INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

ALTHOUGH MOST CARDIAC HYPERTROPHY occurs as a secondary response to hemodynamic overload or other mechanical factors, hypertrophy can also be induced by inherited mutations (28). A number of different missense mutations in the head and head/rod region of the -cardiac myosin heavy chain (MHC) gene can cause familial hypertrophic cardiomyopathy (FHC) (7, 8, 10, 24). One of these mutations, Arg⁴⁰³  Gln has nearly 100% penetrance and dramatically reduces the life expectancy of affected individuals (8). However, the mechanism by which the Arg⁴⁰³  Gln mutation and other mutations in sarcomere protein genes cause cardiac hypertrophy and sudden death is not certain.

To further understand the physiological and pathological mechanisms by which the Arg⁴⁰³  Gln missense mutation causes FHC, a murine model was created by introducing this missense mutation into the -cardiac MHC gene (9). Heterozygous mice bearing the Arg⁴⁰³ Gln missense mutation (designated -MHC^403/+) exhibit alterations in cardiac function and histopathology characteristic of human FHC, homologous to human -cardiac MHC, with 92% identity overall (9). The -MHC^403/+ mouse hearts exhibit lower cardiac output and abnormal left ventricular (LV) relaxation, and they also demonstrate fibrosis. Whether the alteration in cardiac function of mutant hearts is directly due to defective myocyte function or whether these differences in function are secondary to changes in cardiac structure and/or loading conditions remains uncertain.

To address this question, we have studied the structure and function of isolated cardiac myocytes bearing the Arg⁴⁰³  Gln mutation from left ventricles of -MHC^403/+ mice. We examined the morphology of the myocytes and observed that many of the mutant myocytes were shorter and wider than wild-type (WT) myocytes, with irregular cellular borders and myofibrillar disarray. We measured contractile and relaxation properties by means of computer-assisted video motion edge detection. Measuring these indexes from single cardiac myocytes provides intrinsic contractile and relaxation function, independent of the extracellular matrix and systemic hemodynamic and neurohormonal effects. After we determined that myocyte relaxation was delayed and accompanied by delayed recovery of Ca²⁺ transients, we measured the message for sarco(endo)plasmic reticulum Ca²⁺-ATPase (SERCA) to see whether it was altered in the mutant.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Heterozygous -MHC^403/+ mice were bred by mating -MHC^403/+ mice as described previously (9). WT mice were the littermates of -MHC^403/+ mice.

Preparation of mouse ventricular myocytes. Cardiac myocytes were prepared from 15-wk-old male wild-type (n = 17) and -MHC^403/+ mice (n = 19). In brief, the heart was rapidly excised and submerged in Ca²⁺-free Tyrode solution containing (in mM) 140 NaCl, 5.4 KCl, 1 MgCl₂, 0.33 NaH₂PO₄, 10 glucose, and 5 HEPES at pH 7.4. The aorta was cannulated with a blunt-tip needle (20 gauge) on a perfusion apparatus. The heart was perfused for 3 min with Tyrode solution and was then perfused with Tyrode solution with 2% calf serum (Sigma) and 75 U/ml each of collagenase types 1 and 2 (Worthington). Perfusion pressure was monitored continuously through an in-line pressure transducer during digestion. Perfused LV free walls were excised when perfusion pressure reached ~20-30 mmHg from an initial pressure of 60-70 mmHg, which provided a consistent quality of myocytes (>50% yield). Perfusion was identical for both groups. The digested tissue was then placed in the separate petri dish filled with Tyrode solution with 2% calf serum and 10% bovine albumin (fraction V, Sigma). All perfused media were maintained at 37°C, and all solutions were continuously bubbled with 95% O₂-5% CO₂. After the tissue was minced and gently aspirated, the cells were teased apart to disperse myocytes, and the myocyte-containing solution was placed in a tube to allow viable myocytes to settle by gravity. Myocytes were stored at room temperature before use, and the protocol was completed within 3 h after isolation to avoid any deterioration due to prolonged storage of cells.

Measurement of contractile performance. Myocytes were transferred to a warmed (37°C) and continuously perfused cell chamber located on the stage of an inverted microscope (Nikon). The chamber was perfused with physiological buffer containing (in mM) 120 NaCl, 2.6 KCl, 1.2 MgCl₂, 1.2 KH₂PO₄, 11 glucose, 5 HEPES, 25 NaHCO₃, 2 taurine, 1 pyruvate, and 1 Ca²⁺. Myocyte contraction was induced once per second (1 Hz) by platinum field electrodes placed in the cell chamber that were attached to a stimulator (Grass S48, Grass Instruments). Cell images were continuously acquired through a ×40 objective lens (Nikon) and transmitted to a 240 sample/s charge-coupled device (CCD) video camera (TM-640, Pulnix). The output from the CCD camera was displayed on a video monitor (PVM-135, Sony). Myocytes were selected for study according to the following criteria (6): a rod-shaped appearance with clear striations and no membrane blebs, no spontaneous contractions when unstimulated in 1 mM Ca²⁺, and steady diastolic length and contractile amplitude at basal stimulation rates. Myocyte length was measured using a video motion edge detector (VED103, Crescent Electronics), and the data were acquired at 240 samples/s, stored, and analyzed on a Dell 433 computer. Myocyte dimensions were calibrated with a hemacytometer grid placed on the microscope stage.

