Syndecans in heart fibrosis
- First Online:
- Cite this article as:
- Lunde, I.G., Herum, K.M., Carlson, C.C. et al. Cell Tissue Res (2016) 365: 539. doi:10.1007/s00441-016-2454-2
- 524 Downloads
Heart disease is a deadly syndrome affecting millions worldwide. It reflects an unmet clinical need, and the disease mechanisms are poorly understood. Cardiac fibrosis is central to heart disease. The four-membered family of transmembrane proteoglycans, syndecan-1 to -4, is believed to regulate fibrosis. We review the current literature concerning syndecans in cardiac fibrosis. Syndecan expression is up-regulated in response to pro-inflammatory stimuli in various forms of heart disease with fibrosis. Mice lacking syndecan-1 and −4 show reduced activation of pro-fibrotic signaling and increased cardiac rupture upon infarction indicating an important role for these molecules. Whereas the short cytoplasmic tail of syndecans regulates signaling, their extracellular part, substituted with heparan sulfate glycosaminoglycan chains, binds a plethora of extracellular matrix (ECM) molecules involved in fibrosis, e.g., collagens, growth factors, cytokines, and immune cell adhesion proteins. Full-length syndecans induce pro-fibrotic signaling, increasing the expression of collagens, myofibroblast differentiation factors, ECM enzymes, growth factors, and immune cell adhesion molecules, thereby also increasing cardiac stiffness and preventing cardiac rupture. Upon pro-inflammatory stimuli, syndecan ectodomains are enzymatically released from heart cells (syndecan shedding). Shed ectodomains affect the expression of ECM molecules, promoting ECM degradation and cardiac rupture upon myocardial infarction. Blood levels of shed syndecan-1 and −4 ectodomains are associated with hospitalization, mortality, and heart remodeling in patients with heart failure. Improved understanding of syndecans and their modifying enzymes in cardiac fibrosis might contribute to the development of compounds with therapeutic potential, and enzymatically shed syndecan ectodomains might constitute a future prognostic tool for heart diseases with fibrosis.
KeywordsExtracellular matrix Wound healing Myofibroblast Inflammation Infarction
Heart disease is common, costly, and deadly
Heart disease is a leading cause of morbidity and mortality worldwide (Bui et al. 2011; Ho et al. 1993; Lloyd-Jones et al. 2010). Several forms of heart disease, including myocardial infarction, valvular heart disease, hypertensive heart disease, myocarditis and cardiomyopathies, can lead to heart failure. Heart failure is a syndrome in which the heart is unable to pump blood at a rate that meets the requirements of the metabolizing tissues (Braunwald and Bristow 2000). Because of an ageing population and improved treatment of acute cardiac events, the number of patients with heart failure has increased rapidly in the Western world and has reached epidemic proportions (Bui et al. 2011; McCullough et al. 2002). Accordingly, its socio-economic impact is alarming, and heart failure comprises the most costly illness with health care costs twice as high as those for cancer (O’Connell and Bristow 1994). Today, heart failure is the most common reason for adults to be admitted to the hospital in the US (Bui et al. 2011; Levy et al. 2002; Lloyd-Jones et al. 2010). A growth in heart failure cases is expected in developing countries in the years to come, because of a shift from acute to chronic heart disease and the increased pervasiveness of risk factors (Yusuf et al. 2001). Despite advances in patient management and therapy, heart failure remains a deadly syndrome with a 5-year mortality of 50 % (Bui et al. 2011; Levy et al. 2002). These statistics reflect that important pathophysiological mechanisms, such as fibrosis, are not sufficiently modified by current treatment (Kaye and Krum 2007; Mudd and Kass 2008). An improved understanding of the molecular mechanisms governing cardiac fibrosis will most likely contribute to the development of new therapeutics with a reduction in suffering, disability, and death attributable to heart disease.
Fibrosis is an important pathophysiological process in diseased heart
Cardiac fibrosis, namely the formation and accumulation of extracellular matrix (ECM) connective tissue in a reactive or reparative process, is one of the main and most common responses of the heart to pathological stimuli (Mann et al. 2014). Importantly, the fibrotic response and the effects of fibrosis on cardiac function are dependent on the etiology or form of the disease.
In the early phase following coronary occlusion and myocardial infarction, a highly regulated process of cardiac repair follows the necrotic loss of cardiomyocytes. This process is characterized by the activation of ECM-degrading enzymes, e.g., matrix metalloproteinases (MMPs), and invasion of inflammatory cells drawn by immune cell adhesion molecules and cytokines peaking at the first and second weeks after injury in humans (Mann et al. 2014). Increased synthesis of fibrillar types I and type III collagens follows in the infarcted area, with scar tissue being found in regions in which cardiomyocytes have been lost.
