OUP user menu

Developmental expression of myostatin in cardiomyocytes and its effect on foetal and neonatal rat cardiomyocyte proliferation

Godfrina McKoy, Katrina A. Bicknell, Ketan Patel, Gavin Brooks
DOI: http://dx.doi.org/10.1016/j.cardiores.2007.02.023 304-312 First published online: 1 May 2007


Objectives: Myostatin, a member of the transforming growth factor-beta (TGF-β) family, plays a key role in skeletal muscle myogenesis by limiting hyperplastic and hypertrophic muscle growth. In cardiac muscle, myostatin has been shown to limit agonist-induced cardiac hypertrophic growth. However, its role in cardiac hyperplastic growth remains undetermined. The aim of this study was to characterise the expression of myostatin in developing myocardium, determine its effect on cardiomyocyte proliferation, and explore the signalling mechanisms affected by myostatin in dividing cardiomyocytes.

Methods: We used quantitative PCR and Western blotting to study the expression of myostatin in cardiomyocytes isolated from rat myocardium at different developmental ages. We determined the effect of recombinant myostatin on proliferation and cell viability in dividing cardiomyocytes in culture. We analysed myostatin's effect on cardiomyocyte cell cycle progression by flow cytometry and used Western blotting to explore the signalling mechanisms involved.

Results: Myostatin is expressed differentially in cardiomyocytes during cardiac development such that increasing expression correlated with a low cardiomyocyte proliferation index. Proliferating foetal cardiomyocytes, from embryos at 18 days of gestation, expressed low levels of myostatin mRNA and protein, whereas isolated cardiomyocytes from postnatal day 10 hearts, wherein the majority of cardiomyocytes have lost their ability to proliferate, displayed a 6-fold increase in myostatin expression. Our in vitro studies demonstrated that myostatin inhibited proliferation of dividing foetal and neonatal cardiomyocytes. Flow cytometric analysis showed that this inhibition occurs mainly via a block in the G1-S phase transition of the cardiomyocyte cell cycle. Western blot analysis showed that part of the mechanism underpinning the inhibition of cardiomyocyte proliferation by myostatin involves phosphorylation of SMAD2 and altered expressions of the cell cycle proteins p21 and CDK2.

Conclusions: We conclude that myostatin is an inhibitor of cardiomyocyte proliferation with the potential to limit cardiomyocyte hyperplastic growth by altering cardiac cell cycle progression.

  • Cardiomyocytes
  • CDK2
  • Myostatin
  • p21
  • SMAD2
  • TGF-β

1. Introduction

The proliferative capacity of mammalian foetal cardiomyocytes diminishes dramatically during early neonatal development [1–3]. From this point onward, most of the adaptive growth needed to accommodate increased cardiac workload is accomplished by myocyte hypertrophy [3]. The inability of mature cardiomyocytes to proliferate, but instead grow by hypertrophy, has serious consequences for patients who suffer from heart failure post-infarct [4] or genetic cardiomyopathies that lead to progressive fibro fatty replacement of cardiac tissue, such as arrythmogenic right ventricular cardiomyopathy [5]. Identifying the key regulators of cardiomyocyte proliferation and understanding the molecular mechanisms involved in limiting the ability of cardiomyocytes to divide, therefore, is crucial if we are to develop new therapies to treat such heart diseases.

The loss in the ability of differentiated adult cardiomyocytes to divide has been the focus of interest for a number of researchers [6,7]. During cardiac development, it has been observed that changes in the proliferative capacity of cardiomyocytes coincide with a biochemical block in cardiac cell cycle progression leading to terminal differentiation [2,8]. In an attempt to stimulate cardiomyocytes to undergo cell division and repair damaged myocardial tissue, considerable efforts have been focussed on identifying putative regulators and molecular mechanisms that alter the expression and activity of cell cycle molecules involved in myocyte proliferation [7].

