VEGF Contributes to Mesenchymal Stem Cell-Mediated Reversion of Nor1-Dependent Hypertrophy in iPS Cell-Derived Cardiomyocytes

Myocardial hypertrophy is present in many heart diseases, representing a strong predictor of adverse cardiovascular outcomes. Regarding therapeutic intervention, mesenchymal stem cells (MSCs) have been suggested to significantly reduce cardiac hypertrophy and progression to heart failure. Preconditioning of MSCs was previously demonstrated to highly improve their paracrine activity resulting in modulation of immune responses and the progression of diseases. Here, we studied the effects of bone marrow-derived preconditioned MSCs on hypertrophied induced pluripotent stem cell-derived cardiomyocytes (iPS-CM) and also sought to identify MSC-derived antihypertrophic molecules. Phenylephrine (PE) was used to induce hypertrophy in murine iPS-CM, and markers of hypertrophy were identified by microarray analysis. Murine MSCs were treated with IFN-γ and IL-1β to enhance their paracrine activity, and transcriptional profiling was performed by microarray analysis. Hypertrophied iPS-CM were subsequently cocultured with preconditioned MSCs or MSC-conditioned medium (CM), respectively. Effects on hypertrophied iPS-CM were studied by cell area quantification, real-time PCR, and western blot. In some experiments, cells were incubated with fractions of MSC-CM obtained by ultrafiltration or by MSC-CM supplemented with inhibitory antibodies. Intracellular and extracellular levels of vascular endothelial growth factor (VEGF) were evaluated by western blot and ELISA. PE-induced hypertrophy in iPS-CM was associated with an upregulation of neuron-derived orphan receptor (Nor1) expression, activation of Akt, and inhibition of both strongly prevented hypertrophy induction in iPS-CM. VEGF secreted by preconditioned MSCs provoked hypertrophy regression in iPS-CM, and a negative correlation between Nor1 expression and hypertrophic growth could be evidenced. Our results demonstrate that Nor1 expression strongly supports hypertrophy in iPS-CM. Moreover, the secretome of preconditioned MSCs triggered regression of hypertrophy in iPS-CM in a VEGF-dependent manner. We suggest that the delivery of the MSC-derived secretome may represent a therapeutic strategy to limit cardiac hypertrophy. However, additional in vivo studies are needed to prove this hypothesis.


