Epigenetic changes: The missing link

Best Practice & Research Clinical Rheumatology xxx (2014) 1e11

Contents lists available at ScienceDirect

Best Practice & Research Clinical Rheumatology journal homepage: www.elsevierhealth.com/berh

4

Epigenetic changes: The missing link Diego Kyburz a, b, c, *, Emmanuel Karouzakis c, Caroline Ospelt c a

Division of Rheumatology, University Hospital of Basel, Basel, Switzerland Department of Biomedicine, University of Basel, Basel, Switzerland c Center of Experimental Rheumatology, University Hospital of Zurich, Zurich, Switzerland b

a b s t r a c t Keywords: Epigenetic microRNA miR Methylation Acetylation HDAC Histone Environment Rheumatoid arthritis

The association of rheumatoid arthritis (RA) with a number of genetic risk loci is well established; however, only part of the risk to develop the disease is based on genetics. Environmental factors significantly contribute to the pathogenesis. A geneeenvironment interaction for smoking and certain major histocompatibility complex (MHC) class II alleles has been shown to promote anticitrullinated protein antibody (ACPA)-positive RA; however, the molecular mechanisms of interaction remain unclear. In contrast to the genetic background, epigenetic factors are responsive to external stimuli and can modulate gene expression. Therefore, epigenetic mechanisms may function as intermediaries between genetic risk alleles and environmental factors. In this review, epigenetic mechanisms are explained and the evidence for epigenetic changes relevant for the pathogenesis of RA and potential therapeutic applications are discussed. © 2014 Elsevier Ltd. All rights reserved.

Introduction Sequencing technology has allowed to perform whole genome sequencing approaches to determine the genetic basis of rheumatoid arthritis (RA). A number of new genes were found to be associated with the disease, in addition to the long-known association with HLA-DR genes. However, considering a concordance rate in monocygotic twins of only 15% it follows that only a part of

* Corresponding author. Division of Rheumatology, University Hospital of Basel, Petersgraben 4, 4031 Basel, Switzerland. Tel.: þ41 61 265 90 20; fax: þ41 61 265 90 21. E-mail address: [email protected] (D. Kyburz).

http://dx.doi.org/10.1016/j.berh.2014.10.014 1521-6942/© 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Kyburz D, et al., Epigenetic changes: The missing link, Best Practice & Research Clinical Rheumatology (2014), http://dx.doi.org/10.1016/j.berh.2014.10.014

2

D. Kyburz et al. / Best Practice & Research Clinical Rheumatology xxx (2014) 1e11

the disease risk can be attributed to genetic factors [1]. On the other hand, several environmental factors were identified to impact the disease risk. Among those, smoking is of particular interest, as a geneeenvironment interaction was demonstrated in anti-citrullinated protein antibody (ACPA)positive patients with RA. Smoking greatly increased the disease risk when the shared epitope of HLA-DRB1 was present [2]. How smoking and shared epitope are molecularly linked is unknown. In contrast to genetic risk alleles, epigenetic marks are sensitive to external factors. Therefore, environmental exposures may potentially impact the function of the immune system by modulating epigenetic modifications. The term “epigenetics” denotes heritable changes in gene function without alterations of the DNA sequence. These include histone modifications such as methylation and acetylation, as well as DNA methylation. The gene expression profile of a cell is critically dependent on DNA and histone modifications. In a broader sense, also noncoding RNA are included in the definition of epigenetics. In this review, epigenetic changes of cells are described and their role in the disease pathogenesis of RA is discussed, with focus on the implications on geneeenvironment interactions. Epigenetic control of gene expression In the past 10 years, epigenetics research revealed exciting new pathways in the regulation of gene expression. The biochemical modification of DNA and histones was found to have an important role in gene expression and human disease [3]. DNA methylation is often associated with transcriptional repression in CpG-rich genomic regions [4]. Two major mechanisms have been proposed. Firstly, the direct inhibition of transcription factor binding and secondly the recognition of the methylated regions by specific proteins that bind to methylated DNA (MeCP2, MBD1, MBD2, MBD3, MBD4) and inhibit the access of DNA polymerases and transcription factors. Another field of epigenetics is the study of chromatin biology [5]. Chromatin consists of DNA wrapped around a complex protein network of histone proteins that form the nucleosome. The nucleosome is an octamer consisting of the histones H2A, H2B, H3 and H4 around which DNA is packed in a tight conformation. Posttranslational modification of histones often occurs at specific transcriptionally active or repressed sites in the genome. Histone modifications alter the structural conformation of nucleosomes and allow access of transcriptional activators. DNA methylation and histone modifications The addition of a methyl group to the cytosine base pair is known as DNA methylation and is catalysed by DNA methyltransferases. Three DNA methyltransferases, DNMT1, DNMT3A and DNTM3B, have been associated with DNA methylation in humans. DNMT1 is involved in the maintenance of DNA methylation during somatic cell DNA replication. It interacts with the Ubiquitin-like, containing PHD and RING finger domains 1 (UHRF 1) protein and targets hemimethylated 50 CpG 30 DNA sequences [6]. The DNMT1/UHRF1 complex transfers a methyl group from S-adenosylmethionine to the unmodified cytosine in the opposite DNA strand. DNMT3A and DNMT3B are called the de novo methyltransferases. During embryonic development, DNA methylation undergoes extensive demethylation [7]. De novo methyltransferases are mainly involved in the restoration of DNA methylation during cell development. A characteristic of the human genome is the presence of DNA sequences that are CpG rich and remained unmethylated. These are called CpG islands and often associate with promoter and transcriptional starting sites of genes [8]. However, a known feature of DNA methylation is that the CpG islands remain methylated in the inactive X chromosome, imprinted genes and tissue-specific genes [9]. Transposable elements such as LINE-1 have the ability to integrate randomly to the genome and cause genomic instability. DNA methylation is known to silence these elements. In addition, DNA methylation can actively be reversed by the action of the TET family of DNA dioxygenases that convert 5-methylcytosine (5-mC) to 5-hydroxylmethylcytosine (5-hmC), 5formylcytosine (5-fC) and 5-carboxylcytosine (5-caC) [10]. There are three TET genes currently known as TET-1, TET-2 and TET-3. Please cite this article in press as: Kyburz D, et al., Epigenetic changes: The missing link, Best Practice & Research Clinical Rheumatology (2014), http://dx.doi.org/10.1016/j.berh.2014.10.014

