Inflammation and Syndecan-4 Shedding from Cardiac Cells in Ischemic and Non-Ischemic Heart Disease

Circulating biomarkers reflecting cardiac inflammation are needed to improve the diagnostics and guide the treatment of heart failure patients. The cardiac production and shedding of the transmembrane proteoglycan syndecan-4 is upregulated by innate immunity signaling pathways. Here, we investigated the potential of syndecan-4 as a blood biomarker of cardiac inflammation. Serum syndecan-4 was measured in patients with (i) non-ischemic, non-valvular dilated cardiomyopathy (DCM), with (n = 71) or without (n = 318) chronic inflammation; (ii) acute myocarditis (n = 15), acute pericarditis (n = 3) or acute perimyocarditis (23) and (iii) acute myocardial infarction (MI) at day 0, 3 and 30 (n = 119). Syndecan-4 was investigated in cultured cardiac myocytes and fibroblasts (n = 6–12) treated with the pro-inflammatory cytokines interleukin (IL)-1β and its inhibitor IL-1 receptor antagonist (IL-1Ra), or tumor necrosis factor (TNF)α and its specific inhibitor infliximab, an antibody used in treatment of autoimmune diseases. The levels of serum syndecan-4 were comparable in all subgroups of patients with chronic or acute cardiomyopathy, independent of inflammation. Post-MI, syndecan-4 levels were increased at day 3 and 30 vs. day 0. IL-1Ra attenuated IL-1β-induced syndecan-4 production and shedding in vitro, while infliximab had no effect. In conclusion, syndecan-4 shedding from cardiac myocytes and fibroblasts was attenuated by immunomodulatory therapy. Although its circulating levels were increased post-MI, syndecan-4 did not reflect cardiac inflammatory status in patients with heart disease.


Introduction
Heart failure is a common syndrome of multiple etiologies, carrying high morbidity, mortality and costs for society. Despite therapeutic advancement, the management of heart failure remains a significant challenge, and research efforts are increasingly focused on identifying novel cardiovascular biomarkers to aid in diagnosis, risk stratification, predict prognosis and guide treatment [1]. The innate immune system is activated in tissue injury, in response to hemodynamic stress or during infection. Inflammation is a central pathophysiological process across the entire spectrum of human or experimental heart failure, whether acquired or genetic, acute or chronic, and is associated with a poor outcome independent of etiology [2][3][4][5]. Nevertheless, trials attempting immunomodulating therapy have not been consistent, and although there is hope generated from certain trials [6,7], an improved mechanistic understanding of cardiac inflammation, finding new drug targets, developing drugs, and designing appropriate trials are needed, if anti-inflammatory treatment is to be implemented for heart failure [2,8]. To this end, circulating biomarkers reflecting cardiac inflammation are sought to provide clinicians with valuable information and to reduce the use of biopsies to describe the inflammatory status of the heart.
Syndecan-4 is a transmembrane proteoglycan that is part of an evolutionary old, fourmembered family encoded by the SDC1-4 genes [9][10][11][12]. They consist of a core protein substituted with three-five covalently attached heparan sulfate (HS) glycosaminoglycan (GAG) chains on the extracellular domain. The short cytoplasmic signaling domain, the transmembrane domain, the attachment sites for the GAGs and the extracellular cleavage domain are highly conserved. Syndecans have been implicated in a wide variety of functions in numerous cell types and tissues, and typically function as co-receptors. The MW of syndecan-4 is around 20 kDa, although it normally forms homodimers and is found in protein gels at around 35-40 kDa. The HS-substituted ectodomains can be enzymatically cleaved from cell surfaces and released into the local milieu and circulation [13]; termed shedding. Enzymes responsible for shedding are called sheddases and include matrix metalloproteinases (MMPs), thrombin and endopeptidases such as seralysins, astacins and a disintegrin and metallo-proteinases (ADAMs).
