Epigenetics of scleroderma: Integrating genetic, ethnic, age, and environmental effects

Abstract

Scleroderma or systemic sclerosis is thought to result from the interplay between environmental or non-genetic factors in a genetically susceptible individual. Epigenetic modifications are influenced by genetic variation and environmental exposures, and change with chronological age and between populations. Despite progress in identifying genetic, epigenetic, and environmental risk factors, the underlying mechanism of systemic sclerosis remains unclear. Since epigenetics provides the regulatory mechanism linking genetic and non-genetic factors to gene expression, understanding the role of epigenetic regulation in systemic sclerosis will elucidate how these factors interact to cause systemic sclerosis. Among the cell types under tight epigenetic control and susceptible to epigenetic dysregulation, immune cells are critically involved in early pathogenic events in the progression of fibrosis and systemic sclerosis. This review starts by summarizing the changes in DNA methylation, histone modification, and non-coding RNAs associated with systemic sclerosis. It then discusses the role of genetic, ethnic, age, and environmental effects on epigenetic regulation, with a focus on immune system dysregulation. Given the potential of epigenome editing technologies for cell reprogramming and as a therapeutic approach for durable gene regulation, this review concludes with a prospect on epigenetic editing. Although epigenomics in systemic sclerosis is in its infancy, future studies will help elucidate the regulatory mechanisms underpinning systemic sclerosis and inform the design of targeted epigenetic therapies to control its dysregulation.

Introduction

Systemic sclerosis (SSc, scleroderma) is a multisystem autoimmune disease characterized by cutaneous and visceral fibrosis, immune dysregulation, and vasculopathy. It is often considered a fibrosing disease preceded and accompanied by autoimmunity and vasculopathy, ¹ as immune system activation is an early and key event in SSc pathogenesis. ²

As reviewed elsewhere, SSc shows great variation in reported prevalence estimates (varying between 31 and 659 cases per million), a strong sex bias (between three and 14 affected women for one man), and marked ethnic disparities.^3,4 While twin and family studies support a high genetic contribution to the development of autoimmunity in SSc, multiple lines of evidence support a modest genetic etiology in SSc, ³ suggesting a substantial role for epigenetic or environmental factors in SSc susceptibility.

Epigenetics refers to chromatin modifications that influence gene function but not involving changes in the DNA sequence that can be propagated through cell division. By allowing or preventing molecules’ access to regulatory sequences, the chromatin’s structure provides gene regulation. Commonly studied epigenetic mechanisms include DNA methylation, histone modification, non-coding RNA, and chromatin accessibility representing transcription factor-binding sites. These epigenetic marks can change the spatial conformation of chromatin to modulate or regulate gene expression. The epigenetic architecture is influenced by genetic variation and environmental exposures, and it changes with chronological age and between populations. Given that epigenetic changes reflect influences of both genetic and environmental factors, and epigenetics thus provides a molecular mechanism linking genetic background and environmental exposures to disease, there is much interest in understanding the role of epigenetic regulation in disease.

Comprehensive reviews of epigenetic research in SSc have been recently published.^5–8 In the first part, this review summarizes the changes in DNA methylation, histone modification, and non-coding RNAs associated with SSc. It then discusses the role of genetic, ethnic, age, and environmental effects on epigenetic regulation, followed by a prospect on epigenetic editing. Since immune system dysregulation critically contributes to early pathogenic events in the progression of fibrosis and SSc, and epigenetic regulation is critically involved in immune cell differentiation and response to stimuli,^9,10 the second part of this review focuses on epigenetic regulation of immune cells.

DNA methylation

DNA methylation consists of the addition of a CH3 methyl group to cytosine phospho-guanine dinucleotides (CpG). CpG-rich sequences, called CpG islands, often locate within gene promoter regions. CpG methylation is essential for multiple biological processes, including gene regulation, genomic imprinting, X chromosome inactivation or repression of transposable elements. Methylation is catalyzed by a family of DNA methyltransferases (DNMTs). Although generally associated with transcriptional silencing, ¹¹ the precise relationships between DNA methylation, genetic factors, and gene expression are complex and poorly understood.^12–14 The correlation between DNA methylation and gene expression can be positive or negative, is tissue-specific, and context-specific, in that the local DNA sequence and genomic features largely account for local patterns of methylation.^11,15–17

In 2006, Wang and colleagues ¹⁸ first implicated DNA methylation mechanisms as mediators of the fibrotic manifestations of SSc by unveiling hypermethylation and histone deacetylation at the friend leukemia integration 1 (FLI1) promoter. As thoroughly reviewed elsewhere,^5,6 a total of 18 studies have documented changes in enzyme (including methyl-CpG-binding domain proteins (MBD) and DNMTs), global methylation, and candidate gene methylation levels across multiple cell types (blood, CD4+ T cells, fibroblasts, endothelial cells). Among these, three genome-wide DNA methylation studies^19–21 on skin, lung fibroblasts, CD4+ T cells, and CD8+ T cells, further confirm the role of aberrant DNA methylation on SSc pathogenesis.