D. Kyburz et al. / Best Practice & Research Clinical Rheumatology xxx (2014) 1e11

3

The posttranscriptional modification of histone tails is the other epigenetic mark that can regulate gene expression [5]. There are 16 posttranscriptional modification marks described so far, including methylation, acetylation, phosphorylation and ubiquitylation [11]. H4 acetylation and H3K4me3 are found in active genes. By contrast, H3K9 and H3K27me3 are associated with gene silencing. A variety of specialized proteins are involved in deposition of the chromatin marks such as the histone acetyl transferases (HATs) and the histone methyltransferases that add the acetyl and methyl group to the histone tails. Histone deacetylases (HDACs) and demethylases constitute another group of chromatin regulator proteins that remove the acetylation and methylation marks. The classical HDAC family consists of HDAC class I, II, while the SIR2 family of NAD-dependent HDACs form the HDAC class III [12]. The HDAC class I comprises HDAC 1, 2 and 3 which mainly target the histone proteins and HDAC 8 that has the structural maintenance of chromosome 3 (SMC) as a specific protein substrate [13]. The HDAC class II includes the HDACs 4, 5, 6, 7, 9, 10 and 11 that target mainly nonhistone substrates such as transcription factors. Lastly, the HDAC class III family consists of different sirtuins members (SIRT1, 2, 3, 4, 5, 6 and 7) with distinct cell functions [14]. Inhibition of the HDACs can cause alterations in gene expression, affect transcription factor activity, dysregulate signalling pathways and inhibit protein degradation. The molecular targets of HDACs play important roles in apoptosis, cell differentiation, cell cycle, inflammation and angiogenesis. Dysregulation of HDACs is known in different cancer types and a variety of HDAC small molecules inhibitors are in phase II and III clinical trials. [15] An additional group of proteins has the function to read the histone modifications. Members of this group are the bromodomain proteins (BRD2, BRD3, BRD4) and BRDT in germ cells. These proteins have bromodomains, which interact with acetylated histones and influence gene expression, cell cycle regulation and development [11]. Currently, next-generation sequencing technologies coupled with chromatin and methylation immunopreciptation assays can determine which histone modification or chromatin modifiers are found in a specific genomic region. These methods have been also used to identify alterations in different diseases. Tumour suppressor genes have been shown to be silenced by promoter methylation or to have hyperacetylated promoters [16]. Noncoding RNA Recent years have brought major progress in understanding the expression and function of noncoding RNA. Since the early 1990s when the H19 and Xist long noncoding RNAs were discovered and even more since the year 2000 when let-7 was discovered as one of the first microRNAs (miRNAs), more and more types of regulatory noncoding RNAs have been described and their substantial role in gene regulation elucidated [17e19]. The variety and complexity of RNA molecules and functions suggest that we are still at the very beginning of understanding the interconnected network by which noncoding RNA regulates gene expression. Long noncoding RNA By definition, every noncoding RNA that is longer than 200 bp is a long noncoding (lnc) RNA. Naturally, this rather crude definition includes noncoding RNAs with totally different structures and functions. The first discovered lncRNA was H19, which is interesting in many aspects. The H19 lncRNA, which was described to be overexpressed in synovial tissues of patients with RA compared to osteoarthritis (OA) or healthy individuals, belongs to the group of imprinted genes [20]. These genes are exclusively expressed from one allele only, in the case of H19 from the maternal, while the other allele is silenced by DNA methylation. As the monoallelic expression of imprinted genes is strongly dependent on regular DNA methylation, it is easily disturbed by disease-related hyper-and hypomethylation, respectively. The function of H19 is not fully clarified up to now. It is important in embryonic growth and development and has been described to work as tumour suppressor as well as to have oncogenic properties [21,22]. On the molecular level, H19 was found to associate with the histone methyltransferase EZH2, similar to many of the currently described lncRNAs, and therefore might have a role in changing histone methylation and subsequently gene transcription [23]. In addition, the first exon of Please cite this article in press as: Kyburz D, et al., Epigenetic changes: The missing link, Best Practice & Research Clinical Rheumatology (2014), http://dx.doi.org/10.1016/j.berh.2014.10.014

4

D. Kyburz et al. / Best Practice & Research Clinical Rheumatology xxx (2014) 1e11

H19 harbours the sequence of miR-675. To what extent the different functions of H19 are conferred by miR-675 still has to be analysed. microRNA From the group of small noncoding RNAs, miRNAs are by far the most studied by now. Mature miRNAs are ~22 nucleotides (nt) long and are processed from longer precursor transcripts. Most miRNAs are encoded in intergenic regions of the genome. However, they can also lie within introns or even exons of protein coding genes. Quite often, host genes of miRNAs are lncRNAs, as already mentioned for miR-675 which is encoded in H19. The primary microRNA transcript (pri-miR) is around 70 nt long. Often miRNA genes are clustered and are transcribed as polycistrons that can reach a length of several hundred nucleotides. The best characterized of these clusters is the miR-17/92 cluster which is around 800 nt long and comprises six miRNAs, miR-17, miR-18a, miR-19a, miR-20a, miR-19b-1 and miR-92a-1, which are transcribed together. The primary miRNA transcript is further processed by the endonuclease drosha to form a hairpin structured miRNA precursor, called pre-miR. Pre-miRs are exported to the cytoplasma, where they are rapidly cleaved by the endonuclease dicer. By cleavage of the loop of the hairpin, dicer creates a short, intermediate double-stranded miRNA product consisting of a 50 (5p) strand and a 30 (3p) strand. Both miR strands can be functional, even though mostly one of the strands is more stable and the only one incorporated in the RNA-induced silencing complex (RISC). The miRNA guides RISC to its target mRNA, which generally results in reduced protein translation via various mechanisms. The short length of miRNAs in combination with the fact that only partial complementarity between miRNA and target mRNA is necessary for regulation results in hundreds of potential target mRNAs for every single miRNA. In the rare case, that mRNA and miRNA sequences match perfectly, cleavage of the mRNA is induced. In the more common case of imperfect sequence matching, destabilization of the target mRNA or blockade of translation reduces the protein levels of the target mRNA. Interestingly, lncRNAs, in particular pseudogenes, with sequences that match to miRNAs have been found to work as decoys for miRNAs [24]. These competing endogenous RNAs (ceRNAs) draw miRNAs from their mRNA targets and thereby regulate miRNA function. More than two-thirds of protein coding genes are believed to be regulated by miRNAs, and miRNAs are involved in the regulation of practically all cellular pathways reaching from proliferation to apoptosis, and from cell differentiation to inflammatory response. It is not surprising that in various diseases, altered expression of miRNAs was detected and miRNAs have been suggested as therapeutic targets. Targeting one miRNA could regulate a whole network of proteins that are dysregulated in disease. The first miRNA that is targeted in disease is miR-122, which was shown to be critical for the replication of hepatitis C virus (HCV) in the liver [25]. Miravirsen is an intravenously applied miR-122 inhibitor that is currently in phase II for treatment of patients with HCV infections [26]. Remarkably, miRNAs are also found circulating in the blood stream, where they are protected from degradation by microparticles or RNA-binding proteins. Several studies could show strong correlations between miRNA serum levels and disease severity, outcome, or subtypes, in particular in cancer and inflammatory diseases. However, before an miRNA biomarker can be used in routine tests, problems regarding reproducibility, accuracy and simplicity of measuring miRNAs in serum need to be addressed. An overview of miRNA biogenesis and epigenetic modifications of DNA and histones is given in Fig. 1. Epigenetic changes in inflammatory arthritis DNA methylation First evidence for changes in the DNA methylation pattern in RA was presented already in 2000 when the expression of LINE-1 retrotransposons was shown in RA synovial fibroblasts (RASFs) but not in OA synovial fibroblasts (OASFs) [27]. Retrotransposons are mobile DNA sequences that are normally repressed by DNA methylation. In support of the assumption that demethylation may be responsible for the LINE-1 expression, it was shown that the DNA of RASFs was globally hypomethylated in Please cite this article in press as: Kyburz D, et al., Epigenetic changes: The missing link, Best Practice & Research Clinical Rheumatology (2014), http://dx.doi.org/10.1016/j.berh.2014.10.014

