Pathogenesis of systemic sclerosis (scleroderma)

INTRODUCTION — The pathogenesis of systemic sclerosis (SSc; scleroderma) is complex and remains incompletely understood. Immune activation, vascular damage, and excessive synthesis of extracellular matrix with deposition of increased amounts of structurally normal collagen are all known to be important in the development of this illness [1,2]. These mechanisms result from cell-cell, cell-cytokine, and cell-matrix interactions. The heterogeneity in the clinical features of patients with SSc is most likely a reflection of the variable contributions from each of these pathogenic factors.

Most hypotheses of the pathogenesis of SSc focus on the interplay between early immunological events and vascular changes, which result in the generation of a population of activated fibrogenic fibroblasts generally considered to be the effector cell in the disease (figure 1) [2-4]. There is no doubt that vascular and immunologic processes are central to the pathogenesis of scleroderma, although it is unclear what the initial events are and how different processes respectively trigger, amplify, and facilitate the development of the skin- and organ-based fibrosis with vasculopathy that is the hallmark of the disease.

Considering the clinical differences between scleroderma and other autoimmune rheumatic diseases and the relatively modest effects that have been observed in clinical trials of immunosuppressive agents (eg, cyclophosphamide), it is perhaps surprising that genetic and serologic approaches to understanding scleroderma pathogenesis have highlighted the importance of cellular and humoral immunity. Many of the genetic loci associated with scleroderma susceptibility in large-scale genetic analyses are also associated with systemic lupus and other autoimmune conditions [5]. Moreover, even when non-immune genes are associated in subphenotype analyses of genetic loci, these associations are often defined by hallmark scleroderma autoantibodies [6], again supporting the key role of the immune system in the development and clinical features of the disease.

The multiple factors felt to be involved in the pathogenesis of SSc are reviewed here. The possible causes of SSc, as well as the clinical manifestations, diagnosis, and treatment of this disorder, are discussed separately. (See "Risk factors for and possible causes of systemic sclerosis (scleroderma)" and "Clinical manifestations and diagnosis of systemic sclerosis (scleroderma) in adults" and "Overview of the treatment and prognosis of systemic sclerosis (scleroderma) in adults".)

VASCULAR AND ENDOTHELIAL CHANGES — Vascular and endothelial cell changes, primarily mediating vascular tone, appear to precede other features of systemic sclerosis (SSc).

Vasoconstriction — The following mediators of vascular tone may be important in the pathogenesis of SSc [7-9]:

●Endothelins (ETs)

●Nitric oxide (NO, or endothelium-derived relaxing factor)

●Endothelium-derived constricting factors (EDCF)

●Neural, humoral, and inflammatory mediators

●Hypoxia

●Physical stress

Derangements of endothelins, NO, and superoxide anions (which are inflammatory mediators) are felt to be the most significant mediators of altered vascular tone in SSc.

Endothelin — ET-1, a potent vasoconstrictor, is also fibrogenic and is potentially of great importance in the pathogenesis of SSc [8]. The following observations are consistent with this hypothesis:

●ET-1 may contribute both to vascular dysfunction and to the development of the fibrotic lesion in SSc [10].

●The basal secretion of ET-1 from the endothelial cell may provide an important early link between endothelial cell damage and fibroblast activation.

●Significantly elevated circulating levels of ET-1 have been found both in the minimally fibrotic subset of SSc patients with primary vascular disease and associated pulmonary hypertension and in the diffuse fibrotic subset in whom widespread fibrosis is the major hallmark of the disease [11-13].

ET-1 may play an important role in the initiation of fibrosis in SSc as suggested by immunohistochemical and autoradiographic studies. ET-1 has been found in association with the following structures using these special techniques [14]:

●Superficial vessels in both clinically involved and uninvolved skin.

●Its putative receptor, to which it is bound, in the "yet to be involved" skin. This binding decreased with increasing tissue fibrosis.

Nitric oxide — NO balances the vasoconstrictive action of ET-1 in normal blood vessels; a change in the relative amounts of these two compounds is postulated to play a pathogenic role in SSc. However, conflicting data exist concerning the role of nitric oxide. In a preliminary study, total NO-producing compounds were decreased in the plasma of patients with SSc [15]; furthermore, the increase in plasma NO seen after exposure of normal subjects to a cold challenge was absent in these patients. In addition, levels of exhaled nitric oxide correlated inversely with the severity of elevation of pulmonary artery pressure in patients with SSc lung disease [16]. In another study, however, levels of plasma NO were significantly elevated in SSc patients compared with those in controls [17].

NO synthase (NOS) is present in some neurons, as well as within endothelial cells. A progressive decline in cutaneous neurons containing NOS has been noted in patients with diffuse and limited SSc; however, it is uncertain whether this is a cause or a result of diminished tissue perfusion [18].

Superoxide anions — Superoxide anions released from the endothelium may damage the endothelium by being able to neutralize NO and by oxidizing circulating low-density lipoproteins (LDLs). Oxidized LDLs may be the source of the cytotoxicity of stored SSc serum on cultured endothelial cells [19]. Studies have shown that LDLs in patients with SSc are much more susceptible to oxidation (perhaps via free radical attack) than are those from patients with primary Raynaud phenomenon or other rheumatic diseases [20]. In addition, increased urinary and serum levels of markers of oxidative free radical injury (F2-isoprostanes and 8-isoprostane, respectively) have been found in patients with SSc when compared with healthy controls [21,22].

