The contribution of cardiac myosin binding protein-c Ser282 phosphorylation to the rate of force generation and in vivo cardiac contractility

Abstract

Cardiac myosin binding protein-C phosphorylation plays an important role in modulating cardiac muscle function and accelerating contraction. It has been proposed that Ser282 phosphorylation may serve as a critical molecular switch that regulates the phosphorylation of neighbouring Ser273 and Ser302 residues, and thereby govern myofilament contractile acceleration in response to protein kinase A (PKA). Therefore, to determine the regulatory roles of Ser282 we generated a transgenic (TG) mouse model expressing cardiac myosin binding protein-C with a non-phosphorylatable Ser282 (i.e. serine to alanine substitution, TG^S282A). Myofibrils isolated from TG^S282A hearts displayed robust PKA-mediated phosphorylation of Ser273 and Ser302, and the increase in phosphorylation was identical to TG wild-type (TG^WT) controls. No signs of pathological cardiac hypertrophy were detected in TG^S282A hearts by either histological examination of cardiac sections or echocardiography. Baseline fractional shortening, ejection fraction, isovolumic relaxation time, rate of pressure development and rate of relaxation (τ) were unaltered in TG^S282A mice. However, the increase in cardiac contractility as well as the acceleration of pressure development observed in response to β-adrenergic stimulation was attenuated in TG^S282A mice. In agreement with our in vivo data, in vitro force measurements revealed that PKA-mediated acceleration of cross-bridge kinetics in TG^S282A myocardium was significantly attenuated compared to TG^WT myocardium. Taken together, our data suggest that while Ser282 phosphorylation does not regulate the phosphorylation of neighbouring Ser residues and basal cardiac function, full acceleration of cross-bridge kinetics and left ventricular pressure development cannot be achieved in its absence.

Key points

• Phosphorylation of cardiac myosin binding protein-C Ser282 has been proposed to modulate the phosphorylation of Ser273 and Ser302, and thereby the contractile response to increased β-adrenergic stimulation, yet the precise functional role of Ser282 is unknown.

• Protein kinase A phosphorylation of Ser273 and Ser302 was unaffected by Ser282 phospho-ablation, suggesting that Ser282 phosphorylation is not required for full phosphorylation of neighbouring residues.

• Mice with Ser282 phospho-ablation (TG^S282A) displayed normal basal in vivo cardiac function but impaired rates of pressure development in response to β-adrenergic stimulation.

• Basal rates of cross-bridge kinetics were unaffected by Ser282 phospho-ablation; however, the protein kinase A-mediated acceleration of cross-bridge recruitment was blunted in TG^S282A myocardium.

• Collectively, our data suggests that Ser282 phosphorylation is critical to achieve complete acceleration of cardiac contractile function in response to increased β-adrenergic stimulation, but also implicates Ser273 and Ser302 phosphorylation as important modulators of the cardiac myosin binding protein-C-mediated contractile response.

Introduction

To meet constantly changing systemic demand, the heart must modulate its output on a beat-to-beat basis. Increased cardiac output is achieved via increased β-adrenergic stimulation, which results in enhanced protein kinase A (PKA)-mediated phosphorylation of several key cardiac proteins. The principal targets of β-adrenergic signalling at the level of the myofilament are troponin I (TnI) and cardiac myosin binding protein-C (cMyBP-C) (Hartzell & Titus, 1982; Garvey et al. 1988), which together contribute to accelerated rates of contraction and relaxation at the myocyte level, and accelerated left ventricular pressure development and relaxation at the whole heart level (Layland et al. 2005; Barefield & Sadayappan, 2010).

cMyBP-C, which is highly phosphorylated under basal conditions, becomes increasingly dephosphorylated in patients with hypertrophic cardiomyopathy and heart failure (El-Armouche et al. 2007; Jacques et al. 2008; Copeland et al. 2010), suggesting that phosphorylation of cMyBP-C is critical for normal cardiac function. In agreement, transgenic (TG) mouse models that express cMyBP-C with non-phosphorylatable serine residues 273, 282 and 302 (i.e. serine to alanine substitutions at Ser 273, 282 and 302) develop pathological cardiac hypertrophy and dysfunction and display a reduced contractile reserve in response to β-adrenergic stimulation (Sadayappan et al. 2005; Tong et al. 2008). Although it appears there are other cMyBP-C residues that can be substrates for kinase activity (Copeland et al. 2010; Jia et al. 2010; Kuster et al. 2013), Ser273, 282 and 302 have been shown to be critical in modulating cardiac function in both health and disease (Nagayama et al. 2007; Tong et al. 2008).

Phosphorylation of cMyBP-C by PKA has been proposed to relieve a constraint on the myosin heads that allows increased interaction between myosin and the thin filament (Gruen et al. 1999; Colson et al. 2008, 2012). cMyBP-C phosphorylation accelerates the rate of force redevelopment, a measure of cross-bridge (XB) cycling kinetics, at submaximal Ca²⁺ activation (Stelzer et al. 2006b, 2007), the activation range in which force generation is highly dependent on cooperative recruitment of XBs, which progressively activate neighbouring thin filament subunits (reviewed by Moss et al. 2004). Previous investigations of the regulatory roles of PKA-mediated cMyBP-C phosphorylation on cardiac muscle contraction have been performed by site-directed mutagenesis of all three phosphorylatable serine residues (273, 282 and 302, i.e. TG3SA) (Tong et al. 2008; Colson et al. 2012) and have shown that preventing cMyBP-C phosphorylation by PKA eliminates the acceleration of XB kinetics. However, it remains largely unknown what the contribution of each of the individual Ser residues in the contractile response to PKA phosphorylation is and how each site contributes to the in vivo response to β-adrenergic stimulation. In this regard, earlier work suggested that there was a hierarchical phosphorylation pattern in which Ser282 phosphorylation was permissive for the phosphorylation of the remaining serine residues (Gautel et al. 1995). When Ser282 was mutated to an alanine residue (ablating phosphorylation at Ser282) and incubated with PKA, phosphorylation of the remaining phosphorylatable serine residues was decreased to 9% of wild-type (WT) levels (Gautel et al. 1995). An analysis of the time course of PKA-induced phosphorylation revealed that cMyBP-C was initially phosphorylated at one site, presumably at Ser282, with full phosphorylation of all three sites occurring only at a later time point (Gruen et al. 1999). A more recent study has also supported the idea that Ser282 phosphorylation is required for phosphorylation of the other remaining serine residues in vitro (Sadayappan et al. 2011). Wang et al. 2014 presented evidence that non-phosphorylatable Ser282, when coupled with phosphomimetic Ser273 and Ser302 residues, impairs cardiac contractile function by negatively impacting tension development and altering XB kinetics, suggesting that Ser282 could be critical in mediating the accelerating effects of PKA on XB kinetics (Wang et al. 2014). To date, however, no study has been conducted to examine the role of individual Ser residues in the PKA-mediated acceleration of XB kinetics, a fact that limits our understanding of cMyBP-C's phosphorylation-dependent regulation of contraction.

Therefore, the goals of this study were to determine if Ser282 phosphorylation is critical for the cMyBP-C-mediated regulatory response to PKA, to test whether Ser273 and Ser302 also contribute to changes in the rate of force generation in response to PKA and to establish if Ser282 phosphorylation is critical for acceleration of in vivo cardiac function. To achieve these goals we have generated a TG mouse model that expresses a constitutively non-phosphorylatable cMyBP-C Ser282 by mutating Ser282 to an alanine residue (S282A) and we investigated the effects of S282A on in vitro and in vivo cardiac contractile function, before and after PKA and dobutamine treatments, respectively.

Methods

Ethical approval and generation of mouse models

This study was conducted in accordance with the Guide for the Care and Use of Laboratory Animals (NIH Publication No. 85–23, Revised 1996), and the procedures for anaesthesia, surgery and general care of the animals were approved by the Institutional Animal Care and Use Committee at Case Western Reserve University. Mice of both sexes (aged 3–6 months) were used for all studies. TG mice expressing cMyBP-C with a serine to alanine substitution at residue 282 (TG^S282A) were generated on a cMyBP-C^−/− background (SV/129 strain) using procedures described previously (Tong et al. 2008). Mice expressing TG WT cMyBP-C (TG^WT) were previously generated and characterized (Tong et al. 2008). The cMyBP-C M-domain sequence containing Ser273, Ser282 and Ser302 is illustrated in Fig. 1 along with the TG S282A construct sequence. Non-transgenic (NTG) mice of the SV/129 strain were used as controls.

Figure 1. Transgenic cMyBP-C expression and phosphorylation.