Myostatin, also known as Growth and Differentiation Factor-8, is a member of the TGF-beta (TGF-β) super family of secreted proteins and a potent inhibitor of skeletal muscle mass [9]. Studies in cattle and mice [9,10] have demonstrated that absence or impaired functioning of myostatin results in hyperplastic and/or hypertrophic growth of skeletal muscle, whilst increased levels of myostatin in the serum and muscles of patients with cachexia and HIV infection is linked to the muscle wasting disease seen in these patients [11]. Recently, a naturally occurring myostatin mutation has been identified in humans such that a male patient showed increased muscle mass compared to other boys of his age [12]. Together, these studies demonstrate that myostatin affects both hyperplastic and hypertrophic skeletal muscle growth. Myostatin mRNA and protein are also expressed in cardiac muscle [13] and it has been shown that myostatin affects adaptive hypertrophic growth in serum-starved neonatal rat cardiomyocytes [14]. However, the effect of myostatin on cardiomyocyte proliferation remains undetermined.

To study the role of myostatin in cardiac muscle proliferation, we firstly determined its expression in foetal cardiomyocytes with high proliferation index, in neonatal cardiomyocytes with lower proliferation index that decreases with age and in adult rat hearts wherein the majority of cardiomyocyte have lost their ability to divide [1–3]. We observed that myostatin is expressed developmentally. Our in vitro studies demonstrated that myostatin limited cardiomyocyte proliferation and caused cell cycle arrest. Myostatin also induced SMAD2 phosphorylation and altered the expression of cell cycle proteins critical for S-phase entry in proliferating cardiomyocytes. These findings highlight the potential of myostatin to limit cardiomyocyte proliferation and suggest that this inhibition involves SMAD2 signalling and cell cycle regulation.

2 Materials and methods

2.1 Animals

Wistar rats were obtained from Charles River Laboratories, UK. All animal experiments complied with the United Kingdom Animal Scientific Procedures Act 1986 as well as 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) and were approved by the ethical committee of the University of Reading, UK.

2.2 Recombinant myostatin

Bacterially-expressed recombinant myostatin was a kind gift from Professor Ravi Kambadur, University of Waikato, New Zealand [13]. Recombinant myostatin, purified from mouse myeloma cell line, NS0, was purchased from R&D Systems. Variations in the relative amounts of correctly folded bioactive protein in purified preparations of the recombinant protein have been described for members of the TGF-β [19] including myostatin [20]. Hence, concentrations of recombinant myostatin from different sources/lots that resulted in greater than 45% inhibition of proliferation in P2 rat cardiomyocytes after 72 h of treatment were empirically determined (see Table 1) and used in this study.

View this table:
Table 1

Effect of recombinant myostatin on P2 cardiomyocyte proliferation

Myostatin concentrationMean cell no.±SD (n=3)% Growth
Recombinant myostatin(E. coli)
0 μg/ml210,552±1661100
1 μg/ml174,173±4475*82.7
5 μg/ml88,635±805*42.1
Recombinant myostatin (mouse myeloma, lot EZ11612A, R&D)
0 ng/ml206,667±14,434100
250 ng/ml155,000±5000*75
500 ng/ml110,000±5000*53.2
  • Proliferating cardiomyocytes were treated with indicated concentrations of myostatin for 72 h. Viable cells were identified by trypan blue exclusion and counted using a haemocytometer. * indicates significant difference to untreated control cultures.

2.3 Isolation of primary cardiomyocyte

Cardiomyocytes from, embryonic age 18 (E18), postnatal ages 0, 2, 5 and 10 (P0, P2, P5 and P10, respectively) rat ventricles were obtained by serial enzymatic digestions. Adult (Ad) cardiomyocytes were prepared by Langendorff perfusion [21]. Isolated cardiomyocytes preparations were >95% pure as revealed by staining with cardiac-specific sarcomeric actin (Abcam, UK) antibody.

2.4 Cardiomyocyte culture and growth inhibition studies

Isolated E18 and P0 rat cardiomyocytes were seeded at a density of 2×105/well in 6 well dishes and grown in myocyte medium (DMEM/M199 (4:1) containing 5% foetal bovine serum (FBS) for 24 h. Recombinant myostatin was then added to appropriate cultures. At various time points thereafter, myostatin treated and untreated control cardiomyocytes were harvested by trypsinisation for analysis. Viable cells identified by trypan blue exclusion were counted using a haemocytometer.

