Role of adipokines in systemic sclerosis pathogenesis

Abstract

Systemic sclerosis (SSc) is a chronic autoimmune connective tissue disease with manifestations in multiple organs, including the skin, lung, heart, joints, gastrointestinal tract, kidney, and liver. Its pathophysiology is characterized by inflammation, fibrosis, and vascular damage, with an increased expression of numerous cytokines, chemokines, and growth factors. However, besides these growth factors and cytokines, another group of molecules may be involved in the pathogenesis of SSc: the adipokines. Adipokines are proteins with metabolic and cytokine-like properties, which were originally found to be expressed by adipose tissue. However, their expression is not limited to this tissue, and they can also be found in other organs. Therefore, this review will describe the current knowledge regarding adipokines in the context of SSc and try to elucidate their potential role in the pathogenesis of SSc.

Introduction

Systemic sclerosis (SSc) (1) is a chronic progressive autoimmune connective tissue disorder affecting the skin as well as internal organs. Although the thick, hardened skin (scleroderma=hard skin) is one of its most noticeable symptoms, other symptoms can be very diverse owing to the multiorgan involvement, most commonly related to the gastrointestinal (GI) tract (2), lung, vascular system, and heart (1). This results in various manifestations, including myocardial disease, digestive problems, pulmonary arterial hypertension (PAH), and interstitial lung disease. Articular manifestations with joint/muscle pain or stiffness are also frequent, whereas kidney involvement (3) leading to renal dysfunction is relatively rare. There are two major subgroups of SSc: diffuse cutaneous (dc) and limited cutaneous (lc) SSc (4, 5). These two forms differ in the areas of the skin that are mainly affected, and their risk for developing internal organ damage is also different, which is higher for dcSSc patients compared with lcSSc patients. In addition, disease progression is generally faster in dcSSc patients and the symptoms are more severe. Approaches toward a more comprehensive subsetting of SSc have already been made (6).

SSc pathophysiology is characterized by inflammation, fibrosis, and vascular damage (1, 7). Specifically, inflammation and fibrosis are related to early and established SSc, respectively. This is mainly caused by the activated endothelial cells (EC) undergoing apoptosis, lymphocytes promoting inflammation, and highly proliferative fibroblasts producing increased amounts of extracellular matrix. In SSc, the fibroblasts or -to be more precise- the myofibroblasts are mainly responsible for the increased and sustained fibrosis (8). Myofibroblasts are transformed cells of fibroblastic lineage but are not restricted to originate from the tissue-resident fibroblasts (9). Pericytes, smooth muscle, and epithelial cells can be other sources of myofibroblasts. This is also reflected in the observation that myofibroblasts in SSc patients are a rather heterogeneous cell population with varying gene expression profiles (9). Activated B cells producing SSc-associated antibodies, such as antitopoisomerase (10), represent the main autoimmunity component of SSc.

In SSc, the production and secretion of numerous cytokines, chemokines, and growth factors is increased (7, 11, 12). Transforming growth factor β (TGF-β) is one of the major profibrotic factors in SSc (13), whereas proinflammatory cytokines include interleukin (IL)-6 and TNF-α (11). IL-17 is another important cytokine in SSc, which has an effect on both inflammation (14) and fibrosis (15, 16). However, besides these growth factors and cytokines, another group of molecules may be involved in SSc: the adipokines (17). Therefore, this review will address the current knowledge regarding adipokines in the context of SSc.

Adipose tissue and adipokines

Adipokines are adipose tissue-derived factors with metabolic and cytokine-like properties, hence their name (17). However, most adipokines are not only produced and secreted by adipose tissue but also by other tissues. Nonetheless, the discovery of these bioactive factors has put adipose tissue into a new light showing that it is not merely an immunologically inactive type of connective tissue but an important immunoendocrine organ with more functions than previously assumed (17–20). For example, with respect to the effector cells of rheumatoid arthritis (RA), proinflammatory and matrix degrading properties could be shown for adipokines, adiponectin, and visfatin (21–23). Adipokines also play a role in the cardiovascular system in which at least adiponectin appears to be rather protective (24), whereas, for example, chemerin can act as a chemoattractant and an angiogenic factor (25). The number of factors that are classified as adipokines have been growing continuously with an increasing number of publications regarding their potential as biomarkers or therapeutic targets in various rheumatic diseases including RA and osteoarthritis (26, 27). In this review, however, the focus will be on the potential role of adipokines in the pathogenesis of SSc based on clinical as well as cellular and molecular observations.

