OUP user menu

Bone morphogenetic protein-2 acts upstream of myocyte-specific enhancer factor 2a to control embryonic cardiac contractility

Yue-Xiang Wang, Lin-Xi Qian, Dong Liu, Ling-Ling Yao, Qiu Jiang, Zhang Yu, Yong-Hao Gui, Tao P. Zhong, Hou-Yan Song
DOI: http://dx.doi.org/10.1016/j.cardiores.2007.02.007 290-303 First published online: 1 May 2007

Abstract

Objective: Cardiac contractility is regulated tightly as an extrinsic and intrinsic homeostatic mechanism to the heart. The molecular basis of the intrinsic system is largely unknown. Here, we test the hypothesis that bone morphogenetic protein-2 (BMP-2) mediates embryonic cardiac contractility upstream of myocyte-specific enhancer factor 2A (MEF2A).

Methods: The BMP-2 and MEF2A expression pattern was analyzed by RT-PCR, Western blotting, whole-mount in situ hybridization, and an in vivo transgenic approach. The cardiac phenotype of BMP-2 and MEF2A knock-down zebrafish embryos was analysed. Cardiac contractions were recorded with a video camera. Myofibrillar organization was observed with transmission electron microscopy. Gene expression profiles were performed by quantitative real-time PCR analysis.

Results: We demonstrate that BMP-2 and MEF2A are co-expressed in embryonic and neonatal cardiac myocytes. Furthermore, we provide evidence that BMP-2 is required for cardiac contractility in vitro and in vivo and that MEF2A expression can be activated by BMP-2 signaling in neonatal cardiomyocytes. BMP-2 is involved in the assembly of the cardiac contractile apparatus. Finally, we find that exogenous MEF2A is sufficient to rescue ventricular contractility defects in the absence of BMP-2 function.

Conclusions: In all, these observations indicate that BMP-2 and MEF2A are key components of a pathway that controls the cardiac ventricular contractility and suggest that the BMP2-MEF2A pathway can offer new opportunities for the treatment of heart failure.

Keywords
  • Developmental biology
  • Contractile function
  • Heart failure
  • Cardiomyopathy

1. Introduction

In vertebrates, the heart is the first organ formed in the developing embryo [1,2]. Cardiac contractility is regulated tightly as an essential homeostatic mechanism. Some of the control is neuro-humoral, but much appears intrinsic to the heart. The molecular basis of this intrinsic system is less clear. Heart failure is a complex disorder in which cardiac contractility is insufficient to adequately supply blood to the other organs. An important hallmark of heart failure is reduced myocardial contractility. This syndrome is a common complication ensuing from a wide variety of cardiovascular pathologies [3–6].

We have previously shown that myocyte-specific enhancer factor 2A (MEF2A), a transcription factor involved in cardiac development [7], plays a role in zebrafish embryonic cardiac contractility [8]. The signaling mechanism regulating the expression of MEF2A is largely unclear.

Bone morphogenetic proteins (BMPs) are growth and differentiation factors of the transforming growth factor-β superfamily involved in embryogenesis and morphogenesis of various tissues and organs [9–11]. BMPs and downstream BMP signaling effectors are essential for cardiovascular development [12]. BMP-2, a member of this family of proteins, is an important growth and differentiation factor. Deletion of BMP-2 in mice results in lethality at embryonic day 7.0–10.5 due to malformation of the amnion/chorion and cardiac malformations [13]. The early embryonic lethality of the BMP-2 germline null allele hinders further investigation into BMP-2 function at later stages. The mechanism by which BMP-2 regulates cardiac development and cardiac gene expression has remained obscure.

The zebrafish, Danio rerio, offers several distinct advantages as an embryological and genetic model system [14]. In addition, zebrafish embryos are not completely dependent on a functional cardiovascular system. Even in the total absence of blood circulation, they receive enough oxygen by passive diffusion to survive, thereby allowing a detailed analysis of animals with severe cardiovascular defects [2,15–19].

The neonatal rat cardiomyocyte model is conducive to a broad spectrum of experiments, such as studies of contraction, ischaemia, hypoxia and the toxicity of various compounds [20,21].

To gain insights into the molecular mechanism by which BMP-2 regulates cardiac development and cardiac gene expression, we sought to identify novel targets of BMP-2 signaling in neonatal rat cardiomyocytes. In the process of this screening, we found that MEF2A expression could be specifically increased by BMP-2 treatment in the neonatal rat cardiomyocytes. In this study, we dissect the biological function of this regulation employing zebrafish and neonatal rat cardiomyocyte model systems. We demonstrate that BMP-2 and MEF2A are co-expressed in embryonic and neonatal cardiac myocytes. Furthermore, we provide evidence that BMP-2 is required for cardiac contractility in vitro and in vivo. We further display that MEF2A controls embryonic cardiac contractility. Biochemical studies reveal that MEF2A expression can be activated by BMP-2 signaling in neonatal cardiomyocytes. Finally, we prove that exogenous MEF2A activity is unable to rescue the dorsalized phenotype in zebrafish embryos lacking BMP-2 function, while ventricle contractility defects can be significantly restored with MEF2A overexpression, suggesting that BMP-2 mediates embryonic cardiac contractility upstream of MEF2A.

2 Methods

All experiments conformed 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).