DNA methylation changes associated with SSc are summarized in Table 1. In CD4+ T cells there is global DNA hypomethylation and concomitant underexpression of methylation-regulating genes (e.g. DNMT1, MBD3, and MBD). ²² Overexpression of CD40L, CD70, and ITGAL (aka CD11a) has been attributed to hypomethylation of the genes’ promoters.^28–30 Demethylation of the costimulatory CD70 (aka TNFSF7) and X-chromosome-encoded CD40L genes leads to their overexpression on activated CD4+ T cells, and ultimately B-cell autoantibody production. In the case of the integrin subunit alpha L (ITGAL), the gene’s overexpression contributes to increased CD4+ T cell proliferation, IgG production, and collagen production, suggesting a role for epigenetic dysregulation in immune and fibrotic dysfunction is SSc. ³⁰ Forkhead box P3 (FOXP3), a transcription factor required for generation of regulatory T (Treg) cells that is hypomethylated in Treg cells and hypermethylated in conventional T cells, ³⁶ shows hypermethylation in CD4+ T cells of patients with SSc, which has effect in the number of Tregs. ³¹

Table 1.

Summary of DNA methylation changes in SSc.

Modification	Cell type/tissue	Refs
GlobalZ
DNMTs levels	CD4+ T cells	22
	Fibroblasts	18,23
MBDs levels	CD4+ T cells	22
	Fibroblasts	18
TET levels	Fibroblasts	24,25
MECP2 levels	Fibroblasts	18,25
Global methylation levels	Whole blood	26
	CD4+ T cells	22
	Fibroblasts	24
Gene-specific
RORC1, RORC2	PBMCs	27
CD40L	CD4+ T cells	28
CD70	CD4+ T cells	29
CD11a	CD4+ T cells	30
FOXP3	CD4+ T cells	31
BMPR2	Endothelial cells, skin	32
FLI1	Fibroblasts	18
KLF5	Fibroblasts	33
C8orf4/TCIM	Lung fibroblasts	21
DKK1, SFRP1	PBMCs, fibroblasts	34
Several X chromosome genes	PBMCs	35
ITGAL, CD11a, CD70, FOXP3, and other	Whole blood	26
ITGA9, COL4A2, TNXB, CDH11, and other	Fibroblasts	19
Type I interferon genes	CD4+ and CD8+ T cells	20

SSc: scleroderma or systemic sclerosis; Refs: references; PBMCs: peripheral blood mononuclear cells; DNMTs: DNA methyltransferases; MBDs: methyl-CpG-binding domain; TET: ten-eleven translocation; MECP2: methyl-CpG binding protein 2; RORC: RAR-related orphan receptor C; FOXP3: forkhead box P3; BMPR2: bone morphogenetic protein receptor type 2; FLI1: Fli-1 proto-oncogene, ETS transcription factor; KLF5: Kruppel like factor 5; TCIM: transcriptional and immune response regulator; DKK1: dickkopf WNT signaling pathway inhibitor 1; SFRP1: secreted frizzled related protein 1; ITGAL: integrin subunit alpha L; ITGA9: integrin subunit alpha 9; COL4A2: collagen type IV alpha 2 chain; TNBX: tenascin XB; CDH11: cadherin 11.

In contrast to CD4+ T cells, fibroblasts from SSc patients show high levels of methylation-regulating genes such as DNMT1, MBD1, and MECP2. There is hypermethylation in the promoter region of FLI1, an inhibitor of collagen expression. ¹⁸ In lung fibroblasts, hypermethylation of the transcriptional and immune response regulator (TCIM, aka C8orf4) gene, which functions as a positive regulator of the Wnt/β-catenin signaling pathway, leads to its underexpression, resulting in a decreased capacity of the fibrotic lung fibroblasts to up-regulate cyclooxygenase-2 (COX2) and COX-2-derived prostaglandin E2 (PGE2) synthesis. ²¹ It is important to note that many of these epigenetic modifications (e.g. epigenetic silencing of FLI1 or TCIM) might drive fibroblast activation and transition to myofibroblasts. Recent reviews summarize the role of DNA methylation, histone modifications, and non-coding RNAs involved in myofibroblast activation.^37,38 For example, MeCP2 promotes myofibroblast differentiation in mice and human,^39,40 and hypomethylation of genes such as integrin subunit alpha (ITGA9) and cadherin 11 (CDH11) in SSc skin fibroblasts ¹⁹ might contribute to myofibroblast differentiation.

