MARJA PERTOVAARA, MIKKO HURME, JAAKKO ANTONEN, AMOS PASTERNACK, and JANARDAN P. PANDEY

ABSTRACT.

Objective. To investigate whether polymorphism of immunoglobulin (Ig) genes affects susceptibility to or severity of primary Sjögren's syndrome (pSS).

Methods. Ig gene kappa (KM) and gamma (GM) polymorphisms were analyzed by a polymerase chain reaction restriction fragment length polymorphism (PCR-RFLP) based method in 65 Finnish Caucasian patients with pSS and in 66 healthy controls matched for sex, ethnic origin, and area of residence. Clinical and immunological data on the pSS patients were analyzed in relation to Ig genotypes.

Results. The genotype frequencies of Ig KM and GM genes did not differ between pSS patients and controls. Anti-SSB antibodies were encountered significantly more frequently in pSS patients carrying the KM1 allele than in those without (100% vs 48%, p = 0.016). The pSS patients with the KM1 allele had several signs of immunologically active disease: they had significantly higher erythrocyte sedimentation rate, serum IgA, serum ß₂-microglobulin (ß₂-m), and plasma IgG1 concentrations than patients without this allele. The pSS patients carrying the GM z allele had a milder form of pSS than those without this determinant. They had less severe labial salivary gland histological findings (grade 3-4 in 60% vs 93%, p = 0.004) and lower plasma IgG3 and serum ß₂-m concentrations than those without GM z allele.

Conclusions. Ig KM and GM genes do not contribute to susceptibility to pSS. The Ig KM1 allele is associated with several markers of immunologically active disease, whereas the Ig GM z allele is associated with milder pSS. (J Rheumatol 2004;31:2175-80)

Key Indexing Terms:

IMMUNOGLOBULIN
GENE POLYMORPHISM
PRIMARY SJÖGREN'S SYNDROME

Supported in part by funds from the US Department of Energy cooperative agreement DE-FC09-02CH11109 and by grants from the Medical Research Fund of Tampere University Hospital, Tampere, Finland, and the Scandinavian Rheumatology Research Foundation.

M. Pertovaara, MD, PhD, Department of Internal Medicine, Tampere University Hospital; M. Hurme, MD, PhD, Professor, Centre for Laboratory Medicine, Tampere University Hospital and the University of Tampere Medical School; J. Antonen MD, PhD; A. Pasternack MD, PhD, Professor, Department of Internal Medicine, Tampere University Hospital and University of Tampere Medical School; J.P. Pandey, PhD, Professor, Department of Microbiology and Immunology, Medical University of South Carolina.

Address reprint requests to Dr. M. Pertovaara, Department of Internal Medicine, Section of Rheumatology, Tampere University Hospital, P.O. Box 2000, FIN-33521 Tampere, Finland. E-mail: marja.pertovaara@pshp.fi

Submitted February 24, 2004; revision accepted June 21, 2004.

Primary Sjögren's syndrome (pSS) is a chronic autoimmune exocrinopathy causing symptoms of dry eyes and mouth but also various extraglandular symptoms. Hypergammaglobulinemia and abundant autoantibody production, including rheumatoid factor (RF), antinuclear antibody (ANA), and anti-SSA and anti-SSB antibodies are characteristic features of the disease. The etiology of pSS is largely unknown, but an interaction between genetic and environmental factors is thought to lead to autoimmunity in SS¹. HLA class II genes are known to affect susceptibility and autoantibody production in pSS² , but non-HLA genes, e.g., the genes encoding for cytokines, have also been found to influence susceptibility^3,4, clinical presentation^5,6, and autoantibody production in pSS⁷.

Immunoglobulin (Ig) genes are polymorphic, and their polymorphism has been found to have an effect on various autoimmune and infectious diseases. The Ig kappa (KM) and gamma (GM) genes are located on human chromosomes 2 and 14, respectively. Allelic variation at the KM and GM loci results in amino acid differences in the constant regions of the kappa and gamma 1, 2, and 3 polypeptide chains, respectively⁸. The role of constant region allotypes in antibody response could be explained by possible linkage disequilibrium with particular variable region determinants or by effects on antibody affinity or idiotype formation⁹. An association between anti-SSB antibodies and serologically detected Ig KM allotypes has been observed¹⁰. However, it was recently reported that KM light chain alleles detected by a polymerase chain reaction (PCR) and restriction fragment length polymorphism (RFLP)-based method were not associated with pSS in an Australian population; moreover, the KM gene polymorphism was not associated with anti-SSA or anti-SSB antibody production¹¹. The association of KM or GM allotypes with the clinical presentation and severity of pSS has not been previously investigated.