Clinical observations for adipokines in systemic sclerosis

Table 1 provides an overview of the studies in which the relationship between serum or plasma adipokine levels in SSc patients and controls has been analyzed. Table 2 shows the studies in which correlation analyses have been performed for the adipokine levels and clinical SSc-related parameters. The results are not always consistent among the studies. This is most likely because of the patient populations or the subgroups analyzed. Selection of these subgroups was based on gender, SSc type (dc or lc), disease activity, disease duration, renal dysfunction, or digital ulcers. It should be noted that even when the serum or plasma adipokine levels are comparable between SSc patients and healthy controls, adipokine expression might still be different locally within the tissue or in other body compartments.

Table 1.

Relationship between serum or plasma adipokine levels in systemic sclerosis patients in comparison with controls.

Adipokine	Change compared with healthy controls (unless indicated otherwise)	Sample type	References
Adiponectin	↓ in dcSSc patients	Serum	(28, 29, 31)
	↓ in dcSSc patients	Plasma	(33)
	↓ in SSc patients (meta-analysis)	Serum	(30, 34)
	↓ in dcSSc patients compared with lcSSc patients	Serum	(31)
	↑ in SSc patients (early stage)	Bronchoalveolar lavage	(51)
	no difference in SSc patients	Serum	(44)
	↓ in SSc patients with pericardial effusion compared with SSc patients without pericardial effusion	Serum	(47)
	↓ in SSc patients with active ILD compared with healthy controls	Serum	(83)
	↑ in SSc patients with active ILD after treatment with an intravenous pulse of cyclophosphamide compared with their initial levels	Serum
Adipsin	↑ in lcSSc patients	Serum	(38)
	↑ in early-stage SSc patients compared with mid-stage SSc patients	Serum	(36)
	no difference in SSc patients	Serum	(36)
Chemerin	↑ in SSc patients with digital ulcers compared with those without	Serum	(35)
	↓ in chemerin levels in dcSSc patients especially during the early phase (with levels increasing over time)	Serum	(84)
CTRP9	↑ in CTRP9 levels
	(but no difference for CTRP3)	Serum	(48)
Leptin	↓ in SSc patients	Serum	(45, 46)
	↓ in SSc patients with higher disease activity compared with those with lower disease activity	Serum	(43)
	↑ in idiopathic PAH patients with or without SSc	Serum	(52)
	no difference in SSc patients	Serum	(44)
	no difference in SSc patients (meta-analysis)	Serum	(30, 34)
Omentin	↓ in dcSSc patients compared with lcSSc patients (but no difference in dcSSc patients compared with healthy controls)	Serum	(49)
Resistin	↑ resistin levels in (female) SSc patients	Plasma	(42)
	↑ resistin levels in SSc patients with digital ulcers compared with those without	Serum	(44)
	no difference in SSc patients	Serum	(41, 44)
	no difference in SSc patients (meta-analysis)	Serum	(30)
Vaspin	↓ in SSc patients with digital ulcers compared with those without	Serum	(50)
Visfatin	↑ in visfatin levels in late-stage dcSSc patients (disease duration>6 years) when skin sclerosis regresses	Serum	(66)
	no difference in SSc patients	Serum	(66)
	↑ in SSc patients with pericardial effusion compared with those without	Serum	(47)

dcSSc: diffuse cutaneous systemic sclerosis; SSc: systemic sclerosis; lcSSc: limited cutaneous systemic sclerosis; ILD: interstitial lung disease; CTRP: C1q/tumor necrosis factor-related protein; PAH: pulmonary arterial hypertension.

Table 2.

Correlations/associations between serum or plasma adipokine levels in SSc patients with clinical SSc-related parameters.

