Advances in lupus genetics and epigenetics

Abstract

Purpose of review

Genome-wide association studies (GWAS) have identified more than 50 robust loci associated with SLE susceptibility, and follow-up studies help reveal candidate causative genetic variants and their biological relevance contributing to the development of SLE. Epigenetic modulation is emerging as an important mechanism for understanding how the implicated genes interact with environmental factors. We review recent progress towards identifying causative variants of SLE-associated loci and epigenetic impact to lupus, especially genetic-epigenetic interactions that modulate expression levels of SLE susceptibility genes.

Recent findings

A few SLE-risk loci have been refined to localize likely causative variants responsible for the observed GWAS signals. Few of such variants disrupt coding sequences resulting in gain or loss of function for the encoded protein, while most fall in noncoding regions with potential to regulate gene expression through alterations in transcriptional activity, splicing, mRNA stability and epigenetic modifications. Multiple key pathways related to the SLE pathogenesis have been indicated by the identified genetic risk factors, including type I interferon signaling pathway that can also be regulated by epigenetic changes occurred in SLE.

Summary

These findings provide novel insights of the disease pathogenesis, and promise better diagnostic accuracy and new therapeutic targets for patient management.

INTRODUCTION

The GWAS approach based on its ability to screen hundreds of thousands of single nucleotide polymorphisms (SNPs) across the genome in a hypothesis-free manner has accelerated the discovery of genetic variations contributing to systemic lupus erythematosus (SLE). Eight GWAS in SLE (four in European-derived [1–4] and four in Asian populations [5–8]) and subsequent meta-analysis and large-scale replication studies have identified a rapidly expanding number of risk loci exceeding the genome-wide significance level (P<5×10⁻⁸) that are located within or near genes encoding products contributing to the pathogenesis of SLE. In some cases, follow-up fine mapping using increased dense markers, imputation and target resequencing help localization of candidate causative variants that explain observed GWAS signals. Only few of such variants disrupt protein-coding sequences and most lie in noncoding regions of the genome. The emergence of genomic and epigenomic datasets, such as eQTL (expression quantitative-trait loci) and ENCODE (Encyclopedia of DNA Elements) [9], have revolutionized functional annotation of noncoding variants with potential impact on gene regulation through alterations in transcriptional activity, splicing, mRNA stability and epigenetic modifications. Here, we discuss non-HLA SLE-associated loci grouped by biological pathways, highlighting those that have causative variants implicated with their functional roles experimentally investigated.

Environmental exposures of genetically-predisposed individuals may modify SLE risk genes epigenetically, resulting in the development of SLE manifestations. Epigenetic processes refer to heritable modifications that regulate gene expression and alter cellular functions without changes in the genomic sequence. Commonly described epigenetic mechanisms include DNA methylation, histone modification and microRNAs (miRNAs). A recent study has revealed epigenomic elements enriched in the promoters of autoimmunity susceptibility genes [10^▪▪], suggesting that epigenetic modifications may amplify genetic risks for autoimmune diseases. We summarize important findings in epigenetic changes and highlight genetic-epigenetic interactions that operate in variants associated with SLE susceptibility.

SLE SUSCEPTIBILITY GENES STRATIFIED BY BIOLOGICAL PATHWAYS

Non-HLA SLE-associated genes are summarized in Table 1. We discuss genes that have the most likely causative variants implicated and functional consequences relevant to the disease pathogenesis.

Table 1.

