A Critical Function for Ser-282 in Cardiac Myosin Binding Protein-C Phosphorylation and Cardiac Function

Abstract

Rationale

Cardiac myosin binding protein-C (cMyBP-C) phosphorylation at Ser-273, Ser-282 and Ser-302 regulates myocardial contractility. In vitro and in vivo experiments suggest the nonequivalence of these sites and the potential importance of Ser-282 phosphorylation in modulating the protein's overall phosphorylation and myocardial function.

Objective

To determine whether complete cMyBP-C phosphorylation is dependent on Ser-282 phosphorylation and to define its role in myocardial function. We hypothesized that Ser-282 regulates Ser-302 phosphorylation and cardiac function during β-adrenergic (β-AR) stimulation.

Methods and Results

Using recombinant human C1-M-C2 peptides in vitro, we determined that protein kinase A can phosphorylate Ser-273, Ser-282 and Ser-302. Protein kinase Cε can also phosphorylate Ser-273 and Ser-302. In contrast, Ca²⁺-calmodulin-activated kinase II (CaMKII) targets Ser-302 but can also target Ser-282 at non-physiological calcium concentrations. Strikingly, Ser-302 phosphorylation by CaMKII was abolished by ablating Ser-282's ability to be phosphorylated via alanine substitution. To determine the sites’ functional roles in vivo, three transgenic lines, which expressed cMyBP-C containing either Ser-273-Ala-282-Ser-302 (cMyBP-C^SAS), Ala-273-Asp-282-Ala-302 (cMyBP-C^ADA) or Asp-273-Ala-282-Asp-302 (cMyBP-C^DAD), were generated. Mutant protein was completely substituted for endogenous cMyBP-C by breeding each mouse line into a cMyBP-C null (t/t) background. Serine to alanine substitutions were used to ablate the residues’ abilities to be phosphorylated while serine to aspartate substitutions were used to mimic the charged state conferred by phosphorylation. Compared to control non-transgenic mice, as well as transgenic mice expressing wild-type cMyBP-C, the transgenic cMyBP-C^SAS(t/t), cMyBP-C^ADA(t/t) and cMyBP-C^DAD(t/t) mice showed no increases in morbidity and mortality and partially rescued the cMyBP-C^(t/t) phenotype. The loss of cMyBP-C phosphorylation at Ser-282 led to an altered β-adrenergic response. In vivo hemodynamic studies revealed that contractility was unaffected but that cMyBP-C^SAS(t/t) hearts showed decreased diastolic function at baseline. However, the normal increases in cardiac function (increased contractility/relaxation) as a result of infusion of β-agonist was significantly decreased in all of the mutants, suggesting that competency for phosphorylation at multiple sites in cMyBP-C is a prerequisite for normal β-adrenergic responsiveness.

Conclusions

Ser-282 has a unique regulatory role in that its phosphorylation is critical for the subsequent phosphorylation of Ser-302. However, each residue plays a role in regulating the contractile response to β-agonist stimulation.

Introduction

Cardiovascular disease, particularly ischemia, myocardial infarction and heart failure (HF) constitutes a growing health and economic problem, afflicting about 5 million people in the U.S. each year at an estimated cost of $29.6 billion.¹ HF is associated with diminished β-adrenergic (β-AR) receptor responsiveness, loss of cardiac contractility, abnormalities in Ca²⁺-handling^2,3 and altered contractile protein phosphorylation.⁴ Altered phosphorylation of the contractile proteins may partially underlie cardiac dysfunction in HF.^4,5 Cardiac myosin binding protein-C (cMyBP-C) phosphorylation at multiple sites can impact myocardial function,^6,7 metabolism⁸ and cardioprotection,⁹ but the precise functional consequences of phosphorylation are not characterized fully. cMyBP-C is a 140 kDa structural protein that is localized at the inner two-thirds of the A-band in the cardiac sarcomere.^10,11 cMyBP-C binds myosin at the S2 region¹² and light meromyosin (LMM),^13,14 where it modulates myosin assembly¹⁰ and actin-myosin interaction.^15,16 It also binds to titin via domains C8-C10,¹⁷ and presumably to actin as well although this interaction is less well defined,¹⁸ suggesting that these interactions may be necessary for the stability or function of cMyBP-C in the sarcomeres.

The cardiac isoform differs from the skeletal isoform in that it contains an extra domain at the N-terminus (C0) and a phosphorylation motif (M-domain) localized between the immunoglobulin I-like C1 and C2 domains,^19,8 in which Ser-273, Ser-282 and Ser-302 can each be phosphorylated.²⁰ The 3 sites are differentially phosphorylated by the enzymes PKA,²¹ PKC,²¹ CaMKII,¹⁹ PKD²² and the 90kDa ribosomal s6 kinase (RSK).²³ To begin to explore the physiological roles of cMyBP-C phosphorylation, we previously established two transgenic (TG) mouse models in which the three-phosphorylation sites were mutated to either alanine (phospho-ablation, AllP-) or aspartate (phospho-mimetic, AllP+).^7,9,24,25 Our findings suggested that cMyBP-C phosphorylation was essential for normal cardiac function and that charged residues at 273, 282 and 302 sites conferred cardioprotection against ischemia-reperfusion injury in either the α-myosin heavy chain or β-myosin heavy chain backgrounds. However, these data did not resolve the individual sites’ functional roles. Previously, hierarchical phosphorylation patterns for cMyBP-C have been defined in vitro,¹⁹ where Ca²⁺ dependent CaMKII phosphorylation at Ser-282 appeared to be necessary for phosphorylation of Ser-273 and Ser-302.^19,21,26 Furthermore, mono-phosphorylation of cMyBP-C by CaMKII is associated with myocardial stunning²⁰ and enhanced contractility²⁷ under certain conditions.²⁶ However, the critical role of Ser-282 phosphorylation in vivo remains to be systematically investigated.