Adipokine	Correlation or association	Sample type	References
Adiponectin	inverse correlation with SSc skin score (modified Rodnan skin score)	Serum	(29, 32)
	inverse correlation with changes in skin fibrosis over time	Serum	(29, 32)
	↓ in adiponectin levels in SSc patients associated with ↑ in total skin thickness score and ↑ in the incidence of pulmonary fibrosis	Serum	(28)
	positive correlation of adiponectin levels with vital capacity of the lung	Serum	(44)
	positive correlation with disease activity score (Valentini score)	Serum	(44)
Adipsin	associated with PAH	Serum	(38)
	no association with SSc skin score (modified Rodnan skin score) or pulmonary function parameters	Serum	(38)
	↓ in the frequency of anticentrome antibody positivity in dcSSc patients compared with lcSSc patients	Serum	(38)
Apelin	inverse correlation with SSc skin score (modified Rodnan skin score)	Serum	(40)
	↑ in the prevalence of intractable skin ulcers, scleroderma renal crisis, and PAH in patients with elevated serum apelin levels than in those without	Serum	(36)
Chemerin	inverse correlation with estimated glomerular filtration rate in SSc patients with renal dysfunction	Serum	(35)
Leptin	association with the duration of SSc symptoms	Serum	(44)
	positive correlation with the severity of lung fibrosis	Serum	(37)
	no association with skin score, activity score, and disease duration in SSc patients	Serum	(45)
Lipocalin-2	↑ in the prevalence of scleroderma renal crisis in SSc patients with elevated lipocalin-2 levels	Serum	(39)
	inverse correlation with estimated glomerular renal filtration rate in SSc patients with renal dysfunction	Serum	(39)
	positive correlation with SSc skin score in patients with dcSSc (+ disease duration of <3 years)	Serum	(39)
Omentin	no correlation with fibrotic or systemic inflammatory markers	Serum	(49)
Resistin	inverse correlation with estimated glomerular filtration rate in SSc patients with renal dysfunction	Serum	(41)
	positive correlation with ILD, arthralgia, esophageal involvement, and CRP levels	Plasma	(42)

SSc: systemic sclerosis; PAH: pulmonary arterial hypertension; dcSSc: diffuse cutaneous systemic sclerosis; lcSSc: limited cutaneous systemic sclerosis; ILD: interstitial lung disease; CRP: C-reactive protein.

Based on the majority of the clinical data, it can be speculated that decreased adiponectin levels contribute to skin fibrosis and PAH (28–34), which would imply a protective role of adiponectin in SSc. This view is supported by the results from some of the in vitro experiments and particularly animal models as described further. In contrast, clinical data (Table 1 and Table 2) suggest the other adipokines to be mainly detrimental, potentially promoting PAH and renal dysfunction (35–42). Although leptin serum levels are decreased in SSc patients with higher disease activity compared to SSc patients with lower disease activity (43), the associations of leptin levels with clinical SSc-related parameters (37, 44, 45) imply negative effects in SSc patients. Aside from this, differences in the leptin serum levels of SSc patients compared to healthy controls are still divisive (30, 34, 43–46). Results regarding cardiac effects of adiponectin in the context of SSc (47) as well as SSc-related results on CTRP9 (48), omentin (49) and vaspin (50) are still too scarce to draw any major conclusions.

Cellular expression of adipokines in SSc-related tissues and cells

As mentioned earlier, differences in the local distribution of adipokines within human tissues may not always be reflected by serum or plasma adipokine levels. Therefore, analyses of the local expression levels within different tissues may provide further cues on how adipokines affect certain tissues within SSc pathogenesis. Table 3 shows an overview of the expression of adipokines in SSc-related tissues and cells. The decreased serum or plasma adiponectin levels in SSc are reflected by their expression in the skin (28), lung, and gastric tissue (51), which may be another indication toward a potential antifibrotic effect of adiponectin in these tissues. In contrast, the increased expression of resistin and visfatin within the lymphocyte infiltrates in the lung tissue (51) points to a proinflammatory role of these adipokines within the lung, whereas the expression of chemerin (35), leptin (52), and lipocalin-2 (39) in the blood vessel cells might promote vascular inflammation.

Table 3.

Expression of adipokines in SSc-related tissues and cells.

