Introduction

The term ‘collagen diseases’ was originally proposed for a group of systemic diseases characterized by widespread fibrinoid degeneration of collagen and includes systemic autoimmune diseases such as rheumatoid arthritis (RA), systemic lupus erythematosus (SLE) and systemic sclerosis (SSc).1 Although the etiology of these systemic autoimmune diseases is still unknown, it is thought that susceptibility to systemic autoimmune diseases is associated with genetic, environmental and stochastic causes. Predisposing genetic factors for systemic autoimmune diseases include the human leukocyte antigen (HLA) class II alleles,2, 3, 4 which are the strongest genetic factors in almost all systemic autoimmune diseases.

HLA class II gene cluster is encoded in the 0.9 M base region on human chromosome 6 and are composed of >30 loci.5 The genes coding HLA class II molecules are located in this region and >3500 alleles were reported. The HLA alleles coded on these loci are in strong linkage disequilibrium (LD), making it highly difficult to determine the functionally relevant protective gene in this region. There are at least six loci for HLA class II genes; DRA, DRB1 (DRB3, DRB4 or DRB5 located in some haplotypes as copy number variations), DQA1, DQB1, DPA1 and DPB1 encode the α and β chains of HLA-DR, -DQ and -DP molecules, respectively (Figure 1). HLA-DR, -DQ and -DP molecules were heterodimers formed by the 32 kD α and 28 kD β chains. HLA class II molecules present primarily exogenous peptides to T-cell receptors of CD4+ T cells, stimulating acquired immunity.

Figure 1
figure 1

HLA class II gene organization of each haplotype. DRB3 is located in the haplotype of DRB1*03, DRB1*11, DRB1*12, DRB1*13 and DRB1*14. DRB4 is included in the haplotype of DRB1*04, DRB1*07 and DRB1*09. DRB5 is found in the haplotype of DRB1*15 and DRB1*16.

Full size image

Predisposing effects of HLA on systemic autoimmune diseases

Skewed HLA class II allele frequencies are associated with systemic autoimmune diseases. SSc is a chronic systemic autoimmune disease that is featured by skin and internal organ fibrosis. Antinuclear antibodies are frequently detected in SSc patients. Genetic risk factors for SSc include HLA class II alleles as the strongest ones. HLA-DRB1*11:04, DQB1*03:01 and DQB1*26 epi (DQB1 alleles encoding a non-leucine residue at position 26 of the HLA-DQβ chain) are associated with SSc susceptibility in Europeans6 and DRB1*15:02, DPB1*03:01 and DPB1*09:01 in Asians.7, 8, 9, 10 It was also known that DRB1*08:04 and DQB1*03:01 are associated with SSc in African-American and that DRB1*11:04 and DQB1*03:01 are associated in Hispanic populations.6

Anti-centromere antibodies (ACA)11 and anti-topoisomerase I antibodies (ATA)12 are detected in SSc patients and suggested to define subsets of SSc. It has been shown that ACA-positive SSc was associated with DQB1*05:01 and DQB1*26 epi in European populations6 but with DRB1*01:01, DRB1*10:01, DRB1*15:02, DQB1*05:01, DPB1*03:01 and DPB1*04:02 in Asians.8, 9, 10, 13 DPB1*13:01 is reported to be associated with ATA-positive SSc in Europeans6 and DRB1*15:02, DRB1*16:02, DQB1*06:01, DPB1*03:01, DPB1*09:01 and DPB1*13:01 in Asians.8, 9, 10, 13 Thus different HLA class II alleles are associated with the risk of overall SSc or subsets of SSc in different ethnic groups.

SLE is a prototypic and systemic autoimmune disease that affects multiple organs. Several different autoantibodies are detected in SLE patients. HLA is one of the important genetic risk factors for SLE. DRB1*03:01 and DRB1*15:01 are associated with SLE susceptibility in European14, 15 and DRB1*09:01, DRB1*15:01 and DRB1*15:02 in Asian populations.16, 17, 18, 19 It was also known that DRB1*15:03 is associated with SLE in African-American populations20 and that DRB1*08:02 is associated with Hispanic populations.21 With respect to the specific autoantibodies to ribonucleoprotein, DPB1*05:01 was associated with SLE patients with anti-Ro/SS-A or anti-La/SS-B antibodies.22 The association of polymorphisms of amino-acid residues 11 and 13 of DRβ molecule with SLE was reported.23 These amino-acid residues form the HLA-DR peptide-binding groove.24 Thus different DRB1 alleles are associated in different ethnic groups also in the case of SLE.