Non-HLA loci associated with SLE in the disease pathways^*

Pathway	Chr	Gene	OR	Population	Likely Causative Variant (SLE-risk allele)	Ref
Type I IFN signaling
	2q24	IFIH1	1.1–1.4	EA,AA	rs13023380 (A allele) results in decreased IFIH1 transcript levels; rs1990760 (A946T; A allele) and rs10930046 (H460R; A allele) confer increased apoptosis and elevated inflammation-related gene expression.	11▪▪, 12
	2q32	STAT4	1.4–1.8	EU,EA,AS,HS,AA		3, 5–6, 13–14
	5q34	miR146a	1.3	AS		15
	7q32	IRF5	1.3–1.9	EU,EA,AA,AS,HS	Four functional SNPs define risk haplotypes associated with increased expression of IRF5 and IFN-α.	16–21
	11p15	IRF7	1.3–2.0	EU,EA,AA,AS	rs1131665 (Q412R; A allele) confers increased IRF7 and downstream IFN pathway activation.	3, 20, 22
	12q24.32	SLC15A4	1.1–1.3	EA,AS		5, 20, 23
	16q24	IRF8	1.2–1.3	EU,EA		12, 23–24
	19p13	TYK2	1.3	EA		12
	Xp22	TLR7	1.2–2.3	AS,EA,AA,HS	rs3853839 (G allele) confers increased TLR7 expression and IFN response.	25, 26▪▪
NFκB signaling
	5q33.1	TNIP1	1.2–1.4	EA,AS		5, 20
	6q23	TNFAIP3	1.7–2.3	EU,EA,AS	TT>A polymorphic dinucleotide (deletion T followed by A transversion) with decreased NFκB binding to the promoter attenuates TNFAIP3 expression.	4–5, 27▪▪
	22q11.21	UBE2L3	1.2–1.4	EU,EA,AS		3, 5, 20
	Xq28	IRAK1/MECP2	1.1–1.6	EA,AS,AA,HS	rs1059702 (S196F; A allele) tags a risk haplotype and confers increased NFκB activity.	28▪▪, 29▪▪
B and T cell signaling
	1p13.2	PTPN22	1.4–2.4	EU,HS	rs2476601 (R620W; T allele) alters TCR and BCR signaling, and interferes with the removal of developing autoreactive B cells.	30–34
	1q25	TNFSF4	1.2–1.5	EU,EA,AS,AA,HS		3, 5, 35
	1q31-q32	IL10	1.2–1.3	EU,EA	rs3122605 (G allele) tags a risk haplotype associated with increased IL10 expression.	20, 36▪▪
	2p25-p24	RASGRP3	1.2–1.4	AS,EU		5, 37
	3q13	CD80	1.3	AS		38▪
	4q21	AFF1	1.2	AS		7
	4q24	BANK1	1.2–1.4	EU,EA,AS,AA	rs10516487 (R61H; G allele) and rs3733197 (A383T; G allele) encode mutant BANK1; rs17266594 (T allele) alters splicing efficiency of BANK1 transcript isoforms.	2, 6, 14
	4q26-q27	IL21	1.1–1.6	EA,AA		39
	6q21	PRDM1	1.2	EA		20
	7p12.2	IKZF1	1.2–1.4	EU,AS	rs4917014 (T allele) alters IKZF1 levels in cis and regulates expression of C1QB and five type I IFN response genes in trans.	5, 12, 40▪▪
	8p23	BLK	1.2–1.6	EU,EA,AS,AA	rs922483 (T allele) and tri-allelic SNP rs1382568 (A, C alleles) reduce BLK promoter activity.	1, 3, 5, 41▪▪
	8q13	LYN	1.2–1.3	EU		3
	10q21	ARID5B	1.2	AS		38▪
	11p13	PDHX/CD44	1.2–1.4	EA,AA,AS		42
	11q23.3	ETS1	1.2–1.4	AS,EU		5, 6, 37
	13q13	ELF1	1.3	AS		43
	15q24.1	CSK	1.3	EU	rs34933034 (A allele) is associated with increased CSK expression, Lyn phosphorylation, BCR-mediated activation of mature B cells and expansion of transitional B cells.	44
	16p11.2	PRKCB	1.2	AS		45
	17q21	IKZF3	1.2–1.9	EU,AA		24
Neutrophil and monocyte function
	19p13	ICAM1/4/5	1.2	EU,AA,HS,AS		46
Immune complex clearance
	1q23	FCGR2A	1.3–1.4	EU,EA,AA,AS	rs1801274 (H166R; T allele)	reviewed in 47
		FCGR3A	1.2–1.5	EU,AA	rs396991 (F176V; T allele)
		FCGR2B	1.3–2.5	AS	rs1050501 (I232T; C allele)
		FCGR3B	1.7–2.3	EU,AA	decreased copy number
	16p11.2	ITGAM	1.3–2.1	EA,EU,AA,AS,HS	rs1143679 (R77H; A allele) impairs leukocyte phagocytosis	1, 3, 47–50
NADPH oxidase-dependent ROS pathway
	1q25	NCF2	1.2–2.8	EA,AA,HS	rs17849502 (H389Q; A allele) confers decreased NADPH oxidase activity and ROS production	51–52
Others
	2p13	TET3	1.3	AS		38▪
	3p14.3	PXK	1.2–1.3	EU,EA		3, 20
	3q13.33	TMEM39A	1.2–1.4	EU,AA,AS		24
	5q35	PTTG1	1.2	EU		3
	6p21	UHRF1BP1	1.2–1.5	EU,AS		20, 53
	6q21	ATG5	1.2–1.3	EU,AS		3, 5
	7p15.2	JAZF1	1.2	EA		20
	8p23.1	XKR6	1.2–1.3	EU		3
	10q11.23	WDFY4	1.2–1.3	AS		6
	12p13	CDKN1B	1.2	AS		38▪
	12q23	DRAM1	1.2	AS		38▪
	16p13.13	CLEC16A	1.2	EU,AS		20, 54