In the present study, we explored the necessity and sufficiency of Ser-282 phosphorylation in mediating cardiac structure and function using three TG mouse models in which mutated cMyBP-C's replaced the endogenous cardiac protein. cMyBP-C's carrying the mutations Ser273-Ala282-Ser302 (SAS), Ala273-Asp282-Ala302 (ADA) or Asp273-Ala282-Asp302 (DAD) were generated and expressed specifically in cardiomyocytes. Our data demonstrate that Ser-282 phosphorylation is a critical determinant of Ser-302 phosphorylation in vitro and in vivo and that phosphorylation at each of the residues participates in the contractile response to β-AR stimulation in vivo.

Methods

TG Constructs

cMyBP-C wild-type cDNA was subjected to site-directed mutagenesis to generate cMyBP-C^SAS, cMyBP-C^DAD and cMyBP-C^ADA constructs. Cardiomyocyte-specific TG mice were generated using the mouse cardiac-specific α-myosin heavy chain promoter.^7,9 An N-terminal myc-tag, which has no effect on cMyBP-C function and stability,^7,9,28 was introduced in order to differentiate TG protein from endogenous cMyBP-C. All biochemical and functional assays used hearts from 10-12 week-old mice. Animals were handled in accordance with the principles and procedures of the Guide for the Care and Use of Laboratory Animals. The Institutional Animal Care and Use Committee at Cincinnati Children's Hospital approved all experimental procedures.

Characterization of TG Mice

TG mice were identified by PCR.⁷ Transgene copy number was determined by Southern blotting (Figure 1, Online Supplement) and transgene expression in the cardiomyocytes determined by real-time PCR and confirmed in some cases by northern blot analysis with a gene-specific γ³²P-labeled probe,⁷ using actin and GAPDH as loading controls (Figure 1, Online Supplement). To ensure that the observed phenotype was not due to insertional mutagenesis, at least two different TG lines were used from each group when assessing the functional consequences of complete cMyBP-C mutant protein replacement in the cMyBP-C^(t/t) background.^7,9 The presence of necrosis, fibrosis and disarray, as well as localization and integration of cMyBP-C mutant proteins into the sarcomere were evaluated by immunofluorescence microscopy as described.⁷ To determine sarcomeric A-band, I-band and M-line organization and thick filament distances, structural analyses at the light and electron microscopy levels were performed.^7,9

Evaluation of cMyBP-C Phosphorylation

To define the phosphorylation status of Ser-273, Ser-282 and Ser-302, myofibrils from non-transgenic (NTG) and TG mouse hearts were treated with either PKA, PKCε, CaMKIIδ or phosphatase (Sigma), electrophoresed on SDS-PAGE and Western blots performed using an anti-cMyBP-C^2-14 antibody that was raised against the first 14 residues of cMyBP-C (ProSci, Inc) and phospho-specific cMyBP-C antibodies as described previously.^22,24 In addition, the importance of each phosphorylation site was confirmed using recombinant human mutant C1-M-C2 peptides and site-specific cMyBP-C phospho-antibodies.^22,24

Cardiac Function

Cardiac function was determined in the intact closed chest model as described.⁷ Increasing doses of dobutamine were administered during 3-minute constant infusions (0.1 μl/min/gm body weight) to determine β-AR responsiveness. To observe changes in left ventricular (LV) chamber size and fractional shortening, noninvasive M-mode echocardiography was used.⁹

Statistical Analysis

All biochemical and functional assays were performed on mice obtained after 10-12 weeks of age along with mixed-gender controls. One-way ANOVA was used to test for overall significance, followed by the Student–Newman–Keuls test for multiple comparison testing between the various groups (Sigma Plot 11). Data are presented as means±standard error of the mean (S.E).