Several studies have investigated syndecan-1 [22][23][24][25][26][27][28] and -4 [29][30][31][32] as potential circulating biomarkers in heart disease, with a focus on ischemia and MI, dilated cardiomyopathy (DCM), and acute and chronic heart failure. We here hypothesized that the amounts of circulating syndecan-4 were increased in patients with inflammatory heart disease. To test our hypothesis, we used a commercially available enzyme-linked immunosorbent assay (ELISA) kit detecting syndecan-4, and recruited a relatively high number of patients for blood sampling with detailed phenotyping. This included separating the patients into inflammatory and non-inflammatory DCM based on the quantification of infiltrating cells through immunohistochemistry analysis of biopsies. Moreover, we hypothesized that immunomodulatory therapy would reduce syndecan-4 levels and shedding from cardiomyocytes and cardiac fibroblasts, and thereby reduce the pro-inflammatory syndecan-4-mediated effects locally in the heart. To test this, we virally overexpressed syndecan-4 in primary heart cultures and used a custom-made syndecan-4 antibody to readily detect syndecan-4 and syndecan-4 fragments.
Serum samples were obtained from n = 119 patients with myocardial infarction at university hospitals in Leuven, Belgium. In brief, patients presenting with MI, with or without ST-elevation, and scheduled for coronary angiography between October 2013 and April 2015 were prospectively enrolled and extensively phenotyped. Serum samples were obtained at hospital admission and at day 3 and day 30 post-MI. See [35] for the detailed method description, exclusion criteria and follow-up care of patients.
Serum samples were obtained during aortic valve replacement surgery at Oslo University Hospital Ullevål, Oslo, Norway, from n = 15 patients with severe, symptomatic aortic stenosis (AS). These patients constitute the subgroup receiving cold blood cardioplegia of a previously published cohort [36]. Blood was drawn from a coronary sinus catheter and from the cannulated radial artery immediately after onset of cardiopulmonary bypass and before clamping of the aorta. See [36] for detailed method description and patient characteristics.

Myocardial Tissue Samples from Patients
Interventricular myocardial biopsies were obtained during septal myectomy at Oslo University Hospital Rikshospitalet, Oslo, Norway, from n = 3 patients with hypertrophic obstructive cardiomyopathy (HOCM). Detailed protocol description and patient characteristics, including genetics, have been previously reported [37].

ELISA Analyses of Serum Samples
We detected human syndecan-4 in serum using a commercially available human syndecan-4 ELISA kit (27188, IBL, Gunma, Japan) according to protocol, by one experienced researcher blinded to patient group. The two specific antibodies used in this ELISA were generated against residues 27-39 and residues 19-145 of the recombinant syndecan-4 ectodomain fused to the maltose binding protein [29]. The detection limit is 20 pg/mL.
A literature search was conducted in PubMed, National Institutes of Health (NIH), with the search words "syndecan-4" and "biomarker". Abstracts of the resulting hits (n = 91) were read to identify those reporting ELISA data of syndecan-4 from human blood samples. The next and final selection criteria were papers specifying the use of serum samples, the IBL human syndecan-4 ELISA, and including a control group of healthy subjects. These data were used to generate a table overviewing absolute levels of serum syndecan-4 in various cohorts, to compare to our data.

RNA Isolation and Quantitative Real-Time PCR (qRT-PCR)
Total RNA was isolated from cells using RNeasy mini (74106, Qiagen Nordic, Oslo, Norway), as previously described [20]. In brief, reverse transcription and cDNA synthesis were performed with the iScript cDNA Synthesis Kit (Bio-Rad Laboratories, Inc., Hercules, CA, USA). Pre-designed TaqMan assays (Rn00561900_m1 and Rn06291926_g1) were used with Taq-Man Universal PCR Master Mix (#4304437, both from Applied Biosystems, Foster City, CA, USA) to quantify syndecan-4 mRNA levels by qRT-PCR. Results were detected on the QuantStudio 3 Real-Time PCR System (Applied Biosystems). mRNA data were normalized to expression of ribosomal protein L4 (Rpl4).

Statistics
Data are expressed as dot plots with bar group means ± S.D. Statistical analyses were performed using GraphPad Prism 9 and statistical differences considered significant for p < 0.05. Significance was determined using Student's paired and unpaired t-test or oneway ANOVA followed by Tukey's post hoc test, as detailed in the figure and table legends.

