OUP user menu

Peroxisome proliferator-activated receptor β stimulation induces rapid cardiac growth and angiogenesis via direct activation of calcineurin

Nicole Wagner, Chantal Jehl-Piétri, Pascal Lopez, Joseph Murdaca, Christian Giordano, Chantal Schwartz, Pierre Gounon, Stéphane N. Hatem, Paul Grimaldi, Kay-Dietrich Wagner
DOI: http://dx.doi.org/10.1093/cvr/cvp106 61-71 First published online: 7 April 2009

Abstract

Aims Peroxisome proliferator-activated receptors (PPARs) are ligand-activated transcription factors. PPARβ agonists were suggested as potential drugs for the treatment of metabolic syndrome, but effects of PPARβ activation on cardiac growth and vascularization are unknown. Thus, we investigated the consequences of pharmacological PPARβ activation on the heart and the underlying molecular mechanisms.

Methods and results Male C57/Bl6 mice were injected with the specific PPARβ agonists GW0742 or GW501516, or vehicle. Cardiomyocyte size and vascularisation were determined at different time points. Expression differences were investigated by quantitative reverse transcriptase–polymerase chain reaction and western blotting. In addition, the effects of PPARβ stimulation were compared with hearts of mice undergoing long-term voluntary exercise or pharmacological PPARα activation. Five hours after GW0742 injection, we detected an enhanced angiogenesis compared with vehicle-injected controls. After 24 h, the heart-to-body weight ratios were higher in mice injected with either GW0742 or GW501516 vs. controls. The increased heart size was due to cardiomyocyte enlargement. No signs of pathological cardiac hypertrophy (i.e. apoptosis, fibrosis, or deteriorated cardiac function) could be detected. The effects are mediated via calcineurin A (CnA) activation as: (i) CnA was upregulated, (ii) GW0742 administration or co-transfection of PPARβ significantly stimulated the activity of the CnA promoter, (iii) PPARβ protein bound directly to the CnA promoter, (iv) the CnA target genes NFATc3, Hif-1α, and Cdk 9 were upregulated in response to PPARβ stimulation, and (v) the inhibition of CnA activity by cyclosporine A abolished the hypertrophic and angiogenic responses to PPARβ stimulation.

Conclusion Our data suggest PPARβ pharmacological activation as a novel approach to increase cardiac vascularization and cardiac muscle mass.

KEYWORDS
  • Angiogenesis
  • Calcineurin
  • Gene transcription
  • Hypertrophy
  • Peroxisome proliferator-activated receptor

1. Introduction

Peroxisome proliferator-activated receptors (PPARs) are ligand-activated transcription factors.1 They have come into focus as targets for the treatment of glucose and lipid abnormalities.24 PPARs are members of the nuclear receptor superfamily. They exist in three isoforms: PPARα, PPARβ (formerly PPARδ), and PPARγ. All PPARs form heterodimers with retinoid X receptors, and upon ligand binding, use the basal transcriptional machinery to augment gene expression.5 Via this mechanism, PPARs regulate lipoprotein and glucose metabolism and modulate adipose tissue, liver, endothelium, and skeletal muscle phenotypes. Synthetic PPARα and PPARγ agonists are already in clinical use for the treatment of hyperlipidaemia and type 2 diabetes, respectively (reviewed in 6).

More recently, PPARβ activation came into focus as an interesting novel approach for the treatment of metabolic syndrome and associated cardiovascular diseases.7,8 Using pharmacological stimulation and transgenic animal models, it has been shown that PPARβ activation increases oxidative metabolism, insulin responsiveness, fatty acid burning,9,10 and induces a fibre-type transition in skeletal muscle.11 Furthermore, PPARβ activation reduces fat mass and normalizes adipokine secretion and insulin responsiveness in adipose tissue.12 Thus, PPARβ activation mimics the beneficial effects of physical exercise. For the heart, it has been shown that PPARβ activation stimulates fatty acid oxidation,13 has anti-inflammatory effects on the vascular system,14 and reduces ischaemia/reperfusion-induced injury in transgenic mice.15 Whether pharmacological specific PPARβ activation acts on the heart in vivo and how these potential effects are mediated in molecular terms remain to be elucidated.

We show here that specific PPARβ stimulation rapidly induces cardiac angiogenesis and myocyte growth and that these effects are mediated at least in part by the direct transcriptional activation of calcineurin.

2. Methods

2.1 Animals

The investigation conforms with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996). Experiments were approved by the University of Nice Animal Care and Use Committee.