In microvascular endothelial cells, hypermethylation of the promoter of the bone morphogenetic protein type II receptor (BMPR2) gene leads to reduced gene expression, resulting in endothelial cell enhanced apoptosis. ³²

Histone modification

Chromatin is formed by DNA wrapped around histone octamers formed by two copies of the core histones (H2A, H2B, H3, and H4) and linker histones (H2 and H5). Covalent post-translational modifications at specific amino acid residues in the side chains of histone tails can affect chromatin structure and influence its accessibility. Histone tail modifications include methylation, phosphorylation, acetylation, ubiquitylation, sumoylation, and ADP ribosylation. Histone acetyltransferases (HATs) (e.g. EP300/p300, KAT/MYST) and histone deacetylases (HDACs) (e.g. HDAC1-11, SIRT1-7) regulate histone acetylation required for biological processes such as DNA replication and transcription.

Histone acetylation and histone methylation are the most common modifications. Histone acetylation typically occurs at lysine residues on histone H3 or H4 and can present as mono-, di-, or tri-acetylation. Because it leads to an open chromatin conformation accessible to transcription factors, histone acetylation marks are associated with transcriptional activation. Histone methylation can be associated with either transcriptional repression or activation depending on the number of methyl groups that are added and the location of the amino acid being methylated. For example, enhancers marked by acetylation of histone H3 at lysine 27 (H3K27ac) and methylation of H3 at lysine 4 (H3K4me1) promote transcription, ⁴¹ while H3K27 trimethylation (H3K27me3) is a repressive epigenetic mark.

The first evidence for histone modification involvement in SSc came from a study investigating the role of the transcriptional coactivator histone acetyltransferase p300 in the profibrotic responses elicited by TGF-beta in fibroblasts. ⁴² The authors found that elevated p300 expression contributes to the intensity of profibrotic responses, establishing the role of the HAT p300 is mediating stimulation of collagen synthesis in fibroblasts. ⁴² Soon after another study in fibroblasts reported increased levels of epigenetic mediators, as well as normalized collagen expression upon addition of HDAC inhibitors. ¹⁸

Table 2 summarizes the histone modification changes associated with SSc. Most studies of histone modification in SSc have focused on fibroblasts. For example, the histone deacetylase sirtuin 1 (SIRT1) is reduced in SSc skin and fibroblasts. ⁴³ The histone demethylase lysine demethylase 6B (KDM6B, aka Jumonji domain-containing protein 3 (JMJD3)) is increased in fibroblasts, ⁴⁶ resulting in accumulation of H3K27me3 at the promoter of the profibrotic FOS like 2, AP-1 transcription factor (FOSL2, aka FRA2). ⁴⁹ Thus, JMJD3 and H3K27me3 modulate fibroblast activation by regulating FOSL2 expression. As reviewed elsewhere,^37,38 histone modifications are known contributors to myofibroblast activation. For example, JMJD3 regulates myofibroblast activation, ⁵² p300 regulates fibrosis through myofibroblast transformation, ⁵³ and elevated levels of the histone methyltransferase enhancer of zeste homolog 2 (EZH2) contribute to myofibroblast transdifferentiation in SSc. ⁵⁴

Table 2.

Summary of histone modification changes in SSc.

Modification	Cell type/tissue	Gene	Refs
Global
HDACs levels	Fibroblasts	NA	18,43,44
	B cells	NA	45
	Endothelial cells	NA	46
HATs levels	Fibroblasts	NA	42
	B cells	NA	47
HDMs levels	CD4+ T cells	NA	48
	B cells	NA	45
HMTs levels	B cells	NA	45
H4 acetylation levels	B cells	NA	45
H3K9 methylation levels	B cells	NA	45
H3K27me3	Fibroblasts	NA	49
	CD4+ T cells	NA	48
Gene-specific
H3, H4 hypoacetylation	Fibroblasts	FLI1	18
H3, H4 hypoacetylation	Fibroblasts	KLF5	33
H3K27me3	Fibroblasts	FRA2/FOSL2	49
H4 hyperacetylation	Fibroblasts	COL1A2	47
H3, H4 hyperacetylation	Fibroblasts	NR4A1	50
Histone hypoacetylation	Fibroblasts	WIF1	51
HDAC5 levels	Endothelial cells	CCN1, NECTIN2, FSTL1	46

SSc: Scleroderma or systemic sclerosis; Refs: references; HDACs: histone deacetylases; HATs: histone acetyltransferases; HDMs: histone demethylases; HMT: histone methyltransferase; H3: histone 3; H4: histone 4; H3K9: histone 3 lysine 9; H3K27me3: histone 3 lysine 27 trimethylation; FLI1: Fli-1 proto-oncogene, ETS transcription factor; KLF5: Kruppel like factor 5; FOSL2: FOS like 2, AP-1 transcription factor subunit; COL1A2: collagen type I alpha 2 chain; NR4A1: receptor subfamily 4 group A member 1; WIF1: WNT inhibitory factor 1; CCN1: cellular communication network factor 1; NECTIN2: nectin cell adhesion molecule 2; FSTL1: follistatin like 1.