Results

Phosphorylation of Ser-282 is Critical for Ser-302 Phosphorylation

We focused on 3 sites in the cardiac-specific insertion region of cMyBP-C in order to determine their roles in modulating cardiac function. Initially, to determine the substrate specificity of these sites by PKA,²¹ PKC²¹ and CaMKII,¹⁹ recombinant human His-tagged C1-M-C2 fragments of cMyBP-C were generated in which the three phosphorylatable sites were individually mutated to alanine, as well as a wild-type control (Figure 1A).²² Peptides were treated with PKA, PKCε or CaMKIIδ and their phosphorylation status determined by Western blot (IB) analysis using site-specific cMyBP-C phospho-antibodies (Figure 1B and 1C). Experiments with PKA- and PKC-mediated phosphorylation confirmed that PKA phosphorylates all 3 sites, whereas PKC phosphorylates only Ser-273 and Ser-302 (Figure 1C). At low Ca²⁺ concentrations, CaMKIIδ targets Ser-302 while at high Ca²⁺ concentrations, Ser-282 can also be a substrate (Online Supplement Figure I), confirming that Ca²⁺-dependent CaMKII differentially phosphorylates Ser-282 and Ser-302. Phosphorylation at Ser-282 appears to be required for Ser-302 phosphorylation in vitro (Figure 1B, 4^th panel down, anti-p302).

Figure 1. Differential substrate specificities and regulation of Ser-273, Ser-282 and Ser-302 phosphorylation.

A, Shown is the His₆-tagged C1-M-C2 recombinant human cMyBP-C peptide that includes the C1-M-C2 domains with the 3 phosphorylation sites. In the normal, endogenous protein, each of these 3 serines is phosphorylatable. B, The ability of each residue to be phosphorylated by PKA, CaMKIIδ or PKCε is shown in Panels B and C, respectively. A dash indicates that the endogenous fragment with all 3 serines intact is present, and immunoblotting is indicated by “IB:” The serine (S) to alanine (A) mutation at each residue is indicated as Ser to Ala. Phosphorylation of each residue with each enzyme was assayed using our phospho-residue-specific antibodies via Western blot analyses. Ponceau-S staining of the membranes confirmed equal amounts of protein were loaded. D, E, Western blot analyses, using site-specific phospho cMyBP-C antibodies, of neonatal rat cardiomyocytes infected with constitutively active (CA) and dominant negative (DN) CaMKIIδ and (F, G) PKCε adenoviruses (Ad) at an MOI of 100 (n=5). C; negative control using an empty vector for infection, ISO; 50 nM isoproterenol treated (ISO) neonatal rat cardiomyocytes. H, I, At different time points after transverse aortic constriction (TAC) in mice, Western blot analysis of total heart proteins, using site-specific phospho cMyBP-C antibodies demonstrated that phosphorylation increased significantly after 30 minutes, but that Ser-282 phosphorylation was significantly reduced after 24 hours compared to Ser-273 and Ser-302 sites (n=6). *P<0.01 versus control or sham (n=5).

To explore further the site-specificity of Ser-273, -282 and -302 for PKA, PKC and CaMKII, we infected neonatal rat cardiomyocytes with adenovirus expressing constitutively active (CA) and dominant negative (DN) forms of rat CaMKIIδC²⁹ and rabbit PKCε.³⁰ Western blot analysis showed 10±2 fold overexpression of CaMKIIδ and 4±1.2 overexpression of PKCε in neonatal rat cardiomyocytes (Figure 1D and 1F). Cardiomyocytes were treated with isoproterenol as a positive control for the PKA activation. We then performed western blot analysis using our site-specific cMyBP-C phospho-antibodies²⁴ and compared those results with isoproterenol-treated cardiomyocytes as a positive control.²⁴ Results show that overexpression of CA CaMKIIδ specifically increased Ser-302 phosphorylation compared to controls, confirming that Ser-302 is a substrate for CaMKIIδ ex vivo, while the PKCε transfections confirmed that Ser-273 and Ser-302 are both PKC substrates (Figure 1D-G). To demonstrate active regulation of these sites during cardiac stress, a transverse aortic constriction (TAC) induced pressure-overload mouse model was used. Western blot analyses, using myofilament proteins from the TAC induced pressure-overload model, showed that Ser-273 and Ser-302 phosphorylation was significantly increased after 5 minutes, but decreased after 4 weeks, of TAC (Figure 1H and 1I). In contrast, Ser-282 phosphorylation was significantly increased after 5 minutes, but decreased after 24 hours, showing the sites were differentially regulated in terms of their steady state phosphorylation levels. These in vivo experiments are consistent with the hypothesis that Ser-282 phosphorylation is an early event in the stressed heart but does not resolve the question of whether its post-translational modification is necessary for the subsequent phosphorylation of the other sites.^19,21 To resolve this question, we turned to the intact animal and replaced endogenous cMyBP-C with protein that could not be phosphorylated at Ser-282.

Three Transgenic Mouse Models

Three separate constructs were used to generate multiple lines of TG mice in order to test the hypothesis that Ser-282 phosphorylation influenced the phosphorylation levels of residue 302 (Figure 2A and 2B). In the cMyBP-C^SAS mouse model, Ser-282 of cMyBP-C was mutated to a non-phosphorylatable alanine, preventing Ser-282 phosphorylation. In the cMyBP-C^ADA mouse model, the PKC sites (Ser-273 and Ser-302) were mutated to non-phosphorylatable alanines and Ser-282 was replaced with aspartate, resulting in a Ser-282 phospho-mimetic. In contrast, the third mouse model, cMyBP-C^DAD, contains charged residues (Aspartates) at amino acids 273 and 302, mimicking constitutive PKC-mediated phosphorylation in the absence of phosphorylation at Ser-282, resulting in Ser-282 phospho-ablation. All TG mice were characterized prior to crossing into the cMyBP-C^(t/t) background to assess any effects of TG expression. In each TG mouse model, we obtained three lines with variable transgene expression resulting in 25-45% replacement of the endogenous cMyBP-C with TG protein (Online Supplement Figure II). None of the lines showed any obvious phenotype or increased morbidity and mortality over a 10-month period, suggesting that the SAS, ADA and DAD mutations are benign in the unstressed animals (data not shown).