2.1 Zebrafish strains and maintenance

Wild-type (AB* strain) zebrafish embryos were obtained from natural spawning of wild-type adults. Zebrafish were raised, maintained and staged as previously described [22,23].

2.2 Morpholino modified antisense oligonucleotide and microinjections

We designed one morpholino modified antisense oligonucleotide (MO, Gene Tools, LLC) against the splice donor site of MEF2A exon 8 to interfere with splicing (MEF2A-MO). The sequence for the MEF2A-MO was the 5′-GTCGTTTGTGCTCACCAGAGTTGTA-3′. We designed one MO directed against the 5′ sequence around the putative start codon to block BMP2b translation (BMP-2b-MO). The sequence for the BMP-2b-MO was the 5′-CGGTCTGCGTTCCCGTCGTCTCCTA-3′. A standard control morpholino was designed for control microinjections. The sequence for the standard control morpholino was 5′-CCTCTTACCTCAGTTACAATTTATA-3′. Wild-type embryos were injected at the one-to-two cell stage with 0.2 ng BMP-2b-MO or 10 ng MEF2A-MO per embryo.

2.3 mRNA overexpression experiments

Sense-capped RNA was synthesized using the mMESSAGE mMACHINE system (Ambion) from the following linearized plasmids: pT7TS-MEF2A (zebrafish MEF2A, digested with EcoRI, transcribed with T7); pT7TS-BMP-2b (zebrafish BMP-2b, digested with EcoRI, transcribed with T7); pT7TS-BMP-2a (zebrafish BMP-2a, digested with EcoRI, transcribed with T7);pT7TS-GFP (GFP; EcoRI digestion and T7 transcription). RNA was diluted at 40 μg/ml in solution A (0.1% phenol red, 0.2 M KCl) and microinjected into the blastomeres at the one-two-cell stage of embryos.

2.4 Isolation and culture of neonatal rat ventricular cardiomyocytes, adult rat ventricular cardiomyocytes and embryonic rat cardiac myocyte

Ventricular cardiomyocytes were isolated from the hearts of 1-day-old Sprague Dawley rats or adult rats as described previously [21,24–26]. Rat cardiomyocytes were purified from embryonic day 15 (E15) embryos as described previously [25,27]. Cardiac myocytes were seeded in 60-mm gelatin-coated culture dishes (2.5×106) in Dulbecco's modified Eagle medium containing 10% fetal bovine serum. Using this method, we routinely obtained cultures containing more than 95% cardiac myocytes, as determined by immunohistochemical staining with the monoclonal antibody (MF20) against sarcomeric myosin heavy chain.

2.5 RNA interference and transfection

siRNAs for rat BMP-2 and FITC-labeled negative control siRNA were designed and synthesized by Shanghai GeneChem Inc (Shanghai, China). 24 h after transfection, knock-down in gene expression was identified after isolating the RNA and performing RT-PCR and western blot analysis.Lists of siRNA used in the experiment:

  • rat BMP-2 siRNAs

    • anti-sense:5′-UAG UCU GGU CAC AGG AAA UdTdT-3′

    • sense: 5′-AUU UCC UGU GAC CAG ACU AdTdT-3′

  • rat BMP-2 control siRNAs

    • anti-sense: 5′-UAG CCU UGU GAC UGC UAA UdTdT-3′

    • sense: 5′-AUU AGC AGU CAC AAG GCU AdTdT-3′

  • FITC-labeled negative control siRNA

    • anti-sense: 5′-UUC UCC GAA CGU GUC ACG UdTdT-3′

    • sense: 5′-ACG UGA CAC GUU CGG AGA AdTdT-3′

Neonatal rat ventricular cardiomyocytes were grown in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum. Transfections were performed with RNAi-Mate (BioChain Institute, Inc) according to the manufacturer's instructions. FITC-labeled negative control siRNA, a non-functioning oligo-ribonucleotide, serves as an internal standard of transfection efficiency. For confocal immunocytochemistry, cells plated on glass coverslips in six-well dishes were fixed with 2% paraformaldehyde 24 h after transfection. For transmission electron microscopy, cells seeded in 60-mm gelatin-coated culture dishes were fixed 2% paraformaldehyde and 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer 24 h after transfection.

2.6 In situ hybridization and photography

In situ hybridization experiments with BMP-2b antisense were performed as previously described [28]. Stained embryos were examined with Olympus BX61 and SZX12 microscopes, and photographed with a DP70 digital camera. Images were processed using Adobe Photoshop software.

2.7 Ventricular/atrial contractility analysis

Embryos were anesthetized and transferred to a recording chamber perfused with modified Tyrode's solution (136 mM NaCl, 5.4 mM KCl, 0.3 mM NaH2PO4, 1.8 mM CaCl2, 1 mM MgCl2, 10 mM HEPES, 5 mM glucose, pH 7.3). Cardiac contractions were recorded with a video camera (JVC, TK-C1381) as described [29,30,8]. The lengths of ventricles/atria in diastolic and systolic conditions were measured to calculate the ventricular/atrial shortening fraction (VSF/ASF). Values are presented as mean±S.D.

