Introduction

With a prevalence of 4 per 10,000 and a sex ratio (F/M) ranging from 5:1 to 12:1 [1,2], systemic sclerosis (SSc) is a rare and severe autoimmune disorder characterized by progressive fibrosis, vascular damage, and immune abnormalities [3–5]. SSc is a multifactorial disease, and genetic and environmental factors are known to interplay in the onset and progression of SSc [6]. Breaking of epigenetic homeostasis by environmental agents causes alterations in gene expression that would contribute to the development of SSc.

Sex plays an important and determinant role in the development of such a disease. SSc, as many other autoimmune diseases, shows a significant female preponderance. However, the reason for this female preponderance is incompletely understood.

Different factors may be involved in this sexual dimorphism, such as sex and reproductive hormones [7,8], fetal microchimerism [9,10], and different environmental exposure [11].

This review focuses mainly on the mechanisms—exclusively linked to the X chromosome—which contribute to the development of systemic sclerosis. For a detailed discussion on other mechanisms involved in the susceptibility to autoimmune diseases, including SSc, we suggest the reading of other excellent reviews [2,12,13].

The first part of this review deals with epigenetic regulation changes, such as X chromosome inactivation and loss of epigenetic control, which could contribute to the susceptibility of SSc. Then, a section is dedicated to the sex hormones, which could be potentially responsible for the sex-dependent discrepancy of disease. Sex hormones are known to influence genes on sex chromosome, thus exerting an effect on autoimmunity. Finally, a brief description on X-linked single nucleotide polymorphisms associated with the susceptibility to SSc is given.

The Female Preponderance in Systemic Sclerosis

SSc shows a significant female preponderance [14]. However, as shown in Table 1, sex ratio (F/M) for this disease ranges from 4 to 11. The female preponderance, such as other SSc clinical features, varies in different geographical areas and ethnicities [31], thus reflecting the possibility that genetic factors may underlie these differential phenotypic aspects. Another important factor to be considered is the population age. In fact, a female population in childbearing age is known to increase the F/M ratio [32]. Moreover, different criteria for definition and classification of SSc may make sex ratios imprecise and biased [23,33].

Table 1 Representative summary of sex ratios (F/M) in systemic sclerosis reported in the last 10 years and chronologically ordered
Full size table

It is noteworthy that female preponderance reflects the different incidence of the disease between the sexes, and not severity. Male SSc patients have a more severe prognosis compared to female ones [15,34]. In conclusion, we cannot exclude that high rate of mortality in male population could influences the sex ratio of SSc.

X Chromosome and Dosage Compensation

The X chromosome contains 1,098 genes, 70 % of which have been associated to various diseases, whereas Y chromosome encodes only a small number of genes, about 100 genes, most of which are different from the ones encoded by the X chromosome [35].

In order to achieve a dosage compensation in females, one copy of X chromosome is randomly inactivated by transcriptional silencing during early development [36,37]. More than 70 % of genes on the X chromosome are silenced. In females, the effect of dosage compensation is the occurrence of a mosaic expression of maternally or paternally derived X chromosome in different cell populations.

The mechanisms involved in this inactivation have not been fully clarified. DNA methylation, histone acetylation, methylation, and phosphorylation, as well as microRNAs, contribute to the initiation and maintenance of X chromosome inactivation [38–40].

Briefly, X chromosome inactivation begins at position Xq13.3, in the X Inactivation Center, a complex locus producing a large untranslated RNA, the X-inactive Specific transcript (XIST) and its antisense (TSIX) [37]. The XIST gene is expressed in the chosen inactivated X chromosome, and its RNA transcript binds to chromatin in order to inactivate it [41]. As proposed by Lyon [42], long interspersed repeat elements in the X chromosome may serve as specific signals for gene inactivation. However, this model is not sufficient to explain the mechanisms underlying gene inactivation, since these elements seem not to be essential for Xist localization [43]. Long terminal repeats and inverted repeat sequence represent additional regulatory sequence which may be involved in XCI [44,45].

Subsequently, several epigenetic modifications which contribute to the X chromosome inactivation and maintaining take place [46]. Following its recruitment by Xist, the polycomb complexes cause some histone changes on the X chromosome [47]. DNA methylation is required for stable maintenance of XCI only in embryo.

