Recent Findings on Genetics of Systemic Autoimmune Diseases

Summary

Association studies of over 1 million SNPs capturing most of the human genome common variation became possible thanks to the information provided by the HapMap International project and the development of high-throughput genotyping technologies at accessible prices. Genome-wide scans analyzing thousands of individuals have now identified most if not all of the major genes involved in susceptibility for several systemic autoimmune diseases. In particular, results for rheumatoid arthritis (RA), systemic lupus erythematosus (SLE) and systemic sclerosis (SSc) are reviewed here. While most genes are shared between diseases, few seem to be unique reflecting that we still are long before knowing all genes, their interactions with other genes and the environment and their impact on biological functions.

Introduction

Previous studies on the genetics of complex diseases used small sample sizes leading to inconclusive results, with the exception of the strong genetic association between genes within the major histocompatibility complex (MHC) and the various autoimmune diseases. While GWAS have provided some surprises, they have also confirmed a few of the old results, but not all. The most important genes identified previously, IRF5 and STAT4 for SLE and PTPN22 and STAT4 for RA stand [1]. In this review we focus only on those genes identified in the last two to three years. As GWAS with extensive coverage have been used, except for a few examples (PDCD1 is a relevant one, as SNPs within this gene are lacking in the GWAS arrays), most of the genome has been screened with thousands of samples providing an overall picture of genetic susceptibility. We present only those genes that have been clearly replicated and whose role in genetic susceptibility is beyond doubt or the result of powerful GWAS with replication using large sample sizes.

Genetics of SLE

SLE is considered as a prototypic autoimmune disease, characterized by production of antinuclear autoantibodies, immune complex deposition and subsequent multiple organ damage.

The understanding of the genetic basis of SLE has expanded enormously over the past couple of years, driven principally by technological advances and the assessment of six genome-wide association studies (GWAS) in the last two years, in Caucasians and Asians [2–7].

The great majority of the identified genes are involved in innate and adaptive immune responses. The convincingly associated genes, as summarized in Table 1, are mainly implicated in immune-complex (IC) processing, T/B cell signaling and/or and Toll-like receptor (TLR) and type I interferon signaling. The strongest association in the GWAS era is undoubtedly the MHC region, but because it consists of strongly correlated polymorphisms in a ~6Mb region involving more than 400 genes it remains a challenge. Recent observations by Fernando, et al have shown at least two independent genetic effects within the MHC region in SLE: one signal is provided by the HLA-DRB1*0301 allele, and the second within the class III region specifically detecting the 6^th intron of the SKIVL2 gene [8]. A more recent study genotyping over 1400 variants within the MHC region and over 10,000 individuals with various autoimmune diseases including SLE and RA showed the top disease-specific signal for SLE to be SNP [8,9] located between TNXB and CREBL1 and the HLA DRB1*0301 [9].

Table 1.

