Research ArticleCARDIAC FUNCTION

Cardiac myosin binding protein-C Ser302 phosphorylation regulates cardiac β-adrenergic reserve

+ See all authors and affiliations

Science Advances  10 Mar 2017:
Vol. 3, no. 3, e1602445
DOI: 10.1126/sciadv.1602445

Abstract

Phosphorylation of cardiac myosin binding protein-C (MyBP-C) modulates cardiac contractile function; however, the specific roles of individual serines (Ser) within the M-domain that are targets for β-adrenergic signaling are not known. Recently, we demonstrated that significant accelerations in in vivo pressure development following β-agonist infusion can occur in transgenic (TG) mouse hearts expressing phospho-ablated Ser282 (that is, TGS282A) but not in hearts expressing phospho-ablation of all three serines [that is, Ser273, Ser282, and Ser302 (TG3SA)], suggesting an important modulatory role for other Ser residues. In this regard, there is evidence that Ser302 phosphorylation may be a key contributor to the β-agonist–induced positive inotropic responses in the myocardium, but its precise functional role has not been established. Thus, to determine the in vivo and in vitro functional roles of Ser302 phosphorylation, we generated TG mice expressing nonphosphorylatable Ser302 (that is, TGS302A). Left ventricular pressure-volume measurements revealed that TGS302A mice displayed no accelerations in the rate of systolic pressure rise and an inability to maintain systolic pressure following dobutamine infusion similar to TG3SA mice, implicating Ser302 phosphorylation as a critical regulator of enhanced systolic performance during β-adrenergic stress. Dynamic strain–induced cross-bridge (XB) measurements in skinned myocardium isolated from TGS302A hearts showed that the molecular basis for impaired β-adrenergic–mediated enhancements in systolic function is due to the absence of protein kinase A–mediated accelerations in the rate of cooperative XB recruitment. These results demonstrate that Ser302 phosphorylation regulates cardiac contractile reserve by enhancing contractile responses during β-adrenergic stress.

Keywords
  • cross-bridge kinetics
  • cardiac contractile function
  • hemodynamic function
  • contractile reserve

INTRODUCTION

Activation of the β-adrenergic signaling pathway rapidly enhances ventricular contraction to increase cardiac output in response to enhanced systemic demand, via protein kinase A (PKA) activation, which phosphorylates key myofilament proteins involved in the regulation of myocardial force generation (1, 2). Previous studies have demonstrated that cardiac myosin binding protein-C (MyBP-C) phosphorylation regulates myofilament function by accelerating the rates of cross-bridge (XB) cycling and force generation in response to PKA treatment (3, 4). Animal models expressing phospho-ablated MyBP-C display an abolished PKA-mediated enhancement of the magnitude and rate of cooperative XB recruitment and force generation in skinned myocardium (5, 6), which limits the magnitude and the acceleration of left ventricular pressure generation in response to an acute β-agonist infusion (68). Thus, increased MyBP-C phosphorylation appears to be a primary mechanism by which the heart activates its contractile reserve under conditions of increased cardiac stress (7).

Despite growing evidence demonstrating the importance of MyBP-C phosphorylation in regulating in vivo cardiac contractile function, the respective roles of individual MyBP-C phosphorylation residues are not yet clear. In addition to PKA, which targets four M-domain MyBP-C residues (Ser273, Ser282, Ser302, and Ser307 mouse MyBP-C numbering) (913), individual MyBP-C residues are also selectively targeted by non-PKA kinases (9, 14). Previous studies have demonstrated a role for Ser282 phosphorylation in accelerating XB recruitment and force generation (15, 16), which contributes to the in vivo acceleration of systolic pressure development following β-adrenergic stimulation (16, 17). However, significant acceleration of in vivo pressure development was observed in Ser282 phospho-ablated mice following infusion of the β-agonist dobutamine, whereas a complete MyBP-C phospho-ablation (that is, Ser273, Ser282, and Ser302) abolished the acceleration in the rate of pressure development (6, 7, 16), demonstrating that phosphorylation of other PKA-targeted MyBP-C residues (Ser273 or Ser302) must be critical for modulation of the contractile response to increased β-adrenergic stimulation.

In addition to PKA, previous studies have shown that Ser302 can be phosphorylated in vitro by several kinases, which are known to modulate cardiac function, such as PKC-ɛ (18), PKD (19), and glycogen synthase kinase-3β (20). Furthermore, previous studies have suggested that increased Ca2+-dependent protein kinase II (CaMKII) activity due to increased cardiac pacing frequency enhances cardiac contractility, in part, due to increased Ser302 phosphorylation (21). We have previously demonstrated that Ser302 phosphorylation was lower than Ser273 or Ser282 phosphorylation at baseline and was increased by PKA treatment to a greater extent than Ser273 and Ser282 phosphorylation (7, 16). Thus, our hypothesis is that among MyBP-C phosphorylation residues, Ser302 phosphorylation may be the principal MyBP-C residue, which modulates the inotropic contractile response to enhanced β-adrenergic activation, and that preventing Ser302 phosphorylation impairs β-agonist–mediated enhancements in contractile response and reduces cardiac contractile reserve. However, to date, no study has examined the specific contribution of Ser302 phosphorylation to in vivo contractile function. Therefore, to elucidate the precise roles of Ser302 phosphorylation in modulating myofilament and in vivo contractile function, we generated a novel transgenic (TG) mouse model expressing MyBP-C with a nonphosphorylatable Ser302 (that is, Ser302 to Ala302; TGS302A) and quantified the in vitro contractile response of skinned myocardium isolated from TGS302A hearts to PKA treatment and in vivo contractile function of TGS302A mice in response to β-agonist infusion.