2.5 Measurement of cell death

Propidium iodide (PI) uptake was used as a marker of dying cells that have lost their cell-membrane integrity to determine whether myostatin induced cell death in proliferating cardiomyocytes. Briefly, E18 cardiomyocytes cultured for 72 h with or without myostatin were harvested. Adherent and non-adherent cells from respective cultures were pooled, washed with PBS and incubated in PI buffer (PBS containing 3% FBS and PI; 20 μg/ml) at 4 °C for 20 min. Sorting of PI positive cells was performed using a FACSCalibur flow cytometer (BD Biosciences). Loss of plasma membrane permeability was quantified as the percentage of cells positively stained by PI.

2.6 Fluorescence activated cell sorting (FACS) analysis

Two-colour flow cytometric analysis was performed on ethanol-fixed cardiomyocytes. Briefly, foetal cardiomyocytes seeded at a density of 1×106 cells/10 cm2 dish were grown in 5% FBS myocyte media with or without myostatin for 48 h. Cells then were pulse-labelled with 10 μM bromodeoxyuridine (BrdU) for 30 min at 37 °C and prepared for FACS analysis as described [21]. Labelled nuclei were stained with anti-BrdU antibody and PI and analysed using a FACS Calibur flow cytometer and the CellQuest software (BD Biosciences). Flow cytometric measurements were performed on samples containing 10,000–30,000 cells. Cardiomyocytes were gated into the respective cell cycle phases (G0/G1, S and G2/M) as described [21].

2.7 Protein synthesis analysis

The effect of myostatin on protein synthesis in cultured cardiomyocytes was evaluated using [14C]-phenylalanine incorporation as an index. For proliferating cardiomyocytes, E18 myocytes were cultured in 5% FBS myocyte media with or without myostatin for 24 h. [14C]-phenylalanine (0.1 μCi/ml) was then added to each culture and allowed to grow for a further 24 h. For non-dividing cardiomyocytes, P2 myocytes cultured in 5% FBS myocyte media for 24 h were serum starved for 48 h. Phenylephrine (PE) (100 μM) with or without 500 ng/ml myostatin (R&D) and 0.1 μCi/ml [14C]-phenylalanine was then added to respective cultures and grown for a further 24 h. Cells were washed with PBS and ice-cold trichloroacetic acid (10%) and the precipitated proteins solubilized in 0.1 N NaOH and 0.01% SDS at 37 °C. Samples were mixed with scintillation fluid (3 ml) and quantified using a liquid scintillation counter.

2.8 Western blotting

To assess the expression of myostatin protein in cardiomyocytes isolated from foetal (E18), neonatal (P2, P5 and P10) and adult rat hearts, protein extracts prepared as described previously [21] were analysed by Western blotting by using the anti-myostatin rabbit polyclonal antibody previously described [13]. For analysis of phospho protein and cell cycle proteins, myostatin treated and control E18 cardiomyocytes from triplicate experiments were washed with ice cold PBS and scraped directly in protein lysis buffer. Pooled protein extracts from each treatment group were analysed by Western blotting using antibodies against phospho-SMAD2 (Chemicon), SMAD2/3 (Upstate biotechnology), p21 and CDK2 (Santa Cruz Biotechnology Ltd). GAPDH (Abcam) was used to correct for protein loading. Relative protein abundance was determined by densitometric analysis, in which signal intensities within the linear range on X-ray film were converted to numerical values using Quantity One® densitometric analysis software (Bio-Rad).