Syndecan-4 Is Shed in and from the Human Heart
To measure syndecan-4 shedding from cells or tissue and in blood, we performed Western blotting and ELISA, respectively ( Figure 1A). The custom-made syndecan-4 antibody against the cytoplasmic tail of syndecan-4 ( Figure 1A, left panel) detects both full-length (FL) syndecan-4 and the cellular fragment (CF) left in the cell after release of the ectodomain in mouse, rat and human samples (see [16,20] for details). Thus, the levels of the CF reflect syndecan-4 shedding were measured in tissue and cells. The detection of syndecan-4 FL (35-40 kDa) and the CF (10-15 kDa) in diseased human myocardium is shown here for demonstration of the protein bands and the ongoing syndecan-4 shedding observed in heart disease ( Figure 1B), with biological variation in human samples, i.e., a stronger syndecan-4 signal in two out of the three samples tested. To measure the levels of the syndecan-4 shed ectodomain (SE) in blood samples, we used an ELISA kit equipped with an antibody against the ectodomain ( Figure 1A, right panel). To test whether circulating syndecan-4 originated from the heart, we used ELISA to measure the venous-arterial difference in serum from the coronary sinus, i.e., venous blood from the heart, and the radial artery, of aortic stenosis patients during open heart surgery. Consistent with syndecan-4 ectodomains being released into the circulation from the human heart, syndecan-4 levels were higher in the coronary sinus (10.73 ± 4.24 vs. 6.50 ± 2.66 ng/mL in the radial artery) ( Figure 1C). To study syndecan-4 shedding from major cell types in the heart, we prepared primary cardiomyocyte and cardiac fibroblast ( Figure 1D) cultures from neonatal rats. To provide a robust assay for the measurement of syndecan-4 shedding in cardiac cells, cells were transduced with a virus encoding mouse syndecan-4 (AdSdc4) or vehicle (AdNull). Overexpression of syndecan-4 resulted in the expected increase in FL (at or somewhat below 37 kDa in fibroblasts vs. cardiomyocytes, respectively) and CF, the latter by constitutive shedding from cardiomyocytes ( Figure 1E) and fibroblasts ( Figure 1F). We considered our methods suited to detect syndecan-4 shedding from cardiac cells, in myocardial tissue, and in blood from patients.

Serum Syndecan-4 Levels Do Not Reflect the Inflammatory Status of the Heart in Patients with Heart Disease
To test our hypothesis that syndecan-4 could constitute a blood biomarker of cardiac inflammation, syndecan-4 was measured by ELISA in serum samples from a total of 549 well-characterized patients with cardiomyopathies of various etiologies and stages of disease, or acute MI. Flow diagrams of the patient selection process for the two cohorts are shown in Figure 2. The patient characteristics were previously reported [33][34][35].
To investigate whether circulating syndecan-4 levels were elevated in patients with a high degree of inflammation in the myocardium, we separated the DCM patients into inflammatory (chronic myocarditis, n = 71) and non-inflammatory (n = 318) DCM, where inflammation was defined as >14 infiltrating cells/mm 2 on immunohistochemistry. However, we found no differences in serum syndecan-4 levels between the two groups (average 16.26 ng/mL vs. 15.20 ng/mL in inflammatory and non-inflammatory DCM, respectively) ( Figure 3).
To examine the possibility that syndecan-4 levels could differentiate between subgroups of cardiomyopathies characterized by acute immune responses, we sub-divided patients based on their specific diagnoses, i.e., acute myocarditis (n = 15), acute pericarditis (n = 3) and acute perimyocarditis (n = 23). Serum syndecan-4 was on average 13.66 ng/mL in myocarditis, 13.54 ng/mL in pericarditis and 17.47 ng/mL in perimyocarditis, and again, no differences were observed between the groups ( Figure 3). Apart from the very small group of pericarditis patients (n = 3), every patient group analyzed so far displayed a comparable variation of syndecan-4 levels, with comparable minimum (min.; 2.44-6.95 ng/mL) and maximum (max; 31.28-38.72 ng/mL) levels between the groups, and an average range of 30.90 ng/mL within each group. Thus, our results suggest that circulating syndecan-4 is of limited value in the clinic with regard to reflecting cardiac inflammation.  Table S1. (D) Schematic illustrating the adenoviral vectors used to transduce cultures of prim neonatal rat cardiac cells. Cells were transduced with an Adenovirus serotype 5 (dE1/E3) with RG fiber modification encoding mouse syndecan-4 (AdSdc4) or vehicle (AdNull). (E,F) Representat immunoblots of mouse syndecan-4 FL and CF in primary cultures of neonatal rat cardiomyocy (n = 9, (E)) and cardiac fibroblasts (n = 12, (F)). Vinculin was used for loading control. Data are p sented as mean ± S.D. Statistical differences were tested using Student's paired t-test (C).