RESULTS

Sarcomeric protein expression and phosphorylation

To examine the role of MyBP-C Ser302 phosphorylation in regulating β-adrenergic enhancement of ventricular contraction and relaxation, we generated a novel TG mouse line in which Ser302 was mutated to an Ala residue to abolish phosphorylation (Fig. 1A). Expression of TG MyBP-C in TGS302A myocardial samples was confirmed by Western blot analysis (Fig. 1B) and was detected as a full-length MyBP-C protein band at a molecular mass of ~150 kDa. Ser273 and Ser282 phosphorylation was detected in TGS302A samples and was similar to non-TG (NTG) phosphorylation (TGS302A Ser273 phosphorylation was 106 ± 6% and Ser282 phosphorylation was 104 ± 14% of NTG phosphorylation; n = 5), but Ser302 phosphorylation was absent, confirming Ser302 phospho-ablation in TGS302A mice (Fig. 1B). TG MyBP-C expression in TGS302A mice was determined to be 79 ± 15% of NTG MyBP-C expression levels, similar to expression levels of MyBP-C TG mice in previous studies (6, 16). MyBP-C phosphorylation at Ser273, Ser282, and Ser302 was not detected in TGS3A mice, as previously reported (Fig. 1B) (6, 7). The percentage of β-myosin heavy chain (MHC) expression in myocardium isolated from TGS302A (3.6 ± 1.5%) was found to be similar to NTG myocardium (2.2 ± 1.2%), whereas β-MHC levels were slightly elevated in TG3SA myocardium compared with NTG and TGS302A myocardium (8.3 ± 1.7%; P < 0.05) (Fig. 1C). The expression and phosphorylation of other sarcomeric proteins, including troponin I (TnI), TnT, and regulatory light chain (RLC), were similar among all three groups (Fig. 1, D and E).

Fig. 1 Determination of expression, PKA-mediated phosphorylation of MyBP-C, and other sarcomeric proteins.

(A) MyBP-C is composed of eight immunoglobulin (ovals) and three fibronectin III (rectangles) domains labeled C0 to C10 (N to C terminus). The conserved M-domain in the linker between domains C1 and C2 contains three serines (S273, S282, S302; mouse numbering) in the NTG sequence that are targets for PKA phosphorylation. The substitution used to prevent S302 phosphorylation in TGS302A is shown in red. (B) Representative Western blot showing S273, S282, and S302 phosphorylation, before and after PKA incubation. No Ser302 expression was detected in TGS302A samples, whereas fully phosphorylatable Ser273 and Ser282 were observed. (C) Representative 5% tris-HCl gel, stained with silver stain, showing MHC isoform expression in the myocardial samples. (D) Representative gels shown are stained by Pro-Q (left) for protein phosphorylation, and the same gel is shown for total protein (right) stained with Coomassie Blue. cMyBP-C, cardiac myosin binding protein-C; pMyBP-C, phosphorylated cardiac myosin binding protein-C. (E) Relative protein phosphorylation (phosphorylated signal/total protein signal) was calculated for each protein and is expressed as % of PKA-treated NTG values for that protein. Values are expressed as means ± SEM, from three to six hearts in each group. *P < 0.05, different from non–PKA-treated samples from the same line; P < 0.05, different from the corresponding NTG group.

Following PKA treatment, Ser273 and Ser282 phosphorylation levels were similar between TGS302A and NTG samples (TGS302A Ser273 phosphorylation was 97 ± 10% and Ser282 phosphorylation was 95 ± 15% of PKA-treated NTG phosphorylation; n = 3 to 5), demonstrating normal PKA-mediated phosphorylation of these sites in the absence of Ser302 phosphorylation (Fig. 1B). Ser302 phosphorylation was not detected in TGS302A samples following PKA treatment, confirming Ser302 phospho-ablation. TnI and TnT phosphorylation levels were similar between TGS302A and NTG samples following PKA treatment. Phosphorylation of RLC was unaffected by PKA treatment and was similar between TGS302A and NTG samples under all conditions. Ser273, Ser282, and Ser302 phosphorylation was not detected in TG3SA samples following PKA treatment. These results demonstrate that TG expression of S302A MyBP-C does not disrupt phosphorylation of other MyBP-C Ser residues or phosphorylation of other sarcomeric proteins.

Assessment of cardiac morphology

Representative formalin-fixed hearts and cross sections for NTG, TG3SA, and TGS302A hearts are presented in Fig. 2A. TGS302A hearts exhibited similar overall size and morphology with no noticeable alterations in chamber geometry or increased cardiac fibrosis (Fig. 2B) when compared to control NTG hearts. In contrast to the cardiac hypertrophy produced by complete MyBP-C phospho-ablation (Fig. 2 and Table 1), left ventricle (LV) mass and wall thickness were similar between TGS302A and NTG mice as measured by echocardiography. These results demonstrate that Ser302 phospho-ablation does not result in pathological cardiac hypertrophy.

Fig. 2 Analysis of cardiac morphology.

(A) Representative NTG, TG3SA, and TGS302A formalin-fixed hearts. TGS302A hearts showed a similar overall morphology compared to NTG hearts, in contrast to the enlarged heart size observed in TG3SA hearts. Scale bar, 1 mm. (B) Representative cross sections of NTG, TG3SA, and TGS302A hearts from the mid-LV were stained with Masson’s trichrome and imaged at ×100 magnification, demonstrating no observable increase in fibrosis in TGS302A hearts. Scale bars, 25 μm.

Table 1 LV morphology and in vivo cardiac performance measured by echocardiography.

BW, body weight; LV mass/BW, ratio of LV and body weight; HR, heart rate; PWd, posterior wall thickness in diastole; PWs, posterior wall thickness in systole; IVRT, isovolumic relaxation time; EF, ejection fraction. Values are expressed as means ± SEM from 10 mice per group.