A, cMyBP-C is composed of eight immunoglobulin (oval shaped) and three FnIII (rectangle shaped) domains labelled C0 (N-terminus) to C10 (C-terminus) and has a conserved M-domain linker between C1 and C2 domains. The M-domain contains three serine residues (Ser273, Ser282 and Ser302, underlined) in the WT sequence that are targets for protein kinase A phosphorylation. The substitution used to prevent phosphorylation of Ser282 in S282A TG protein is shown. B, Western blots of NTG and TG^S282A samples were probed with antibodies specific for phosphorylation of Ser273, Ser282 or Ser302, total cMyBP-C, myc tag or HSC70. Western blot with an antibody to myc tag shows that a band corresponding to cMyBP-C with S282A was only detected in TG^S282A samples. C, NTG and TG^S282A samples were stained with coomassie blue to determine the relative protein incorporation into the sarcomere. D, band intensities corresponding to protein expression were normalized to actin content to determine the relative protein expression. Values are expressed as means ± s.e.m., from six hearts per group. FnIII, fibronectin type III; cMyBP-C, cardiac myosin binding protein-C; NTG, non-transgenic; TG, transgenic; WT, wild-type.

Histological analysis of cardiac tissue

Histological examination of hearts was carried out as described previously (Cheng et al. 2013). Mouse hearts were formalin fixed for 4 h and subsequently sectioned at their mid-left ventricles (LVs). Dissected hearts remained in formalin overnight, paraffin embedded and sectioned at 5 μm thickness using a microtome. Staining of the cross-sections was done with haematoxylin and eosin (H&E) (Cheng et al. 2013).

In vivo cardiac morphology and haemodynamic function

Evaluation of overall cardiac morphology and in vivo cardiac function was performed by echocardiography on anaesthetized mice (1.5–2.0% isoflurane) (n = 8 mice per group), using a Sequoia C256 system (Siemens Medical, Malvern, PA, USA), according to a protocol described previously (Cheng et al. 2013). Pressure–volume (PV) loop analysis was performed on anaesthetized (1.5–2.0% isoflurane) and ventilated mice to asses LV contractile properties as previously described (Cheng et al. 2013). Ventricular pressure and heart rate (HR) were allowed to stabilize before recording baseline parameters, after which an i.p. injection of dobutamine (10 μg g⁻¹) was administered to measure the response to β-adrenergic stimulation. PV loop analysis was performed offline using LabChart7 software P–V Loop and Peak analysis modules (ADInstruments, Dunedin, New Zealand).

Preparation and PKA phosphorylation of myocardium for SDS gel and Western blot analysis

Cardiac myofibrils were isolated from frozen mouse hearts on the day of the experiment. A section of the frozen tissue was placed in a fresh relaxing solution, homogenized and myofibrils were skinned for 15 min using 1% Triton X-100 (Cheng et al. 2013). Skinned myofibrils were resuspended in fresh relaxing solution containing protease and phosphatase inhibitors (PhosSTOP and cOmplete ULTRA Tablets; Roche Applied Science, Indianapolis, IN, USA) and were kept on ice until use. Myofibrils, PKA and all solutions were brought to room temperature (22°C) before initiating the reaction. Myofibrils were incubated for 1 h, at 30°C in a solution containing the catalytic subunit of bovine PKA (Tong et al. 2008) with the final concentration of PKA at 0.15 U μg⁻¹ myofibrils. Control myofibrils were incubated under the same conditions in the absence of PKA. To assess the effects of S282A phospho-ablation on the time course of PKA phosphorylation, myofibrils isolated from TG^WT and TG^S282A hearts were incubated in PKA for 15 min, 30 min, 45 min and 60 min, and subsequently the levels of cMyBP-C phosphorylation were quantified. The PKA reactions were stopped by the addition of Laemmli buffer and samples were heated to 90°C for 5 min and stored at −20°C until use for Western blots and Pro-Q analysis.

Western blot and Pro-Q analysis of myocardial samples

Ten micrograms of the solubilized myofibrils were loaded and electrophoretically separated using 4–20% Tris-glycine gels (Lonza Walkersville Inc., Rockland, ME, USA) at 180 V for 70 min. Western blots were performed as previously described (Cheng et al. 2013). PVDF membranes were incubated overnight with one of the following antibodies: total cMyBP-C (Santa Cruz Biotechnology, Santa Cruz, CA, USA), phospho-serine antibodies specific for Ser273, Ser282 or Ser302 phosphorylation (21st Century Biochemical, Marlborough, MA, USA; Cheng et al. 2013), myc tag (Santa Cruz Biotechnology) or heat shock cognate protein 70 (HSC70) (Santa Cruz Biotechnology). To determine the level of total protein phosphorylation, gels were fixed and stained with Pro-Q diamond phosphoprotein stain (Invitrogen, Carlsbad, CA, USA) and coomassie blue. Densitometric scanning of Western blots and stained gels was performed using Image J software (US National Institutes of Health, Bethesda, MD, USA).

Myofilament contractile function

Solutions for experiments with skinned ventricular myocardium

Preparation of skinned ventricular myocardium was performed according to procedures described previously (Cheng et al. 2013). In brief, ventricular tissue was homogenized in a relax solution and skinned for 30 min using 1% Triton-X 100. A computer program (Fabiato, 1988) and known stability constants (Godt & Lindley, 1982) were used to calculate the composition of Ca²⁺ activation solutions. All solutions contained the following (in mm): 100 N,N-bis-(2-hydroxyethyl)-2-aminoethanesulphonic acid, 15 creatine phosphate, 5 dithiothreitol, 1 free Mg²⁺ and 4 MgATP. The maximal activating solution (pCa 4.5; pCa = −log [Ca²⁺]_free) also contained 7 EGTA and 7.01 CaCl₂; while the relaxing solution (pCa 9.0) contained 7 EGTA and 0.02 CaCl₂; and the pre-activating solution contained 0.07 EGTA. The pH of the Ca²⁺ solutions was adjusted to 7.0 with KOH (at 22°C) and the ionic strength was 180 mm. A range of pCa solutions, containing different amounts of [Ca²⁺]_free, were made by mixing appropriate volumes of pCa 9.0 and 4.5 solutions.

Apparatus and experimental protocols

Skinned myocardial preparations were mounted in between a motor arm (312C; Aurora Scientific Inc., Aurora, Ontario, Canada) and a force transducer (403A; Aurora Scientific Inc.) as described previously (Merkulov et al. 2012; Cheng et al. 2013). Changes in the motor position and signals from the force transducer were sampled at 2.0 kHz using sarcomere length control software (Campbell & Moss, 2003). For all mechanical measurements, the sarcomere length of the muscle preparations was set to 2.1 μm and the experiments were performed at 22°C (Merkulov et al. 2012; Cheng et al. 2013).

Force–pCa relationships

Force–pCa relationships were obtained and analysed as described previously (Desjardins et al. 2012; Cheng et al. 2013). Submaximal force (P) developed at each pCa was normalized to maximal force (P_o, at pCa 4.5), i.e. P/P_o to construct the force–pCa relationships. The steepness of a Hill plot transformation of the force–pCa relationship was used to estimate the apparent cooperativity of force development. The force–pCa data were fit using the equation P/P_o = [Ca²⁺]^nH/(k^nH + [Ca²⁺]^nH), where n_H is the Hill coefficient and k is the [Ca²⁺] required to reach half-maximal activation (i.e. pCa₅₀).

Measurement of the rate of force redevelopment (k_tr)

Measurement of k_tr in Ca²⁺-activated muscle fibres was performed according to a mechanical slack–restretch manoeuvre described previously (Stelzer et al. 2006a; Chen et al. 2010; Cheng et al. 2013). k_tr was measured to assess the rate of XB transitions from weak to strong binding states (Brenner & Eisenberg, 1986; Campbell, 1997). Skinned preparations were transferred from relaxing (pCa 9.0) to an activating solution (pCa ranging from 6.2 to 4.5) and were allowed to develop a steady-state force. Once the muscle fibre achieved a steady-state isometric force, it was rapidly slacked by 20% of its original muscle length, which resulted in a rapid decline in force. After a brief period of unloaded shortening (10 ms), the preparation was rapidly restretched back to its original length and the time course of force redevelopment was measured. k_tr for each slack–restretch manoeuvre was estimated by linear transformation of the half-time of force redevelopment, i.e. k_tr = 0.693/t_1/2, as described previously (Chen et al. 2010; Cheng et al. 2013).