Introduction
Left ventricular hypertrophy (LVH) is a common manifestation of many forms of cardiac disease, including myocardial infarction and hypertension [1]. It is widely recognized as a major risk factor for the development of heart failure and cardiac sudden death representing an established predictor of cardiovascular morbidity and mortality. LVH is a complex and multifactorial condition involving a process of adaptive remodeling, which is usually a compensatory mechanism in response to increased hemodynamic load [2]. Although called compensatory, prolonged hypertrophy increases the risk for progression into heart failure (decompensated hypertrophy). LVH is accompanied by various cardiac and molecular changes, including increase in protein synthesis, cell sizes, and inflammatory and fibrogenic responses which lead to myocardial remodeling and subsequent ventricular dysfunction [1]. Inflammation was shown to be a prominent hallmark of LVH in animal [3] and human studies [4]. Increased wall stress associated with LV overload promotes remodeling processes characterized by increased proinflammatory cytokine expression, leukocyte infiltration, and myocardial destruction. Since a variety of signaling pathways have been shown to be involved in the progression of cardiac hypertrophy [5], understanding the mechanisms underlying the development of cardiac hypertrophy is essential for the treatment of cardiovascular disease. Accordingly, reversal of myocardial hypertrophy has been considered a therapeutic strategy for the reduction of adverse remodeling and prevention of heart failure [6].
The NR4A orphan nuclear receptor subfamily consists of the highly homologous receptors Nur77 (NR4A1), Nurr1 (NR4A2), and Nor1 (NR4A3) which can be regulated by various stimuli, including growth factors, hormones, and inflammatory signals. Recently, there has been much attention paid to the function of these receptors in the cardiovascular system [7]. Nor1 was shown to regulate a number of essential biological processes including inflammation [8], and the first evidence has already been provided for its prohypertrophic properties [9,10]. Therapeutic strategies targeting hypertrophic markers to dampen the inflammatory response may hold promise for beneficial intervention in LVH. In this context, modulation of a macrophage phenotype has been proposed as a new therapeutic approach, as these cells are recognized as key players in cardiovascular diseases. Exacerbated local inflammation in the myocardium mediated by proinflammatory microRNA-155-expressing M1 macrophages induced cardiac hypertrophy and failure in response to pressure overload [11]. In turn, microRNA-155 knockout in macrophages, but not in cardiomyocytes, markedly reduced inflammation, hypertrophy, and dysfunction in pressure overload. Recently, regression of cardiac hypertrophy by stem cell transplantation and subsequent macrophage polarization towards an anti-inflammatory M2 phenotype was demonstrated [12]. Besides, stem cell-based therapies are currently considered for the treatment of heart failure. Several studies, including clinical studies, have shown that the transplantation of mesenchymal stem cells (MSCs) may enhance tissue perfusion and angiogenesis and preserve or regenerate the myocardial tissue [13][14][15][16]. In this context, the prior research of our group has demonstrated that transfer of bone marrow-derived MSCs to the myocardium improves global function and remodeling in mice [17]. MSCs are able to secrete cytokines, growth factors, micro-RNAs, and exosomes which are potent for improving neovascular formation, attenuating inflammation and adverse remodeling, and inducing endogenous cardiac regeneration. Rather, the secretome of bone marrow-derived MSCs was found to favor an immunosuppressive environment [18][19][20]. At present, MSCs are still the most suitable for clinical treatment [21]. Preconditioning of MSCs highly increases their efficacy in vitro and in vivo, thus preventing loss of biological function, enhancing their biological activity, and representing a more attractive strategy for clinical use [22].
In this study, we have established an in vitro model of cardiac hypertrophy on the basis of murine iPS cell-derived cardiomyocytes (iPS-CM) and studied the hypertrophic response. We further investigated whether preconditioned MSCs might reverse hypertrophy in iPS-CM and identified VEGF as a major contributor of MSC-mediated hypertrophy regression.

Materials and Methods
2.1. Cell Culture Conditions for Murine iPS-CM and Hypertrophy Induction. Differentiated, nonproliferative murine iPS-CM were kindly provided by Dr. Kurt Pfannkuche (Institute of Neurophysiology, University of Cologne). These cells exhibit spontaneous contraction and unimpaired physiological function as evidenced by action potential generation, functional hormonal regulation, and sensitivity to ion channel blockers [23]. Cells were seeded on fibronectincoated culture dishes (2.6 μg/cm 2 ) and cultured in IMDM medium with GlutaMAX (Gibco), 20% FCS (PAN), 1% nonessential amino acids (Gibco), 0.1 mM 2-mercaptoethanol (Gibco), and 100 U/ml penicillin/streptomycin (Sigma-Aldrich). The medium was renewed every day. To induce hypertrophy, cells were starved in IMDM medium with Glu-taMAX supplemented with 1% FCS for 24 h and further treated with L-phenylephrine (PE, 100 μM, Sigma-Aldrich) for additional 24 h. In some experiments, hypertrophied iPS-CM were cultured in medium supplemented with MSC-conditioned medium (CM) or fractions from MSC-CM and PE (100 μM) in the presence or absence of polyclonal goat anti-mouse VEGF 164 antibodies (R&D Systems).