Adipokine	Change compared with control	Tissue/cells	References
Adiponectin	↓ in adiponectin levels in dcSSc	Skin tissue	(28)
	↓ in the expression of adiponectin in SSc and IPF	Lung tissue	(51)
	↓ in the expression of adiponectin in SSc	Gastric tissue	(51)
	increased amounts of CD4+ T helper cells in adiponectin-positive tissue	Gastric tissue	(51)
Apelin	↓ in the expression of apelin	SSc dermal fibroblasts	(40)
Chemerin	↑ in the expression in small blood vessels of SSc
	↓ in the expression in dermal fibroblasts surrounded with thickened collagen bundles	SSc lesional skin	(35)
Leptin	↑ in the secretion by ECs from IPAH patients	ECs	(52)
Lipocalin-2	↑ in the expression in dermal fibroblasts and ECs	SSc lesional skin	(39)
Resistin	↑ in the expression of visfatin within lymphocyte infiltrates	Lung tissue	(51)
Visfatin	↑ in the expression of visfatin within lymphocyte infiltrates	Lung tissue	(51)

dcSSC: diffuse cutaneous systemic sclerosis; SSC: systemic sclerosis; IPF: idiopathic pulmonary fibrosis; EC: endothelial cells; IPAH: idiopathic pulmonary arterial hypertension.

Effects of adipokines on SSc-related cells

Dermal fibroblasts, vascular cells, and cells of the lung, GI tract, liver, and kidney are associated with the pathogenesis of SSc. Table 4 summarizes the results from experiments with SSc-related cells, which may provide hints to the involvement of these adipokines in SSc pathogenesis. In vitro results for adiponectin are inconsistent as far as its potential effect on fibrosis is concerned, (53–55) and a potential effect on SSc-associated fibrosis cannot be deducted from these results. Proliferation of pulmonary artery smooth muscle cells and liver cells (56, 57) as well as EC activation (58, 59) was inhibited by adiponectin. This suggests a protective effect of adiponectin in SSc, whereas based on the in vitro results, leptin (37, 53, 60) and resistin (61, 62) would promote the disease. The proinflammatory effects of visfatin have been shown for many cell types, including SSc dermal fibroblasts (63), synovial fibroblasts (23), and ECs (64, 65). However, one study showed a reversal of the profibrotic phenotype of SSc dermal fibroblasts with visfatin (66), which is clearly beneficial.

Table 4.

Effects of adipokines on SSc-related cells or factors and factors influencing adipokine expression.

Adipokine	Effect	Cell type	Type of effect	References
Adiponectin	↑ in the production of HA, HA synthase 2 and collagen	human dermal fibroblasts	Profibrotic	(53)
	↑ in the synthesis of extracellular matrix	human normal dermal fibroblasts	Profibrotic	(55)
	↓ in the production of collagen and α-smooth muscle actin both at basal level and after TFG-β or LPS induction	control and/or scleroderma dermal fibroblasts	Antifibrotic	(54)
	↑ in the expression of fibronectin 1, MMP1, MMP3, TIMP1, TIMP3	human normal dermal fibroblasts	Mixed	(55)
	↓ in proliferation	pulmonary artery smooth muscle cells from PAH rats	Antiproliferative	(57)
	↓ in proliferation/↑ in apoptosis
	↓ in migration	rat HSC	Antiproliferative/proapototic	(56)
	↓ in (cytokine-induced) EC activation	human ECs	Anti-inflammatory	(58, 59)
Apelin	depletion of apelin from fibroblasts significantly ⇨ ↑ in fibrosis-related gene expression	human dermal fibroblasts	Antifibrotic	(40)
	↓ in autophagosome formation	dermal fibroblasts	–	(40)
Chemerin	experimental ↓ in TGF-β ⇨ ↓ in chemerin expression	dermal fibroblasts from bleomycin-treated mice	–	(35)
CTRP3	↓ in production of TGF-β, CTGF and collagen I	human CLPF	Antifibrotic	(85)
	↓ in LPS-induced IL-8 secretion (but no effect on IL-6 and TNF-α release)		Anti-inflammatory	(85)
Leptin	↑ in the production of HA and HA synthase 2	human dermal fibroblasts	Profibrotic	(53)
	↑ in the expression of collagen I, TIMP-1, TGF-β1, CTGF, α-SMA by incubation with medium from leptin-stimulated Kupffer cells	rat HSC	Profibrotic (indirect)	(60)
	↑ in the expression of collagen I, α-SMA	A549 cell line (type II pulmonary epithelial cell model)	Profibrotic	(37)
	no direct activation by leptin	rat HSC	–	(60)
Resistin	↑ in EC activation	human ECs	Proinflammatory	(62)
	↑ in endothelial permeability (=potential promotion of vascular lesion formation)	human ECs	–	(61)
Visfatin	↑ in IL-6 secretion	SSc dermal fibroblasts	Proinflammatory	(63)
	reversal of profibrotic phenotype	SSc dermal fibroblasts	Antifibrotic	(66)
	↑ in EC activation	human ECs	Proinflammatory	(64, 65)