Stretch activation experiments

Stretch activation experiments were carried out as previously described (Cheng et al. 2013). Fibres were placed in pCa solutions that yielded submaximal force (∼50% maximal force) and were allowed to develop a steady-state tension. Fibres were rapidly stretched by 2% of the initial muscle length and held at the increased length for 5 s before being returned to relaxing solution. The stretch activation variables are shown in Fig 6 and the characteristic features of the stretch activation have been described previously (Stelzer et al. 2006c; Ford et al. 2010). In brief, a sudden 2% stretch of the muscle fibre produces an instantaneous increase in force (P₁), which results from the strain of bound XBs. The strained XBs then rapidly detach, with a characteristic rate constant k_rel, and result in a rapid decline in force to reach minimum amplitude (P₂). Following this rapid decline, force develops gradually, with a characteristic rate constant k_df, because of length-induced recruitment of additional XBs into the force-producing state (Stelzer et al. 2006c) and reaches a new steady-state level (P₃). Stretch activation amplitudes were normalized to the pre-stretch Ca²⁺-activated force and were measured as described previously (Desjardins et al. 2012, Cheng et al. 2013). The time course of force decay was fit to a single exponential to yield k_rel and k_df was determined by a linear transformation of the half-time of force redevelopment as previously described (Stelzer et al. 2006c).

Measurement of PKA-mediated effects on contractile function

For experiments assessing the functional effects of PKA phosphorylation on XB kinetics and force–pCa relationships, skinned preparations were incubated with the catalytic subunit of bovine PKA (Sigma-Aldrich, St Louis, MO, USA), for 1 h (at 22°C), in a pCa 9.0 solution (final concentration 0.25 U μl⁻¹) (Cheng et al. 2013). Experimental protocols following PKA treatment were repeated as described above.

Statistical analysis

Comparisons between groups was performed using a one-way analysis of variance (ANOVA) followed by Tukey–Kramer post hoc test, and comparisons of baseline and PKA or dobutamine treatments within groups was performed using a Student's t test as appropriate. All data are presented as means ± s.e.m. The criterion for statistical significance was set at P < 0.05.

Results

Western blot and Pro-Q analysis of non-transgenic and transgenic myofilament protein content and phosphorylation

To determine the relative expression of cMyBP-C in TG^S282A hearts, myocardial samples were electrophoretically separated on a 4–20% SDS gel and transferred on to a PVDF membrane for Western blot analysis (Fig. 1B). TG expression was confirmed by the presence of a myc-tag band in TG^S282A samples, which was absent in NTG samples. Densitometric scanning of the SDS gels (Fig. 1C) indicated that the expression of cMyBP-C with S282A in the hearts of TG^S282A mice was 85 ± 3% when compared to the expression of cMyBP-C in the NTG hearts (Fig. 1D). The expression of cMyBP-C in TG^WT hearts was previously characterized as 72 ± 3% of NTG control (Tong et al. 2008). The expression level of cardiac TnI, another key sarcomeric protein, was not altered in TG^S282A hearts (Fig. 1D). Ser282 phospho-ablation in TG^S282A hearts was confirmed by Western blot using an antibody specific for phosphorylated Ser282 (Fig. 1B), which showed a robust band in NTG samples that was absent in TG^S282A samples (Fig. 1B). Phosphorylation of Ser273 and Ser302 was detected in both NTG and TG^S282A samples (Fig. 1B). HSC70 was used as a loading control to show that equal amounts of the samples have been loaded on to the gels (Fig. 1B).

Next, we assessed if abolishing phosphorylation of Ser282 prevents the PKA-mediated phosphorylation of neighbouring phosphorylatable sites (Ser273 and Ser302) within the M-domain of cMyBP-C (Fig. 2). All phosphorylation values in TG^S282A myocardium were normalized to TG^WT phosphorylation levels after PKA treatment (Fig. 2A and B) to facilitate comparisons between the two groups. Densitometric analysis of Western blots (Fig. 2A) indicated that the basal phosphorylation of Ser273 was not significantly different between the groups (47 ± 4%, 54 ± 8% and 61 ± 5% for NTG, TG^WT and TG^S282A respectively) (Fig. 2B). The basal phosphorylation of Ser282 was also similar in NTG and TG^WT groups (67 ± 9% and 74 ± 11% for NTG and TG^WT respectively), but was absent in TG^S282A samples (Fig. 2A and B). Ser302 phosphorylation levels were low at baseline for all three groups, with phosphorylation of Ser302 in TG^S282A myofibrils being slightly elevated (5 ± 1%, 6 ± 1% and 11 ± 1% for NTG, TG^WT and TG^S282A respectively; P < 0.05, Fig. 2A and B). To determine whether PKA-mediated Ser273 and Ser302 phosphorylation could still be achieved in the absence of Ser282 phosphorylation, myofibrils isolated from NTG, TG^WT and TG^S282A hearts were incubated with PKA for 1 h (see Methods). Phosphorylation of both Ser273 (102 ± 5%, 100 ± 12% and 99 ± 7% for NTG, TG^WT and TG^S282A respectively) and Ser302 (91 ± 8%, 100 ± 7% and 94 ± 3% for NTG, TG^WT and TG^S282A respectively) increased with PKA treatment in TG^S282A samples, indicating that abolishing phosphorylation at Ser282 does not prevent PKA-mediated phosphorylation of Ser273 or Ser302 (Fig. 2A and B). PKA phosphorylation of Ser282 was similar in NTG and TG^WT groups (106 ± 11% and 100 ± 8% for NTG and TG^WT respectively) and was absent in TG^S282A samples. To determine if the rate of PKA-mediated cMyBP-C phosphorylation was altered by Ser282 phospho-ablation, TG^WT and TG^S282A myofibrils were treated with PKA for 15, 30, 45 or 60 min, and Ser273 and Ser302 phosphorylation was examined by Western blot (Fig. 2C). No differences in the rate of Ser273 and Ser302 phosphorylation were observed at any time point examined, demonstrating that the rate of PKA-mediated Ser273 and Ser282 phosphorylation was unaffected by Ser282 phospho-ablation.

Figure 2. Determination of PKA-mediated phosphorylation of cMyBP-C and other sarcomeric proteins.

A, myofibrils isolated from NTG, TG^WT and TG^S282A hearts were treated with PKA and the phosphorylation of Ser273, 282 and 302 was assessed by Western blot. Band intensity of the phosphorylated signal for each residue in non-PKA treated (–PKA, left) and PKA treated (+PKA, right) samples was normalized to total cMyBP-C expression. HSC70 was used to verify equal sample loading. B, relative protein phosphorylation (phosphorylated signal/total cMyBP-C) was calculated for each residue and is expressed as a percentage of PKA-treated TG^WT values for that residue. C, representative Western blot showing the rate of PKA-mediated Ser273 and Ser302 phosphorylation. Samples were taken before the addition of PKA (0 min) and every 15 min after the addition of PKA. No differences in the rate of Ser273 and Ser302 phosphorylation were observed between TG^WT and TG^S282A samples. Representative gels shown are stained by Pro-Q stain for protein phosphorylation (D) and coomassie for total protein (E). cMyBP-C, TnT, TnI and RLC are labelled in each image, and the same gel is shown in (A) and (B). F, intensities of the phospho-bands normalized to total protein (i.e. phosphorylated TnI/total TnI) are expressed as a percentage of PKA-treated TG^WT phosphorylation. Values are expressed as mean ± s.e.m., from four NTG hearts, and eight TG^WT and eight TG^S282A hearts. *Significantly different from non-PKA treated samples from the same line, P < 0.05; †significantly different from the non-PKA-treated TG^WT, P < 0.05. cMyBP-C, cardiac myosin binding protein-C; NTG, non-transgenic; PKA, protein kinase A; RLC, regulatory light chain; TG, transgenic.

We also assessed whether abolishing phosphorylation of Ser282 affects the phosphorylation of other important regulatory sarcomeric proteins such as troponin T (TnT), TnI and myosin regulatory light chain (RLC) (Fig. 2C and D). The relative protein phosphorylation values (phosphorylation signal/total protein signal) are presented in Fig. 2E. Our results indicate that the baseline phosphorylation of TnT and TnI was similar between all the groups: 62 ± 6%, 64 ± 6% and 61 ± 6%, for TnT and 66 ± 6%, 73 ± 0% and 80 ± 6%, for TnI in NTG, TG^WT and TGS^282A samples, respectively. Furthermore, PKA phosphorylation of TnT and TnI was unaffected by the presence of S282A substitution in cMyBP-C as increases in phosphorylation following PKA treatment were similar in all the groups (99 ± 7%, 100 ± 1% and 107 ± 4% for TnT and 102 ± 9%, 100 ± 2% and 108 ± 8%, for TnI in NTG, TG^WT and TG^S282A samples respectively). Phosphorylation of RLC was unaffected by PKA treatment and thus was similar between all three groups under both basal and PKA-treated conditions.