The main cells responsible for the production of collagens in the diseased heart are cardiac myofibroblasts. Myofibroblasts are differentiated cardiac fibroblasts that are characterized by a contractile phenotype with the expression of α-smooth muscle actin (α-SMA) and excessive ECM production (Darby et al. 1990; Hinz et al. 2007; Porter and Turner 2009). Although the steps in the fibroblast-to-myofibroblast differentiation process are not well understood, an increase in focal adhesion proteins, including integrins and syndecan proteoglycans, has been shown (Goffin et al. 2006; Woods et al. 2000). Mechanical stress and cytokines, e.g., transforming growth factor (TGF) β, are major signals responsible for differentiation (Dobaczewski et al. 2011; Petrov et al. 2002).
The collagen fibers produced by myofibroblasts after a myocardial infarction form scar tissue. In addition to the fibrotic scar in the infarcted region, fibrosis and the growth of cardiomyocytes occur in the non-infarcted regions causing alterations in the shape and size of the heart, a process termed cardiac remodeling (Mann et al. 2014). Fibrosis in the non-infarcted region develops around intramural coronary arteries (perivascular fibrosis) or in the interstitial space (interstitial fibrosis) between myocytes. Often, in the late phase of the remodeling, progressive cardiomyocyte death causes scarring (replacement or focal fibrosis). Clinically, we consider it important to note that, in the early phase after myocardial infarction, ventricular rupture or an aneurysm may occur, with death as a consequence. In the late phase, failure of the remodeling process to normalize wall stress results in progressive dilation, deterioration of cardiac function, and heart failure with systolic dysfunction. In these settings, fibrosis is important to counteract rupture, aneurysm, or dilation.
Pressure overload of the ventricles caused by valvular disease or hypertension leads to the remodeling of the heart with thickening of the ventricular walls (concentric hypertrophy; Mann et al. 2014). Hypertrophy is caused by the growth of cardiomyocytes and interstitial fibrosis. Hypertrophy of cardiomyocytes and increased interstitial fibrosis cause an impairment of cardiac filling because of the increased stiffness of the myocardium and reduced chamber volume. In the late phase, cardiomyocyte cell death and replacement fibrosis might develop, which further reduces diastolic and systolic cardiac function.
Not only pressure overload leads to concentric hypertrophic remodeling with fibrosis and diastolic dysfunction. Mutations in genes encoding the main contractile proteins in cardiomyocytes, e.g., myosin heavy chain (MHC7) and myosin-binding protein (MYBPC3), cause familial hypertrophic cardiomyopathy (HCM), a relatively common genetic cardiac disease with a prevalence of 1:500 individuals in the general population worldwide (Mann et al. 2014; Seidman and Seidman 2011). HCM is the leading cause of sudden cardiac death in young people. It is a primary disease of the cardiomyocytes, which often are hyper-contractile with supra-normal systolic function. This milieu activates pro-fibrotic pathways such as TGFβ, and both interstitial and focal myocardial fibrosis is present (Ho et al. 2010). Fibrosis is believed to be the arrhythmogenic substrate in HCM, causing sudden cardiac death (Almaas et al. 2013).
Myocarditis or inflammatory cardiomyopathy is an acute inflammatory disease of the heart often characterized by a cellular infiltrate (Mann et al. 2014). Immune cells such as T-cells are known to increase pro-fibrotic processes in the heart and in other tissues, linking cardiac inflammation to fibrosis and dysfunction (Hartupee and Mann 2016; Ramos et al. 2016; Wei 2011; Yu et al. 2010). Myocarditis causes chest pain, heart failure, or sudden death, most often attributable to an infection by common viruses. If the inflammation of the myocardium does not resolve during the acute stage, loss of cardiomyocytes with replacement fibrosis might occur, causing a reduction in systolic function; it might also progress into overt dilation with severe heart failure.
Heart failure is frequently associated with reduced systolic function, most commonly characterized by reduced ejection fraction (HFrEF). However, heart failure with preserved ejection fraction (HFpEF) is also common (Mann et al. 2014) and has been referred to as diastolic heart failure, implying that the filling of the heart is impaired. HFpEF is associated with aortic stenosis, diabetes, hypertension, age, obesity, and female gender. Both an increase in the amount of collagen and post-translational modification of collagens through cross-linking have been shown to increase diastolic stiffness in these patients (Kasner et al. 2011; López et al. 2012).