^*These loci are identified through GWAS, meta-analysis, candidate gene studies or replication studies. All achieve p<5×10⁻⁸ in at least one study.

Abbreviation: AA, African American; AS, Asian; EA, European American; EU, European; HS, Hispanic.

ARID5B, AT rich interactive domain 5B; AFF1, AF4/FMR2 family member 1; ATG5, autophagy related 5; BANK1, B-cell scaffold protein with ankyrin repeats 1; BLK, B lymphoid tyrosine kinase; CD44, CD44 molecule; CD80, CD80 molecule; CDKN1B, cyclin-dependent kinase inhibitor 1B; CLEC16A, C-type lectin domain family 16 member A; CSK, c-src tyrosine kinase; C1QB, complement component 1 q subcomponent B chain; DRAM1, DNA-damage regulated autophagy modulator 1; ELF1, E74-like factor 1; ETS1, v-ets avian erythroblastosis virus E26 oncogene homolog 1; FCGR2A, Fc fragment of IgG low affinity IIa receptor; FCGR3A, Fc fragment of IgG low affinity IIIa receptor; FCGR2B, Fc fragment of IgG low affinity IIb receptor; FCGR3B, Fc fragment of IgG low affinity IIIb receptor; ICAM1, intercellular adhesion molecule 1; ICAM4, intercellular adhesion molecule 4; ICAM5, intercellular adhesion molecule 5; IFIH1, interferon induced with helicase C domain 1; IKZF1, IKAROS family zinc finger 1; IKZF3, IKAROS family zinc finger 3; IL10, interleukin 10; IL21, interleukin 21; IRF5, interferon regulatory factor 5; IRF7, interferon regulatory factor 7; IRF8, interferon regulatory factor 8; IRAK1, interleukin-1 receptor-associated kinase 1; ITGAM, integrin alpha M; JAZF1, JAZF zinc finger 1; LYN, v-yes-1 Yamaguchi sarcoma viral related oncogene homolog; MECP2, methyl CpG binding protein 2; miR146a, microRNA 146a; NCF2, neutrophil cytosolic factor 2; PDHX, pyruvate dehydrogenase complex component X; PRDM1, PR domain containing 1 with ZNF domain; PRKCB, protein kinase C beta; PTPN22, protein tyrosine phosphatase non-receptor type 22; PTTG1, pituitary tumor-transforming 1; PXK, PX domain containing serine/threonine kinase; RASGRP3, RAS guanyl releasing protein 3; SLC15A4, solute carrier family 15 (oligopeptide transporter) member 4; STAT4, signal transducer and activator of transcription 4; TET3, tet methylcytosine dioxygenase 3; TLR7, toll-like receptor 7; TMEM39A, transmembrane protein 39A; TNFAIP3, tumor necrosis factor alpha-induced protein 3; TNIP1, TNFAIP3 interacting protein 1; TNFSF4, tumor necrosis factor superfamily member 4; TYK2, tyrosine kinase 2; UBE2L3, ubiquitin-conjugating enzyme E2L3; UHRF1BP1, UHRF1 binding protein 1; WDFY4, WDFY family member 4; XKR6, XK Kell blood group complex subunit-related family member 6.