Figure 2. Transgenic expression of cMyBP-C mutants in the cMyBP-C^(t/t) null background.

A, Schematic diagram of cMyBP-C illustrates domain structure and regions that interact with the other filament systems in the sarcomere. B, Mutations in the cMyBP-C transgenes. C, SDS-PAGE analysis of myofibrillar proteins from non-transgenic (NTG), cMyBP-C null (t/t), TG cMyBP-C^WT (WT), cMyBP-C^SAS (SAS) cMyBP-C^ADA (ADA) and cMyBP-C^DAD (DAD) hearts. D, Expression levels of total cMyBP-C and the presence of myc-tagged (TG) cMyBP-C in the same mice were confirmed by Western blot analysis. E, Western blot analyses using phospho-specific antibodies on the in vitro dephosphorylated and phosphorylated NTG, cMyBP-C^WT and cMyBP-C^SAS myofibrils with phosphatase (control), PKCε, PKA and CaMKIIδ treatment. PKA phosphorylates all three sites, while PKC phosphorylates Ser-273 and Ser-302. CaMKIIδ was used at two different Ca²⁺ concentrations. F, Quantitative analysis of Western blot data (panel E). No significant differences were found between NTG and WT(t/t) groups. Values represent means ± S.E for each group (n=3), and designations denote significant differences (P<0.001) from WT(t/t) (* and #) controls.

To obtain complete replacement of endogenous cMyBP-C with the TG mutant proteins, the individual cMyBP-C^SAS, cMyBP-C^ADA and cMyBP-C^DAD lines were bred into the cMyBP-C null background (t/t).³¹ SDS-PAGE confirmed normal levels of cMyBP-C in this background, driven by transgenic expression of the mutant cMyBP-C and levels of other sarcomeric proteins, such as actin and myosin, were unaltered (Figure 2C). Using antibodies against the myc-tag as well as to the first 14 N-terminal residues of cMyBP-C, the incorporation of the mutant cMyBP-C into the sarcomere was confirmed as well as the ability of the protein to maintain the proper stoichiometry by Western blotting (Figure 2D).

To determine if elimination of Ser-282 phosphorylation affected Ser-302 phosphorylation, we treated myofilaments derived from cMyBP-C^SAS(t/t) hearts with PKA, PKC and CaMKII and determined the phosphorylation status of each residue by Western analyses using the residue-specific phospho-antibodies (Figure 2E). As expected, the Ser-282 site-specific phospho-antibody showed no reactivity under any conditions with cMyBP-C^SAS(t/t)-derived protein. Strikingly, the loss of Ser-282 phosphorylation impacted significantly on the ability of PKA, PKC and CaMKII to phosphorylate Ser-302 Figure 2E). Specifically, CaMKII-mediated phosphorylation was completely abolished in the SAS myofilament preparations (Figure 2E and Figure 4). To determine whether increased Ca²⁺ concentrations could overcome this block in the phospho-ablated cardiac sample, 25 mM Ca²⁺ was used with CaMKII to treat the cMyBP-C^SAS(t/t) myofilaments. Western blot analyses showed that increased Ca²⁺ concentrations can indeed stimulate Ser-282 phosphorylation in the NTG and cMyBP-C^WT(t/t) myofilaments, while Ser-302 phosphorylation in the absence of Ser-282 phosphorylation was reduced, with the site being hypo-phosphorylated relative to normal cMyBP-C (Figure 2F).

Figure 4. cMyBP-C phosphorylation and cardiac function.

A, B, Phosphorylation status of Ser-273, Ser-282 and Ser-302 sites before and after dobutamine treatment was determined by Western blots using phospho-specific antibodies (n=5). C, Contractile function was measured at baseline and dobutamine (32ng/ml min). Maximal rate of ventricular contraction (dP/dt_max) and relaxation (dP/dt_min). Data are presented as mean±S.E (n=6). *P<0.05, **P<0.005 compared with NTG; ¶P<0.05, ¶¶P<0.005, ¶¶¶P<0.0005 compared with (t/t), §P<0.05 compared with basal (absence of dobutamine stimulation) measurements.