Embedded Image

2.8 Reverse transcription quantitative real-time PCR

Total RNAs were extracted from cultured cells using the TRIZOL method and reverse transcribed using oligo-dT primer following the manufacturers' instructions. For all experiments, cDNA was quantified using Applied Biosystems Sequence Detection System 7300. The SYBR green method was used to quantify cDNA. The sequence-specific primers (see Supplemental Table 1 for primers sequences) were designed using Primer Express2.0 software (Applied Biosystems, USA). No-RT controls and water controls gave similar high threshold cycle values, demonstrating that contamination contributed to less than 0.1% of quantified product. The input cDNA was normalized for PCR by using primers specific for β-actin.

2.9 Statistical analysis

Results are expressed as mean±S.E.M. Comparisons between groups were made with ANOVA (t test with Bonferroni correction). Results were considered significant at a value of p<0.05.

3 Results

3.1 Co-expression of BMP-2 and MEF2A in embryonic and neonatal cardiac myocytes

We confirmed by RT-PCR expression of BMP-2b and MEF2A in the zebrafish embryonic (Fig. 1A, lane 1) and adult (Fig. 1A, lane 2) heart. Expression of BMP-2 and MEF2A was also analyzed in purified neonatal rat ventricular cardiomyocytes (NRVCs). Western blot analysis found that both BMP-2 and MEF2A proteins were expressed in the NRVCs (Fig. 1B).

Fig. 1

BMP-2 and MEF2A are co-expressed in embryonic and neonatal cardiac myocytes. (A) RT-PCR analysis demonstrates expression of BMP-2 and MEF2A in zebrafish embryonic (lane 1) and adult (lane 2) hearts, and in purified NRVCs (lane 3). M, molecular size markers (in base pairs (bp)). (B) Western blot analysis of cultured NRVCs (lane 1) using rabbit polyclonal anti-human BMP-2 antibody and rabbit polyclonal anti-human MEF2A antibody. Lysate from human embryonic kidney (HEK) 293T cells transiently transfected with rat MEF2A cDNA (pcDNA3-MEF2A) and recombinant human BMP-2 (lane 2) serve as a positive control, respectively. Protein size standards are indicated in kilodaltons. (C–F) Confocal microscopy identification of MEF2A (C,E) and BMP-2 (D,F) in NRVCs (E,F) and in rat cardiomyocytes from embryonic day 15 (E15) embryos (C,D). (C,E) Expression of MEF2A protein in embryonic and neonatal cardiac myocytes. Cultured cardiac myocytes were co-immunostained with rabbit polyclonal anti-MEF2A and mouse monoclonal anti-myosin primary antibodies. Secondary antibodies were the anti-rabbit IgG FITC-conjugated secondary antibody (green signal for MEF2A) and anti-mouse TRITC-conjugated secondary antibody (red signal for myosin). The nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI) (blue) and the cytoplasm was stained with monoclonal anti- myosin. Note that the MEF2A signal co-localizes with DAPI in the nuclei of NRVCs. Scale bar, 40 μm. (D,F) Expression of BMP-2 protein in embryonic and neonatal cardiac myocytes. Cultured cardiac myocytes were immunostained with rabbit polyclonal anti-human BMP-2 and mouse monoclonal anti-myosin as the primary antibodies. The secondary antibodies are the anti-rabbit IgG FITC-conjugated secondary antibody (green signal for BMP-2) and anti-mouse TRITC-conjugated secondary antibody (red signal for myosin). Note that BMP-2 signal localizes in the cytoplasm of cardiac myocytes. Scale bar, 40 μm.

MEF2A is known to be a transcription factor belonging to the myocyte enhancer factor family [31,32]. Immunostaining for MEF2A revealed strong MEF2A protein signal within the nuclei of the embryonic and neonatal cardiac myocytes (Fig. 1C,E). In contrast, BMP-2 was predominantly localized to the cytoplasmic area of the cardiomyocytes at the same stage of development (Fig. 1D,F). These data suggest that BMP-2 and MEF2A proteins co-express in the developing cardiac myocytes.

The similar expression pattern of BMP-2 and MEF2A in embryonic and neonatal cardiac myocytes suggests a possible role of each protein in the other's expression.

3.2 BMP-2 is required for cardiac contractility

To investigate the function of BMP-2 on zebrafish cardiac development, we performed gene knock-down experiments. A morpholino modified antisense oligonucleotide (MO) against the start codon was designed to block translation of the BMP-2 mRNA.

At the end of gastrulation and at early segmentation stages, the embryos injected with 0.2 ng of BMP-2b MO displayed an elongated shape with a broadened notochord (Fig. 2B, left panels). Most of these embryos auto-lysed during somitogenesis and the rest displayed truncated and twisted body axis at 24 hpf (Fig. 2B, right panels). These correspond to class 5 (C5) dorsalized phenotypes according to the scale developed by Mullins et al [33]. Injection of the BMP-2b MO caused obvious cardiac contractility defects in live embryos. We quantified cardiac contractility by measuring the ventricular and atrial chamber shortening fraction. Ventricular shortening fraction (VSF) is reduced from wild-type values of 24%±2% (n=30) to 16%±2% (n=30) in BMP-2b MO-injected embryos after 48 hpf (p<0.05). By 72 hpf, the ventricle in the morphant becomes silent (VSF=0%, n=30), while the atrium continues to contract normally (Fig. 2C, see Supplementary Movie 1). To confirm the knock-down specificity, we evaluated whether BMP-2b mRNA can restore ventricular contractile function of BMP-2b morphant. BMP-2b morphant ventricle contractility defects can be significantly restored with BMP-2b mRNA (Fig. 2C; 35 out of 49 embryos in two experiments). These data suggest that BMP-2b may play a critical role in mediating ventricular contractility.