D. Kyburz et al. / Best Practice & Research Clinical Rheumatology xxx (2014) 1e11

5

Fig. 1. Epigenetic modifications and microRNA biogenesis. Histone modifications are placed on histone tails by writer proteins such as histone acetyl transferases (HATs) and the DNA methyltransferase EZH2, which is bound by long non-coding RNAs (lncRNAs). Reader proteins, e.g. bromodomains (BRDs), bind to these modifications and promote the formation of multiprotein complexes. Erasers, such as HDAC, remove the modifications. On the DNA, the DNMT1/UHRF1 complex mediates methylcytosine (mC) formation, while the TET proteins convert mC to carboxylcytosine (caC), formylcytosine (fC), and hydroxylmethylcytosine (hmC). MicroRNAs are transcribed as primary transcripts (pri-miR) that can be embedded in host genes or lncRNA. Drosha processes the primiR and Dicer cleaves the resulting precursor miR (pre-miR). The miR-Duplex is loaded onto the RNA induced silencing complex (RISC) where the separated strands bind their target mRNA. Ac ¼ acetylation, Me ¼ methylation, P ¼ phosphorylation, Ub ¼ ubiquitination.

comparison to OASFs or synovial fibroblasts of healthy individuals [28]. In addition, it was demonstrated that proliferating RASFs showed a relative lack of the DNA methyltransferase DNMT1, suggesting a methylation deficiency in these cells. Several studies have analysed the methylation status of specific promoter regions. The constitutively increased expression of CXCL12 was shown to correlate with a relative hypomethylation of its promoter region [29]. In addition, hypermethylation of DR3, EBF3 and IRX1 gene promoters has been found in RASFs [30,31]. A recent genome-wide study of the DNA methylome signature in RASFs from six versus OASFs from five patients revealed a number of differentially methylated genomic regions [32]. Totally, 203 genes with multiple differentially methylated loci were identified, playing key roles in inflammation, matrix regulation, leukocyte recruitment and immune responses. Hypomethylation was generally associated with overexpression, and hypermethylation inversely with decreased expression. These results were confirmed by another genomewide study of methylation in RASFs [33]. A much larger study on peripheral blood mononucleated cells (PBMCs) of more than 354 ACPA-positive patients versus 335 controls analysed the methylation pattern and the risk for RA [34]. To account for confounding factors, the authors have corrected for differences in cellular composition and used a statistical analysis to filter out methylation changes that were probable to be a consequence of the disease rather than causally related. Ten differentially methylated positions were identified to confer genetic risk for RA. Of these 10, three could be confirmed in a replication experiment on sorted monocytes of 12 patients with untreated ACPApositive RA. It remains to be shown whether changes in DNA methylation may account for changes in expression of HLA molecules which are associated with disease risk in RA. Previous studies had also analysed the promoter of CD40L and IL6 in CD4-positive cells and macrophages, respectively, and found specific differential methylated regions [35,36]. A different approach was chosen by Glant et al. who have analysed the expression of chromatin-modifying enzymes in B- and T-cells of patients with RA as well as of arthritic mice [37]. The most strongly upregulated genes were Aurora kinase A and B. Aurora kinases are enzymes that phosphorylate the tail of histone H3, important for NF-kB recruitment Please cite this article in press as: Kyburz D, et al., Epigenetic changes: The missing link, Best Practice & Research Clinical Rheumatology (2014), http://dx.doi.org/10.1016/j.berh.2014.10.014