View this table:

Basal cardiac function

Cardiac contractile and hemodynamic function was assessed by echocardiography and pressure-volume (P-V) loop analysis to determine the functional effects of Ser302 phospho-ablation. No changes in systolic function or diastolic function as evidenced by measurements of ejection fraction (EF) or isovolumic relaxation time (IVRT) were detected in TGS302A mice compared to NTG controls (Table 1). Similarly, no differences were detected in the rate of pressure development (dP/dtmax) or relaxation (τ) at baseline between TGS302A mice and NTG controls (Table 2). In contrast, complete MyBP-C phospho-ablation in TG3SA hearts resulted in slowed IVRT and reduced EF at baseline (Table 1), consistent with previous studies (6, 7, 22).

Table 2 Left ventricular hemodynamic function measured by P-V loop analysis.

HR, heart rate; Pmax, maximal systolic pressure; EDP, end diastolic pressure; dP/dtmax, maximum rate of pressure development; τ, time constant of pressure relaxation; DOB, dobutamine. Values are expressed as means ± SEM. n = 9 for NTG, 8 for TG3SA, and 10 for TGS302A.

View this table:

β-Adrenergic reserve

The contribution of Ser302 phosphorylation to the contractile response to β-adrenergic stimulation was assessed in TGS302A mice following dobutamine infusion. The postdobutamine dP/dtmax in TGS302A mice was slower than in dobutamine-treated NTG mice (P < 0.005; Table 2 and Fig. 3A), and the rate of pressure development throughout early systole in dobutamine-treated TGS302A mice was lower when compared to the rate of pressure development observed in dobutamine-treated NTG mice (Fig. 3B), demonstrating that Ser302 phosphorylation is required to accelerate pressure development following β-adrenergic stimulation. The acceleration in pressure development in NTG mice following dobutamine increased early systolic developed pressure levels compared to baseline (Fig. 3), but the impaired acceleration in pressure development in TGS302A mice prevented an increase in developed pressure, resulting in developed pressure levels that were similar to NTG mice before dobutamine treatment (Fig. 3). Additionally, maximal systolic pressure was lower in TGS302A mice following dobutamine treatment compared to NTG controls (P < 0.005; Table 2). The impaired acceleration in pressure development and decreased maximal systolic pressure observed in TGS302A mice were also observed in TG3SA mice, consistent with our previous study (7). There were no differences in the rate or magnitude of pressure development following dobutamine treatment between TG3SA and TGS302A mice (Table 2), suggesting that phosphorylation of Ser273 and Ser282 cannot compensate for the loss of Ser302 phosphorylation following β-adrenergic stimulation. Similar to previous reports of individual MyBP-C Ser residue phospho-ablation (16), Ser302 phospho-ablation did not disrupt normal acceleration of diastolic pressure relaxation because dobutamine accelerated relaxation to a similar extent in TGS302A and NTG mice (Table 2), in contrast to TG3SA mice, which displayed slowed relaxation following dobutamine treatment.

Fig. 3 Analysis of the β-adrenergic acceleration in pressure development.

(A) Developed pressure, where end diastolic pressure was subtracted from instantaneous ventricular pressure, was plotted during early systolic contraction, to yield developed pressure. Dobutamine (DOB) increased the developed pressure in NTG mice starting at 8 ms after the start of contraction compared to baseline (P = 0.015) but did not increase developed pressure in TG3SA or TGS302A mice. Following dobutamine treatment, developed pressure in TGS302A mice was not higher than that in NTG mice before dobutamine treatment at any point examined (that is, NTG at baseline). (B) The rate of pressure development (dP/dt) was measured during early systole. Dobutamine accelerated dP/dt starting at 6 ms after the start of contraction in NTG (P < 0.005). In TGS302A mice, the acceleration in dP/dt was blunted compared to that in NTG mice and was only faster than baseline after 9 ms (P < 0.05). dP/dt in TGS302A mice was not faster after dobutamine than NTG dP/dt before dobutamine treatment (that is, NTG at baseline) at any time point examined. *P < 0.05, NTG + dobutamine versus NTG without dobutamine treatment; P < 0.05, TGS302A + dobutamine versus TGS302A without dobutamine treatment; P < 0.05, NTG + dobutamine versus TGS302A + dobutamine.

Basal myofilament contractile function

The effects of Ser302 phospho-ablation on myofilament contractile properties were assessed in skinned myocardium isolated from TGS302A hearts (Fig. 4, A to D). Stretch activation experiments were performed to measure dynamic strain-induced XB behavior. Figure 4 illustrates a typical stretch activation response in skinned myocardium isolated from NTG, TG3SA, and TGS302A hearts and highlights the parameters assessed in the resultant force transients following the imposed stretch. Under basal conditions, the rate of delayed force development following the imposed stretch (that is, kdf) and the magnitude of XB recruitment (that is, Pdf) were not different between the NTG and TGS302A groups, whereas TG3SA myocardium displayed diminished Pdf (Table 3). However, the rate of force decay (krel) in TGS302A myocardium was slower than that in NTG myocardium (P = 0.01) and was not different from TG3SA skinned myocardium (Table 3). Additionally, P2 values in TGS302A skinned myocardium were higher than those in NTG myocardium (P = 0.03) and were similar to TG3SA values, indicating that the magnitude of strain-induced XB detachment was lower in the TGS302A and TG3SA groups. No differences in Ca2+-activated maximal force (Fmax), Ca2+-independent force (Fmin), Ca2+ sensitivity of force development (pCa50), or cooperativity of force development (nH) were observed between the NTG, TG3SA, and TGS302A groups (Table 4), indicating that differences in dynamic XB behavior were not due to changes in Ca2+-mediated thin filament activation.