RA is a chronic systemic autoimmune disease that mainly affects synovial joints, but extra-articular manifestations are often complicated. Association between RA and HLA has been known for 40 years.25 RA risk is associated with some HLA-DRB1 alleles.26 Amino-acid sequence at positions 70–74 (QKRAA, RRRAA or QRRAA) of the HLA-DRβ chain is conserved among these alleles and was referred to as the ‘shared epitope’ (SE).26 DRB1*04:01 and DRB1*04:05 are associated with RA in European and Asian populations, respectively.26, 27 Such difference could be explained by the different frequencies of these RA risk alleles in different ethnicities. A gene dosage effect was reported in RA but not in SLE; having two copies of SE alleles confer higher RA risk than those with one copy of SE. Although all the known genetic risk factors could explain 16% of RA risk, HLA alleles can explain 11%.28, 29 Anti-citrullinated peptide antibody (ACPA) is specifically detected in RA patients and is suggested to be pathogenic. SE alleles are strongly associated with ACPA-positive RA but only weakly with ACPA-negative RA.30 The association of HLA-DRB1 amino-acid residues with RA was also analyzed, and the important role of the polymorphisms in amino-acid positions 11 and 13 of DRβ molecule was reported.31 The amino-acid residues of positions 11 and 13 form the HLA-DR peptide-binding groove.24 Thus HLA is the most important genetic risk factor for RA, and the main predisposing alleles are different among ethnic groups. In addition, the major predisposing alleles are different among systemic autoimmune diseases.

Protective effects of HLA on systemic autoimmune diseases

Although many studies reported susceptible associations of HLA class II alleles with systemic autoimmune diseases,2, 3, 4 few attempts have been made on the protective association of HLA. DRB1*07:01, DRB1*15:01, DQB1*02:02 and DQB1*06:02 alleles were reported to be protectively associated with SSc in European populations6 and DRB1*01:01, DRB1*04:06, DRB1*07:01, DRB1*13:02, DRB1*14:06, DQB1*03:01 and DPB1*02:01 in Asians.9, 10, 13 With respect to SLE, DRB1*13 is protective against European SLE,32, 33 and DRB1*13:02 and *14:03 have been shown to be protective in Japanese population.19

In the case of RA, it had been suggested that an amino-acid sequence (DERAA) at positions 70–74,34 isoleucine at position 67 (I67),35 aspartic acid at position 70 (D70)36 or a conserved amino-acid sequence at positions 71–74 (S1; ARAA or ERAA) 37, 38 in the HLA-DRβ chain were shown to be protective.

Of particular interest, DRB1*13:01 and DRB1*13:02 are commonly present in all these protective allele groups. DRB1*13 alleles were reported to be protectively associated with RA in European populations.39, 40 DRB1*13:01 allele was protective against ACPA-positive RA in European populations.41 The protective effect was attributed to DRB1∗13 rather than DERAA, D70 or I67.41 It was recently shown that HLA-DRB1*13 affects the onset of ACPA-positive RA but not protective against ACPA production in individuals without RA.42 DRB1*13 was also protective against RA in Turkish43 and Asian populations.44 DRB1*13:02 is protectively associated against ACPA-positive and ACPA-negative RA in Japanese.27, 45 Thus HLA is one of the important resistant factors for systemic autoimmune diseases, and DRB1*13 are the shared protective alleles against multiple diseases,

The role of HLA-DRB1*13:02 in systemic autoimmune diseases

DRB1*13:02 is carried by the extended haplotype A*33:03-C*14:03-B*44:03-DRB1*13:02-DQB1*06:04-DPB1*04:01, which has been reported to be positively selected in Japanese.46 LD between DRB1*13:02 was also observed in HLA-G located on the telomeric side of HLA-A;47 thus LD with DRB1*13:02 extends across almost the whole MHC region. Therefore, certain allele(s) on this haplotype is thought to have a causative protective for systemic autoimmune diseases. In fact, the causative allele may not necessarily be single, and multiple protective alleles of different loci may have a protective role independently.