6

D. Kyburz et al. / Best Practice & Research Clinical Rheumatology xxx (2014) 1e11

[38]. Interestingly, the Aurora-kinase specific inhibitor VX-680 had a preventive effect and significantly reduced disease severity in proteoglycan-induced arthritis in mice. Histone modifications Studies on the activity of HDAC in RA have provided conflicting results. A decreased activity of HDAC in RA synovial tissue compared to OA synovial tissues has been reported. Levels of HDAC1 and HDAC2 were found to be decreased [39]. By contrast, Kawabata et al. have found increased HDAC activity in RA synovial tissues and an increase in HDAC1 expression, dependent on the local concentration of TNFa [40]. On the level of a single gene, it was shown that acetylation of the histone 4 in the promoter region of the matrix metalloproteinase 1 (MMP1) gene was increased in RASFs compared to OASFs, but could be normalized by overexpression of the sentrin-specific protease SENP-1, which cleaves the small ubiquitin-like modifier (SUMO)-1 [41]. Experiments using inhibitors of HDAC (HDACi) support the concept of an increased activity of HDACs in RA. Trichostatin A (TSA) a class I/II HDACi was shown to block IL-6 production induced by lipopolysaccharide (LPS) or tumour necrosis alpha (TNFa) treatment of macrophages derived from PBMCs from patients with RA and from RA synovial fluid macrophages [42]. Similar results were obtained when nicotinamide, a nonspecific inhibitor of sirtuins, class III HDACs, was used. This result was somewhat unexpected as sirtuins have been shown to have beneficial metabolic effects and promote longevity [43]. In addition, deletion of SIRT1 on myeloid cells in mice resulted in increased nuclear facto kappa-B (NF-kB) activity and inflammatory activity [44]. Analysis of SIRT1 expression in SF and monocytes revealed a significantly increased expression in patients with RA compared to OA. Interestingly, inhibition of SIRT1 reduced LPS-induced TNF production by monocytes and IL-6 production by RASFs. The same result was found when the pan-sirtuin inhibitor Sirtinol was used, suggesting an overall proinflammatory effect of sirtuins on RA monocytes [45]. At the same time, SIRT1 mediated apoptosis resistance of RASFs, adding to the proinflammatory effects of this class III HDAC. In contrast to the results with SIRT1, it has been reported that adenoviral overexpression of SIRT6 in mice had a preventive as well as therapeutic effect on collagen-induced arthritis (CIA). SIRT6 overexpression led to a decrease in the clinical arthritis score, the production of proinflammatory cytokines in the synovium as well as a reduced severity of radiographic changes in the hind paws. The anti-inflammatory effect of SIRT6 was shown to be mediated by deacetylation of H3K9 at NF-kB target gene promoters [46]. microRNA In 2008, first reports of dysregulated miR expression in RA have appeared, describing upregulation of miR-155 and -146 in the rheumatoid synovium and in PBMCs [47e49]. MiR-155 expression has previously been shown to be induced by Toll-like receptor (TLR) ligands [50]. Among several putative targets of miR-155, the SH2 domain-containing inositol-50 -phophatase 1 (SHIP-1) was validated [51,52]. High expression of miR-155 in synovial fluid CD14 þ cells from patients with RA correlated with low expression of SHIP-1. Functional studies revealed that overexpression of miR-155 in human PBMCs resulted in an induction of proinflammatory cytokines, including TNF and IL-6 [51]. In agreement with the in vitro data with human cells, mice deficient for miR-155 were protected from CIA in vivo [51,53]. The absence of miR-155 led to a decrease of cytokine production including TNF and IL-6 and to a lower titre of collagen type II-specific IgG antibodies. However, miR-155-deficient mice developed joint inflammation after passive transfer of collagen-specific autoantibodies, suggesting that miR-155 is dispensable once autoantibodies are present. Interestingly, in the KxB/N serum transfer arthritis model, miR-155-deficient mice were protected from bone erosions, suggesting a regulatory role for miR-155 on antibody-mediated osteoclastogenesis [53]. A very recent publication has reported that activated T cells treated with methylprednisolone displayed decreased miR-155 expression and increased suppression of cytokine signalling 1 (SOCS1), leading to decreased JAK/STAT signalling and IL-12 production. The authors concluded that the anti-inflammatory effects of methylprednisolone may be at least in part be mediated by a suppression of miR-155 [54]. These findings add to the previous data indicating that miR-155 is an attractive therapeutic target in RA as well as other inflammatory diseases. Please cite this article in press as: Kyburz D, et al., Epigenetic changes: The missing link, Best Practice & Research Clinical Rheumatology (2014), http://dx.doi.org/10.1016/j.berh.2014.10.014

D. Kyburz et al. / Best Practice & Research Clinical Rheumatology xxx (2014) 1e11

7

In contrast to miR-155, miR-146a seems to have an anti-inflammatory role. Validated targets in PBMCs include interleukin-1 receptor-associated kinase 1 (IRAK-1) and TNF receptor associated factor 6 (TRAF-6) which are key adaptor molecules of signalling pathways of the innate immunity [48]. The fact that miR-146a-deficient mice develop a fatal immune-mediated disease supports the overall antiinflammatory effects of this miR [55]. In RA, increased expression of miR-146a has been shown in PBMCs, synovial fluid, RASFs and CD þ T-cells in comparison to OA samples [48,49,56,57]. In cultured RASFs, miR-146a was found to be constitutively upregulated and expression was increased by treatment with IL-1b or LPS [47,49]. The in vivo administration of miR-146a to mice with CIA resulted in a reduced bone destruction but did not significantly affect joint inflammation [58]. This in vivo effect correlated with an inhibition of osteoclast differentiation in vitro in human PBMCs transfected with miR-146a. Therefore, miR-146a may have a counter-regulatory role in joint destruction. Interestingly, a positive association of an IRAK-1 polymorphism with RA, ankylosing spondylitis and psoriatic arthritis has been described [59]. Although this variant is not located in the miR-146a target site, it may influence miR-146a binding by affecting the folding of the 3-untranslated region (UTR) of IRAK-1. MiR-155 and miR-146a are the best-characterized miR and stand out because of their functional importance for inflammation. While there are many more miR which have been reported to be dysregulated in RA (Table 1), for many of these the target mRNA and their functional role are not known.

Impact of environmental factors on epigenetics in arthritis Smoking is the strongest environmental risk factor for the development of RA. It is estimated that around one-fifth of all RA cases could be prevented by smoking cessation [60]. A strong geneeenvironment interaction of the risk alleles at the HLA-DRB1 locus with smoking has been found, in that the risk to develop ACPA-positive RA increases in smokers carrying the shared epitopes, whereas neither the risk to develop ACPA-negative disease nor individuals without shared epitopes are affected by smoking [2]. Based on these findings it was hypothesized that smoking may be associated with citrullination and the formation of auto-antibodies against citrullinated peptides. As citrullination has been found to be increased in bronchoalveolar lavage (BAL) cells of smokers compared to non-smokers, break of tolerance was suggested to be initiated in the lungs [61]. However, smoking can also directly induce changes in the joints. Increased expression of CYP1A1 and aryl hydrocarbon receptor repressor (AHRR) and lower expression of interleukin (IL)-17 were found in synovial tissues of smokers [62]. In addition, smokers as well as mice exposed to cigarette smoke have increased expression of heat shock proteins in their synovial tissues [63]. Several studies analysed the impact of smoking on epigenetic modifications and could show that smoking influences the expression of HDACs, histone modifications and DNA methylation in lung tissue [64e66]. In joints, increased expression of Sirtuin 6 could be shown in mice exposed to cigarette smoke and in RA patients who smoked compared to non-smoking patients [67]. Sirtuin 6 functions as protein deacetylase and as mono-ADP ribosyltransferase. As silencing of Sirtuin 6 in synovial fibroblasts Table 1 microRNA dysregulated in RA. miR