Fig. 4 Effect of PKA treatment on the stretch activation responses in NTG, TG3SA, and TGS302A skinned myocardium.

Representative force responses elicited by a sudden 2% stretch in muscle length (ML) in isometrically contracting (A) NTG, (B) TG3SA, and (C) TGS302A myocardial preparations before (black) and following (red) PKA incubation. (D) Representative stretch activation traces following PKA treatment for NTG (red), TG3SA (green), and TGS302A (blue) myocardial preparations. Following PKA treatment, NTG preparations exhibited faster XB contractile behavior, whereas TG3SA exhibited slower XB contractile behavior and TGS302A exhibited intermediate XB contractile behavior. In (A), highlighted are the important phases of the force transients and the various stretch activation parameters that are derived from the elicited force response (see Materials and Methods for explanation of each phase and parameters). PKA incubation accelerated both krel (P < 0.005) and kdf (P < 0.005) in NTG skinned myocardium. However, in the TGS302A skinned myocardium, the PKA-mediated acceleration in krel is less pronounced (P = 0.02) than that in the NTG coupled with no PKA-mediated accelerations in kdf. There were no PKA-mediated accelerations in krel and kdf in TG3SA skinned myocardium. AU, arbitrary units.

Table 3 Dynamic stretch-activation parameters measured in NTG, TG3SA, and TGS302A skinned myocardium.

Twelve preparations isolated from four hearts were used for all the groups. P1, XB stiffness; P2, magnitude of XB detachment; P3, the new steady-state force attained in response to the imposed stretch in ML; Pdf, magnitude of XB recruitment; P0, prestretch isometric force; krel, rate of XB detachment; kdf, rate of XB recruitment. Traces in the top row highlight the portion of the stretch activation trace that the particular parameter represents. Values are expressed as means ± SEM.


Embedded Image
View this table:
Table 4 Steady-state contractile parameters measured in NTG, TG3SA, and TGS302A skinned myocardium.

Fmin, Ca2+-independent force measured at pCa 9.0; Fmax, Ca2+-activated maximal force measured at pCa 4.5; nH, Hill coefficient of the force-pCa relationship; pCa50, pCa required for the generation of half-maximal force. Values are expressed as means ± SEM.

View this table:

Myofilament contractile function following PKA treatment

Following PKA treatment, both TGS302A and TG3SA skinned myocardium displayed slower rates of stretch-induced delayed force development (kdf) compared to NTG myocardium (Table 3 and Fig. 4D). PKA treatment accelerated kdf by ~65 ± 17% (P < 0.005) in NTG myocardium but had no effect on kdf in TGS302A or TG3SA myocardium, demonstrating that Ser302 phospho-ablation impairs PKA-mediated accelerations in the rate of XB recruitment to the same extent as complete MyBP-C phospho-ablation. Following PKA incubation, P3 was increased by ~43 ± 9% (P < 0.005) and ~25 ± 6% (P = 0.005) in the NTG and TGS302A groups but not in the TG3SA group (Fig. 5A). TGS302A skinned myocardium also displayed an impaired magnitude of XB recruitment (Pdf) following PKA treatment compared to NTG myocardium (Table 3). Whereas Pdf was increased by PKA incubation by ~84 ± 14% (P < 0.005), ~61 ± 12% (P < 0.005), and ~61 ± 9% (P < 0.005) in the NTG, TG3SA, and TGS302A groups (Fig. 5B), respectively (Table 3), Pdf was higher in the NTG group compared to the TGS302A group (P < 0.005), demonstrating that Ser302 phospho-ablation blunts the PKA-mediated enhancement in the overall magnitude of XB recruitment.

Fig. 5 Effect of PKA on stretch activation parameters.

Measurements in the skinned ventricular preparations were first made under basal conditions (no PKA). Measurements were again made on the same preparations following PKA incubation. The net changes in stretch activation parameters after PKA incubation were calculated and are expressed as % increases from baseline for (A) the magnitude of new steady-state force attained in response to stretch in ML, P3; (B) the magnitude of overall XB recruitment in response to stretch in ML, Pdf; and (C) the rate of XB detachment, krel in the NTG, TG3SA, and TGS302A groups. PKA treatment enhanced the Pdf in all the groups, but the PKA-mediated enhancements in Pdf were lower in the TG3SA and TGS302A groups when compared to the NTG group, indicating that the magnitude of XB recruitment occurs to a lesser extent in the TG3SA and TGS302A groups. NTG preparations displayed accelerations in krel following PKA incubation; however, this acceleration in krel was less pronounced in the TGS302A group and is completely absent in the TG3SA group. Similar effects were observed for P3. Values are expressed as means ± SEM. *P < 0.05; 12 preparations isolated from four hearts were used for all the groups.

In addition to enhancing the rate of delayed force development, PKA treatment also accelerated the rate and magnitude of force decay following stretch. In NTG myocardium, PKA treatment accelerated the rate of force decay (~80 ± 13% increase in krel; P < 0.005) and produced a greater magnitude of force decay (that is, more negative P2 values). However, the PKA-mediated acceleration in force decay was blunted in TGS302A myocardium (~45 ± 8% increase in krel; P < 0.005) and resulted in krel values similar to pre-PKA krel values in NTG skinned myocardium (Table 3), suggesting that Ser302 phosphorylation is required for complete acceleration of the rate of XB detachment following PKA treatment. PKA incubation also decreased P2 values in the TGS302A group but to a lesser extent than in the NTG group (Table 3). PKA treatment did not accelerate krel (Fig. 5C) or decrease P2 amplitude in TG3SA myocardium when compared to the NTG group, as previously reported (5).