In general, adequate timing and the quantity and quality of fibrosis are absolutely crucial for heart function and survival by determining cardiac stiffness, contractility, compliance, probability of rupture, dilation, and diastolic and systolic function. Molecular mechanisms orchestrate pro- and anti-fibrotic signaling pathways and cellular processes, including inflammation, fibroblast-to-myofibroblast differentiation, myofibroblast function, deposition and degradation of ECM molecules (Barry and Townsend 2010; Kehat and Molkentin 2010).
Cardiac extracellular matrix and fibrosis: role of proteoglycans
Two main families of cell surface proteoglycans are found in mammals: the four syndecans (syndecan-1 to -4) and the six glypicans (glypican-1 to -6). These constitute the main sources of cell surface GAGs (Bernfield et al. 1992; Filmus et al. 2008). Syndecans are transmembrane, in contrast to glypicans, which are attached to the extracellular part of the membrane by a glycosylphosphatidylinositol (GPI) anchor. In addition to the cell surface proteoglycans, ECM-localized proteoglycans such as the small leucine-rich proteoglycans (SLRPs) are found within the cardiac ECM (Engebretsen et al. 2013; Iozzo and Schaefer 2010).
The role of cardiac proteoglycans in fibrosis and heart failure is not well understood. The unstable and altered microenvironment in the heart during disease results in a different make-up of syndecans, affecting the levels of, processing of, and signaling related to these transmembrane proteoglycans. Syndecan expression has been shown in several publications to be upregulated in the heart in response to pathological stimuli and to be associated with important aspects of fibrosis. The role of syndecans in the development of cardiac fibrosis is addressed in the current review.
Syndecans in heart
The transmembrane proteoglycans, including the syndecans, were discovered 30 years ago (Kjellén et al. 1981), and all animal cells express at least one of the fewer than ten known transmembrane proteoglycans (Bernfield et al. 1992; Couchman 2010). Invertebrates express one syndecan, reflecting a long evolutionary history, and two rounds of gene duplication produced the four vertebrate isoforms. Syndecans are highly conserved among species, and certain domains are conserved among syndecan isoforms. Mice lacking syndecans are viable and fertile and have a normal heart function and phenotype; however, phenotypes develop when they are subjected to stress or injury. Syndecans have been implicated in numerous cellular functions relevant to the development of cardiac fibrosis, e.g., cellular adhesion, differentiation, signaling, migration, growth, metabolism, ECM production, and recently, calcium homeostasis (Gopal et al. 2015).
Increased expression of all four syndecans in the hearts of mice upon myocardial infarction was detected by our group in 2004 in a filter array study (Finsen et al. 2004), linking the syndecans to heart disease. Since then, we have shown that the four syndecan family members are also up-regulated upon pressure overload in mice (Strand et al. 2013), suggesting they all play a role across the various forms of heart disease. Of the four isoforms, mainly syndecan-1 and −4 have been studied in the heart and will be discussed, with particular focus on their effects and roles in cardiac fibrosis.
Role of syndecan-1 in myocardial infarction and fibrosis of the heart
Several studies suggest that syndecan-1 expression is induced by tissue injury and regulates inflammatory and reparative responses. Syndecan-1 has emerged as a pro-fibrotic molecule in the lungs (Kliment et al. 2009) and the heart (Frangogiannis 2010). Syndecan-1 is up-regulated in the non-infarcted hypertrophic region (Finsen et al. 2004), in the border zone (Lei et al. 2012), and in the infarct region (Li et al. 1997; Vanhoutte et al. 2007) of hearts of mice and rats subjected to myocardial infarction. Syndecan-1 is also rapidly upregulated in hearts of mice after pressure-overload by aortic banding, i.e., as early as 24 hrs after overload (Strand et al. 2013). The role of syndecan-1 in heart disease has been investigated by using syndecan-1 knock-out mice subjected to myocardial infarction. Syndecan-1 is thought to play an important role for infarct healing and for counteracting cardiac rupture, dilation, and systolic heart failure (Vanhoutte et al. 2007). Syndecan-1 knock-out mice show functionally adverse infarct healing via the attenuated formation of collagen matrix and altered recruitment of immune cells to the infarct area. Importantly, the lack of syndecan-1 results in collagen disorganization and fragmentation, the latter being related to the increased activity of MMP-2 and −9. In accordance with the results from syndecan-1 knock-out mice, adenoviral overexpression of syndecan-1 in vivo improves collagen matrix quality and protects against cardiac dilation and failure.