HA: hyaluronic acid; TFG-β: transforming growth factor-β; LPS: lipopolysaccharid; MMP: matrix metalloproteinase; TIMP: tissue inhibitor of metalloproteinase; PAH: pulmonary arterial hypertension; HSC: hepatic stellate cells; EC: endothelial cell; CTRP: C1q/tumor necrosis factor-related protein; CTGF: connective tissue growth factor; CLPF: colonic lamina propria fibroblasts; LPS: lipopolysaccharide; IL: interleukin; TNF-α: tumor necrosis factor α; α-SMA: α-smooth muscle actin; HSC: hepatic stellate cells.

Animal models

A number of animal models try to emulate the various manifestations of SSc (67); however, the results from the animal models providing insight specifically into the potential role of adipokines in the pathogenesis of SSc are still relatively scarce. With respect to liver fibrosis, animal models showed that leptin has profibrotic and adiponectin has antifibrotic effects (68). For example, adiponectin knockout mice displayed advanced liver fibrosis compared with wild-type mice when liver fibrosis was induced by carbon tetrachloride (CCl4) (56). Bleomycin-treated mice serve as a common model for SSc-like dermal fibrosis. In this model, bleomycin-induced dermal fibrosis could be inhibited by administration of the adipokine apelin (40). The potentially protective effect of adiponectin as suggested by the clinical data is supported by the results from animal studies. Interestingly, experimentally induced fibrosis in mice is associated with an intradermal white adipose tissue loss along with the conversion of adiponectin-positive intradermal progenitor cells into dermal myofibroblasts (69, 70). Adiponectin-deficient mice show defects in the lung as observed in SSc (71, 72). This includes symptoms of PAH at the pathophysiological (72) as well as the cellular level (71). Consistent with the findings for adiponectin deficiency, adiponectin overexpression in a mouse model of pulmonary hypertension attenuated the symptoms (73). Regarding skin manifestation, mice with a loss-of-function mutation of adiponectin showed an exaggerated dermal fibrotic response, whereas mice with constitutively elevated adiponectin levels were protected from skin fibrosis (74). In contrast, leptin is critically involved in the development of bleomycin-induced lung fibrosis because mice with defective leptin receptor signaling are resistant to this type of experimentally induced lung fibrosis (71).

Adipokine signaling and potential therapeutic approaches

For therapeutic approaches, adipokine-mediated effects are either to be inhibited or enhanced depending on their effects within the pathophysiology of SSc. Direct approaches would involve adipokine antagonists or agonists, whereas indirect approaches would involve inhibiting or enhancing the signaling pathways activated by the adipokines. An example for the latter would be nintedanib (75), a small molecular tyrosine-kinase inhibitor, targeting vascular endothelial growth factor receptor, fibroblast growth factor receptor, and platelet-derived growth factor receptor, which is currently used in the therapy of idiopathic pulmonary fibrosis as well as lung cancer and may find its way into SSc therapy (76). Adipokine signaling involves many different signaling molecules and pathways as can be seen in Table 5, which provides a list of those molecules and pathways that are involved in adipokine-mediated effects related to SSc pathogenesis. Some of these molecules or pathways might represent viable therapeutic targets for SSc treatment, particularly TGF-β signaling (77, 78) and the nuclear receptor peroxisome proliferator-activated receptor (PPAR) γ (79). TGF-β is a primary inducer of fibrosis and may be inhibited in its effect by various other molecules such as A20, a ubiquitin-editing enzyme, which again is increased in its expression in fibroblasts by adiponectin (80). Because of their antifibrotic effect, adiponectin and its signaling pathways lend themselves to be employed therapeutically. In this respect, adiponectin receptor agonists, i.e., molecules mimicking adiponectin and activating the adiponectin receptor(s), have already shown promising results in vitro and in animal models: ex vivo fibrotic responses in explanted SSc fibroblasts and in three-dimensional human skin equivalents were abolished (74), adipocyte-to-myofibroblast transdifferentiation was decreased, and the profibrotic phenotype of SSc fibroblasts was reversed (81). In mice, experimentally induced organ fibrosis could be prevented and reversed (74, 81). This clearly demonstrates the potential of adiponectin agonists to be used in the SSc therapy. For leptin, on the contrary, the signaling pathway analyses suggest a profibrotic effect (37, 60, 82). Hence, leptin might represent a therapeutic target for an antagonistic approach for inhibiting fibrosis in SSc.