PKA treatment similarly decreased the pCa50 in all the groups, and no differences were observed in Fmax, Fmin, and nH between groups before or following PKA treatment (Table 4). Therefore, our data indicate that Ser302 phospho-ablation does not affect the Ca2+-sensitivity of force generation and that the impaired acceleration in force decay and delayed force development in TGS302A skinned myocardium are due to a diminished PKA-mediated enhancement of XB kinetics in the absence of Ser302 phosphorylation.

DISCUSSION

Although augmented ventricular function during β-adrenergic stimulation requires MyBP-C phosphorylation, it is unclear how Ser302 phosphorylation contributes to enhanced myofilament and whole-heart contractile function. Therefore, to define the effects of Ser302 phosphorylation on in vitro and in vivo contractile function, we generated novel TG mice expressing nonphosphorylatable MyBP-C Ser302 (that is, TGS302A). Our findings show that Ser302 phospho-ablation did not disrupt PKA-mediated phosphorylation of neighboring MyBP-C Ser residues (that is, 273 and 282) or other sarcomeric proteins and did not induce significant maladaptive cardiac remodeling or altered basal ventricular function. However, Ser302 phospho-ablation largely prevented β-adrenergic enhancement of the rate of systolic pressure development and decreased maximal systolic pressure following β-adrenergic stimulation, demonstrating that Ser302 phosphorylation is the predominant MyBP-C residue responsible for regulation of cardiac systolic β-adrenergic reserve. Measurements of dynamic XB behavior revealed that loss of β-adrenergic reserve at the whole-heart level was mirrored by the loss of PKA-mediated accelerations in cooperative XB recruitment into the force-generating states at the myofilament level. Collectively, our data show that MyBP-C Ser302 phosphorylation is required to enhance force and pressure generation during enhanced β-adrenergic signaling and uniquely equips the heart with a functional reserve to meet higher circulatory demands during β-adrenergic stimulation.

Ser302 phosphorylation is required for complete cooperative recruitment of XBs during submaximal Ca2+ activation

MyBP-C Ser302 phosphorylation is relatively lower than Ser273 and Ser282 under basal conditions but is highly phosphorylated following PKA incubation (16, 23); therefore, it is expected that MyBP-C Ser302 phosphorylation substantially contributes to enhanced contractile performance following β-adrenergic stimulation. Our data show that TGS302A skinned myocardium displayed a slowed rate and diminished magnitude of XB detachment (krel and P2), along with a reduced magnitude of cooperative XB recruitment (Pdf) compared to NTG skinned myocardium, suggesting that ablation of even low levels of basal Ser302 phosphorylation impairs dynamic XB behavior (Table 3). Following PKA incubation, functional differences between TGS302A and NTG skinned myocardium were amplified. Although we observed PKA-induced accelerations in krel and enhancements in P2 (indicating increases in both the rate and magnitude of XB detachment in TGS302A skinned myocardium), the effects were smaller than NTG skinned myocardium (Table 3 and Fig. 4). Notably, unlike NTG skinned myocardium, following PKA incubation, TGS302A skinned myocardium did not display significant accelerations in kdf, an indicator of the rate of cooperative XB recruitment into the force-generating XB pool. This finding is in agreement with an earlier study, which showed that selective phosphorylation of Ser302 by PKD accelerates the rate of tension redevelopment in wild-type mouse myocardium (19). Additionally, PKA-induced enhancement in Pdf, which is an indicator of the magnitude of XB recruitment into the force-generating XB pool, was also blunted in TGS302A skinned myocardium when compared to the NTG skinned myocardium, indicating that Ser302 phosphorylation modulates the overall number of cooperatively recruited force-generating XBs during PKA-mediated myofilament activation. Previous studies using intact mouse papillary fibers showed that increased pacing frequency (an activator of CaMKIIδ, which targets Ser302 primarily and Ser282 to a lesser degree) accelerates the rates of force development and relaxation in wild-type fibers but these responses are diminished in TG3SA fibers, indicating that phospho-ablation of MyBP-C depresses CaMKIIδ-mediated enhancements in contractile function (21). However, under normal physiological conditions, the immediate/short-term demand for increased cardiac inotropy is primarily met by PKA activation, which provides an acute boost in contractility due to phosphorylation of all three Ser residues, whereas the need for prolonged increases in cardiac contractile function [that is, prolonged increases in pacing frequency (21)] may be met by increased CaMKII activation, which phosphorylates only a subset of Ser residues (24). Thus, it is likely that PKA-mediated Ser302 phosphorylation is the main contributor to β-adrenergic stimulation–mediated enhancements in cardiac contractility under normal physiological conditions.

The magnitude of cooperative XB recruitment and binding to the thin filaments is influenced both by Ca2+ binding to TnC and by XB binding to actin. Ca2+ binding to TnC rapidly turns on the thin filament regulatory units and opens up the myosin binding sites on actin by displacing the tropomyosin to facilitate XB binding, which allows for propagation of XB-mediated cooperative XB recruitment along neighboring thin filament regulatory units (25). Cooperative XB-mediated recruitment of additional XBs is further facilitated by PKA phosphorylation of MyBP-C (26), which accelerates XB transitions to force-bearing states and enhances the probability of actomyosin interactions by displacing and orienting the XBs away from the thick filament backbone toward the actin sites (27) and also by reducing XB on time, which allows for accelerated XB turnover and recruitment of additional XBs to vacated myosin binding sites on actin (26). The mechanism by which MyBP-C phosphorylation enhances cooperative XB recruitment may be due to enhanced thin filament activation upon binding of phosphorylated MyBP-C to actin, which induces displacement of tropomyosin to the open state (28), or due to reduced binding of MyBP-C to myosin S2, which facilities XB binding to actin (2932), or both. Thus, at submaximal Ca2+ activations, MyBP-C phosphorylation not only accelerates the rate of force development but also expands the spread of thin filament activation beyond regulatory units that have already been activated by Ca2+, thereby shifting the net equilibrium of myosin XBs toward the force-bearing states and greatly enhancing myocardial force generation (26). In contrast, a reduction in MyBP-C Ser302 phosphorylation would be expected to slow cooperative XB recruitment and turnover, thereby disrupting the equilibrium of XB transitions into the force-generating pool, which in turn limits systolic pressure generation during β-adrenergic stimulation. However, as the level of Ca2+-activation increases, myofilament force generation becomes less reliant on cooperative XB recruitment because more thin filament regulatory units are directly activated by Ca2+, and thus, the modulatory influence of MyBP-C phosphorylation on XB behavior diminishes (5, 12).