Sarcomere Architecture in the SAS, ADA and DAD Mutant Hearts

We reported previously that expression of cMyBP-C^WT(t/t) and cMyBP-C^AllP+(t/t) effectively rescued the cMyBP-C^(t/t) phenotype, but that expression of the nonphosphorylatable cMyBP-C^AllP– failed to rescue the null, suggesting the necessity of cMyBP-C phosphorylation for normal sarcomere and cardiac function.^7,9 To explore this observation at the individual residue level, we first determined whether the phospho-specific mutants incorporated normally into the sarcomere. Immunohistochemical analyses were performed with both anti-myc and anti-cMyBP-C-antibodies (Figure 3A and 3B). Results show that SAS, ADA and DAD proteins were incorporated normally into the sarcomere in the cMyBP-C^(t/t) mouse background with striations indistinguishable from those seen in the NTG and cMyBP-C^WT(t/t) sarcomeres (Figure 3A and 3B). Histological analyses showed that, as expected, the cMyBP-C^(t/t) hearts displayed cardiac hypertrophy with fibrosis and disarray. However, there were no obvious abnormalities, fibrosis, calcification or disarray observed in the cMyBP-C^SAS(t/t) and cMyBP-C^ADA(t/t) samples, compared to NTG and cMyBP-C^WT(t/t) controls (Figure 3C and 3D). In contrast with these data, the cMyBP-C^DAD(t/t) mouse hearts showed fibrosis and disarray, mimicking the cMyBP-C^t/t phenotype. Considering that this mouse contains the cMyBP-C phospho-mimetic in which residues 273 and 302 are constitutively charged, the data emphasize the importance of Ser-282 phosphorylation for cMyBP-C function. Transmission electron microscopy confirmed the previous observations of subtle organizational changes at the sarcomere's center, at the M-line, in the cMyBP-C nulls, with the characteristic electron dense M-line largely absent or irregular (Figure 3E).^7,9,31 Consistent with the lack of disarray and fibrosis observed in the cMyBP-C^WT(t/t), cMyBP-C^SAS(t/t) and cMyBP-C^ADA(t/t) hearts, the center of those sarcomeres contained a well-defined M line that was indistinguishable from those observed in normal sarcomeres in 81-91% of the sarcomeres. In contrast, 83% of the cMyBP-C^DAD(t/t) sarcomeres displayed significant M-line abnormalities, suggesting that Ser-282 phosphorylation is necessary for the structural functions played by cMyBP-C.

Figure 3. Phenotypic analyses of sarcomere architecture.

A, Incorporation of Myc-tagged mutant cMyBP-C was confirmed by immunofluorescent staining of cMyBP-C with either an anti-myc (A) or anti-cMyBP-C antibody (B). C, Longitudinal sections of the heart stained with hematoxylin/eosin (20x) or D, Masson-trichrome (20x) to assess fibrosis. Note the areas of disrupted myocyte organization and fibrosis in the (t/t) and DAD(t/t) hearts. E, Transmission electron micrographs of sarcomeres show loss of the normal M-line structures in both cMyBP-C^(t/t) and cMyBP-C^DAD/(t/t)hearts. For each line between 240-539 individual sarcomeres were scored. For the control hearts, WT(t/t), 544 sarcomeres were counted with 9% blindly scored as “abnormal.” For the SAS/KO or ADA/KO animals, 15-18% had abnormal M lines, while for the DAD/KO animals, 244 sarcomeres were counted with 83% scored as having abnormal M-lines.

The heart/body weight ratios were normal in cMyBP-C^SAS(t/t) (0.57±0.02) and cMyBP-C^ADA(t/t) (0.57±0.02) mice, compared to cMyBP-C^WT:(t/t) (0.56±0.02) and NTG (0.50±0.013) controls at 3 months. Consistent with the histology, cMyBP-C^DAD(t/t) mouse hearts showed significantly increased heart/body weight ratios (0.70±0.015, P<0.0001 vs NTG) as did the cMyBP-C^(t/t) hearts (0.81±0.02, P<0.0001 vs NTG),⁷ but with normal systolic function (Table 1). In contrast, cMyBP-C^ADA(t/t) mice showed normal heart/body weight ratios, but reduced systolic function, suggesting that phospho-specific mutations have differential effects on cardiac structure and function. M-mode echocardiographic measurements showed that expression of normal cMyBP-C (cMyBP-C^WT(t/t)) rescued the cMyBP-C^(t/t) phenotype as previously shown (Table 1).^7,9 In contrast to those data, cMyBP-C^SAS(t/t), cMyBP-C^ADA(t/t) and cMyBP-C^DAD(t/t) mice showed only partial rescue of the null phenotype. Each model had significantly increased intraventricular septal thickness, LV end-diastolic and end-systolic dimensions (Table 1), indicating that none were as effective at restoring normal cardiac hemodynamics as the wild type protein (Table 1). The cMyBP-C^SAS(t/t), cMyBP-C^DAD(t/t) hearts and cMyBP-C^ADA(t/t) all showed reduced fractional shortening although only the cMyBP-C^ADA(t/t) hearts reached statistical significance. These data suggest that phosphorylation at Ser-273, Ser-282 and Ser-302 individually or in combination, is required for normal cardiac function to be maintained long term.

Table 1.

