Faster cross-bridge detachment and increased tension cost in human hypertrophic cardiomyopathy with the R403Q MYH7 mutation

Abstract

The first mutation associated with hypertrophic cardiomyopathy (HCM) is the R403Q mutation in the gene encoding β-myosin heavy chain (β-MyHC). R403Q locates in the globular head of myosin (S1), responsible for interaction with actin, and thus motor function of myosin. Increased cross-bridge relaxation kinetics caused by the R403Q mutation might underlie increased energetic cost of tension generation; however, direct evidence is absent. Here we studied to what extent cross-bridge kinetics and energetics are related in single cardiac myofibrils and multicellular cardiac muscle strips of three HCM patients with the R403Q mutation and nine sarcomere mutation-negative HCM patients (HCM_smn). Expression of R403Q was on average 41 ± 4% of total MYH7 mRNA. Cross-bridge slow relaxation kinetics in single R403Q myofibrils was significantly higher (P < 0.0001) than in HCM_smn myofibrils (0.47 ± 0.02 and 0.30 ± 0.02 s⁻¹, respectively). Moreover, compared to HCM_smn, tension cost was significantly higher in the muscle strips of the three R403Q patients (2.93 ± 0.25 and 1.78 ± 0.10 μmol l^–1 s⁻¹ kN⁻¹ m⁻², respectively) which showed a positive linear correlation with relaxation kinetics in the corresponding myofibril preparations. This correlation suggests that faster cross-bridge relaxation kinetics results in an increase in energetic cost of tension generation in human HCM with the R403Q mutation compared to HCM_smn. Therefore, increased tension cost might contribute to HCM disease in patients carrying the R403Q mutation.

Key points

• The R403Q mutation, located in the S1 domain of the β-myosin heavy chain, is associated with a severe phenotype of hypertrophic cardiomyopathy (HCM).

• Increased cross-bridge relaxation kinetics caused by the R403Q mutation might underlie increased energetic cost of sarcomeric tension generation; however, direct evidence is absent.

• We studied the relationship between cross-bridge kinetics and energetics in single cardiac myofibrils and multicellular cardiac muscle strips in human HCM tissue with and without the R403Q mutation.

• In human HCM with the R403Q mutation, cross-bridge relaxation was faster and correlated well with a rise in energetic cost of tension generation.

• Our data suggest that an increase in tension cost is one of the causes underlying cardiomyopathy development in patients with the R403Q mutation.

Introduction

Familial hypertrophic cardiomyopathy (HCM) is an inherited cardiac disease with an incidence of 1:500. It is most frequently caused by mutations in genes encoding sarcomeric proteins (Maron et al. 2006a). The first identified mutation (R403Q) is located in the gene (MYH7) encoding cardiac β-myosin heavy chain and has been associated with high penetrance and a severe clinical phenotype. This malignant MYH7 mutation results in conversion of a highly conserved arginine residue to a glutamine (Geisterfer-Lowrance et al. 1990) at position 403 in the globular head of myosin (subfragment 1, S1). Based on the location of residue 403 at the base of a surface loop known to interact with actin it is most likely that the R403Q mutation directly interferes with motor and sarcomere function (Rayment et al. 1995; Volkmann et al. 2007).

A previous study in myofibrils from an HCM patient harbouring the R403Q mutation revealed faster generation and relaxation rates of tension development compared to healthy controls (Belus et al. 2008). In the two-state model of acto-myosin interactions (Brenner, 1988) the rate of force redevelopment (k_tr) in myofibrils is equal to f_app + g_app in which f_app and g_app represent the apparent rate constants of the transition of the cross-bridges into the force-generating states and back into the non-force generating states, respectively. k_tr can be derived from quick tension recovery after a short period of unloaded shortening or from the time course of force development (k_act), as can be measured in myofibrils (Poggesi et al. 2005; Belus et al. 2008). The relaxation of myofibrils is characterized by a slow, linear relaxation phase (slow k_rel) and a fast exponential relaxation component (fast k_rel). It has been shown that the isometric slow k_rel is equal to g_app because at low [Ca²⁺], strongly bound cross-bridges leave the force-generating states upon ADP-release and ATP-rebinding whereas weakly-bound cross-bridges are unable to go into the force-generating states (Brenner, 1988; Stehle et al. 2002; Tesi et al. 2002; Poggesi et al. 2005).

Moreover, g_app is equivalent to the energetic cost of tension generation or the ratio of ATPase activity over force, tension cost = ATPase activity/force (Brenner, 1988; Piroddi et al. 2007; de Tombe & Stienen, 2007). The human myofibrils with the R403Q mutation showed a lower maximal tension, a 2-fold increase in k_act and a 3-fold increase in slow k_rel compared to non-failing donor myofibrils (Belus et al. 2008). The prominent enhancement of relaxation kinetics indicates increased energetic cost of tension generation, and may be central in the pathogenesis of HCM caused by the R403Q mutation. However, direct demonstration that high cross-bridge detachment kinetics under isometric conditions of contraction caused by the R403Q mutation underlies increased energetic cost of tension generation (i.e. tension cost) is absent.

Therefore, we extended the previous observations by investigating if tension cost is increased in cardiac samples from three different HCM patients harbouring the R403Q mutation and how this correlates with slow relaxation kinetics. We simultaneously measured isometric force development and ATPase activity in cardiac muscle strips, providing a direct measure of tension cost at the level of the sarcomere. Cross-bridge kinetics was assessed during force measurements in myofibrils isolated from the same samples. R403Q samples were compared to cardiac tissue of eight sarcomere mutation-negative HCM patients (HCM_smn). The HCM_smn patients present with the same clinical phenotype as the R403Q patients and underwent cardiac surgery under similar conditions, which makes them a representative reference group.