For the presence of one copy of X chromosome, males are more exposed to deleterious mutations on the X chromosome [48]. Therefore, having two copies of X chromosome, as in females, may be advantageous. Furthermore, a drastic skewing of X chromosome inactivation could facilitate, in females, the silencing of a harmful mutation. However, as discussed in the next section, the presence of an altered X chromosome inactivation in females may play an important role in the increased risk of autoimmunity. Furthermore, it has been shown, through transgenic mouse models, that XX complement confers greater susceptibility to autoimmune diseases, as compared with XY [49]. One explanation, discussed in detail in next sections, of this discrepancy between males and females is that female X-linked self-antigens may bypass their presentation in the thymus, thus avoiding the central tolerance induction.

X chromosome contains many important immune-related genes, such as interleukin 2 receptor γ (IL2RG), forkhead box P3 (FOXP3), CD40 ligand (CD40L, CD154), and interleukin-1 receptor-associated kinase 1 (IRAK1) [50–53]. As shown in Table 2, many of these X-linked and immuno-related genes have been associated with SSc.

Table 2 X-linked genes associated with SSc with relative references
Full size table

Skewed X Inactivation

Skewed X chromosome inactivation (sXCI) is the preferential inactivation of one X chromosome in females, leading to a deviation from the theoretical 1:1 ratio [66]. Although skewed X inactivation is considered a normal process, it can cause hemizygosity, a possible condition which could give rise to some X-linked recessive disease [67]. Moreover, although this phenomenon is considered to be stable for all descendant cells, the frequency of sXCI, present in peripheral blood cells, increases with age [68].

In sXCI, some specific genes show different patterns of inactivation, which could contribute to the heterogeneity in gene expression among females [69], thus influencing immune regulation. Until now, there is no evidence for a genetic mechanism involved in human sXCI, and most of observations on X chromosome inactivation skewing suggest the occurrence of mechanisms acquired secondarily [70].

It has been hypothesized that in random X inactivation, the tolerance mechanisms are well performed because of the expression of self-antigens on both X chromosomes. However, under nonrandom X inactivation conditions, such as in sXCI, thymic dendritic cells which express one copy of X chromosome fail in recognizing thymic cells, thus causing some thymocytes to escape from negative central selection [71–73] (Fig. 1a).

Fig. 1
figure 1

Several hypotheses, linked to X chromosome abnormalities, have been proposed to explain the preponderance of autoimmune disorders in females. Such hypothesis may be applied also for SSc. a sXCI may induce a loss of mosaicism in females, leading ultimately to autoimmunity. According to this model [43–45], females with a normal (random) XCI have two subpopulations of thymic dendritic cells (DC) expressing maternal (m) or paternal (p) X-linked self-antigens. Under this condition, potential autoreactive thymocytes (thy) undergo a negative selection. On the contrary, in females with a severe sXCI, thymic dendritic cells expressing self antigens of one parent, will tolerize only the thymocytes containing the self-antigens from other parent. Such thymocytes will escape thymic negative selection. b Reactivation of genes from inactive X chromosome could cause an overexpression of X immunorelated genes [74,75]. c Acquired haploinsufficiency, caused by X monosomy, may induce autoantibodies production by peripheral B cells (BC) [76]. Each mechanism is not sufficient to explain SSc development, and one of them must be regarded as a possible cofactor, together with other genetic and epigenetic factors, in the onset of the disorder

Full size image

sXCI could be involved in the pathogenesis of SSc. Ozbalkan et al., with the analysis of the methylation status in the androgen receptor gene, have shown an association between X inactivation and female predisposition to systemic sclerosis. In particular, they found a high rate of a skewed pattern of X chromosome inactivation in DNA of peripheral blood cells from patients with SSc compared to the control subjects [77]. Broen et al., showing a sXCI pattern in female patients affected by SSc, have found that such a skewed pattern was not restricted to a specific cell population, thus suggesting the occurrence of this inactivation in precursor cell populations. Moreover, they observed that the sXCI in SSc was associated with decreased expression of FOXP3, a member of the forkhead family of transcription factors and essential for regulatory CD4+ T cells [78].