Our aim was to establish in a well-characterized group of Finnish Caucasian patients with pSS whether polymorphism of the Ig genes (KM and GM) affects susceptibility, clinical presentation, or autoantibody production in pSS.

MATERIALS AND METHODS

Patients. All patients fulfilling 3 or more modified California criteria for pSS¹² (salivary flow determinations were not performed, histological findings were graded on the Chisholm-Mason scale¹³, grades 3 and 4 being regarded as diagnostic) were selected from the records of patients with sicca symptoms examined in the Department of Internal Medicine, Section of Rheumatology, at Tampere University Hospital, Finland, during the years 1977 to 1992 (n = 111). Those alive were invited by letter to attend for gene polymorphism determinations. Samples for genotyping were collected after obtaining informed consent from 65 pSS patients (63 female, 2 male, mean age 60 ± 12 years). Sixty-two of the patients fulfilled the revised American-European consensus group criteria for SS¹⁴. Demographic and clinical data are presented in Table 1.

Clinical methods. A careful clinical examination together with an in-depth interview covering previous and concurrent diseases and duration of pSS had recently been conducted with these patients¹⁵. Special emphasis was laid on possible extraglandular symptoms of pSS (dermatological, endocrine, gastrointestinal, lymphoproliferative, musculoskeletal, neurological, renal, respiratory, and vascular symptoms). Purpura was defined as a history of typical episodic palpable purpura lesions in the lower limbs or skin biopsy histology. Lymphadenopathy was defined as lymph node enlargement so persistent as to have indicated a nodal biopsy. Arthritis was defined as articular swelling observed by a clinician. Peripheral and central neurological symptoms were recorded from the history given by the patients as well as from data on possible neurological investigations from case histories. Diagnosis of pulmonary fibrosis was based on findings in chest radiographs; the diagnosis of alveolitis had been established by thorough investigations in a pulmonary unit. Urinary total protein excretions had been measured¹⁵ and proteinuria was defined as urinary total protein excretion ≥ 0.15 g/24 h. Labial salivary gland biopsies had been taken from all 65 patients and the histological grade according to Chisholm-Mason scale was 0 in one, 1 in 2, 2 in 14, 3 in 15, and 4 in 33 of the patients.

Controls. Sixty-six healthy Finnish Red Cross Transfusion Service blood donors matched for sex (64 female, 2 male), ethnic origin (Finnish Caucasian), and area of residence (Tampere, Finland) served as a control group for DNA studies of pSS patients. The mean age of the control subjects was 53 ± 7 years.

GM and KM genotyping. DNA samples were typed for G3M b/5 and g/21 alleles using a PCR-RFLP method as described¹⁶. For the determination of G1M f/3 and z/17 alleles, the CH1 region of the g1 chain was amplified by PCR, using primers described by Balbin, et al¹⁷, and purified double-stranded PCR product was subjected to automatic DNA sequencing on an ABI PRISM 377. k chain determinants KM1 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 (538 bp) with the restriction enzyme Acc1, products corresponding to the 3 genotypes were detected: KM1 (538 bp), KM1/3 (538 bp, 390 bp, 148 bp), and KM3 (390 bp, 148 bp).

The Ig KM, GM (b,g), and GM (f,z) genotypes were successfully determined in 62, 59, and 62 patients with pSS, respectively. KM and GM (f,z) genotyping was successfully performed in all 66 control subjects, and GM (b,g) genotyping in 63 of them.

Statistical analysis. Mann-Whitney U-test and chi-square test were used in comparisons of continuous and dichotomous variables, respectively. Findings were considered statistically significant at p < 0.05. P values were not corrected for multiple comparisons. Statistical analyses were performed with SPSS 10.1 for Windows.

Ethical approval. The study protocol was approved by the Ethical Committee of Tampere University Hospital.

RESULTS

Table 1. Demographic, clinical, and immunological characteristics of 65 patients with pSS. The values indicate numbers of patients, unless otherwise indicated, and the numbers in parentheses are percentages of cases.

