Shifts in the myosin heavy chain isozymes in the mouse heart result in increased energy efficiency

Abstract

Cardiac-specific transgenesis in the mouse is widely used to study the basic biology and chemistry of the heart and to model human cardiovascular disease. A fundamental difference between mouse and human hearts is the background motor protein: mouse hearts contain predominantly the αα-myosin heavy chain (MyHC) isozyme while human hearts contain predominantly the ββ-MyHC isozyme. Although the intrinsic differences in mechanical and enzymatic properties of the αα- and ββ-MyHC molecules are well known, the consequences of isozyme shifts on energetic of the intact beating heart remain unknown. Therefore, we compared the free energy of ATP hydrolysis (|ΔG_~ATP|) determined by ³¹P NMR spectroscopy in isolated perfused littermate mouse hearts containing the same amount of myosin comprised of either >95% αα-MyHC or ~83% ββ-MyHC. |ΔG_~ATP| was ~2 kJ mol⁻¹ higher in the ββ-MyHC hearts at all workloads. Furthermore, upon inotropic challenge, hearts containing predominantly ββ-MyHC hearts increased developed pressure more than αα-MyHC hearts whereas heart rate increased more in αα-MyHC hearts. Thus, hearts containing predominantly the ββ-MyHC isozyme are more energy efficient than αα-MyHC hearts. We suggest that these fundamental differences in the motor protein energy efficiency at the whole heart level should be considered when interpreting results using mouse-based cardiovascular modeling of normal and diseased human heart.

1. Introduction

Myosin is the primary mechanoenzyme responsible for muscle contraction, coupling the energy released from ATP hydrolysis (|ΔG_~ATP|) to force generation. It is also the primary ATPase in muscle, consuming most of the ATP used for contraction. Myosin is a highly conserved protein composed of two myosin heavy chains (MyHC) and four light chains [1]. In vertebrate hearts, two isozymes of MyHC exist. The αα-homodimer predominates in cardiac ventricles of small mammals [1], the ββ-homodimer in ventricles of large mammals and man [2]. Despite the fact that the two isozymes are highly homologous, with 93% amino acid identity [3], their physiologic and biochemical properties differ. Actin- and Ca²⁺-activated ATPase activities of the different myosin isozymes determine the velocity of sarcomere contraction [4].

In cardiac ventricles, developmental stage, thyroid status, intermittent exercise, chronic work overload, and diseased states all alter MyHC composition. In end-stage failing human heart, for example, αα-MyHC decreases until it is not detectable, leading to a relative increase in ββ-MyHC [5]. As ββ-MyHC is the more economical ATPase, this isoform shift may be energetically beneficial for the diseased heart. In mouse hearts, thyroid hormones induce αα-MyHC expression whereas hypothyroidism and pressure overload result in an increase in ββ-MyHC composition. An in vitro mobility assay has been used to demonstrate that αα-MyHC has twice the actin-activated and Ca²⁺-stimulated ATPase activity and three-times the actin filament sliding velocity whereas it produces only one-half the average cross-bridge force when compared with ββ-MyHC [6]. Consistent with this, myocyte shortening and relengthening are slower for hypothyroid rat myocytes containing more ββ-MyHC [7] and maximal velocity of unloaded shortening is slower in papillary muscles with more ββ-MyHC [8].

One limitation of these studies is that the cardiac isozyme shift was achieved by endocrine intervention, changing the expression of many proteins ([9] and references therein). Therefore, whether these functional changes are induced solely by the MyHC isozyme shift or by other biochemical and structural changes is uncertain. A different strategy for altering the composition of MyHC isozymes is to use transgenesis in mice. Cardiac-specific expression of the protein of interest has the advantage that any cardiac phenotype must be due, either directly or indirectly, to that protein. Tardiff et al. [10] investigated the functional significance of cardiac myosin isozyme diversity using transgenic mice in which 12% of the ventricular αα-MyHC was replaced by rat-ββ-MyHC. Krenz et al. [11] studied mice with 73% of ventricular αα-MyHC replaced by mouse-ββ-MyHC. Both studies demonstrated no overt morphological phenotype, but systolic function and the rate of relaxation in hearts containing more ββ-MyHC were decreased compared with non-transgenic hearts.