One adverse effect of superoxide anions, inactivation of NO, has been mentioned above. Another is nitrosylation of proteins by peroxynitrite, a product of the reaction of NO and superoxide. Significantly increased concentrations of nitrosylated proteins have been noted in the plasma of patients with diffuse skin involvement but have not been noted in those with limited skin disease or in people with primary Raynaud phenomenon [23].

Defective vasculogenesis — A deficiency of circulating endothelial cell precursors and a defect in the ability of such cells to proliferate and mature into endothelial cells may be an important pathogenic mechanism in SSc. Preliminary evidence suggests that there are significantly fewer circulating endothelial cell precursors in such patients than in healthy controls or in those with rheumatoid arthritis [24]. In addition, cells with surface markers typically associated with a vasculogenic potential (ie, CD34+CD133+ and vascular endothelial growth factor type 2 bearing) that are present in SSc patients less often differentiate into endothelial cells in culture than cells with the same markers obtained from controls.

VASCULAR CYTOTOXIC FACTORS — The sera from some patients with systemic sclerosis (SSc) have been found to be cytotoxic to endothelial cells. The following antibodies, cytokines, proteases, and complement factors may be responsible for this activity:

●Twenty to 30 percent of patients with SSc have circulating antiendothelial cell antibodies. These antibodies by themselves are not thought to be critical, since they are found in the sera of patients with other rheumatic diseases. However, they are capable of upregulating the expression of adhesion molecules on endothelial cells and (in in vitro and in vivo animal experiments) of inducing apoptosis of these cells [25-28].

●The vascular cytotoxicity of some SSc sera is blocked by monoclonal antibodies to tumor necrosis factor (TNF)-alpha or -beta.

●The activity of some SSc sera is blocked by prior incubation with protease inhibitors. The source of these proteases and their nature are unknown, but retroviral proteases and the granzyme family of proteases produced by activated T cells are thought to be possible candidates [29]. Proteolytic activity of granzymes may also degrade angiogenic plasmin and may increase the levels of angiostatin [30].

●Deposition of the complement membrane attack complex (C5b-C9) has been demonstrated immunohistochemically in the microvasculature of involved skin of patients with early and late SSc [31].

ADHESION MOLECULES — Adhesion molecules such as intercellular adhesion molecule-1 (ICAM-1), vascular cell adhesion molecule-1 (VCAM-1), and endothelial leukocyte adhesion molecule-1 (E-selectin) are upregulated in response to cytokines and other factors following inflammation and damage to the vascular endothelium. These endothelial adhesion molecules bind to specific integrins on T and B cells, neutrophils, monocytes, natural killer (NK) cells, and platelets [32]. This interaction results in adhesion and subsequent migration of these cells through leaky endothelium and into the extracellular matrix. (See "Leukocyte-endothelial adhesion in the pathogenesis of inflammation".)

Patients with systemic sclerosis (SSc) have both an increase in endothelial cell surface expression of adhesion molecules and an elevation in circulating levels of their soluble forms [33], as shown in the following studies:

●Positive immunostaining for E-selectin and ICAM-1 has been demonstrated on endothelial cells in SSc skin but not in control skin [34-36].

●Increased circulating levels of E-selectin, VCAM-1, and ICAM-1 have also been observed in small cross-sectional studies of patients with SSc [17,37,38].

●Elevated levels of soluble VCAM-1, E-selectin, and ICAM-1 were found in patients with SSc renal crisis, but not in those with SSc-associated pulmonary disease [39].

It is unlikely that a single measurement will be of particular clinical use or predictive value. However, one study of serial serum samples has indicated a correlation between changes in soluble VCAM-1 and in soluble E-selectin and clinical deterioration or improvement [40]. Such serial studies need to be expanded, as proven markers for disease activity are not available but are urgently required.

Increased expression of messenger ribonucleic acid (mRNA) for some adhesion molecules in the peripheral blood cells of patients with SSc may lead to novel diagnostic tests for SSc. Upregulated expression of genes that encode ligands for P-selectin (SELPLG gene) and receptors for von Willebrand factor (ITGA2B and GP1BB genes) are among a total of 382 differentially regulated genes identified in a study that utilized gene expression profiling of unseparated peripheral blood cells to distinguish between SSc patients and healthy controls [41].

IMMUNOLOGIC ACTIVITY — The continuing activation of endothelial cells, resulting in the upregulation of adhesion molecules, leukocyte adhesion, and leukocyte migration out of the vasculature, probably contributes to the pathogenesis of systemic sclerosis (SSc). It is likely that both innate immunity and adaptive immunity are perturbed in SSc and share a role in pathogenesis. The following studies support this conclusion:

●The expression of endothelial leukocyte adhesion molecule-1 (E-selectin) on endothelial cells correlates with the degree of mononuclear cell infiltration in the early inflammatory lesion of SSc [40].

●Subpopulations of lymphocytes from patients with SSc, namely activated/cytotoxic/inducer T cells, natural killer (NK) cells, and some helper T cells, have a markedly increased ability to adhere to endothelium [42]. The number of highly adherent cells appears to be diminished in the peripheral blood of patients with SSc, as suggested by an overall reduction in adhesion of peripheral mononuclear cells to endothelium in an in vitro assay. It was postulated that this subpopulation of lymphocytes was eliminated from the circulation through enhanced in vivo adhesion aided by chemotactic forces and by a highly permeable endothelium.