Methods

Ethical approval

The study protocol is in line with the principles outlined in the Declaration of Helsinki and was approved by the local Ethics Committees. Written informed consent was obtained at the involved medical centres (Erasmus Medical Centre, Rotterdam, The Netherlands; Brigham and Women's Hospital, Cardiology, Boston, USA; and the Careggi University Hospital, Florence, Italy).

Cardiac tissue

Interventricular tissue was obtained from ten HCM patients during myectomy surgery to relieve left ventricular (LV) outflow tract obstruction, and LV tissue from two HCM patients was obtained during heart transplantation (HT) surgery. The tissue was snap frozen in liquid nitrogen. Three patients harboured the MYH7 R403Q mutation [one myectomy, R403Q(1), and two HT, R403Q(2) and R403Q(3)], while nine HCM_smn patients (myectomies) served as control [HCM_smn; no identified sarcomeric gene mutation after screening of nine HCM-associated genes, MYBPC3, MYH7, TNNT2, TNNI3, MYL2, MYL3, ACTC, TPM1, CSRP2 (Hershberger et al. 2009)]. Patient characteristics are shown in Table 1. Manifest HCM was evident from the high septal thickness (normal value <13 mm) (Shub et al. 1994) and high LV outflow tract pressure gradient (normal value <30 mmHg) (Maron et al. 2006b).

Table 1.

Patient characteristics

Patient	Age (years)	Sex	LVOT	ST
HCMR403Q
R403Q(1)	24	M	85	34
R403Q(2)	35	M	HT	HT
R403Q(3)	59	F	HT	HT
HCMsmn
1	35	M	76	30
2	54	M	100	33
3	42	M	75	26
4	52	M	169	22
5	65	F	85	19
6	72	F	88	24
7	49	M	61	20
8	46	M	81	19
9	59	M	85	18

Abbreviations: M, male; F, female; LVOT, left ventricular outflow tract pressure gradient (mmHg); ST, septal thickness (mm); HT, heart transplantation.

Multicellular cardiac muscle strips were isolated from three R403Q and six HCM_smn patients. In addition, cardiac myofibril preparations were isolated from the same three R403Q patients and three HCM_smn patients. Data of HCM_smn muscle strips (Witjas-Paalberends et al. 2014a) and myofibril data from patient R403Q(1) (Belus et al. 2008) have been published previously.

Sarcomere energetics in multicellular muscle strips

All muscle strips were immersed in cold relaxing solution (pH 7.0; in mmol l^–1: free Mg²⁺ 1, KCl 145, EGTA 2, ATP 4, Imidazole 10) and cut parallel to the long axis of the fibres in cold relaxing solution to minimize damage. Subsequently, muscle strips were permeabilized overnight in relaxing solution with 1% Triton X-100 at 4°C, which allowed us to control the intracellular Ca² concentration during measurements without interference of Ca²⁺ handling proteins.

The experimental procedures, solutions and equipment used for functional measurements were as described previously (Potma et al. 1994; Witjas-Paalberends et al. 2014a). The length of the preparations was adjusted on the basis of passive tension by stretching them until a level of 1–2 kN m^–2 was reached, which corresponds to a sarcomere length of about 2.2 μm (Narolska et al. 2005). The ventricular muscle strips were 1.52 ± 0.09 mm (mean ± SEM) in length, 368 ± 12 μm in width and 330 ± 9 μm in depth (no significant differences between groups). Isometric force and ATPase activity were measured at maximal and submaximal [Ca²⁺] at 20°C. During the measurement the muscle strips were kept inside a small chamber with a volume of 30 μl with thin quartz windows. Original registrations of force development and ATPase activity are shown in Fig. 1A–D. Maximal force was determined after the force signal reached a plateau and normalized to the cross-sectional area (CSA) of the muscle strip to calculate tension (force/CSA) (Fig. 1A, B). The CSA of the preparation was estimated in the set-up assuming an elliptical shape, i.e. CSA = (width × depth × π)/4. ATPase activity was measured using an enzyme coupled assay. In this assay the ATP regeneration from ADP and phosphoenol-pyruvate, catalysed by the enzyme pyruvate kinase, is coupled to the oxidation of NADH to NAD⁺ and the reduction of pyruvate to lactate, catalysed by l-lactic dehydrogenase. NADH oxidation was measured photometrically from the absorbance at 340 nm of near-UV light. The absorbance signal was calibrated by adding 0.5 nmol of ADP (arrow) to the solution in the measuring chamber. Using this calibration, ATPase activity in the muscle strip could be derived from the slope of the absorbance signal (Fig. 1C, D). The maximal Ca²⁺-activated ATPase activity was calculated by subtracting basal ATPase activity (measured in relaxing solution) from the maximal NADH oxidation and normalization to the volume of the muscle strip. The economy of muscle contraction is expressed as tension cost, i.e. ATPase activity during force development (ATPase activity/tension).

Figure 1.

Original registrations of a multicellular muscle strip and myofibril preparation

A and B, the generation of force normalized to cross-sectional area (CSA) (tension) in an HCM_smn and R403Q muscle strip. The muscle strip was activated in a saturating [Ca²⁺] solution (pCa 4.5) until force reached a steady state and subsequently relaxed in a low [Ca²⁺] solution (pCa 9). C and D, corresponding NADH absorbance recorded during the activations shown in A and B. E and F, contraction–relaxation cycles of myofibril preparations from an HCM_smn (black curve) and R403Q (red curve) patient. G, the time course of tension activation following sudden [Ca²⁺] increase of the HCM_smn and R403Q(1) myofibril preparations shown in E and F superimposed on a faster time base after normalization for maximal tension. H, tension relaxation kinetics following sudden Ca²⁺ removal of the HCM_smn and R403Q(1) myofibril preparations superimposed on a faster time base after normalization for maximal tension. k_act is the rate constant of tension generation and k_rel represents the rate constant of relaxation.