2.9 Quantitative RT-PCR

Real-time quantitative RT-PCR (qPCR) was used to evaluate myostatin mRNA expression in E18, P2, P5, P10 and adult rat cardiomyocytes, as well as p21 mRNA expression in E18 cardiomyocytes following myostatin treatment. RNA isolation and cDNA synthesis was performed as previously described [22]. Real-time quantitative PCR analysis was performed using ABsolute™ QPCR ROX mix (ABgene) and a GeneAmp® 5700 sequence detector (Applied Biosystems). Myostatin cDNA was amplified using forward primer, 5′-CGCTACCACGGAAACAATCATT-3′, and reverse primer, 5′-GCTTTCCATCCGCTTGCAT-3′ and the amplified product detected using a myostatin-specific Taqman® probe, 5′-(FAM)-CCATGCCTACCGAGTCTGACTTTC-(TAMRA)-3′. p21 cDNA was amplified using forward primer, 5′-AGAGCCACAGGCACCATGTC-3′, and reverse primer, 5′-CGAACAGACGACGGCATACTT-3′ and the amplified product detected using a p21-specific Taqman® probe, 5′-(FAM)-ATCCTGGTGATGTCCGACCTGTTCCAC-(TAMRA)-3′. Real-time PCR analysis was normalized using the rodent GAPDH control reagent kit (Applied Biosystems).

2.10 Statistical analysis

At least three independent experiments were performed for all quantitation analysis. Student's t-test was used for paired data analysis and for multiple data analysis one-way analysis of variance followed by Bonferroni t-test was used. Significance was accepted at p<0.05.

3 Results

3.1 Myostatin is expressed developmentally in cardiomyocytes

We and others have reported previously that there is a greater percentage of proliferating cardiomyocytes in E18 rodent hearts compared to P2 hearts and that this decreases with age [1–3]. We determined whether there was a correlation between these observed age-related changes in cardiomyocyte proliferative index and the temporal expression of myostatin, by studying the expression of myostatin in cardiomyocytes isolated from rat hearts at different developmental ages. Fig. 1 shows that myostatin is differentially expressed in rat cardiomyocytes at various developmental stages. Accordingly, E18 foetal cardiomyocytes display low levels of myostatin expression compared to P2 cardiomyocytes which display significantly higher level of myostatin mRNA (Fig. 1a). By postnatal age 10, when the majority of cardiomyocytes have lost their ability to proliferate [1–3], a 6-fold increase in myostatin mRNA expression is observed compared to that in E18 cardiomyocytes (p<0.001). Western blot analysis (Fig. 1b and c) demonstrates that the levels of active myostatin protein correlate strongly with myostatin mRNA levels, up to the age of P10. Thereafter in adult cardiomyocyte, a 2-fold decrease in the myostatin protein levels and a 15-fold decrease in mRNA levels were observed. Since the majority of cardiomyocytes lose their ability to proliferate during early postnatal development [1–3], the increasing expression of myostatin (particularly during the early postnatal period of cardiomyocyte growth) suggests a potential role for myostatin during this developmental growth phase.

Fig. 1

Developmental expression of myostatin in isolated rat cardiomyocytes. Myostatin (a) mRNA and (b) protein levels are expressed differentially in foetal (E 18), neonatal (P2, P5 and P10) and adult (Ad) rat cardiomyocytes as determined by qPCR and Western blot analysis, respectively. For qPCR analysis, cardiomyocytes isolated from hearts of pups (n=4–6 separate litters) were used to generate total RNA analysed for each data point. For protein analysis, 20 μg of total protein isolated from foetal (E 18), neonatal (P2, P5 and P10) and adult (Ad) rat cardiomyocytes were subjected to Western blot analysis and probed with a rabbit anti-myostatin polyclonal antibody. The position of myostatin precursor protein and active C-terminal myostatin band is indicated as previously described [13]. GAPDH was used as loading control. (c) Relative abundance of the processed, active form of myostatin as determined by densitometric analysis of at least 4 independent experiments. * indicates p<0.05 relative to myostatin levels in E18 cardiomyocytes.