These results support the classic view that the mechanical properties of hearts containing different MyHC isozymes are due to innate differences in the mechanoenzyme. However, the energetic consequences of replacing one MyHC isozyme with the other in the intact beating heart remain undefined. Here we compare the energetics of non-transgenic mouse hearts containing >95% αα-MyHC [2] to genetically engineered mouse hearts with ~83% ββ-MyHC. We defined the energetic state of the isolated perfused heart using the non-invasive tool of ³¹P NMR spectroscopy to determine |ΔG_~ATP| while simultaneously measuring contractile performance. We found that mouse hearts bearing predominantly ββ-MyHC have an energetic phenotype that is distinct from hearts with native αα-MyHC.

2. Materials and methods

2.1. Mouse model

Female mice (FVB/N) with ~83% replacement of ventricular αα-MyHC with mouse ββ-MyHC were generated as described [12]. The isozyme content of MyHC was determined using glycerol gels as described in [11]. Four- to six-month-old female non-transgenic littermates were used as controls. Animals were maintained in accordance with NIH guidelines for the use and care of laboratory animals. The experimental protocols of this study were all in compliance with the recommendations of current NIH and American Physiological Society guidelines.

2.2. Isolated perfused mice hearts

Hearts were isolated and perfused in the Langendorff mode with phosphate-free Krebs-Henseleit buffer as described [13]. Systolic function was assessed as systolic pressure (SP), developed pressure (DevP), rate pressure product (RPP), and wall stress as global average stress [14]. Contractility was estimated as +dP/dt (the maximum value of the first derivative of LV pressure). Diastolic function was assessed as EDP (end-diastolic pressure) and -dP/dt (the minimum value of the first derivative of LV pressure). EDP was set to 10 ± 1 mm Hg at baseline conditions and then allowed to change with changing conditions. For a description of the measurement of indices of isovolumic contractile performance, see [13].

2.3. Experimental protocols

Hearts containing >95% αα-MyHC (referred to as αα-MyHC hearts) and ~83% ββ-MyHC (referred to as ββ-MyHC hearts) were used for ³¹P NMR spectroscopy experiments; measures of contractile performance were made throughout. Hearts were paced at 7 Hz for 20 min, and then challenged with dobutamine (300 nM; unpaced) for 11 min, followed by 30 min recovery without dobutamine delivery. At the end of each protocol hearts were blotted dry, weighed and stored at −80 °C.

2.4. ³¹P NMR spectroscopy

³¹P NMR free induction decays (FIDs) were acquired at 161.4 MHz using a Varian Inova spectrometer (Varian, Palo Alto, CA), averaging 200 FIDs (60° pulse, 2.5 sec recycle time). The resonance areas and chemical shifts (intracellular pH) were quantified using the MacNuts–Utility transform software (Acorn NMR, Livermore, CA). Values of ΔG_~ATP are expressed as their absolute numbers and were calculated as shown in the following equation:

$$\left|\mathrm{\Delta }{G}_{\sim \mathit{ATP}}\right|(kJ/\mathit{mol})=|\mathrm{\Delta }{G^0}_{\sim \mathit{ATP}}+RT\hspace{0.17em}ln\hspace{0.17em}([\mathit{ADP}][Pi]/[\mathit{ATP}])|$$

where ΔG⁰ is the standard free energy change of ATP hydrolysis under standard conditions of molarity, temperature, pH, and [Mg2+]; R is the gas constant (8.3 J/mol K), and T is in Kelvin. Free cytosolic [ADP] was calculated using the equilibrium expression for the creatine kinase (CK) reaction [13] and measured metabolite contents.