Tissue

Targets

References

miR-155 miR-146a miR-203 miR-34a* miR-124 miR-223 miR-16 miR-23b miR-132 miR-24 miR-221 miR-323-3p

FLS, ST, PBMC FLS, ST, PBMC FLS FLS FLS FLS, SF, PB CD3 Tcells, ST SF, PBMC ST PBMC, Plasma Plasma FLS FLS

SHIP-1, MMP-1, MMP-3 IRAK-1, TRAF-6

[48,49], [51], [56] [47e49], [56,57], [76e78] [79] [80] [76] [56], [76], [81,82] [48, 56] [78] [48, 56] [83] [84] [84]

X-linked inhibitor of apoptosis CDK2, MCP-1

TAB3, IKK-a, TAB2

FLS, fibroblast-like synoviocytes; ST, synovial tissue; PB, peripheral blood; PBMC, peripheral blood mononuclear cells.

Please cite this article in press as: Kyburz D, et al., Epigenetic changes: The missing link, Best Practice & Research Clinical Rheumatology (2014), http://dx.doi.org/10.1016/j.berh.2014.10.014

8

D. Kyburz et al. / Best Practice & Research Clinical Rheumatology xxx (2014) 1e11

increased the expression of MMP1, it is assumed that the smoke-induced upregulation of Sirtuin 6 is a protective mechanism trying to counterregulate smoke-induced MMP1 expression. Whether this effect is mediated by changes in histone acetylation or ADP ribosylation is however not clarified up to now. Whereas the effects of smoking on the RA risk and severity of disease are well known, it is less clear what component of the smoke is conveying the risk. Nicotine does not seem to be responsible, as it has been shown that consumption of nicotine in the form of moist snuff is not associated with an increased risk for RA [68]. On the other hand, environmental exposure to silica dust was found to increase the risk for ACPA-positive RA alone as well as in synergy with smoking, suggesting that different irritants may interact with MHC risk alleles [69]. Whether ambient air pollution contributes to the risk of RA is debated, but a clear association has not been established [70,71]. As seen in other autoimmune diseases RA occurs more often in females than in males. Sex hormones have been proposed to be responsible for this difference. In a cohort of over 500 patients with early arthritis, it was shown that hormonal replacement therapy significantly reduced the risk of HLADRB1 *04/01-positive patients to develop RA. The patients who had hormonal replacement therapy were significantly more often ACPA-negative. The authors concluded, that in a geneeenvironment interaction, hormonal replacement therapy might protect HLA-DRB1 *04/01-positive individuals from RA via inhibition of the production of ACPA [72]. The molecular mechanisms remain unclear; however, an involvement of miR may be considered. Dai et al. have reported a selective regulation of various miR in splenic lymphocytes from oestrogen-treated mice. Among others, they showed a decreased expression of miR-146a [73], suggesting that sex hormones may impact immune function at least in part via epigenetic pathways. At this time, the available evidence for a direct link between environmental factors and epigenetic changes impacting the pathogenesis of RA is still scarce. More studies are needed to understand how epigenetics may link the environment with a predisposing genetic background. Therapeutic targeting of epigenetic mechanisms in arthritis As outlined above, in vitro data with human cells from patients with RA as well as studies with animal models of arthritis have established an important role for epigenetic mechanisms in the pathogenesis of arthritis and support the development of drugs targeting key epigenetic factors. So far, clinical development has focused on cancer. Several drugs have been approved until today. 5azacytidine and decitabine, targeting the methyltransferase DNMT1, have been licensed for the treatment of myelodysplastic syndromes. The HDAC inhibitors vorinostat and romidepsin are licensed for the treatment of cutaneous T-cell lymphoma. Many more compounds targeting methyltransferases, HDAC inhibitors, histone acetyl transferases as well as inhibitors/activators of sirtuins are currently in clinical trials for various neoplastic diseases (for review Ref. [74]). For autoimmune rheumatic diseases, there is less clinical data available. The HDAC inhibitor givinostat was used in a small open-label trial in 17 patients with active systemic onset juvenile idiopathic arthritis. Two-thirds of the patients achieved a 70% improvement of defined disease activity criteria (ACRPedi 70) with only mild-to-moderate adverse events related to the study drug [75]. These results clearly show the potential of drugs targeting epigenetic mechanisms in chronic inflammatory arthritis. Despite the preclinical data suggesting important effects of miRNA in the pathogenesis of RA, there are as yet no clinical trials with miRNA targeting agents in this disease. Successful development of miRbased therapeutics for inflammatory arthritis will need to find solutions for targeted delivery and the avoidance of effects on bystander cells potentially causing unwanted side effects. Summary RA is an autoimmune disease with a complex multifactorial pathogenesis. Genome-wide association studies have led to the identification of a number of genetic risk loci, most prominently MHC alleles. However, genetics explain only part of the disease risk. Environmental factors contribute in addition to genetic factors to the disease risk. Groundbreaking studies have shown that smoking increases the risk for ACPA-positive RA in individuals with certain HLA alleles. The molecular Please cite this article in press as: Kyburz D, et al., Epigenetic changes: The missing link, Best Practice & Research Clinical Rheumatology (2014), http://dx.doi.org/10.1016/j.berh.2014.10.014

D. Kyburz et al. / Best Practice & Research Clinical Rheumatology xxx (2014) 1e11

9

mechanisms of these geneeenvironment interactions however remain unknown. As epigenetic changes can occur in response to external stimuli, they may represent a link between genetics and the environment. There is evidence for epigenetic changes in RA. Alterations of DNA methylation, histone modifications and dysregulation of miRNA expression have been demonstrated in synovial tissues and in peripheral blood cells. In animal models of arthritis, key roles for certain miRNA and therapeutic efficacy of drugs targeting epigenetic modifications have been found. These data clearly document the importance of epigenetic mechanisms for the pathogenesis of RA and the potential for therapeutic application. However, we are only beginning to understand how epigenetic changes induced by environmental factors such as smoking may impact the expression of genes associated with RA.

