A common genetic background for selected autoimmune disorders, such as GD, SLE, T1D, CD and potentially MG and RA, involving both MHC and non-MHC encoded genes, has previously been suggested (14
). There is also a considerable overlap in concomitant diseases; for example, T1D is prevalent in patients with GD, SLE, CD and RA. Furthermore, CD is overrepresented in patients with GD (168
), and patients with SLE also show a higher prevalence of thyroid disorders (169
). In this report, we have added IgAD to this group of diseases, since it shows a markedly increased prevalence in GD, SLE, T1D and CD ().
Overlap between immune-mediated diseases.
IgAD is thought to be present from birth in most cases. Theoretically, the increased frequency of infections associated with IgAD could therefore precipitate autoimmune disorders such as GD and SLE. However, in CD, IgAD has occasionally been reported to occur after the onset of the gastrointestinal symptoms. Thus, the common genetic background is likely to be the main contributor to the different autoimmune disorders where environmental factors determine if, and when, the primary and subsequent diseases will appear.
The gene(s) involved are primarily located within the MHC region, where the population-attributable risk is strong in T1D, CD and RA and moderate in GD and SLE (168
). Our recent preliminary work in IgAD, based on 100 multicase families, suggests a similarly strong MHC-associated risk in these patients. However, owing to strong linkage disequilibrium within the MHC region, the gene(s) involved in disease pathogenesis, with the possible exception of HLA-DQ in CD, have not yet been identified.
It is well documented that IgAD is strongly associated with the MHC region, in particular, the HLA-B8, DR3, DQ2 haplotype (9
). This haplotype is also associated with GD, SLE, T1D and CD (reviewed in [16
]). Although single loci within the MHC region were initially thought to confer susceptibility or resistance to different autoimmune diseases, the current picture is markedly more complex, since multiple loci/genes have been shown to confer independent risk both in SLE (12
) and T1D (91
) (). This also appears to be true for IgAD where genes both within the class II (9
) and class III region (176
) have been shown to be associated with the defect (). Multiple loci have also been suggested to be involved in MG (17
), another 8.1 haplotype–associated disease, as well as immune-mediated diseases associated with other HLA haplotypes, including multiple sclerosis (MS) (179
) and RA (18
Markers across the HLA region associated with risk and protection to IgAD and selected autoimmune disorders. The HLA-DRB1*1501 allele constitutes a risk factor for SLE, but is a protective factor for IgAD and T1D.
Case-control studies in European SLE patients have shown a consistent association with the HLA-DRB1*1501 allele and its linked haplotypes (45
). An association with the DRB1*1501 allele (rs3135391 serving as a tagging SNP) has also been observed in patients with MS (12
). Although a few case reports on IgAD in MS patients have been published, no large-scale screening studies have been performed to date. In view of the almost complete protection against IgAD by the DRB1*1501 allele (12
), its frequency among MS patients would be expected to be quite low. Similarly, the DRB1*1501-DQB1*0602 haplotype confers protection from T1D (181
). Thus, the same allele confers risk or protection in different immune-mediated diseases, suggesting its involvement at a crucial step in pathogenesis.
The association between a given disease and the MHC could either be due to coding mutations/variations in a given gene, directly influencing its function, or, alternatively, mutations/variations in regulatory sequences, affecting the expression of the gene. Examples of the former include copy number variation of functional C4 alleles in SLE (46
) and potentially IgAD (182
) and promoter polymorphisms in nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor-like 1
, also known as IKBL
), DEAD (Asp-Glu-Ala-Asp) box polypeptide 39B
, also known as BAT1
) and HLA-DQA1 (185
) in the latter.
We previously suggested that an IgAD predisposing (mutated/variant) haplotype (B8, DR3, DQ2) is present at a frequency of 0.46% in the general Swedish population (10
). Preliminary experiments have indeed shown multiple sequence differences between patients and controls homozygous for the 8.1 haplotype. Similar findings have previously been reported by Yan et al.
) in Alaskan natives with RA, where multiple mutations/variations were identified in the disease-associated haplotype. The presence of such a predisposing haplotype may explain the overlap in HLA between IgAD and autoimmunity, since the affected variant of the 8.1 haplotype might constitute a risk factor for developing additional disorders, including GD, SLE, T1D, CD and potentially MG and RA. However, additional studies in cohorts of patients with different auto-immune diseases will be necessary to substantiate this notion, and full sequencing of the MHC region of the 8.1 haplotype may ultimately be required to identify the potential mutations/variants involved.