2.5. Myocardial oxygen consumption (MVO₂)

MVO₂ was measured using Clark type microelectrodes (Microelectrodes Inc., Bedford, NH). The pulmonary artery was cannulated with a PE-10 catheter connected via a T-piece to an electrode. Perfusate was withdrawn from the pulmonary artery with a constant flow of 0.1 ml/min by a pulsation-free pump. The T-piece in front of the O₂ electrode allowed the outflow of excessive pulmonary flow. Inflowing O₂ tension was measured with a second electrode at the same flow rate. Both electrodes were kept at 37°C. MVO₂ was calculated using the difference of inflowing and outflowing O₂ tensions, coronary flow rate, heart weight, and the molar O₂ solubility coefficient [15].

2.6. Biochemical measurements

Five to 10 mg of ventricular tissue was homogenized for 10 sec at 4°C in K-phosphate buffer (0.1 M) containing 1 mM EDTA and 1 mM β–mercaptoethanol, pH 7.4 (final concentration of 5 mg wet weight/ml). Aliquots were removed to determine protein [16] and total creatine contents [17]. Triton X-100 was added to the homogenate at a final concentration of 0.1 % for analysis of activities of CK [18]; of the major glycolytic enzymes phosphofructokinase (PFK) [19] and lactate dehydrogenase (LDH) [20]; and of the mitochondrial markers citrate synthase (CS) [21] and cytochrome C oxidase (COX, measured with a Sigma assay kit, [22]).

2.7. Statistical analysis

Results are expressed as means ± SEM. Comparisons were tested with one-way ANOVA followed by a multiple comparison test using Fisher’s protected least significant difference; P<0.05 was considered significant. Linear regressions and tests whether slopes and intercepts differed were calculated with Prism (GraphPad, San Diego, CA).

3. Results

3.1. Properties of the hearts

To determine whether whole heart energetics differed or were the same in αα- and ββ-MyHC hearts, it was necessary to make the comparison under identical experimental conditions, using pairs of 4–6 month-old littermate mouse hearts, one from each group, perfused in the Langendorff mode. Cardiomyocytes rigidly control the stoichiometry of the contractile proteins within the sarcomere, and the total amount of myosin in the two types of hearts was the same [11]. Furthermore, previously analyzed transcript levels of a number of heart failure/hypertrophy related genes such as atrial natriuretic factor, brain natriuretic peptide, alpha-skeletal actin, and phospholamban in right- and left-ventricular tissue samples from αα- and ββ-MyHC mice did not show discernable differences (Krenz & Robbins, unpublished data). Body weights of 25.9 ± 0.2 g and 26.8 ± 0.3 g and heart weights of 94.8 ± 1.6 mg and 105.3 ± 1.5 mg were indistinguishable for the αα- and ββ-MyHC mice (n=13 for both groups), respectively. Left and right ventricular weights were also comparable (data not shown); atrial weights were higher in the ββ-MyHC hearts (10.2 ± 0.35 mg vs. 6.2 ± 0.06 mg, P<0.001). Basal perfusion conditions were identical for both heart groups: perfusion pressure of 80 mmHg, pacing rate of 7 Hz, placing and filling the balloon (volume for αα-MyHC: 13.8 ± 1.3 μl; for ββ-MyHC: 15.3 ± 1.0 μl, n=8) in the same way to set the EDP at ~10 mmHg, and coronary flow of ~2 ml min⁻¹. The stimulus to increase workload was the inotropic drug dobutamine, delivered to a final concentration of 300 nM for each heart.

3.2. Energetics

To determine whether the changes in the MyHC isozyme composition lead to differences in the energetic state of the intact beating heart, we obtained sequential ³¹P NMR spectra at three time points. We determined ATP, ADP, phosphocreatine (PCr), and inorganic phosphate (Pi) concentrations in the hearts at baseline, during high workload, and during recovery from the high work. Figure 1 shows summed ³¹P NMR spectra acquired during baseline perfusion and inotropic challenge from αα- and ββ-MyHC hearts and Table 1 summarizes the results.