Pro-fibrotic syndecan-1 signaling in cardiac fibroblasts
Role of syndecan-4 in myocardial infarction and fibrosis of the heart
Syndecan-4 is found at equal expression levels in cardiac myocytes and fibroblasts (Herum et al. 2013; Strand et al. 2013). Syndecan-4 and syndecan-1 are rapidly up-regulated in both the non-infarcted hypertrophic region (Finsen et al. 2004) and in the infarcted region (Li et al. 1997; Matsui et al. 2011) of the heart in response to myocardial infarction in a rat diabetes model of diastolic dysfunction (Strunz et al. 2011) and in pressure-overloaded and failing human hearts (Finsen et al. 2011; Strand et al. 2013) and mouse hearts (Finsen et al. 2011; Herum et al. 2013; Strand et al. 2013).
Studies manipulating syndecan-4 expression clearly demonstrate the importance of syndecan-4 for fibrosis in the heart. Syndecan-4 knock-out mice show increased cardiac rupture and death, hampered granulation tissue formation (i.e., inflammation), attenuated myofibroblast differentiation, and reduced contractile function after myocardial infarction (Matsui et al. 2011). Accordingly, when syndecan-4 is overexpressed in the myocardium by using a viral approach, mice subjected to myocardial infarction show improved survival and function (Xie et al. 2012). Following infarction, syndecan-4 is thought to be important for granulation tissue formation (Matsui et al. 2011) and cardiac remodeling (Echtermeyer et al. 2011).
When syndecan-4 knock-out mice are subjected to mechanical stress by increasing left ventricular pressure, they do not develop the expected concentric myocardial hypertrophy seen in wild-type mice, but instead exhibit dilation and exacerbated heart failure (Finsen et al. 2011; Herum et al. 2013). Importantly, myocardial stiffness is reduced when syndecan-4 is lacking because of attenuated collagen maturation, i.e., collagen cross-linking (Herum et al. 2015). We have found a reduced number of myofibroblasts in myocardial tissue at an early stage after induction of pressure overload in syndecan-4 knock-out mice, providing an explanation for the observed reduced stiffness. This reduction in myofibroblast differentiation and collagen cross-linking when syndecan-4 is not present is likely related to attenuated T-lymphocyte infiltration, indicated by reduced expression levels of T-cell surface markers (e.g., CD3, CD4 and CD8) in syndecan-4 knock-out hearts after pressure overload (Strand et al. 2013). T-cells are known to regulate fibroblast-to-myofibroblast differentiation and the production of lysyl oxidase (LOX) collagen cross-linking enzymes (Hartupee and Mann 2016; Yu et al. 2010). At later and more advanced stages of pressure-overload-induced remodeling, no difference in α-SMA or collagen I and III mRNA has been found between syndecan-4 knock-out mice and controls, indicating that syndecan-4-independent mechanisms regulating fibrosis (e.g., integrins and TGFβ) take over. Thus, upon pressure overload, ECM remodeling and fibrosis are dysregulated when syndecan-4 is lacking, and a role for syndecan-4 in fibrotic pathways is indicated during the response to mechanical stress.
Pro-fibrotic syndecan-4 signaling in cardiac fibroblasts
Syndecan-4 is expressed in focal adhesions of many cell types, including fibroblasts. These structures are important for signaling across the membrane and for physically anchoring the ECM to the cytoskeleton. Thus, syndecan-4 has been proposed to be part of a cellular mechanical stress-sensing apparatus (Bellin et al. 2009; Finsen et al. 2011; Herum et al. 2013; Samarel 2013).
The major part of syndecans resides in the ECM. However, syndecans are involved in cellular signaling events through their short cytoplasmic tail. The cytoplasmic domain is highly conserved from zebrafish to humans (Whiteford et al. 2008). It is divided into three regions: two that are highly conserved among syndecans and across species (C1 and C2) and a central variable region (V), which is also conserved across species and which is unique for each syndecan and thus expected to perform isoform-specific interactions.
Syndecans interact directly with signaling molecules, and syndecan-4-mediated signaling events regulate cardiac fibrosis. Although not studied in cardiac fibroblasts, syndecan-4 in non-cardiac fibroblasts regulates the localization, activity, and stability of protein kinase C alpha (PKCα) through a direct binding to the V-region (Keum et al. 2004; Oh et al. 1997). This direct interaction regulates downstream signals for the formation of focal adhesions and actin stress fibers, important aspects of fibroblast function. The regulation of the Rho GTPases (Dovas et al. 2006), molecules with known effects on cardiac fibrosis (Hartmann et al. 2015), occur downstream of syndecan-4-dependent PKCα-activation.