Type I interferon (IFN) pathway

Integration of data from GWAS, gene expression microarrays and serologic studies supports that dysregulation of type I IFN severs as a central driver of SLE pathogenesis [55]. The diverse mechanisms by elevated type I IFN affecting SLE development include promoting differentiation of monocytes and plasmacytoid dendritic cells, activation of autoreactive B and T cells, production of autoantibodies and expression of proinflammatory cytokines and chemokines [56]. Over half of the currently identified SLE susceptibility genes encode proteins that can be directly or indirectly linked to type I IFN production or signaling [55].

Binding of immune complexes containing self-antigens and nucleic acids by pattern recognition receptors, such as TLR7, is a likely major trigger of type I IFN production in SLE. The importance of TLR7 upregulation in mediating autoimmune responses has been clearly demonstrated in murine lupus models [57–59]. Our group has identified a functional SNP at 3′untranslated region (UTR) exhibiting consistent association with SLE in four major ancestries [25,26^▪▪]. This SNP maps within a binding site for miRNA-3148 conferring allelic modulation of TLR7 expression. The SLE-risk allele confers decreased degradation of TLR7 transcripts, resulting in elevated TLR7 levels and heightened downstream IFN response [26^▪▪].

IRF5, IRF7 and IRF8, a family of transcription factors downstream of endosomal TLRs, are required for activating transcription of IFN-α and IFN-inducible genes. Genetic variants in or near these three genes have been associated with SLE susceptibility [3,16–20,22,24]. In particular, SLE-associated variants of IRF5 and IRF7 have functional impact on increased serum IFN-α and such impact depends on the presence of specific autoantibodies [21,60]. A cis-eQTL SNP located within the SLE-associated haplotype not only is associated with IRF7 expression but also confers trans-eQTL effect on regulating type I IFN response in activated but not unstimulated dendritic cells [61^▪▪], highlighting the importance for using appropriate activated ex vivo cells to explore functional roles of the disease-associated variants.

Other genes related to the type I IFN pathway that are associated with increased risk for SLE include IFIH1 [11^▪▪], STAT4 [3,5,13,14] TYK2 [12] and SLC15A4 [5,20,23]. IFIH1, a cytosolic sensor of dsRNA, promotes IRF7/3 phosphorylation and type I IFN production. Three independent variants at IFIH1 (one intronic and two missense SNPs) account for genetic association in multiple ancestries [11^▪▪]. Risk alleles of both missense variants confer increased apoptosis and elevated expression of inflammation-related genes. The intronic risk allele leads to decreased IFIH1 transcript levels by disruption of binding to protein complex including nucleolin and lupus autoantigen Ku70/80, which would be expected to promote autoantibody generation.