Cross-bridge kinetics in myofibril preparations

Thin strips of the patient cardiac samples were incubated overnight in ice-cold relaxing solution (pCa 8.0, pH 7.0, in mmol l^–1: ionic strength 200, EGTA 10, Tris 10, MgATP 5, free Mg²⁺ 1) with 1% Triton X-100 added. Triton was then removed and the strips were homogenized in the same relaxing solution to produce myofibril suspensions that were stored at 0–4°C and used for experiments for up to 4 days. Techniques for mechanical measurements in human cardiac myofibrils were as previously described (Piroddi et al. 2007; Belus et al. 2008, 2010). Briefly, myofibrils were transferred to a temperature controlled chamber filled with relaxing solution (pCa 8.0, 15°C). The selected myofibril was mounted (initial sarcomere length around 2.2 μm) between a cantilever force probe and a glass needle mounted on the shaft of a length control motor. Myofibrils were maximally activated and fully relaxed by rapid solution switching between two continuous streams of solutions (pCa 8.0 and 4.5) (Fig. 1E–H). All solutions contained an MgATP-regenerating system and a cocktail of protease inhibitors. Slow k_rel is estimated from the slope of the regression line fitted to the tension trace normalized to the entire amplitude of the tension relaxation transient. The rate constant for fast k_rel is estimated from a mono-exponential fit.

Histology

From each R403Q patient and three HCM_smn patients, 10 myocardial cryosections of 5 μm were formalin fixed. The CSA of cardiomyocytes and the amount of fibrosis per cryosection were analysed using Wheat Germ Agglutinin (WGA) staining and Picrosirius Red staining, respectively. Photomicrographs of the cryosections stained with WGA were obtained (Leitz DMRXE; Leica, Wetzlar, Germany) and the CSA of at least 100 representative delineated cardiomyocytes per patient was quantified as width × depth × π/4 using ImageJ 1.45s software (National Institutes of Health, Bethesda, MD, USA). Photomicrographs of the cryosections stained with Picrosirius Red were obtained (Leica DM 2000) and fibrosis of each cryosection was quantified as a percentage of the total slide area using ImageJ 1.45s software.

Allelic mRNA expression analysis

RNA was extracted from 40 mg frozen cardiac tissue using the PeqGold Total RNA Kit (PeqLab, Fareham, UK) according to the supplier's protocol. RNA integrity was analysed using a PicoKit bioanalyser (Agilent, Santa Clara, CA, USA) and revealed a RNA integrity number above 6 for all patient samples. For the reverse transcription 2 μl RNA was incubated with 1× reaction buffer, 2.5 mmol dNTPs each, 8 pmol MYH7-specific primer (5′-GCCTTGCCTTTGCCCTTCTCA-3′), 20 U RiboSafe (Bioline, London, UK), and 100 U Tetro RT (Bioline) in 20 μl for 1 h at 42°C. For quantitative PCR, 1× reaction buffer S (Hot Taq DNA Polymerase, PeqLab), 25 mmol dNTPs each, 10 pmol forward primer (5′-TGACAGGCGCCATCATGCACTTTGGA-3′), 10 pmol reverse primer (5′-CTGCTTGGTCTCCAGGGTGGCA-3′), 1.25 mmol MgCl₂, 1 U Hot Taq DNA Polymerase (PeqLab) in a final volume of 25 μl were applied to an initial denaturation at 95°C for 5 min, 25 cycles of denaturation at 95°C for 30 s, annealing at 70°C for 30 s, elongation at 72°C for 30 s and a final elongation step at 72°C for 2 min.

To eliminate heteroduplexes we performed a reconditioning PCR as described previously (Tripathi et al. 2011). In brief, the PCR products were diluted 1:100 and applied to eight parallel PCR reactions that were terminated in the linear phase of the PCR. The reactions were pooled, precipitated and subjected to a mutation-specific DdeI treatment by adding 1× buffer 4 and 10 U DdeI (New England Biolabs, Inc., Ipswich, MA, USA) to a final volume of 10 μl and incubated at 37°C for at least 3 h or overnight. The treatment yielded a fragment of 119 bp for both alleles, a wildtype-specific fragment of 180 bp and a mutation-specific fragment of 148 bp. The fragments were separated in 2.5% sieving agarose gels (Biozym, Hessisch Oldendorf, Germany) stained with ethidium bromide and quantified as described for other MYH7 mutations (Tripathi et al. 2011). In brief, the integrated optical density was determined for each band and normalized for the respective size. The fraction of each allele was calculated as the percentage from the common 119 bp band.

Statistical analysis

Statistical analysis of the data was performed using Prism version 5.0 (Graphpad Software, Inc., La Jolla, CA, USA) and SPSS version 20.0 (IMB, Armonk, NY, USA). The data are presented as mean ± SEM. The statistics regarding functional measurements are based on the individual myofibril preparations and multicellular muscle strips. Using the Shapiro–Wilk test the data sets were analysed regardless of whether the assumption of normality was violated. Normality was assumed when P > 0.05 and the variances were equal. As the repeated samples assessments had to be taken into account (multiple myofibril preparations and muscle strips were measured from one patient), multilevel analysis was performed using a linear mixed model procedure to investigate differences between the groups as published previously (Coppini et al. 2013; Sequeira et al. 2013; Witjas-Paalberends et al. 2013). For this, the data sets violating the normality assumption (P < 0.05) were logarithmically transformed. The histological data were analysed with the non-parametric Mann–Whitney–Wilcoxon test or Kruskal–Wallis test followed by a post hoc Dunn's test due for violation of the normality assumption. Regarding the mRNA analyses, a one-way analysis of variance (ANOVA) was performed with Bonferroni's Multiple Comparison as a post hoc test. P < 0.05 was considered significant.