Ten-week-old male C57/BL6J (Janvier, France) mice were subcutaneously injected with GW0742 (Glaxo Smith Kline) or GW501516 (Alexis Biochemicals) dissolved in DMSO at 1 mg/kg once daily (2 p.m.) according to the duration of treatment (24, 48, 96 h). Controls received DMSO injections. Cyclosporine A (CsA) (Sigma) dissolved in DMSO was administered daily (9 a.m.) by subcutaneous injection in a dose of 20 mg/kg. Clofibric acid (Sigma) was injected one time a day (2 p.m.) in a dose of 300 mg/kg. Propranolol (Sigma) dissolved in PBS was injected subcutaneously in a dose of 0.5 mg/kg. Physical training was performed by keeping the animals in cages for voluntary physical exercise (Techniplast, Italy), equipped with a running wheel (25 cm diameter), for 5 weeks.

2.2 Histological and immunohistochemical analysis

Paraffin sections (5 µm) were stained with haematoxylin and eosin. Mean cardiomyocyte diameters were determined by a blinded investigator measuring at least 30 cells per heart of four hearts for each condition. Collagen-specific Sirius Red (Waldeck Gmbh, Münster, Germany) staining for the measurement of interstitial fibrosis was performed according to manufacturer’s instructions. Antibodies, dilutions, and kits used for immunohistochemistry are described in Supplementary material online. Slides were viewed under an epifluorescence microscope connected to a digital camera with the Spot software (Universal Imaging Corp.).

2.3 Terminal deoxynucleotidyl transferase dUTP nick end labelling of apoptotic cells

Apoptotic cells were detected by terminal deoxynucleotidyl transferase dUTP nick end labelling (TUNEL) staining in paraffin-embedded heart sections using the In situ Cell Death Detection Kit (Roche Molecular Biochemicals) as described previously.16 TUNEL-positive cells/field were counted. Five tissue sections were analysed from three different controls and three hearts each at different time points of GW0742 treatment or physical exercise for 5 weeks. Heart sections of mice with hereditary cardiomyopathy due to the injection of miR 1 in the oocyte16 served as a positive control.

2.4 Sodium dodecyl sulfate–polyacrylamide gel electrophoresis and western blot

Total heart lysates from the mice treated with GW0742 or vehicle alone (DMSO) were prepared, electrophoresed, and blotted as described.16 Subcellular fractionation of heart tissues was performed using the Qproteome™ Cell Compartment Kit (Quiagen) according to manufacturer’s instructions. Antibodies for immunodetection are mentioned in Supplementary material online.

2.5 Electron microscopy

For ultrastructural analysis of cardiac muscle, small pieces were fixed in 1.6% glutaraldehyde in 0.1 M phosphate buffer (pH 7.5), then washed with 0.1 M cacodylate buffer (pH 7.5), and post-fixed with 1% osmium tetroxide in the same buffer. Samples were embedded in epoxy resin. Thin sections were contrasted with uranyl acetate and lead citrate and observed with a Philips CM12 electron microscope fitted with a CCD camera (Morada, Olympus SIS).

2.6 Cell culture and transient transfection experiments

Neonatal mouse cardiomyocytes were isolated as described,17 with minor modifications mentioned in Supplementary material online. Adult cardiomyocytes were isolated using the Adumyts Kit (Cellutron) according to manufacturers instructions. Adult cardiomyocytes were maintained for 24 h in serum-free medium in the presence of 200 nM GW0742 or vehicle. To investigate the effect of PPARβ expression on calcineurin promoter activity, a 2.3 kb fragment of the calcineurin promoter in the pGl3 basic luciferase expression vector18 was co-transfected with PPARβ and RxR expression constructs.19 Alternatively, the calcineurin Aβ (CnAβ) promoter construct was co-transfected only with the β-galactosidase reporter plasmid and the cells cultured for 48 h in the presence of 100 nM GW0742 or vehicle in the presence or absence of a dominant-negative PPARβ isoform.19 Transient co-transfections and assays for luciferase and β-galactosidase activity are described elsewhere.20 The putative PPAR-responsive element (PPRE) (position –1249 to –1224 bp of the published promoter) was deleted from the CnAβ promoter construct using the QuikChange II site-directed mutagenesis kit (Stratagene). Details are provided in Supplementary material online.

2.7 Chromatin immunoprecipitation assay

Chromatin immunoprecipitation (ChIP) assay was performed on neonatal cardiomyocytes using manufacturers instructions (Upstate). Antibodies and primer sequences are described in Supplementary material online.