3.2 Myostatin inhibits cardiomyocyte proliferation

In skeletal muscle, mature myostatin is a potent inhibitor of skeletal muscle myoblast [23] and satellite cell [24] proliferation. Therefore, we determined whether myostatin would inhibit cardiomyocyte proliferation by adding recombinant myostatin to proliferating cultures. Fig. 2 shows that the addition of recombinant myostatin to proliferating E18 (Fig. 2a) and P0 (Fig. 2b) rat cardiomyocyte cell cultures significantly inhibited their proliferation. The inhibitory effect of myostatin on cardiomyocyte proliferation was evident 24 h after treatment and at all time points studied thereafter. Quantitation of our results indicated that proliferation of E18 and P0 cardiomyocytes was inhibited by 29% and 21%, respectively after 24 h of myostatin treatment. The effect of myostatin treatment was sustained since, at 72 h, proliferation in E18 and P0 cultures was inhibited by 46% and 58%, respectively relative to untreated controls. The inhibition was not accompanied by increased cell death as determined by PI cell exclusion staining (Fig. 2c).

Fig. 2

Myostatin mediated inhibition of rat cardiomyocyte proliferation. Myostatin treatment inhibits (a) foetal (E18) and (b) neonatal (P0) cardiomyocyte proliferation. Cardiomyocytes were cultured in 5% FBS with or without 5 μg/ml recombinant E. coli overexpressed myostatin and harvested at the various time points indicated. Viable cells were identified by trypan blue exclusion and counted using a haemocytometer. Day 0 time point represents the average number of adherent cells/well (n=6) 24 h after plating. Data are mean values±SD of 4 independent experiments. (c) Inhibition of cardiomyocyte proliferation was not due to loss of cell viability as determined by PI cell exclusion staining. (d) Inhibition by myostatin was reversible following myostatin release. In the release experiments, cells were treated with 5 μg/ml of recombinant E. coli-expressed myostatin for 24 h then washed with PBS and cultured in 5% FBS myocyte medium for a further 48 h. * p<0.05 for myostatin-treated compared with untreated control cultures.

It has been observed that myostatin acts in a novel manner in that it blocks proliferation but that cells maintain their ability to carry on dividing once myostatin is withdrawn [25]. Thus, we investigated the ability of E18 cardiomyocytes to resume proliferation once myostatin was removed. E18 proliferating cardiomyocytes were cultured in the myostatin-containing media for 24 h, this media was then replaced with 5% FBS myocyte media and cardiomyocytes cultured for a further 48 h. Inhibition of myocyte proliferation by myostatin was found to be reversible as demonstrated by the reactivation of cardiomyocyte proliferation following replacement of myostatin-containing media with 5% FBS myocyte media (Fig. 2d).

3.3 Myostatin blocks G1 to S phase cell cycle progression in cardiomyocytes

Flow cytometric analysis was performed to determine what phase of the myocyte cell cycle myostatin affected. Fig. 3 shows that myostatin inhibited cardiomyocyte proliferation by blocking S phase entry during myocyte cell cycle progression. Results in Fig. 3b indicates that treatment of proliferating cardiomyocytes with myostatin for 48 h caused a 57–86% decrease in the S phase population (p<0.05), and a concomitant increase of 7–14%, and 15–17% (though not statistically significant) increase the G0/G1 and G2/M populations, respectively.

Fig. 3

Myostatin alters the cell cycle profile of E18 cardiomyocytes. Representative cell cycle profiles of proliferating E18 cardiomyocytes cultured with or without myostatin for 48 h. (a) BrdU treated cardiomyocytes were stained with anti-BrdU/FITC labelled secondary antibody and propidium iodide then analysed by FACS. X-axis represents FL3-A (red flouresence) and Y-axis represents FL1-H (green fluorescence). (b) Graphical representation of the percentage of cells in the G0/G1, S, or G2/M phases of the cell cycle following myostatin treatment for 48 h versus control. Data are mean values±SD from 3 independent experiments. * indicates p<0.05.