Research agenda:  To understand the interaction of genes and environment, there is a need for better characterization of DNA and histone modifications as well as miRNA expression associated with RA in the context of the presence of genetic and environmental risk factors.  Functional studies with agents inducing or inhibiting epigenetic changes are required to elucidate the impact on the disease pathogenesis and to define potential targets for therapeutic intervention.

Acknowledgements DK is supported by the Swiss National Fund (grant 310030_144254) and the Clinical Research Focus Program of the University of Zurich. CO and EK are supported by IMI-BTCure, IAR Epalinges und FP7 EuroTEAM.

References [1] Bax M, van Heemst J, Huizinga TW, et al. Genetics of rheumatoid arthritis: what have we learned? Immunogenetics 2011; 63(8):459e66. *[2] Klareskog L, Stolt P, Lundberg K, et al. A new model for an etiology of rheumatoid arthritis: smoking may trigger HLA-DR (shared epitope)-restricted immune reactions to autoantigens modified by citrullination. Arthritis Rheum 2006;54(1): 38e46. *[3] Portela A, Esteller M. Epigenetic modifications and human disease. Nat Biotechnol 2010;28(10):1057e68. [4] Jones PA. Functions of DNA methylation: islands, start sites, gene bodies and beyond. Nat Rev Genet 2012;13(7):484e92. [5] Bannister AJ, Kouzarides T. Regulation of chromatin by histone modifications. Cell Res 2011;21(3):381e95. [6] Bostick M, Kim JK, Esteve PO, et al. UHRF1 plays a role in maintaining DNA methylation in mammalian cells. Science 2007;317(5845):1760e4. *[7] Smith ZD, Meissner A. DNA methylation: roles in mammalian development. Nat Rev Genet 2013;14(3):204e20. [8] Deaton AM, Bird A. CpG islands and the regulation of transcription. Genes Dev 2011;25(10):1010e22. [9] Norris DP, Patel D, Kay GF, et al. Evidence that random and imprinted Xist expression is controlled by preemptive methylation. Cell 1994;77(1):41e51. [10] Wu H, Zhang Y. Reversing DNA methylation: mechanisms, genomics, and biological functions. Cell 2014;156(1e2):45e68. [11] Dawson MA, Kouzarides T, Huntly BJ. Targeting epigenetic readers in cancer. N Engl J Med 2012;367(7):647e57. *[12] Chapman-Rothe N, Brown R. Prospects for epigenetic compounds in the treatment of autoimmune disease. Adv Exp Med Biol 2011;711:150e61. [13] Deardorff MA, Bando M, Nakato R, et al. HDAC8 mutations in cornelia de lange syndrome affect the cohesin acetylation cycle. Nature 2012;489(7415):313e7. [14] Giblin W, Skinner ME, Lombard DB. Sirtuins: guardians of mammalian healthspan. trends in genetics. TIG 2014;30(7): 271e86. [15] West AC, Johnstone RW. New and emerging HDAC inhibitors for cancer treatment. J Clin Invest 2014;124(1):30e9. [16] Issa JP. CpG island methylator phenotype in cancer. Nat Rev Cancer 2004;4(12):988e93. [17] Brannan CI, Dees EC, Ingram RS, et al. The product of the H19 gene may function as an RNA. Mol Cell Biol 1990;10(1): 28e36. [18] Brockdorff N, Ashworth A, Kay GF, et al. The product of the mouse Xist gene is a 15 kb inactive X-specific transcript containing no conserved ORF and located in the nucleus. Cell 1992;71(3):515e26. [19] Reinhart BJ, Slack FJ, Basson M, et al. The 21-nucleotide let-7 RNA regulates developmental timing in caenorhabditis elegans. Nature 2000;403(6772):901e6.

Please cite this article in press as: Kyburz D, et al., Epigenetic changes: The missing link, Best Practice & Research Clinical Rheumatology (2014), http://dx.doi.org/10.1016/j.berh.2014.10.014