Figure 1.

Summed ³¹P NMR spectra from four αα-MyHC (A and B) and ββ-MyHC (C and D) hearts. A and C show spectra at baseline; B and D during inotropic challenge. Resonances correspond to (from left to right) Pi, PCr, and the three phosphate groups of ATP (γ, α, β). The height of the PCr resonance for αα-MyHC hearts at baseline is set to 1.0 to allow comparisons. Note that the PCr resonance for ββ-MyHC hearts at baseline is higher than for the αα-hearts. During dobutamine treatment, areas of ATP resonances do not change, but PCr resonance area decreases while Pi increases more in the αα-MyHC than in the ββ-MyHC hearts. *: P ≤ 0.05 for αα-MyHC vs. ββ-MyHC, ⁺: P ≤ 0.05 for baseline vs. inotropic challenge within the groups.

Table 1. Cardiac energetics.

	αα-MyHC		ββ-MyHC
	baseline	dobutamine	baseline	dobutamine
Pi (mM)	2.8* ± 0.54	10.1+ ± 0.58	1.2 ± 0.30	5.7 ± 0.39
PCr (mM)	18.6* ± 0.34	12.5+ ± 0.56	20.5 ± 0.89	14.9 ± 0.40
pHi	7.13 ± 0.20	7.10 ± 0.01	7.14 ± 0.01	7.10 ± 0.01
ATP (mM)	10.1 ± 0.01	9.7 ± 0.27	10.3 ± 0.32	9.6 ± 0.41
ADP (μM)	64 ± 5	123 ± 11	61 ± 6	101 ± 8
|ΔG~ATP|(kJ mol−1)	59.3* ± 0.7	53.7+ ± 0.3	61.6 ± 0.8	55.7 ± 0.2

All results are given as mean ± SEM (n=8) at baseline and during inotropic challenge (dobutamine) in both MyHC-groups.

^*P<0.05 baseline αα-MyHC vs. ββ-MyHC,

⁺P<0.05 inotropic challenged αα-MyHC vs. ββ-MyHC.

During baseline perfusion, [ATP], [ADP] and pH_i for ββ- and αα-MyHC hearts were indistinguishable. However, [PCr] was 10% higher (P<0.05) while [Pi] was 56% lower (P<0.05) in the ββ-MyHC hearts. Based on previous experiments comparing baseline and inotropic challenge in isolated mouse hearts [23], we expected [PCr] to decrease, [Pi] and [ADP] to increase, and [ATP] and intracellular pH (pH_i) not to change with high workload. This pattern demonstrating increased energy utilization with increased workload at the whole heart level was observed here, i.e. the CK reaction supplied sufficient phosphoryl groups to match ATP utilization at high workload by hydrolyzing PCr. Of interest, the greater [PCr] observed in the ββ-MyHC hearts at baseline persisted even at high workloads.

Changes in [ATP], [ADP], and [Pi] can be expressed as |ΔG_~ATP|, representing the chemical driving force for the ATPase reactions in the cell and describing its energy state. |ΔG_~ATP| was higher in ββ-MyHC hearts (n=8 for both groups) at baseline (61.6 vs. 59.3 kJ mol⁻¹, P=0.03, Figure 2), during high workload (55.7 vs. 53.7 kJ mol⁻¹, P<0.001, Figure 2) and for recovering hearts (63.5 vs. 60.4 kJ mol⁻¹, P<0.009). Note that the difference in |ΔG_~ATP| of 2–3 kJ mol⁻¹ requires large differences in [ATP], [ADP], and/or [Pi]. Since the maximum difference in |ΔG_~ATP| observed for the intact heart caused by an abrupt increase in workload is ~ 6 kJ mol⁻¹ within each group, the difference of ~2–3 kJ mol⁻¹ between the two groups is ~33–50 % of the maximum.

Figure 2.