3.4 Myostatin treatment alters the expression of cell cycle proteins necessary for G1 to S phase transition

In skeletal muscle myoblasts, myostatin treatment alters the expression of cell cycle proteins, such as p21 and CDK2 [18,23]. We therefore examined whether myostatin treatment affected the expression of cell cycle proteins critical for the G1 to S-phase transition. Fig. 4a and b shows that in proliferating E18 cardiomyocytes, myostatin treatment decreases the expression of CDK2 protein significantly at 48 h. However, by 72 h, as proliferating control cultures become confluent and the rate of cell division ceases, a decreased expression of CDK2 is also observed in untreated cultures and at comparable levels to that observed in non-confluent myostatin-treated cultures (Fig. 4b). Interestingly myostatin mRNA is also upregulated as proliferating cardiomyocyte cultures become confluent (data not shown). Myostatin treatment also increased the mRNA (Fig. 4c) and protein (Fig. 4d and e) expression of the cell cycle inhibitor p21 in the proliferating E18 cultures significantly. A decreased (but insignificant) expression of cyclin E protein was also observed in some myostatin-treated cardiomyocyte cultures (data not shown).

Fig. 4

Myostatin affects the expression of G1/S cell cycle proteins in E18 cardiomyocytes. Pooled protein extracts from triplicate experiments for each treatment group were blotted onto PVDF membranes and probed with the antibodies indicated. GAPDH was used as loading control. (a) Representative Western blot analysis of CDK2 expression shows that myostatin treatment decreases its expression in proliferating cardiomyocytes. (b) Densitometric analysis (n=5) demonstrates that CDK2 expression is downregulated significantly in 48 h myostatin-treated cultures. (c) Myostatin-treatment increases the expression of p21 mRNA and (d) protein in E18 cardiomyocytes cultures. (e) Densitometric analysis (n=4) demonstrates that p21 protein expression is significantly upregulated within 48 h of myostatin treatment.

3.5 Effects of myostatin on protein synthesis in cardiomyocytes

Recently, we demonstrated that myostatin inhibits PE-induced protein synthesis rate per cell in serum-starved P2 neonatal cardiomyocytes thereby limiting agonist-induced hypertrophy [14]. Thus, we compared the effect of myostatin on protein synthesis in proliferating foetal cardiomyocytes and serum-starved non-proliferating P2 neonatal cardiomyocytes by [14C]-phenylalanine incorporation. Fig. 5a shows that total protein synthesis in E18 cardiomyocytes exposed to myostatin was significantly lower than that in untreated control cultures. However, as shown in Fig. 5b the protein synthesis rate per cell (normally an indication of hypertrophic growth) was not significantly different between myostatin treated and control cultures. This is in contrast to the effect of myostatin on serum-starved P2 cardiomyocytes subjected to PE-induced hypertrophy (Fig. 5c) wherein a significant inhibition on the rate of protein synthesis per cell and subsequently hypertrophic growth [14] is evident within 24 h of myostatin treatment. This suggests that the decrease in total protein synthesis observed in myostatin-treated E18 cultures (Fig. 5a) reflects proliferation inhibition and hence lower cardiomyocyte numbers relative to untreated controls.

Fig. 5

Effects of myostatin on protein synthesis in proliferating and PE stimulated non-proliferating cardiomyocytes. The effect of myostatin on the rate of [14C]-phenylalanine incorporation in proliferating E18 cardiomyocytes culture with or without myostatin for 24 h. Data expressed as (a) total [14C]-phenylalanine incorporation shows significant difference in myostatin treated cultures compared to controls. Data expressed as (b) the rate of phenylalanine incorporation per cell shows comparable protein synthesis rate between myostatin-treated and control cultures. (c) In contrast the effect of myostatin on the rate of [14C]-phenylalanine incorporation per cell in non-proliferating serum starved P2 cardiomyocytes following PE stimulation are significantly different to PE stimulated controls. Data are means±SD from triplicate experiments. * indicates p<0.05.

3.6 Acute myostatin treatment stimulates SMAD2 phosphorylation in E18 cardiomyocytes

Previous studies have demonstrated that members of the TGF-β growth factor family of ligands including myostatin, can elicit a functional response through the classical SMAD2/3 signalling pathway [15–17]. We determined whether myostatin activates a similar signalling pathway in proliferating cardiomyocytes by Western blot analysis. Fig. 6 demonstrates that myostatin induced SMAD2 phosphorylation in cardiomyocytes within 30 min of treatment and that this phosphorylation persisted (although at a reduced level) for up to 1 h. SMAD2 phosphorylation, however, dropped to basal levels by 3 h and no further increase in SMAD2 phosphorylation was evident up to 24 h of myostatin treatment. In contrast, Western blot analysis using anti-SMAD2/3 specific antibody shows that total SMAD2 protein remains unchanged in both myostatin-treated and control cultures.