Haplotype association analyses provide valuable information in elucidating the causatively associated alleles among a group of alleles in LD. LD between DRB1*13:02 and DQB1*06:04 or DQB1*06:09 is especially strong in the Japanese population.48 The haplotype carrier frequencies of both DRB1*13:02-DQB1*06:04 and DRB1*13:02-DQB1*06:09 in systemic autoimmune disease patients were obtained from secondary analyses based on the previously published data,13, 19, 27 suggesting that DRB1*13:02 rather than DQB1 alleles has the protective role (Table 1). In addition, two-locus analysis and conditional logistic regression analysis between these alleles revealed that the primary protective effect is neither DQB1*06:04 nor DQB1*06:09 but DRB1*13:02 (Tables 2 and 3). Thus the primary protective role of DRB1*13:02 with Japanese SSc, SLE and RA was suggested.

Table 1 HLA allele and haplotype carrier frequencies in SSc, SLE and RA patients and controls
Full size table
Table 2 HLA-DRB1 or DQB1 allele carrier frequency in the SSc, SLE and RA patients with specific DRB1 or DQB1 alleles (two-locus analysis)
Full size table
Table 3 Conditional logistic regression analysis between the protective HLA alleles in SSc, SLE and RA patients
Full size table

It was known that DRB1*13:02 and DQA1*01:02 alleles are in strong LD in the Japanese population.48 DRB1*15:01 or DRB1*16:02 alleles are also in strong LD with DQA1*01:02; however, neither DRB1*15:01 nor DRB1*16:02 conferred protective effects for systemic autoimmune diseases.13, 19, 27 Thus DQA1*01:02 is unlikely to be the functional protective allele for systemic autoimmune diseases. Similarly, DRB1*13:02 and DRB3*03:01 alleles are in strong LD in the Japanese population.48 DRB1*12:02 and DRB3*03:01 alleles are also in strong LD, but DRB1*12:02 did not show protective association against systemic autoimmune diseases.13, 19, 27 Thus DRB3*03:01 allele is unlikely to be the primary protective allele for systemic autoimmune diseases.

These haplotype analyses supported the primary protective role of DRB1*13:02 among the HLA class II genes in the Japanese. However, the possibility that other gene(s) on the DRB1*13:02-extended haplotype, including those on the class I and class II regions, has a functional role cannot be excluded. In fact, it is possible that multiple genes on this haplotype may independently have a functional role. Such possibility will be addressed by comparison of the re-sequencing data of the MHC region of the DRB1*13:02 haplotype.

In the protective associations of DRB1*13:02 with the systemic autoimmune diseases (Table 4), a gene dosage effect was observed. Homozygosity of DRB1*13:02 more effectively prevents the development of systemic autoimmune diseases than heterozygous DRB1*13:02 genotypes.

Table 4 HLA-DRB1 genotype frequency in the SSc, SLE and RA patients and controls
Full size table

DRB1*13 has also been shown to be protective for other systemic autoimmune diseases, including anti-neutrophil cytoplasmic antibody-associated vasculitis,49, 50, 51 mixed connective tissue disease52 and polymyositis/dermatomyositis.53 In addition, it was also reported that DRB1*13:02 confers protection in organ-specific autoimmune diseases, such as psoriasis,33, 54 autoimmune hepatitis,55 primary biliary cirrhosis,56 Graves’ disease and Hashimoto’s thyroiditis.57 DRB1*13:02 is also a protective allele for cervical cancer caused by human papilloma virus infection,58 severe malaria59 and chronic hepatitis B infection.60 In addition, DRB1*13:02 is associated with slow disease progression in HIV infection.61 As DRB1*13 molecules proficiently stimulate CD4+ T cells,62, 63 it appears possible that DRB1*13:02 might be protective for putative undiscovered infectious diseases that trigger autoimmune diseases. Such a hypothesis might explain the role of DRB1*13:02 for the protection of systemic and organ-specific autoimmune diseases.