What is the relationship between skewed X inactivation and SSc? According to Ozcelik, the failure of self-tolerance could arise from the skewing and not vice versa because of the extreme degree of skewing in SSc patients. Moreover, this author speculated that any mutation, which compromises cell survival, could induce skewed X inactivation [73].

In conclusion, it is noteworthy that there is a strong interplay between non-X-linked genes and X inactivation. In particular, non-X-linked mutations may influence X inactivation and expression of X-linked genes, through their ability in regulating transcriptional and/or translational events. Similarly, molecular abnormalities on the X chromosome may influence gene expressions from other chromosome [79].

Escaping Gene Dosage Compensation

It is known that about 15 % of genes escape X chromosome silencing in female humans. Furthermore, an additional 10 % of X-linked genes show a variable individual- and/or tissue-dependent profile of inactivation [69]. Thus, females may show substantial heterogeneity in X-linked gene expression. Although in normal conditions X chromosome inactivation is stably maintained, many environmental factors may induce to a loss of epigenetic control, which can reactivate silenced X-linked genes.

Gene escaping inactivation is a nonrandom event. X-linked gene escaping inactivation is known to be clustered and preferentially located in the distal portion of the X chromosome short arm [69]. The mechanisms involved in XCI escaping are unclear. In addition to the long interspersed repeat elements and the long terminal repeats, AT-rich motifs seem to be involved in escaping XCI with the control of Xist RNA recruitment [80]. Furthermore, other noncoding RNA may play an important role in gene escaping inactivation [81]. Finally, boundary elements, acting as insulator sequences, may arrest the spreading of heterochromatin into region escaping inactivation [82].

Several genes escaping from XCI have been identified, including tissue inhibitor of metalloproteinases 1 (TIMP1), Il-1R associated kinase (IRAK1), and inhibitor of nuclear factor-κB kinase-γ (IKBKB), which have been shown to be associated to the development and/or to the different clinical subsets of SSc [59,83,84].

These phenomena, resulting in sex-biased gene expression, could explain sex-specific phenotypes in complex diseases [85]. An altered gene dosage of X-linked genes may be implicated in the loss of immune tolerance and in the female preponderance to develop autoimmune disorders [48,86].

The reversal of X chromosome inactivation, X chromosome reactivation (XCR), is a critical step for diversifying X-linked gene pool in germ cell formation [87,88]. Mechanisms involved in X chromosome reactivation remain unclear. In some autoimmune diseases, it has been demonstrated that X-linked genes may be reactivated by DNA methyltransferase inhibitors [89].

Selmi et al., with a genomewide fine mapping of DNA methylation, found gene hyper- or hypomethylation exclusively located on X chromosome of SSc monozygotic twins [61].

The interaction between CD40 and its ligand (CD40L) plays a role in SSc by stimulating B cells and fibroblasts [90,91]. CD40 and CD40L expression has been shown to be increased in fibroblast from SSc patients [91,92]. Furthermore, CD40L expression was demonstrated to be elevated only in SSc female patients [75]. Such a gene reactivation on the inactive X chromosome would double the expression of related proteins in women [74], thus providing an explanation for the female predominance of SSc [75] (Fig. 1b).

Hypermethylation of the CpG islands and histone deacetylation in the promoter region FLI1, a collagen suppressor gene, were shown to occur in SSc fibroblasts and skin biopsy specimens by Wang et al. [93]. On the contrary, a reduced global DNA methylation was observed in CD4+ T cell DNA from SSc patients [60].

X Chromosome Monosomy

In sXCI, some specific genes show different patterns of inactivation, which could contribute to the heterogeneity in gene expression among females [69], thus influencing immune regulation. By using fluorescence in situ hybridization, a higher rate of monosomy was found in peripheral blood cells from SSc patients in comparison to healthy subjects [76]. Moreover, these increased monosomy rates have been mainly found in lymphocytes.

An unusual case of morphea, a localized form of SSc, has been found to be associated with a low-grade mosaic Turner’s syndrome (karyotype 45, X0) [94]. In conclusion, X chromosome haploinsufficiency, caused by X chromosome monosomy, may be involved in the development of SSc (Fig. 1c).