MyBP-C Ser302 phosphorylation is required to enhance systolic pressure development

Although all three M-domain MyBP-C residues are targets of PKA phosphorylation, Ser302 phosphorylation is observed at low levels in vivo and is lower than the other two PKA-targeted sites, Ser273 and Ser282 (16, 23). Because the pool of MyBP-C that can be phosphorylated at Ser302 following β-adrenergic stimulation (~95% of total MyBP-C) is larger than for Ser273 and Ser282, loss of phosphorylation at Ser302 would be expected to reduce the recruitable pool of force-generating XBs to a greater extent than loss of phosphorylation at the other PKA sites in response to β-adrenergic signaling. The rate of pressure development and maximal developed pressure were not different between TG3SA and TGS302A mice following dobutamine treatment, suggesting that phosphorylation of Ser302 was needed to recruit the reserve force-generating XBs to enhance pressure development above basal levels. Previous studies have suggested that in addition to the known PKA phosphorylation residues Ser273, Ser282, and Ser302, a fourth residue, Ser307, may also be a target of PKA phosphorylation (913). However, TG3SA hearts did not exhibit systolic enhancements following dobutamine treatment despite expressing a phosphorylatable Ser307 residue, suggesting that Ser307 is unlikely to be involved in modulating the molecular and in vivo behaviors that we measured in this study. Together, our data suggest that Ser302 is the primary MyBP-C residue that modulates the positive inotropic response to enhanced β-adrenergic stimulation.

By preventing the recruitment of additional XBs following β-adrenergic stimulation, Ser302 phospho-ablation slowed pressure development and reduced maximal systolic pressure development without disrupting basal systolic pressure development. Loss of phosphorylation at Ser302 did not produce pathological cardiac hypertrophy or notable changes in basal unstressed ventricular function, perhaps because the primary role of Ser302 phosphorylation is to promote greater XB recruitment at higher workloads when β-adrenergic signaling is enhanced. This is also consistent with the observation that myocardial force generation is unaltered at baseline but cannot be enhanced by PKA or CaMKII (which targets Ser302) in myocardium expressing nonphosphorylatable MyBP-C (21) because the reserve pool of recruitable XBs is reduced by Ser302 phospho-ablation.

It was proposed that Ca2+ is released from ryanodine receptors near the Z line in myocytes and diffuses toward the center of the sarcomere (13): first, activating thin filament regulatory units in the region of the sarcomere that does not contain MyBP-C and, subsequently, activating the thin filament in MyBP-C–containing regions. Increased Ca2+ release following β-adrenergic stimulation would be expected to increase XB recruitment, in part, by increasing thin filament activation closer to the center of the sarcomere, in the region of the sarcomere regulated by MyBP-C. Thus, MyBP-C phospho-ablation would limit the recruitment of additional XBs in this region of the sarcomere following β-adrenergic signaling, thereby reducing the pool of recruitable XBs compared to the myofilaments expressing phosphorylatable MyBP-C. MyBP-C phosphorylation– and Ca2+-dependent regulation of XB recruitment could also explain the temporal pattern of enhanced pressure development following β-adrenergic signaling (Fig. 3). The recruitment of XBs not regulated by MyBP-C would produce an initial increase in pressure as cytosolic Ca2+ levels increase, followed by dobutamine-mediated acceleration of pressure development (observed after several milliseconds) due to recruitment of newly available XBs following MyBP-C phosphorylation to force-generating states.

Previously, it has been suggested that cardiac thick filaments harbor a subpopulation of super-relaxed (SRX) myosin heads that are characterized by substantially slower adenosine 5′-triphosphate (ATP) turnover rates (33, 34) and thereby equipping the cardiac contractile apparatus with a reserve of inactive myosin heads that can potentially be recruited into the force-bearing state under conditions of increased cardiac stress (35). Previous findings demonstrate that MyBP-C modulates the SRX state of myosin heads via its ability to interact with and position the myosin heads closer to the thick filament backbone and that a loss of MyBP-C decreases the number of SRX myosin heads (36). Thus, ablation of Ser302 phosphorylation will potentially maintain the interactions of MyBP-C–SRX heads and possibly prevent them from being recruited toward the myosin binding sites on actin filaments, effectively reducing the number of XBs that can be recruited during thin filament activation. This suggests that a possible role of MyBP-C is to regulate the recruitable pool of XBs to fine-tune force generation during a cardiac twitch, with enhanced β-adrenergic signaling accelerating the initial rate of pressure development at higher workloads by increasing XB recruitment through MyBP-C Ser302 phosphorylation. MyBP-C dephosphorylation limits this recruitable pool and prevents enhanced contraction in response to increased pacing or β-adrenergic signaling (7, 21), and loss of this regulation leads to altered timing of the cardiac cycle and a mismatch between ejection and filling (7).