The free energy of ATP hydrolysis (|ΔG_~ATP|) at baseline and during dobutamine treatment for individual hearts. Also shown is the average value marked for each condition (solid line). Note that |ΔG_~ATP| is higher for ββ-MyHC hearts (n=8) than for αα-MyHC hearts (n=8) at both work states. Means ± SEM.

3.3 Isovolumic contractile performance

To determine whether switching the MyHC isozyme composition from αα- to ββ-MyHC alters cardiac function, we measured isovolumic contractile performance during baseline, high workload, and recovery. Representative pressure tracings of αα- and ββ-MyHC hearts and mean values for indices of systolic and diastolic function are shown in Figure 3.

Figure 3.

Representative tracings of left ventricular isovolumic contractile performance. A, αα-MyHC and B, ββ-MyHC hearts at baseline; C, αα-MyHC and D, ββ-MyHC hearts during increased cardiac work. Balloon volume was set at an EDP of 10 mmHg at baseline and then allowed to vary. Hearts were paced at 420 bpm (7 Hz) only during baseline. DevP is the difference between the EDP and SP (dashed line); RPP (the product of HR times DevP). Also shown are the maximum and the minimum values of positive and negative dP/dt (solid line) as well as their ratio (+/− dP/dt) and the half time of relaxation at baseline. *P<0.05 basal αα-MyHC vs. ββ-MyHC, ⁺P<0.05 inotropic challenged αα-MyHC vs. ββ-MyHC, RPP during inotropic challenge in αα-MyHC vs. ββ-MyHC differ with a significance of ^#P=0.06. The quantitative changes within the groups after adding dobutamine are all significant except for the +/− dP/dt in the αα-MyHC mouse hearts. Note that the x-axes differ between C and D. Means ± SEM (n=8 for both groups).

During low work conditions (baseline), all indices of systolic contractile performance, namely SP, DevP, +dP/dt, and RPP, were similar for the two groups. Diastolic function, assessed as the rate of relaxation (−dP/dt) was >30% slower in the ββ-MyHC hearts, resulting in a slower half time of relaxation.

To determine if the ββ-MyHC hearts respond similarly to high workload conditions as the αα-MyHC hearts, we measured contractile performance in hearts challenged with 300 nM dobutamine, a dose producing maximal change in contractile performance. This resulted in increases in RPP in both groups (to 82,600 mmHg min⁻¹ in the ββ- and 74,000 mmHg min⁻¹ in the αα-MyHC hearts), but how it was achieved differed: In the αα-MyHC hearts, the heart rate (HR) increased by 24% and DevP by 71%; in the ββ-MyHC HR increased by only 15% while DevP increased by 128%. Thus, increases in HR and DevP both contributed to the increased RPP for αα-MyHC hearts whereas increased DevP was the main contributing factor for the increase in ββ-MyHC hearts. The increase in wall stress (to 176 ± 12 and 150 ± 10 kdynes (cm²)⁻¹ for ββ-MyHC hearts vs. αα-MyHC hearts) was also substantially greater in ββ-MyHC than in the αα-MyHC hearts (123% vs. 70%). Negative dP/dt increased so much during inotropic challenge in the ββ-MyHC hearts (248% vs. 150% for the αα-MyHC hearts) that the large difference in this parameter between the groups at baseline was abolished.

3.4. Relationship between energetics and contractile performance

The relationships between physiologic performance assessed as RPP and the energetic state assessed as |ΔG_~ATP| for the αα- and ββ-MyHC hearts using data obtained during baseline, inotropic challenge, and recovery are shown in Figure 4. Both groups demonstrated the expected decrease in |ΔG_~ATP| as RPP increased in response to the inotropic challenge and |ΔG_~ATP| returned to baseline values during recovery from high workload. Data are well fit by linear relationships. The slopes are indistinguishable for the two groups, but the y-intercept is higher (P < 0.001) for the ββ-MyHC hearts. Thus, for any given RPP, the energetic state is ~2–3 kJ mol⁻¹ higher for ββ-MyHC hearts, and for any given |ΔG_~ATP|, the ββ-MyHC hearts yield more isovolumic contractile work.