In Vivo Cardiac Function Assayed by M-Mode Echocardiography

	NTG (n=7)	(t/t) (n=15)	WT(t/t) (n=7)	SAS(t/t) (n=12)	ADA(t/t) (n=8)	DAD(t/t) (n=8)
IVSd (mm)	0.73 ±.12	1.15 ± .15***	0.74 ±.10¶¶¶	1.02 ± .13***	1.14 ± .15***	1.21 ± .10***
LVPW (mm)	.80 ± .06	1.06 ± .20*	.79 ± .12¶	.87 ± .13	.96 ± .10	1.04 ± .18*
LVIDd (mm)	3.86 ± .17	4.73 ± .49***	3.93 ± .32¶¶¶	4.29 ± .34¶	4.33 ± .38	4.53 ± .38*
LVIDs (mm)	2.63 ± .23	3.86 ± .53***	2.85 ± .38¶¶	3.27 ± .38	3.40 ± .32*	3.46 ± .51*
FS (%)	32 ± 4	18 ± 5***	28 ± 5¶	24 ± 5	22 ± 4*	24 ± 7
LV VOLd (mm)	64 ± 7	105 ± 25***	68± 12¶¶¶	83 ± 15	85 ± 18	95 ± 18**
LV VOLs (mm)	26 ± 6	66 ± 21***	31 ± 9¶¶¶	44 ± 13	48 ± 11*	51 ± 17*
LV EF (%)	60 ± 5	38 ± 9***	54 ± 8¶	48 ± 8	44 ± 7*	47 ± 11
LV Mass	84 ± 17	193 ± 44***	87 ± 18¶¶¶	134 ± 31¶	153 ± 29	184 ± 21***
Body weight (gm)	25.9±0.36	25.7±0.42	26.3±0.39	24.95±0.51	26.17±0.37	25.6±0.46

Measurement values in millimeter (mm) were averaged from at least three separate cardiac cycles: LVIDd and LVIDs, left ventricular inner diameter in diastole and systole; inter-ventricular septum (IVSd) thickness in diastole; LV VOLd and VOLs, left ventricular (LV) volume in diastole and systole; EF, ejection fraction; FS, fractional shortening. All values given as mean ± standard deviation.

^*P<0.05 vs NTG

^**P<0.01 vs NTG

^***P<0.001 vs NTG

^¶P<0.05 vs (t/t)

^¶¶P<0.01 vs (t/t)

^¶¶¶P<0.001 vs (t/t).

β-AR Stimulation and cMyBP-C Phosphorylation

Phospho-ablation at Ser-282 attenuates Ser-302 phosphorylation and we hypothesized that this would critically impact the heart's ability to respond to β-AR stimulation. The cMyBP-C^SAS(t/t) hearts contain only cMyBP-C^SAS cMyBP-C, allowing us to determine the functional consequence(s) of the inability of Ser-282 to be phosphorylated in vivo during β-AR stimulation. We noted that Ser-273 appeared to be hyper-phosphorylated in the cMyBP-C^SAS hearts at baseline, in contrast to the hypo-phosphorylation observed at Ser-302 (Figure 4A and 4B). At baseline, all groups showed normal systolic function (dP/dt_max), suggesting that mutating the three selected phosphorylation sites does not affect systolic function. Upon β-AR stimulation (dobutamine infusion), a reduced response was observed in all of the TG hearts, with the exception of those animals expressing the normal, cMyBP-C protein (cMyBP-C^WT/(t/t)). The cMyBP-C^SAS(t/t) hearts further showed significantly decreased phosphorylation at Ser-302 during β-agonist stimulation compared to NTG and cMyBP-C^WT:(t/t) controls, confirming the importance of Ser-282 phosphorylation in mediating Ser-302 phosphorylation as well (Figure 4C). Interestingly, the mice containing phospho-mimetics (cMyBP-C^DAD:(t/t)) and phospho-ablation (cMyBP-C^ADA:(t/t)) at residues 273 and 302 could not respond normally to dobutamine, emphasizing the importance of normal and, potentially, reversible phosphorylation at each of the 3 sites in mediating normal function and β-agonist-induced stimulation.

Discussion

Increasing data suggest that cMyBP-C phosphorylation can regulate myocardial function,⁷ sarcomere integrity⁹ and, in some cases, can be cardioprotective.^9,25 It is currently the only thick filament protein that is differentially phosphorylated by five important kinases (Figure 5) that can impact on cardiac contraction: PKA,²¹ PKC,²¹ CaMKII,¹⁹ PKD²² and RSK.²³ Our present study focused on deciphering the phosphorylation pattern of cMyBP-C and determining which phosphorylatable sites are necessary or, perhaps sufficient for normal cardiac function. Although cMyBP-C is extensively phosphorylated under basal conditions, the level of cMyBP-C phosphorylation decreases in animal models during development of HF, myocardial injury, pathologic hypertrophy and myocardial stunning.^{5,20,24,32,33} Previously, to better understand the mechanisms and significance of cMyBP-C phosphorylation, we established two TG mouse models to determine the necessity and sufficiency of cMyBP-C for normal cardiac function.^7,9 Our findings showed that cMyBP-C phosphorylation is essential for normal cardiac function⁷ and that a phospho-mimetic at the Ser-273, Ser-282 and Ser-302 phosphorylation sites conferred cardioprotection against myocardial ischemia reperfusion injury in either the α-myosin heavy chain⁹ or β-myosin heavy chain backgrounds.²⁴