Ig KM gene polymorphism. The Ig KM genotype frequencies in pSS patients and controls were not different (Table 2). The pSS patients carrying the rare KM1 allele yielded several laboratory findings suggestive of a more active form of the disease (Table 3). Anti-SSB antibodies were encountered significantly more frequently in pSS patients with the KM1 allele than in patients without this allele (100% vs 48%, p = 0.016). The titer of anti-SSB antibodies was higher in KM1 carriers compared to non-carriers, albeit not significantly (Table 3). The KM1 allele carriers had significantly higher mean erythrocyte sedimentation rate, serum IgA, serum beta-2 microglobulin (ß₂-m), and plasma IgG1 subclass concentrations than patients not carrying this allele.

Table 2. The frequencies of immunoglobulin kappa (KM) and gamma (GM) b,g and f,z genotypes in patients with pSS and controls. The numbers in parentheses are percentages of cases. Statistical analysis was performed using the chi-square test.

Table 3. Demographic, clinical, and immunological characteristics of patients with pSS carrying or not carrying the Ig kappa (KM) 1 allele. The values indicate numbers of patients, unless otherwise indicated, and the numbers in parentheses are percentages of cases. Statistical analysis was performed using the Mann-Whitney U-test or chi-square test. P values were not corrected for multiple comparisons.

No differences were observed in the frequencies of such extraglandular manifestations of pSS as a history of arthralgia, arthritis, salivary gland swelling, Raynaud's symptoms, purpura, lymphadenopathy, pleuritis, alveolitis or pulmonary fibrosis, or peripheral or central nervous system symptoms (data not shown). Neither the histological grade in labial salivary gland biopsies nor the number of diagnostic criteria for pSS differed between pSS patients with or without KM1 allele (Table 3).

Ig GM gene polymorphism. The Ig GM genotype frequencies in pSS patients and controls did not differ (Table 2). A strong, but not absolute, linkage between the GM b,g and GM f,z allotypes was observed in both pSS patients and controls. The GM g allele occurred more frequently in both controls and pSS patients carrying the GM z allele compared to non-carriers of z (94% vs 0, p < 0.0001, and 76% vs 16%, p < 0.0001, in the respective groups).

Patients with pSS carrying the GM allele z had findings suggestive of a milder form of the disease than patients without this determinant. They had less severe labial salivary gland histological findings (grade 3-4 in 60% vs 93%, p = 0.004) and lower plasma IgG3 and serum ß₂-m concentrations than those without GM z allele (Table 4).

Table 4. Demographic, clinical, and immunological characteristics of patients with pSS carrying or not carrying the Ig gamma (GM) z allele. Values indicate numbers of patients, unless otherwise indicated; numbers in parentheses are percentages of cases. Statistical analysis was by Mann-Whitney U-test or chi-square test. P values were not corrected for multiple comparisons.

The pSS patients with GM allele g had proteinuria significantly more frequently than those not carrying allele g. They had lower plasma IgG3 concentrations than those without this allele (Table 5).

Table 5. Demographic, clinical, and immunological characteristics of patients with pSS carrying or not carrying the Ig gamma (GM) g allele.Values indicate numbers of patients, unless otherwise indicated; numbers in parentheses are percentages of cases. Statistical analysis was by Mann-Whitney U-test or chi-square test. P values were not corrected for multiple comparisons.

Pulmonary fibrosis or alveolitis occurred less frequently in both Ig GM z carriers (3% vs 22%, p = 0.017) and Ig GM g carriers (3% vs 21%, p = 0.049) compared with respective non-carriers. Purpura occurred less frequently (10% vs 31%, p = 0.045) in Ig GM g carriers compared with g non-carriers. No differences were observed in the frequencies of other extraglandular manifestations of pSS between Ig GM z or g carriers and corresponding non-carriers (data not shown).

DISCUSSION

The main findings in our study were that Ig KM genes contribute to the severity of pSS and to anti-SSB antibody production and that the GM z allele seems to be a protective factor in respect of pSS severity. Ig gene polymorphism does not affect susceptibility to pSS in Finnish Caucasian patients. Previously, the association of KM or GM allotypes with the clinical presentation and severity of pSS has not been investigated, except for the presence of anti-SSA and anti-SSB antibodies.