10

D. Kyburz et al. / Best Practice & Research Clinical Rheumatology xxx (2014) 1e11

[20] Stuhlmuller B, Kunisch E, Franz J, et al. Detection of oncofetal h19 RNA in rheumatoid arthritis synovial tissue. Am J Pathol 2003;163(3):901e11. [21] Hao Y, Crenshaw T, Moulton T, et al. Tumour-suppressor activity of H19 RNA. Nature 1993;365(6448):764e7. [22] Rachmilewitz J, Elkin M, Rosensaft J, et al. H19 expression and tumorigenicity of choriocarcinoma derived cell lines. Oncogene 1995;11(5):863e70. [23] Luo M, Li Z, Wang W, et al. Long non-coding RNA H19 increases bladder cancer metastasis by associating with EZH2 and inhibiting E-cadherin expression. Cancer Lett 2013;333(2):213e21. [24] Pink RC, Wicks K, Caley DP, et al. Pseudogenes: pseudo-functional or key regulators in health and disease? RNA 2011; 17(5):792e8. [25] Jopling CL, Yi M, Lancaster AM, et al. Modulation of hepatitis C virus RNA abundance by a liver-specific microRNA. Science 2005;309(5740):1577e81. *[26] Janssen HL, Reesink HW, Lawitz EJ, et al. Treatment of HCV infection by targeting microRNA. N Engl J Med 2013;368(18): 1685e94. [27] Neidhart M, Rethage J, Kuchen S, et al. Retrotransposable L1 elements expressed in rheumatoid arthritis synovial tissue: association with genomic DNA hypomethylation and influence on gene expression. Arthritis Rheum 2000;43(12): 2634e47. [28] Karouzakis E, Gay RE, Michel BA, et al. DNA hypomethylation in rheumatoid arthritis synovial fibroblasts. Arthritis Rheum 2009;60(12):3613e22. [29] Karouzakis E, Rengel Y, Jungel A, et al. DNA methylation regulates the expression of CXCL12 in rheumatoid arthritis synovial fibroblasts. Genes Immun 2011;12(8):643e52. [30] Park SH, Kim SK, Choe JY, et al. Hypermethylation of EBF3 and IRX1 genes in synovial fibroblasts of patients with rheumatoid arthritis. Mol Cells 2013;35(4):298e304. [31] Takami N, Osawa K, Miura Y, et al. Hypermethylated promoter region of DR3, the death receptor 3 gene, in rheumatoid arthritis synovial cells. Arthritis Rheum 2006;54(3):779e87. [32] Nakano K, Whitaker JW, Boyle DL, et al. DNA methylome signature in rheumatoid arthritis. Ann Rheum Dis 2013;72(1): 110e7. [33] de la Rica L, Urquiza JM, Gomez-Cabrero D, et al. Identification of novel markers in rheumatoid arthritis through integrated analysis of DNA methylation and microRNA expression. J Autoimmun 2013;41:6e16. *[34] Liu Y, Aryee MJ, Padyukov L, et al. Epigenome-wide association data implicate DNA methylation as an intermediary of genetic risk in rheumatoid arthritis. Nat Biotechnol 2013;31(2):142e7. [35] Liao J, Liang G, Xie S, et al. CD40L demethylation in CD4(þ) T cells from women with rheumatoid arthritis. Clin Immunol 2012;145(1):13e8. [36] Nile CJ, Read RC, Akil M, et al. Methylation status of a single CpG site in the IL6 promoter is related to IL6 messenger RNA levels and rheumatoid arthritis. Arthritis Rheum 2008;58(9):2686e93. [37] Glant TT, Besenyei T, Kadar A, et al. Differentially expressed epigenome modifiers, including aurora kinases A and B, in immune cells in rheumatoid arthritis in humans and mouse models. Arthritis Rheum 2013;65(7):1725e35. [38] Wolter S, Doerrie A, Weber A, et al. c-Jun controls histone modifications, NF-kappaB recruitment, and RNA polymerase II function to activate the ccl2 gene. Mol Cell Biol 2008;28(13):4407e23. [39] Huber LC, Brock M, Hemmatazad H, et al. Histone deacetylase/acetylase activity in total synovial tissue derived from rheumatoid arthritis and osteoarthritis patients. Arthritis Rheum 2007;56(4):1087e93. [40] Kawabata T, Nishida K, Takasugi K, et al. Increased activity and expression of histone deacetylase 1 in relation to tumor necrosis factor-alpha in synovial tissue of rheumatoid arthritis. Arthritis Res Ther 2010;12(4):R133. [41] Maciejewska-Rodrigues H, Karouzakis E, Strietholt S, et al. Epigenetics and rheumatoid arthritis: the role of SENP1 in the regulation of MMP-1 expression. J Autoimmun 2010;35(1):15e22. *[42] Grabiec AM, Krausz S, de Jager W, et al. Histone deacetylase inhibitors suppress inflammatory activation of rheumatoid arthritis patient synovial macrophages and tissue. J Immunol 2010;184(5):2718e28. [43] Bordone L, Guarente L. Calorie restriction, SIRT1 and metabolism: understanding longevity. Nat Rev Mol Cell Biol 2005; 6(4):298e305. [44] Schug TT, Xu Q, Gao H, et al. Myeloid deletion of SIRT1 induces inflammatory signaling in response to environmental stress. Mol Cell Biol 2010;30(19):4712e21. [45] Niederer F, Ospelt C, Brentano F, et al. SIRT1 overexpression in the rheumatoid arthritis synovium contributes to proinflammatory cytokine production and apoptosis resistance. Ann Rheum Dis 2011;70(10):1866e73. [46] Lee HS, Ka SO, Lee SM, et al. Overexpression of sirtuin 6 suppresses inflammatory responses and bone destruction in mice with collagen-induced arthritis. Arthritis Rheum 2013;65(7):1776e85. [47] Nakasa T, Miyaki S, Okubo A, et al. Expression of microRNA-146 in rheumatoid arthritis synovial tissue. Arthritis Rheum 2008;58(5):1284e92. [48] Pauley KM, Satoh M, Chan AL, et al. Upregulated miR-146a expression in peripheral blood mononuclear cells from rheumatoid arthritis patients. Arthritis Res Ther 2008;10(4):R101. [49] Stanczyk J, Pedrioli DM, Brentano F, et al. Altered expression of microRNA in synovial fibroblasts and synovial tissue in rheumatoid arthritis. Arthritis Rheum 2008;58(4):1001e9. [50] O'Connell RM, Taganov KD, Boldin MP, et al. microRNA-155 is induced during the macrophage inflammatory response. Proc Natl Acad Sci U S A 2007;104(5):1604e9. [51] Kurowska-Stolarska M, Alivernini S, Ballantine LE, et al. microRNA-155 as a proinflammatory regulator in clinical and experimental arthritis. Proc Natl Acad Sci U S A 2011;108(27):11193e8. [52] O'Connell RM, Chaudhuri AA, Rao DS, et al. Inositol phosphatase SHIP1 is a primary target of miR-155. Proc Natl Acad Sci U S A 2009;106(17):7113e8. *[53] Bluml S, Bonelli M, Niederreiter B, et al. Essential role of microRNA-155 in the pathogenesis of autoimmune arthritis in mice. Arthritis Rheum 2011;63(5):1281e8. [54] Davis TE, Kis-Toth K, Tsokos GC. A114: methylprednisolone-induced inhibition of miR-155 expression increases SOCS1Driven suppression of cytokine signaling. Arthritis Rheumatol 2014;66(Suppl. 11):S151.