Serum Syndecan-4 Levels Do Not Reflect the Inflammatory Status of the Heart in Patients with Heart Disease
To test our hypothesis that syndecan-4 could constitute a blood biomarker of card inflammation, syndecan-4 was measured by ELISA in serum samples from a total of 5 well-characterized patients with cardiomyopathies of various etiologies and stages of d ease, or acute MI. Flow diagrams of the patient selection process for the two cohorts a shown in Figure 2. The patient characteristics were previously reported [33][34][35].  Supplementary  Table S1. (D) Schematic illustrating the adenoviral vectors used to transduce cultures of primary neonatal rat cardiac cells. Cells were transduced with an Adenovirus serotype 5 (dE1/E3) with RGD fiber modification encoding mouse syndecan-4 (AdSdc4) or vehicle (AdNull). (E,F) Representative immunoblots of mouse syndecan-4 FL and CF in primary cultures of neonatal rat cardiomyocytes (n = 9, (E)) and cardiac fibroblasts (n = 12, (F)). Vinculin was used for loading control. Data are presented as mean ± S.D. Statistical differences were tested using Student's paired t-test (C).
Acute MI is another condition with a high degree of inflammation. We obtained serum samples at hospital admission (day 0), and at day 3 and day 30 post-MI from each patient (n = 119). Syndecan-4 showed slightly higher mean values at day 3 (16.29 ng/mL, p = 0.0002) and day 30 (15. To investigate whether circulating syndecan-4 levels were elevated in patients with a high degree of inflammation in the myocardium, we separated the DCM patients into inflammatory (chronic myocarditis, n = 71) and non-inflammatory (n = 318) DCM, where inflammation was defined as >14 infiltrating cells/mm 2 on immunohistochemistry. However, we found no differences in serum syndecan-4 levels between the two groups (average 16.26 ng/mL vs. 15.20 ng/mL in inflammatory and non-inflammatory DCM, respec-    Lastly, after demonstrating that circulating syndecan-4 levels did not differentiate between the cardiac pathologies investigated, we wondered whether syndecan-4 levels were elevated in patients vs. healthy controls. On our end, we had access only to a small number of control subjects (n = 5), and found that these showed an average level of serum syndecan-4 of 13.30 ng/mL (Figure 3). The range of syndecan-4 values (min. value of 7.88 and max. value of 21.09 ng/mL) in healthy controls overlapped with patients of all etiologies, apart from the pericarditis group. Thus, we were unable to conclude whether serum syndecan-4 was increased vs. the controls in our cardiomyopathy or acute MI patient cohorts.

Interleukin 1β-Increased Syndecan-4 Levels Is Attenuated by Immunomodulatory Therapy in Cultured Cardiomyocytes
We capitalized on our previous finding that mediators of innate immunity, i.e., TNFα and IL-1β, upregulate production and shedding of syndecan-4 from cardiac myocytes and fibroblasts [18,20], in order to investigate whether immunomodulatory therapy affected syndecan-4 levels in such cells. Primary neonatal rat cardiomyocytes were treated with recombinant IL-1β and TNFα of rat origin. Treatment with IL-1β resulted in a 3.4-fold increase in rat syndecan-4 mRNA ( Figure 4A). Interestingly, we found that co-treatment with the immunomodulating IL-1 receptor antagonist IL-1Ra resulted in an attenuated syndecan-4 increase. Consistent with the upregulated syndecan-4 mRNA, immunoblots showed that IL-1β increased syndecan-4 FL and CF 3.8-and 2.5-fold vs. the non-treated, respectively, in cardiomyocytes overexpressing mouse syndecan-4 ( Figure 4B). Upon cotreatment with IL-1Ra, the FL levels were not increased vs. the non-treated; however, the levels of CF appeared not to be affected. When performing a separate statistical analysis comparing FL and CF levels in IL-1β vs. IL-1β + IL-1Ra only, IL-1Ra resulted in a 25% reduction in FL (p < 0.05) and a tendency for reduced CF (p = 0.07, 15%).