Potential molecular mechanisms of HLA-DRB1*13

DRB1*13:01 is protective against ACPA-positive RA40; 41 and SSc64 in European populations, while DRB1*13:02 allele was protectively associated with RA27 and SSc13 in Japanese populations. The only difference in the amino-acid sequence between these two alleles is at position 86 (V in DRB1*13:01 and G in DRB1*13:02) of the HLA-DRβ chain. It is plausible that common protective mechanisms are present between DRB1*13:01 and DRB1*13:02 against systemic autoimmune diseases. In support of this hypothesis, a common peptide (TPKIQVYSRHPAENGKSN) derived from β2-microglobulin has been shown to be presented by DRB1*13:01 and DRB1*13:02 molecules.65

When we examined the association of each amino-acid residue with systemic autoimmune diseases, the protective role was mapped to the amino-acid position 13S of the HLA-DRβ chain (Figure 2). This amino-acid residue constitutes the HLA-DR peptide-binding groove24 and is shared between DRB1*13:01 and DRB1*13:02 molecules.

Figure 2
figure 2

Associations of amino-acid residues in the DRβ chain with systemic autoimmune diseases, including SSc, SLE and RA. Corrected P (Pc) values were calculated by multiplying the P-value by the number of amino-acid residues tested. Associations were analyzed by Fisher’s exact test using two-by-two contingency tables in the comparison of systemic autoimmune disease patients (n=2397) and healthy controls (n=413). Protective associations are shown by open circles and predisposing associations by filled circles.

Full size image

The protective effects of the DRB1*13 alleles can overcome the predisposing effects of susceptible alleles in DRB1 heterozygous RA patients.27, 41 Similar tendency was also observed in SLE19 and SSc13 patients heterozygous for DRB1 in Japanese populations. To explain the dominant effects of protective DRB1*13 alleles, it was hypothesized that resistant DRB1*13 molecules are recognized by the T-cell receptors of autoreactive regulatory T cells with high affinity.66

A recent study made an attempt to explain the protection mediated by DRB1*13 molecules for RA by the DERAA motif at positions 70–74 of the HLA-DRβ chain.67 This motif was shared with vinculin and microbe-derived proteins. Citrullinated vinculin is one of the autoantigens of ACPA and self-reactive CD4+ T cells. The motif is presented by predisposing DQ molecules and stimulates self-reactive CD4+ T cells, resulting in the triggering of arthritis. However, these self-reactive CD4+ T cells could not be found in DRB1*13 possessing individuals. The motif of DRB1*13 molecules was thought to mediate the central tolerance, explaining the protective mechanisms of DRB1*13 molecules.

Non-inherited maternal antigen was reported to have protective roles in the pathogenesis of RA. It was observed that the resistant DRB1 alleles with DERAA epitope at positions 70–74 of the HLA-DRβ chain, including DRB1*13, have protective effects on children, though these alleles were not inherited from their mothers.68 This could be explained by the maternal micro-chimerism in the circulation of the children.

Thus several hypotheses have been proposed to explain the protective molecular mechanisms of DRB1*13 with systemic autoimmune diseases. These hypotheses, along with other possibilities, need to be validated in future studies.

Conclusion

Recent studies demonstrated the protective effect of DRB1*13 with systemic autoimmune diseases. DRB1*13:01 is protective in European populations and DRB1*13:02 allele in Japanese. As the ethnic difference of HLA allele distributions is well known, the protective effects of HLA alleles for systemic autoimmune diseases in other populations should be explored. Because DRB1*13 is carried by extended haplotypes formed by HLA class II and class I alleles, it is quite difficult to identify which of the alleles is functionally responsible for the protective effects; however, several lines of evidence suggest that DRB1*13 alleles themselves, at least in part, have a role, although independent effects from other genes in LD cannot be excluded. Several hypotheses have been proposed to explain the protective molecular mechanisms of DRB1*13 molecules against systemic autoimmune diseases, some of which can explain the dominant effects of DRB1*13 molecules. The precise understanding of the protective mechanisms of DRB1*13 might eventually lead to cellular, molecular or genetic targets for the permanent curative treatment of systemic autoimmune diseases.