Fig. 2

BMP-2 is required for cardiac contractility. (A) Bright field (top row) and fluorescent (bottom row) pictures of wild-type embryos at the 75%-epiboly stage injected with 0.1 ng of BMP2b-GFP RNA plus 3 ng of control MO (left panels); or 0.1 ng of BMP2b-GFP RNA plus 3 ng of BMP-2b MO (middle panels); or 0.1 ng of BMP2a-GFP RNA plus 3 ng of BMP-2b MO (right panels). (B) Dorsalized phenotype of wild-type embryos injected with 0.2 ng of BMP-2b MO (top row) or 5 ng of control MO (bottom row) at 12 hpf (right panels) and 24 hpf (left panels). (C) Shortening fraction of the atrial (filled boxes) and ventricular (filled circles) chambers of the BMP-2b morphant and control embryos at 48, 60, and 72 hpf. BMP-2b morphant ventricle contractility defects can be significantly restored with BMP-2b mRNA. (D) Embryos at 24 hpf following treatment with 1.6 μg/ml Noggin (left) or acetic acid vehicle (right). Note the severe expansion of somites and curling of the trunk and tail. (E) Shortening fraction of the atrial (filled boxes) and ventricular (filled circles) chambers of Noggin-treated and control embryos (treated with acetic acid vehicle) at 48, 60, and 72 hpf. (F) Distribution of siRNA on the NRVCs obtained 12 h after transfection. (right) Bright field image of a typical neonatal ventricular cardiomyocyte. (middle) Green signal indicates the FITC-labeled control siRNA, a non-functioning oligo-ribonucleotide duplex which serves as an internal standard of transfection efficiency. (left) Merged image of right and middle. (G) Immunocytochemistry staining to show the effects of BMP-2 siRNA on cellular BMP-2 protein distribution. BMP-2 protein immunostaining (green) and nuclei staining (blue) of control siRNA-transfected (right) and BMP-2 siRNA-transfected (left) NRVCs. BMP-2 protein was visualized (green) using anti-BMP-2 antibody followed by FITC-conjugated secondary antibody. Representative images obtained 48h after transfection are shown. (H) Effect of BMP-2 siRNA on expression of BMP-2, BMP-4 and BMP-7. (right panels) Quantitative PCR analysis of BMP-2, BMP-4 and BMP-7 mRNAs. (left panels) Western blot analysis for BMP-2, BMP-4 and BMP-7 proteins. β-Actin serves as a loading control. (I) Noggin significantly blocks the shortening fraction of the NRVCs and ARVCs (J) BMP-2 depletion in the NRVCs disrupts the ventricular contractility.

To confirm that these morphant phenotypes specifically resulted from the inhibition of BMP-2b function and that BMP-2b is essential for zebrafish cardiac contractility, we investigated the efficacy of BMP-2b MO in blocking the translation of the BMP-2b and BMP-2a gene. We constructed a reporter construct by fusing between the coding region of BMP-2b or BMP-2a and the green fluorescent protein (GFP) reporter gene, BMP2b-GFP or BMP2a-GFP. Injection of 0.1 ng of BMP2b-GFP RNA in the presence of 3 ng of BMP-2b MO resulted in the absence of GFP protein expression (Fig. 2A, right panels; 100%, n=60). By contrast, there was no effect on the expression of GFP in embryos co-injected with 0.1 ng of BMP2b-GFP RNA and 3 ng of control MO (Fig. 2A, middle panels, 92%, n=80). There was also no effect on the expression of GFP in embryos co-injected with 0.1 ng of BMP2a-GFP RNA and 3 ng of BMP-2b MO (Fig. 2A, left panels, 98%, n=50). These experiments demonstrate that BMP-2b MO is able to specifically block translation of the BMP-2b gene, while translation of the BMP-2a gene is uninhibited. Additionally, these studies suggest that BMP-2b is essential for zebrafish cardiac contractility. Our data also suggest that BMP-2a cannot completely compensate for the BMP-2b deficiency in zebrafish cardiac function.

It is possible that BMP-2b morphant cardiac contractility defect is secondary to earlier developmental defects. To further confirm that BMP-2b plays an essential role in cardiac contractility, we set out to use a chemical genetic strategy to study BMP-2 signal transduction in the zebrafish embryo. This approach can easily achieve rheostatic control of signal strength at a special time window [34,35,19]. We first tested the effectiveness of Noggin, a BMP-specific antagonist protein [36,37], on the zebrafish BMP-2 signaling. To determine if the Noggin-treated embryos would recapitulate phenotypes of known BMP-2b mutants, we treated wild-type embryos with various concentrations of the chemical from 0.2 μg/ml to 3.2 μg/ml at 3 hpf. The majority of embryos exposed to Noggin at a concentration of 1.6 μg/ml were in the C5 dorsalization class, with severe expansion of somites and curling of the trunk and tail, resembling BMP-2b morphant (Fig. 2D, left panel, comparing with Fig. 2B). These data demonstrate that Noggin-treated embryos can phenocopy zebrafish mutants with severe reductions in BMP-2b signaling. When wild-type zebrafish embryos are exposed to Noggin at a concentration of 1.6 μg/ml after 48 h of development, a time point when both heart chambers usually exhibit vigorous, rhythmic contractions and propel blood cells to the vascular system, ventricular cardiomyocytes become silent within 24 h (Fig. 2E, see Supplementary Movie 2). The atrium in Noggin-treated embryos continues to contract, even long after the ventricle becomes silent. These findings suggest that along with the established cardiogenic role of BMP-2, this growth and differentiation factor also plays an essential role in controlling ventricular cardiomyocyte contractility, even in later stages of development.