Isolation, Characterization, and Activation of Bone
Marrow-Derived MSCs. Isolation and characterization of mouse bone marrow-derived MSCs were previously reported by our group in detail. Cells were positive for classical MSC markers including Sca-1, CD49e, CD44, and CD29 and did not express CD45 and CD11b, respectively. The MSC phenotype was confirmed by the analysis of its differentiation potential into chondrocytes, adipocytes, and osteoblasts [20]. For experiments, MSCs were stimulated with 30 ng/ml recombinant murine IFN-γ (PeproTech) and 3 ng/ml recombinant murine IL-1β (PeproTech) in MSC culture medium (Pan-Biotech) supplemented with 2.5 ng/ml human basic fibroblast growth factor FGF (FGF-b, PeproTech), 100 U/ml penicillin, and 10 μg/ml streptomycin (Sigma-Aldrich) for 24 h. Preconditioned MSCs possess anti-inflammatory and immunomodulatory capacities [20] and were used for coculture experiments.
To gain MSC-CM, preconditioned MSCs were further cultured in serum-free IMDM medium with stable glutamine (Gibco) for additional 24 h. Then, supernatants were collected and stored at -80°C until further processing. To prove if MSCs retain their immunomodulatory properties upon serum starvation, nitrite levels in culture supernatants were quantified using a modified Griess reagent (Sigma-Aldrich) as already described [20]. Supernatants containing high levels of nitrite, indicating upregulated iNOS expression, were used for experiments. 2.8. Western Blot. Cells were lysed in RIPA buffer supplemented with protease and phosphatase inhibitor cocktail (Cell Signaling Technology) and sonicated, and protein concentration was quantified by using Pierce BCA Protein Assay Kit (Thermo Fisher Scientific). Protein mixture was separated by SDS-PAGE and transferred to a nitrocellulose membrane (0.2 μm). Membranes were blocked with blocking buffer and further incubated overnight at 4°C with the following antibodies: rabbit anti-mouse phospho-Akt (Ser473), NF-κB p65, pNF-κB p65 (Ser536), HIF-1α (Cell Signaling Technology), polyclonal rabbit anti-mouse Nor1 (Abcam), and polyclonal goat anti-mouse VEGF 164 antibody (R&D Systems). After incubation with a HRP-conjugated goat anti-rabbit (Dako) or horse anti-goat (Vector Labs) secondary antibody membranes were developed with UptiLight HRP Blot Chemiluminescent ECL Substrate (Uptima). To normalize protein expression, blots were stripped and reprobed for GAPDH or β-actin and Akt, respectively, using a mouse anti-GAPDH antibody (Novus Biological) and anti-mouse β-actin and rabbit anti-mouse Akt antibody (Cell Signaling Technology).
2.9. ELISA. The levels of VEGF were quantified by ELISA using mouse DuoSet kits according to the manufacturer's instructions (R&D Systems).

Ultrafiltration. Ultrafiltration units with various cutoffs
(Vivaspin 6-20, Sartorius, Germany) were first rinsed by centrifuging pure water through the membrane. The medium to be analyzed was then centrifuged at 6000 × g until the supernatant reached a tenth of its initial volume, and both the concentrated supernatant and filtrate fractions were collected in plastic tubes and used for experiments.
2.11. Statistical Analysis. Data were analyzed with GraphPad Prism 5 software. Experimental data are presented as means with standard error of the mean (SEM). Unpaired data of two groups were analyzed using the unpaired t-test. The onesample t-test was used when samples were compared with the reference control sample (set as 1). Normally distributed unpaired data of multiple groups were analyzed using oneway ANOVA with the Newman-Keuls post hoc test. A p value less that 0.05 was considered statistically significant.

Identification of Hypertrophy Markers in PE-Stimulated
iPS-CM. Cardiac hypertrophy is manifested by numerous transcriptional, biochemical, and structural transformations including increased cell size, enhanced protein synthesis, and remodeling of the actin cytoskeleton. However, an in vitro hypertrophy model based on murine iPS-CM has not been described so far. Thus, we first started to identify gene markers in iPS-CM regulated by the hypertrophic substance PE. By performing microarray analysis, we found 154 genes to be regulated (fold change in expression >2 or <-2) by PE (Table 1, Figure 1(a)). Among the 46 upregulated 3 Stem Cells International    [24], was lower when compared to Nor1/Nr4a3 expression levels (Figure 1(b)). On the basis of these results, we decided to use Nor1 as a valuable marker of hypertrophic cell growth in iPS-CM. The upregulation by PE was confirmed by real-time PCR (Figure 1(c)). In addition, Nor1 protein was significantly increased in hypertrophied iPS-CM, and increased activity of the kinase Akt, but not Erk, could be detected (Figure 1(d)). Nor1 expression showed a positive correlation with the cell surface area of iPS-CM evidenced by actin staining and immunofluorescence (Figure 1(e)).