●The mononuclear cells from patients with SSc which migrate into the extracellular matrix express differentiation markers of activated T cells including CD3, CD4, CD45, human leukocyte antigen (HLA)-DR, and lymphocyte function associated antigen-1 (LFA-1) [43]. These cells also have subsets of integrin molecules on their surface, including those of the beta 1 and beta 2 class, that facilitate binding to other cells including fibroblasts and tissue components such as type I and IV collagens, fibronectin, and laminin [44,45].

●The peripheral blood of patients with SSc disorders contains a higher proportion of such activated CD3 and CD4 double-positive cells that migrate in vitro into an endothelial cell-covered collagen matrix than do healthy controls (71 versus 56 percent) [46].

●There is growing appreciation that the innate immune system is likely to be involved in pathogenesis of SSc. Genetic and functional studies have provided evidence for infiltration of cells of the monocyte or macrophage lineages in the skin and lung of patients with SSc, as well as evidence of altered expression and function of toll-like receptors [47,48]. Different subpopulations of macrophages may be critical to different stages and sites of disease [49].

Further inflammatory agents including histamines, kinins, complement, antibodies, free radicals, thromboxanes, leukotrienes, oxidized low-density lipoproteins (LDLs), and cytolytic T cells are all possible additional mediators of the deranged immunologic processes present in SSc.

Autoantibodies — Approximately 95 percent of patients with SSc have circulating autoantibodies directed against one or more of a number of antigens. These include topoisomerase I (formerly called Scl-70) (20 to 45 percent) [50], centromere antigens (12 to 44 percent) [50], fibrillarin, ribonucleic acid (RNA) polymerase, PM-Scl, and fibrillin-1 [51], as well as RNA I, II, and III (20 percent) (table 1) [52]. (See "Clinical manifestations and diagnosis of systemic sclerosis (scleroderma) in adults", section on 'Laboratory testing'.)

Although not very sensitive, antitopoisomerase-I (Scl-70) antibodies are highly specific for SSc (98 to 100 percent) and correlate with a higher risk of interstitial lung disease [1]. Higher concentrations or titers are also associated with more extensive skin involvement and with higher disease activity [53]. Anticentromere antibodies (ACA) are associated with limited cutaneous involvement.

The inciting stimulus for the production of these autoantibodies is unknown. One possibility is that the targets for these autoantibodies are autoantigens that have been fragmented via reactive oxygen species and specific metals, such as copper or iron [54]. Immunogenic peptides are subsequently formed which are capable of breaking tolerance, inducing antibodies, and, through epitope spreading, developing antibodies to other epitopes. In addition, the involvement of metals provided a rationale for the use of penicillamine, a metal chelator, in the treatment of SSc.

It is also possible that, in patients with cancer-associated SSc, tumors may induce these autoantibodies by producing somatic mutations, which then induce antibody formation against the mutated protein. As an example, somatic mutations in the POL3RA gene, which encodes RNA polymerase III, may induce anti-RNA polymerase III autoantibodies that may be involved in development of SSc. (See "Clinical manifestations and diagnosis of systemic sclerosis (scleroderma) in adults", section on 'Laboratory testing'.)

Another possibility is that these antibodies arise in response to an infection and, via molecular mimicry, crossreact with a native antigen. In one report, for example, antibodies directed against cytomegalovirus and found in patients with SSc-induced apoptosis in endothelial cells via an interaction with a cell surface protein complex [55].

Immunoglobulin G (IgG) antibodies generally mirror T-cell response to antigen, and most workers believe it is the T cells which will be pathogenic. However, antibodies can, under different circumstances, be directly pathogenic (as described above), can be additive to T-cell damage, or can block T cells [1].

Autoantibodies that interact with fibroblasts could potentially play a pathogenic role in SSc. One study of sera from 69 patients with diffuse and limited SSc, 30 patients with sarcoidosis, and 50 healthy controls found elevated levels of antifibroblast antibodies of IgG-class in 58 percent and IgM-class in 33 percent of those with SSc, but found the same in only 1 of 30 patients with sarcoidosis (3 percent) and in none of the healthy controls [56]. Sera-containing antifibroblast antibodies promoted increased expression of the intercellular adhesion molecule-1 (ICAM-1) on cultured human fibroblasts and increased both messenger RNA (mRNA) and secretion of a number of proinflammatory cytokines.

The autoantibodies discussed above are not necessarily mutually exclusive. Indeed, a strong correlation was noted in one study between the intensity of antifibroblast staining with immunoglobulins from patients with SSc and reactivity of Scl-70 antibodies [57]. Further investigation demonstrated that affinity-purified antitopoisomerase antibodies were able to bind to the cell surface of cultured fibroblasts. Some data suggest that topoisomerase I on the fibroblast surface mediates antifibroblast activity [58].

The presence of antibodies to receptors for platelet-derived growth factor (PGDF) in the serum of patients with SSc may be a relatively specific finding, and profibrotic effects of agonistic antibody binding to the fibroblast PGDF receptor suggest a potentially pathophysiological role for such antibodies. This was illustrated in a study of 46 patients with SSc, 75 controls, and 10 patients who underwent allogeneic bone marrow transplants and who developed graft-versus-host disease (GVHD) with skin involvement [59]. A biologic assay that assessed the stimulatory effect of immunoglobulin G (IgG) was increased (>99 percentile for healthy persons) in all of those with SSc but in none of those with systemic lupus erythematosus, rheumatoid arthritis, primary Raynaud phenomenon, or interstitial lung disease. The stimulatory activity was present in Ig obtained from the 10 patients with GVHD [60].