X-linked Noncoding RNA

Recently, noncoding RNAs (ncRNAs) have attracted attention for its involvement in the pathogenesis of various diseases, such as autoimmune disorders. Noncoding RNA, regulatory elements of gene expression, includes microRNA (miRNA), long noncoding RNA (lncRNA), circular RNA [95,96].

About 10 % of microRNA, a family of small double-stranded noncoding RNA, is localized on X chromosome, suggesting a possible role in regulating gene expression, as by transcriptional gene silencing [97,98]. X-linked miRNA may be involved in immune regulation. It is possible that X-linked miRNAs, which escape inactivation or are subject to skewed X inactivation, may influence immune response in females [99].

Recently, many miRNAs—some of which are involved in autoimmune and fibrotic processes, and located on X-chromosome—were observed to be dysregulated in skin tissue from SSc patients, and different regulation patterns were associated to the different subtypes of the disease [100–102].

ncRNAs, together with the above discussed epigenetic mechanisms, are known to be important players in the homeostasis and function of the immune system. In particular, ncRNAs are molecular regulators involved in many autoimmune diseases. Since many studies on these molecules rely on their expression profile in patient peripheral blood mononuclear cells and target tissues, it is hoped that further studies on various cell subpopulations, such as on lymphocyte subpopulations, may help to elucidate the mechanisms responsible of the pathogenesis of autoimmune diseases, including SSc.

Sex Hormones and X Chromosome

Sex chromosomes and sex hormones are profoundly interconnected. It is known that some X- and Y-linked genes are needed for the differentiation of gonads, which in turn are needed for the synthesis of sex hormones, as testosterone and estrogens. Sex hormone-induced signalling pathways may modulate genes on sex chromosomes [103].

It is known that females have stronger immune response compared with males. Such differences may be caused by sex hormones, through their modulation of Th1/Th2 response [11]. Generally speaking, androgens could promote a Th1 response and activation of CD8+ cells. On the contrary, estrogens could favor a Th2 response [103–105].

It was suggested that the deep partnership between sex chromosomes and sex hormones would reciprocally antagonize their effects in order to minimize the differences between male and female immune response [106]. Estrogens play an important role in lymphocyte maturation and activation, as well as in the synthesis of cytokines and antibodies [107,108]. Moreover, estrogen receptors participate in innate and adaptive immunity, regulating activity of antigen-presenting cells and dendritic cells [109].

Such a sex-biased difference in immune response could explain the predominance of autoimmune disorders in females. It is known, in fact, that estrogens, inducing an extramedullary hematopoiesis, could favor the autoreactive B cell escaping from negative selection. Moreover, estrogens are able to promote the survival of the autoreactive T cells [110].

Since SSc occurs in women in childbearing age more frequently than in men, it was suggested a pathogenic role of sex hormones in this autoimmune disease. It is well known that SSc patients show an altered hormonal status. In particular, 17β-estradiol would exert a profibrotic effect on normal and SSc fibroblasts [111].

Decreased levels of serum dehydroepiandrosterone in males, as well as in females in childbearing age, have been associated with SSc [112]. Similarly, decreased serum levels of testosterone and increased levels of prolactin were observed only in female SSc patients [113]. Serum levels of prolactin were found to be significantly elevated in SSc female patients than in control subjects. Such a hyperprolactinemia correlated positively with aggressiveness of skin involvement [114]. Afterwards, other authors showed that lymphocytes from SSc patients were additional active sources of prolactin. In these patients, lymphocyte-derived prolactin would stimulate the same lymphocytes to increase the synthesis of interleukin-2 receptor [115].

X-Linked Single Nucleotide Polymorphisms

It is well established that genetic predisposition is an important factor for SSc development. Single nucleotide polymorphisms of genes located on the sex chromosomes can directly influence the SSc susceptibility. These X-linked genes may have a predominant role in sex bias of SSc. Moreover, many of these SNPs may characterize distinct clinical SSc phenotypes.

Significant association was found between the rs638376 of interleukin-13 receptor subunit α2 (IL13RA2) and SSc in a female Caucasian population [58]. These authors have shown also an association between the SNP rs5946040 of IL13RA2 and dcSSc subtype. It is known that interleukin-13 plays an important role in normal tissue repair, as well as in fibrosis [116,117], and its serum level was found increased in SSc patients compared to control healthy subjects [118].