The mechanism underlying the induction of IgAD still remains elusive. It is however likely that the pathophysiological process involves a break of tolerance against IgA itself (since 30% of IgAD patients have demonstrable titers of IgG antibodies against IgA) or one of the factors involved in the switching process (such as a proliferation-inducing ligand [APRIL] or B-cell activating factor [BAFF]).
A number of non-MHC genes have also been shown to be associated with GD, SLE, T1D and CD. These genes generally cluster into pathways involved in T-cell differentiation, cell activation/signaling and innate immunity and have been shown to be associated with single or several diseases (). It is however noteworthy that there is also a lack of association of given “autoimmune” genes such as PTPN22, which shows association with a large number of diseases, including some of the above disorders, suggesting differences in the pathophysiological pathways.
SNPs reported to be associated with IgAD and selected autoimmune disorders.a
is located on chromosome 2q24 and encodes the interferon induced with helicase C domain 1 protein. Together with retinoic acid-inducible gene I
), it functions as a sensor for viral infections. IFIH1
expression is highly upregulated in activated immune cells in response to type 1 interferon induced by viral infections, suggesting that it could potentially be involved in the pathogenesis of autoimmune diseases including T1D (187
). Several SNPs within the IFIH1
gene and its 3′ untranslated region show an association with T1D (rs2111485, rs13422767, rs1990760 and rs3747517) (188
), where the rs1990760 marker (Ala946Thr) is the most strongly associated (reviewed in [189
]). Most of these rare IFIH1
alleles are protective against disease (190
A recent large-scale study also showed a highly significant association of rs1990760 to SLE in Caucasian (191
) but not in Japanese patients (192
). Moreover, GD has also been shown to be associated with rs1990760 (193
), although this point remains controversial (194
). CD, on the other hand, does not appear to be associated with rs1990760 (74
In 2010, we performed a GWAS on a large cohort of patients with IgAD (13
) and identified an association with rs1990760 in the IFIH1
gene. Taken together, there is thus ample evidence for the implication of IFIH1
both in IgAD and in several autoimmune disorders, although the mechanism involved remains elusive.
, located on chromosome 16p13, is widely expressed on B lymphocytes, natural killer and dendritic cells (197
). Several SNPs (rs2903692, rs17673553, rs725613 and rs12708716) within the CLEC16A
gene have been shown to be associated with T1D in different populations (196
). Dubois et al.
) recently also identified an association between the class II, major histocompatibility complex, transactivator
)–suppressor of cytokine signaling 1
region and CD, although it just reached borderline genome-wide significance. Another study, however, suggested that CLEC16A
(rs2903692) is not involved in susceptibility to CD development (205
). rs12708716 was also suggested to be associated with SLE (191
) where the A allele confers susceptibility. Awata et al.
, on the other hand, showed that rs2903692 was not associated with GD in Japanese patients (206
). In IgAD, we recently showed suggestive evidence for association with CLEC16A
(rs6498142 and rs7201845) (13
), again suggesting a common genetic link.
One remaining question is whether the MHC- and non-MHC–associated susceptibility genes are acting independently or synergistically in the pathophysiological processes underlying auto-immune diseases. A number of studies have addressed this question, and Hodge et al.
) and Jacobson et al.
) showed that interaction between HLA-DRB1*03 and different thyroglobulin
variants conferred an increased risk for GD. A subsequent study in Japanese subjects (208
) suggested that there was also an interaction between selected HLA-A and DP alleles and CTLA4
in a subgroup of GD patients, although Kula et al.
) suggested that the interaction with CTLA4
was due to HLA-DR encoding genes.
In T1D, a study in Belgian patients showed no interaction between HLA-DQ and the IFIH1
rs1990760 SNP (210
). However, PTPN22
Trp620 (rs2476601) has a higher relative risk in T1D patients carrying low-risk MHC class II genotypes (non-DR3/DR4) than in those carrying the high-risk ones, suggesting a potential interaction (211
). Moreover, a Norwegian study indicated a weak synergistic effect between FOXP3
and the HLA-DR3, DQ2, haplotype both in T1D and CD patients (214
). A recent case-control collection study also showed a significant interaction between HLA-DR3 and IRF5
in patients with SLE (215
). Similar interaction studies in IgAD have not yet been performed.
In summary, IgAD is markedly more prevalent in patients with a variety, albeit not all (such as MG [14
]), of 8.1 haplotype–associated autoimmune diseases. Similarities in the genetic susceptibility suggest involvement of common pathophysiological pathways, implicating that IgAD, as recently suggested by Ferreira et al.
), may in fact be an autoimmune disease. However, additional dense SNPing and sequencing of the implicated genes may be required to fully understand the mechanisms involved.