In B cells from SSc patients, global hyperacetylation of histone H4, and hypomethylation of H3K9 were correlated with levels of histone enzymes like HDAC2 and HDAC7, disease activity, and skin thickness. ⁴⁵ Similarly to what was observed in fibroblasts, KDM6B (aka JMJD3) overexpression in CD4+ T cells of SSc patients was associated with lower levels of the repressive H3K27me3 mark. ⁴⁸

Endothelial cells from SSc patients show increased expression of HDAC5, which alters the regulation of key proangiogenic and profibrotic genes, resulting in impaired angiogenesis. ⁴⁶

Non-coding RNAs

Non-coding RNAs (ncRNAs) are functional RNA molecules that mediate various intracellular processes. ⁵⁵ Classes of ncRNAs include microRNAs (miRNAs), long non-coding RNAs (lncRNAs), small nucleolar RNAs (snoRNAs), and circular RNAs (circRNAs), all with different regulatory functions. The best characterized ncRNAs are miRNAs and lncRNAs. miRNAs (19–24 nucleotides in length) regulate gene silencing of more than 60% of human protein-coding genes at the transcriptional and/or translational level. lncRNAs (over 200 nucleotides in length) are found in sense or antisense orientation to protein-coding genes, or within intergenic regions. lncRNAs function as chromatin remodelers, transcriptional regulators, and post-transcriptional regulators.

Comprehensive lists of ncRNAs dysregulated in SSc patient tissues have been recently reviewed.^5,6 Table 3 summarizes the ncRNAs with target genes that have been reported in SSc; several studies that did not report target genes for the ncRNAs are not included, including all studies evaluating panels of candidate circulating RNAs in serum, plasma, or exosomes (recently reviewed elsewhere ⁵ ). The vast majority of studies in SSc has focused on the role of miRNAs, especially in fibroblasts.

Table 3.

Summary of non-coding RNA with reported target genes in SSc.

Modification	Cell type/tissue	Target genes	Refs
miRNAs
miR-618	pDCs	IRF8	56
miR-7	Fibroblasts, skin, serum	COL1A1, COL2A1	57,58
miR-21	Fibroblasts, skin	SMAD7	59,60
miR-27a-3p	Fibroblasts	ACTA2, SMAD2, SMAD4	61
miR-29a	Fibroblasts, hair	COL1A1, COL3A1	59,62–65
miR-30a-3p	Fibroblasts	TNFSF13B	66
miR-30b	Skin, serum	PDGFRB	67
miR-92a	Fibroblasts, serum	MMP1	68
miR-129-5p	Fibroblasts	COL1A1	69
miR-130b	Fibroblasts, skin	PPARG	70
miR-135b	Fibroblasts, serum, monocytes	STAT6	25
miR-145	Fibroblasts, skin	SMAD3	59
miR-150	Fibroblasts, skin, serum	ITGB3	71
miR-155	Fibroblasts, skin	CSNK1A1, INPP5D	72
mir-193b	Fibroblasts, skin	PLAU	73
miR-196a	Fibroblasts, skin, serum, hair	COL1A1, COL2A1	74–77
miR-202-3p	Fibroblasts, skin	MMP1	78
miR-5196	Monocytes, serum	FOSL2	79
let-7a	Fibroblasts, skin, serum	COL1A1, COL2A1	80
lncRNAs
TSIX	Fibroblasts, serum	COL1A1, COL1A2	81

SSc: scleroderma or systemic sclerosis; Refs: references; pDCs: plasmacytoid dendritic cells; IRF8: interferon regulatory factor 8; COL1A1: collagen type I alpha 1 chain; COL2A1: collagen type II alpha 1 chain; SMAD7: SMAD family member 7; ACTA2: actin alpha 2, smooth muscle; COL3A1: collagen type III alpha 1 chain; TNFSF13B: TNF superfamily member 13b; PDGFRB: platelet-derived growth factor receptor beta; MMP1: matrix metallopeptidase 1; PPARG: peroxisome proliferator activated receptor gamma; STAT6: signal transducer and activator of transcription 6; ITGB3: integrin subunit beta 3; CSNK1A1: casein kinase 1 alpha 1; INPP5D: inositol polyphosphate-5-phosphatase D; PLAU: plasminogen activator, urokinase; FOSL2: FOS like 2, AP-1 transcription factor subunit; TSIX: TSIX transcript, XIST antisense RNA; COL1A2: collagen type I alpha 2 chain.