Other studies have demonstrated presence of agonist antibodies to vascular receptors for endothelin (ET) and angiotensin was associated with poor long-term survival, although the contribution to pathogenesis remains unclear [61]. However, a report demonstrated antibodies against G protein coupled receptors in a number of diseases including SSc and suggested that these may provide a key link between pathobiology in target cells and the adaptive immune system [62].

Other autoantibodies that have been observed include antinucleolar, anti-Th/To, anti-Ku, antiphospholipid, anti-Sm, anti-RNP, anti-Ro, antineutrophil cytoplasmic antibody (ANCA), anti-U1 RNA, antinucleosome, anti-fibrillin-1 (FBN1), anti-matrix metalloproteinases 1 and 3, and anti-endothelial cell specificities [25,63-66].

It is becoming apparent that the clinical diversity seen between the key antinuclear antibody (ANA) subgroups in SSc are associated with relevant differences in gene and protein expression. This may also be relevant in determining potential differential response to targeted therapies such as tocilizumab that have been observed in clinical trials [67,68].

Graft-versus-host disease or microchimerism — Some investigators have postulated that SSc is a variant of GVHD, since both disorders share clinical features such as prominent involvement of the skin, lung, and esophagus. An animal model of GVHD replicates many of the features of human disease [69,70]. Indirect support for this hypothesis is provided both by an increased frequency of persistent fetal cells among women with SSc and a history of pregnancy (as compared with normal women with a history of pregnancy) and by the presence of fetal cells in the involved skin of such patients [71-74]. Cells derived from the fetus may also be found in other tissues of women with SSc, most often in the spleen [75]. These associations, however, do not prove causality [76].

GROWTH FACTORS AND CYTOKINES — The cell-cell and cell-matrix interactions mentioned above may stimulate the production and release of growth factors and cytokines able to mediate the proliferation and activation of vascular and connective tissue cells, particularly fibroblasts. The importance of cytokines in the pathogenesis of systemic sclerosis (SSc) has been strengthened by the realization that the endothelial cell and the fibroblast are both capable of producing factors essential for the development of this illness (table 2).

It had been previously assumed that the major sources of paracrine factors in SSc were inflammatory cells and lymphocytes. However, fibroblasts and endothelial cells can also produce and release cytokines, and they can mobilize growth factors, such as basic fibroblast growth factor, from the extracellular matrix [77]. These observations raise the possibility of reciprocal interactions between all of these cell types, which may be central to disease pathogenesis.

Autocrine or paracrine loops of cytokines may also be responsible for the generation and/or persistence of the SSc phenotype seen after several cell passages in conventional tissue culture [78,79]. A number of possible interactions involving various mediators and several cell types may occur (figure 2). Different loops may also operate within clinical subsets and may be variously expressed at different stages of the disease.

Potential pathogenic cytokines — Large numbers of cytokines have been examined as potential effectors of fibrosis in SSc (table 3). Included in this group are:

●Transforming growth factor-beta (TGF-beta)

●Platelet-derived growth factor (PDGF)

●Interleukin (IL) 1, IL-4, IL-6, IL-8, IL-13, and IL-17

●Connective tissue growth factor

●Insulin-like growth factors (IGF)

●Basic fibroblast growth factor (bFGF)

●Interferon-gamma (IFN-gamma)

●CXCL4

●Monocyte chemotactic protein-1 and -3

●Tumor necrosis factor (TNF)

It is highly unlikely that any one mediator can account for the pathogenesis of SSc, since cytokines are produced by several cell types that may interact through autocrine and paracrine pathways. The following sections address the biological properties of those growth factors and cytokines felt to be most likely to contribute to the development of SSc.

Effects on cell-cell interactions — Cytokines can modulate cell-cell interactions in addition to the direct cellular effects previously mentioned. As an example, IL-1 and TNF-alpha both enhance leukocyte adhesions to endothelial cells, their subsequent extravasation into the extracellular matrix, and the number of activated endothelial cells in SSc lesions.

Effects on adhesion molecules — Cytokines may directly upregulate key adhesion molecules, including intracellular adhesion molecule (ICAM)-1. As an example, fibroblasts from patients with SSc bind to lymphocytes in vitro, through interactions between ICAM-1 and lymphocyte function-associated antigen (LFA)-1 [80]. In addition, IL-1, TNF, and IFN-gamma have been shown to modulate the expression of ICAM-1 by SSc fibroblasts, an effect enhanced by the female sex hormone B-estradiol. This may be an important observation in view of the female predominance of this disease [81].

Effects on the SSc fibroblast phenotype — Many studies have focused on the possible paracrine factors that may initiate or maintain the typical SSc fibroblast phenotype. Early investigations demonstrated the inherent ability of SSc cells to grow in low serum environments and their reduced responsiveness to exogenous PDGF. These findings were explained by the observation that SSc fibroblasts secrete PDGF, thereby rendering them less sensitive to exogenous growth factors [82]. Subsequent work has shown that there is considerable enhancement of the fibroblasts’ mitogenic response to other mediators, including TGF-beta, IL-6, and endothelin (ET)-1.