2.8 Electrophoretic mobility shift assays

The putative PPRE from the CnAβ promoter contained the following sequence: 5′-TTGGCCCCTGAACCATTCAACACTGC-3′. PPRE from the acyl-CoA oxidase (Aco) gene (5′-CCCGAACGTGACCTTTGTCCTGGTCC-3′) served as positive control. Annealed oligonucleotides were 32P-end-labelled in a T4 polynucleotide kinase reaction (New England Biolabs). PPARβ and RxRα proteins were generated from full-length cDNAs in pSG5 vector (Stratagene) using the coupled TNT in vitro-transcription–translation system (Promega). For supershift assays, the identical antibodies as for the ChIP experiments were used. Details are described in Supplementary material online.

2.9 Quantitative reverse transcriptase–polymerase chain reaction

Reverse transcriptase–polymerase chain reaction (RT–PCR) was performed with 2 µg of total RNA as described.16 Primers for quantitative RT–PCR are listed in Supplementary material online.

2.10 Echocardiography

Echocardiography was performed on lightly anaesthetized (1% isoflurane in oxygen) mice, using a GE Medical System VIVI7 device equipped with an 8 to 14 MHz phased array transducer. The left ventricle (LV) was imaged in parasternal long-axis view to obtain measurements of the LV in time-motion mode. The following measurements were performed: LV end-diastolic diameter (mm/g), interventricular septum diastolic thickness (mm/g), LV posterior wall diastolic thickness (mm/g), and shortening fraction (%). Aortic outflow was estimated by measuring the pulse-wave Doppler velocity in the LV at the level of the aortic annulus and using the following equation: (πd2/4) × VTI × HR, d is the diameter of the aortic annulus, VTI is the velocity time integral, HR is the heart rate.

2.11 Statistics

Data are expressed as means ± SEM. ANOVA with the Bonferroni test as post hoc test was used vs. control. Differences between two groups were tested using the Mann–Whitney test for non-parametric samples. A P-value <0.05 was considered statistically significant.

3. Results

3.1 PPARβ activation results in a rapid induction of cardiac growth without pathological aspects

Wild-type mice (C57/BL6J strain) were treated by subcutaneous injection with the specific PPARβ agonist GW0742. The animals showed, 24 h after the injection of the specific PPARβ agonist, an increased heart size when compared with vehicle-injected controls (Figure 1A). Heart weights were significantly higher in mice treated with GW0742 for 24 and 48 h (Figure 1B), whereas body weights remained unchanged (Figure 1C). Histological analysis revealed a marked increase in cardiomyocyte diameters 24 h after GW0742 injection (Figure 1DF) with a shift of the cardiomyocyte diameter distribution to higher values. This was even more pronounced 48 h after GW0742 treatment (Figure 1F). The cardiac growth-promoting effect of PPARβ stimulation was evident regardless of whether the animals were treated for 24, 48 (Figure 1A, B, D, and E) or 96 h (Figure 2AC) with GW0742. Five hours after GW0742 injection, this effect was not yet detectable (Figure 1B, D, E, and F). Echocardiographic examination in anaesthetized animals confirmed the growth-promoting effect, which is indicated by an increased septum thickness after GW0742 treatment (Figure 1Ga). Interestingly, no functional disadvantages could be detected after PPARβ stimulation, i.e. aortic outflow, the ventricular shortening fraction, and left ventricular end-diastolic diameters were completely normal (Figure 1Gb, c, and d). Cardiac hypertrophy was still present 10 days after the end of a 48 h treatment, but heart-to-body weight ratios and cardiomyocyte diameters returned to control values 4 weeks after the end of treatment (Figure 2AC). The degree of cardiac hypertrophy induced by short-term PPARβ stimulation was comparable to that achieved by long-term voluntary exercise (training) for 5 weeks (Figure 2AC). Also on the ultrastructural level, short-term PPARβ stimulation resembled the long-term training effect with the addition of sarcomeres, dense mitochondrial cristae, normal sarcomere lengths, and no signs of cardiomyocyte damage or fibrosis (Figure 2D). As qualitative measure for collagen accumulation, we performed Sirius Red staining of cardiac sections. Sirius Red staining showed a decrease in collagen accumulation after long-term training or PPARβ stimulation (Figure 2E and F). TUNEL labelling as a marker for DNA damage or repair, which is increased in human and mouse pathological hypertrophy,16,21 was not evident after PPARβ activation or long-term moderate exercise (Figure 2G and H). Our recently established mouse model for cardiomyopathy,16 which is due to an miR-1-induced paramutation and characterized by ultrastructural signs of hypertrophic cardiomyopathy and a large number of apoptotic cardiomyocytes, served as a positive control (Figure 2G and He).