miRNAs were first implicated in SSc in 2010, when miR-29a was shown to be downregulated in SSc fibroblasts and skin, leading to an increase in type I and type III collagen in fibroblasts. ⁶² Further studies in dermal fibroblasts have suggested that miR-29a is an inducer of apoptosis and an attenuator of extracellular matrix (ECM) production, ⁶⁵ that it reduces tissue inhibitor of metalloproteinases 1 (TIMP1) secretion, a key player in ECM deposition, via targeting TGF-beta activated kinase 1 binding protein 1 (TAB1) repression, ⁶³ and that it increases functional matrix metallopeptidase 1 (MMP1) production resulting in collagen degradation. ⁶³ As summarized in Table 3, expression of miR-7, miR-129-5p, miR-196a, and let-7a in fibroblasts also affect the expression of collagen. It should be noted that the antifibrotic miR-7 was upregulated in SSc ⁵⁷ but downregulated in patients with limited cutaneous SSc, ⁵⁸ suggesting different mechanistic roles for miR-7 depending on specific disease manifestations. Other downregulated, antifibrotic miRNAs include miR-30b, miR-135b, and miR-150. Overexpressed, profibrotic miRNAs include miR-21, miR-92a, miR-130b, miR-155, and miR-202-3p. Among these studies focused on fibroblasts, two miRNAs were shown to target immune genes, and one to target a vascular gene. miR-30a-3p, which was downregulated in SSc fibroblasts, is a basal repressor of the TNF superfamily member 13b gene (TNFSF13B, aka BAFF), thus leading to decreased B cell survival. ⁶⁶ miR-135b, which was downregulated in SSc fibroblasts, targets the transcription factor signal transducer and activator of transcription 6 gene (STAT6), which in turn regulates IL-13-mediated collagen production by fibroblasts. ²⁵ miR-135b is thus important to reduce collagen induction independent from TGFβ1. Interestingly, the authors also showed that methylation regulates miR-135b, underscoring the joint role of multiple epigenetic factors in gene regulation. Finally, the downregulation of miR-193b in SSc skin and fibroblasts promoted the expression of the plasminogen activator, urokinase gene (PLAU, aka u-PA), resulting in proliferation of vascular smooth muscle cells, and thus contributing to the proliferative vasculopathy with intimal hyperplasia characteristic for SSc. ⁷³ The effects of ncRNAs in myofibroblast activation in SSc have been reviewed in detail elsewhere. ³⁷

Analyses of the role of miRNAs in immune cells are emerging. In monocytes, miRNA-5196 was increased and positively correlated with C-reactive protein (CRP) level in SSc patients, which led to inhibition of the FOSL2 gene (aka FRA2) and TIMP1 expression ⁷⁹ . As mentioned above, accumulation of H3K27me3 at the promoter of FOSL2 in fibroblasts also regulates FOSL2 expression, ⁴⁹ once again highlighting the orchestrated control of gene regulation involving multiple epigenetic mechanisms. miRNA expression profiling of plasmacytoid dendritic cells (pDCs) revealed that the observed overexpression of miR-618 suppressed the development of pDCs and increased their ability to secrete interferon α (IFNα), potentially contributing to the type I IFN signature observed in SSc patients. ⁵⁶

A large number of studies have investigated a broad range of circulating miRNAs in SSc.^5,6,82 The resulting profiles of miRNAs can help discriminate patients from controls, patients with different disease subtypes, and with different clinical criteria or disease manifestations. Circulating miRNAs can be found in body fluids such as serum or plasma, which given the easy tissue accessibility, makes them good candidates for disease biomarkers. Circulating miRNAs can also be found in extracellular vesicles such as exosomes, which mediate intercellular communication by releasing their macromolecular content into target cells. Interestingly, serum exosomes from SSc patients contain miRNA displaying a markedly profibrotic profile and induce a profibrotic phenotype in target normal fibroblasts. ⁸³ This ability to alter the phenotype of distant target cells could explain the extension of the SSc-associated fibrotic process to unaffected cells and tissues. ⁸³

Very little is known about the involvement of lncRNA in SSc. In 2016, TSIX (TSIX transcript, XIST antisense RNA) was reported as upregulated in the serum and dermal fibroblasts of SSc patients. ⁸¹ More recently, RNA-seq of skin tissue uncovered 676 lncRNAs differentially expressed between patients and controls. ⁸⁴

Genetic, ethnic, and age effects on the immune epigenome

The epigenome refers to the collection of chemical modifications of chromatin, alterations in chromatin constituents, and changes in the spatial chromatin organization that regulate gene expression. Dynamic regulation of these epigenetic marks underlies cell function and plasticity and provides a heritable response to developmental and environmental cues. This epigenomic regulation is critical for immune cell development and differentiation, immune function in host defense mechanisms, and immune dysregulation exhibited during autoimmune disease. ⁸⁵ Given the need of the immune system to respond prompt and efficiently to stimuli by eliciting rapid immune gene expression, immune cells may be more susceptible to epigenetic reprogramming that might poise them to a sustained, activated phenotype. Evidence shows that many immune gene promoters exist in an epigenetically primed, or accessible state prior to activation, which permits robust gene transcription upon activation by transcription factors. ⁸⁶ As recently reviewed elsewhere, ⁸⁷ this primed epigenetic state is also observed in innate immune genes; interestingly, monocytes in a resting state can be epigenetically primed, and many tumor necrosis factor (TNF)-responsive innate immune genes are engaged in preformed enhancer-promoter chromosomal looping interactions required for gene expression. Since immune system dysregulation critically contributes to early pathogenic events in the progression of fibrosis and SSc, and epigenetic phenomena may explain this early immune system dysregulation, there has been increasing interest in identifying differential epigenetic patterns in autoimmune diseases. ⁸⁵ In SSc, for example, the hypomethylation and overexpression of CD40L in CD4+ T cells of SSc patients might be contributing to the overproduction of autoantibodies that drive disease pathogenesis. ²⁸