Nuclear factor κB (NFκB) pathway

Genes that function in the NFκB pathway downstream of TLR engagement have been associated with increased SLE risk in multiple ancestries. For example, TNFAIP3 encodes a deubiquitinating enzyme (A20) that participates in the termination of NFκB signaling. A pair of tandem polymorphic dinucleotides (TT>A), downstream of the TNFAIP3 promoter, have been nominated as causal variants responsible for disease association with TNFAIP3 [27^▪▪]. The SLE-associated TT>A risk alleles with inefficient delivery of NFκB to the TNFAIP3 promoter via long-range DNA looping attenuate A20 expression, leading to enhanced NF-κB pathway activity that contributes to SLE. Located on the X chromosome, the S196F variant of IRAK1 captures the association signals of a SLE risk haplotype shared by multiple ancestries [28^▪▪,29^▪▪]. The risk 196F allele confers increased NFκB activity and is associated with decreased mRNA levels of MECP2, but not IRAK1, suggesting contributions of both IRAK1 and MECP2 to SLE susceptibility [28^▪▪,62]. Studies in murine lupus models indicate a pivotal role of Irak1 in activation of NFκB and induction of IFN-α/γ [63]. MECP2 is a transcriptional regulator involved in modulating expression of methylation-sensitive genes [29^▪▪]. Functional consequences of the MECP2 risk haplotype will be discussed in DNA methylation section.

B and T cell signaling

Multiple loci that mediate signaling transduction in B and T cells are associated with SLE, supporting a central role of dysregulated lymphocytes in the pathogenesis of SLE.

The R620W SNP of PTPN22 is a well-characterized functional variant associated with multiple autoimmune diseases including SLE. Both humans carrying the risk 620W allele and knockin mice expressing the analogous 619W mutation show altered T-cell receptor (TCR) and BCR signaling as well as enhanced B cell autoreactivity, illustrating the consequences of the risk variant in modulating T and B cell tolerance and autoimmunity [30–34]. Efforts to elucidate the role of R620W in nonlymphocytes have revealed that the risk 620W allele is associated with reduced TRAF3 (TNF receptor-associated factor 3) autopolyubiquitination and diminished TLR-induced type I IFN production in myeloid cells [64], but with enhanced functions in neutrophils including increased transendothelial migration, Ca²⁺ release and ROS (reactive oxygen species) production [65 ]. Taken together, this PTPN22 polymorphism affects multiple signaling pathways in diverse immune cell types contributing to autoimmunity.

CSK encoding c-Src tyrosine kinase can regulate activation of Src kinases (e.g. LYN) by physically interaction with LYP (encoded by PTPN22) in lymphocytes. Association of CSK with SLE identified in European population has been localized to an intronic SNP as causal [44]. The SLE-risk allele is associated with high CSK expression, increased phosphorylation of Lyn, enhanced BCR-mediated activation of mature B cells and expansion of transitional B cells, suggesting CSK involved in multiple development stages of B cells [44]. Integrating gene-expression data with SLE-associated loci has implicated transitional B cells to be appropriate for investigating functional consequences of risk alleles [66], such as CSK.

BANK1, encoding an adaptor/scaffold protein, facilitates the release of intracellular calcium and alter the B cell activation threshold [2]. Three functional BANK1 variants including one intronic and two missense SNPs have been associated with SLE in multiple ancestries, which contributes to sustained BCR signaling and B cell hyperactivity characteristic of SLE [2]. BLK encodes a member of the Src family kinases functioning in intracellular signaling and the regulation of B cell proliferation, differentiation and tolerance [1]. Risk alleles of the BLK promoter variants account for genetic associations in multiethnic populations, and reduce promoter activity in B cell lines representing different B cell developmental stages, which suggests that decreased BLK expression may affect B cell development and impair functional responses in B cells [41^▪▪].

IL-10 is a pivotal cytokine which inhibits T cells and antigen-presenting cells while enhances B cell survival and activity [67]. Our group has identified an IL10 upstream SNP that tags a risk haplotype associated with SLE in European population [36^▪▪]. The SLE-risk allele is associated with increased IL10 expression by preferentially binds to the phosphorylated transcription factor Elk-1. Interestingly, SLE patients with active disease show enhanced phosphorylated Elk-1 in nuclei and elevated co-expression of IL-10 and phosphorylated Elk-1 in peripheral lymphocytes, suggesting the involvement of aberrant Elk-1 activation in development of SLE.