Interestingly, the syndecan-4 V-region contains a PIxIxIT-similar motif, a specific motif found in calcineurin-interacting proteins such as nuclear factor of activated T-cells (NFAT; Aramburu et al. 1998). Calcineurin-NFAT signaling is well known as one of the main pro-hypertrophic and pro-fibrotic pathways in the heart, playing a central role in cardiac remodeling, dysfunction, and failure (Chen et al. 2012; Davis et al. 2012; Molkentin 2004, 2013; Wang et al. 2016). Calcineurin binds directly to the V-region of syndecan-4 (Finsen et al. 2011). The syndecan-4 binding site has been located in the auto-inhibitory domain (AID) of calcineurin, suggesting that the phosphatase is activated upon binding. Calcineurin dephosphorylates NFAT transcription factors when active, and accordingly, syndecan-4 activates downstream NFAT signaling in cardiac fibroblasts in response to mechanical stress (Fig. 4; Echtermeyer et al. 2011; Finsen et al. 2011; Herum et al. 2013). Thus, calcineurin-NFAT signaling represents a syndecan-4-associated molecular mechanism likely to underlie the observed reduced myofibroblast numbers, collagen production, and myocardial stiffness (Herum et al. 2013) in syndecan-4 knock-out mice in response to pressure overload or myocardial infarction.
Phosphorylation of serine 179 (pS179) in C1 of syndecan-4 has been shown to constitute a molecular switch regulating the conformation of the cytoplasmic tail, affecting PKCα binding and activity, cell mobility, and growth (Horowitz and Simons 1998; Koo et al. 2006). pS179 inhibits the interaction with calcineurin, and levels of pS179 are reduced in pressure-overloaded murine and human hearts, favoring the binding and activation of calcineurin by syndecan-4 in these forms of heart disease (Finsen et al. 2011).
Recent evidence suggests that syndecan-4 regulates pro-fibrotic responses in fibroblasts by inhibiting calcium influx through the transient receptor potential canonical (TRPC) 7 cell membrane channel. This is accomplished through phosphorylation of TRPC7 by PKCα in a syndecan-4-dependent manner, and thus, fibroblasts lacking syndecan-4 have increased cytosolic calcium levels accompanied by hampered myofibroblast phenotypic features such as the presence of osteoblast-cadherin in adherence junctions and the lack of α-SMA in microfilament bundles (Gopal et al. 2015). Although the regulation of TRPC channels by syndecan-4 has not been demonstrated in the heart, important roles for TRPC channels in cardiac fibroblast activation have been shown, i.e., calcium influx through TRPC6 is necessary for TGFβ-induced myofibroblast differentiation (Davis et al. 2012). Thus, these mechanisms might also be relevant for syndecan-4-regulated fibrosis in the heart.
Activation of NFAT increases the expression of the transcription factor MRTF-A. Being an essential co-factor for the transcription factor serum response factor (SRF), MRTF-A has also recently been found to be necessary for myofibroblast differentiation (Velasquez et al. 2013). The collagen 1a2 promoter has been identified as a target of MRTF-A/SRF, and MRTF-A knock-out mice have diminished fibrosis and scar formation following myocardial infarction or angiotensin II treatment (Small et al. 2010). Interestingly, MRTF-A activity is known to be regulated by actin polymerization, which is involved in its nuclear translocation (Mack et al. 2001; Velasquez et al. 2013). Since syndecan-4 has been found to regulate both the transcription of MRTF-A through NFAT (Herum et al. 2013) and the polymerization of actin (Matsui et al. 2011) by RhoA in cardiac fibroblasts, this signaling pathway is probably also affected by syndecan-4 and therefore might contribute to the impaired myofibroblast differentiation of cardiac fibroblasts lacking syndecan-4.
The heparan sulfate GAG chains of syndecans can interact with numerous ligands including growth factors, and syndecans in this respect are thought to act as co-receptors (Bernfield et al. 1999). Although not studied in the heart, the best-known example is the growth factor interaction of syndecan-4 with fibroblast growth factor 2 (FGF2; Rapraeger et al. 1991; Yayon et al. 1991; Zhang et al. 2003). Since FGF2 is released in the heart during stress (Clarke et al. 1995), and since FGFs are believed to promote cardiac fibrosis by activating mitogen-activated protein kinase (MAPK) signaling (Itoh and Ohta 2013), syndecan-4 probably modulates FGF signaling and thereby might affect cardiac fibrosis.
Syndecan shedding in cardiac fibrosis
Syndecans carry mainly heparan, but also some chondroitin sulfate GAGs on their extracellular domain (Bernfield et al. 1992; Bishop et al. 2007). GAGs are unbranched linear polysaccharides consisting of repeated disaccharide units of various lengths that are attached to specific serine residues in the proteoglycan core proteins. GAGs carry a highly negative charge allowing their interaction with, binding, and retention of a variety of positively charged ECM molecules, and the GAG type on proteoglycans dictates much of their function (Rienks et al. 2014). The importance of heparan sulfate GAGs and their electrostatic interaction with ligands has been clearly demonstrated in development, normal physiology, and pathology of all animal cells and organ systems (Bishop et al. 2007).
