Molecular and cellular basis of scleroderma
- First Online:
- Cite this article as:
- Eckes, B., Moinzadeh, P., Sengle, G. et al. J Mol Med (2014) 92: 913. doi:10.1007/s00109-014-1190-x
- 717 Views
Systemic sclerosis (scleroderma) is a chronic inflammatory disease that leads to fibrosis of the skin and involved internal organs. No efficient therapy is currently available. This review summarizes recent progress made in basic as well as clinical science and concludes with a concept that therapy targeting fibrosis in scleroderma needs to take into account the global microenvironment in the skin with its diverse cellular players interacting with a complex extracellular matrix environment and matrix-associated growth factors.
KeywordsSystemic sclerosisSclerodermaExtracellular microenvironmentPathophysiologySignaling
The clinical spectrum of sclerodermatous disease
Systemic sclerosis (scleroderma)
Scleroderma is not an inherited disease and does not follow Mendelian inheritance; however, several independent investigations suggest that a genetic background predisposes to the development of the disease. Family studies indicate an increased probability to develop scleroderma in families with other autoimmune diseases . Moreover, inbred populations exist with a very high frequency of certain subsets of scleroderma [8–11].
More recently, genetic linkage studies and genome-wide association studies (GWAS) have identified polymorphisms associated with the predisposition of patients to develop systemic sclerosis [11–16]. Some of those have identified genes associated with the metabolism of extracellular matrix (ECM) molecules [11, 17, 18]. As for other autoimmune diseases, most investigations provide evidence for an association with genes coding for proteins involved in the control of innate immunity, macrophage activation, and T-cell functions [13, 14, 19–21]. For further improvement, several large registries have been developed worldwide [22, 23], which provide a detailed characterization of patients, defining subsets and allowing controlled follow-up clinical studies. These networks enable additional validation in large well-characterized patient cohorts and provide the basis for mechanistic molecular studies.
Environmental risk factors and trigger mechanisms
In addition to a predisposing genetic background, an early trigger is likely required to initiate the disease process. Various reports have correlated environmental factors with the incidence of scleroderma or related diseases [24–26]. These include environmental challenges such as vinyl chloride, silica, viral (e.g., cytomegalovirus and EBV) or bacterial infections, or certain drugs and chemicals. Since some aspects in the pathophysiology of scleroderma resemble the changes occurring in graft-versus-host disease, the hypothesis was raised that fetal and maternal lymphocytes can cross the placenta barrier during pregnancy, thereby initiating a chronic inflammatory reaction leading to fibrosis [27, 28]. These studies will require further validation.
Microvascular damage as an early step in the pathophysiology
Clinical symptoms and also the histological investigation of early disease stages clearly indicate that microvascular damage is an important feature in the early phases of the disease [29, 30]. Although this is more pronounced in certain subsets (limited disease), similar alterations are found in all other forms (e.g., diffuse scleroderma). For clinical routine alterations of the nail fold, capillaries can be visualized by capillaroscopy (for review, see ). Vascular damage includes swelling and morphological alterations in endothelial cells as well as the presence of endothelial cell toxic factors in the circulation [32–34]. More recently, circulating autoantibodies targeting vascular receptors have been detected .
At the ultrastructural level, gaps between adjacent endothelial cells and vacuolization as well as duplications of the basement membrane have been reported . Hypoxia and reactive oxygen radicals contribute to endothelial cell damage . Endothelial cells were found to induce vascular cell adhesion molecule (VCAM)-1 expression, secrete a number of cytokines and chemokines, release potent factors such as the vasoconstrictor endothelin-1, and to participate in the remodeling of vascular structures . Progressive thickening of the vessel wall with duplications of the basement membrane results in a narrowing of the lumen of the capillaries and finally in the loss of the microvasculature, which is a specific feature detected by capillaroscopy. Interestingly, a recent report initially based on extensive genome-wide association studies describes elevated CXCL4 levels in the circulation of SSc patients. CXCL4 is a chemokine (also known as platelet factor 4) with anti-angiogenic activities secreted by plasmacytoid dendritic cells .