Interleukin-1 receptor-associated kinase 1 (IRAK1), with its modulating activity on NF-KB, is known to be involved in TLR pathway [119]. An association between some IRAK1 SNPs and diffuse cutaneous SSc and anti topo-I positive SSc patients was found [59]. However, the frequencies of IRAK1 SNP rs1059702 have been found associated only to the susceptibility to SSc with lung fibrosis [120].

The X-linked methyl-CpG-binding protein 2 (MECP2) encodes a protein that binds specifically to methylated DNA, thus participating in the epigenetic mechanism, which may be responsible cofactors for the onset of autoimmune disorders [121,122]. The rs17345 of MEPC2 was found to be associated to diffuse cutaneous SSc subtype in females of Caucasian ancestry [120].

It is noteworthy that IRAK1 and MEPC2 activities are strictly interrelated; thus, a variation of one gene may influence the expression of the other one, and vice versa [123]. Furthermore, since these two genes are in high linkage disequilibrium, the real contribution of these genes to the susceptibility to SSc remains still controversial [59,123].

The transcription factor forkhead box P3 (FOXP3), a member of the forkhead/winged-helix family, is widely known for its role in the development and function of regulatory T cells, which play an important role in peripheral tolerance by suppressing immune response of self-reactive T cells [124]. Moreover, alterations in regulatory T cells number as well as FOXP3 expression have been shown in SSc [125–129] and its clinical subtypes [130].

Recently, we have shown that rs2280883 of FOXP3 gene could be associated to the susceptibility to SSc in a Caucasian female population, with a preferential association with ACA+ patients and with those suffering from lcSSc. Unfortunately, for the low number of recruited patients, we cannot rule out such a possibility for male patients [131].

Tissue inhibitor of metalloproteinases 1 (TIMP-1) belongs to the TIMP gene family. The main function of its encoded protein is to inhibit the matrix metalloproteinases, peptidases involved in degradation of the extracellular matrix [132]. Dysregulation of TIMP-1 has been found in serum and/or skin biopsies of SSc patients [62,133,134].

An association between the rs4898 of the TIMP1 gene with SSc as well as with SSc-associated digital ulcer formations has been shown in a female Italian Caucasian population [135]. As mentioned above, TIMP-1 is known to undergo a polymorphic inactivation in females, thus leading to a gene dosage skewing in females [83]. In agreement with polymorphic X chromosome inactivation, rs4898 of TIMP-1 has been shown to be accompanied with a lower TIMP-1 protein expression in males but not in females in inflammatory bowel disease [136].

Conclusions

SSc is a complex, heterogeneous, and multifactorial disease. Both genetic and epigenetic factors would contribute to the sex-specific difference in SSc incidence. As already mentioned, many sex- and immune-related genes, crucial for the maintenance of physiological levels of sex hormones, as well as of immune tolerance, are located on the X chromosome. Epigenetic X-linked abnormalities, such as skewing of X inactivation and/or gene reactivation, could play an important role in predisposing females to develop SSc (Fig. 2).

Fig. 2
figure 2

Interplay between genetic and epigenetic factors in the influence of susceptibility to SSc. Only X-linked factors are shown. Genetic backgrounds are source of predisposition to the disease. Epigenetic and environmental factors are needed for the trigger of SSc (SNPs single nucleotide polymorphisms; XCI X chromosome inactivation)

Full size image

Thus, the advantage of having a stronger immunoresponse to infections in females reflects the possibility of having a higher susceptibility to SSc. Understanding the mechanisms which underlie X chromosome abnormalities would be very useful for developing therapeutic strategies for autoimmune disorders, including SSc. Moreover, because of reversible nature of epigenetic modifications, potential therapeutics treatments could be developed for reversing these modifications, which could be involved in the pathogenesis of SSc. For example, modulation of miRNAs could offer a new potential benefit for selective therapies.

Finally, in the development of therapeutic strategies differences in the immune response between the sexes must be considered because they may differently affect pharmacokinetics and -dynamics in a sex-dependent manner.