We also examined the effect of BMP-2 inhibition on the shortening of isolated NRVCs and adult rat ventricular cardiomyocytes (ARVCs). The cells were incubated with Noggin and cardiomyocyte shortening was measured using a phase-contrast microscope attached to an optical video system in which the cardiomyocyte motion was analyzed by a computer. Noggin significantly decreased the shortening fraction of isolated NRVCs and ARVCs (Fig. 2I, see Supplementary Movie 3). These data demonstrate that BMP-2 plays a pivotal role in cardiac myocytes shortening.

Aside from inhibiting BMP-2 signaling, Noggin also blocks BMP-4 and BMP-7 signaling [36,37]. To determine the potential role of BMP-2 in neonatal cardiac myocyte shortening, we knocked-down endogenous BMP-2 expression in NRVCs by using siRNA. BMP-2 depletion in BMP-2 siRNA transfected cells was corroborated by immunocytochemistry assay. Fig. 2G shows strong BMP-2 staining in neonatal cardiac myocytes transfected with the control siRNA (right panel) cells compared to BMP-2 siRNA transfection (left panel). Using immunocytochemical analysis, we found that ∼95% of the cells were effectively transfected (based on knock-down of BMP-2). These data gave us confidence that nearly all cells utilized for single cell studies were highly likely to be transfected. Quantitative real time-PCR analysis demonstrated that BMP-2 siRNA resulted in a significant reduction in BMP-2 mRNA levels compared with control siRNA (p<0.01, Fig. 2H, left panel). Furthermore, BMP-4 and BMP-7 mRNA levels were unaffected by any of the above treatments (Fig. 2H, left panel). Changes in BMP-2 protein expression were consistent with the decrease in mRNA levels (Fig. 2H, right panel). Again, BMP-4 and BMP-7 protein expression was unaffected by BMP-2 siRNA (Fig. 2H, left panel). The finding that RNAi of BMP-2 triggers the specific knock-down of BMP-2 expression in NRVCs led us to examine the biological role of endogenous BMP-2 in NRVCs. We find that shortening fraction of the NRVCs is significantly impaired by siRNA directed to BMP-2. Shortening fraction of the NRVCs is reduced from normal values of 4.7%±0.4% (n=30) to 0.1%±0.1% (n=30) (p<0.05) (Fig. 2J, see Supplementary Movie 4). These data strongly suggest that BMP-2 is critical to ventricular myocyte contractility.

3.3 BMP-2 is involved in assembly of the cardiac contractile apparatus

The functional defects in BMP-2b morphants suggest that the cardiac contractile apparatus is abnormal. We compared sarcomere assembly in control zebrafish embryos and BMP-2 inhibition embryos at 72 hpf. At this stage, the control heart exhibits nascent sarcomeres containing both thick and thin filaments, and the nascent myofibrils assemble into higher-order sarcomere structures (Fig. 3A). By contrast, the differences between embryonic cardiomyocytes from, BMP-2b morphants and Noggin-treated embryos were the myofibrillar disorganization in the cardiac muscle. Detailed ultrastructural analysis of zebrafish cardiomyocytes revealed that the z-disc and other structures were irregular in BMP-2 inhibition embryos compared to controls. (Fig. 3B,C). Similarly, sarcomeric structures are disrupted in BMP-2 siRNA-transfected NRVCs (Fig. 3D–G).

Fig. 3

BMP-2 is essential for sarcomeric organization. (A-C) Transverse sections of zebrafish embryonic ventricle cells at 72 hpf are shown by transmission electron microscopy. In control embryos, nascent myofibrils (MF) assemble into higher-order sarcomere structures (A). Myofibrils are evident in the cardiomyocytes of BMP-2b knock-down (B) and Noggin-treated (C) embryos. The differences between embryonic cardiomyocytes from MEF2A morphants, BMP-2b morphants and Noggin-treated embryos were the myofibrillar disorganization in the cardiac muscle. MF, myofibril. Scale bar, 300 nm. (D,E) Transmission electron microscopy of NRVCs transfected with BMP-2 siRNA (D) or control siRNA (E). Sarcomeric structures are disrupted in BMP-2 siRNA-transfected NRVCs. Scale bar, 300 nm. (F,G) Confocal immunocytochemistry of neonatal cardiomyocyte cellular architecture and sarcomeric organization. Cultured NRVCs were immunostained with mouse monoclonal anti-myosin as the primary antibodies. The secondary antibody is the anti-mouse TRITC-conjugated secondary antibody (red signal for myosin). BMP-2 inhibition leads to disorganization in myosin (G). As a control, control siRNA transfection does not induce the same phenotypic disorganization in sarcomeres (F). Scale bar, 100 μm.