The inflammatory response
Histologically, endothelial cell damage is accompanied by the presence of perivascular lymphohistiocytic infiltrates [29, 32, 36] (Fig. 1d). These consist mainly of activated CD3- and CD4-positive mononuclear cells, which release proinflammatory and fibrogenic cytokines. Mast cells and eosinophils have also been discussed to contribute to the fibrotic reaction, however, recent studies show that the early work, based mainly on electron microscopy studies, has led to an overinterpretation and indicate that mast cells are dispensable for the development of fibrosis [37, 38]. Also eosinophils, which play an important role in combating parasite infections, are probably not directly responsible for the fibrotic reaction often arising subsequently to infection and severe tissue damage. More recently, the shift from Th1 to Th2 lymphocytes and a concomitant modulation of macrophage subsets has been highlighted (for review see ). This is supported by increased serum levels of cytokines released from these inflammatory cells. More recently, activation of macrophages was identified as a major factor for the development of SSc-related progressive lung fibrosis . By immunohistochemistry, B cells have further been detected in the inflammatory infiltrates, and an activated B cell signature was found by microarray analysis [40, 41]. Interestingly, depletion of B cells in a mouse model of scleroderma was reported to lead to reduced fibrosis , yet clinical studies in patients using rituximab led to conflicting results [43–49].
Cytokines and growth factors in systemic sclerosis
Th1 cells, memory T CD8+ cells, dendritic and NK cells
Th1 differentiation, activation of B cells
Th2 cytokine inhibition
Macrophages, APC, skin mast cells and keratinocytes
Pro-inflammatory and pro-apoptotic
Th1 cells, monocytes, macrophages, dendritic, and endothelial cells
Pro-inflammatory, production of IL-6 and PDGFα
Th1 (together with IL-1β, IL-23) and Th17 (together with IL-1β, IL-6, and IL-23) differentiation
Stimulates NK and T CD8+ cells
Lymphocyte proliferation and differentiation to Th1 cells in the presence of IL-12 and IFNγ as well as Th2 cells together with IL-4
Anti-inflammatory response associated with Th17 inhibition and Treg activation
Th2 cells, macrophages, mast cells, and NK cells
Th2 differentiation, Th1 cytokine inhibition as well as Treg differentiation
Fibroblast proliferation, chemotaxis, and collagen production
Production of TGFβ, CTGF, and TIMP-1
Th2 cells, as well as mast cells and eosinophils
B cell differentiation
Th2 cells, macrophages, and B lymphocytes as well as epithelial cells and fibroblasts
Th2 differentiation, Th17 differentiation (together with TGFβ and IL-21)
Inhibition of Treg differentiation
Stimulation of collagen production and inhibition of collagenase production
Alveolar macrophages, lung fibroblasts and skin fibroblasts
Stimulates fibroblast chemotaxis, chemoattractant and activator of neutrophils
Activated B cells and monocytes
Induces Th2 immune response and collagen synthesis
Th2 cells, NK and mast cells
B cell proliferation and differentiation, anti-inflammatory response and fibrosis
Monocytes, macrophages, dendritic cells as well as B lymphocytes
Th1 differentiation, inhibits Th17 differentiation
Th17 cells and NK cells
Pro-inflammatory and pro-fibrotic
Th17 cells and NK cells
Th17 response expansion
Mature Th17 cells and Th2 and NK cells
Keratinocytes, macrophages, fibroblasts, T and B cells, platelets and endothelial cells
Stimulates fibroblast proliferation and expression of TGFβ and PDGF receptors, production of CTGF and endothelin-1
Stimulates synthesis of collagen, fibronectin, proteoglycans, and TIMP (inhibition of ECM degradation)
Fibroblasts, endothelial cells, and smooth muscle cells
Induced by TGFβ, IL-4, and VEGF
Induces fibroblast proliferation and chemotaxis of fibroblasts, stimulates ECM production
Platelets, macrophages, fibroblasts, endothelial cells
Mitogen/chemoattractant for fibroblasts, synthesis of collagen, fibronectin, and proteoglycans
Stimulates secretion of TGFβ, MCP-1, and IL-6
Dose-dependent vasoconstrictor and dose dependent inducer of fibroblast proliferation and collagen synthesis
Autoantibodies, their diagnostic value, and contribution to pathophysiology
Autoantibodies with diagnostic relevance in systemic sclerosis and associated overlap syndromes
Frequent in patients with limited SSc