Transforming growth factor-beta — The marked matrix stimulatory properties of the transforming growth factors, particularly TGF-beta, have implicated these proteins as potentially important mediators in SSc. However, studies of skin biopsies, bronchoalveolar lavage (BAL), and blood samples from patients with SSc have produced conflicting results on the levels of TGF-beta messenger ribonucleic acid (mRNA) and protein, despite the fact that TGF-beta can be readily demonstrated in both involved skin and the fibrotic lung [83-85]. These unexpected results may be explained by the different clinical characteristics of the SSc patients enrolled in the two studies. These patients may have been representative of unique SSc subsets, may have been studied at different stages of their disease process, or may have had variable rates of disease progression. These variables make exact comparison of results from different studies exceedingly difficult.

Even if TGF-beta is an important mediator in the pathogenesis of SSc, it is likely to operate in concert with other cytokines. Moreover, different downstream-signaling pathways may be invoked for short- or long-term (chronic) TGF-beta-mediated effects. A central role for the Smad family of proteins is postulated. Smads are evolutionarily conserved proteins that mediate transcriptional activation by members of the TGF-beta superfamily of cytokines. Some (eg, Smad3) activate intracellular signaling, whereas others (eg, Smad7) are inhibitory.

The role of TGF-beta in the acquisition and maintenance of the SSc phenotype is also unclear. Attempts to induce long-term enhancement of collagen-1 production by fibroblasts via regular pulsed exposure to TGF-beta have failed, as have attempts to demonstrate coordinated regulation of collagen-1 and TGF-beta by SSc fibroblasts [86-88]. Furthermore, collagen and TGF-beta type II receptor mRNA appear to be inducible by TGF-beta in both SSc cells and control cells, indicating that the lack of a sustained elevation in collagen synthesis is not due to lack of responsiveness by the fibroblasts; it may, however, reflect the transient nature of TGF-beta-induced fibrogenesis [89].

The following evidence suggests that TGF-beta may indirectly influence other cytokines, primarily PDGF, to promote fibrogenesis:

●TGF-beta is an indirect mitogen for fibroblasts, acting via interaction with the PDGF-alpha receptor and inducing expression of other mitogenic growth factors including connective tissue growth factor (CTGF, also termed CCN-2).

●Fibroblasts from patients with SSc, in contrast to normal adult newborn foreskin fibroblasts, respond to TGF-beta-1 by expressing greater numbers of PDGF-alpha receptors on their surface, increasing PDGF-alpha receptor protein, and elevating mRNA levels [90].

●TGF-beta increases the mitogenic responses to platelet-derived growth factor (PDGF)-AA. PDGF-AA binds only to the PDGF-alpha receptor, in contrast to PDGF-beta, which binds to both the alpha and beta receptors of SSc cells. PDGF-AA has been found immunohistochemically in SSc dermis near blood vessels and hair follicles but not in normal skin.

Transforming growth factor-beta receptors — The possibility that receptors for TGF-beta on fibroblasts have a role in SSc pathogenesis has also been explored. Expression of type I and type II TGF-beta receptors were increased more than twofold in fibroblasts cultivated from the skin of patients with diffuse SSc when compared with control fibroblasts [91]. Production of collagen mRNA by cultivated fibroblasts could be decreased by anti-TGF-beta antibodies and by transfection with TGF-beta receptor antisense oligonucleotides.

Upregulation of type I and type II TGF-beta receptors also occurs in fibroblasts obtained from lesional skin of patients with localized forms of SSc, such as morphea and linear SSc [92]. The stimulus and mechanism for the increased expression of these receptors are unknown.

Decreased turnover of TGF-beta receptors and a failure of control over receptor signaling could also result in an increased sensitivity to TGF effects. This mechanism was suggested by the finding, in cultured skin fibroblasts of patients with diffuse cutaneous SSc, of impaired ubiquitin-mediated degradation of TGF-beta receptors; and by an absence of negative feedback via inhibitory molecules (Smad7) [93]. Defective induction of Smad7 signaling may also contribute to altered TGF-beta responsiveness in SSc [94].

Increased expression of endoglin, a non-activating TGF-beta receptor, as indicated by modest elevation in fibroblast endoglin mRNA and membrane-bound protein, occurs in sclerodermatous skin, and there is preliminary evidence that these levels increase with the duration of disease. [95]. It is unclear what role, if any, endoglin may play in SSc. Endoglin is important in angiogenesis, and mutations in the endoglin gene are responsible for some forms of hereditary hemorrhagic telangiectasia. (See "Clinical manifestations and diagnosis of hereditary hemorrhagic telangiectasia (Osler-Weber-Rendu syndrome)", section on 'Pathophysiology'.)

Platelet-derived growth factor — These results suggest that TGF-beta may increase the expression of fibroblast populations in SSc by activating the PDGF-AA ligand or PDGF-alpha receptor pathway. PDGF, in turn, has been shown to increase the rate of transcription of bFGF in normal fibroblasts. Both PDGF and bFGF are deposited around blood vessels in the lower dermis of early lesions, suggesting that these growth factors may be related to endothelial injury in SSc.

PDGF also regulates the expression of genes that are important in the mitogenic response. As an example, an increase in ras oncogene product in early lesional skin was noted in association with vascular endothelium and mononuclear cells [96]. Other studies have established that fibroblasts expressing the ras oncogene produce more bFGF, suggesting that bFGF may also play an autocrine role leading to growth stimulation.

Connective tissue growth factor — There is growing evidence for upregulation of the TGF-beta-induced matricellular protein known as CTGF in various forms of fibrosis [97]. Increased production by SSc fibroblasts has been consistently found. Mean plasma levels of an N-terminal fragment of CTGF are markedly elevated in patients with diffuse skin involvement and are moderately increased in those with limited cutaneous disease when compared with healthy controls [98].