Treatment with TNFα resulted in 2-fold higher levels of endogenous rat syndecan-4 mRNA ( Figure 4C). However, co-treatment with its inhibitor, the anti-TNFα antibody infliximab, did not affect syndecan-4 production. Similarly, treatment with TNFα resulted in the upregulation of syndecan-4 with 2.9-and 3.1-fold increases in FL and CF protein, respectively, in cardiomyocytes overexpressing mouse syndecan-4 ( Figure 4D), and cotreatment with infliximab did not affect TNF-α-induced syndecan-4 protein levels, whether compared to non-treated or in a separate statistical analysis comparing FL and CF levels in TNFα vs. TNFα + infliximab only.

Interleukin 1β-Increased Syndecan-4 Levels and Shedding Is Attenuated by Immunomodulatory Therapy in Cultured Cardiac Fibroblasts
Primary neonatal rat cardiac fibroblasts were treated with recombinant IL-1β and TNFα of rat origin. Treatment with IL-1β resulted in a 7.7-fold increase in syndecan-4 mRNA ( Figure 5A). Importantly, co-treatment with IL-1Ra resulted in a 50% reduction in IL-1β-induced syndecan-4 production. These effects of IL-1β and IL-1Ra were also evident when we investigated syndecan-4 protein levels. IL-1β induced a 20.2-fold increase in syndecan-4 FL and a 14.8-fold increase in CF ( Figure 5B). Co-treatment with IL-1Ra resulted in a clear reduction of IL-1β-induced syndecan-4 FL protein, while CF levels showed no change. When performing a separate statistical analysis comparing FL and CF levels in IL-1β vs. IL-1β + IL-1Ra only, IL-1Ra resulted in a 39% reduction in FL (p < 0.001) and a 27% reduction in CF (p < 0.01).
Treatment with TNFα resulted in a 2.4-fold increase in syndecan-4 mRNA ( Figure 5C). Interestingly, the TNFα-induced increase was attenuated when the cardiac fibroblasts were co-treated with infliximab. In line, TNFα resulted in a 2.4-and 4.3-fold increase in syndecan-4 FL and CF protein, respectively, in cardiac fibroblasts overexpressing mouse syndecan-4 ( Figure 5D). However, infliximab resulted in a non-significant tendency for increased syndecan-4 FL and CF protein levels, and consequently, did not have an attenuating effect on TNF-α-induced syndecan-4 protein levels when compared to non-treated or in the separate statistical analysis of TNFα vs. TNFα + infliximab.
in the upregulation of syndecan-4 with 2.9-and 3.1-fold increases in FL and CF protein, respectively, in cardiomyocytes overexpressing mouse syndecan-4 ( Figure 4D), and cotreatment with infliximab did not affect TNF-α-induced syndecan-4 protein levels, whether compared to non-treated or in a separate statistical analysis comparing FL and CF levels in TNFα vs. TNFα + infliximab only. Figure 4. Attenuation of interleukin 1β-increased syndecan-4 levels immunomodulatory therapy in cultured cardiomyocytes. Primary cardiomyocyte cultures were prepared from neonatal rat hearts. (A) Relative mRNA expression of endogenous rat syndecan-4 (Sdc4) in cardiomyocytes following 24 h treatment with interleukin (IL)-1β (10 ng/mL) compared to non-treated control or co-treatment with the IL-1 receptor antagonist (IL-1Ra; 100 ng/mL), n = 6. (B) Representative immunoblot and quantification of full-length (FL) syndecan-4 and the cellular fragment (CF) seen in cells after release of the ectodomain, in cardiomyocytes transduced with an adenovirus encoding mouse syndecan-4 (AdSdc4) and after 24 h treatment with IL-1β compared to non-treated or co-treatment with IL-1Ra, n = 9. (C) Relative mRNA levels of endogenous rat syndecan-4 in cardiomyocytes after 24 h treatment with tumor necrosis factor (TNF)α (10 ng/mL) compared to non-treated control or co-treatment Figure 4. Attenuation of interleukin 1β-increased syndecan-4 levels immunomodulatory therapy in cultured cardiomyocytes. Primary cardiomyocyte cultures were prepared from neonatal rat hearts. (A) Relative mRNA expression of endogenous rat syndecan-4 (Sdc4) in cardiomyocytes following 24 h treatment with interleukin (IL)-1β (10 ng/mL) compared to non-treated control or co-treatment with the IL-1 receptor antagonist (IL-1Ra; 100 ng/mL), n = 6. (B) Representative immunoblot and quantification of full-length (FL) syndecan-4 and the cellular fragment (CF) seen in cells after release of the ectodomain, in cardiomyocytes transduced with an adenovirus encoding mouse