These data suggest that BMP-2 inhibition produces sarcomere assembly defects. The failure to generate normal sarcomeres possibly underlies the inability of the BMP-2 inhibited heart to generate significant systolic force. The disorganization of sarcomeres could be due to a specific function of BMP-2 in regulating the expression of a subset of sarcomeric genes, or other specific alterations in gene expression that secondarily lead to sarcomeric disorganization.

3.4 BMP-2 controls MEF2A expression in the neonatal cardiac myocytes

We recently reported that MEF2A is required for zebrafish embryonic cardiac contractility [8]. Similar ventricular contractility defects due to BMP-2 or MEF2A inhibition led us to explore whether BMP-2 and MEF2A share a common pathway that governs cardiac contractility in the embryonic heart. It has been demonstrated that BMP-2 induces differentiation of P19 teratocarcinoma cells with concomitant expression of cardiac-specific transcription factors such as NKX2.5 [38]. Ghosh-Choudhury et al reported that BMP-2 can increase MEF2A expression in CL6 cells, a derivative of P19 embryonal carcinoma cells, in vitro [39]. We therefore asked whether stimulation of neonatal cardiac myocytes with BMP-2 could induce MEF2A gene expression. We examined MEF2A expression levels by western blot analysis. After 12 h of culture in serum free medium, cardiac myocytes were stimulated with BMP-2 (100 ng/ml) for the indicated times (Fig. 4A). MEF2A protein was rapidly increased within 30 min in cardiac myocytes. BMP-2 did not induce global protein synthesis, as β-tubulin protein expression was unchanged in BMP-2-treated cardiomyocytes (Fig. 4A). To further address whether BMP-2 increases MEF-2A protein expression in a dose-dependent manner, we incubated the NRVCs with different concentrations of BMP-2. The lysates were immunoblotted with anti-human MEF2A antibody. BMP-2 concentrations, ranging from 25-200 ng/ml, increased MEF-2A protein expression in a dose-dependent manner (Fig. 4B).

Fig. 4

Effects of BMP-2 on MEF2A expression in NRVCs. (A) NRVCs were treated with 100 ng/ml BMP-2 for the indicated periods of time before harvesting for western blot analysis with indicated antibodies. Equal amounts of proteins were immunoblotted with the MEF2A antibody (top panel). The bottom panel shows the loading controls with anti-β-Actin antibody to demonstrate equal loading. Note that MEF2A protein was rapidly increased within 30 min. (B) BMP-2 increases MEF2A protein expression in a dose-dependent manner. The NRVCs were incubated with various concentrations of BMP-2 for 30 min. (C) BMP-2 increases MEF2A transcriptional activity in NRVCs in a PI 3-kinase dependent manner. The NRVCs were incubated with/without Ly294002 for 1 h before stimulation with various concentrations of BMP-2 for 30 min. Nuclear extracts from NRVCs were tested for MEF2A activity using the MEF2 assay kit. pcDNA3-MEF2A serves as a positive control.

To directly examine the effects of BMP-2 on MEF2A transcriptional activity in neonatal cardiomyocytes, we incubated the NRVCs with various concentrations of BMP-2. Transfection of MEF2A cDNA resulted in a significant increase in the transcriptional activity demonstrating that the MEF2 assay kit is responsive to the MEF2A transcription factor (Fig. 4C). We found that BMP-2 increased MEF2A transcriptional activity in the NRVCs (n=4, p<0.05) (Fig. 4C). BMP-2 can increase MEF2A expression in CL6 cells in a phosphatidylinositol (PI) 3-kinase dependent manner [39]. To test the effect of PI 3-kinase on BMP-2-induced MEF2A transcriptional activity in NRVCs. The NRVCs were incubated with Ly294002 for 1 h before stimulation with various concentrations of BMP-2 for 30 min. Inhibition of PI 3-kinase activity abolished BMP-2-induced MEF2A transcriptional activity (Fig. 4C). These data suggest that PI 3-kinase is associated upon BMP-2 activation of the MEF2A transcriptional activity in the neonatal cardiac myocytes.

These biochemical studies have demonstrated that not only could MEF-2A expression be activated by BMP-2 signaling in neonatal cardiomyocytes, but its transcriptional activity could also be increased by BMP-2.

3.5 BMP-2 mediates embryonic cardiac contractility upstream of MEF2A

We hypothesized that BMP-2 acts upstream of MEF2A to control cardiac contractility in the embryonic heart based on the biochemical studies that BMP-2 signaling increases MEF2A expression in the NRVCs. Therefore, we explored whether exogenous MEF2A was sufficient to rescue ventricular contractility defects in the absence of BMP-2 function.