We found anti-SSB antibodies significantly more frequently in pSS patients with the KM1 allele than in patients without this allele. The titer of anti-SSB antibodies also tended to be higher in KM1 carriers compared to non-carriers. This finding is analogous to original findings of an association of serologically detected KM allotypes with anti-SSB antibodies¹⁰, but is in contrast to a recent report from Australia, in which KM alleles were not associated with anti-SSA or anti-SSB antibody responses in patients with pSS¹¹. The difference between those results and ours is probably explained by different genetic backgrounds of study populations. Frequencies of GM and to a lesser extent KM allotypes vary widely among racial groups. Within a race, GM allotypes show strong linkage disequilibrium¹⁸. KM and GM markers are located on the constant region of the Ig, but there is a growing body of evidence for involvement of these regions in antibody specificity usually associated with the variable region of the Ig molecule. The underlying mechanisms are not completely understood, but there are several well-documented examples of constant domains influencing the specificity of antibodies to various antigens^19-21. Previously, associations have been found with KM alleles and production of various autoantibodies such as anti-DNA antibodies in systemic lupus erythematosus (SLE)²², anti-fibrillin antibodies in scleroderma²³, and anti-Mi antibodies in inflammatory myositis²⁴. Ig GM f,z genotypes and interleukin 6 –174 genotypes have recently been shown to exert an epistatic effect on the production of autoantibodies to heat-shock protein²⁵.

The pSS patients with the KM1 allele had several other laboratory findings uniformly suggesting immunologically active disease: they had significantly higher mean erythrocyte sedimentation rate, serum IgA, and serum ß₂-m concentrations than patients not carrying the KM1 allele. Increased levels of serum ß₂-m have been associated with lymphoproliferative and renal complications of pSS^6,26,27. The association of serum ß₂-m concentrations with KM or GM allotypes in pSS has not previously been investigated. However, in hematological malignancies the phenotype facilitating ß₂-m shedding from cell surface has been found to be independent of specific IgG heavy chain allotypes²⁸.

IgG1 levels were associated with KM allotypes, and IgG3 levels were lower in patients with GM z or GM g than in patients with GM f or GM b, respectively. These findings are in agreement with previously observed strong associations of IgG subclasses with Ig gene loci^29,30.

For the first time, GM z was found to constitute a protective factor with respect to the severity of pSS. The pSS patients carrying GM allele z had less severe labial salivary gland histological findings and lower serum ß₂-m concentrations than those without this allele.

Pulmonary fibrosis or alveolitis occurred less frequently in both Ig GM z carriers and Ig GM g carriers compared with the corresponding non-carriers. Purpura occurred less frequently in Ig GM g carriers compared with g non-carriers. Patient numbers in these subgroups were very small and therefore reliable conclusions cannot be drawn. It is nonetheless of interest that in a family study with 12 subjects with clinical evidence of pulmonary fibrosis, all affected siblings were carrying the Ig haplotype GM1, and based on this observation a dominantly inherited gene located on chromosome 14 close to the loci of GM was suggested to be associated with familial fibrosing alveolitis³¹.

The pSS patients with GM allele g had proteinuria significantly more frequently than those not carrying allele g. In an earlier study no difference was observed in the distribution of GM and KM alleles among patients with IgA nephropathy and controls³² or between SLE patients with or without nephritis³³, but contrasting results regarding GM allotypes and renal manifestations in patients with SLE have been found³⁴.

Neither the Ig KM nor GM genotype frequencies in pSS patients and controls differed. This is similar to findings reported in the recent Australian study¹¹. No association with disease susceptibility and KM or GM allotypes has been found in rheumatoid arthritis^35,36 or in scleroderma²³. KM genotypes have been found to affect susceptibility to SLE, but conflicting associations have been reported^22,37. A strong association between the KM1 allele and susceptibility to SLE was observed in one study²² but in contrast, in a population-based case-control study the KM3,3 was associated with an increased risk, whereas KM1,3 with a lower relative risk of SLE in Caucasians³⁷.

Recently HLA class II markers were reported to confer genetic susceptibility to Sjögren's syndrome only in patients with anti-SSA and anti-SSB antibodies³⁸. Hence, the role of KM allotypes in the pathogenesis of pSS appears to be similar to that of HLA antigens; neither genetic system is strongly associated with the disease itself, but both contribute significantly to the generation of particular autoimmune responses.

In conclusion, Ig allotypes were found to be responsible for many determinants of the clinical presentation and severity of pSS as well as for anti-SSB antibody production, but did not affect disease susceptibility.

ACKNOWLEDGMENTS

We thank Mr. Keith Rocca for expert technical assistance.