The Secretome of MSCs Induces Hypertrophy Regression
in PE-Treated iPS-CM. Stimulation of bone marrowderived MSCs with IFN-γ and IL-1β was accompanied by increased activation of NF-κB as well as visible, but not significant, increase in intracellular HIF-1α levels ( Figure S2ab). As NF-κB represents the main transcription factor controlling iNOS expression in response to IL-1β stimulation, nitrite levels were quantified in culture supernatants to validate preconditioning of MSCs [20]. Activation of NF-κB usually results in the upregulation of genes involved in apoptosis, inflammation, and cellular adhesion, among others [25]. To further find out if preconditioned MSCs (acMSCs) influence hypertrophy in iPS-CM, PE-treated iPS-CM were cocultured with preconditioned MSCs by using a transwell system that prevents direct cell-cell interactions. The cell area of iPS-CM was significantly reduced upon coculture with preconditioned MSCs, and similar results were obtained after medium supplementation with 20% MSC-conditioned medium (Figure 3(a)). Of note, the observed effects of the MSC secretome on hypertrophied cells were found to be dose-dependent as medium supplementation with 10% conditioned medium or less did not result in hypertrophy regression ( Figure S3). These results indicate that hypertrophy regression is mediated by MSC-derived soluble factors. Notably, hypertrophy regression was associated with downregulation of Nor1 protein expression in iPS-CM (Figure 3(b)). We next aimed to identify MSC-derived molecules involved in the regression of hypertrophy. For this, we performed microarray analysis to enclose genes regulated by IFN-γ and IL-1β. A total of 2026 differentially regulated genes were identified; among them, 1037 were upregulated and 989 were downregulated (Figures 4(a)-4(c)). The top 100 list of significantly regulated genes (>2-fold; <-2-fold change) is given in Table S1. In addition to many genes encoding for chemokines, iNOS and IL-6 were ranked in the top 30 upregulated genes (Figure 4(b)), supporting MSC activation and expression of immunomodulatory factors. Gene ontology (GO) enrichment analyzed by Metascape was used to categorize the upregulated gene networks in preconditioned MSCs (Figure 4(c)). The results suggest that the identified genes are largely involved in cellular responses to interferon-beta, interferon-gamma, cytokine production, and cytokine-mediated signaling pathways. In addition, some of the genes were involved in the regulation of the apoptotic pathway, regulation of DNA-binding transcription factor activity, and myeloid leukocyte activation.
To further identify MSC-derived molecules which are potentially involved in hypertrophy regression, the supernatant of preconditioned MSCs was concentrated and fractionated by using Vivaspin centrifugal devices with a molecular mass cutoff of 30 kDa, 50 kDa, and 100 kDa. Both the concentrates and the filtrates, containing small molecules, were used to explore their effects on hypertrophied iPS-CM. Histological analysis revealed that all concentrates used for    Stem Cells International differences in protein expression were not significantly altered (Figure 4(b)).

VEGF Secreted by Preconditioned MSCs Reverses Hypertrophy in PE-Treated iPS-CM and Downregulates
Nor1 Gene Expression. VEGF secreted by bone marrowderived MSCs was recently reported to reverse hypertrophic cell growth [26,27]. Having found that molecules involved in hypertrophy reduction seem to be larger than 30 kDa, we next determined if preconditioned MSCs represent a potential source for VEGF. Significant 5.8-fold upregulation of Vegfa expression after MSC preconditioning could be detected by microarray analysis (data not shown). VEGF levels were visibly elevated in the supernatants of preconditioned MSCs 24 h as well as 48 h after cytokine stimulation. Moreover, VEGF was detected in all ultrafiltration concentrates (Figure 6(a)) which were previously found to induce hypertrophy regression in iPS-CM ( Figure 5(a)). The antihypertrophic effects of MSC-conditioned medium as well as of the medium fraction containing molecules > 30 kDa were strongly mitigated after VEGF inactivation by VEGF-neutralizing antibodies (Figure 6(b)). VEGF neutralization in MSC-conditioned medium increased Nor1 expression in iPS-CM, although significant increase in Nor1 protein levels could not be detected (Figure 6(c)).
As these results do not exclude the possibility that VEGF may derive from the hypertrophied iPS-CM themselves, we further tested if VEGF expression in iPS-CM becomes regulated by the MSC-conditioned medium. As shown in Figure 6(d), intracellular VEGF protein could be detected in both MSCs and iPS-CM; however, no effect of MSCconditioned medium on VEGF expression in iPS-CM could be observed. Additionally, neither PE treatment nor culture in the presence of MSC-conditioned medium provoked an increase in VEGF secretion by iPS-CM (Figure 6(e)), suggesting that preconditioned MSCs represent the main source for VEGF.