CONCLUSIONS

Dephosphorylation of MyBP-C, and reduced Ser302 phosphorylation (37, 38), in particular, has been documented in conditions of chronic human heart failure (HF) (37, 39), including patients exhibiting HF with preserved EF (HFpEF) (40), a detrimental condition caused by a complex interplay of deficits in both diastolic and systolic reserves (41). Notably, cardiac dysfunction in HFpEF is often not apparent at rest but becomes noticeable during stress or increased workloads, suggesting that an inability to modulate cardiac output leads to exercise intolerance in HFpEF patients (41). Myocardial samples isolated from HF patients also display desensitization of β-adrenergic signaling pathway (42) that produces a reduction in PKA-mediated phosphorylation of myofilament contractile proteins, including MyBP-C (37, 43). Recent evidence demonstrates that the loss of MyBP-C phosphorylation directly blunts enhanced myocardial force generation and ventricular pressure development in response to increased workloads (7, 21). Here, we show that abolishing Ser302 phosphorylation greatly diminishes the cardiac β-adrenergic reserve by blunting PKA-mediated enhancements in XB kinetics and dobutamine-mediated enhancements in systolic pressure generation, similar to the effects of abolishing all three PKA phosphorylation residues. These results establish MyBP-C Ser302 as a critical modulator of cardiac output at higher workloads and suggest that therapeutic strategies designed to boost Ser302 phosphorylation could be beneficial for contractile function and enhance cardiac output in HF patients with reduced cardiac β-adrenergic reserve.

MATERIALS AND METHODS

Ethical approval and animal treatment protocols

Experiments were conducted according to the procedures laid out in the Guide for the Care and Use of Laboratory Animals [National Institutes of Health (NIH) Publication No. 85-23, Revised 1996] and as per the guidelines of the Institutional Animal Care and Use Committee at the Case Western Reserve University. NTG wild-type mice expressing full-length MyBP-C and TG mice expressing nonphosphorylatable MyBP-C with Ser-to-Ala substitutions at Ser273, Ser282, and Ser302 (that is, TG3SA) (6, 16) were used as control groups. TG mice expressing MyBP-C with Ser-to-Ala substitution at Ser302 (that is, TGS302A) were generated on a MyBP-C–null background such that there is no endogenous Ser302 phosphorylation. All mouse lines were of the SV/129 strain, and adult mice (3 to 6 months old) of both sexes were used for this study.

Determination of sarcomeric protein phosphorylation and MHC expression

Western blot and Pro-Q Diamond phosphoprotein stain (Life Technologies) were used to assess myofilament protein phosphorylation, as described previously (16, 44). On the day of the experiment, cardiac myofibrils were prepared by briefly homogenizing frozen mouse ventricular tissue for 15 s using a handheld homogenizer (PowerGen 500, Thermo Fisher Scientific) in relaxing solution that contains protease and phosphatase inhibitors (PhosSTOP and cOmplete ULTRA Tablets, Roche Applied Science). Myofibrils were chemically skinned for 15 min using 1% Triton X-100, centrifuged at 10,000g for 5 min, and resuspended in fresh relaxing solution on ice until further use. For PKA treatment, 100 μg of myofibrils isolated from hearts of each mouse line was incubated with the catalytic subunit of bovine PKA (Sigma-Aldrich) to a final concentration of 0.15 U PKA/μg myofibrils for 1 hour at 30°C, and the reaction was stopped by the addition of Laemmli buffer (16). Control myofibrils were incubated under the same conditions without PKA. For Western blots, myofibrils were separated on 4 to 20% tris-glycine gels (Lonza), transferred to polyvinylidene difluoride (PVDF) membranes, and incubated overnight with one of the following primary antibodies: MyBP-C (Santa Cruz Biotechnology); MyBP-C phospho–serine 273, 282, or 302 (detects phosphorylation of Ser273, Ser282, or Ser302 of MyBP-C; 21st Century Biochemicals); or HSC70 as a loading control (Santa Cruz Biotechnology). PVDF membranes were then incubated with appropriate secondary antibodies and imaged. For estimating total myofilament protein phosphorylation, myofibrils were separated on 4 to 20% tris-glycine gels and stained with Pro-Q phosphostain. Pro-Q–stained gels were counterstained with Coomassie Blue to determine total protein content. Myocardial samples were run on 5% tris-HCl gels (Bio-Rad, #3450001), followed by staining with silver stain plus (Bio-Rad, #1610449) to quantify the MHC expression as β-MHC percentage of the total MHC expression in the sarcomere (n = 6 per group). MHC gels were run at 150 V for 4 hours at 4°C (6). Densitometric scanning of the stained gels was done using ImageJ software (NIH, Bethesda, MD) (16).

Histological analysis of cardiac tissue

To examine cardiac morphology, excised hearts were submerged in formalin for 4 hours and sectioned at the mid-LV, as previously described (7). Hearts were paraffin-embedded and sectioned at 5 μm thickness before staining with Masson’s trichrome.

Cardiac morphology, contractile function, and in vivo hemodynamic function

Cardiac morphology (that is, left ventricular wall thickness) and in vivo systolic (that is, EF) and diastolic (that is, IVRT) function were evaluated by echocardiography, as previously described (7). In vivo P-V loop analysis was used to evaluate hemodynamic function, as previously described (7, 16). For in vivo experiments, stable baseline measurements of maximal systolic pressure, end diastolic pressure, the maximum rate of systolic pressure development (dP/dtmax), and the time constant of diastolic pressure relaxation (τ) were made before intraperitoneal injection of dobutamine (10 μg/g) to measure the response to increased β-adrenergic stimulation. P-V loops were analyzed offline using LabChart 7 software (ADInstruments).