Histological assessment of cardiac morphology

To examine if TG incorporation of cMyBP-C affects cardiac morphology, we examined histological sections of tissue prepared from NTG, TG^WT and TG^S282A hearts that were stained with H&E. Representative cross-sections for NTG, TG^WT and TG^S282A hearts are presented in Fig. 3, along with representative formalin-fixed hearts from each group. TG^S282A hearts exhibited similar overall size and morphology with no noticeable alterations in their chamber geometry when compared to control NTG and TG^WT sections (Fig. 3), indicating that TG expression of S282A cMyBP-C did not induce any overt cardiac pathology.

Figure 3. Analysis of cardiac morphology.

A, representative NTG, TG^WT and TG^S282A formalin-fixed hearts showing overall cardiac morphology, with the left ventricle indicated by an arrow. The scale bar is 1 mm. Representative cross-sections of NTG, TG^WT and TG^S282A hearts from the mid left ventricle were stained with haematoxylin and eosin and imaged at 100× magnification, with the digital zoom used to reduce the image in (B), to examine the cardiac morphology. The scale bars in (B) are 1 mm and in (C) are 50 μm. NTG, non-transgenic; TG, transgenic.

Echocardiography

Echocardiography was used to further confirm the effects of TG S282A expression on LV morphology and to test the impact of Ser282 phospho-ablation on in vivo LV function (Table 1). Cardiac morphology was assessed by LV mass (normalized to body weight) and posterior wall thickness in diastole and systole. TG^S282A hearts displayed no significant increase in heart weight or wall thickness, indicating an absence of pathological hypertrophy. Taken together, these results along with the observations from histological sections suggest that abolishing Ser282 phosphorylation does not affect overall cardiac morphology. Measures of systolic and diastolic function were evaluated to assess the impact of Ser282 phosphorylation on cardiac performance further. Ejection fraction and fractional shortening, measures of the systolic function, were not different in TG^S282A hearts compared to TG^WT controls (Table 1). Additionally, the isovolumic relaxation time, a measure of diastolic function, showed no differences between any of the groups (Table 1). These results indicate that Ser282 phospho-ablation had no significant impact on overall in vivo cardiac performance at baseline.

Table 1.

LV morphology and in vivo cardiac performance measured by echocardiography

	NTG	TGWT	TGS282A
BW (g)	26.6 ± 1.5	24.0 ± 2.3	25.8 ± 1.0
LV mass/BW	3.8 ± 0.2	4.1 ± 0.3	4.3 ± 0.2
HR (beats min−1)	412 ± 14	401 ± 22	420 ± 13
PWd (mm)	0.9 ± 0.1	0.9 ± 0.1	1.0 ± 0.1
PWs (mm)	1.1 ± 0.1	1.1 ± 0.1	1.2 ± 0.1
IVRT (ms)	19.8 ± 1.6	21.3 ± 1.5	21.9 ± 1.6
FS (%)	36.0 ± 2.3	33.9 ± 2.2	32.1 ± 1.7
EF (%)	72.6 ± 2.5	71.1 ± 2.1	68.5 ± 2.0

Values are expressed as means ± s.e.m. from eight mice per group. Abbreviations: BW, body weight; EF, ejection fraction; FS, fractional shortening; HR, heart rate; IVRT, isovolumic relaxation time; LV, left ventricular; LV mass/BW, ratio of LV and body weight; NTG, non-transgenic; PWd, posterior wall thickness in diastole; PWs, posterior wall thickness in systole.

Haemodynamics

To assess the impact of Ser282 phospho-ablation on the contractile response to β-adrenergic stimulation, cardiac function was measured by P–V loop catheterization in response to dobutamine administration (Table 2). In agreement with echocardiography results, TG^S282A mice displayed no differences in baseline function as assessed by the peak rate of pressure development (dP/dt _max) or τ, the relaxation time constant. In NTG and TG^WT hearts, dobutamine significantly accelerated dP/dt_max to a similar extent above baseline rates. Interestingly, in TG^S282A mice dP/dt_max was significantly accelerated from baseline by dobutamine, but the acceleration was significantly less than TG^WT values after dobutamine (Table 2), suggesting that ablation of Ser282 phosphorylation prevents full acceleration of pressure development in response to dobutamine administration. Post-dobutamine τ was similar in all groups. No differences in HR or end-systolic and end-diastolic pressure (EDP) were observed between groups at baseline or after dobutamine administration.

Table 2.

Left ventricular haemodynamic function measured by pressure–volume loop analysis

Group	HR (beats min−1)	ESP (mmHg)	EDP (mmHg)	dP/dtmax (mmHg s−1)	τ (ms)
–DOB
NTG	434 ± 11	97.0 ± 2.8	7.51 ± 0.35	7013 ± 409	9.15 ± 0.50
TGWT	418 ± 6	100.8 ± 4.4	7.97 ± 1.22	6861 ± 517	10.08 ± 0.51
TGS282A	447 ± 11	92.5 ± 1.5	7.62 ± 1.14	6064 ± 144	10.62 ± 0.36
+DOB
NTG	509 ± 12*	93.7 ± 3.4	5.16 ± 0.57*	12,132 ± 298*	7.60 ± 0.67*
TGWT	509 ± 12*	99.8 ± 4.0	6.63 ± 1.07	11,188 ± 581*	8.62 ± 0.32*
TGS282A	545 ± 10*	93.3 ± 2.1	4.97 ± 0.68*	8993 ± 421*†	8.99 ± 0.42*

Values are expressed as mean ± s.e.m. Abbreviations: DOB, dobutamine; dP/dt_max, maximum rate of pressure development; EDP, end-diastolic pressure; ESP, end-systolic pressure; HR, heart rate; τ, time constant of pressure relaxation; NTG, non-transgenic. n = 6 for NTG, 9 for TG^WT and 8 for TG^S282A.

^*Significantly different from the corresponding baseline group (without DOB treatment), P < 0.05.

^†Significantly different from TG^WT, P < 0.05.

To characterize the impact of Ser282 phospho-ablation further on the acceleration of pressure development in response to dobutamine, we assessed the rate of pressure change (dP/dt) over a range of developed pressures in the LV before ejection (Fig. 4A). EDP was subtracted from instantaneous ventricular pressure to calculate developed pressure (P_d) during isovolumic contraction before the onset of ejection. In all groups, dP/dt increased as pressure developed in the LV until dP/dt reached a maximum at about 40–50 mmHg, at which point the rate of pressure development began to slow slightly until the onset of ejection. Dobutamine significantly accelerated dP/dt during pressure development as indicated by an upward shift in the relationship between dP/dt and P_d (Fig. 4A). In TG^S282A hearts dobutamine significantly accelerated the rate of pressure development at every level of P_d examined; however, the acceleration of dP/dt was significantly less in TG^S282A hearts when compared to TG^WT hearts, reflected by a smaller upward shift in the dP/dt –P_d relationship (Fig. 4A). The rate of pressure development at baseline was unaltered by phospho-ablation of Ser282, and the rate of pressure development in NTG hearts was not different when compared to TG^WT hearts at any ventricular pressure assessed, either at baseline or following dobutamine administration.

Figure 4. Analysis of the rate of pressure development.

A, rate of pressure development (dP/dt) was measured over a range of developed pressures during isovolumic contraction. End-diastolic pressure was subtracted from instantaneous ventricular pressure to yield developed pressure. Dobutamine significantly increased dP/dt in both TG^WT and TG^S282A; however, the shift was blunted in TG^S282A hearts. *TG^WT + dobutamine significantly different from the TG^WT baseline (without dobutamine treatment), P < 0.05; †TG^S282A + dobutamine significantly different from the TG^S282A baseline (without dobutamine treatment), P < 0.05; ‡TG^S282A + dobutamine significantly different from TG^WT + dobutamine, P < 0.05. B, time taken to reach dP/dt_max t_d measured from end-diastole. Dobutamine shortened t_d in both TG^WT and TG^S282A, but TG^S282A hearts took significantly longer time to reach dP/dt_max. C, duration of ejection was unaltered by TG S282A expression, as TG^S282A hearts showed no differences either at baseline or after dobutamine compared to TG^WT hearts. (B) and (C): *significantly different from the corresponding baseline (non-dobutamine-treated) group, P < 0.05; †significantly different from the dobutamine-treated TG^WT group, P < 0.05. dob, dobutamine.