Multiple genome-wide association (GWA) and other large-scale studies identified genetic variants associated with SSc. ⁵ These results, coupled with the regions shared with other autoimmune diseases, emphasize the role of dysregulation of immune pathways as a key etiologic factor in SSc pathogenesis. ³ However, the molecular mechanisms linking each variant to the disease phenotype are largely unknown. This is because most variants, including 90% of causal autoimmune disease variants, lie in non-coding regions of the genome and are concentrated in regulatory DNA whose mechanisms are not yet well characterized.^88,89 Since the effects of environmental exposures on phenotypes might be mediated by epigenetic changes, there has been great interest in understanding the role of the epigenome in gene regulation, cellular differentiation, and human disease. Several large-scale initiatives such as the ENCODE, ⁹⁰ Epigenomics Roadmap, ⁹¹ and BLUEPRINT ⁹² projects have mapped gene regulatory elements across a wide range of tissues and cell types using a variety of assays.

The DNA sequence variation can exert effects of gene expression though epigenetic modifications. The influence of genetic factors on methylation levels is well established, as multiple studies have mapped DNA sequence variants associated with DNA methylation variation, that is, methylation quantitative trait loci (meQTLs), across a variety of tissues.^13,93–97 About ~ 20% of the inter-individual variation in DNA methylation has been attributed to genetics.^98,99 In addition, genetic variants associated with complex traits by genome-wide association studies (GWAS) often overlap both expression QTLs and meQTLs, suggesting that disease risk can be mediated, directly or indirectly, by variation in DNA methylation.^100–105

Despite the disproportionate burden on African Americans, the majority of SSc studies have been conducted in populations of European ancestry. This underrepresentation in research continues to exacerbate the health disparities gap, as epigenetic changes have been shown to vary between populations,^106–113 and might be involved in the heightened autoimmune dysregulation in African ancestry relative to European ancestry SSc patients.^114,115 Although differential methylation between ethnic groups is partially explained by their distinct genetic ancestry, environmental factors not captured by ancestry are significant contributors to variation in methylation. ¹⁰⁸ Evidence for the causal impact of ancestry-related differences in DNA methylation on immune gene regulation is emerging. Husquin et al. ¹⁰⁹ unveiled extensive population differences in DNA methylation between individuals of African and European descent, detected in primary monocytes. They report higher levels of DNA methylation in genes involved in the activation and regulation of immune responses in individuals of European ancestry, mostly owing to genetic control. Since up to 16% of immune-related genes that are hypermethylated in Europeans are also differentially expressed between populations, ¹¹⁰ the ancestry-related differences in transcriptional responses to bacteria reported in macrophages, where European ancestry is associated with lower inflammatory responses, ¹¹⁶ might be partially due to these ancestry-related epigenetic differences. ¹⁰⁹ At the time of writing, only one epigenetic study has been conducted in SSc patients of African ancestry: Matatiele and colleagues ²⁶ assessed DNA methylation levels of selected genes in black South African SSc patients, revealing patterns similar to those observed in studies of other ancestries (Table 1).

In addition to genetic background and ancestry, aging can also influence epigenetic patterns. Specifically, thousands of CpG sites whose tissue-specific methylation levels are strongly correlated with chronological age have been previously identified,^117,118 and many of these CpG sites are shared and replicate across a broad range of human populations. ¹¹¹ In SSc, premature activation of aging-associated molecular mechanisms is emerging as an important contributor to the autoimmune, vascular, and fibrotic disease pathogenesis. ¹¹⁹ Notably, the HDACs sirtuins (SIRTs), which are key regulators of the aging process and decline during normal aging, show decreased levels in SSc, ¹¹⁹ supporting the notion that epigenomic dysregulation contributes to disease pathogenesis.