Associated with vasculopathy/digital ulcers, calcinosis cutis and upper GI involvement
Associated with diffuse SSc
Increased risk for lung fibrosis, renal crisis, severe heart involvement, and digital ulcers
Anti-RNA polymerase antibodies
Associated with diffuse SSc
Increased risk for renal crisis
Associated with diffuse SSc
Associated with skin involvement, vasculopathy, pulmonary hypertension and musculoskeletal involvement
Associated with limited skin involvement, interstitial lung disease and isolated pulmonary arterial hypertension
Associated with limited SSc, SSc myositis overlap and CK elevation
No severe internal organ manifestations
High titers are frequent in patients with MCTD
Associated with less skin involvement but increased risk for pulmonary hypertension
Associated with puffy fingers, RP, arthritis, and upper GI involvement
Main clinical associations are overlaps with SLE and myositis
Associated with SSc/myositis overlap syndromes
Associated with sicca syndromes
Secondary Sjoegren's syndrome
Autoantibodies with potential role in the pathogenesis of SSc
Stimulate expression of type I collagen and alpha SMA in fibroblasts through Ha-Ras-ERK1/2-ROS
Anti-endothelial cell antibodies
Seem to play an important role in vascular injury/damage associated with SSc
Acts against phospholipids and/or proteins that interact with phospholipids, includes lupus anticoagulant, anticardiolipin antibodies and anti-β2 GPI
May induce the expression of ICAM-1 and other pro-inflammatory cytokines/chemokines
Anti-extracellular MMP antibodies
Dysregulate ECM turnover and prevent degradation of ECM proteins
May release TGFβ by blocking latent TGFβ binding sites
This was demonstrated by Gabrielli et al. who detected antibodies directed against the PDGF receptor in several patients . Binding of these antibodies induced activation of PDGF signaling with subsequent activation of fibroblasts. Similarly, antibodies against endothelial cells [35, 57–59], against fibroblasts [60, 61], fibrillin , and matrix metalloproteinases (MMP)-1 and MMP-3 [63, 64] were found, which are also believed to carry biological activities. Unfortunately, some of the data are still contradictory [65–67] and the methods to identify these antibodies are by far more complex than those employed by routine diagnostic testing. Confirmation in large independent cohorts of patients is still missing. However, initially based on the findings by Baroni et al. several therapeutic approaches were initiated either to interfere with PDGF signaling or to inhibit the generation of autoantibodies by depleting B cells (for review, see ).
Recently, an interesting discussion was generated by the occurrence of an increased risk of cancer in a distinct group of scleroderma patients [69, 70]. Most of these patients were positive for autoantibodies directed against RNA polymerase III. Based on these observations, Joseph et al.  identified genetic alterations of the POLR3A locus in six of eight patients and suggested that the mutations would trigger cellular immunity. Although clearly extended studies are required these data have raised very interesting questions concerning autoimmunity and cancer in general .
Development of fibrosis
Following microvascular damage and the inflammatory response, the final step in the pathophysiology of scleroderma is the fibrotic reaction, which determines many of the clinical symptoms and the overall outcome of the disease.
Although fibrosis can occur in all involved organs, first symptoms are usually observed in the skin. Here, fibrosis starts in the deep dermis with involvement of the subcutaneous fat tissue . Reduction of the microvasculature [29, 73] and in later stages loss of appendages and of rete ridges is observed which leads to progressive disruption of the dermal architecture. This is due to an accumulation of ECM structures produced by activated fibroblastic cells and deposition in an irregular arrangement . Although it is obvious that de novo synthesis of ECM components is an important factor in the excessive deposition  there is accumulating evidence that disturbed degradation and remodeling also contribute to the fibrotic reaction . More recently, it has been recognized that an altered composition and macromolecular organization of the ECM control the availability and activation of growth factors and cytokines and also determine tissue stiffness [50, 77–79].