Additionally, we examined the time course of dP/dt to determine the role of Ser282 phosphorylation in accelerating early pressure development. Figure 4B shows the time taken to reach dP/dt_max (t_d) for TG^WT and TG^S282A hearts after the start of pressure development. Baseline t_d was not different between the groups (11.4 ± 0.6 ms and 12.0 ± 0.4 ms for TG^WT and TG^S282A, respectively), and while dobutamine shortened the time to dP/dt_max in both TG^WT and TG^S282A, TG^S282A mice reached dP/dt_max significantly slower than TG^WT after dobutamine (8.4 ± 0.4 ms and 9.8 ± 0.4 ms for TG^WT and TG^S282A, respectively). The t_d in NTG hearts was not significantly different either at baseline or after dobutamine when compared to TG^WT hearts (data not shown). HR had no effect on our observations, as TG^S282A t_d was still significantly slower when normalized to cardiac cycle duration (data not shown). We also examined the impact of TG S282A cMyBP-C expression on the time course of ventricular ejection (Fig. 4C). Ejection time was similar between TG^WT and TG^S282A hearts at baseline, and dobutamine administration slightly shortened the duration of ejection to a similar extent in both groups (Fig. 4C). No differences were detected between TG^WT and TG^S282A groups when ejection time was normalized to cardiac cycle duration at either baseline or after dobutamine infusion, demonstrating that ejection time is not impacted by the absence of Ser282 phosphorylation. This is in agreement with a previous study (Nagayama et al. 2007), which demonstrated that cMyBP-C plays a structural role in maintaining ejection that is independent of its phosphorylation status. Thus, these observations signify that the absence of a phosphorylatable Ser282 residue significantly alters the acceleration and timing of systolic pressure development following dobutamine administration. Taken together, our analysis of the rate of pressure development further demonstrates that baseline function can be maintained in the absence of Ser282 phosphorylation and, although cardiac contractility can be accelerated in the absence of Ser282 phosphorylation, pressure development in hearts lacking Ser282 phosphorylation cannot be accelerated to the same extent as in hearts that have three phosphorylatable Ser residues.

Measurement of steady-state mechanical properties

Steady-state force generation, cooperativity of force generation (n_H) and myofilament Ca²⁺ sensitivity of force generation (pCa₅₀) were measured to assess the effects of Ser282 phospho-ablation on myocardial mechanical properties. At baseline, there were no differences in Ca²⁺-independent force (measured at pCa 9.0; F_min), maximum Ca²⁺-activated force (measured at pCa 4.5; F_max), n_H, and pCa₅₀ between NTG, TG^WT and TG^S282A (Table 3). Furthermore, the addition of PKA had no effect on the minimum or the maximum force produced and n_H in all the groups tested, in agreement with earlier findings (Stelzer et al. 2007; Tong et al. 2008; Cheng et al. 2013). PKA treatment led to a similar decrease in the pCa₅₀ in all groups when compared to the corresponding non-PKA-treated values (Table 3). Thus, our data indicate S282A expression in the myocardium does not alter myofilament steady-state force generation under basal conditions or in response to PKA treatment.

Table 3.

Steady-state mechanical properties of skinned myocardial fibres isolated from NTG and TG hearts

Group	Fmin (mN mm−2)	Fmax (mN mm−2)	nH	pCa50
–PKA
NTG	0.61 ± 0.25	22.13 ± 1.73	4.10 ± 0.45	5.82 ± 0.02
TGWT	0.69 ± 0.20	21.74 ± 2.02	4.27 ± 0.56	5.81 ± 0.02
TGS282A	0.64 ± 0.21	21.41 ± 1.92	4.34 ± 0.66	5.82 ± 0.02
+PKA
NTG	0.47 ± 0.20	21.31 ± 2.12	3.87 ± 0.47	5.71 ± 0.02*
TGWT	0.52 ± 0.17	20.88 ± 1.71	4.11 ± 0.50	5.70 ± 0.02*
TGS282A	0.51 ± 0.19	20.47 ± 1.84	3.79 ± 0.63	5.73 ± 0.03*

Values are expressed as means ± s.e.m., from eight skinned fibres per group. Abbreviations: F_min, Ca²⁺-independent force at pCa 9.0; F_max, maximal Ca²⁺-activated force at pCa 4.5; n_H, Hill coefficient of the force–pCa relationship; NTG, non-transgenic; pCa₅₀, pCa required for half-maximal force generation; PKA, protein kinase A; TG, transgenic.

^*Significantly different from the corresponding non-PKA-treated group, P < 0.05.

Measurement of the rate constant of force redevelopment (k_tr)

We measured the effect of TG^S282A on the rate constant of transition of XBs from weak- to strong-bindings states (k_tr) (Brenner & Eisenberg, 1986; Campbell, 1997) using a release–restretch protocol (Stelzer et al. 2007; Chen et al. 2010). k_tr was measured at 50% of maximal Ca²⁺ activation before and after PKA treatment (Fig. 5) to determine the role of Ser282 in accelerating submaximal cooperative recruitment of strongly bound XBs. At baseline, no differences in k_tr were detected in NTG, TG^WT or TG^S282A fibres, suggesting that the absence of Ser282 phosphorylation does not negatively affect basal force redevelopment. Treatment with PKA significantly accelerated k_tr in NTG, TG^WT and TG^S282A fibres from baseline values; however, the acceleration in k_tr observed in TG^S282A was significantly less than TG^WT controls. After PKA treatment, TG^S282A k_tr at 50% of maximal Ca²⁺ activation was 16% less than in TG^WT controls (9.8 ± 0.4 vs. 11.7 ± 0.4, P < 0.05). A representative k_tr trace of TG^S282A-skinned myocardium before and after PKA treatment is shown in Fig. 5B. NTG k_tr was not significantly different from TG^WT after PKA treatment. Collectively, our results from k_tr measurements indicate that while abolishing phosphorylation at Ser282 does not alter k_tr at baseline or prevent an acceleration of k_tr at 50% maximal Ca²⁺-activation induced by PKA, a full acceleration of XB kinetics could not be achieved in the absence of Ser282 phosphorylation.

Figure 5. Effect of PKA-mediated phosphorylation on k_tr at half-maximal Ca²⁺ activation.

A, k_tr was measured at 50% of maximal Ca²⁺ activation, and shows that PKA treatment (open bars) accelerated k_tr in NTG (lane 1), TG^WT (lane 2) and TG^S282A (lane 3) fibres when compared to the corresponding non PKA-treated fibres (solid bars). Values are expressed as means ± s.e.m., from 12 skinned fibres and four hearts per group. *Significantly different from the corresponding baseline (non-PKA treated) group, P < 0.05. †Significantly different from the PKA-treated TG^WT group, P < 0.05. B, representative force traces following a mechanical slack and restretch protocol demonstrating PKA-mediated acceleration of the rate of force redevelopment (k_tr) in TG^S282A-skinned myocardium. Fibres were activated in a pCa solution that produced ∼50% of maximal force before (black trace) and after PKA treatment (grey trace). NTG, non-transgenic; PKA, protein kinase A; TG, transgenic.

Stretch activation

The role of Ser282 phosphorylation on the PKA-mediated response to dynamic strain-sensitive XB behaviour was assessed by stretch activation experiments. A rapid stretch of 2% of initial muscle length was imposed on NTG, TG^WT and TG^S282A myocardium at 50% of maximal Ca²⁺ activation before and after PKA treatment. Figure 6 illustrates a typical response of TG^S282A-skinned myocardium to PKA treatment. PKA treatment accelerated the rate of force decay (k_rel) and the rate of delayed force redevelopment (k_df) in TG^S282A myocardium, results that were qualitatively similar to the PKA-mediated changes in stretch activation in NTG and TG^WT myocardium observed in this study (Table 4) and in previous work (Tong et al. 2008). However, the absolute magnitude of the acceleration in k_df was significantly less in TG^S282A myocardium compared to TG^WT myocardium, as demonstrated by the reduced k_df after PKA treatment (Table 4). No significant differences in k_rel between TG^S282A- and TG^WT-skinned myocardium were observed. The amplitudes of all stretch variables were normalized to pre-stretch isometric force to facilitate comparisons between groups (Stelzer et al. 2006b). The amplitude of the initial rise in force (P₁), which results from stretch-induced strain of attached XBs, was decreased by PKA treatment to a similar extent in all groups. PKA treatment also increased the amplitude of the rapid decline in force (P₂) due to accelerated force decay, but did not alter the magnitude of the steady-state force achieved after stretch (P₃) (Table 4), an effect that was observed in all groups. The increased rate of XB detachment after PKA treatment that led to an increase in the amplitude of P₂ also resulted in an increased amplitude of P_df, which was unaffected by the phosphorylation status of Ser282. All amplitudes were normalized to the pre-stretch isometric force. No differences in k_tr or k_rel, or the amplitudes of P₁, P₂, P₃ or P_df were observed at baseline between groups. Our measurements of PKA-mediated acceleration of XB kinetics after a rapid stretch further demonstrate that while baseline kinetics can be maintained in the absence of Ser282 phosphorylation, the complete PKA-mediated acceleration of XB recruitment was not achieved in TG^S282A myocardium, indicating a role for Ser282 phosphorylation in the PKA-mediated acceleration of XB-induced cooperative activation of the thin filament.