Figure 1

Peroxisome proliferator-activated receptor β (PPARβ) activation stimulates cardiac growth. (A) Hypertrophic development of the ventricular walls 24 h after GW0742 injection (b) compared with a vehicle-injected control (a). (B) Mean values ± SEM of heart-to-body weight ratios. (C) Mean values ± SEM of body weights of control mice and animals injected for different time points with GW0742 (n = 4 each). (D) High magnifications of heart sections. (E) Mean values ± SEM of cardiomyocyte diameters. ***P < 0.001. (F) Distribution of cardiomyocyte diameters. Note the progressive shift of cell sizes to larger diameters with increasing time of PPARβ stimulation. (G) Echocardiographic examination of control mice and animals injected with GW0742 (n = 7 each) confirmed the increased septum thickness (a), but revealed no functional disadvantage as indicated by the normal aortic outflow (b), unchanged shortening fraction (c), and preserved left ventricular end-diastolic diameters (LVEDD, d). (H) Haematoxylin–eosin staining of kidney and liver sections of control mice and animals 24 h after GW0742 injection shows no obvious abnormalities in these organs. Scale bars: 50 µm.

Figure 2

Peroxisome proliferator-activated receptor β (PPARβ)-induced cardiac growth is reversible and resembles exercise-induced cardiac hypertrophy. (A) Mean values ± SEM of heart-to-body weight ratios of control mice, animals injected for different time points with GW0742, and mice performing 5 weeks of voluntary exercise (n = 4 each). Note that the heart-to-body weight ratio was still increased 10 days after the end of the treatment, but returned to control values 4 weeks after the end of the treatment. The degree of cardiac growth due to GW0742 injection was comparable to physiological cardiac hypertrophy induced by 5 weeks of voluntary exercise. **P < 0.01. (B) Mean values ± SEM of cardiomyocyte diameters. *P < 0.05 and **P < 0.01. (C) Distribution of cardiomyocyte diameters from mice after GW0742 injection, voluntary exercise training, and respective controls. (D) Electron microscopic examination of the hearts after GW0742 injection (b and e) revealed no signs of pathological cardiac hypertrophy compared with controls (a and d), but addition of sarcomeres, dense mitochondrial cristae, and normal sarcomere lengths, which was comparable to cardiac hypertrophy induced by voluntary exercise for 5 weeks (c and f). Scale bars: 5 µm in a, b and c; 2 µm in d, e and f. (E) Sirius Red staining was used to detect collagen fibres (red). (F) Quantification of Sirius Red staining. Note that PPARβ activation and exercise training slightly reduced the collagen content. (G) Quantification of the number of TUNEL-positive cells per field as a marker of apoptosis. (H) Representative photomicrographs for TUNEL staining (DAB substrate, brown colour). Note that TUNEL staining revealed no differences between mice after GW0742 injection, voluntary exercise training, and controls. A section from a mouse model with cardiomyopathy16 served as positive control. Scale bars: 50 µm.

The growth-promoting effect seems to be specific for the heart, as kidney weights relative to body weights showed no significant differences (data not shown). Furthermore, histological analysis of different organs (i.e. the kidney and liver) showed no obvious abnormalities (Figure 1H). No increase in cell size was observed in the kidney or liver.

3.2 PPARβ activation induces rapid cardiac angiogenesis

Next, we analysed a potential cardiac angiogenic response to PPARβ stimulation. The expression of hypoxia-inducible factor 1α (Hif-1α), a well-known inducer of angiogenesis,22 was already increased after 5 h and more pronounced after 24 h (Figure 3A and B). PECAM-1 (CD31) as a marker for newly forming vessels was augmented within the same time frame after PPARβ stimulation (Figure 3B and C). This pro-angiogenic effect of PPARβ was also demonstrated by a higher number of BrdU-positive cells in the cardiac vasculature of GW0742-treated animals, whereas no difference in the number of BrdU-positive cardiomyocytes or fibroblasts was observed (data not shown). The quantification of isolectin B4-positive vessels confirmed an increased capillary density after 5 and 24 h, indicating that angiogenesis precedes cardiomyocyte growth in response to PPARβ activation (Figure 3D). Ten days after the end of 48 h GW0742 treatment, the number of isolectin B4-positive vessels was not significantly different from the controls anymore (data not shown). Although angiogenesis preceded cardiomyocyte hypertrophy after pharmacological PPARβ stimulation, it seems not to be a prerequisite for cardiomyocyte growth, because GW0742 increased cell diameters and lengths also in isolated adult mouse cardiomyocytes after 24 h of treatment (see Supplementary material online, Figure S1).