Although probably the majority of activated fibroblasts arises from cells residing in the tissue, a certain fraction can be recruited from the circulation as mesenchymal progenitor cells, which are attracted into the target tissue, where they proliferate and participate in the fibrotic process [84, 85]. Transdifferentiation from other cell types, e.g., from pericytes, mesenchymal progenitor cells, or epithelial cells may also contribute to the myofibroblast population [80, 81] (Fig. 2).
Interestingly, it was recognized many years ago that activated fibroblasts obtained from patients with SSc remain activated in culture following isolation from the tissue . With higher passage numbers, however, these cells tend to loose their characteristics, which was thought to be due to selection of distinct fibroblast populations [86, 87]. More recently, epigenetic modifications have been discussed to contribute to the activation state of these fibroblasts. This is based on the detection of DNA methylation modifications in scleroderma fibroblasts involving the collagen suppressor FLI-1 . These studies were further corroborated by analysis of myofibroblasts in fibrotic mouse kidneys, where hypermethylation led to the silencing of the Rasal-1 suppressor protein, which resulted in spontaneous fibroblast proliferation . Although these studies are only at the beginning, they indicate that epigenetic modifications may also result in persistent induction of collagen synthesis in fibroblasts from SSc patients.
Signaling pathways in scleroderma fibroblasts
Based on these findings, many attempts were initiated to specifically inhibit these signaling pathways, which worked well in cultured cells and also in mice with bleomycin induced fibrosis [93, 99], which is often used as an animal model for scleroderma [90, 100]. However, clinical studies based on these mechanisms have not yet led to a breakthrough. This is best illustrated by successful inhibition of PDGF signaling using imatinib to target activating antibodies against the PDGF receptor in vitro and in animal models; however, clinical studies have so far not been convincing. Similar disappointing results were obtained using inhibitors of TGFβ activity. In those studies, however, it remained unclear whether the compounds actually reached concentrations high enough to be sufficiently active in the vicinity of cells .
Therefore, it is assumed that the concept of direct fibroblast activation exclusively by TGFβ or other cytokines mentioned above does not provide the complete picture and is probably not sufficient to explain the phenotype of scleroderma.
Structural and nonstructural activities of the extracellular matrix
The ECM is a highly organized structure containing a large number of components, e.g., proteoglycans, fibulins, fibrillins, fibronectin, and collagens. They are arranged in a complex macromolecular organization, which is essential for the biological function of tissues (for review, see [79, 101–103]). The ECM determines the rigidity and stiffness of tissues but it also has important nonstructural activities since most ECM components directly bind to cells and thereby modulate their functions. In addition, they are important regulators of cytokine and growth factor activities [79, 104]. It is well established that collagens are the most abundant ECM proteins accumulating during fibrosis. Collagens occur in several subfamilies (review in [102, 105, 106]). The interstitial collagens represent the structural components of connective tissues; their macromolecular organization is strongly influenced by FACIT collagens (fibril-associated collagens with interrupted triple helices) as well as by other noncollagenous components of the ECM. It is important to note that the mechanical stability and dynamics of collagen fibrils is largely determined by the degree and the quality of crosslinks between molecules. Crosslink formation is executed intracellularly via hydroxylation of lysyl residues by lysyl hydroxylase and by lysyl oxidases, which catalyze the formation of intra- and intermolecular covalent crosslinks in the extracellular space. Both, lysyl hydroxylase  and lysyl oxidase  have been implicated in the development of fibrosis  and contribute to the altered tissue architecture.
Different ECM-embedded cell types make contact with ECM molecules via their cell surface receptors, of which the integrins are best characterized. Integrins are heterodimeric molecules composed of one α and one β chain with an extracellular, a transmembrane and a cytoplasmic domain. The extracellular domain specifically recognizes ECM molecules in the pericellular environment and binding induces signaling which modulates specific cellular activities such as proliferation, differentiation, and gene expression (for review see ). In fibroblasts, cell-matrix interactions modulate synthesis of ECM molecules and release of MMPs and their inhibitors. The balance between the latter two determines the turnover of ECM structures [76, 111].
Fibrillin and the pericellular environment
The pericellular matrix is composed of several noncollagenous ECM proteins, which direct cellular behavior. These proteins participate in interconnected networks mainly composed of fibronectin and elastin/ fibrillin together with several associated ligands such as fibulins, small proteoglycans, and LTBPs (latent TGF β binding proteins). There is also evidence that these networks closely interact with the collagenous scaffold and determine its macromolecular organization [114–118].