Figure 5. cMyBP-C domains and kinases that target the three-phosphorylation motifs.

cMyBP-C is the 140-kDa thick filament sarcomeric proteins that consists of 11 domains; 8 immunoglobulin (oval) and 3 fibronectin type III (square) domains. The C0-domain, phosphorylation motif (m) and a cardiac-unique insertion in the C5-domain are unique (light green) to cMyBP-C, when compared to skeletal MyBP-C MyBP-C. The Proline/Alanine rich region (Pro-Ala) and hypothesized actin and myosin binding sites are marked. The three-phosphorylation motifs, Ser-282, Ser-273 and Ser-302 are shown (reference to the mouse identification number), which are targeted by protein kinase A (PKA),²¹ protein kinase C (PKC),²¹ Ca²⁺-calmodulin-activated kinase II (CaMKII),¹⁹ protein kinase D (PKD)²² and the 90kDa ribosomal s6 kinase (RSK).²³ Amino acid sequences are numbered using Uniprot and NCBI reference numbers: note that Ser-282 in the mouse corresponds to Ser-284 in the human sequence.

Phosphorylation of Ser-282 was a focus of the present study since phosphorylation levels of this residue are decreased in patients with HF^32,34 and atrial fibrillation.⁴ Strikingly, Ser-282 is also selectively phosphorylated by RSK, which regulates myofilament function.²³ In vitro studies demonstrated that phosphorylation at Ser-282 is a prerequisite for phosphorylation of the Ser-273 and Ser-302 sites.^19,21 The data obtained from those studies also showed that the ability to phosphorylate Ser-273 and Ser-302 is markedly reduced when Ser-282 is mutated to alanine, suggesting that hierarchical phosphorylation may be involved in cMyBP-C-mediated regulation of contraction in cardiac muscle. In the present study, we removed the ability of Ser-282 to be phosphorylated so as to investigate the importance of Ser-282 phosphorylation in maintaining normal cardiac function at the whole organ function. Our data demonstrate that Ser-282 regulates the subsequent phosphorylation of Ser-302, but not Ser-273.^19,21,22 We hypothesize that Ser-282 is both a regulatory and functional site for phosphorylation with the cMyBP-C^DAD:(t/t) mice showing very significant pathology at the light and ultrastructural levels (Figure 3D and 3E).

cMyBP-C is a substrate for PKC, which can phosphorylate Ser-273 and Ser-302 in cardiomyocytes.²¹ PKCε can also phosphorylate cMyBP-C³⁵ and this series of enzymes, some more selective than others, makes cMyBP-C phosphorylation exquisitely responsive to the changing environment. cMyBP-C phosphorylation, which is partially mediated by PKC, appears to play a critical role in cardiac function.⁹ However, the role(s) of these sites in cardiac physiology have not been characterized. PKC phosphorylation mediates a major mechanism by which the myofilament modulates changes in myocardial function.³⁵ Gautel's group obtained in vitro evidence that phosphorylation of Ser-282 may potentiate the PKC phosphorylation sites (Ser-273 and -302) and thus phosphorylation could function as a switch¹⁹ that might also be triggered in order to mediate accelerated cardiac function during β-AR activation. Evidence from other studies suggests that cMyBP-C phosphorylation influences actomyosin Mg²⁺-ATPase activity, the kinetics of cross-bridge cycling and the rate of relaxation.^19,36-38 Therefore, the second objective of the study was to assess the necessity and sufficiency of PKC-mediated cMyBP-C phosphorylation for sarcomeric integrity and for normal cardiac function. cMyBPC^ADA:(t/t) and cMyBP-C^DAD:(t/t) hearts showed decreased hemodynamics and contractility relative to cMyBP-C^WT:(t/t) controls during β-AR stimulation due to the constitutive phospho-ablation or phospho-mimetic “phosphorylation” (charged residues) of the PKC-phosphorylation sites in cMyBP-C.

Importantly, cMyBP-C^ADA:(t/t) mice unambiguously showed the effect of Ser-282 phosphorylation in the absence of PKC-site phosphorylation on whole organ anatomy and function and partially rescued the cMyBP-C^(t/t) phenotype. Conversely, despite the charged residues at Ser-273 and Ser-302, the cMyBP-C^DAD:(t/t) hearts displayed extensive damage, compared to cMyBP-C^ADA:(t/t) and cMyBP-C^SAS:(t/t) mice, to cellular components (Figure 3) and hypertrophied over time, mimicking the t/t null phenotype. Why do the cMyBP-C^DAD:(t/t) hearts display more overt abnormal pathology than the cMyBP-C^SAS:(t/t) mice when residue 282 is mutated in both? Overexpression of PKC isoforms in the heart causes hypertrophy in adult mice.³⁹ We hypothesize that mimicking constitutive activation of these two PKC-sites in cMyBP-C by replacing the two serines with aspartates in the cMyBP-C^DAD:(t/t) cardiomyocytes is detrimental to the hearts. In contrast, serines 273 and 302 in the cMyBP-C^SAS:(t/t) will not behave as if they were chronically phosphorylated: in fact, serine 302 will be hypophosphorylated in these hearts, and thus not exhibit the effects of chronic phosphorylation, as manifested both by a hypertrophic response and decreased M-line definition.