Figure 6. Effects of PKA treatment on the stretch activation responses of TG^S282A myocardium.

Force transients following a stretch of 2% of muscle length were recorded at [Ca²⁺] yielding a pre-stretch isometric force of ∼50% maximal before (black trace) and after PKA treatment (grey trace) in TG^S282A-skinned myocardium. These representative transients are normalized to pre-stretch isometric force corresponding to the force baseline, which is arbitrarily set at zero, and demonstrate that PKA treatment significantly accelerated the overall stretch activation response in TG^S282A-skinned myocardium. The stretch activation parameters that were investigated are represented in the trace and are described in detail in the text. PKA, protein kinase A; TG, transgenic.

Table 4.

Dynamic stretch-induced responses of skinned myocardial fibres isolated from NTG and TG hearts

Group	kdf (s−1)	krel (s−1)	P1	P2	P3	Pdf
–PKA
NTG	10.07 ± 0.32	442 ± 30	0.588 ± 0.029	0.030 ± 0.004*	0.199 ± 0.016	0.169 ± 0.019
TGWT	10.32 ± 0.37	466 ± 29	0.602 ± 0.025	0.034 ± 0.009*	0.191 ± 0.024	0.157 ± 0.022
TGS282A	9.63 ± 0.34	479 ± 26	0.577 ± 0.027	0.042 ± 0.008*	0.203 ± 0.027	0.161 ± 0.024
+PKA
NTG	15.51 ± 0.42*	761 ± 46*	0.498 ± 0.024*	−0.057 ± 0.018*	0.194 ± 0.021	0.251 ± 0.024*
TGWT	15.97 ± 0.45*	748 ± 52*	0.495 ± 0.028*	−0.040 ± 0.016*	0.205 ± 0.028	0.245 ± 0.026*
TGS282A	13.01 ± 0.41*†	717 ± 40*	0.480 ± 0.024*	−0.042 ± 0.013*	0.214 ± 0.030	0.256 ± 0.027*

Values are expressed as means ± s.e.m., from 12 skinned fibres and four hearts per group. k_df, the rate constant of force development during stretch activation; k_rel: the rate constant of force decay during stretch activation; NTG, non-transgenic; P₁, the peak force attained following a rapid stretch; P₂, the difference between pre-stretch steady-state force to the minimum force value at the end of phase 2 of stretch activation protocol; P₃, the difference between pre-stretch steady-state force to the peak value of delayed force of stretch activation protocol; P_df, the difference between P₃ and P₂; PKA, protein kinase A; TG, transgenic. The comparisons were made at equivalent levels of activation, ∼50% of maximal activation following a 2% stretch of initial muscle length.

^*Significantly different from the corresponding non PKA-treated group, P < 0.05. ^†Significantly different from TG^WT, P < 0.05.

Discussion

Phosphorylation of Ser282 in cMyBP-C has been proposed to act as a molecular switch that dictates the phosphorylation of neighbouring residues, Ser273 and Ser302 (Gautel et al. 1995); however, the functional effects of Ser282 phosphorylation and its role in modulating XB behaviour are not fully understood. Thus, the purpose of this study was to determine the functional roles of Ser282 phosphorylation in the regulation of the contractile response to PKA treatment in vitro, and its role in modulating in vivo cardiac contractility in response to β-agonist stimulation. To delineate the functional effects of Ser282 phosphorylation on myofilament and whole heart function, we generated a TG mouse model (TG^S282A) expressing cMyBP-C with a non-phosphorylatable Ser282 (i.e. an alanine substitution for serine at position 282, S282A). Our results show that phospho-ablation of Ser282 does not prevent PKA-mediated phosphorylation of the neighbouring M-domain phosphorylatable residues, Ser273 and Ser302. TG^S282A mice displayed normal cardiac morphology and baseline contractile function as assessed by in vivo echocardiography and P–V loop analysis. Additionally, baseline steady-state force generation and dynamic XB kinetics as assessed by measurements of the rate constant of force redevelopment (k_tr) following a mechanical slack–restretch manoeuvre and the rates of XB relaxation (k_rel) and delayed force development (k_df) following a rapid acute mechanical stretch, were unaltered by Ser282 phospho-ablation. However, differences were observed in the ability of dobutamine in vivo and PKA in vitro to accelerate fully contractile kinetics in TG^S282A. While treatment of skinned myocardium with PKA or β-agonist infusion in live animals resulted in accelerated XB kinetics and rates of pressure development, respectively, the magnitude of acceleration was attenuated when Ser282 phosphorylation was blocked. Taken together, our results demonstrate that Ser282 phosphorylation does not regulate the phosphorylation status of Ser273 and Ser302, and that acceleration of pressure development and XB kinetics can be achieved in its absence, but full acceleration of contractility in vitro and in vivo probably requires phosphorylation of all three M-domain Ser residues.

Effects of Ser282 phospho-ablation on Ser273 and Ser302 phosphorylation

Early experiments suggested that initial phosphorylation of Ser282 permits subsequent phosphorylation at neighbouring serine residues, because bacterially derived recombinant cMyBP-C protein expressing serine to alanine substitutions at residue 282 nearly eliminated PKA-mediated phosphorylation of Ser273 and Ser302 (Gautel et al. 1995). This idea was later supported by experiments in both mutated recombinant cMyBP-C proteins and myofibrils isolated from TG mouse models expressing S282A cMyBP-C in the heart (Sadayappan et al. 2011), which showed decreased phosphorylation of Ser302 by PKA, PKC and CamKII in the presence of a non-phosphorylatable Ser282 (i.e. S282A). It is not clear, however, why in Gautel et al. (1995) phosphorylation of both Ser273 and Ser302 was severely blunted by S282A, whereas Sadayappan et al. (2011) found only the phosphorylation of Ser302 is abolished in the presence of S282A. In contrast, there is also evidence that phosphorylation of Ser273 and Ser302 could still be achieved in recombinant cMyBP-C proteins expressing a non-phosphorylatable Ser282 residue (i.e. S282A) (Bardswell et al. 2010; Cuello et al. 2011). A recent study (Bardswell et al. 2010) demonstrated that phosphorylation of Ser282 is not required for phosphorylation of other serine residues within the M-domain, as it was shown that protein kinase D can selectively phosphorylate Ser302 in recombinant cMyBP-C proteins expressing a S282A substitution. Similarly, Cuello et al. (2011) observed an increase in Ser273 and Ser302 phosphorylation following PKA treatment in recombinant cMyBP-C proteins expressing a S282A substitution, suggesting that phosphorylation at Ser273 and Ser302 is not necessarily predicated on the phosphorylation status of Ser282. In agreement with the latter studies, our present data show that abolishing Ser282 phosphorylation in the myocardium had minimal effects on the basal phosphorylation of Ser273 or Ser302 and did not affect PKA-mediated increases in phosphorylation of Ser273 and Ser302 (Fig. 2), indicating that Ser282 phosphorylation is not a prerequisite for PKA-mediated phosphorylation of neighbouring serine residues.

The basis for the differences between studies showing the functional effects of Ser282 phosphorylation is not known; however, it is possible that variability in the recombinant proteins utilized, differences in the experimental protocols of kinase phosphorylation assays and variations in the TG mouse models utilized may have contributed to the observed discrepancies in the results reported. For example, purified rabbit cMyBP-C was used by Gautel et al. (1995) and recombinant His-tagged human cMyBP-C C1C2 fragments were used by Bardswell et al. (2010) and Cuello et al. (2011). Furthermore, differences in the experimental conditions of the PKA phosphorylation assays between studies such as the source of the kinase, the concentration of PKA utilized and the length of PKA incubation, could have contributed to the divergent effects on cMyBP-C phosphorylation. There are also differences between the mouse models utilized in previous studies and the present study as the expression of TG S282A cMyBP-C in the hearts of mice employed in Sadayappan et al. (2011) was ∼25–45%, whereas in this study cardiac expression of TG S282A cMyBP-C was ∼85%. Additionally, minor differences in genetic background of the mice employed in different studies could underlie subtle observed differences in phosphorylation of cMyBP-C Ser residues (Tong et al. 2008, 2014), and any resulting functional differences.