Figure 3

Pharmacological peroxisome proliferator-activated receptor β (PPARβ) stimulation induces cardiac angiogenesis rapidly. (A) Expression of hypoxia-inducible factor 1α (Hif-1α), a known regulator of angiogenesis,22 was increased 5 and 24 h after GW0742 injection on an immunohistological level (b and c). (B) Western blot analysis of Hif-1α and PECAM-1 (CD31) as a marker of proliferating endothelial cells. α-Tubulin served as an internal standard. (C) Expression of PECAM-1 (CD31) was stimulated after 5 and 24 h in response to PPARβ activation on the immunohistological level (b and c), which was quantified by counting PECAM-1-positive vessels per field (d). (D) Isolectin B4 staining was used to identify all vascular cells (a–c). Morphometric analysis revealed a higher number of vessels in response to PPARβ activation (d). Scale bars: 50 µm. ***P < 0.001.

The observed cardiac growth and angiogenesis seem to be specific for PPARβ activation, as treatment with GW501516, another PPARβ agonist, induced exactly the same cardiac phenotype as GW0742 after 24 h of treatment (see Supplementary material online, Figure S2). In contrast, the administration of the PPARα agonist clofibrate, a derivative of fenofibrate, which is in wide clinical use for the treatment of metabolic syndrome,2 led to neither enhanced cardiac growth nor angiogenesis in a similar short-time frame (see Supplementary material online, Figure S3).

To investigate whether adrenergic signalling might be involved in GW0742-induced cardiac growth, we performed additional GW0742 treatment experiments in mice in the presence of β-adrenergic blockade using propranolol.23 Co-administration of propranolol affected neither the cardiac growth promoting nor the pro-angiogenic responses to PPARβ stimulation (see Supplementary material online, Figure S4).

3.3 PPARβ-induced cardiac growth is accompanied by an increased calcineurin expression

As molecular markers and inducers of cardiac growth, expression of α- and β-myosin heavy chains (Myh6 and Myh7, respectively) was only transiently increased, angiotensinogen expression was not significantly changed, whereas atrial natriuretic factor and calcineurin became upregulated after 5 h and remained elevated for 96 h of GW0742 treatment (Figure 4A). Pharmacological PPARβ stimulation slightly increased calcineurin protein expression after 5 h, which became more pronounced after 24 and 48 h, whereas PPARβ and tubulin protein expression were not altered (Figure 4B). We were not able to measure directly whether increased calcineurin protein expression was preceded by calcineurin activation. Calcineurin activation results in dephosphorylation and subsequent nuclear translocation of NFAT (for review, see 24). The enhanced nuclear NFATc3 localization 5 h after GW0742 injection (Figure 4C) suggests that calcineurin activation happens much earlier. Using subcellular protein fractionation, increased nuclear NFATc3 localization could be determined quantitatively 5 and 24 h after GW0742 injection (Figure 4D). In agreement with a described regulation of HIF-1 and Cdk9 by calcineurin,25,26 we detected enhanced Hif-1α (Figures 3A and 4D) and Cdk9 protein (Figure 4E) expression after 5 h of PPARβ stimulation, which might contribute to enhanced angiogenesis and cardiac growth.

Figure 4

Markers of cardiac growth and angiogenesis are upregulated in response to peroxisome proliferator-activated receptor β (PPARβ) activation. (A) Quantitative reverse transcriptase–polymerase chain reaction analysis of the levels of α-myosin heavy chain (α-MHC, Myh6), β-MHC (Myh7), angiotensinogen, atrial natriuretric factor (ANF), and calcineurin in the hearts at different time points after PPARβ stimulation. *P < 0.05, **P < 0.01, ***P < 0.001. (B) Western blot analysis of PPARβ and calcineurin expression at different time points after PPARβ stimulation. Note the strong increase in calcineurin expression while PPARβ remains unchanged. α-Tubulin served as an internal standard. (C) Immunohistochemical localization of NFATc3, which becomes nuclearized 5 h after PPARβ activation. (D) Subcellular fractionation in cytoplasmic (C), nuclear (N), and insoluble (I) proteins and subsequent western blot analysis. Note the increased nuclear expression of NFATc3 and hypoxia-inducible factor 1α (Hif-1α) in response to PPARβ activation. CREB (nuclear), Rho GDIα (cytoplasmic), and acetylated histone H3 (H3, chromatin associated) antibodies were used to evaluate the purity of each fraction. (E) Immunohistological and western blot analysis of Cdk9 expression in response to PPARβ activation. Note the increased Cdk9 expression after 5 h. Scale bars: 50 µm.