An important part of syndecan function is related to the enzymatic release of the GAG-substituted extracellular domain into body fluids and the ECM. In the extracellular milieu, syndecans associate with proteases that cleave off the ectodomain in a process termed shedding. Increased shedding has been linked to wound repair, inflammation, tumor progression, and oxidative stress in non-cardiac cell types, emphasizing that syndecan ectodomain shedding represents an important aspect of syndecan function (Couchman 2010; Lambaerts et al. 2009; Manon-Jensen et al. 2010).
Role of syndecan-4 shedding fragments in regulation of cardiac fibrosis
We have shown that shed syndecan-4 fragments are found within the myocardium of human patients with end-stage dilated heart failure (Strand et al. 2013). Similar to syndecan-4 knock-out, viral overexpression of the extracellular domain of syndecan-4 in the heart in vivo has deleterious effects in mice upon myocardial infarction, impairing cardiac function and increasing rupture and mortality (Matsui et al. 2011). These data suggest that shed fragments of syndecan-4 have a net negative impact on infarct healing, resembling the effects of the lack of syndecan-4. Moreover, these data indicate that the overexpression of full-length syndecan-4 vs. the shed ectodomain have opposite effects on infarct healing.
As discussed above, full-length syndecan-4 has pro-fibrotic properties that include the up-regulation of collagen I and III expression in cardiac fibroblasts. Recently, we have reported that cardiac fibroblasts treated with shed syndecan-4 fragments show reduced collagen I and III expression, reduced proliferation, and an increased ratio of MMP-to-tissue inhibitor of MMP (TIMP) expression, favoring matrix degradation (Strand et al. 2015). Thus, the shed ectodomain has an opposite effect to the full-length syndecan-4 with regards to collagen expression.
Cardiac myocytes and fibroblasts show increased expression of important immune cell adhesion molecules, i.e., intercellular adhesion molecule 1 (ICAM-1) and vascular cell adhesion molecule 1 (VCAM-1), together with the activation of NFkB signaling, when treated with shed syndecan-4 fragments (Strand et al. 2015). These data suggest that shed syndecan-4 ectodomains induce pro-inflammatory signaling cascades in the heart, including pathways promoting immune cell infiltration. In accordance with these in vitro findings, reduced immune cell infiltration is indicated in hearts from syndecan-4 knock-out mice after pressure overload, myocardial infarction, and LPS by reduced mRNA expression of specific immune cell surface markers, i.e., T-cell markers CD3, CD4, and CD8 and the macrophage marker F4/80 (Matsui et al. 2011; Strand et al. 2013, 2015). Heparan sulfate is well-known to participate in almost every stage of leukocyte transmigration through the blood-vessel wall and is involved in the adhesion of leukocytes to the inflamed endothelium through a direct binding of the lymphocyte L-selectin adhesion protein (Götte 2003; Parish 2006). For instance, the lack of heparan sulfate in endothelial cells results in an impaired influx of neutrophils into tissues in several inflammatory models (Wang et al. 2005). Thus, we suggest that shed syndecan-4 ectodomains influence immune cell infiltration through two mechanisms: (1) by up-regulating adhesion molecules ICAM-1 and VCAM-1 on cardiac myocytes and fibroblasts and (2) by a direct binding of immune cell adhesion molecules to heparan sulfate. Whether the attenuated influx of immune cell to syndecan-4 knock-out hearts, as indicated by the reduced expression of immune cell surface markers, is a result of the inability to shed syndecan-4 or the lack of the full-length protein, or both, has not been clarified. However, given the effects of immune cells on fibroblast-to-myofibroblast differentiation, collagen, and LOX production (Hartupee and Mann 2016; Ramos et al. 2016; Wei 2011; Yu et al. 2010), syndecan-4 shedding could influence fibrosis through its effects on immune cell infiltration.
Altogether, based on the published data, shed syndecan-4 fragments seem to have a net anti-fibrotic effect that, in the setting of myocardial infarction, contributes to heart rupture and mortality. In other forms of heart disease, such as pressure-overload, the role of syndecan-4 shedding has not been investigated. We speculate that increasing the shedding will result in anti-fibrotic effects opposing the pro-fibrotic effects of full-length syndecan-4.
We have shown that, in cardiac myocytes and fibroblasts, syndecan-4 shedding is determined by LPS, IL-1β, and TNFα through NFkB (Strand et al. 2013; Fig. 4). Thus, cardiac syndecan-4 shedding might be modulated through the manipulation of the IL-1β-NFkB, TNFα-NFkB, (LPS-) TLR-4-NFkB axes.