Environmental effects on the epigenome

Multiple chemical and physical environmental stressors, diet, life habits, and pharmacological treatments that can alter the epigenome and determine disease phenotypes have been thoroughly reviewed elsewhere. ¹²⁰ Briefly, environmental pollutants with epigenetic effects include metals (arsenic, mercury, nickel, lead, and cadmium), air pollutants, asbestos, benzene, and electromagnetic radiation (ultraviolet (UV) light). Endocrine disruptors include pesticides (DDT and methoxychlor), fungicides (vinclozolin), herbicides (atrazine), industrial chemicals (PCBs, dioxins), plant hormones (phytoestrogens), and plastics (bisphenol A (BPA) and phthalates). Diet is one of the most studied and better understood environmental epigenetic factors, and well-known examples of diet affecting epigenetic marks include folate (vitamin B9), caloric restriction, and bioactive dietary compounds such as polyphenols. Lifestyle and environmental conditions of the mother during pregnancy (including diet, chemical stressors, endocrine disruptors, smoking or alcohol consumption) affect the epigenome of the offspring,^120,121 and can have long-term effects on the health of the offspring.^122–124 There is evidence that certain medications can also induce genome-wide epigenetic changes (sodium valproate (VPA), diethylstilbestrol (DES), procaine, pyrazinamide, doxorubicin, gemcitabine, cisplatin, and temozolomide). ¹²⁰ Infectious agents (Leishmania and mycobacterium tuberculosis),^14,125 tobacco smoke,^126,127 diesel exhaust particles, ¹²⁸ and other indoor and outdoor pollutants ¹²⁹ have been shown to affect methylation levels. Psychosocial factors, including measures of traumatic experiences,^130–132 socioeconomic status,^133,134 and general perceived stress ¹³⁵ also affect methylation levels.

Notably, exposures during sensitive periods of immune development can have lasting effects on inflammation, and recent empirical work has implicated environments in infancy and childhood as important determinants of inflammatory phenotypes. For example, individuals born at lower birth weight and infants breastfed for shorter durations have higher concentrations of CRP, and major psychosocial stressors and socioeconomic adversity in childhood are associated with proinflammatory activity in adulthood.^136–138 Several studies have documented associations between early life psychosocial exposures and DNA methylation levels of inflammatory genes.^139–141 Collectively, these data support the notion that early environmental exposures might shape adult inflammatory phenotypes through epigenetic mechanisms.

Several environmental exposures have been implicated in the development of SSc; these include silica, organic solvents, epoxy resins, welding fumes, pesticides, silicone breast implants, viruses, and drugs.^142,143 Despite the paucity of studies investigating the effects of many of these exposures on the epigenome, there is growing evidence that environmental factors have a crucial impact on both alterations and modulation of epigenetic determinants, resulting in SSc onset and progression. ¹⁴³ It has been well established that there is a marked correlation between SSc onset and occupational exposure to crystalline silica. ¹⁴³ Although the interactions between silica exposure and the epigenome have not been assessed in SSc, an investigation of associations between silica particles and SSc-associated genetic variants in a fibroblast model unveiled a genetic variant in TNF alpha induced protein 3 (TNFAIP3) to be associated with silica-induced profibrotic response of fibroblasts. ¹⁴⁴ Interestingly, the variant is thought to affect the binding activities of the transcription factors for TNFAIP3, ¹⁴⁴ underscoring the role of regulatory variation on mediating environmental effects on phenotypes.

Organic solvents also have established associations with SSc onset. These are compounds commonly used in dry cleaning (tetrachloroethylene), paint thinners (toluene), nail polish removers and glue solvents (acetone, methyl acetate, ethyl acetate), and spot removers. ¹⁴³ The effects of organic solvents on the epigenome remains unclear. One study reported that exposure to trichloroethylene resulted in DNA hypermethylation on rat cardiac myoblasts. ¹⁴³

Exposure to heavy metals is also associated with SSc, including antimony, cadmium, lead, and mercury. Cadmium, lead, and mercury exposure have all been associated with DNA methylation changes, and cadmium and mercury with changes in miRNA profiles. ¹²⁰ Interestingly, collagen type I alpha 2 chain (COL1A2) promoter DNA hypomethylation has been associated with high exposure to lead. ¹⁴⁵

Viral infections (Epstein-Barr virus (EBV), hepatitis B virus (HBV)) are known to induce epigenetic changes that involve DNA methylation and histone modifications, reprogramming the infected host cells in ways that persist from one generation to the next and leave long-lasting phenotypes.^146,147 Clearly, the deregulation of gene expression is a phenotypic consequence of epigenome plasticity modulated by the virus. Infections have long been speculated to play a role in SSc onset, as several mechanisms, such as endothelial cell damage, molecular mimicry, and self-reactive antibodies, could explain disease etiology. ¹⁴³ It is not known if the pathogens implicated in SSc (parvovirus B19, cytomegalovirus, Helicobacter pylori ((H. pylori)) cause epigenetically driven changes in gene expression that affect the cellular dysfunction seen in SSc. ¹⁴³

Pesticides and BPA, a carbon-based synthetic compound employed to make, among other plastics, the epoxy resins that line metal food and drink cans, are well documented endocrine disruptors that alter DNA methylation levels. ¹²⁰ Future studies should investigate if the apparently increased risk of SSc after exposure to epoxy resins and pesticides is mediated by epigenetic changes. Similarly, although the extent to which particulate air pollution increases susceptibility to SSc requires further assessment, ¹⁴³ the effects of air pollutants on the epigenome of SSc patients ought to be evaluated.