To exclude the possibility that GW0742 directly affects the mRNA or protein stability of calcineurin and Hif-1, we cultured neonatal cardiomyocytes and inhibited transcription or translation using actinomycin D and cycloheximide, respectively. GW0742 treatment showed no significant effects on calcineurin and Hif-1α mRNA stability (see Supplementary material online, Figure S5). Calcineurin protein stability was also not significantly altered by GW0742 in neonatal cardiomyocytes in vitro. Hif-1α protein was barely detectable in the cultured cells and disappeared completely upon cycloheximide treatment, independently on the presence of GW0742 in the culture medium (see Supplementary material online, Figure S5).

3.4 PPARβ directly activates calcineurin

To test whether PPARβ could directly regulate calcineurin expression, we performed immunohistological double-labelling and showed that both proteins are co-expressed in adult mouse hearts in vivo (Figure 5Aa and b) and in isolated neonatal cardiomyocyte cultures in vitro (Figure 5Ac and d). The addition of 200 nM GW0742 significantly stimulated the activity of the published CnAβ promoter18 in neonatal cardiomyocyte cultures (Figure 5B). Retroviral transduction of neonatal cardiomyocyte cultures with a dominant-negative PPARβ isoform completely abolished the stimulatory effect of GW0742 on CnAβ promoter activity (Figure 5C). Transient co-transfection with a PPARβ expression construct enhanced CnAβ promoter activity more than six-fold (Figure 5D). A sequence region containing a predicted PPRE with 70% identity to the FAT-PPRE27 was identified in the CnAβ promoter. Using ChIP, we show that PPARβ protein associates with this region, whereas no interaction could be detected in the 3′-UTR of CnA. An antibody against acetylated histone H3 was used to check for nucleosome integrity (Figure 5E). Electrophoretic mobility shift assays confirmed binding of PPARβ to a 26 bp oligonucleotide from the CnAβ promoter. Incubation with different PPARβ antibodies supershifted the retardation bands. The published PPARβ-binding oligonucleotide from the Aco gene28 served as positive control (Figure 5F). The deletion of the identified binding site from the CnAβ promoter construct completely abolished activation by the co-transfection of the PPARβ expression construct (Figure 5G).

Figure 5

Peroxisome proliferator-activated receptor β (PPARβ) directly transcriptionally activates calcineurin. (A) Immunohistological localization of PPARβ (red) and calcineurin (green) in adult mouse hearts (a and b) in vivo and in isolated neonatal cardiomyocytes in vitro (c and d). Note the nuclear PPARβ and cytoplasmic calcineurin expression in the same cells. Scale bars: 50 µm. (B) Transient transfections of the calcineurin Aβ promoter in a luciferase vector in the presence of 100 nM GW0742 or vehicle in primary neonatal mouse cardiomyocytes. Luciferase activities were normalized for the activity of co-transfected β-galactosidase (n = 7, *P < 0.05). (C) Transient transfection of the calcineurin Aβ promoter construct in the presence of 200 nM GW0742 or vehicle with additional retroviral transduction of primary neonatal cardiomyocytes with a dominant-negative isoform of PPARβ (PPARβDN). Note that in the presence of PPARβDN, GW0742 did not activate the calcineurin promoter, ensuring the specificity of the GW0742 effect for PPARβ (n = 4, *P < 0.05). (D) Transient co-transfection of the calcineurin Aβ promoter construct together with a PPARβ expression construct in neonatal cardiomyocytes (n = 13, *P < 0.05). (E) Chromatin-immunoprecipitation to analyse PPARβ protein interaction with calcineurin Aβ (CnAβ) regulatory sequences. PPARβ protein binds to the CnAβ promoter (upper panel), but not to the 3′-untranslated region (UTR) of CnAβ (lower panel). Input DNA and immunoprecipitates obtained with acetylated histone 3 antibody served as positive controls. Negative controls were performed with normal serum instead of specific antibodies and with DNase-free water for the PCR. (F) Electrophoretic mobility shift assay demonstrating the binding of the PPARβ/RxRα complex to the predicted consensus element of the CnAβ promoter (arrowhead). Pre-incubation of the reactions with different PPARβ antibodies supershifted the complexes (arrow). Note that the PPARβ antibody from rabbit tended to produce negative supershifts as has been described.28 An oligonucleotide from the acyl-CoA oxidase gene served as positive control.28 (G) Transient co-transfection of the wild-type calcineurin Aβ promoter construct or a construct with deletion of the identified 26 bp consensus motif together with PPARβ expression constructs in neonatal cardiomyocytes (n = 4, *P < 0.05). Note that the 26 bp deletion abolished transactivation by PPARβ.