Susceptibility Genes for SLE

Chromosome	Gene	SNPs	Population	References
6p21	HLA region	DRB1*0301 and several other alleles	European, several Asian, African-American, mixed European-Amerindian	[8,9]
7q32	IRF5	5bp promoter indel, rs2004640, rs2070197, rs10954213	European, several Asian, mixed European-Amerindian, African-American	[3–5,7,20]
2q32	STAT4	rs7574865, rs3821236, rs7601754	European, mixed European-Amerindian, several Asian, African-American	[3–5,7,20]
6q23	TNFAIP3	rs5029939, rs2230926	European, Asian, African- American	[2–5,7,17,20]
16p11	ITGAM	rs9888739, rs1143679, rs4548893	European, mixed European-Amerindian, Asian, African-American	[3–5,7,10,20]
4q24	BANK1	rs10516487, rs17266594, rs3733197	European, European-Amerindian, Asian	[3,6,7]
1p13	PTPN22	rs2476601	European	[4]
8p23	BLK	rs13277113, rs2736340	European, several Asian	[3–5,7,20]
2q37	PDCD1 (CD279)	PD1.3A	European, European-Amerindian, Chinese	[49]
1q25	TNFSF4	Risk haplotype; rs3850641	European, Asian	[3,7,20]
18q22.3	CD226	rs763361, rs727088	European, European-Amerindian	[13]
1q21-23	FCGR2A	ARG131HIS	European, European-Amerindian, African American	[4,5,20]
19p13.2	TYK2	rs280519	European	[20]
3p21.3	TREX1	rs72556554, R114H and other 11 nonsynonimous substitutions	European	[16]
Xq28	MECP2-IRAK1	rs2269368	European, Chinese, Korean, European-Amerindian (Mexican)	[20]
3p14.3	PXK	rs6445975, rs2176082	European	[4,20]
2q24	IFIH1	rs1990760	European	[20]
11p15.5	KIAA1542 (PHRF1)	rs4963128	European	[4]
8p23.1	XKR6	rs6985109	European	[4]
6q21	ATG5-PRMD1	rs6568431, rs2245214	European, Chinese	[4,20]
22q11.2	UBE2L3	rs5754217	European, Chinese	[2,20]
5q33.3	PTTG1	rs2431099	European	[2,20]
6p21	UHRF1BP1	rs11755393	European	[20]
5q32	TNIP1	rs7708392	European, Chinese, Thai	[3,20]
7p15.2	JAZF1	rs849142	European	[20]
7p21.3	ICA1	rs10156091	European	[4,20]
1q24	IL10	rs3024505	European	[20]
1q25.3	NMNAT2	rs2022013	European, Chinese	[3,4]
11q23.3	ETS1	rs6590330	Chinese, Thai	[3,7]
10q11.23	WDFY4	rs877819	Chinese, Thai	[3,7]
7p12.2	IKZF1	rs4917014	Chinese	[3]
12q24.32	SLC15A4	rs10847697, rs1385374	Chinese	[3]
2p22.3	RASGRP3	rs13385731	Chinese	[3]

Impaired IC clearance and deposition is an important pathological aspect in SLE. Susceptibility genes with important roles in IC processing known from previous studies are the FcGR family of genes and more recently ITGAM [10] coding for the surface antigen CD11b (or CR3). Signal transduction in immune cells, particularly T and B cells, is another pathway that has revealed to contain multiple lupus susceptibility genes, modulating T cell signaling such as TNFSF4 (OX40L) [11,12] and CD226 on NKT cells [13]. More recently BANK1 and BLK, thought to be involved in B cell activation and tolerance [5,6], respectively have been now clearly established [3,14,15].

One pathway that has been biologically and genetically strongly related to SLE pathogenesis is the type I interferon (IFN) pathway. Several genes for factors upstream and downstream of IFN production, such as IRF5, STAT4, and more recently TNFAIP3, TYK2, and TREX1, have been associated to susceptibility to SLE [2,16,17]. IRAK1 has been an interesting candidate but it is closely linked to MECP2, a gene that can regulate expression of IRAK1 also associated. It has proven as yet impossible to discern the genetic effects between these X chromosome genes [18–20]. TREX1 is mainly represented by rare but penetrant and mutations leading to high levels of type I interferon found in few patients with lupus, suggesting a potentially important role of rare variants that have remained undetected with the use of common variation mapping.

Other recently identified loci, such as PXK, XKR6 and KIAA1542 close to IRF7 [4], with no known function or correlation to SLE pathology, have the potential to lead to the discovery of novel pathways involved in SLE. It is unclear if the genetic association of KIAA1542 indeed represents an association with IRF7.

Finally, a very large replication study identified and replicated several genes among which are JAZF1, TNIP1, PRDM1 (or BLIMP1), UHRF1BP1, PTTG1 [2], UBE2L3 [2], IL10, IL21 and the IL21R [20]. Importantly, several other genes were also confirmed in this study such as ATG5, ICA1 and NMNAT2 found in a previous GWAS from the SLEGEN consortium [4]. ATG5 is an important component of the autophagy pathway. JAZF1 and UHRF1BP1 are transcription factors, while TNIP1 interacts with TNFAIP3, its function in regulating TNFAIP3 is not known. TNFAIP3 (or A20) regulates inflammation by turning off NFκB through polyubiquitination and degradation.