Table 5.

Signaling molecules and pathways involved in adipokine-mediated effects related to SSc pathogenesis.

Adipokine	Signaling molecules and pathways involved	Potentially mediated effect	References
Adiponectin	AMPK/PPAR-γ	Antifibrotic (in skin)	(29, 54)
	iNOS/nitric oxide/adipoR2/AMPK/JNK/Erk1/2/NF-κB	Antiproliferative/proapopotic (in liver)	(56)
	adipoR1/NF-κB/COX-2/AMPK/nitric oxide/p38 MAPK	Anti- or proinflammatory (on blood vessels) depending on adiponectin isoform	(58)
	A20 (↑ by adiponectin)	Antifibrotic	(80)
Apelin	TGF-β signaling	Antifibrotic (in skin)	(40)
Leptin	PI3K/Akt/mTOR pathway	Profibrotic (in lung)	(37)
	STAT3/Akt/Erk1/2/AP-1/NF-kB	Profibrotic (in liver)	(60)
	PPAR-γ/Erk1/2	Profibrotic (in liver)	(82)
Resistin	STAT3/SOCS3	Proinflammatory (promoting vascular inflammation)	(62)
	p38 MAPK	↑ endothelial permeability	(61)
Visfatin	NF-κB	Proinflammatory (promoting vascular inflammation)	(65)

AMPK: AMP-activated protein kinase; PPAR-β: peroxisome proliferator-activated receptor β; iNOS: inducible nitric oxide synthase; JNK: c-Jun N-terminal kinase(s); Erk1/2: extracellular signal-regulated kinase 1/2; NF-κB: nuclear factor κ-light-chain-enhancer of activated B cells; adipoR1/2: adiponectin receptor 1/2; COX-2: cyclooxygenase-2; MAPK: mitogen-activated protein kinase; TFG-β: transforming growth factor-β; PI3K: phosphatidylinositol-3-kinase; Akt: protein kinase B; mTOR: mechanistic Target of Rapamycin; STAT3: signal transducer and activator of transcription 3; AP-1: activator protein 1; SOCS3: suppressor of cytokine signaling 3.

Conclusion

Inflammation, fibrosis, and vascular damage are the major pathophysiological hallmarks of SSc, which lead to its various manifestations and are targets of the therapeutic approach for this disease. Clinical data as well as data from in vitro and animal experiments suggest that adipokines are involved in these processes. The data also suggest that some adipokines might contribute to the pathogenesis of SSc (e.g., leptin, resistin, and visfatin), whereas others might have a protective effect against the disease (e.g., adiponectin and apelin). Hence, therapeutic targeting would consist of different possible approaches: a) for disease-promoting adipokines: direct neutralization of the adipokines or inhibition of the adipokine-induced signaling pathways, and b) for disease-protective adipokines: development of adipokine mimetics. However, because adipokines are highly multifunctional proteins within the human body, their use as therapeutic targets may be challenging and not without problems. Therefore, further studies, particularly in animal models of SSc, should be performed to elucidate the role of adipokines in SSc and their applicability as therapeutic targets.

Main Points.

• Adipokine expression is not limited to adipose tissue but is also found in other organs including those affected by systemic sclerosis.

• Adipokines, proteins with metabolic and cytokine-like properties, may be involved in the pathogenesis of systemic sclerosis, with data suggesting that some have disease-promoting and some have disease-protective effects.

• Due to their highly multifunctional nature within the human body, the use of adipokines as therapeutic targets represents a very challenging task.