3.5 The calcineurin inhibitor cyclosporine A abolishes PPARβ-induced cardiac growth and vascularization

Finally, to investigate the in vivo relevance of the PPARβ–calcineurin interaction for cardiac growth and angiogenesis, we blocked the calcineurin activity in mice under pharmacological PPARβ stimulation by parallel injection of CsA. Blocking calcineurin activity reduced the relative heart-to-body weight to control values, blocked the increase in cardiac cell diameters in response to GW0742 injection, and normalized the cell size distribution (Figure 6A). Furthermore, NFATc3 nuclear translocation was blocked by CsA (Figure 6B), and the angiogenic response to PPARβ stimulation inhibited as indicated by the lower number of isolectin B4-positive vessels (Figure 6C).

Figure 6

Blocking the calcineurin activity by cyclosporine A (CsA) inhibits both peroxisome proliferator-activated receptor β (PPARβ)-induced cardiac growth and angiogenesis in vivo. (A) Mean values ± SEM of heart-to-body weight ratios of mice injected with GW0742 or a combination of GW0742 and CsA (a), mean values ± SEM of cardiomyocyte diameters (b), and distribution of cardiomyocyte diameters (c). *P < 0.05, **P < 0.01. (B) Immunohistological localization of NFATc3 in hearts of mice injected with GW0742 (a and b) or a combination of GW0742 and CsA (c and d). Note that CsA inhibits PPARβ-induced nuclear translocation of NFATc3. (C) Immunohistological detection of isolectin B4 as a marker for vascular cells. Note that CsA inhibits PPARβ-induced angiogenesis in the heart. Scale bars: 50 µm.

4. Discussion

The major findings of the present study are the rapid induction of cardiac growth and angiogenesis in response to pharmacological PPARβ activation, which are mediated via the direct transcriptional stimulation of calcineurin. The growth-promoting effect seems to be restricted to the heart, as we did not observe any alterations in the liver or kidney. Recently, we reported a fibre remodelling in skeletal muscle in response to PPARβ stimulation, which resulted in an increase in the number of oxidative myofibres and hyperplasia.11 Interestingly, this hyperplasia in tibialis anterior muscle was accompanied by a reduced size of the myofibres, which is the opposite effect evoked in cardiac muscle cells. The morphological enlargement of cardiomyocytes was accompanied by increased expression of molecular markers of cardiac growth, i.e. α- and β-myosin heavy chains, calcineurin, and ANF, whereas angiotensinogen was unchanged. Whether the cardiomyocyte enlargement corresponds to pathological or physiological hypertrophy was not possible to judge from the increase in the earlier-mentioned markers, as the calcineurin-NFAT pathway has been described to be important for normal cardiac growth29,30 and pathological hypertrophy (reviewed in 31). Also, ANF is not only elevated in pathological hypertrophy, but also upregulated in physiological cardiac growth in response to moderate non-exhausting exercise.32 Importantly, echocardiographic examination revealed no functional abnormalities in response to PPARβ activation. Morphological and ultrastructural comparison with hearts from mice undergoing long-term voluntary exercise training showed a high degree of similarity to the effects of short-term PPARβ activation. On the ultrastructural level, PPARβ stimulation resembled physical exercise, i.e. addition of sarcomeres, dense mitochondrial cristae, no signs of cardiomyocyte damage or fibrosis were observed. The slight decrease in collagen content after long-term training or PPARβ stimulation is in agreement with the inhibition of collagen expression in response to PPARβ activation.33 TUNEL labelling as a marker for DNA damage or repair was not evident after PPARβ stimulation or long-term moderate exercise. This corresponds to a previous report showing the prevention of stress-induced apoptosis in cardiomyoblasts upon PPARβ activation.34 The notion of a ‘physiological’ hypertrophy in response to PPARβ stimulation is supported by recent data providing evidence that agonists for PPARβ and AMP-activated protein kinase (AMPK) have exercise mimetic effects in skeletal muscle.35 In this study, long-term pharmacological PPARβ activation in combination with exercise training increased running time of mice by 68% and running distance by 70% over vehicle-treated trained mice. Unfortunately, potential effects on cardiac performance or morphology were not investigated. On the basis of our data, it is tempting to speculate that these modifications occur much more rapidly than within the 5 weeks interval tested in this study.