syndecan-4 (AdSdc4) and after 24 h treatment with IL-1β compared to non-treated or co-treatment with IL-1Ra, n = 9. (C) Relative mRNA levels of endogenous rat syndecan-4 in cardiomyocytes after 24 h treatment with tumor necrosis factor (TNF)α (10 ng/mL) compared to non-treated control or co-treatment with the anti-TNFα antibody infliximab (100 ng/mL), n = 6. (D) Representative immunoblot and quantification of syndecan-4 FL and CF in cardiomyocytes transduced with mouse AdSdc4 and treated for 24 h with TNFα vs. non-treated or co-treatment with infliximab, n = 9. mRNA data were normalized to ribosomal protein L4 (Rpl4). Vinculin was used for loading control. Data are presented as mean ± S.D. Statistical differences were tested using one-way ANOVA with Tukey's post hoc test ((A-D), p-values given) or Student's unpaired t-test ((B) § p < 0.05, comparing Il-1Ra to IL-1β alone, not relative to control). The numerical data supporting the graphs can be found in Supplementary  Tables S3-S6. Taken together, these results from primary heart cultures showed that IL-1β was a strong inducer of syndecan-4 mRNA and protein production in cardiac myocytes and fibroblasts, and that treatment with IL-1Ra reduced the levels of FL syndecan-4 and its shedding. Although TNFα was a potent inducer of syndecan-4 mRNA and protein in cardiac myocytes and fibroblasts, co-treatment with infliximab did not alter the FL or shed protein. shedding of the ectodomain, in rat cardiac fibroblasts transduced with an adenovirus encoding mouse syndecan-4 (AdSdc4) and following 24 h treatment with IL-1β compared to non-treated or co-treatment with IL-1Ra, n = 12. (C) Relative mRNA of rat syndecan-4 in cardiac fibroblasts after 24 h treatment with tumor necrosis factor (TNF)α (10 ng/mL) vs. non-treated control or co-treatment with the anti-TNFα antibody infliximab (100 ng/mL), n = 6. (D) Representative immunoblot and quantification of syndecan-4 FL and CF in cardiac fibroblasts transduced with mouse AdSdc4 and treated for 24 h with TNFα vs. non-treated or co-treatment with infliximab, n = 12. mRNA data were normalized to ribosomal protein L4 (Rpl4). Vinculin was used for loading control. All data are presented as mean ± S.D. Statistical differences were tested using one-way ANOVA with Tukey's post hoc test ((A-D), p-values given) or Student's unpaired t-test ((B) § p < 0.05; § § § p < 0.001, comparing Il-1Ra to IL-1β alone, not relative to controls). The numerical data supporting the graphs can be found in supplementary Table S7-S10. with TNFα vs. non-treated or co-treatment with infliximab, n = 12. mRNA data were normalized to ribosomal protein L4 (Rpl4). Vinculin was used for loading control. All data are presented as mean ± S.D. Statistical differences were tested using one-way ANOVA with Tukey's post hoc test ((A-D), p-values given) or Student's unpaired t-test ((B) § p < 0.05; § § § p < 0.001, comparing Il-1Ra to IL-1β alone, not relative to controls). The numerical data supporting the graphs can be found in Supplementary Tables S7-S10.

Discussion
The present work investigates (i) whether circulating levels of syndecan-4 reflect the cardiac inflammatory status in patients with heart disease, and (ii) the effects of immunomodulatory inhibitors used in clinical practice on syndecan-4 levels and shedding in vitro. The presence of syndecan-4 shedding fragments in diseased myocardial tissue, and the increased veno-arterial difference of shed syndecan-4 ectodomains across the heart, indicated that syndecan-4 ectodomains are released in and from the human heart. Importantly, serum syndecan-4 showed comparable levels in patients with cardiomyopathies of different etiologies, i.e., inflammatory vs. non-inflammatory DCM, myocarditis, pericarditis and perimyocarditis, and following MI, indicating that syndecan-4 has limited value as a clinical circulating biomarker of cardiac inflammation. In cardiac cells in vitro, a decrease in IL-1β-induced syndecan-4 mRNA, protein and shedding following treatment with IL-1Ra suggested that syndecan-4 production and shedding is among the downstream targets of IL-1 blockage. Treatment with an anti-TNFα antibody, infliximab, failed to show a similar effect on the TNFα-induced upregulation of syndecan-4 protein levels and shedding.