Results

Sarcomere energetics

Simultaneous measurements of force generation and ATPase activity were performed in membrane-permeabilized muscle strips from three R403Q patients and five HCM_smn patients. The normality assumption was not violated by any of the datasets (P > 0.05).

Maximal tension (force normalized by CSA) on average was significantly lower (11.4 ± 1.1 kN m⁻²) in the R403Q muscle strips compared to HCM_smn (24.4 ± 1.6 kN m⁻², Fig. 2A). The muscle strips from each individual patient showed a significant decrease in maximal tension compared to HCM_smn as well (Fig. 2B). Maximal ATPase activity, corrected for basal ATPase activity (4.4 ± 0.5 and 6.6 ± 1.2 μmol l⁻¹ s⁻¹, respectively), was significantly lower (30.5 ± 2.7 μmol l⁻¹ s⁻¹) in R403Q muscle strips compared to HCM_smn (41.4 ± 2.4 μmol l⁻¹ s⁻¹, Fig. 2C). When the muscle strips were separated per patient, both R403Q(1) and R403Q(3) muscle strips showed a significantly lower maximal ATPase activity compared to HCM_smn. In addition, R403Q(1) had a significantly lower ATPase activity compared to R403Q(2) (Fig. 2D). Maximal tension cost was calculated by dividing maximal ATPase activity by maximal tension. Tension cost (Fig. 2E) was significantly higher in R403Q compared to HCM_smn (2.9 ± 0.2 and 1.8 ± 0.1 μmol l⁻¹ s⁻¹ kN⁻¹ m⁻², respectively). In addition, maximal tension cost of R403Q(1) was significantly higher than HCM_smn and higher compared to R403Q(2) and R403Q(3) (Fig. 2F).

Figure 2.

Maximal tension, ATPase activity and tension cost in multicellular muscle strips

A, maximal tension on average was significantly lower in R403Q muscle strips compared to HCM_smn (*P < 0.0001). B, the lower maximal tension was evident in each individual R403Q patient compared to HCM_smn (*P < 0.0001). In addition, muscle strips of R403Q(1) revealed a lower maximal tension compared to R403Q(2) (^#P = 0.025) and R403Q(3)(⁺P = 0.026). C, maximal ATPase activity was significantly lower in R403Q muscle strips compared to HCM_smn (*P = 0.004). D, R403Q(1) and R403(3) muscle strips showed significantly lower ATPase activity compared to HCM_smn (*P = 0.002 and *P = 0.017, respectively). ATPase activity in R403Q(1) was significantly lower than R403Q(2) (^#P = 0.04). E, tension cost was calculated by dividing maximal ATPase activity by tension. Maximal tension cost was significantly higher in R403Q compared to HCM_smn (*P < 0.0001). F, the increase in tension cost was significant in R403Q(1) and R403Q(2) compared to HCM_smn (*P < 0.0001 and *P = 0.008, respectively). In addition, tension cost in R403Q(1) muscle strips was higher compared to R403Q(2) (^#P = 0.020) and R403Q(3)(⁺P = 0.001). Average data are shown ± SEM, N = number of patients, n = number of individual muscle strips; data points represent individual muscle strips ± SEM.

Additional measurements of force and ATPase activity were performed at submaximal [Ca²⁺] in a subset of muscle strips. Figure 3A shows tension development and corresponding ATPase activity of the R403Q and HCM_smn muscle strips (not corrected for basal ATPase activity) over a range of [Ca²⁺]. The relationship between tension and ATPase activity over the entire [Ca²⁺] range can be estimated by a linear equation, as shown previously (de Tombe & Stienen, 1995). The slopes of the tension–ATPase activity relationships represent tension cost over the entire range of [Ca²⁺]. The slopes of the R403Q muscle strips are significantly higher compared to HCM_smn (Fig. 3A). When plotting the average curves per R403Q patient, the R403Q(1) showed a significantly higher slope compared to HCM_smn and R403Q(3) (Fig. 3B).

Figure 3.

Tension cost at maximal and submaximal [Ca⁺]

A, tension and ATPase activity in muscle strips were measured at submaximal [Ca²⁺] yielding the following relations for the R403Q patients: y = 2.82x + 4.47 and HCM_smn patients: y = 1.65x + 4.47. The average slope (i.e. tension cost) of the R403Q muscle strips was significantly higher compared to HCM_smn (*P = 0.018). B, ATPase activity–tension relations of muscle strips of individual R403Q patients are as follows: R403Q(1): y = 3.66x + 7.26, R403Q(2): y = 2.59x + 5.40 and R403Q(3): y = 2.22x + 0.75. The R403Q(1) slope reached a significant difference compared to both HCM_smn (*P = 0.001) and R403Q(3) (⁺P = 0.026). Average data are shown ± SEM, N = number of individual muscle strips; data points represent individual muscle strips ± SEM.

Cross-bridge kinetics

The parameters of cross-bridge kinetics were measured in myofibril preparations of the three HCM patients with the R403Q mutation and of three HCM_smn patients. Data for R403Q(1) have been previously published (Belus et al. 2008), but measuring more myofibrils of this patient was not possible as the remaining preserved myectomy tissue has been used for tension cost measurements in multicellular muscle strips, histological analyses and mRNA analysis. Representative contraction–relaxation traces of myofibril preparations are shown in Fig. 1E–H. Both k_act and fast k_rel data sets violated the normality assumption (P < 0.05) and were logarithmically transformed, resulting in data sets not violating the normality assumption.

Figure 4A shows the maximal tension (force normalized to CSA) of the entire R403Q group compared to HCM_smn, which did not differ between the two patient groups. However, when the three patients were analysed individually (Fig. 4B), myofibril preparations of the R403Q(1) patient did show a significantly lower tension compared to the myofibril preparations of HCM_smn and the other R403Q patients.