We injected mRNA encoding GFP, or MEF2A-GFP into embryos lacking BMP-2b activity and assayed ventricular contractility by measuring the VSF at 72 hpf. MEF2A overexpression could not rescue the dorsalized abnormality in the BMP-2 morphant (Fig. 5A; 93%, n=153), however, the dorsalized abnormality could be restored with BMP-2b mRNA (Fig. 5A; 76%, n=79). Wild-type sibling embryos injected with 50 pg of GFP mRNA and subsequently exposed to Noggin exhibited ventricle contractility defects (VSF=0%) (n=30) (Fig. 5B, see Supplementary Movie 6). In contrast, injection with 50 pg of MEF2A-GFP mRNA rescued ventricle contractility defects significantly (VSF=18%±3%) (n=30, p<0.05) (Fig. 5B, see Supplementary Movie 6). The rescue effect of MEF2A-GFP mRNA on the ventricular contractility defects in the absence of BMP-2 function was assayed by measuring VSF at 72 hpf. We assessed the percentage of embryos rescued with MEF2A-GFP mRNA that showed severe ventricular contractility defects (VSF<5%). Approximately, 93% of all embryos exposed to Noggin lost their ventricular contractility (Fig. 5C, 111 out of 120 embryos in three experiments). This percentage dropped to 18% when embryos were injected with MEF2A-GFP mRNA and subsequently exposed to Noggin (Fig. 5C, 15 out of 85 embryos in three experiments). These data demonstrate that exogenous MEF2A is able to rescue the phenotypes derived from the Noggin treatment.

Fig. 5

Loss of BMP-2 function in Noggin-treated zebrafish embryos could be restored with MEF2A mRNA. (A) MEF2A overexpression cannot rescue dorsalized abnormality in BMP-2b morphant. Embryos at 11 hpf were injected with 50 pg of GFP mRNA (first lane), or 50 pg of MEF2A-GFP mRNA plus 0.2 ng of BMP-2b MO (second lane), or 50 pg of GFP mRNA plus 0.2 ng of BMP-2 MO (third lane), or 50 pg of BMP-2b mRNA plus 0.2 ng of BMP-2b MO (fourth lane). (B) Ventricle contractility defects can be significantly restored with MEF2A-GFP mRNA. Ventricular contractility was assayed by measuring the VSF at the indicated time. The ventricular shortening fraction of MEF2A-GFP mRNA-injected embryos subsequently exposed to Noggin (filled boxes) is significantly reduced at 72 hpf in contrast to GFP mRNA-injected embryos (filled circles). (C) 93% of all embryos exposed to Noggin lose their ventricular contractility. This percentage dropped to 18% when embryos were injected with MEF2A-GFP mRNA and subsequently exposed to Noggin. This percentage increased to 95% when embryos were injected with GFP mRNA and subsequently exposed to Noggin.

3.6 Dysregulation of cardiac genes in BMP-2b knockdown, Noggin-treated and MEF2A overexpression zebrafish hearts

To further investigate the molecular basis for the cardiac abnormalities in BMP-2 inhibition embryos, we performed real-time PCR analysis of the gene expression profiles of hearts from BMP-2b knockdown, Noggin-treated and MEF2A overexpression embryos at 72 hpf. From this analysis, we detected alterations in expression of cardiac contractile apparatus genes. In particular, troponin C (Fig. 6A), troponin T (Fig. 6B), troponin I (Fig. 6C), α-actin (Fig. 6D), ventricular myosin heavy chain (vmhc) (Fig. 6E), cardiac myosin light chain 2 (cmlc2) (Fig. 6F), α-tropmyosin (Fig. 6G) were down-regulated in BMP-2 inhibition embryos hearts. MEF2A overexpression can up-regulate the gene expression of these contractile proteins (Fig. 6A–G).

Fig. 6

Altered cardiac gene expression in BMP-2b knockdown, Noggin-treated and MEF2A overexpression zebrafish embryos at 72 hpf. Total RNA was extracted from twenty microdissected hearts of 72 hpf embryos and reverse transcribed using oligo-dT primer following the manufacturers' instructions. Expression of troponin C (A), troponin T (B), troponin I (C), α-actin (D), vmhc (E), cmlc2 (F), α-tropomyosin (G), FGF-8 (H), MEF2C (I), GATA4 (J), GATA5(K), tbx2 (L), tbx5 (M), tbx20 (N), and hand2 (O) was evaluated by quantitative PCR analysis. WT, wild type; KD, knock-down.

Expression levels of the genes for MADS domain factor MEF2C (Fig. 6I), zinc finger factor GATA4 (Fig. 6J), GATA5 (Fig. 6K), T-box transcription factor tbx5 (Fig. 6M), tbx20 (Fig. 6N), and bHLH transcription factor hand2 (Fig. 6O) were not affected by loss of BMP-2 function or MEF2A overexpression.

It has been reported that BMP-2 regulated tbx2 expression, an endogenous inhibitor of cardiac differentiation, in the mouse developing hearts [9]. We also found that tbx2 expression was decreased in BMP-2 inhibited embryo hearts (Fig. 6L). It seems that MEF2A does not suppress tbx2 based on the results that tbx2 was unaffected by MEF2A overexpression (Fig. 6L). FGF-8 was also not affected by loss of BMP-2 function or MEF2A overexpression (Fig. 6H).

4 Discussion

BMP-2 can increase MEF2A expression in CL6 cells in vitro [39]. In this study, we dissect the biological function of this regulation in vivo. There is no direct genetic evidence to confirm that BMP-2 controls cardiac contractility so far. In this study, we describe a signaling pathway made up of well-known signaling factors (Fig. 7) that is responsible for cardiac contractility during heart development. We show that both BMP-2 and MEF2A regulate cardiac contractility. Biochemical studies demonstrate that MEF2A expression can be activated by BMP-2 signaling in neonatal cardiomyocytes. We also show that exogenous MEF2A activity is sufficient to rescue ventricle contractility defects in zebrafish embryos lacking BMP-2 function, suggesting that BMP-2 may mediate embryonic cardiac contractility upstream of MEF-2A.