Published studies have indicated that circulating syndecan-4 holds promise as a cardiac biomarker. Although these include cohorts with relatively few patients, circulating levels of syndecan-4 are increased after acute MI [29], resistant hypertension [39] and chronic heart failure [30]. Indicating that syndecan-4 could be a blood biomarker reflecting cardiac remodeling and function, serum syndecan-4 correlated positively with LV geometry and negatively with pumping function as measured by EF [30,31]. Based on the knowledge that syndecan-4 FL levels and the shedding of its ectodomain are upregulated in response to innate immunity pathways [18,20], we set out to evaluate whether syndecan-4 could also hold promise as a circulating biomarker of cardiac inflammation. We tested this by measuring the serum concentrations of syndecan-4 in patients with DCM of inflammatory and non-inflammatory origin, patients with acute cardiomyopathies, and in patients with acute MI, when admitted and after three and 30 days. However, our results from a high number of patients did not show differences in syndecan-4 levels among patients with chronic, acute or minor inflammation. Overall, we conclude from our findings that there is limited value in the use of syndecan-4 as a clinical circulating biomarker of cardiac inflammation. Despite our disappointing results, we believe there is scientific value in publishing such neutral data. We performed the analyses on two well-characterized patient cohorts that included a high number of patients. Respect for the patients and the valuable samples they provide to research, together with the cost of analyses, certainly justify that information of this scale ought to be shared among researchers to avoid unfruitful replications and to guide future hypotheses.
Interestingly, we did demonstrate that serum syndecan-4 increased during the inflammatory phase in patients with acute MI, with levels peaking at day 3 and remaining elevated at day 30, compared to at hospital admission. These results are supportive of a role for syndecan-4 shedding in the ischemic heart, and are in line with previous studies showing increased levels of circulating syndecan-4 after MI [29,32]. Such a role has also been demonstrated for syndecan-1, which is shed during ischemic conditions of the heart causing endothelial damage, e.g., cardiac arrest [22,23], chronic heart failure [24] and microvascular occlusions in ST elevation MI [40].
Another aspect of our study of serum syndecan-4 in patients is its level in healthy controls. Our study failed to conclude whether circulating syndecan-4 levels are at all increased in heart disease patients vs. healthy controls. In our study, the average level in healthy controls was 13.3 ng/mL. Considering that the mean and range of syndecan-4 levels in our patients overlapped with levels in healthy controls, albeit the latter in a low number, we evaluated the levels of serum syndecan-4 obtained using the same ELISA kit and sample material in published, historical controls (Table 1). Although a number of studies have found significant changes in serum syndecan-4 in patient cohorts vs. healthy subjects, Table 1 demonstrates that there is a spread in the reported mean values for circulating syndecan-4 in control populations, ranging from 4.34 to 18.99 ng/mL [30,39,[41][42][43][44][45]. Of note, syndecan-4 levels measured in healthy controls in one study are comparable to disease states in others. Collectively, these studies are in line with our current results showing overlapping syndecan-4 concentrations among patient groups. Consequently, to advance our understanding of shed syndecan-4 as a biomarker in pathology, efforts should be directed towards truly defining circulating syndecan-4 levels in healthy controls. Historical, published levels of serum syndecan-4 (ng/mL) in patient cohorts and healthy controls measured using the same ELISA kit (IBL 27188, Gunma, Japan) as in the present study. Our color-coding: blue, syndecan-4 in healthy controls; green, syndecan-4 increased in patients vs. healthy controls in given study; red, syndecan-4 reduced in patients vs. healthy controls; grey, no difference patients vs. healthy controls. SD-IIP, stable disease idiopathic interstitial pneumonia; AE-IIP, acute exacerbation-IIP, SCAP, severe community-acquired pneumonia (CAP); NS, non-significant. Data are presented as mean ± S.E.M [39,[41][42][43][44] or mean ± S.D. [30,45].