Several case reports have described the development of SSc or SSc-like manifestations after exposure to several drugs, namely some chemotherapy agents (bleomycin, taxanes, gemcitabin, tegafur-uracil, IFNα, and aldesleukin), appetite suppressants, and tryptophan, but there is insufficient evidence to establish a causative role on the development of SSc. ¹⁴⁸ Nevertheless, chemotherapy drugs such as bleomycin or paclitaxel are considered to be associated with the induction of SSc, ¹⁴⁹ although their roles as epigenetic modifiers have not been investigated. Doxorubicin, gemcitabine, and temozolomide have been found to induce epigenetic changes. ¹²⁰ Furthermore, aberrant DNA methylation patterns have been found in the sperm of patients treated with temozolomide, raising caution about the potential transmission of epigenetic alterations to offspring of individuals treated with chemotherapeutic agents. ¹²⁰

Prospects for epigenetic editing

Since the epigenome exerts a pivotal role during immune cell development and modulates functional immune programs during autoimmune disease, and epigenetic modifications can be reversed or modified, they offer an opportunity for the development of targeted therapies. ⁸⁵ Several Food and Drug Administration (FDA)-approved drugs alter DNA methylation and histone acetylation. The epigenetic compounds investigated in the context of SSc have been recently reviewed. ⁵ However, a limitation of the current epigenetic-targeting approaches (e.g. DNMT, HDAC, or HAT inhibitors, or miRNA mimics) is their lack of specificity, as they affect epigenetic marks at many genomic sites indiscriminately. The most commonly used compounds are the DNMT inhibitors 5-azacytidine and 5-aza-2’-deoxycytidine (collectively known as 5-aza). As reviewed,^5,6 treatment of fibroblasts with 5-aza reduces the expression of alpha-smooth muscle actin, collagen genes, and has potent antifibrotic effects.

An alternative to these global pharmacological compounds is the recently developed CRISPR (clustered regulatory interspaced short palindromic repeat)/Cas9-based RNA-guided DNA endonuclease technology. This gene editing technology has been retooled for epigenome editing, including the potential to modulate epigenetic marks and gene expression, manipulate nuclear architecture and chromatin loops, and visualize chromosome organization and dynamics via chromosome imaging. ¹⁵⁰ Use of epigenome editing technologies for cell programming holds unprecedented potential as a therapeutic approach for durable regulation of disease-related genes and in cellular reprogramming. ¹⁵¹ Notably, Farhang et al. ¹⁵² used CRISPR-based epigenome editing to repress inflammatory cytokine cell receptors, specifically TNFR1 and IL1R1, which mitigated the downstream activation of NF-κB in the presence of TNF-α or IL-1β. The authors thus demonstrated that modulating cell response to inflammatory signaling can be used in engineering cells delivered to inflammatory environments, and as a direct gene therapy to protect endogenous cells. This indicates that deregulated loci identified in SSc might become the targets of epigenome editing to inhibit cell-specific dysregulation in SSc. Targeted regulation of deregulated loci using epigenome editing can thus be a potential therapeutic avenue to control fibrosis and autoimmunity in SSc.^151,153

Despite the role of genetic factors and environmental exposures in SSc, an understanding of their effects on gene regulation is lacking. To characterize genetic and environmental risk factors for SSc, it is critical to investigate gene regulation in a relevant cell type under controlled treatment conditions. A tractable way to study genetic and environmental effects on gene regulation and cellular phenotypes is to integrate genetic, epigenetic, and expression data in cells exposed to environmental perturbations, such as the addition of a drug. In SSc, one study has investigated the interactions between genetic and environmental factors (silica particles) in SSc fibroblasts, revealing a genetic variant associated with the silica-induced profibrotic response of fibroblasts. ¹⁴⁴ Epigenome editing might also be used to dissect the regulatory mechanisms involved in disease etiology, as genetic and epigenetic variants can be incorporated into human induced pluripotent stem cells (iPSCs) via genome editing to characterize phenotypic differences after differentiation, under different environmental conditions. Future studies aimed at identifying active regulatory elements for gene expression response to specific environmental perturbations are needed to advance our understanding of SSc risk. Results from these mechanistic studies integrating genetic, epigenetic, transcriptomic, and exposure data are likely to inform the design of epigenetic therapies to mitigate autoimmunity and fibrosis in SSc.

Conclusion

Multiple epigenetic modifications continue to be implicated in SSc pathogenesis. Similarly, genetic risk loci continue to be identified, but molecular mechanisms linking each variant to the disease phenotype are largely unknown. Multiple environmental exposures are associated with SSc, but only a few have been associated with epigenetic changes. Epigenetic modifications provide a mechanistic link between a cell’s genotype and their functional responses to environmental exposures. However, the lack of experimentally validated regulatory mechanisms is hampering the understanding of the regulatory mechanisms underlying SSc. Studies integrating genetic, epigenetic, and gene expression data, in relevant cell subsets, under different exposures, coupled with new epigenetic editing technologies, will provide a better understanding of the regulatory mechanisms underpinning SSc. This knowledge will ultimately inform the design of targeted epigenetic therapies for SSc.