Figure 4.

Cross-bridge kinetics in myofibril preparations

A, maximal tension on average did not differ between R403Q and HCM_smn myofibrils. B, myofibrils from R403Q(1) showed a significant decrease in maximal tension compared to the other R403Q patients and HCM_smn patients (*P = 0.02 vs. HCM_smn, ^#P = 0.019 vs. R403Q(2) and ⁺P = 0.003 vs. R403Q(3)). C, slow k_rel (∼g_app∼tension cost) was significantly higher in R403Q myofibrils compared to HCM_smn (*P < 0.0001). D, this increase was visible in myofibrils of all R403Q patients compared to HCM_smn (R403Q(1)(2) *P < 0.0001 and R403Q(3) *P = 0.001). In addition, slow k_rel was significantly higher in myofibrils of R403Q(1) compared to R403Q(2)(^#P = 0.006) and R403Q(3)(⁺P = 0.001). Data are represented as individual myofibril preparations ± SEM and N = number of patients, n = number of individual muscle strips. ^aData from Belus et al. (2008).

The rate constant of force development (k_act) was significantly higher in R403Q myofibrils compared to HCM_smn (Table 2 and Fig. 1G). In addition, when the myofibril data were separated per patient (Table 2), each individual patient showed a significantly higher k_act compared to HCM_smn. In addition, k_act of R403Q(2) was significantly higher compared to the other R403Q samples (Table 2).

Table 2.

Kinetic data (mean±SD) for tension generation and fast relaxation

	HCMsmn	R403Q	R403(1)a	R403Q(2)	R403Q(3)
kact (s−1)	0.74 ± 0.03 (45)	1.27 ± 0.04 (92)*	1.4 ± 0.15 (18)*	1.34 ± 0.06 (41)*#	1.11 ± 0.05 (33)*
Fast krel (s−1)	4.62 ± 0.20 (44)	5.32 ± 0.18 (78)+	5.12 ± 0.34 (16)	5.13 ± 0.27 (39)	5.77 ± 0.35 (23)^

^*P < 0.0001 R403Q, R403Q(1), R403Q(2), R403Q(3) vs. HCM_smn

^#P = 0.01 R403Q(2) vs. R403Q(3)

⁺P < 0.017 R403Q vs. HCM_smn

^{^}P < 0.005 R403Q(3) vs. HCM_smn; number of myofibril preparations in parentheses.

^aData from Belus et al. (2008).

The fast component of relaxation (fast k_rel) on average was significantly increased in R403Q myofibrils compared to HCM_smn (Table 2). This difference was due to higher fast k_rel of myofibrils from patient R403Q(3) (Table 2). The slow component of relaxation (slow k_rel) was significantly higher in the myofibrils of all R403Q patients together (Figs. 1H and 4C) and when taken individually compared to HCM_smn (Fig. 4D).

As slow k_rel ∼ g_app ∼ tension cost, the average slow k_rel values of the three R403Q patients and HCM_smn patients (Fig. 4D) should correspond to the maximal tension cost values (Fig. 2F). A clear positive linear relationship is visible when the data of Figs. 2F and 4D are combined with tension cost as a function of slow k_rel (Fig. 5). This shows not only that cross-bridge relaxation kinetics and cross-bridge energetics correlate, but also the differences among the three R403Q patients, suggesting that factors other than the R403Q mutation underlie the changes in sarcomere properties.

Figure 5.

Relation between slow relaxation kinetics and sarcomere energetics and function

The correlation between slow k_rel of the myofibril preparations and tension cost of the muscle strips can be described by the positive linear equation y = 0.13x + 0.1 with R² = 0.97. The slow k_rel measured in myofibrils correlated very well with the energetic cost of isometric tension generation (tension cost) measured in multicellular muscle strips from the same samples. Data are represented as mean ± SEM of HCM_smn (N = 9) and each individual R403Q patient.

Based on k_act and slow k_rel, myofibril force can be estimated using the following formula: (k_act – slow k_rel)/k_act. Figure 6A shows that the estimated myofibril force of R403Q(1) was 15% lower compared to HCM_smn. In addition, the apparent rate constant of the transition of the cross-bridges into the force-generating states, f_app, can be calculated as k_act – slow k_rel. Multilevel analysis revealed that the estimated f_app is significantly higher in the R403Q myofibrils together and for each R403Q patient individually compared to HCM_smn (Fig. 6B).

Figure 6.

Estimated myofibril force and f_app

A, estimated myofibril force of R403Q(1) myofibrils was lower compared to HCM_smn myofibrils. B, estimated f_app was significantly higher in R403Q myofibrils *P < 0.0001 and for myofibrils of each R403Q patient individually compared to HCM_smn (*P < 0.001, *P < 0.0001 and *P < 0.005, respectively).

Histology

Figure 7A shows representative cross-sections of HCM_smn, R403Q(1), R403Q(2) and R403Q(3) tissue stained with WGA. CSA was significantly higher for R403Q cardiomyocytes compared to HCM_smn cells. The higher cellular CSA was mostly attributed to R403Q(1) cardiomyocytes (Fig. 7B, C). Figure 7D shows representative cross-sections of HCM_smn, R403Q(1), R403Q(2) and R403Q(3) tissue stained with Picrosirius Red. The amount of fibrosis tended to be higher (P = 0.06) in R403Q compared to HCM_smn tissue, largely due to the significantly higher percentage of fibrosis in the R403Q(3) sample (Fig. 7E, F).

Figure 7.