Please cite this article in press as: Kyburz D, et al., Epigenetic changes: The missing link, Best Practice & Research Clinical Rheumatology (2014), http://dx.doi.org/10.1016/j.berh.2014.10.014

D. Kyburz et al. / Best Practice & Research Clinical Rheumatology xxx (2014) 1e11

11

[55] Lu LF, Boldin MP, Chaudhry A, et al. Function of miR-146a in controlling treg cell-mediated regulation of Th1 responses. Cell 2010;142(6):914e29. [56] Murata K, Yoshitomi H, Tanida S, et al. Plasma and synovial fluid microRNAs as potential biomarkers of rheumatoid arthritis and osteoarthritis. Arthritis Res Ther 2010;12(3):R86. [57] Li J, Wan Y, Guo Q, et al. Altered microRNA expression profile with miR-146a upregulation in CD4þ T cells from patients with rheumatoid arthritis. Arthritis Res Ther 2010;12(3):R81. [58] Nakasa T, Shibuya H, Nagata Y, et al. The inhibitory effect of microRNA-146a expression on bone destruction in collageninduced arthritis. Arthritis Rheum 2011;63(6):1582e90. [59] Chatzikyriakidou A, Voulgari PV, Georgiou I, et al. A polymorphism in the 3'-UTR of interleukin-1 receptor-associated kinase (IRAK1), a target gene of miR-146a, is associated with rheumatoid arthritis susceptibility. Jt Bone Spine 2010; 77(5):411e3. [60] Klareskog L, Gregersen PK, Huizinga TW. Prevention of autoimmune rheumatic disease: state of the art and future perspectives. Ann Rheum Dis 2010;69(12):2062e6. [61] Makrygiannakis D, Hermansson M, Ulfgren AK, et al. Smoking increases peptidylarginine deiminase 2 enzyme expression in human lungs and increases citrullination in BAL cells. Ann Rheum Dis 2008;67(10):1488e92. [62] Kazantseva MG, Highton J, Stamp LK, et al. Dendritic cells provide a potential link between smoking and inflammation in rheumatoid arthritis. Arthritis Res Ther 2012;14(5):R208. *[63] Ospelt C, Camici GG, Engler A, et al. Smoking induces transcription of the heat shock protein system in the joints. Ann Rheum Dis 2014;73(7):1423e6. [64] Belinsky SA, Palmisano WA, Gilliland FD, et al. Aberrant promoter methylation in bronchial epithelium and sputum from current and former smokers. Cancer Res 2002;62(8):2370e7. [65] Ito K, Ito M, Elliott WM, et al. Decreased histone deacetylase activity in chronic obstructive pulmonary disease. N. Engl J Med 2005;352(19):1967e76. [66] Sundar IK, Nevid MZ, Friedman AE, et al. Cigarette smoke induces distinct histone modifications in lung cells: implications for the pathogenesis of COPD and lung cancer. J Proteome Res 2014;13(2):982e96. [67] Engler A, Niederer F, Klein K, et al. SIRT6 regulates the cigarette smoke-induced signalling in rheumatoid arthritis synovial fibroblasts. J Mol Med 2014;92(7):757e67. [68] Carlens C, Hergens MP, Grunewald J, et al. Smoking, use of moist snuff, and risk of chronic inflammatory diseases. Am J Respir Crit care Med 2010;181(11):1217e22. [69] Stolt P, Yahya A, Bengtsson C, et al. Silica exposure among male current smokers is associated with a high risk of developing ACPA-positive rheumatoid arthritis. Ann Rheum Dis 2010;69(6):1072e6. [70] Hart JE, Kallberg H, Laden F, et al. Ambient air pollution exposures and risk of rheumatoid arthritis. Arthritis Care Res Hob 2013;65(7):1190e6. [71] Hart JE, Kallberg H, Laden F, et al. Ambient air pollution exposures and risk of rheumatoid arthritis: results from the Swedish EIRA case-control study. Ann Rheum Dis 2013;72(6):888e94. [72] Salliot C, Bombardier C, Saraux A, et al. Hormonal replacement therapy may reduce the risk for RA in women with early arthritis who carry HLA-DRB1 *01 and/or *04 alleles by protecting against the production of anti-CCP: results from the ESPOIR cohort. Ann Rheum Dis 2010;69(9):1683e6. [73] Dai R, Phillips RA, Zhang Y, et al. Suppression of LPS-induced Interferon-gamma and nitric oxide in splenic lymphocytes by select estrogen-regulated microRNAs: a novel mechanism of immune modulation. Blood 2008;112(12):4591e7. [74] Di Costanzo A, Del Gaudio N, Migliaccio A, et al. Epigenetic drugs against cancer: an evolving landscape. Archives Toxicol 2014;88(9):1651e68. *[75] Vojinovic J, Damjanov N. HDAC inhibition in rheumatoid arthritis and juvenile idiopathic arthritis. Mol Med 2011; 17(5e6):397e403. [76] Nakamachi Y, Kawano S, Takenokuchi M, et al. MicroRNA-124a is a key regulator of proliferation and monocyte chemoattractant protein 1 secretion in fibroblast-like synoviocytes from patients with rheumatoid arthritis. Arthritis Rheum 2009;60(5):1294e304. [77] Niimoto T, Nakasa T, Ishikawa M, et al. MicroRNA-146a expresses in interleukin-17 producing T cells in rheumatoid arthritis patients. BMC Musculoskelet Disord 2010;11:209. [78] Zhu S, Pan W, Song X, et al. The microRNA miR-23b suppresses IL-17-associated autoimmune inflammation by targeting TAB2, TAB3 and IKK-alpha. Nat Med 2012;18(7):1077e86. [79] Stanczyk J, Ospelt C, Karouzakis E, et al. Altered expression of microRNA-203 in rheumatoid arthritis synovial fibroblasts and its role in fibroblast activation. Arthritis Rheum 2011;63(2):373e81. [80] Niederer F, Trenkmann M, Ospelt C, et al. Downregulation of microRNA-34a* in rheumatoid arthritis synovial fibroblasts promotes apoptosis resistance. Arthritis Rheum 2011;64(6):1771e9. [81] Fulci V, Scappucci G, Sebastiani GD, et al. miR-223 is overexpressed in T-lymphocytes of patients affected by rheumatoid arthritis. Hum Immunol 2010;71(2):206e11. [82] Shibuya H, Nakasa T, Adachi N, et al. Overexpression of microRNA-223 in rheumatoid arthritis synovium controls osteoclast differentiation. Mod Rheumatol 2013;23(4):674e85. [83] Murata K, Furu M, Yoshitomi H, et al. Comprehensive microRNA analysis identifies miR-24 and miR-125a-5p as plasma biomarkers for rheumatoid arthritis. PLoS One 2013;8(7). e69118. [84] Pandis I, Ospelt C, Karagianni N, et al. Identification of microRNA-221/222 and microRNA-323-3p association with rheumatoid arthritis via predictions using the human tumour necrosis factor transgenic mouse model. Ann Rheum Dis 2012;71(10):1716e23.

Please cite this article in press as: Kyburz D, et al., Epigenetic changes: The missing link, Best Practice & Research Clinical Rheumatology (2014), http://dx.doi.org/10.1016/j.berh.2014.10.014