Increased circulating levels of syndecans in patients with heart diseases with fibrosis
Shed syndecan fragments are not only found in the ECM of tissues, but also in body fluids such as wound fluids and blood. The detection of syndecan fragments in blood raises the questions of whether syndecans can serve as biomarkers in disease.
Increased levels of shed syndecan-1 fragments are found in lavage fluid and fibrotic areas of lung samples from patients with idiopathic pulmonary fibrosis (IFP), an interstitial lung disease characterized by severe and progressive fibrosis (Kliment et al. 2009). In alveolar epithelial cells, shed syndecan-1 fragments induce neutrophil chemotaxis and fibrogenesis, and thus, shed syndecan-1 has been suggested as a marker of fibrosis. After myocardial infarction in rats, increased levels of syndecan-1 are detected in the circulation, and serum levels correlate with myocardial expression (Lei et al. 2012). Recently, circulating syndecan-1 levels have been related to cardiac function in a study of almost 600 patients with chronic heart failure (HFrEF and HFpEF; Tromp et al. 2014). Patients with higher circulating syndecan-1 levels had higher levels of the heart failure marker NT-proBNP and worse renal function. Interestingly, multivariable regression analyses showed a positive correlation between circulating syndecan-1 levels and markers of fibrosis and remodeling. A doubling of the syndecan-1 levels was associated with an increased risk of all-cause mortality and re-hospitalization in heart HFpEF, and circulating syndecan-1 levels were useful for risk stratification in this patient group. Thus, syndecan-1 is a promising prognostic biomarker in heart failure when levels are likely to be related to fibrosis in the heart.
We have shown that syndecan-4 shed fragments are found within the myocardium of the failing human heart (Strand et al. 2013) suggesting local effects. Furthermore, syndecan-4 fragments are released from the human and mouse heart into the circulation (Strand et al. 2015). Thus, shed syndecan-4 fragments detected in heart failure patients might originate in the heart, although other sources are also likely because of the ubiquitous expression of this syndecan isoform. In studies with relatively few enrolled subjects, serum levels of syndecan-4 are increased in patients with acute myocardial infarction (Kojima et al. 2001), in patients with chronic failure (Bielecka-Dabrowa et al. 2013; Takahashi et al. 2011), and in patients with hypertension (Bielecka-Dabrowa et al. 2015a, 2015b). Syndecan-4 serum levels correlate positively with left ventricle geometry parameters such as mass and negatively with the ejection fraction, suggesting that circulating syndecan-4 levels reflect cardiac remodeling and function. Although these initial studies show that syndecan-1 and −4 are promising as biomarkers in heart disease with fibrosis, further studies are needed to determine whether they offer clinical value next to existing biomarkers.
ECM enzymes regulating syndecan shedding
The enzymatic environment in the myocardium is dramatically changed upon pathological stimuli and contributes to altering the processing of syndecans in heart disease, i.e., shedding and GAG modulation. Shedding of syndecans is believed to be directed by specific sheddase enzymes (Manon-Jensen et al. 2010), i.e., by MMPs at the highly conserved juxtamembrane domain and by ADAMTS (a disintegrin and metalloproteinase domain with thrombospondin motifs) enzymes close to the GAG attachment sites. We have shown that excision of the juxtamembrane domain (syndecan-4 Δ138-145) or mutation of the three GAG attachment sites (S44A, S62A, S64A) in syndecan-4 reduces its shedding (Strand et al. 2013), supporting this idea. ADAMTS levels are altered in the remodeling and failing rat heart (Vistnes et al. 2014). In particular, ADAMTS-1, −4, and −8 are increased in pressure-overloaded hearts, suggesting a role for these particular ADAMTS enzymes, in addition to MMPs, in syndecan shedding. Upon LPS challenge, a potent signal for syndecan-4 shedding in the heart, a profound up-regulation of ADAMTS4, together with ADAMTS1 and MMP9, suggest that these enzymes mediate syndecan-4 shedding during LPS-induced cardiac dysfunction (Strand et al. 2015). Heparanase, a GAG-modifying enzyme that trims or degrades heparan sulphate GAGs, has been found in tissue fluids and to regulate syndecan-1 function (Edwards 2012); however, its role in cardiac disease remains unknown. Since shed syndecan-4 ectodomains affect pro-fibrotic processes and infarct healing, an understanding of those enzymes that mediate the shedding might contribute to new strategies to modulate the amount of shedding and, thereby, fibrosis.
