carried this allele (44.1 vs 23.3 arbitrary units). At the GM f,z locus, the results of the present investigation show that noncarriers of the z allele have significantly higher levels of anti-hsp60 antibodies than those who carried this allele (33.2 vs 22.7 arbitrary units). And jointly, the high-responder genotypes at the two loci have even more pronounced influence on anti-hsp60 antibody levels (55.6 vs 21.3 arbitrary units).

IL-6 -174 and GM f,z loci could themselves—via their protein products—influence the production of anti-hsp60 autoantibodies. Although we did not find an association between IL-6 genotypes and IL-6 plasma levels,⁷ another study reported the circulating IL-6 concentrations to be approximately twice as high in subjects homozygous for the G allele as compared to those who were homozygous for the C allele.¹⁰ A dose-dependent influence of IL-6 has been shown for the production of anti-DNA topoisomerase I autoantibodies.¹² IL-6 may have a similar influence on the production of anti-hsp60 autoantibodies. GM f,z loci, too, have been implicated in autoimmune responses.^13,14 Thus, B cells carrying particular f,z allotypes on their Ig receptors may be more efficient in the uptake, processing, and subsequent presentation of hsp60-derived peptides to the collaborating T cells, resulting in anti-hsp60 autoantibody production. Although GM markers are constant-region determinants, they can influence immune responsiveness, normally associated with the variable-region, in several ways: direct contribution to the formation of idiotypic determinants, modulation of antibody binding affinity, linkage disequilibrium with alleles coding for the variable-region epitopes.^15,16,17 The CH1 domain—where allelic determinants GM f and z are located—has been shown to modulate the kinetic competence of antigen binding sites.¹⁸

Alternatively, the observed interactive effect of these loci on the production of anti-hsp60 antibodies may be due to other, as-yet-unidentified, autoimmune response genes, which are in linkage disequilibrium with particular alleles at the IL-6 -174 and GM f,z loci. Transmission disequilibrium tests may aid in examining this possibility.¹⁹ Autoantibodies to hsp60 and hsp65 have been implicated in susceptibility to coronary heart disease and carotid atherosclerosis, respectively.^20,21,22 In view of the results reported here, studies to determine whether the IL-6 -174 and GM f,z loci epistatically interact to contribute to the risk of these diseases are warranted.

Finally, although the associations reported here are strong and can, at least partially, be explained by the known immunological properties of IL-6 and GM allotypes; nevertheless, they must be followed by confirmation in an independent study population to be of wider significance. It is especially important to examine the role of GM determinants in autoimmune responses to hsp60 and hsp65 in other ethnic groups, as there is significant difference in the qualitative and quantitative distribution of these markers among races,^1,2 and therefore, different allelic combinations are likely to be implicated in non-Caucasian races. To our knowledge, this is the first report implicating immunoglobulin GM genes in autoimmunity to heat-shock proteins.