In conclusion, our data confirm the critical importance of Ser-282 in maintaining normal cardiac (Figure 3C and 3D) and sarcomere (Figure 3E) architecture regardless of the phosphorylation status of the PKC-sites. These data provide strong evidence that cMyBP-C phosphorylation directly affects both the heart's contractile properties and sarcomere organization.

Table 2.

In Vivo Cardiac Hemodynamic Measurements

		NTG (n=5)	(t/t) (n=8)	WT(t/t) (n=6)	SAS(t/t) (n=6)	ADA(t/t) (n=5)	DAD(t/t) (n=5)
Heart Rate (bpm)	Basal	411 ± 8	383 ± 7*	400 ± 5	363 ± 7**	353 ± 12**	361 ± 20
	Dobutamine (32ng/g/min)	589 ± 10	533 ± 16*	587 ± 10¶	555 ± 20	514 ± 10**	518 ± 33
MAP (mm Hg/s)	Basal	72.8 ± 3	66 ± 2	74 ± 4	68 ± 4	68 ± 5	59 ± 6
	Dobutamine (32ng/g/min)	67.9 ± 10	58 ± 6	70 ± 3	60 ± 5	60 ± 5	61 ± 4
Systolic Pressure (mm Hg/s)	Basal	86.8 ± 7	76 ± 5	88 ± 5	78 ± 5	88 ± 8	70 ± 5
	Dobutamine (32ng/g/min)	85 ± 8	72 ± 7	85 ± 3	70 ± 6	82 ± 9	80 ± 6
LVP (mm Hg/s)	Basal	94.7 ± 5	88 ± 2	98 ± 6	90 ± 4	90 ± 5	81 ± 4*
	Dobutamine (32ng/g/min)	106.9 ± 4	91 ± 4	105 ± 4¶	93 ± 3	93 ± 4	91 ± 4
dP/dtmax (mm Hg/s)	Basal	8611 ± 703	7853 ± 385	7842 ± 464	7472 ± 582	7058 ± 805	6592 ± 964
	Dobutamine (32ng/g/min)	19480 ± 1227	14858 ± 1362*	19016 ± 664¶	15283 ± 1426*	14242 ± 1629*	15268 ± 1328*
dP/dtmin (mm Hg/s)	Basal	-10008 ± 914	-4486 ± 224**	-7720 ± 361¶¶¶	-5868 ± 483**¶	-6112 ± 724*	-6477 ± 824*
	Dobutamine (32ng/g/min)	-14548 ± 2091	-5225 ± 409*	-10138 ± 952¶¶	-8077 ± 495*¶¶	-7143 ± 727*¶	-7264 ± 505*¶
LVEDP (mm Hg/s)	Basal	7.2 ± 0.3	13 ± 2	5 ± 1¶	10 ± 1	15 ± 1**	6 ± 2
	Dobutamine (32ng/g/min)	2.6 ± 0.8	3 ± 0.7	3 ± 1	3 ± 0.8	4 ± 2	3 ± 1
dP/dt40 (mm Hg/s)	Basal	8053 ± 508	7358 ± 350	7106 ± 294	7082 ± 618	6270 ± 1005	5945 ± 901
	Dobutamine (32ng/g/min)	16367 ± 499	13678 ± 995*	16037 ± 284¶	13536 ± 1140*	12831 ± 1364*	13776 ± 902*

Ventricular contraction and relaxation rates were measured at baseline and during β-agonist stimulation in the intact closed chest model including telemetric measurements of heart rate, mean arterial pressure (MAP), left ventricular systolic and end diastolic (LVEDP) pressures, left ventricular pressure (LVP), the maximum rate of heart contraction (+dP/dt_max) and relaxation (-dP/dt_min), and the rate of LV pressure increase determined at an LV pressure of 40 mm Hg (dP/dt₄₀). Cardiovascular stress response to increasing doses of dobutamine was determined during 3-minute constant infusions (0.1 μl/min/g BW) and data were analyzed using a PowerLab system (AD Instruments, Colorado Springs, CO). Data are presented as mean±S.E.

^*P<0.05

^**P<0.005 compared with NTG

^¶P<0.05

^¶¶P<0.005

^¶¶¶P<0.0005 compared with (t/t).

Non-standard Abbreviation and Acronyms

HF

heart failure

RSK

90kDa ribosomal s6 kinase

cMyBP-C

cardiac myosin binding protein-C

TG

transgenic

NTG

non transgenic

AllP-

a cMyBP-C in which 3 phosphorylatable serines have been mutated to alanine

AllP+

cMyBP-C in which 3 phosphorylatable serines have been mutated to aspartate

CaMKII

Ca²⁺-calmodulin-activated kinase II