Studies in Asian populations have identified new susceptibility genes for lupus and replication of the hitherto identified genes in Europeans has revealed the presence of some genes but not others [3,7,14]. Two genes were clearly identified, ETS1 and WDFY4 [3,7]. ETS1 is involved in the development of TH17 cells while WDFY4 codes for a protein of unknown function. STAT4, IRF5, BANK1, BLK, TNFAIP3 and TNFSF4 have been confirmed in Asians [3,7]. Studies on African Americans and European-Amerindian admixed populations are ongoing.

Genetics of RA

Multiple GWAS have corroborated the MHC genes as major genetic contributors to the risk of developing RA [21–25]. Within the MHC, the strongest contribution to risk is given by the HLA-DRB1 gene, which codes for the third hypervariable region of the HLA-DR molecule β chain. Since the original report of Peter Stastny, diverse HLA-DRB1 alleles have been associated in European, Asian, African and European-Ameridian populations [26]. All the associated alleles were unified by Peter Gregersen under the hypothesis of the shared epitope (SE). SE alleles are associated with anti-citrulline antibody production, a major biomarker for RA and determine severity. A study analyzing over 1400 SNPs within the HLA regions from the IMAGEN Consortium found the peak association for RA between the gene BTNL2 and HLA-DRA (SNP rs2395175) and the allele for the DQA1 gene DQA1*0301 [9]. Dense typing of the MHC have revealed several DRB1-independent associations, including a signal at MICA, one in the border between class I and class III region and some in the class I region [27].

Although any other association outside the MHC is rather modest in RA, and the SE accounts for 18 to 37% of the genetic heritability [28], fine mapping of candidate non-MHC linkage regions successfully identified important susceptibility genes such as PTPN22 (1p13), PADI4 (1p36) and STAT4 (2q32). To date, seven GWAS conducted in collections of thousands of patients with RA and healthy controls reliably detect several new susceptibility genes [21–23,25,29–31] (Table 2). The loci supported by the best evidence are the TRAF1-C5 [22,24,30] and the 6q23 region [23–25,29], both with strongest effect in anti-CCP+ patients. GWAS studies have also confirmed the association of PTPN22 and STAT4, as well as previously reported candidate genes identified by studies that did not have enough power such as CTLA4 and CD40. The maximum power has been achieved by a recent meta-analysis of GWAS conducted in a total of 12,307 patients and 28,975 controls of European ancestry [24]. This allowed the identification and confirmation of IL6ST, SPRED2, RBPJ, CCR6 [25], IRF5 and PXK [24]. While SPRED2 has been found to be a negative regulator of the Ras-ERK cascade, IL6ST and CCR6 are inflammation regulators and RBPJ is a transcription factor important in dendritic cell function.

Table 2.