Although we did not find that circulating syndecan-4 levels reflected the inflammatory status of the heart, we found syndecan-4 shedding fragments within the diseased human myocardial tissue and increased veno-arterial difference of shed syndecan-4 ectodomains across the heart, showing that syndecan-4 ectodomains were released in and from the human heart. Using a novel approach with syndecan-4 overexpression, IL-1β and TNFα induced a robust increase in syndecan-4 production in cardiac myocytes and fibroblasts in cultured, in line with our previous findings on endogenous syndecan-4 [20]. Thus, the cardiac increase was induced by activation of innate immunity signaling.
We have shown that when cardiomyocytes and cardiac fibroblasts are exposed to shed ectodomains of syndecan-4, they upregulate key molecules involved in inflammatory responses, i.e., cytokines and cell adhesion receptors involved in the binding of immune cells [18]. Thus, we believe that the increased release of syndecan-4 ectodomains resulting from augmented shedding in diseased hearts generates soluble pro-inflammatory molecules that activate immune cell recruitment pathways and propagate immune responses locally in the cardiac tissue.
We and others have demonstrated that syndecan-4 has pro-hypertrophic effects in cardiomyocytes and pro-fibrotic effects in cardiac fibroblasts [14,16,17,[46][47][48]. This means that treatments reducing syndecan-4 levels could be anti-hypertrophic and anti-fibrotic, as well as anti-inflammatory. After discovering here, and in our previous study [20], that IL-1β is a potent driver of syndecan-4 production, we investigated whether immunomodulatory agents used in current clinical practice could modulate cardiac syndecan-4 levels. To this end, we found a decrease in IL-1β-induced syndecan-4 mRNA, protein and shedding following treatment with IL-1Ra (also called anakinra) in vitro, suggesting that syndecan-4 production and shedding is among the downstream targets of IL-1 blockage. Interestingly, IL-1Ra is an IL-1 inhibitor currently under investigation as therapy after acute MI and chronic heart failure [49]. Similar to other therapies targeting pro-inflammatory cytokines, studies on the effect of IL-1Ra have reported conflicting and oftentimes disappointing results. However, there have been several clinical trials showing promising effects of IL-1 blockage by IL-1Ra with regard to peak oxygen consumption and heart failure outcomes [7,50]. The exploration of immunomodulating therapies in patients with cardiac disease remains a field of continued research efforts. Our results, showing that treatment with IL-1Ra reduced the upregulation and shedding of syndecan-4 from cardiac cells, indicate that syndecan-4 levels will be reduced in hearts of patients receiving immunosuppressive therapies, with likely downstream consequences.

Conclusions
In conclusion, increased syndecan-4 production and the release of its ectodomain take place during the innate immune response of the heart, yet levels of syndecan-4 in the circulation were not different between patients with inflammatory and non-inflammatory heart disease. However, we did observe increased circulating levels 3-30 days post-MI compared to at hospital admission, supporting a role for syndecan-4 shedding in the ischemic heart. Thus, we conclude that syndecan-4 had limited clinical value as a circulating biomarker of cardiac inflammation and that immunomodulating therapies targeting IL-1β likely will affect the molecular processes that shed syndecan-4 ectodomains play locally in cardiac tissue.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/biomedicines11041066/s1, Table S1: Corresponding data to Figure 1C, Table S2: Corresponding data to Figure 3, Table S3: Corresponding data to Figure 4A, Table S4: Corresponding data to Figure 4B, Table S5: Corresponding data to Figure 4C, Table S6: Corresponding data to Figure 4D, Table S7: Corresponding data to Figure 5A, Table S8: Corresponding data to Figure 5B, Table S9: Corresponding data to Figure 5C, Table S10: Corresponding data to Figure 5D. Institutional Review Board Statement: The study was conducted in accordance with the Declaration of Helsinki, and protocols for human blood and tissue sampling in Norway were reviewed and approved by the Regional Committee for Medical Research Ethics (REK approvals IDs 290-06114 and S-02292/2017-570). The study protocols for blood sampling in Belgium and the Netherlands were reviewed and approved by the institutions' ethics committees (approval IDs B322201214942 and METC 12-04-013). The animal study protocol was approved by the Norwegian National Animal Research Committee (approval ID IV 1-17U) and is in accordance with the NIH Guide for the Care and Use of Laboratory Animals and the ARRIVE guidelines for reporting animal research.
Informed Consent Statement: Informed consent was obtained from all subjects in the study.