Measurement of myocyte transient Ca²⁺ concentration. LV myocytes from 9 -MHC^403/+ and 10 WT mice were loaded with 3.8 mM of fura 2-AM (Sigma) dissolved in dimethyl sulfoxide at room temperature (20°C) for 30 min in Tyrode solution with 2% calf serum and 10% bovine albumin (fraction V, Sigma). After cells were loaded, they were washed with Tyrode solution for 30 min and placed in the myocyte perfusion chamber on the microscope, as described in Measurement of contractile performance. The myocytes were excited by ultraviolet light (wavelengths 340 and 380 nm, alternately) and the fura emission wavelength (510 nm) was synchronously monitored by the Photoscan dual-beam spectrofluorophotometer (Photon Technology). Intracellular free Ca²⁺ was measured as the fluorescence ratio (340/380) because high mitochondrial fluorescence affects calculated Ca²⁺ concentration in fura 2-loaded myocytes, and the measurement averages the fluorescent signal from an area within a single cell (29). Loaded myocytes were stimulated as described above. Measurements from an individual myocyte were taken after a steady state for myocyte contraction was reached following stimulation at 1 Hz. The protocol was completed within 3 h after isolation to avoid any deterioration due to prolonged storage of cells.

Morphological evaluation. Light microscopic observation of isolated myocytes was done to evaluate differences in cell morphology. Freshly isolated myocytes from both -MHC^403/+ and WT mice were fixed in 4% paraformaldehyde. An aliquot of cells was washed in PBS with 0.5% Tween 20 and incubated with 10 µg/ml of rhodamine-conjugated phalloidin (Sigma) for 30 min at room temperature. After the cells were stained and washed in PBS, they were mounted and observed with a Leica confocal laser-scanning microscope. Both differential interference contrast and fluorescent images were collected and analyzed using PhotoImage. In addition, two -MHC^403/+ and two WT mouse hearts were fixed with 2% phosphate-buffered paraformaldehyde. The LV myocardial samples were embedded in aldehyde epoxy resin. Sections were cut to a 1-µm thickness and stained with toluidine blue. Light microscopic evaluation was used to determine in situ myocyte morphology independent of the perfusion pressure used in isolated myocytes.

RNA extraction and Northern blot analysis. Total RNA was isolated from the left ventricles of WT and -MHC^403/+ mice using Trizol (GIBCO BRL) according to the manufacturer's protocol. Total RNA (20 µg) was separated on an agarose-formaldehyde gel and subsequently blotted to a nylon hybridization transfer membrane (GeneScreen Plus, NEN Life Science Products). The oligonucleotides used as transcript-specific probes were as follows: SERCA 5'-AACAACGCACATGCACGCACCCGAACACCCTTATATTTCTGCAAATGG and 5'-GGAACATGTAGACCATGTAGTTGAGGTCAATGAAG. The hybridization signal for SERCA was normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) signal intensity.

Data analysis. LV myocytes from four -MHC^403/+ (n = 1,020) and four WT mice (n = 860) were classified morphologically using light microscopy: type I, rod shaped with parallel myofibrils; type II, irregularly shaped, shorter and wider than other types, with parallel myofibrils; and type III, irregularly shaped with disoriented myofibrils. The percentage of myocytes from each type was calculated from each animal, and these data were used for the mean values.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

LV myocytes from both -MHC^403/+ and wild-type mice were classified morphologically. Compared with WT myocytes, -MHC⁴⁰³⁺ myocytes showed more irregular cell outlines with apparently more cell junctional areas and less well-aligned myofibrils within the cells but with no change in sarcomere length, which ranged between 1.81 and 1.87 mm (Figs. 1 and 2). Compared with WT hearts, -MHC^403/+ hearts had fewer type I cells (WT = 74 ± 3%; -MHC^403/+ = 41 ± 4%, P < 0.01) and more type III cells (WT = 12 ± 3%; -MHC^403/+ = 49 ± 7%, P < 0.01). The altered morphology was observed in myocytes isolated in both Ca²⁺-containing (Fig. 1) and Ca²⁺-free (Fig. 2) solutions. Both myocytes and tissue sections (Fig. 1, C and D) from -MHC^403/+ mice showed irregular morphology, i.e., shorter in length and wider than WT, with misaligned myofibrils. Staining with rhodamine-phalloidin for actin filaments and examination by confocal microscopy revealed disorganization of the myofibrils in type III cells (Fig. 2).

View larger version (112K): [in this window] [in a new window]   Fig. 1.   Morphology of wild-type (WT; a and b) and -myosin heavy chain 403/+ myocytes (-MHC403/+; c and d) showing myofibril disorientation in both -MHC403/+ isolated myocytes (c) and in situ (d). Myocyte and tissue section from -MHC403/+ mice showed irregular morphology, i.e., shorter in length and wider with misaligned myofibrils.