Fig. 6

Myostatin stimulates SMAD2 phosphorylation in E18 cardiomyocytes. (a) Representative western blot demonstrates myostatin's ability to induce SMAD2 phosphorylation in proliferating E18 cardiomyocytes within 30 mins of treatment. E18 cardiomyocytes stimulated with 500 ng of recombinant myostatin (R&D Systems) for the length of time indicated were analysed by Western blotting for SMAD phosphorylation.

4 Discussion

This study and previous reports [13,14,26] demonstrates that myostatin mRNA and protein are expressed in cardiomyocytes. However, its physiological role in cardiac muscle growth remains unclear since myostatin null mice do not display the profound alterations in cardiac muscle mass evident in skeletal muscle [9]. Our study shows for the first time that myostatin expression is regulated developmentally in rat cardiomyocytes. E18 foetal cardiomyocytes were shown to express low levels of myostatin that gradually increased 6-fold in P10 rat cardiomyocytes. Given that the majority of rat cardiomyocytes cease to proliferate shortly after birth [1–3], it was interesting to note that the increased expression of myostatin coincided closely with the diminishing proliferative index of neonatal rat cardiomyocytes with age. Hence, we studied the effects of myostatin on cardiac proliferation and observed that myostatin is a potent inhibitor of serum-induced embryonic and neonatal cardiomyocyte proliferation. Taken together, these observations support a developmental role for myostatin in limiting cardiomyocyte proliferation.

Using FACS analysis, we have shown that myostatin inhibits cardiac proliferation in vitro through a block in the G1 to S phase cell cycle transition. Progression through the G1-S phase transition is controlled by the activity of the G1-acting cyclin:cyclin-dependent kinase (CDK) complex, cyclin E:CDK2 [27]. CDK2-associated activity plays a key role in regulating entry into and exit from S-phase of the cell cycle. For instance, it has been demonstrated that microinjection of affinity-purified anti-cdk2 antibodies during G1, inhibited entry into S phase [28]. Therefore, limiting CDK2 availability ultimately can lead to a G1 to S cell cycle arrest. Accordingly, we observed that a decrease in CDK2 expression is part of the mechanism through which myostatin acts to inhibit cardiomyocyte proliferation.

The CDK inhibitor p21 has been shown to inhibit cyclin:CDK complexes, including those that promote cell cycle progression through the G1 and S phase [29]. Furthermore, overexpression of p21 has been shown to arrest cells in G1 [30]. Since p21 mRNA and protein is upregulated in myostatin-treated E18 cardiomyocytes, our results suggests that upregulation of p21 is also part of the mechanism through which myostatin elicits a G1/S cell cycle block in cardiomyocytes. Myostatin-mediated inhibition of skeletal muscle myoblast proliferation has also been shown to involve increased expression of p21, decreased CDK2 activity and protein expression that ultimately results in a G1/S cell cycle arrest [24]. These observations highlight similarities in cell cycle regulation by myostatin in both skeletal muscle and cardiac myocytes.

In this study we observe that in proliferating cardiomyocytes, myostatin induced SMAD2 phosphorylation within 30 min of stimulation and subsequently inhibited cardiomyocyte cell division. It is noteworthy that TGF-β, which also signals via the classical SMAD2/3 pathway, has been reported to inhibit proliferation of immature cardiomyocytes [31]. The anti-proliferative effect of myostatin and other TGF-β family members, in conjunction with SMAD signalling, is well recognised in other cell types [15,16,32]. SMAD phosphorylation following binding of these ligands to their respective serine threonine kinase transmembrane receptor leads to the formation of heteromeric complexes with SMAD4. The complex then translocates into the nucleus to bind promoter regions of target genes or interact with various transcription factors to regulate gene expression [33–35]. SMAD2 proteins are able to interact with a diverse range of proteins that do not share a common motif [35,36]. Hence the phenotypic outcome following SMAD activation is therefore likely to depend on the SMAD-binding transcription factors present within the nucleus at the time of their activation. We propose that SMAD2 activation in myostatin treated E18 cardiomyocyte cultures most likely affects proliferation, as very little hypertrophic growth is evident at this age [1–3].