Discussion
In this study, we demonstrate for the first time that Nor1 represents an essential molecule for hypertrophy induction in  In this study, we established an in vitro model of cardiac hypertrophy on the basis of murine iPS-CM. Stimulation of cells with the hypertrophic substance PE increased cell size and induced significant upregulation of the nuclear receptor Nor1. However, upregulation of other key hypertrophy genes like Nppa (atrial natriuretic factor, ANF), Nppb (brain natriuretic peptide, BNP), Gja1 (connexin 43), and Myh7 (β-myo-sin heavy chain) was much weaker, indicating that Nor1 represents the most feasible hypertrophy marker in iPS-CM. Several reports suggest that Nor1 is involved in the pathophysiology of cardiac hypertrophy and heart failure, although very little is known about the underlying molecular mechanisms. Knockdown of Nor1 has been reported to prevent isoprenaline-induced hypertrophy in rat cardiomyocytes, whereas Nor1 overexpression exacerbated cardiac hypertrophy in vitro and in vivo [9,10]. In this regard, it is assumed that Nor1 upregulates the expression of hypertrophy markers and additionally promotes PARP-1 activation, resulting in functional impairment and progression of cardiac hypertrophy and heart failure [9]. Although we did not investigate the mechanism by which Nor1 regulates hypertrophy in iPS-CM, we found that silencing of Nor1   12 Stem Cells International expression by siRNA completely inhibited hypertrophy induction in a similar way as the pI3K inhibitor wortmannin. Thus, Nor1 expression as well as pI3K/Akt activation is fundamental for hypertrophic cell growth although the involvement of additional factors could be expected.
In the past decade, MSCs have been recognized to represent one of the most promising cellular therapies for cardiovascular diseases due to their ability to self-renew in a longterm manner and the capacity to differentiate into diverse specialized cell types. Additionally, there is a large body of evidence showing that MSCs secrete a variety of biologically active molecules such as growth factors, chemokines, and cytokines [28].
Preconditioning of MSCs greatly improves their proliferative, secretory, and migratory abilities and their abilities to differentiate [22]. Here, we present data demonstrating that preconditioned bone marrow-derived MSCs as well as MSC-conditioned medium promote hypertrophy regression in iPS-CM indicating that this effect is solely dependent on soluble mediators. Importantly, initial experiments demonstrated that MSCs without preconditioning do not lead to hypertrophy regression in iPS-CM after indirect coculture. MSC preconditioning by IFN-γ and IL-1β led to significant upregulation of NF-κB and consequently induced upregulation of over 1000 genes, most of them encoding for chemokines and cytokines. Hence, both IL-1β and IFN-γ were reported to promote NF-κB-dependent accumulation of HIF-1α under normoxic conditions [29,30]. As we further found that the molecular weight of these molecules seems to be larger than 30 kDa, we further concluded that most of the upregulated chemokines (molecular weight~10 kDa) can be excluded as mediators of the protective effects. In fact, many factors such as CCL-2, CCL-3, and CCL-5, whose gene expression was highly increased in preconditioned MSCs, were reported to be upregulated in patients with structural remodeling including hypertrophy and congestive heart disease [31]. Moreover, it has previously been found by Jung et al. that HIF-1α is involved in the regulation of VEGF via a pathway dependent on NF-κB [29].
The role of VEGF in cardiac hypertrophy is controversially discussed. On the one hand, VEGF was demonstrated to drive cardiac hypertrophy [32] whereby, on the other hand, VEGF-mediated hypertrophy regression was largely attributed to its ability to increase cardiac angiogenesis [33,34]. Although MSCs produce remarkable amounts of VEGF [26,35,36], VEGF secretion was significantly upregulated by MSC preconditioning. All concentrates of MSCconditioned medium containing VEGF reversed hypertrophy in iPS-CM, indicating that VEGF induces hypertrophy regression. However, it should be noted that VEGF was identified in all concentrates although concentrators with different cutoffs were used in this study. These results might be explained by the absorbance of the protein to the membrane or the plastic surface of the device. On the basis of the generated results, we cannot exclude that additional MSC-derived soluble molecules may exert antihypertrophic effects by synergistic or additive interactions with VEGF.
VEGF neutralization attenuated the protective effect of MSC-conditioned medium, and upregulation of Nor1 could be detected at the same time. Additionally, our data do not support the findings of Cai et al. who described a synergistic increase in extracellular VEGF after coculturing hypertrophied cardiomyocytes and bone marrow-derived MSCs [26]. Neither intracellular VEGF upregulation in iPS-CM nor increased VEGF secretion was detected after treatment of cells with MSC-conditioned medium or after indirect coculture, respectively. These observations suggest that preconditioned MSCs seem to represent the main source for VEGF which in turn reverses hypertrophy by suppressing Nor1 expression in cardiomyocytes.