Figure 4.

Relationship between RPP and |ΔG_~ATP| for αα-MyHC and ββ-MyHC mouse hearts. Linear regressions are: y = −6267x + 406,700 with R² = 0.82 (αα-MyHC, circles) and y = −7140x + 476,600 with R² = 0.91 (ββ-MyHC, squares). While the slopes are not different, y-intercepts differ (P<0.0001). Filled symbols: baseline and during inotropic challenge; open symbols: during recovery.

3.5. ATP supply

To rule out any differences in the ATP synthesis pathways in hearts in which the αα-MyHC were replaced by ββ-MyHC, we performed two sets of experiments. The first was to determine the relationship between RPP and MVO₂ in the intact heart and the second was to determine Vmax of key enzymes of ATP synthesis pathways in extracts of hearts. There were no differences in the RPP/MVO₂ relationships or in the tissue activities of glycolytic enzymes, and key phosphoryltransferases (Table 2) between the two study groups. One marker of the capacity for oxidative phosphorylation, cytochrome C oxidase, was slightly lower (19%) in the ββ-MyHC than in the αα-MyHC hearts.

TABLE 2. ATP supply.

	αα-MyHC		ββ-MyHC
PFK	339 ± 13		351 ± 10
LDH	669 ± 16		698 ± 20
CS	396 ± 19		420 ± 20
COX	160* ± 8		130 ± 4
total CK	4294 ± 164		4248 ± 111
Creatine (mM)	31.0 ± 0.7		32.5 ± 1.3
MVO2 (μmol O2/gww min)	Baseline 10 ± 1	Dobutamine 16 ± 1	Baseline 10 ± 2	Dobutamine 17 ± 3

Representative enzymes of the glycolytic pathway (PFK, LDH) and oxidative phosphorylation (CS, COX), and total CK activity and creatine content were measured in tissue homogenates of αα-MyHC (n=8) and ββ-MyHC (n=8) hearts. Activities presented as mIU/mg protein. Oxygen consumption (MVO₂) was monitored during baseline and during treatment with dobutamine in both heart groups (n=5) as a measure of ATP synthesis rate. Means ± SE.

^*P<0.05

4. Discussion

In this study, we defined the energetic consequences of a bioengineered isozyme shift from 95% αα̃ MyHC/5%ββ-MyHC to 17% αα-MyHC/83% ββ-MyHC in intact beating mouse hearts. We found that the energetic phenotype of these two hearts differed. The ββ-MyHC hearts exhibited a higher energetic driving force expressed as the free energy available from ATP hydrolysis at all workloads: at baseline, during inotropic challenge and during recovery from high workloads. Since isovolumic contractile performance was greater in the ββ-MyHC than for αα-MyHC hearts for the same change in chemical driving force, we conclude that hearts containing predominantly ββ-MyHC are more efficient.

The different energetic phenotype at the whole heart level for ββ-MyHC hearts is due to the innate differences between the αα- and ββ-MyHC isozymes, and not due to differences in sarcomere length [11], myosin light chain composition [24], local pH or source of ATP used for contraction. Although filament sliding velocity is reduced under both acidic and alkaline pH [25], differences in pH cannot explain the mechanical differences observed here as pH_i was the same in αα- and ββ-MyHC hearts. Finally, neither the rate of oxygen consumed nor the capacity for ATP synthesis differed for αα-and ββ-MyHC hearts. Note that the slightly lower Vmax observed for COX in the ββ-MyHC hearts does not effect the oxygen uptake.