Susceptibility Genes for RA

Chromosome	Gene	SNPs	Population	Reference
6p21.32	HLA-DRB1	rs615672, rs660895, rs64576200, rs6910071, rs13192471	European, Japanese	[21–25]
1p13.2	PTPN22	rs6679677, rs2476601	European	[21–24,30]
2q32.3	STAT4	rs7574865	European, Japanese	[24,25]
9q34	TRAF1-C5	rs3761847, rs881375	European	[22,24,30]
6q23.3	TNFAIP3, OLIG3	rs10499194, rs6920220	European, Japanese	[23–25,29]
6p21.32	HLA-DQA1, HLA-DQA2	rs6457617	European	[31]
18q23	SALL3	rs2002842	European	[31]
20q13.12	CD40	rs4810485	European	[23,24]
9p13.3	CCL21	rs2812378	European	[23]
12q13.3	KIF5A, PIP4K2C	rs1678542	European	[23]
1p36.32	TNFRSF14	rs3890745	European	[23,24]
10p15.1	PRKCQ	rs4750316	European	[23,24]
7q21.2	CDK6	rs42041	European	[23]
2p16.1	REL	rs13017599, rs13031237	European	[24,30]
2q33.2	CTLA4	rs231735, rs3087243	European	[24,30]
8p23.1	BLK	rs2736340	European	[30]
2q11.2	AFF3	rs11676922, rs10865035	European	[24]
5q11.2	ANKRD55, IL6ST	rs6859219	European	[24]
14q24.3	BATF	rs7155603	European	[24]
5q21.1	C5orf30	rs26232	European	[24]
9p13.3	CCL21	rs951005	European	[24]
6q27	CCR6	rs3093023, rs3093024	Japanese	[24,25]
1q24.2	CD247	rs840016	European	[24]
17q12	IKZF3	rs2872507	European	[24]
4q27	IL2, IL21	rs13119723	European	[24]
10p15.1	IL2RA	rs706778	European	[24]
7q32.1	IRF5	rs10488631	European	[24]
15q23	KIF3	rs17374222	European	[24]
1p34.3	POU3F1	rs12131057	European	[24]
3p14.3	PXK	rs13315591	European	[24]
4p15.2	RBPJ	rs874040	European	[24]
12q24.12	SH2B3	rs3184504	European	[24]
2p14	SPRED2	rs934734	European	[24]
21q22.3	UBASH3A	rs11203203	European	[24]

It is important to note that major differences across racial groups have been noticed; for example, PTPN22 and CTLA4 were associated in Europeans whereas PADI4 and SLC22A4 are confirmed only in Asian population groups. In contrast, the STAT4 association is valid for European, Asian, and European-Amerindian but not African Americans [32]. All these findings add evidence of the complexity and the heterogeneity of the genetic basis of RA and justify the study of diverse populations. One GWAS has been performed in Japanese [25] but none in admixed populations of European-Amerindian or African American. The observed heterogeneity highlights the importance of conducting well-powered GWAS in non-European populations in order to dissect all the genetic contribution to the disease.

Genetics of Systemic Sclerosis

Systemic sclerosis (SSc) or scleroderma is an autoimmune disease characterized by an extensive fibrotic process that affects multiple organs and tissues. Until now genetic studies have not been particularly successful in the identification of risk factors for SSc. The controversial results found for the majority of genes, such as, PTPN22, CTGF or TGF-β, suggest that those studies were often limited by small sample size and clinical heterogeneity. Some were finally replicated when large sample sizes were used (Table 3). One of the first discoveries was the contribution, as in other autoimmune diseases, of HLA class II genes, which seem to be predominantly associated with the presence of specific autoantibodies rather than with SSc itself. These associations have been confirmed in a recent GWAS in SSc in the Korean population [33]. The first GWAS in a sample of European ancestry including 2,296 SSc patients and 5,171 controls has firmly established the role of the HLA genes in SSc [34]. STAT4 and IRF5 that had previously been identified as risk factors for SSc through candidate gene studies [35,36], were also identified in the GWAS of European patients with SSc [34]. The consistent association of these genes with SSc susceptibility provides compelling evidence that variation in genes with key functions in the innate immune system are involved in the pathogenesis of the disease.

Table 3.

Susceptibility Genes for SSc

Chromosome	Gene	SNPs	Population	Reference
6p21.32	HLA-DPB1, DPB2 HLA-DQB1	rs3128930, rs7764491, rs7763822, rs3128965, rs3117230, rs7763822, rs7764491, rs3117230, rs3128965 rs6457617	Korean, European European	[33] [33]
2q32.3	STAT4	rs7574865, rs11889341, rs8179673, rs10181656, rs3821236	European, Japanese	[44]
7q32	TNPO3-IRF5	rs2004640, rs2280714, rs10954213, rs10488631, rs12537284, rs4728142	European, Japanese	[35,36]
4q24	BANK1	rs10516487, rs17266594	European	[41]
8p23.1	C8orf13-Blk	rs2736340, rs13277113	European	[42,43]
17q21.32	TBX21	rs11650354, rs17699436	European	[[44]
1q25.1	TNFSF4	rs1234314, rs2205960, rs844644, rs844648	European	[47]
10q24	FAS	rs1800682 (G-670A)	European	[45,46]
6q23.1	CTGF	rs6918698 (G-945C), rs9399005	European, Japanese	[37–40]
1q22-23	CD247	rs2056626	European	[34]