REFERENCES

Search PubMed for:

2. Pease CT, Charles PJ, Shattles W, Markwick J, Maini RN. Serological and immunogenetic markers of extraglandular primary Sjögren's syndrome. Br J Rheumatol 1993;32:574-7. [MEDLINE]

3. Hulkkonen J, Pertovaara M, Antonen J, Lahdenpohja N, Pasternack A, Hurme M. Genetic associations between IL-10 promoter region polymorphism and primary Sjögren's syndrome. Arthritis Rheum 2001;44:176-9. [MEDLINE]

4. Font J, Carcia-Carrasco M, Ramos-Casals M, et al. The role of interleukin-10 promoter polymorphism in the expression of primary Sjögren's syndrome. Rheumatology 2002;41:1025-30. [MEDLINE]

5. Hulkkonen J, Pertovaara M, Antonen J, Pasternack A, Hurme M. Elevated interleukin-6 plasma levels are regulated by the promoter region polymorphism of IL-6 gene in primary Sjögren's syndrome (pSS) and correlate with clinical manifestations of the disease. Rheumatology 2001;40:656-61. [MEDLINE]

6. Pertovaara M, Hulkkonen J, Antonen J, et al. Polymorphism of the tumour necrosis factor-alpha gene at position -308 and renal manifestations of primary Sjögren's syndrome. Rheumatology 2004;43:106-7. [MEDLINE]

7. Gottenberg JE, Busson M, Loiseau P, et al. Association of transforming growth factor ß1 and tumor necrosis factor a polymorphisms with anti-SSB/La antibody secretion in patients with primary Sjögren's syndrome. Arthritis Rheum 2004;50:570-80. [MEDLINE]

8. Dugoujon J, Cambon-Thomsen A. Immunoglobulin allotypes (Gm and Km) and their interactions with HLA-antigens in autoimmune diseases: a review. Autoimmunity 1995;22:245-60. [MEDLINE]

9. Pandey JP. Immunoglobulin GM and KM allotypes and vaccine immunity. Vaccine 2001;19:613-7. [MEDLINE]

10. Whittingham S, Propert D, Mackay I. A strong association between the antinuclear antibody anti-La (SS-B) and the kappa light chain allotype Km(1). Immunogenetics 1984;19:295-9. [MEDLINE]

11. Downie-Doyle S, Lester S, Bardy P, Gordon T, Rischmueller M, Pile K. Immunoglobulin kappa light chain gene alleles are not associated with primary Sjögren's syndrome. Genes Immunity 2002; Suppl 1:S63-S65.

12. Fox RI, Robinson CA, Curd JG, Kozin F, Howell FV. Sjögren's syndrome. Proposed criteria for classification. Arthritis Rheum 1986;29:577-85. [MEDLINE]

13. Chishom DM, Mason DK. Labial salivary gland biopsy in Sjögren's disease. J Clin Pathol 1968;21:656-60.

14. Vitali C, Bombardieri S, Jonsson R, et al. Classification criteria for Sjögren's syndrome: a revised version of the European criteria proposed by the American-European consensus group. Ann Rheum Dis 2002;61:554-9. [MEDLINE]

15. Pertovaara M, Korpela M, Kouri T, Pasternack A. The occurrence of renal involvement in primary Sjögren's syndrome: a study of 78 patients. Rheumatology 1999;38:1113-20. [MEDLINE]

16. Balbin M, Grubb A, de Lange CC, Grubb R. DNA sequences specific for Caucasian G3m(b) and (g) allotypes: Immunogenetics 1994;39:187-93. [MEDLINE]

17. Balbin M, Grubb A, Abrahamson M, Grubb R. Determination of allotypes G1m(f) and G1m(z) at the genomic level by subclass-specific amplification of DNA and use of allele-specific probes. Exp Clin Immunogenet 1991;8:88-95. [MEDLINE]

18. Cavalli-Sforza LL, Menozzi P, Piazza A, et al. The history and geography of human genes. Princeton: Princeton University Press; 1994.