Stem Cells International
In this regard, beside contributing to the formation of new vessels, VEGF has additionally been described to induce M1 macrophages to shift to an M2 phenotype [37] and to attenuate cardiac hypertrophy [26,38]. Thus, the beneficial effects of VEGF in vivo are expected to be manifold including direct antihypertrophic effects on cardiomyocytes, macrophage transpolarization into an anti-inflammatory phenotype [37], and angiogenesis [33]. In the light of these results, cytokine preconditioning should be considered a tool to increase the therapeutic efficacy of bone marrow-derived MSCs.
However, there are some limitations in the present study that need to be addressed. First, our study was limited by the use of adult bone marrow-derived MSCs with reduced selfrenewal and differentiation capacity, high variations between different batches, and progressive cell senescence [39][40][41]. Second, the secretome of MSCs also appears to vary significantly, depending on the age of the host, health status, and

16
Stem Cells International niches where the cells reside. Besides, NF-κB was previously reported to be critical for the secretion of cytokines and growth factors by MSCs, and several studies proposed that genetic modification of MSCs may improve NK-κB signaling and their paracrine activities [42,43]. Although we provided evidence for NF-κB activation in preconditioned MSCs, the onset of NF-κB activation may markedly differ between different MSC preparations. Further, we did not consider the complexity of the MSC secretome including microvesicles, microRNAs, exosomes, and mitochondrial transfer among others [44]. The limitations may be largely omitted by the development of more attractive strategies on the basis of iPS cell-derived MSCs. These cells were demonstrated to have higher proliferative capacity, higher telomerase activity, and lower cell senescence than bone marrow-derived MSCs allowing generation of highly homogenous populations without loss of functional quality [39,45]. Altogether, the results of our study indicate that conditioned medium from IFN-γ-and IL-1β-stimulated MSCs may represent an appropriate therapeutic approach for the treatment of LVH. Nevertheless, further studies are needed to characterize preconditioned iPS cell-derived MSCs and to evaluate the efficacy and safety of secretome components for the treatment of cardiac hypertrophy in vivo.

Conclusions
Preconditioning of bone marrow-derived MSCs with IL-1β and IFN-γ strongly improves their paracrine actions. Moreover, VEGF secreted by MSCs reverses Nor1-dependent hypertrophy in iPS-CM. Altogether, delivery of MSCconditioned medium may represent a promising cell-free therapeutic approach for the treatment of LVH although in vivo efficacy requires further investigation.

Data Availability
The datasets supporting the conclusions of this article are available from the corresponding author upon reasonable request. Stem Cells International