The identification of the association of connective tissue growth factor (CTGF) gene with risk to SSc provides one of the most conflicting results to date. The potential functional −945 G allele of CTGF gene was primarily associated with susceptibility to SSc in Europeans and Japanese [37]. However, three additional studies in Europeans failed to replicate the association [38–40]. A recent study suggests that variant rs9399005 in the 3′UTR region of the CTGF gene is associated with both subtypes of SSc [39]. The lack of replication of the CTGF polymorphisms in three large cohorts of SSc suggests that CTGF may not be a strong genetic determinant for SSc susceptibility.

Genes involved in B cell receptor signaling contribute to SSc susceptibility. Association with diffuse SSc (dcSSc) has been identified with BANK1 [41]. The C8orf13-BLK region has been associated in European and Japanese patients with SSc [42,43]. These findings suggest an important role of B-cells in the pathogenesis of SSc.

A large multicenter study in Europeans found two polymorphisms (rs11650354 and rs17699436) in the TBX21 gene associated with SSc [44]. In addition, they also showed a gene-gene interaction between the TBX21 and STAT4.

Some others genes that appear to play a role in susceptibility to SSc are FAS, TNFSF4 and CD247. The promoter rs1800682 polymorphism in the FAS gene is a confirmed susceptibility variant for SSc in different populations [45,46]. In addition, the SLE susceptibility gene TNFSF4 [11,12,14], was associated with SSc in a large case-control study [47]. The CD247 gene, which encodes the T-cell receptor zeta (CD3ζ) subunit, was a new susceptibility gene for SSc in a GWAS and this association has been confirmed in an independent cohort [34].

Conclusions

The genetics results deriving from GWAS are just the beginning of a new era of research, but new insight on disease genetics has been acquired. First, the risk alleles identified for these diseases explain only between 5–15% of the whole genetic contribution to disease and with odds ratios ranging from 1.01 to 2.4 at the most. Clearly, gene-gene interactions, gene-environment interactions and other genomic structural variation such as copy number variation, the role of rare variants and epigenetics need to be adressed. In this regard, the recent identification of rare, but highly penetrant mutations in the sialic acid acetylestarase gene (SIAE) involved in SLE and RA susceptibility and with an important functional impact in the gene [48] suggests that rare variants may have an important role not yet fully comprehended. Second, most autoimmune diseases, in particular systemic autoimmune diseases share several susceptibility genes. The differences seem to reside on the contribution of each gene in each disease. HLA alleles have been known to be different from disease to disease, but the risk alleles for non-MHC genes appear, until now to be the same. However while IRF5 and STAT4 are prominent genes in SLE, PTPN22 and TNFAIP3 are major genes in RA and CD247 in SSc. Very few genes are unique for each disease (Figure 1) and whether some genes may predominate in individuals with certain clinical manifestations is still a difficult nut to crack. Phenotyping of samples studied in genetics lack the detail required to define the correlation between clinical maifestations and genes. One possible exception is ITGAM in SLE, a unique gene for lupus importantly associated with kidney disease. Third, studies in different populations are important. Differences and similarities will lead to a comprehensive picture of genetic susceptibility, but may also pave the way to the very needed studies on gene-environmental interactions in autoimmune disease, a theme about which we know practically nothing.

Figure 1.

Unique and shared genes between SLE, RA and SSc

Genetics studies are a starting point to cell biology and immunology studies aimed at understanding disease pathogenesis and the influence of susceptibility genes on cell function. Indeed, we have challenging and exciting times ahead of us.