Initially, it was suggested that CTGF may be important in sustaining a profibrotic fibroblast phenotype as a secondary cytokine induced by TGF-beta. However, a body of evidence suggests a more complex activity as a matricellular protein exerting some direct effects on fibroblasts by regulating activity or bioavailability of TGF-beta [99] or effects on cell adhesion and fibroblast-matrix interaction [100].

Interleukin 6 and related cytokines — There has been an increased focus on IL-6 as a mediator in SSc based upon the promising early clinical data for use of anti-IL-6 antibody in diffuse SSc [101] and emerging clinical data suggesting that higher levels of serum IL-6 may associated with more severe skin disease [102] or greater progression of lung fibrosis [103]. SSc cells appear to produce much more IL-6 than controls [104], and IL-6 may represent one of the soluble mitogenic factors transferable in culture medium derived from such cells [105]. The role of IL-6 is unclear, but it may modulate activity of other key profibrotic pathways in fibroblasts as well as its effect on immune cells. This may be especially relevant to the immune system, favoring Th17 differentiation or macrophage M2 polarization. Interestingly, attenuation of a fibrogenic M2 macrophage gene expression signature in skin was associated with possible clinical response to blocking IL-6 in a clinical trial [101], and studies of explant skin fibroblasts have shown remarkable attenuation of molecular activation following therapeutic administration of an anti-IL-6 receptor blocking antibody [106]. Other members of the IL-6 family have also been suggested as possible mediators in SSc pathogenesis, including IL-31, oncostatin M, and IL-11 [107].

Interleukin 1 — Interest has also been directed to the production and regulation of the cytokines IL-1 by SSc fibroblasts. IL-1a is also produced excessively by SSc fibroblast and may contribute, via autocrine or paracrine pathways, to their abnormal properties in culture [108,109]. Some evidence suggests that keratinocyte-derived IL-1a is important in regulating fibroblast properties in SSc [110].

Interleukin 4 — IL-4 may have a profibrotic effect on scleroderma dermal fibroblasts. These cells have IL-4 receptors, and, when cultured in IL-4 containing media, the production of collagen is significantly increased [111]. IL-4 shares biologic properties and receptor specificity with the related cytokine IL-13. Clinical trial results suggest that blocking IL-4 and IL-13 may benefit SSc skin fibrosis [112]. These findings may reflect the role of these cytokines in alternative macrophage activation and the M2 profibrotic phenotype [112].

Chemokines — A number of chemokines have been reported to be elevated in the serum and biopsies of SSc skin. CXCL4, also known as platelet-activating factor 4, belongs to the CXC family of chemokines. CXCL4 is a protein that is a potent antiangiogenic chemokine that influences angiogenesis through an integrin-dependent mechanism [113]. It also inhibits the expression of the antifibrotic cytokine IFN-gamma, and upregulates profibrotic cytokines IL-4 and IL-13 [114].

Further, CXCL4 has been implicated in the pathogenesis of SSc based on findings that plasma levels are elevated in patients with SSc, and high plasma levels correlate with progression of both skin and pulmonary fibrosis, as well as pulmonary arterial hypertension [115]. This was demonstrated in a proteome-wide analysis in which CXCL4 levels were measured in patients with SSc and compared with controls, including patients with rheumatoid arthritis, ankylosing spondylitis, and systemic lupus erythematosus. CXCL4 was the predominant protein secreted by plasmacytoid dendritic cells in patients with SSc, both in the circulation and the skin. CXCL4 was also the only chemokine that correlated with faster progression of skin and lung fibrosis. Although interesting, these observations require further confirmation to assess whether targeting this pathway may be helpful in managing the disorder.

The role of plasmacytoid dendritic cells has been an area of interest in the pathogenesis of SSc as they are the major source of type I interferon, which has been implicated in the pathogenesis of several autoimmune diseases [116-119]. However, this may suggest that altered function of these cells is a generic feature of autoimmunity, and further studies are needed to delineate scleroderma-specific importance.

It has been suggested that CC motif chemokine receptor 2 (CCR2) overexpression of monocyte chemotactic protein (MCP)-1 (CC motif ligand 2 [CCL2]) associates with early-stage disease and may promote fibrosis [120]. Interactions with other cytokines including IL-4 have also been demonstrated [111].

mRNA for MCP-3, a member of the CC chemokine family, is overexpressed by dermal fibroblasts cultured from tight-skin mice, a model of SSc [121]. Immunostaining for this protein in the skin of patients and healthy controls demonstrated more MCP-3 in the lesional skin of those with early onset diffuse disease than in those with limited cutaneous involvement or in healthy controls. This chemokine could contribute to fibrosis through effects on extracellular matrix production as a chemoattractant for leukocytes. Chemokine expression in lung tissue may be relevant to the recruitment of fibroblast precursor cells and to the development of lung fibrosis in SSc [122-124].

The chemokine CCL18 has also been shown to predict progression of lung fibrosis [125] and is a potential marker of M2 macrophage polarization that is attenuated in serum by IL-6 receptor blockade [101].

Lipid mediators — There is emerging evidence for the importance of profibrotic lipid mediators in SSc pathogenesis. In particular, the lysophosphatidic acid (LPA) pathway appears to be important based upon preclinical studies and proof-of-concept trials. LPA1 synthesized from arachidonic acid by autotoxin promotes fibrotic pathways including stimulation of IL-6 production. This is supported by clinical trials showing downregulation of profibrotic genes in skin by an LPA1 receptor antagonist associated with potential clinical benefit in skin fibrosis [126].