Histological and mRNA analysis

A, representative iamges of WGA-stained cryosections. B, CSA was significantly higher (*P < 0.0001) in the R403Q cardiomyocytes, which was mostly caused (C) by the R403Q(1) and R403Q(2) cardiomyocytes (*P < 0.0001 and P < 0.01, respectively). D, representative images of Picrosirius Red-stained cryosections. E, fibrosis tended to be higher for R403Q compared to HCM_smn (P = 0.06). F, fibrosis was highest for R403Q(3) compared to HCM_smn (P < 0.05). G, RNA was extracted three times from a tissue sample of each patient. The relative expression of the R403Q and wildtype alleles was quantified at least in duplicate. Analysis of mutant R403Q mRNA revealed a deviation from the 50:50 ratio (allelic imbalance) in R403Q(1) with a fraction of mutated mRNA being significantly lower than in R403Q(2) (^#P < 0.0001) and R403Q(3) (⁺P < 0.0002), both showing no allelic imbalance. N = number of individual PCR reactions analysed.

Fraction of mutated MYH7 mRNA

As the three R403Q patients revealed different cross-bridge kinetics and energetics we studied the amount of mutant MYH7 mRNA present in the tissue of these patients. Three RNA extractions per patient were performed. Interestingly, Fig. 7G shows that R403Q(1) has a significant lower amount of R403Q mRNA compared to both R403Q(2) and R403Q(3), while tension cost and slow k_rel were highest in this patient (Figs. 2F and 4D).

Discussion

This is the first study in which cross-bridge kinetics are directly compared to cross-bridge energetics in cardiac samples from the same patients. Our measurements of ATPase activity during sarcomere contraction in multicellular strips revealed a significantly higher tension cost in R403Q compared to HCM_smn. High tension cost correlated with a significantly higher apparent rate of cross-bridge detachment under isometric conditions (measured from the isometric relaxation rate of myofibrils). Interestingly, changes in cross-bridge kinetics and energetics were not the same in all three R403Q patients. Our study provides evidence that the R403Q MYH7 mutation decreases the economy of myocardial contraction at the level of the sarcomeres and may indeed represent one of the causal factors of cardiomyopathy in patients carrying the R403Q mutation.

Decrease in maximal force

The R403Q mutation is located in a surface loop of the myosin head domain that forms an actin–myosin interface. It has been called the cardiomyopathy loop (Liu et al. 2005) as mutations in this area are known for their detrimental consequences on the cross-bridge cycle and sarcomere function. In both multicellular muscle strips and myofibril preparations of HCM with the R403Q mutation, maximal force tended to be lower compared to HCM_smn (Figs. 2A, B and 4B). Decreased maximal force generating capacity was also reported in human skeletal R403Q muscle fibres (Lankford et al. 1995; Malinchik et al. 1997) and cardiac muscle fibres from transgenic rodents (Blanchard et al. 1999) compared to controls as well. Due to the R403Q mutation the cross-bridges leave the force generating states faster, which is supported by the findings of the present study by the enhanced relaxation kinetics, i.e. increase in g_app (Fig. 4C, D). This should lead to a decline in force generation capacity of the sarcomere (Brenner, 1988; Belus et al. 2008) unless the increase in g_app is compensated for by an increase in f_app. Kinetic data in Table 2 and Fig. 4D support the idea that the R403Q mutation does increase f_app besides increasing g_app, blunting the potential impact on force generation of the increased cross-bridge detachment rate, which was confirmed when myofibril force and f_app were estimated based on kinetic data (Fig. 6A, B). In contrast to the multicellular preparations (Fig. 2B), in myofibril preparations the force reduction was only seen in R403Q(1) (Fig. 4B). This is in line with the larger effects on slow k_rel, and hence g_app, in this patient compared to R403Q(2) and R403Q(3) (Fig. 4D). In R403Q(2) and R403Q(3) the increase in g_app is not that large and its effect on maximal force is compensated for by the increase in f_app (Fig. 6A, B).

In multicellular preparations other than myofibrils, cellular remodelling such as cellular hypertrophy, a lower myofibrillar density and fibrosis may also contribute to the reduced force generation (Fig. 2A). Indeed, a decrease in myofibrillar density has previously been correlated with the maximal force generating capacity in single human HCM cardiomyocytes with MYH7 mutations (Kraft et al. 2013; Witjas-Paalberends et al. 2013). In the present study we show that cellular hypertrophy (Fig. 7B) and the extent of fibrosis (Fig. 7E) are higher in the R403Q group compared to the HCM_smn. Cellular hypertrophy was highest in the R403Q(1) myectomy sample, which showed the lowest force generating capacity in muscle strips (Fig. 2B), while fibrosis was most prominent in the R403Q(3) sample, which was obtained during HT surgery from an end-stage failing patient. A combination of cellular remodelling and faster g_app could explain the large decline in force in muscle strips of R403Q(1), while remodelling (i.e. fibrosis) seems to be the major determinant of reduced maximal force generation in R403Q(2) and R403Q(3) muscle strips. Another patho-mechanism that may underlie reduced force generating capacity and warrants further research may be protein modifications caused by oxidative stress such as actin carbonylation. Actin carbonylation was found to be increased in human end-stage heart failure tissue (Canton et al. 2011). Interestingly, actin carbonylation was slightly higher in MYH7_mut tissue compared to HCM_smn and significantly higher compared to non-failing donor tissue (Witjas-Paalberends et al. 2014b).

Increased relaxation kinetics underlie higher tension cost

A significantly higher tension cost at maximal [Ca²⁺] was found in multicellular muscle strips of three HCM patients harbouring the R403Q mutation compared to muscle strips from HCM_smn patients (Fig. 2E, F). The same increase in tension cost was seen at submaximal [Ca²⁺], illustrated by the steeper ATPase activity – tension relationships of R403Q muscle strips compared to HCM_smn (Fig. 3). Overall, our data indicate that the R403Q mutation lowers economy of cardiac muscle contraction at the level of the sarcomere at maximal and submaximal (physiological) [Ca²⁺].