Functional role of Ser282 phosphorylation in modulating cross-bridge kinetics

We studied the impact of TG expression of cMyBP-C with constitutively non-phosphorylatable Ser282 (S282A) on XB turnover kinetics by measuring k_tr, the rate constant of force redevelopment following a mechanical release–restretch manoeuvre (Brenner & Eisenberg, 1986). k_tr is the sum of the forward (f) and reverse (g) rate constants of the transition of XBs from weak, non-force-bearing state to a strong, force-bearing state (Brenner & Eisenberg, 1986; Campbell, 1997) and can be used as an index of XB turnover kinetics in the myocardium. At low levels of Ca²⁺ activation, force development in cardiac muscle is highly dependent on XB-mediated cooperative activation where initial strong binding of XBs to actin cooperatively recruits additional XBs into strongly bound states, thereby further activating the thin filament. XB-induced cooperative recruitment of additional XBs tends to slow the overall rate of force development at low levels of Ca²⁺ activation (Campbell, 1997), whereas at high levels of Ca²⁺ activation the transition of XBs into the strongly bound state is less reliant on XB-mediated cooperativity, as more thin filament regulatory units are directly activated by the binding of Ca²⁺ to troponin C. An acceleration of k_tr in skinned myocardium following PKA treatment is not a universal finding as some studies showed no change in XB turnover kinetics following PKA treatment (Hofmann & Lange, 1994; Janssen & de Tombe, 1997; Walker et al. 2011), and some other studies showed a decrease in XB turnover kinetics following PKA treatment (Hanft & McDonald, 2009, 2010). However, in agreement with previous studies from our laboratory, and others (Stelzer et al. 2006b; Bardswell et al. 2010; Chen et al. 2010; Cuello et al. 2011; Cheng et al. 2013), here we show that PKA treatment of WT-skinned myocardium significantly accelerated k_tr at 50% of maximal Ca²⁺ activation when compared to the non-treated preparations (Fig. 5). The disparities in the reported effects of PKA phosphorylation on k_tr are unclear; however, they may be attributed to differences in the myocardial preparations, the temperature at which the k_tr measurements were assessed, and the concentrations of the PKA used for incubating the cardiac preparations. For example, Hanft & McDonald (2009, 2010) used single myocytes, conducted the experiments at 13°C and incubated their preparations with 0.125 U μl⁻¹ of PKA for 45 min. On the other hand, in the present study we used multicellular myocardial preparations, performed the k_tr measurements at 22°C and incubated our preparations with 0.250 U μl⁻¹ of PKA for 1 h. In the physiological context, slowed k_tr following PKA treatment (Hanft & McDonald, 2010) has been interpreted to result from enhanced cooperative activation of the thin filaments, which would progressively enhance the number of force-producing XBs working against a constant afterload resulting in faster loaded shortening and, therefore, enhanced systolic ejection and increased stroke volume (McDonald et al. 2012). In contrast, results from the present study are consistent with the idea that PKA treatment accelerates k_tr due to acceleration in the rate of cooperative activation of the thin filament and transitions of XBs into the force-producing state, which results in faster pressure development during isovolumic contraction, thereby leaving more time for systolic ejection in vivo, which enhances stroke volume and cardiac output.

Our results show that basal k_tr at 50% maximal Ca²⁺activation was not different between TG^S282A-skinned myocardium and TG^WT-skinned myocardium (Fig. 5), demonstrating that basal XB kinetics were not affected by Ser282 phospho-ablation. Basal k_tr values were also not significantly different between TG^S282A and TG^WT myocardium at all levels of Ca²⁺ activation (data not shown). Following PKA treatment, k_tr in TG^S282A-skinned myocardium was significantly accelerated compared to baseline; however, the acceleration in k_tr was modestly (16%) but significantly less compared to TG^WT-skinned myocardium (Fig. 5), suggesting that a complete PKA-mediated acceleration of k_tr requires Ser282 phosphorylation.

We further investigated the functional effects of Ser282 phospho-ablation in the myocardium on dynamic strain-dependent XB behaviour using a stretch activation mechanical perturbation in which a small stretch (2% of initial muscle length) is applied to an otherwise isometrically contracting muscle fibre (Stelzer et al. 2006c). The initial rise in force due to acute stretch (i.e. P₁) is related to the strain of strongly bound XBs, and can be indicative of XB stiffness (Cheng et al. 2013). Following PKA treatment P₁ values were decreased in all groups (Table 4) despite similar levels of pre-stretch steady-state force before and after PKA treatment, suggesting that PKA phosphorylation reduced XB stiffness. Furthermore, PKA phosphorylation resulted in greater XB detachment as displayed by the decline in the amplitude (i.e. P₂) and rate of force relaxation (i.e. k_rel) following acute stretch (Table 4), which is also consistent with an increase in XB compliance in response to increased strain. These results are consistent with previous studies showing that cMyBP-C and its phosphorylation are important for modulation of radial and longitudinal stiffness in the sarcomere (Palmer et al. 2011; Cheng et al. 2013).

Phospho-ablation of Ser282 did not affect the amplitudes of P₁, P₂ or the acceleration of k_rel following PKA treatment (Table 4); however, TG^S282A-skinned myocardium displayed a reduced acceleration in stretch-induced delayed XB recruitment (i.e. k_df) compared to TG^WT-skinned myocardium (Table 4). This would suggest that the attenuation of PKA-mediated acceleration in TG^S282A-skinned myocardium is probably due to impaired XB recruitment and transitions to force generating states resulting in an overall slowing of the stretch activation response (Stelzer et al. 2007). The diminished acceleration of k_df in TG^S282A-skinned myocardium in response to PKA treatment compared to TG^WT-skinned myocardium nearly mirrored the differences observed in k_tr measurements between the two groups suggesting that slower k_tr in PKA-treated TG^S282A-skinned myocardium was probably due to incomplete acceleration of XB recruitment and transition to force generating states (the forward rate constant, f) rather than changes in the rate of XB detachment (the reverse rate constant, g). Our results suggest that because Ser282 phosphorylation does not modulate the phosphorylation of Ser273 and Ser302, phosphorylation of these residues can accelerate XB kinetics independent of Ser282 phosphorylation. In agreement, Bardswell et al. (2010) reported that phosphorylation of Ser302 by protein kinase D (which does not phosphorylate Ser273 or Ser282) can significantly accelerate XB kinetics at 50% maximal activation. If phosphorylation of each site can contribute to the modulation of XB kinetics, increasing phosphorylation levels at any of these three residues may be sufficient to accelerate XB kinetics above basal levels. Our finding that XB kinetics can be accelerated in the absence of Ser282 phosphorylation does not mean that Ser282 is not functionally important. In agreement with previous studies (Wang et al. 2014) we did not observe differences in basal XB kinetics between TG^S282A and TG^WT-skinned myocardium; however, Ser282 phospho-ablation diminished the acceleration of XB kinetics in response to PKA treatment. Although the contribution of Ser282 to this response was modest, because force generation in cardiac muscle is a highly cooperative process, even a relatively small decrease in the rate of XB recruitment in response to PKA phosphorylation can significantly amplify the functional response to increased β-adrenergic stimulation in vivo. The functional significance of cMyBP-C residue-specific modulation of cardiac muscle contraction is unclear, but may be important in conditions of heart failure where a downregulation of β-adrenergic signalling may result in increased activation of non-PKA kinases (Bardswell et al. 2012), which can target individual cMyBP-C residues, thereby providing an alternate mechanism for regulating myofilament function in response to acute sympathetic activation.

In vivo consequences of Ser282 phospho-ablation

The importance of cMyBP-C phosphorylation for normal in vivo cardiac function is underscored by studies employing animal models in which all three phosphorylatable serine residues were mutated to alanine residues (i.e. TG^3SA), which show impaired systolic and diastolic cardiac function coupled with compensatory hypertrophy (Sadayappan et al. 2005; Tong et al. 2008). Furthermore, cardiac haemodynamic assessment in mice expressing non-phosphorylatable cMyBP-C revealed that cMyBP-C phosphorylation is critical for systolic preload-dependent early pressure rise and rate-dependent and adrenergic contractile reserve in response to dobutamine administration (Nagayama et al. 2007). In this study, TG^S282A hearts did not display noticeable hypertrophy, as left ventricular mass and wall thickness were not different from NTG and TG^WT hearts (Table 1). The lack of pathological hypertrophy was further confirmed by examination of formalin-fixed hearts and cardiac sections stained with H&E that revealed no alteration in wall size or overall morphology. Additionally, no significant differences were observed in basal cardiac contractile function as assessed by echocardiography (ejection fraction and fractional shortening), and P–V catheterization (maximal pressure development, dP/dt_max or ventricular relaxation, τ) in TG^S282A mice when compared to NTG and TG^WT mice (Table 2). This is consistent with a previous report in which a similar model of Ser282 phospho-ablation did not result in altered basal systolic function or result in increased heart weight/body weight ratios, although that study reported an increase in wall thickness (Sadayappan et al. 2011).