Our study also shows that in adult cardiomyocytes, where adaptive growth occurs mainly by hypertrophy [1], myostatin expression is maintained but at a much lower level. Since the majority of adult cardiomyocytes have lost their ability to divide and express very low levels of key molecules involved in cell division, we propose that high levels of myostatin, perhaps, are not necessary at this age to actively inhibit cell proliferation. We speculate that the expression of myostatin in adult cardiomyocytes might reflect a potential role in modulating hypertrophic growth rather than hyperplastic growth. Indeed, as observed for myostatin [14], adenovirus-mediated expression of growth differentiation factor 15 (GDF15), another member of the TGF-β family, has been shown to antagonise agonist-induced hypertrophy in P2 cardiomyocytes in vitro. Interestingly, this was shown to involve SMAD2/3 activation [37]. Furthermore, overexpression of SMAD2 in serum-starved cultured neonatal cardiomyocytes inhibits agonist-induced cardiomyocyte hypertrophy in vitro [37]. Thus the ability of myostatin to affect both hyperplastic and hypertrophic growth, depending on the physiological status of the cardiomyocyte, might be partly related to the transcriptional pathways influenced following SMAD activation by myostatin.

Myostatin knockout mice studies indicate that the cardiac muscle in these mice is not as grossly enlarged as their skeletal muscle [9,14]. This suggests that other regulators can compensate for the absence of myostatin in cardiac muscle. Nevertheless, relatively smaller cardiomyocytes are observed in myostatin knockout mice compared to their wildtype counterparts [14]. The observation that cardiac-specific overexpression of CDK2 increases cardiomyocyte proliferation and results in significantly smaller cardiomyocytes [38], suggests that the relatively smaller cardiomyocytes observed in myostatin knockout mice might be linked to abnormal proliferation. Thus, it will be interesting to determine whether the proliferative capacity of cardiomyocytes obtained from myostatin knockout mice is enhanced. It also is worth noting that satellite cells isolated from myostatin knockout mice have an enhanced proliferation rate [25] and that their cardiomyocytes display exuberant cell enlargement in response to the hypertrophic agonist, phenylepherine [14].

Our observations emphasise the fact that molecules that regulate cell cycle gene expression play a critical role in the proliferation of cardiomyocytes and that their identification and manipulation might help to extend the proliferative capacity of cardiomyocytes. A role for myostatin in cardiomyocyte growth is further supported by several other observations. Firstly, IGF-1 mediates stretch-induced upregulation of myostatin in neonatal rat cardiomyocytes [26]; secondly, myostatin is induced in cells immediately surrounding a myocardial infarct for up to 30 days post infarct [13]; and finally, myostatin is upregulated dramatically in the hearts of AKT knockout mice, a genetic model of cardiac hypertrophy [14]. These observations, together with those from our study, demonstrate that myostatin plays an important role in modulating cardiac growth. The emerging picture is that myostatin affects both hyperplastic and hypertrophic growth in cardiomyocytes, depending on their physiological and growth status.


We thank Dr. Ian Kavanagh for his help with the protein synthesis studies and Dr. Fleur Moseley, Carmen H. Coxon and Wendy Lockwood for their help with the cardiomyocyte cultures. This project was funded by grants from the British Heart Foundation project grant no PG/04/053/17031 (awarded to G.B. and K.P.) and a BHF Intermediate Fellowship FS/03/024/15491(awarded to K.A.B.).


  • 1 These authors contributed equally to this study.

  • Time for primary review 21 days


  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]
View Abstract