Potential therapeutic targeting and clinical use of syndecans as biomarkers in cardiac fibrosis
Assessment of cardiac fibrosis is currently not part of the routine clinical work-up of patients with heart disease. Despite numerous experimental studies showing the effects of cardiac fibrosis on heart function and disease progression, precise and reliable techniques to define the types and amounts of fibrosis that are associated with the outcome in patients are lacking. Tissue section histology has for long been the gold standard for the analysis of fibrosis (i.e., collagen), offering the advantages of established protocols, distribution assessment (e.g., perivascular, interstitial, or focal accumulation), and quality assessment (e.g., collagen cross-linking). Since cardiac biopsies from patients are not readily available, histology has low clinical applicability with regards to the assessment of cardiac fibrosis. Thus, an unmet need exists for the clinical evaluation of cardiac fibrosis. Such evaluation is currently provided by the research community by the development of imaging techniques and blood biomarkers as surrogate read-outs for fibrosis in heart disease patients. Late gadolinium enhancement (LGE) magnetic resonance imaging (MRI) is currently being tested together with collagen blood biomarkers, i.e., N- and C-terminal collagen I and III fragments (PINP/PIIINP and PICP/PIIICP, respectively).
Blood biomarkers could be useful for the estimation of early stage cardiac fibrosis in well-defined patient groups. We speculate, based on previous studies showing increased syndecan shedding in heart disease, that, next to PINP, PIIINP, PICP and PIIICP, the use of circulating shed syndecan-1 and −4 fragments has potential for this purpose.
Therapeutic targeting of cardiac fibrosis is based on the assumption that fibrosis remains reversible during disease progression. However, some studies suggest that fibrosis becomes irreversible or partly irreversible after prolonged remodeling, for instance, in mice overexpressing calcineurin in cardiomyocytes (Berry et al. 2011) and in aortic stenosis patients years after aortic valve replacement (Lund et al. 2003). These studies suggest that prevention of fibrosis is a therapeutic goal in heart disease. Prevention of cardiac fibrosis could be achieved by a reduction of risk factors, e.g., reduced hypertension and improved stratification of patients at risk, or by anti-fibrotic drug pipelines to be offered to the patients at an early time-point after diagnosis, with close follow-up.
One way of developing anti-fibrotic drugs is to target one molecule or one pathway, for instance, the syndecan-4-calcineurin-NFAT pathway. This could be made possible by developing small molecules targeting the protein-protein interaction, preferably in a cell-type-specific manner. However, many interdependent and redundant pathways lead to fibrosis, suggesting that “dirty” anti-fibrotic drugs are more likely to work than those targeting one molecule, one interaction, or one pathway. This is exemplified by mineralcorticoid receptor antagonists (MRAs) that inhibit aldosterone and the RAAS axis and reduce cardiac fibrosis, as currently used in clinics (Vizzardi et al. 2014). To speculate with regards to the inhibition of the pro-fibrotic effects of syndecan-1 and −4 by using a broader approach than small molecules, the inhibitors of TGFβ, TLR-4, TNFα, or IL1β might be a promising direction in which to proceed. The last-mentioned are used clinically in inflammatory diseases such as rheumatic arthritis, e.g., the monoclonal TNFα antibodies such as infliximab and the IL1β receptor antagonist anakinra. Although immunomodulatory clinical studies have been disappointing and shown no effect in heart disease, e.g., randomized placebo-controlled anti-TNFα studies, study design has been criticized (Gullestad et al. 2006; Heymans et al. 2009). Thus, more targeted studies are needed, and we believe that immunmodulatory treatment in heart disease, among many other effects, is also likely to modulate syndecan levels.
Future studies are needed to establish whether syndecans can be used as biomarkers with a clinical value for the assessment of cardiac fibrosis in heart disease patients. Moreover, much more information is needed to enable us to understand and establish how, how much, when, and in which forms of heart disease to inhibit or stimulate syndecan-dependent fibrotic pathways.
Summary, concluding remarks, and future directions
Cardiac fibrosis is one of the main and common responses of the heart to pathological stimuli. Adequate timing and the quantity and quality of fibrosis are crucial for heart function and survival. The transmembrane proteoglycan family of syndecans extends into the cardiac ECM and mediates ECM-cell responses and signaling events important for the regulation of cardiac fibrosis. We have reflected here upon the roles of full-length and shed syndecans in these processes. An improved understanding of syndecans might lead to the development of new therapeutic and prognostic tools for heart disease with fibrosis, thus affecting millions of people worldwide.
We are grateful to Dr. Mari E. Strand for technical assistance with the graphical abstract.
Compliance with ethical standards
Conflict of interest
The authors declare that they have no conflict of interest.