IKZF1 regulates lymphocyte differentiation, proliferation and BCR signaling [68]. The first evidence supporting IKZF1 as SLE susceptibility gene comes from a Chinese GWAS [5]. Recently, a trans-eQTL study has revealed the SLE-associated SNP at upstream of IKZF1 that alters the IKZF1 levels in cis and regulates expression of C1QB and five type I IFN response genes in trans, thereby reinforcing the importance of IKZF1 in SLE [40^▪▪]. This study illustrates how trans-eQTL mapping helps yield insights into downstream effects of the disease-associated variants.

Other genes with roles in B and T cell functions associated with SLE susceptibility are summarized in Table 1, and most of them await delineation of functional mechanisms by which the associated variants affect disease risk.

Immune complex (IC) clearance

Defective clearance of ICs and apoptotic cells leads to initiation and maintenance of autoimmune responses and chronic inflammation in SLE. Polymorphisms in multiple members of this pathway have been associated with SLE, for example missense variants of FCGR2A/FCGR3A and ITGAM and decreased copy number variations of FCGR3B (reviewed in [47]). CD11b encoded by ITGAM together with CD18 form the complement receptor 3 (CR3) with functions in binding to tissue deposited ICs as well as leukocyte phagocytosis, apoptosis, adhesion and migration via interaction with a range of ligands, such as ICAM1 that has also show association with SLE susceptibility [46]. The SLE-risk allele of ITGAM which encodes an amino acid change from Arg to His at position 77 (R77H) in CD11b impairs the CR3-mediated phagocytosis in monocytes, neutrophils and macrophages [48–50]. Given its location within an active chromatin region in ITGAM, the SLE-risk 77H allele confers reduced transcriptional enhancer activity and is associated with decreased ITGAM expression at both mRNA and protein levels in monocytes [69].

NADPH oxidase-dependent ROS pathway

ROS produced by NADPH (nicotinamide adenine dinucleotide phosphate) oxidase complex are essential in defense against invading pathogens and are likely to promote cellular and tissue damage owing to their highly reactive nature [70]. Considering the pro-inflammatory effect, ROS have been implicated in promoting autoimmune diseases, such as SLE, rheumatoid arthritis (RA), thyroiditis and type 1 diabetes [71^▪, 72–74]. A surprising finding of a polymorphism in the rat Ncf1 gene (encoding a subunit of NADPH oxidase complex) leading to low NADPH oxidase activity and reduced ROS production associated with severe arthritis in rodents provides new insight of an anti-inflammatory role for ROS involved in autoimmune diseases [75,76]. In fact, increasing evidence supports that ROS are not only proinflammatory byproducts of cellular responses to infectious or inflammatory stimuli but act as fine-tuning regulators of autoimmune responses varying in time, location and extent of the ROS production (reviewed in [77]). Consistently, a mutation (H389Q) in human NCF2 gene (encoding another subunit of NADPH oxidase complex) has been identified to associate with SLE risk in European population, which is also accompanied by decreased NADPH oxidase activity and ROS production [51]. A large-scale trans-ancestral replication study has confirmed this mutation with SLE susceptibility in African-American and Hispanic populations but not in Koreans (due to the absence of minor allele), and identified additional independently associated variants at NCF2 in all of the three non-European populations [52].

Other Loci

Other pathways implicated by genetic findings have potential relevance to the SLE pathogenesis, including cell-cycle regulation (CDKN1B), autophagy (ATG5 and DRAM1) and DNA demethylation (TET3); however, further functional characterization is required [38 ]. In addition, SLE-associated loci (e.g. TMEM39A, UHRF1BP1 and CLEC16A) with unknown immune function require further fine mapping and characterization of functional pathways contributing to the development of SLE [24,53,54].