Summary of the effects of growth factors and cytokines — It seems likely that the interplay between a number of cytokines and growth factors will be found to underlie the pathogenesis of SSc. Increased levels of circulating IL-1, IL-2, IL-2R, IL-4, IL-8, IL-17, TNF-alpha, interferon, and antibodies to IL-6 and IL-8 have been found in patients with SSc [127-130]. However, the presence of high levels of a particular cytokine does not necessarily imply a causative role. In addition, it has become clear that the interpretation of these studies is dependent upon the assay used (bioassay or immunoassay) and upon a realization that many factors interfere with each system. Despite these reservations, it may be inferred that the increased levels of the various cytokines support a cellular immune mechanism in SSc and an ongoing expansion of T cells and of their secreted products.

It is important to remember that, because cytokines influence each other, the pattern of total cytokine production may be more important than any individual cytokine. The soluble cytokines are not exclusive to SSc, being found in many other diseases. However, their presence, pattern of expression, and known interactions may provide clues to the correct cell-type and to relevant tissues involved in disease development and may aid in the development of directed therapy.

FIBROBLAST ACTIVITY AND COLLAGEN SYNTHESIS — The development of the fibroblast capable of excess matrix production and deposition is the hallmark of systemic sclerosis (SSc) (table 4). Fibroblast cultures from both the papillary and the reticular dermis of patients with SSc have been found to produce increased amounts of collagen and other extracellular matrix components; this has been demonstrated in vivo, in biopsy specimens, and in vitro in fibroblast cultures derived from lesional SSc skin. These abnormalities persist through many passages in tissue culture. Normal dermal fibroblasts, in contrast, are relatively quiescent, showing little tendency to proliferate or to elaborate extracellular matrix in the absence of positive stimuli.

Abnormal fibroblast activity in SSc can also be identified in organs other than skin. As an example, increased collagen synthesis by fibroblasts isolated from lung biopsy specimens of patients with SSc lung disease has been demonstrated, thereby suggesting a similar pathogenic mechanism in lung and skin [131].

Increased collagen production is not a feature of all SSc skin fibroblast strains or of all fibroblasts within a culture population; it may, therefore, be limited to discrete subpopulations [132]. Indeed, differences between normal and SSc populations are often more apparent when individual subpopulations are studied rather than mass populations. In situ hybridization studies have demonstrated that high collagen-producing subpopulations of fibroblasts are frequently observed in close proximity to mononuclear cells and adjacent to blood vessels in patients with SSc [133,134].

Activating and inhibitory signals — Increased matrix production by fibroblasts may result from a net excess of positive activating signals and/or from a reduction in inhibitory signals. Positive paracrine mediators may be derived from immune cells, from endothelial cells, or from the fibroblasts themselves. In turn, inhibitory influences may arise from interactions with noncellular extracellular matrix components such as collagen fibers and fibronectin. Such interactions appear to be potent down-regulators of collagen synthesis by normal dermal fibroblasts and also modulate cell morphology, proliferative capacity, and cytokine responsiveness. It has been suggested that down-regulation of nuclear hormone receptors, including peroxisome proliferator-activated receptor gamma [135] and others, may provide a mechanism by which there is a failure of appropriate feedback inhibitory signals in activated fibroblasts in SSc [136]. Despite promising preclinical data, a clinical trial did not suggest benefit in SSc from a peroxisome proliferator-activated receptor (PPAR) agonist, suggesting this may not be a critical primary defect in SSc [137].

The SSc fibroblast — One potential mechanism for the development of the activated fibroblast places the initiating event in the vascular bed, leading to growth factor and cytokine production with resulting fibroblast activation and with subsequent fibrosis. Other studies have added complexity by demonstrating that fibroblasts involved in tissue repair may be derived from diverse origins, including bone marrow-derived progenitor cells, local epithelial cells, or resident progenitor or stem cell populations; however, a unifying mechanism may be the dependence of these cell types on transforming growth factor-beta (TGF-beta) and its regulated pathways and mediators [138]. Thus, deletion of TGF-beta responses by genetic modification in animal models appears to attenuate fibrosis and scar formation [139,140]. This may have implications for novel treatment strategies [141]. While this unifies several of the different pathogenic events seen in SSc, it cannot completely explain the abnormal phenotype of the SSc fibroblast as indicated by the following observations:

●Although dermal mononuclear cell infiltrates are common in vivo, they are not universal, and in the later stages of the disease there is ongoing dermal fibrosis in the absence of apparent immune cell activity. Furthermore, SSc fibroblasts in vitro are often less dependent upon serum supplementation and other growth promoting factors than normal, healthy fibroblast cultures [82].

●In another study, the collagen-secreting ability of SSc fibroblasts continued at the same rate whatever their feeding conditions, unlike control fibroblasts which increased their collagen output briskly in response to frequent serum replenishment [142].

It therefore appears that immune cell or other influences may lead to the development of a relatively autonomous activated fibroblast strain. Several of the following possibilities may explain such behavior:

●Soluble fibroblast products, which maintain the SSc phenotype, may have been induced, allowing an autocrine feedback loop to become established. However, no such mechanism has been conclusively shown, although several cytokines have been considered, particularly interleukin (IL) 1, IL-6, platelet-derived growth factor (PDGF), and TGF-beta.