Fibrillin monomers are 350 kDa glycoproteins, which assemble into small diameter microfibrils. These macromolecular scaffolds form the core but are also found on the surface of elastic fibers and are therefore crucial for conferring tensile strength and elasticity to tissues. An involvement of fibrillin in skin fibrosis was first suggested when a large in-frame duplication in Fbn1 was identified in the Tight skin (Tsk) mouse . Tsk mice develop cutaneous fibrosis and serve as a disease model for scleroderma. It could be demonstrated that the expression of elongated, 418 kDa Tsk fibrillin-1 leads to microfibrils with compromised ultrastructure . Such structurally abnormal microfibrils were proposed to be functionally deficient, however, how this translates into disease mechanisms remains unclear. An important clue came from the finding that the rare inborn disorder Stiff skin syndrome is caused by mutations in the 8-cys domain of fibrillin-1 harboring the only integrin-binding site . Stiff skin microfibrils were found in large aggregates within which individual microfibrils appeared to be short. This suggests abnormal fibrillin ultrastructure in combination with impaired integrin binding as the underlying disease cause. Recent studies in mice demonstrated that point mutations in the integrin-binding domain of fibrillin-1 are sufficient to induce a Stiff skin syndrome-like phenotype. Interestingly, these mutations coincided with an activation of the immune system and with autoantibody production . However, skin fibrosis is also caused by several other fibrillin-1 mutations, which do not affect the integrin-binding site but interfere with other functions of fibrillin microfibrils such as the concentration of members of the ADAMTS/-like family [122, 123]. Together, these findings suggest that an ECM-defined fibroblast microenvironment serves as an active driver of fibrotic processes (Fig. 4). Thereby, interactions with the ECM via integrins and the bioavailability of ECM-bound TGFβ are very important. This concept was further extended by demonstrating that tissue stiffness and the transmission of mechanical tension not only contribute to an activated fibroblast phenotype but also induce TGFβ production and activation [124–128].
There is no doubt that considerable progress was made during the last decade towards a better understanding of the cellular and molecular basis of scleroderma. This was accomplished by combined efforts using large GWAS approaches, analyzing gene and protein expression patterns in affected tissues and also by conducting hypothesis-driven investigations. Generation of mouse models allowed the identification of the main players in different phases of the pathophysiology.
Several therapeutic approaches originating from the understanding of the molecular mechanisms have been developed. Many studies were based on the assumption that blocking a defined pathway could inhibit the fibrotic process. This was often supported by in vitro experiments and then also verified in mouse models. Unfortunately, in clinical studies, much less beneficial activity was experienced. This can be due to the fact that the currently available reagents are not sufficiently effective in fully blocking the activity of mediators (e.g., TGFβ).
However, in this context, we also have to consider that although the clinical differentiation of subsets of SSc has been improved, at the cellular and molecular level we might be confronted with different diseases that result in similar clinical features. We should also note that recent studies indicate a more globally disturbed regulation of cellular activities. This could result from the disturbed control of mediators with very general activities as it has been shown for TGFβ. However, we have to be aware that various cell types involved in the different pathophysiological events must be considered as part of an environment in which cell-cell and cell-ECM communication control cellular activities. Therefore, a detailed understanding of fibrotic reactions at the molecular and cellular level is crucial for the development of novel therapeutic tools. We need to unravel a global network, in which tissue stiffness, mechanical tension, and modulation of cytokine activity by the ECM and communication between mesenchymal, endothelial, and immunocompetent cells have to be considered as important players.
The authors thank all members of the Dermatology Department in Cologne and the SFB 829 for stimulating discussion. Many thanks go to Dr. Monique Aumailley for creative artwork. Our work is supported by Deutsche Forschungsgemeinschaft (KR558 to TK) and through SFB 829 (BE, TK, and GS), Deutsche Stiftung Sklerodermie (NH and PM), Edith Busch Foundation (NH) and by the Koeln Fortune Program/Faculty of Medicine, University of Cologne (PM).