This study directly shows, in agreement with previous studies, that slow k_rel is equal to g_app and energetic cost of tension development (Brenner, 1988; Piroddi et al. 2007; de Tombe & Stienen, 2007) as in addition to a ∼80% increase in tension cost, we found a ∼67% higher slow k_rel of human R403Q myofibrils compared to HCM_smn (Figs. 2 and 4). Moreover, the increase in slow k_rel for the three patients with R403Q correlated very well with the respective increase in tension cost (Fig. 5). Our data provide evidence that an increase in cross-bridge detachment rate enhances the energetic costs of sarcomere contraction in human HCM with the R403Q mutation.

R403Q leads to a loss of function

Over the years, extensive research has been performed on the functional consequences of the R403Q mutation. In vitro actin sliding velocity (V_actin) and ATPase assays were used to analyse the effects of the R403Q mutation on cross-bridge mechanics and enzymatic properties of human skeletal muscle (Cuda et al. 1993, 1997) and human cardiac muscle expressing the mutant protein (Cuda et al. 1993, 1997; Palmiter et al. 2000; Belus et al. 2008), or on recombinant proteins/transgenic animal models (Sweeney et al. 1994; Sata & Ikebe, 1996; Roopnarine & Leinwand, 1998; Tyska et al. 2000; Yamashita et al. 2000). It appears that the use of different models leads to a variety of outcomes regarding the functional consequences of the R403Q mutation, of which an overview is provided in Table 3. Specifically, the results obtained in rodents regarding the effects of the R403Q mutation cannot directly be extrapolated to the human situation as rodents mostly express α-MyHC in the ventricles compared to mostly β-MyHC in human ventricles (Lompre et al. 1981; Mercadier et al. 1983). Studies using transgenic R403Q mice with either an α-MyHC or a β-MyHC background revealed increased V_actin, actin-activated ATPase activity and MgADP release rate in the α-MyHC background. However, expressing the R403Q mutation in β-MyHC background mice did not enhance any of the three parameters (Lowey et al. 2008, 2013), showing the impact of the specific MyHC background on the mechanical and enzymatic properties.

Table 3.

Summary of the literature regarding R403Q

	Type of tissue	Analyses	Parameters	Conclusions
Cuda et al. (1993)	Soleus, pectoralis of human R403Q patients and healthy controls, cardiac biopsies from human R403Q patients	In vitro motility assay	Actin sliding velocity (Vactin)	Decreased
Lankford et al. (1995)	Soleus of human R403Q patients and healthy controls	Skinned fibres functional measurements	Isometric force Shortening velocity	Decreased Decreased
Sweeney et al. (1994)	Recombinant R403Q α-MyHC and WT α-MyHC	ATPase assay and In vitro motility assay	ATPase Vmax ATPase km Vactin	Decreased Increased Decreased
Malinchik et al. (1997)	Soleus of human R403Q patients and healthy controls	Skinned fibres functional measurements	Isometric force	Decreased
Sata and Ikebe (1996)	Recombinant R403Q β-MyHC	ATPase assay and In vitro motility assay	ATPase Vmax Actin dissociation constant (ka) Vactin	Decreased Increased Decreased
Cuda et al. (1997)	Soleus of human R403Q patients and healthy controls, cardiac biopsies from human R403Q patients	In vitro motility assay	Vactin	Decreased
Roopnarine & Leinwand (1998)	Recombinant R403Q α-MyHC	ATPase assay	ATPase Vmax ATPase km	Decreased Increased
Blanchard et al. (1999)	Cardiac LV α-MyHC403/+ tissue and WT tissue from rodents	Skinned fibres functional measurements	Dynamic stiffness at rest Maximal tension Submaximal tension Maximal cross-bridge oscillating power Submaximal cross-bridge oscillating power	Increased Decreased Increased Decreased Increased
Tyska et al. (2000)	Cardiac tissue from rodents with α-MyHC403+/−, α-MyHC403+/+ and WT	ATPase assay and In vitro motility assay	ATPase Vmax ATPase km Vactin	Increased Increased Increased
Yamashita et al. (2000)	Recombinant smooth muscle R403Q myosin using insect expression systems	ATPase assay and In vitro motility assay	ATPase Vmax ATPase km Vactin	Increased Decreased Increased
Palmiter et al. (2000)	Cardiac biopsies from R403Q patients	In vitro motility assay and Laser trap assay	Vactin Average force production Single myosin-single actin interaction using Vactin = d/ton × ton	Increased No differences Decreased
Belus et al. (2008)	Cardiac tissue from a R403Q patient and donor tissue	Functional experiments in myofibrillar preparations	ktr Slow krel Fast krel Maximal tension	Increased Increased Increased Decreased
Lowey et al. (2008)	Cardiac tissue from rodents with α-MyHC403+/−, β-MyHC403+/− and WT	ATPase assay and In vitro motility assay	α-MyHC403+/: ATPase Vmax & ATPase km β-MyHC403+/−:ATPase Vmax & ATPase km α-MyHC403+/−: Vactin β-MyHC403+/−: Vactin	Increased No change Increased No change
Lowey et al. (2013)	Cardiac tissue from rodents with α-MyHC403+/−, β-MyHC403+/− and WT	Stopped flow kinetics	α-MyHC403+/−: MgADP release rate β-MyHC403+/−: MgADP release rate	Increased Minor reduction

An in vitro motility study using human cardiac tissue showed an increased V_actin (Palmiter et al. 2000). Single molecule analyses revealed that the increase in V_actin of human R403Q β-MyHC S1 was due to the lower period that the myosin head was connected to actin (t_on) without changes in unitary step displacement (d) of the myosin head, suggesting that the R403Q mutation leads to destabilization of the actomyosin interaction, and hence to a more rapid myosin detachment (Palmiter et al. 2000). In a previous (Belus et al. 2008) and in the present study faster relaxation kinetics from isometric steady-state force generation and thus an increase in g_app was found in myofibril preparations of human cardiac R403Q tissue. Overall it appears that the R403Q mutation leads to faster cross-bridge kinetics in combination with an overall loss of function characterized by decreased force production by the cross-bridges and reduced economy of contraction.