CONNECTING GENOTYPE TO PHENOTYPE IN SLE

Recent genotype-phenotype association studies provide insights into genetic effects on clinical subphenotypes. Multiple established SLE-risk loci are associated with anti-dsDNA autoantibody (e.g. HLA-DR2/DR3, STAT4, IRF5/IRF7, IFIH1, ITGAM and UBE2L3) [78] and/or renal disorders (e.g. ITGAM, STAT4, TNFSF4, TNFAIP3 and TNIP1) [79,80]. In addition, the apolipoprotein L1 nephropathy risk alleles G1/G2 strongly impact the risk and the time of end-stage renal disease (ESRD) in African Americans with lupus nephritis (LN), which may partly explain the increased risk of LN-ESRD in African Americans due to the high frequency of these alleles in African Americans but near absence in European Americans [81]. An analysis of specific manifestations with 22 validated SLE-susceptibility loci has showed cumulative genetic associations with age at diagnosis, anti-dsDNA autoantibody, oral ulcers, immunological and hematologic disorders, single marker association with renal disorders and arthritis, and no genetic association with malar/discord rash, photosensitivity, serositis and neurological disorder [82]. Such measures will be improved as more new SLE-risk loci are emerging, which help enhance the link between genotype and phenotype in SLE.

EPIGENETICS IN SLE

The disease concordance rate of SLE in monozygotic twins (24–58%) suggests epigenetic contribution to the disease development. Dysregulated epigenetic modifications, including DNA methylation, histone modification and noncoding RNA, can affect the expression and function of genes participating in pathogenic processes of SLE.

DNA hypomethylation

DNA methylation usually occurs at the 5′ position of cytosine in the context of CpG dinucleotides distributed in genomes exhibiting suppressive effects on gene expression. A role of DNA hypomethylation in SLE has been demonstrated in studies showing demethylating drugs-induced T cell autoreactivity and lupus-like symptoms in mice and impaired DNA methylation in CD4⁺ T cells from active SLE patients (reviewed in [83]). Many methylation-sensitive genes are overexpressed in SLE CD4⁺ T-cells, functionally contributing to SLE development, including CD11A causing autoreactivity, perforin increasing apoptosis, CD70 and CD40L enhancing B-T cell interaction for autoantibodiy production (reviewed in [84]). Genome-wide DNA methylation analysis have showed severe hypomethylation of type I IFN-regulated genes in SLE CD4⁺ T cells, CD19⁺ B cells and CD14⁺ monocytes, suggesting epigenetically-mediated type I IFN hyper-responsiveness in SLE patients [85^▪▪,86]. Interestingly, healthy individuals carrying the SLE-risk IRAK1-MECP2 haplotype (mentioned in the NFκB pathway section) with increased levels of a specific MECP2 transcript isoform in stimulated T-cells showed a similar hypomethylation of IFN-regulated genes, providing evidence for genetic-epigenetic interaction associated with SLE-risk variants [29^▪▪]. How DNA hypomethylation occurs in SLE patients are discussed in a recent review [87^▪].

Histone modification changes

Various combinations of the histone modifications, including acetylation, phosphorylation, methylation, ubiquitylation and citrullination of histone tails, can affect DNA replication, transcription and chromatin structure [88]. In some cases, specific histone modifications associated with gene expression are defined, such as H3 and H4 hyperacetylation or H3 trimethyl-lysine4 (H3K4me3) present in many active genes, while H3 trimethyl-lysine9 or trimethyl-lysine27 in repressed genes [89]. Changes in histone modifications have been reported in SLE cells, leading to aberrant gene expressions that may contribute to SLE pathogenesis. For example, monocytes from SLE patients are characterized by global H4 hyperacetylation in parallel with increased expression of genes involved in type I IFN and NFκB pathways [90]. By integrating genetic and epigenetic datasets, active histone marks are enriched in the promoters of SLE- and RA-susceptibility genes in EBV-transformed lymphoblastoid cell lines, suggesting a regulatory role of epigenomic elements for functions and cellular specificity of these autoimmune-associated genes [10^▪▪,91^▪▪].