The rate of early (isovolumic) contraction is dependent on the rate of thin filament activation by both Ca²⁺ and strongly bound XBs (reviewed in Hanft et al. 2008) such that an impaired XB recruitment can significantly affect the magnitude and rate of pressure development in vivo. Dobutamine accelerates the maximum rate of early pressure development (dP/dt_max, Table 2) in part by increasing the rate of XB-induced cooperative activation of the thin filament through phosphorylation of cMyBP-C (Nagayama et al. 2007). In this study, we observed an increase in dP/dt_max compared to baseline values after dobutamine administration in all of the groups; however, dP/dt_max was significantly slower in TG^S282A hearts compared to TG^WT controls (Table 2) after dobutamine, consistent with the in vitro attenuation of accelerated XB recruitment in TG^S282A myocardium. The attenuation of accelerated pressure development in TG^S282A was observed not only as a slower dP/dt_max but was also evident as a slower submaximal dP/dt, as demonstrated by plotting the rate of pressure development over various levels of developed pressure before ejection (Fig. 4A). In both TG^WT and TG^S282A animals dobutamine administration significantly influenced the relationship between dP/dt and developed pressure as demonstrated by an upward shift of the curve, reflecting an increase in the contractile state of the myocardium (Fig. 4); however, the shift was blunted in TG^S282A hearts compared to WT controls (Fig. 4) reflecting an impaired enhancement of pressure development after dobutamine due to Ser282 phospho-ablation. Changes in EDP can have modest effects on the relationship between dP/dt and developed pressure (an increase in preload will enhance the rate of pressure development) (Mason et al. 1971), but this probably did not influence our results as TG^S282A and TG^WT hearts had no differences in EDP under similar conditions. The finding that Ser282 phosphorylation modulates the rate of early pressure development is consistent with previous findings (Nagayama et al. 2007) highlighting the importance of cMyBP-C phosphorylation-mediated regulation of cooperative XB activation of the thin filament to accelerating pressure development during isovolumic contraction in response to dobutamine. The acceleration in pressure development after dobutamine administration results in dP/dt_max being reached sooner when compared to baseline, consistent with an overall shortening of the cardiac cycle. By measuring the time to dP/dt_max (t_d), a measure of the contractile state of the myocardium independent of preload (Adler et al. 1996a,b), we were able to assess whether Ser282 phospho-ablation would prolong the length of time required to reach dP/dt_max in TG^S282A animals. At baseline, there were no differences in t_d between TG^WT and TG^S282A hearts (Fig. 4B). After dobutamine administration, t_d was significantly shorter in both the TG^S282A and TG^WT hearts; however, the reduction in t_d was blunted in TG^S282A hearts compared to TG^WT hearts (i.e. t_d was significantly prolonged in TG^S282A). The differences in t_d between TG^WT and TG^S282A hearts were not due to differences in HR as there were no differences between groups either at baseline or following dobutamine infusion, and t_d was still significantly slower in TG^S282A compared to TG^WT post-dobutamine when normalized to cardiac cycle duration (data not shown). Our results suggest that having two functional phosphorylation residues may be sufficient to maintain cardiac function under normal conditions; however, in conditions of increased cardiac workload (such as with increased β-adrenergic stimulation), all three phosphorylated sites may be required to fully accelerate pressure development. In agreement with this idea, other studies show that mutating more than one serine residue, whether to mimic phospho-ablation or pseudo-phosphorylation, can result in cardiac contractile dysfunction and maladaptive remodelling (Sadayappan et al. 2005; Tong et al. 2008; Sadayappan et al. 2011; Gupta et al. 2013).

Potential molecular and structural implications of Ser282 phosphorylation

The molecular basis for the PKA-mediated acceleration of XB kinetics in cardiac muscle is related to the phosphorylation-dependent reversible interactions of cMyBP-C with both myosin S2 and actin. In the non-phosphorylated state, the cMyBP-C's N-terminal C0–C2 domains interact with myosin S2 and actin (Gruen et al. 1999; Shaffer et al. 2009) and PKA phosphorylation of serines in the M-domain reduces these interactions (Gruen et al. 1999; Shaffer et al. 2009). Phosphorylation of cMyBP-C relieves the sterical constraint imposed by cMyBP-C on the myosin heads and accelerates XB interactions with actin (Gruen et al. 1999; Previs et al. 2012), in part due to an increase in the radial displacement of XBs towards actin, which enhances the probability of actomyosin interactions (Colson et al. 2008, 2012). Phosphorylation of cMyBP-C also reduces the binding affinity of cMyBP-C for actin (Shaffer et al. 2009) thereby reducing the inhibition on myofilament sliding velocity allowing for accelerated XB cycling (Shaffer et al. 2009; Weith et al. 2012; Coulton & Stelzer, 2012). Thus, through decreased binding and inhibition of cMyBP-C interactions with myosin and actin, PKA-mediated phosphorylation of cMyBP-C provides a mechanism for accelerating XB kinetics in the myocardium. However, the role of phosphorylation of individual M-domain serine residues in the cMyBP-C-mediated regulation of actomyosin interaction is still not completely understood, and there is no definitive structural model that describes the M-domain. It has been proposed that Ser273 and Ser282 are in close proximity to each other on a β-strand of the immunoglobulin-like core of the M-domain directly opposite a patch of negatively charged amino acids on the surface of myosin S2 (Blankenfeldt et al. 2006; Ababou et al. 2008) that is a part of the cMyBP-C-myosin S2 interaction interface, while Ser302 lies on the opposite side of the β-strand and in a loop that extends away from the immunoglobulin-like core region (Jeffries et al. 2008). It was proposed that the introduction of negative charges due to phosphorylation of Ser273 and Ser282, which lie opposite to myosin S2 will repel this region away from myosin S2 and result in a structural rearrangement in the core region of the M-domain, whereas subsequent phosphorylation of Ser302 may result in movement of the opposing loop region (Jeffries et al. 2008). Furthermore, another recent structural study employing NMR and modelling approaches predicted that phosphorylation of Ser273 primarily stabilizes and extends the helical structure of the M-domain, whereas phosphorylation of Ser282 and Ser 302 has little effect on the M-domain's helical stability (Howarth et al. 2012), suggesting that phosphorylation of each site could have a disparate impact on the structural conformation of the M-domain. In the context of findings from our study, because phospho-ablation of Ser282 did not prevent the effects of PKA phosphorylation on XB kinetics (Table 4; Figs 5 and 6), our results are consistent with structural models that predict that the effects of cMyBP-C phosphorylation on the M-domain structure may be specific to the individual serine residue that is phosphorylated (Howarth et al. 2012), rather than a hierarchical model in which initial phosphorylation of Ser282 dictates structural rearrangements that allow subsequent phosphorylation of other Ser residues (Gautel et al. 1995). Thus, based on previous results (Bardswell et al. 2010; Cuello et al. 2011) and data presented here, it is possible that each of the three individual M-domain serine phosphorylation residues are functionally important and capable of modulating contractile function independently of the phosphorylation status of neighbouring serine residues.

Glossary

cMyBP-C

cardiac myosin binding protein-C

dP/dt

rate of LV pressure development

dP/dt_max

peak rate of left ventricular pressure development

EDP

end-diastolic pressure

F_max

maximum Ca²⁺ activated force

F_min

Ca²⁺ independent force

H&E

haematoxylin and eosin stain

HR

heart rate

HSC70

heat shock chaperone 70

k_df

rate constant of delayed force development

k_rel

rate constant of force decay

k_tr

rate constant of force redevelopment

LV

left ventricle

n_H

Hill coefficient

NTG

non-transgenic

pCa₅₀

[Ca²⁺] required for 50%-maximum activation

PKA

protein kinase A

P

submaximal force

P_ο

maximal force

P–V

pressure–volume

τ

rate constant of pressure relaxation

RLC

regulatory light chain

TG

transgenic

TG^S282A

mice expressing transgenic cMyBP-C with a serine to alanine mutation at residue 282

TG^WT

mice expressing transgenic WT cMyBP-C

TnI

troponin I

TnT

troponin T

WT

wild-type

XB

cross-bridge

Additional information

Competing interests

There are no conflicts of interest.

Author contribution

All experiments were performed at the Department of Physiology and Biophysics, Case Western Reserve University, Cleveland, Ohio, USA. K.S.G., R.M. and J.E.S participated in performing the experiments and data collection. K.S.G. and J.E.S. contributed to the conception and design of the experiments, and K.S.G., R.M. and J.E.S. contributed to the analysis and interpretation of data and in writing and revising the article. All authors approved the final version of the manuscript.

Funding

This work was supported by the National Heart, Lung, and Blood Institute Grant (HL-114770-01).