●A second possibility is the induction of a permanent dysfunction of some intracellular regulatory gene(s), such as a proto-oncogene or its product, which then acts to increase matrix synthesis. As an example, cultured fibroblasts from patients with SSc have more immunoreactive protein kinase C-delta (PKC-delta) than those of healthy controls; increased transcription of type I collagen genes in cells derived from these patients can be abrogated in vitro by a selective inhibitor of PKC-delta, rottlerin [143].

●A third potential mechanism is an immune cell-mediated selection of a subpopulation of fibroblasts already committed to high levels of collagen production or preferential proliferation. The clonal expansion of such fibroblast subsets under immune, hypoxic, or other influences may perpetuate the fibrotic processes in vivo as well as the experimental situation in vitro.

The metabolic heterogeneity of fibroblasts is well-established, and analysis of fibroblast clonal populations shows marked heterogeneity of cell surface markers, proliferation, collagen production, collagenase synthesis, and prostaglandin E2 biosynthetic responses to IL-2 or other mediators [132]. Several studies on clonal selection in SSc fibroblasts have been performed with varying results.

●A fourth possibility is that the interaction of fibroblasts with the extracellular matrix alters cellular function, including the control of matrix deposition and the response to cytokines. The question has been raised as to whether SSc fibroblasts may differ in their adhesion to matrix. In general, interactions of cells with matrix proteins involve specific cell membrane proteins that belong to different families of receptors such as integrins, cell surface proteoglycans, and anchorins. This is a very complex process involving different domains (eg, RGD [arginine-glycine-aspartate] sequences) and various receptor molecules which then mediate different cellular signals that are transmitted to the nucleus. It has been suggested that stiffness of extracellular matrix may be important in generation or persistence of myofibroblasts phenotype. This is based on observations that suggest that mechanical stiffness of fibrotic skin or lung may promote further fibroblast activation or myofibroblasts persistence [144]. It has been suggested that altered matrix proteins may also contribute to innate immune activation relevant to persistent fibrosis [145].

Microfibrils of fibrillin-1 produced by fibroblasts from uninvolved skin of patients with SSc may be less stable than fibrils produced by similar cells of those without disease [146]. The mechanism for the observed instability is uncertain. Findings that suggest that the fibrillar instability may be pathogenic include the association between SSc and genetic markers near the fibrillin-1 gene among Choctaw Indians [147], as well as a tandem repeat in the fibrillin gene of the tight skin mouse [148]. (See "Risk factors for and possible causes of systemic sclerosis (scleroderma)".)

●A fifth possibility, for which there is some support, is that a stable epigenetic change leads to a profibrotic phenotype of the fibroblast. Hypermethylation of deoxyribonucleic acid (DNA) and deacetylation of histone proteins, changes that decrease the transcriptional activity in the region of a promoter of a gene (FLI1 gene), have been noted in cultured fibroblasts from patients, and hypermethylation of the promoter was present in samples of skin from patients but not in three control cell lines [149].

●As in other complex diseases, new insights are emerging from studies using high-dimensional integrated analysis of proteomics and transcriptomics from skin and blood. These approaches are providing powerful insight into disease pathobiology and also identifying potential molecular markers of key pathogenic mechanisms [150]. Single cell approaches are proving especially powerful in defining potentially important fibroblast subpopulations and elucidating critical regulators and pathways that may be targeted therapeutically [151,152].

Models of extracellular matrix — Another area of investigation uses models to simulate the extracellular milieu seen by SSc fibroblasts. The extracellular matrix may be simulated by the use of three-dimensional collagen gels. When normal fibroblasts are seeded in such gels, they contract the gels, change their morphology, and modulate several of their specific functions, including protein synthesis and a marked reduction in collagen-specific mRNA levels. When SSc cells are seeded in a similar fashion, they contract the gel but do not downregulate collagen synthesis compared with normal cells.

This impairment in downregulation may reflect one of the following [153]:

●Changes in the a1b1 integrin expression on the fibroblast cell surface

●Altered intracellular signaling secondary to integrin-ligand binding

●Altered regulation of collagen gene transcription in these cells

It has also been suggested that altered stability of the collagen mRNA transcript may be involved in determining steady-state mRNA levels in three-dimensional collagen gel versus monolayer fibroblast cultures [154].

Other relevant areas of investigation include studies of the complex regulation of extracellular matrix degradation. Early reports suggested that matrix degradative processes were not altered in SSc; however, several studies suggest that this may not be the case [155,156]. As an example, one product of matrix degradation, endostatin (a fragment of type XVIII collagen and an inhibitor of angiogenesis), was increased significantly in patients with SS compared with healthy controls (53.2 versus 9.9 ng/mL) [157].

SUMMARY

●The pathogenesis of systemic sclerosis (SSc) is better defined at a molecular level and involves multiple cell types from the innate or adaptive immune system, vasculature, and connective tissue.

●Fibrogenic fibroblasts ultimately mediate the fibrotic pathology of SSc, but the factors determining inappropriate activation and the relative importance of enhanced activating influences or reduced down-regulatory signals are not well understood.

●Treatment for almost all aspects of SSc remains inadequate. Elucidation of the precise interplay between the immunologic, vascular, and fibrotic components of SSc is likely to be an important first step toward more effective therapy. As key growth factors or chemokines are identified, there may be new opportunities for a more targeted approach to therapy. The ultimate goal of targeted subset and stage-specific therapy will depend upon progress in the understanding of the complex pathogenesis of this disease.