As ‘physiological’ hypertrophy is accompanied by increased vascularization whereas pathological hypertrophy is, on the contrary, characterized by a relative lack of capillaries,36 we analysed a potential cardiac angiogenic response to PPARβ stimulation. In agreement with the reported endothelial cell proliferation and angiogenesis in vitro,37 we observed an increased capillary density in the hearts after GW0742 and GW501516 injection. The angiogenic response to PPARβ stimulation preceded cardiomyocyte growth, but does not seem to be necessary for cardiomyocyte enlargement, as also in adult cardiomyocytes in vitro, PPARβ stimulation-induced hypertrophy. The fast angiogenic response might contribute to the protection against ischaemia/reperfusion injury detected in PPARβ transgenic mice15 and, to our knowledge, represents the most rapid pharmacological approach to induce cardiac vascularization. Whether the PPARβ-induced angiogenesis is sufficient to protect against ischaemia/reperfusion injury and might improve cardiac function after myocardial infarction is subject of further studies. However, the reversibility of cardiac growth and vascularization 4 weeks after the end of GW0742 treatment points to a potential pharmacological use.

The cardiac growth promoting and pro-angiogenic effects seem to be specific for PPARβ, as GW0742 and GW501516 had similar effects, whereas the PPARα agonist clofibrate did not evoke alterations in a similar time frame. Furthermore, the notion of physiological cardiac growth in response to specific PPARβ activation is in agreement with a recent report showing that cardiac-specific PPARβ knockout mice develop cardiomyopathy.38

As the most likely molecular mechanism for GW0742- and GW501516-induced cardiac growth and angiogenesis, we identified the calcineurin-NFAT pathway. Several lines of evidence suggest that calcineurin represents a direct relevant target of PPARβ in cardiac growth and angiogenesis. PPARβ activation resulted in the upregulation of calcineurin on the RNA and protein level. Both proteins are co-expressed in adult mouse hearts in vivo and in neonatal cardiomyocyte cultures. Furthermore, the calcineurin target genes NFATc3, Hif-1α, and Cdk9 were stimulated in response to PPARβ activation. We cannot exclude the possibility that the rapid induction of calcineurin and Hif-1 in response to PPARβ stimulation in adult animals in vivo involves mRNA or protein stabilization, but the results obtained in neonatal cardiomyocytes in vitro do not support this hypothesis. The activation of the calcineurin promoter in response to pharmacological PPARβ stimulation and co-transfection with a PPARβ expression construct suggests transcriptional regulation. The effect of GW0742 on calcineurin promoter activity is specific for PPARβ, as co-transduction with a dominant-negative PPARβ isoform abolished the activation of the promoter construct. In addition, in ChIP experiments, we could detect the binding of PPARβ protein to a promoter sequence, but not to the 3′-UTR of calcineurin. This promoter sequence contained a PPRE with high homology to the published FAT-binding site.27 In electrophoretic mobility shift assays, we observed the binding of PPARβ protein to this predicted PPRE. The deletion of the PPRE from the calcineurin promoter abolished activation by PPARβ in transient co-transfection experiments. Finally, although we cannot exclude that CsA might, besides calcineurin, target other signalling pathways, inhibition of the cardiac growth-promoting and pro-angiogenic responses to PPARβ activation in vivo by CsA underlines the importance of the calcineurin-NFAT pathway for the observed phenotype.

Taken together, we provide evidence that pharmacological PPARβ stimulation rapidly induces cardiac angiogenesis and cardiomyocyte growth, which is at least in part mediated via the direct transactivation of calcineurin. Echocardiographic examination revealed no functional abnormalities in response to PPARβ activation. The short-term PPARβ stimulation was able to mimic in the heart the beneficial effects of long-term voluntary physical exercise. Finally, as PPARβ activation is known to reduce the size of experimental myocardial infarctions, which is most likely due to our observed rapid induction of angiogenesis, it could have therapeutic potential in treating chronic ischaemic heart disease and myocardial infarction.

Funding

The work was supported by grants to K.-D.W. from Fondation Recherche Medicale, Fondation Cœur et Artères, and Association pour la Recherche sur le Cancer, to P.G. from Agence National de la Recherche (ANR-05-PCOD-012) and Fondation Cœur et Artères. N.W. was the recipient of a fellowship from the Fondation de France.

Acknowledgements

GW0742 was a generous gift from T.M. Willson (GlaxoSmithKline). The CnAβ promoter was kindly provided by J. Molkentin. The CREB antibody was a gift of M. Montminy. We thank M. Aupetit, M. Radjkhumar, F. Millot, J. Paput, G. Manfroni, and G. Visciano for technical assistance.

Conflict of interest: none declared.

References

View Abstract