19. Cooper LJ, Robertson D, Granzow R, Greenspan NS. Variable domain-identical antibodies exhibit IgG subclass-related differences in affinity and kinetic constants as determined by surface plasmon resonance. Mol Immunol 1994;31:577-84. [MEDLINE]

20. Pritsch O, Hudry-Clergeon G, Buckle M, et al. Can immunoglobulin C(H)1 constant region domain modulate antigen binding affinity of antibodies? J Clin Invest 1996;98:2235-43. [MEDLINE]

21. McLean GR, Torres M, Elguezabal N, Nakouzi A, Casadevall A. Isotype can affect the fine specificity of an antibody for a polysaccharide antigen. J Immunol 2002;169:1379-86. [MEDLINE]

22. Hoffman RW, Sharp GC, Irvin WS, Anderson SK, Hewett JE, Pandey JP. Association of immunoglobulin Km and Gm allotypes with specific antinuclear antibodies and disease susceptibility among connective tissue disease patients. Arthritis Rheum 1991;34:453-8. [MEDLINE]

23. Pandey JP, Page GP, Silver RM, LeRoy EC, Bona CA. Anti-fibrillin-1 autoantibodies in systemic sclerosis are GM and KM allotype-restricted. Exp Clin Immunogenet 2001;18:123-9. [MEDLINE]

24. Rider LG, Shamim E, Okada S, et al. Genetic risk and protective factors for idiopathic inflammatory myopathy in Koreans and American whites: a tale of two foci. Arthritis Rheum 1999;42:1285-90. [MEDLINE]

25. Pandey JP, Prohászka Z, Veres A, Füst G, Hurme M. Epistatic effects of genes encoding immunoglobulin GM allotypes and interleukin-6 on the production of autoantibodies to 60- and 65-kDa heat-shock proteins. Genes Immun 2004;5:68-71. [MEDLINE]

26. Michalski J, Daniels T, Talal N, Grey H. Beta-2 microglobulin and lymphocytic infiltration in Sjögren's syndrome. N Engl J Med 1975;293:1228-31.

27. Pertovaara M, Pukkala E, Laippala P, Miettinen A, Pasternack A. A longitudinal cohort study of Finnish patients with primary Sjögren's syndrome. Clinical, immunological and epidemiological aspects. Ann Rheum Dis 2001;60:467-72. [MEDLINE]

28. Nakao Y, Matsumoto H, Miyazaki T, et al. Genetic and clinical studies of serum beta-2 microglobulin in hematologic malignancies. Clin Exp Immunol 1981;46:134-41. [MEDLINE]

29. Rautonen N, Sarvas H, Julkunen I, Pyhälä R, Mäkelä O. Gm allotypes influence the production of IgG3 but the effect is age-dependent. Hum Immumol 1991;32:72-7. [MEDLINE]

30. Pandey JP, French MA. GM phenotypes influence the concentrations of the four subclasses of immunoglobulin G in normal human serum. Hum Immunol 1996;51:99-102. [MEDLINE]

31. Musk AW, Zilko PJ, Manners P, Kay PH, Kamboh ML. Genetic studies in familial fibrosing alveolitis. Possible linkage with immunoglobulin allotypes (Gm). Chest 1986;89:206-10. [MEDLINE]

32. Luger AM, Komathireddy G, Walker RE, Pandey JP, Hoffman RW. Molecular and serologic analysis of HLA-genes and immunoglobulin allotypes in IgA nephropathy. Autoimmunity 1994;19:1-5. [MEDLINE]

33. Hartung K, Coldewey R, Rother E, et al. Immunoglobulin allotypes are not associated with HLA-antigens, autoantibodies and clinical symptoms in systemic lupus erythematosus. Members of the SLE Study Group. Rheumatol Int 1991;11:179-82. [MEDLINE]

34. Stenszky V, Kozma L, Szegedi G, Farid NR. Interplay of immunoglobulin G heavy chain markers (Gm) in predisposing to systemic lupus nephritis. J Immunogenet 1985;13:11-7. [MEDLINE]

35. Sanders PA, de Lange GG, Dyer PA, Grennan DM. Gm and Km allotypes in rheumatoid arthritis. Ann Rheum Dis 1985;44:529-32. [MEDLINE]

36. Eberhardt K, Grubb R, Johnson U, Pettersson H. HLA-DR antigens, Gm allotypes and anti-allotypes in early rheumatoid arthritis: their relation to disease progression. J Rheumatol 1993;20:1825-9. [MEDLINE]

37. Pandey JP, Cooper GS, Treadwell EL, Gilkeson GS, St Clair EW, Dooley MA. Immunoglobulin GM and KM allotypes in systemic lupus erythematosus. Exp Clin Immunogenet 2001;18:117-22. [MEDLINE]

38. Gottenberg JE, Busson M, Loiseau P, et al. In primary Sjögren's syndrome, HLA class II is associated exclusively with autoantibody production and spreading of the autoimmune response. Arthritis Rheum 2003;48:2240-5. [MEDLINE]