4.1. Basis for energetic phenotype of ββ-MyHC mouse hearts

In this study of the intact heart using the non-invasive tool of NMR spectroscopy, we found that under identical conditions of perfusion and oxygen consumed, ββ-MyHC hearts consistently had a higher |ΔG_~ATP|. In the two-state cross-bridge model of muscle contraction [26], the ββ-MyHC cross-bridge remains attached to the thin filament for a longer period of time [6, 7]. Thus, if whole heart contractile performance is determined primarily by the properties of the major motor protein, intact hearts containing more ββ-MyHC should have slower rates of relaxation and develop more force (systolic pressure) for the same energy cost, or equivalently have higher |ΔG_~ATP| for a given SP, DevP or RPP. Here we show that this is the case. In general agreement with results of others using smaller MyHC replacements [10, 11], we also found large differences in the rate of relaxation (-dP/dt, >30% slower) and the half-time of relaxation (~25% longer) for ββ-MyHC hearts. Importantly, the change in relaxation rate cannot explain the greater efficiency or the greater |ΔG_~ATP| observed in ββ-MyHC hearts because the differences in relaxation rate were abolished during inotropic challenge while the differences in |ΔG_~ATP| between the two types of hearts remained.

Upon inotropic challenge, increases in SP and DevP were greater in the ββ-hearts compared to the αα-hearts (123% vs. 74% and 128% vs. 71%, respectively, as the percentage change over baseline). Heart rate did not increase as high in the ββ-hearts as it did in the αα-hearts (15% vs. 24%). These results on the whole heart level can be explained by the intrinsic properties of the αα- and ββ-MyHC: the hearts of the ββ-MyHC isozyme have a higher force-generation associated with a lower heart rate. While cardiac performance is influenced by neurohumoral factors and by variations in Ca²⁺ regulation, we suggest that the major determinant of whole heart energetics when these factors are held constant is the nature of the motor protein.

The correlation between HR and the content of MyHC isozyme merits further comment. A trend towards lower HR has also been observed in other isolated heart studies [10, 11] as well as in propranolol-treated mice with predominantly ββ-MyHC [11]. In small animals, accumulating high levels of αα-MyHC is considered to be an adaptation because its faster hydrolysis rate facilitates the extremely rapid rates of cardiac contraction and relaxation [27]. Therefore, the lower heart rate found in the ββ-MyHC hearts may be the result of a long-term adaptation to the presence of slow MyHC isozyme.

In studies using hypothyroidism to shift the cardiac MyHC composition, it has not been possible to distinguish whether differences in tension development and relaxation are caused by the αα–to ββ-MyHC shift in the sarcomere or are due to altered Ca²⁺ handling properties changed by the thyroid state, or both [7, 9]. Here, using transgenesis to alter MyHC composition, we found that the different intrinsic rates of cross-bridge detachment and ATPase activities of the MyHC isozymes are sufficient to explain the differences in rates of tension development and relaxation and in |ΔG_~ATP| observed for intact beating hearts containing either predominantly ααMyHC or ββ-MyHC. This conclusion is also supported by the findings of others showing that the [Ca]-myofibrillar ATPase activity curve does not shift for ββ-MyHC[11] compared to αα-MyHC, that calcium transients were not altered in rat ventricular myocytes expressing human ββ-MyHC [28], and that there is no shift in MyHC isozyme composition in mouse cardiomyocytes with a decreased SERCA pump level with demonstrated changes in intracellular Ca²⁺ homeostasis and contractility[29].

4.2. ββ-MyHC mouse model

In all studies using mouse models with MyHC replacements, 12% [10], 73% [11, 12] and 83% (this study, [12]), the isozyme switch was found to be well tolerated and did not induce disease. However, the ability of hearts with high percentage replacement of endogenous αα-MyHC with ββ-MyHC (73 and 83%) to tolerate stress differs depending on whether the stress is acute, intermittent or chronic. The consequence of acute stress was investigated in both isolated hearts (this study) and in in vivo anesthetized animals [12]. Both αα- and ββ-MyHC hearts increased systolic performance in response to a treatment of the inotrope dobutamine. In the absence of neural and humoral control, isolated ββ-MyHC hearts responded with an even higher increase in systolic performance: RPP increased by 161% compared to 112% for the αα-MyHC hearts. In anesthetized intact mice, infusing dobutamine elicited directionally similar but much smaller increases (~25%) in systolic function. With intermittent stress caused by two swimming sessions of 90 min/day for two weeks, both αα- and ββ-MyHC hearts showed comparable hypertrophic response without signs of heart failure [12].