MiRNA dysregulation

MiRNAs representing a large class of small noncoding RNAs regulate gene expression by targeting specific mRNAs for degradation or translation inhibition [92]. Increasing evidence supports miRNAs playing an essential role in the development, homeostasis and function of innate and adaptive immunity that dysregulation of these processes may induce autoimmunity and inflammation (reviewed in [93]). Studies of miRNA expression profiles in blood cells, plasma and target tissues from SLE patients have revealed unique miRNA signatures when compared with healthy individuals and dysregulation of miRNAs associated with disease activity and major organ involvement, suggesting miRNAs as biomarkers for SLE patients [94^▪▪,95^▪]. These dysregulated miRNAs are mainly involved in three types of biological processes relevant to the SLE pathogenesis (Table 2): (1) hyperactivation of type I IFN pathway (e.g. miR-146a) [96], (2) exacerbation of inflammatory responses by increasing cyto/chemokines secretion (e.g. miR-125a, miR-23b) [97,98], and (3) reduction of DNA methylation by directly or indirectly inhibiting DNA methyltransferase 1 (e.g. miR-126, miR-148a, miR-21) [101,102]. Of interest, the SLE-risk allele of miR146a promoter variant confers lower levels of its transcript due to decreased binding of transcription factor Ets-1 [15]. Given of the great potential for variation in miRNA target sites, some SLE-associated SNPs located in 3′UTR could regulate gene expression through introducing or abolishing miRNA binding sites (e.g. TLR7) [26^▪▪], providing a miRNA-based genetic-epigenetic interaction for dissecting functional consequences of SLE-associated variants.

Table 2.

MiRNA dysregulation in SLE

Function	miRNA	Expression in SLE patients	Target gene	Ref
Hyperactivation of type I IFN pathway
	miR-146a	downregulated in PBMCs	IRAK1, TRAF6, IRF5, STAT1	96
Aberrant cyto/chemokines release
	miR-125a	downregulated in PBMCs	KLF13	97
	miR-23b	downregulated in kidney tissue	TAB2, TAB3, CHUK	98
	miR-21	upregulated in CD4+ T cells	PDCD4	99
	miR-31	downregulated in CD4+ T cells	RHOA	100
DNA hypomethylation
	miR-126	upregulated in CD4+ T cells	DNMT1	101
	miR-21	upregulated in CD4+ T cells	RASGRP1	102
	miR-148a	upregulated in CD4+ T cells	DNMT1	102

Abbreviation: CHUK, conserved helix-loop-helix ubiquitous kinase; DNMT1, DNA methyltransferase 1; IRF5, interferon regulatory factor 5; IRAK1, interleukin-1 receptor associated kinase 1; KLF13, Kruppel-like factor 13; PDCD4, programmed cell death 4; RASGRP1, RAS guanyl releasing protein 1; RHOA, ras homolog family member A; STAT1, signal transducer and activator of transcription 1; TAB2, TGF-β activated kinase1/MAP3K7 binding protein 2; TAB3, TGF-β activated kinase1/MAP3K7 binding protein 3; TRAF6, tumor necrosis factor receptor-associated factor 6.

CONCLUSION

GWAS and follow-up studies have rapidly advanced our understanding of SLE genetics. Future challenges include discovery of new associations responsible for the missing heritability of SLE and identification of causal variants with functional characterization at each locus. It is likely that genetic variants may not explain all variations in gene functions contributing to SLE. Identifying epigenetic modulation and genetic-epigenetic interaction will unravel new mechanisms to understand how SLE susceptibility genes affect disease risk and phenotypic expression. Integration of genetic knowledge with diverse epigenomic and biological studies has already guided drug discovery for RA [103^▪▪], and such systematic strategy will be applied to SLE in the near future.

KEY POINTS.

• Over 50 loci associated with SLE susceptibility have been identified across multiple ancestries, and most encode gene products participating in the key pathways relevant to SLE pathogenesis.

• Few of candidate causative variants are located in the coding sequences to affect protein functions, whereas many reside within noncoding regions to regulate gene expression through transcriptional and posttranscriptional mechanisms.

• Epigenetic changes such as DNA methylation, histone modification and miRNA regulation in SLE represent an important layer to link genetics and gene expression with disease risk.

• Integration of genetic knowledge into diverse biological studies promises to identify specific immune cells and important (or novel) pathways that present therapeutic targets for SLE drug development.