Fig. 7

Model of BMP-2-MEF2A signaling pathway responsible for embryonic cardiac contractility. BMP-2 is produced and secreted by cardiomyocytes during cardiac development. Secreted BMP-2 might bind to a tetrameric complex of type I and type II receptors, which feed the signal from BMP-2 into cells and thereby influence cardiomyocytes behavior. Extracellular inhibitors, such as Noggin (black), can regulate BMP-2 signaling by interact with BMP-2 to sequester it in an inactive complex.

Molecular pathways essential for heart development are difficult to dissect genetically in mammalian embryos because the disturbance of heart development leads to early embryonic death. BMP-2 deficient mice die in early embryogenesis due to cardiac malformations [13]. Since BMP-2 is essential for early development, it has been very tough to investigate its specific functions at later stages. In contrast to mammals, intact cardiovascular function is not essential for the early development of the zebrafish embryo because it obtains adequate oxygen by diffusion from its environment. Zebrafish embryos, mutant in cardiovascular gene function, survive for days to late larval stages. This permits analysis of the role of BMP-2 in heart development. BMP-2 is produced and secreted by cardiomyocytes during cardiac development (Figs. 1 and 7). In zebrafish, we find a separate role for BMP-2 in the control of heart contractility. Therefore, BMP-2 signaling is not only required for dorsal-ventral patterning of the early embryo, but also for the maintenance of ventricular contractility.

The BMP signalling pathway is required for proper tissue patterning and formation during embryogenesis, but because it is essential for early development, it has been very difficult to investigate its specific functions at later stages. To circumvent this problem, we established the Noggin-treated embryo model. Noggin is a BMP-specific antagonist protein found to rescue dorsal structures in ventralized Xenopus embryos [36,37]. Noggin inhibits BMP signalling by blocking the molecular interfaces of the binding epitopes for both type I and type II receptors (Fig. 7) [36]. We tested the effectiveness of Noggin on the zebrafish BMP-2 signaling. Our function data strongly demonstrate that Noggin-treated embryos can phenocopy zebrafish mutants with severe reductions in BMP-2 signaling (Fig. 2D). More recently, Pyati et al [40] generated transgenic zebrafish containing an inducible dominant-negative BMP receptor with which they inactivated BMP signaling at selected time points in development. These excellent reagents will no doubt put insights into the novel function of BMP signaling in zebrafish development.

Heart failure is a result of multiple signaling pathways, which regulate physiologically complex biological responses including myocytes contractility. Progressive decrease in myocyte contractility ultimately leads to chronic heart failure. A characteristic feature of heart failure is the progressive deterioration of left ventricular cardiomyocyte function. The mechanisms that mediate the pathogenesis of heart failure are poorly understood [3–6]. The use of medicines that improve contractility are complicated by concomitant propensity to arrhythmias. No currently used agents safely enhance cardiac contractility. The general patterns of the vertebrate heart, as well as the basic signaling factors that regulate its development, have been conserved throughout evolution. Thus, it is likely that our present findings will be relevant to the description of cardiac function during heart development in other vertebrates. Our results indicate that BMP2-MEF2A pathway may be used as therapeutic molecule in restoring cardiomyocyte function where mechanical contractility is the cause of reduced heart function. It will be of interest to examine whether the BMP2-MEF2A pathway will offer new opportunities for heart failure treatments. Our results open avenues for understanding the molecular mechanism of heart development and diseases. It is likely that identification and characterization of additional mutants with similar phenotypes will yield further insights into the components of the BMP2-MEF2A pathway required for embryonic cardiac contractility.

Acknowledgments

We are grateful to Dr. Yang Shi for constructive advice and Ms. Wen-Bo Bi for the critical reading of this manuscript. This work was supported in part by grants from “211” project to H. S., from National Natural Science Foundation of China to T. P. Z. and Y. W.(30328009, 30600489), from basic clinical joint grant of Fudan University to H. S. and Y. G., from an innovation grant of Fudan University to Y. W.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.cardiores.2007.02.007.

Footnotes

  • The GenBank accession numbers for rat MEF2A and zebrafish MEF2A are DQ 323505 and DQ 323506, respectively.

    Time for primary review 20 days

  • Abbreviations:
    Abbreviations
    MEF2A
    myocyte-specific enhancer factor 2A
    hpf
    hours post fertilization
    GFP
    green fluorescent protein
    MO
    morpholino
    V
    ventricle
    A
    atrium
    BMP
    bone morphogenetic protein
    NRVC
    neonatal rat ventricular cardiomyocytes
    VSF
    ventricular shortening fraction
    PLCγ1
    phospholipase C γ1
    VEGF
    vascular endothelial growth factor
    vmhc
    ventricular myosin heavy chain
    amhc
    atrial myosin heavy chain
    cmlc2
    cardiac myosin light chain 2
    ANF
    atrial natriuretic factor
    FGF-8
    fibroblast growth factor-8.

References

  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
  33. [33]
  34. [34]
  35. [35]
  36. [36]
  37. [37]
  38. [38]
  39. [39]
  40. [40]
View Abstract