Preparation of skinned myocardial preparations and Ca2+ solutions for mechanical experiments

Myocardial preparations were skinned, as previously described (16). In brief, frozen ventricular tissue pieces were homogenized in a relaxing solution, followed by detergent skinning using 1% Triton X-100 (Thermo Scientific) for 1 hour. Multicellular ventricular preparations measuring ~100 μm in width and ~400 μm in length were selected for the mechanical experiments. The composition of Ca2+ activation solutions prepared for the experiments was based on a computer program (45) and established stability constants (46). Ca2+ solutions contained the following: 14.5 mM creatine phosphate, 7 mM EGTA, and 20 mM imidazole. The maximal activating solution (pCa 4.5; pCa = −log [Ca2+]free) also contained 65.45 KCl, 7.01 CaCl2; 5.27 MgCl2, 4.81 ATP, whereas the relaxing solution (pCa 9.0) contained 72.45 mM KCl, 0.02 mM CaCl2; 5.42 mM MgCl2, 4.76 mM ATP. The pH and the ionic strength of the Ca2+ solutions were set to 7.0 and 180 mM, respectively. A range of pCa solutions (pCa 6.3 to 5.4) containing varying amounts of [Ca2+]free were then prepared by mixing appropriate volumes of pCa 9.0 and 4.5. stock solutions, and the mechanical experiments were performed at 23°C.

Experimental setup for measurement of dynamic and steady-state contractile properties in skinned myocardium

Detergent-skinned multicellular ventricular preparations were attached between a motor arm (312C, Aurora Scientific Inc.) and a force transducer (403A, Aurora Scientific Inc.), as described previously (47). Changes in the motor arm position and force transducer signals were sampled at 2000 Hz using a custom-built sarcomere length (SL) control software program developed by Campbell and Moss (48). For all mechanical measurements, SL of the ventricular preparations was set to 2.1 μm (16). Force-pCa relationships were determined by incubating the skinned myocardial preparations in pCa solutions ranging from pCa 6.3 to 4.5. The apparent cooperativity of force generation was estimated from the steepness of Hill plot transformation of the force-pCa relationships (44). The force-pCa data were fit using the equation P/P0 = [Ca2+]nH/(knH + [Ca2+]nH), where nH is the Hill coefficient and k is the pCa needed to elicit half-maximal force (that is, pCa50) (44).

Stretch activation experiments to measure dynamic XB contractile parameters

The stretch activation protocol used in these studies was described earlier (3, 16, 49). Skinned myocardial preparations were placed in Ca2+ solutions (pCa 6.1) that generate ~35% of the maximal force. Once the preparations attained a steady-state force, they were rapidly stretched by 2% of their initial ML, held at the new ML for 8 s, and then returned back to their initial ML. The key features of the stretch activation responses in cardiac muscle have been described earlier (50, 51), and various stretch activation parameters measured are illustrated in Fig. 4A. In brief, a sudden 2% stretch in ML causes an instantaneous spike in the force response (P1), which is due to the sudden strain of elastic elements within the strongly bound XBs (phase 1). The force then rapidly decays (phase 2) because of the detachment of the strained XBs into a non–force-bearing state, with a dynamic rate constant krel (an index of XB detachment). The lowest point of phase 2 (nadir) is indicated by P2 and is an index of the magnitude of XB detachment. Following phase 2, the preparations exhibit a gradual rise of force (phase 3), with a dynamic rate constant kdf (an index of the rate of XB recruitment). The delayed force rise in phase 3 is due to the sudden stretch-mediated recruitment of new XBs into the force-bearing state (50, 51). Stretch activation amplitudes, P3 and Pdf, were normalized to prestretch Ca2+-activated force where P3 was measured from prestretch steady-state force to the peak force value of the delayed force attained in phase 3, whereas Pdf was measured as the difference between P3 and P2 values, as described previously (3, 4, 47).

krel was measured by fitting a single exponential equation to the time course of force decay using the equation: F(t) = a(−1 + exp(−krel*t)), where “a” is the amplitude of the single exponential phase and krel is the rate constant of the force decay, as done earlier (5). kdf, which represents the rate of recruitment of all XBs that give rise to the delayed force transient following the sudden stretch in ML (that is, Pdf), was estimated by linear transformation of the half-time of force redevelopment (5), that is, kdf = 0.693/t1/2, where t1/2 is the time (in milliseconds) taken from the nadir (that is, the point of force reuptake at the end of phase 2) to the point of half-maximal force in phase 3 of the force response, where maximal force is indicated by a plateau region of phase 3 (that is, P3) (Fig. 4A) (5).

Stretch activation experiments were repeated following incubation of the myocardial preparations with PKA. Because PKA treatment decreases myofilament Ca2+ sensitivity of force generation, we used a pCa solution with slightly higher [Ca2+]free (pCa 6.0) to closely match the activation levels before PKA treatment, as done in earlier studies (5, 16).

Data analysis

Data were analyzed using two-way analysis of variance (ANOVA), and multiple pairwise comparisons were made using Fisher’s least significant difference (Fisher’s LSD) method, as previously reported (44). One-way ANOVA was used for analyzing the data reported for echocardiography experiments. Values are reported as means ± SEM. The criterion for statistical significance was set at P < 0.05, and the asterisks in the figures and tables denote statistical significance using post hoc Fisher’s LSD comparisons.

This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial license, which permits use, distribution, and reproduction in any medium, so long as the resultant use is not for commercial advantage and provided the original work is properly cited.

REFERENCES AND NOTES

Funding: This work was supported by the NIH (HL-114770 and P30 EY011373 to J.E.S.) and the American Heart Association (16POST30730000 to R.M.) grants. Author contributions: R.M., K.S.G., and J.E.S. contributed to the conception and design of the experiments. R.M., K.S.G., J.L., and J.E.S. participated in performing the experiments, in acquiring, analyzing, and interpreting data, and in drafting and revising the manuscript. All authors approved the final version of the manuscript. Competing interests: The authors declare that they have no competing interests. Data and materials availability: All data needed to evaluate the conclusions in the paper are present in the paper. Additional data related to this paper may be requested from the authors.
View Abstract

Navigate This Article