In contrast to tolerating acute and intermittent stresses, ββ-MyHC hearts did not tolerate chronic severe stress as well as αα-MyHC hearts [12]. In response to continuous infusion of the inotrope isoproterenol to mice for two weeks, SP was lower in mice with ββ-MyHC (130 vs. 150 mm Hg). Moreover, ββ-MyHC mice developed more pronounced cardiac dysfunction in a post-infarction failure model. Since high levels of αα-MyHC with fast ATP hydrolysis rate support rapid rates of contraction and relaxation of hearts of small animals [27], we suggest that replacing fast myosin with slow myosin leads to a chronic mismatch between enzyme kinetics and demand to increase work. This would contribute to the reduced ability of the hearts to withstand chronic severe stress despite being more energy efficient on a beat-to-beat basis.

4.3. ββ-MyHC mouse model and human disease

In end-stage failing hearts, the low αα-MyHC protein content decreases until it is undetectable [5], resulting in a relatively increased ββ-MyHC content. While the decrease in αα-MyHC composition may contribute to systolic dysfunction [30], based on results presented here, we suggest that hearts containing a greater fraction of the more energy efficient ββ-MyHC isozyme would be compensatory as long as the workload of the heart remains relatively low. This is supported by studies showing that pharmacological intervention having a negative inotropic effect (such as ACE inhibitors, angiotensin blockers, and β-blockers) improves outcome in heart failure ([31], references therein).

Genetic modification in the mouse is commonly used as a tool in the investigation of cardiac disease. One potentially important limitation of these animal models is that the primary energy consuming ATPase is not the isozyme present in the human heart [32]. The amino acid composition of mouse and rat αα- and ββ-MyHC differ in the same five functionally important domains of the myosin head and rod (for details see [1, 3]). Importantly, missense mutations resulting in Familial Hypertrophic Cardiomyopathy (FHC) are also found in these same domains [1, 33]. These mutations have unpredictable effects on mechanical properties of human as well as mouse hearts ([34–36], references therein). Although the underlying molecular mechanisms are not completely understood ([37], references therein), it has been proposed that altered cardiac energetics is a general mechanism underlying the phenotypic expression of malignant FHC-associated mutations ([38], references therein). We suggest that it may be informative to study this disease using a mouse model with a high replacement of αα-MyHC by ββ-MyHC.

5. Conclusion

Using intact beating mouse hearts genetically modified to replace αα-MyHC with ββ-MyHC, we found that hearts containing predominantly ββ-MyHC have higher mechanical-energetic efficiency. Thus, the mechanical and energetic properties of the whole heart differ depending on only a few amino acid changes in MyHC. We suggest that this should be taken into account when interpreting results using mouse-based cardiovascular modeling of normal and diseased hearts.

Abbreviations

CK

creatine kinase

COX

cytochrome C oxidase

CS

citrate synthase

|ΔG_~ATP|

free energy of ATP hydrolysis

DevP

developed pressure

− dP/dt

minimum value of the first derivative of LV pressure

+ dP/dt

maximum value of the first derivative of LV pressure

EDP

end-diastolic pressure

FHC

familial hypertrophic cardiomyopathy

FIDs

free induction decays

HR

heart rate

LDH

lactate dehydrogenase

LV

left ventricle

MVO₂

myocardial oxygen consumption

MyHC

myosin heavy chain

PCr

phosphocreatine

PFK

phosphofructokinase

pH_i

intracellular pH

Pi

inorganic phosphate

³¹P-NMR

phosphorus-31 nuclear magnetic resonance spectroscopy

SP

systolic pressure

RPP

rate pressure product