Expression of R403Q and sarcomere changes

MYH7 mutations are usually heterozygous, producing both an affected and an unaffected allele, resulting in poison peptides which are incorporated in the sarcomere (Becker et al. 1997, 2007; Nier et al. 1999). Therefore, both healthy and mutated myosin proteins are present in the sarcomere. Analysis of R403Q mRNA revealed an almost 50% expression of the mutant in the two samples obtained during HT surgery, while mRNA expression level of R403Q was significantly lower in the sample (R403Q(1)) obtained during myectomy surgery (Fig. 7G). Previously, Tripathi et al. (2011) and Montag et al. (2012) revealed that heterozygous MYH7 mutations are subject to allelic imbalance, and mRNA and protein levels of mutant myosin correlated with HCM disease severity. Allelic imbalance due to cis-effects leading to different allelic expression of each parental chromosome was previously discovered in humans and linked not only to physiological differences between humans but also to the severity of diseases (Hoffmeyer et al. 1994; Buckland, 2004). Our data on mRNA expression support these previous studies and suggest that expression of mutant mRNA is higher at a more severe stage of cardiac disease, i.e. may increase during HCM development to end-stage heart failure.

The question remains, however, why both slow k_rel and tension cost differed among the individual R403Q patients. While the fraction of R403Q mRNA was lowest in R403Q(1) compared to both end-stage failing samples (Fig. 5A), the R403Q(1) heart tissue showed the highest slow k_rel and tension cost values. The fraction of mutated mRNA therefore did not explain the differences found in kinetics and energetics in the individual R403Q patients. We cannot exclude that the protein levels of the mutant protein do not correspond to the observed mRNA levels. However, a previous study (Tripathi et al. 2011) showed similar expression of several mutants at both mRNA and protein level determined in both cardiac and skeletal muscle tissue. The influence of modifier genes, other than the causal mutated gene, could be a likely candidate as well. Modifier genes have been associated with differences in severity, penetrance and age of onset of HCM disease between patients with the same mutation or even from the same family (Marian, 2000, 2002). This could be of potential interest in R403Q(1) as this patient underwent a myectomy surgery at 25 years of age. Indeed, when age was correlated with tension cost and slow k_rel (Fig. 8A, B) a correlation was present between age and tension cost (P = 0.03; R² = 0.51), which clearly cannot be due to HCM_smn as the correlation disappeared when the R403Q patients were excluded (P = 0.6; R² = 0.09). There was no significant correlation for slow k_rel and age (P = 0.23; R² = 0.33) for the entire group. However, when the HCM_smn group was excluded there tended to be a correlation (P = 0.06; R² = 0.71). Interestingly, single myofibril measurements in a large number of non-failing donors and aortic stenosis patients do not show any effect of age on slow k_rel either (our unpublished results). Although the groups are too small to draw definite conclusions, based on the present data is seems that the influence of the R403Q mutation on function might be age-dependent and clearly requires future research.

Figure 8.

Correlation with age

A, there was a significant correlation between tension cost and age (P = 0.03; R² = 0.51), which was not due to HCM_smn. B, there was no significant correlation for slow k_rel and age taking both HCM_smn and R403Q into account (P = 0.23; R² = 0.33).

Conclusion

We investigated the direct influence of the MYH7 mutation R403Q on the economy of sarcomere contraction using cardiac muscle from three HCM patients. For comparison we used tissue of HCM patients with a similar clinical phenotype without a sarcomeric gene mutation after thoroughly screening nine genes. Our data show that the presence of a sarcomeric gene mutation is an important determinant of myocardial energy utilization. Our study shows faster cross-bridge kinetics and a direct increase in energetic costs of tension generation of sarcomeres harbouring the R403Q MYH7 mutation compared to HCM_smn sarcomeres. This indicates that the presence of the R403Q mutation can perturb sarcomere energetics, which may reflect one factor causing HCM disease.

Glossary

CSA

cross-sectional area

HCM

hypertrophic cardiomyopathy

HT

heart transplantation

LV

left ventricle

MyHC

the protein myosin heavy chain

MYH7

the gene encoding β-myosin heavy chain

WGA

wheat germ agglutinin

Additional information

Competing interests

None

Author contributions

Our study represents a collaborative effort between different preclinical and clinical groups with diverse experimental expertise. The original conception and design of our study was born in the ‘Big-HEART’ network funded by the 7th Framework Program of the European Union. All authors approved the final version of the manuscript. Conception and design of the experiments: E.R.W.-P., J.v.d.V., C.P. Collection, analysis and interpretation of data: Myofibril preparations data: C.F., B.S., N.P., C.T., C.P. Multicellular muscle strips data: E.R.W.-P., G.J.M.S., J.v.d.V. mRNA data: J.M., T.K. Clinical and patient data: C.H., M.M., C.P. Drafting the article or revising it critically for important intellectual content: E.R.W.-P., J.M., C.T., G.J.M.S., M.M., C.Y.H., T.K., C.P., J.v.d.V.

Funding

We acknowledge support from the 7th Framework Program of the European Union (‘Big-HEART’, grant agreement 2415577), the Netherlands organization for scientific research (NWO; VIDI grant), Telethon Italy (grant GGP07133) and the Deutsche Forschungsgemeinschaft (DFG) (KR1187119-1).