View larger version (102K): [in this window] [in a new window]   Fig. 2.   Confocal fluorescent (a and c) and differential interference contrast (b and d) images of left ventricular myocytes from WT (a and b) and -MHC403/+ mice (c and d) showing disorganization of myofibrils with more cell junctional areas in -MHC403/+ myocytes.

Indexes of contractile and relaxation function are summarized in Table 1. In -MHC^403/+ LV myocytes, diastolic and systolic lengths were significantly less than in WT myocytes (WT = 130 ± 3, -MHC^403/+ = 118 ± 3 µm, P < 0.05), but percent contraction was similar (Fig. 3). Both contractile and relaxation properties of -MHC^403/+ mice were significantly impaired (P < 0.05) compared with those of WT mice. For example, the rate of contraction (dL/dt_max), contraction slope, and relaxation times (TR 50% and TR 70%) were altered by 20-25% in the -MHC^403/+ myocytes compared with WT myocytes (Table 1 and Fig. 4).

View this table: [in this window] [in a new window]   Table 1.   Contraction and relaxation in male WT and -MHC403/+ myocytes

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Representative contraction recordings from WT left ventricular myocyte compared with -MHC403/+ myocyte demonstrating that resting length (left) from mutant myocyte is shorter and velocity of shortening and relengthening (dL/dt; right) are also significantly delayed. Numbers in left panel indicate %contraction.

View larger version (18K): [in this window] [in a new window]   Fig. 4.   Representative contraction (left) and dL/dt (right) recordings from male WT left ventricular myocyte compared with -MHC403/+ myocyte. Calculation for contraction (cont) slope, which was calculated between 10 and 90% contraction, and times to 50% and 70% relaxation (TR 50% and TR 70%) are also shown.

Because there were more type III cells, which were shorter and wider with disoriented myofibrils, in the -MHC^403/+, it was important to determine whether this alone could account for the differences in contractile function. Accordingly, we divided the data from the mutant myocytes into those myocytes with less than the mean end-diastolic length (<120 µm) and those with an end-diastolic length >120 µm. The longer cells were predominantly type I, rod-shaped myocytes with parallel myofibrils. Whereas both subgroups of myocytes demonstrated impaired contractile and relaxation function (Table 2), the impairment was greater in shorter cells.

View this table: [in this window] [in a new window]   Table 2.   Effects of cell morphology on contraction and relaxation in male WT and -MHC403/+ myocytes

Changes in Ca²⁺ concentration during relaxation and contraction. To begin to understand why -MHC^403/+-derived LV myocytes had impaired function, Ca²⁺ concentrations were recorded during the contraction-relaxation cycle of isolated myocytes. The changes in intracellular Ca²⁺ concentration associated with contraction were estimated by comparing the emission spectra of fura 2 after excitation of loaded myocytes at 340 and 380 nm. The levels of diastolic free Ca²⁺ concentration and Ca²⁺ signal during contraction, monitored as the 340/380 ratio, was similar in WT and -MHC^403/+ myocytes (Table 3). However, the times required for decay of this signal by 50 and 70% resting lengths (TRC 50% and TRC 70%, respectively) were significantly prolonged (P < 0.001), suggesting an impaired ability to sequester Ca²⁺ into the sarcoplasmic reticulum during relaxation (Table 3 and Fig. 5).

View this table: [in this window] [in a new window]   Table 3.   Ca2+ transient in left ventricular myocytes from WT and -MHC403/+ mice

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Representative Ca2+ transient recordings from male WT left ventricular myocyte compared with -MHC403/+ myocyte. Calculations for recovery of Ca2+ signal to 50 and 70% of peak Ca2+ change (TRC 50% and TRC 70%) are also shown. In mutant myocytes, TRC 50% and TRC 70% were significantly delayed.

Northern blot analysis of SERCA. Expression of SERCA mRNA from WT and -MHC^403/+ mice was similar, i.e., no changes in the 28S and 18S bands were observed (Fig. 6). The normalization to GAPDH also demonstrated no differences (WT = 2.01 ± 0.18, -MHC^403/+ = 2.13 ± 0.13).

View larger version (44K): [in this window] [in a new window]   Fig. 6.   Northern blot analysis of total RNA from WT (+/+) and -MHC403/+ (+/403) mice. A: total RNA (20 µg) showing 28S and 18S rRNA bands. B: Northern blot of sarco(endo)plasmic reticulum Ca2+-ATPase (SERCA). C: Northern blot of glyceraldehyde-3-phosphate dehydrogenase (GAPDH).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Understanding the effects of the Arg⁴⁰³  Gln missense mutation on myocyte contraction and relaxation is central to understanding how mutations in the cardiac MHC gene cause FHC. In the present study, we demonstrated that many of the myocytes from heterozygous -MHC^403/+ mice, which bear the FHC-causing mutation Arg⁴⁰³  Gln, are morphologically altered, i.e., shorter and wider with marked misalignment of myofibrils, compared with myocytes from WT mice.

In an earlier study by Maron et al. (23) in 52 patients with hypertrophic cardiomyopathy, cellular disorganization was widely distributed throughout the left ventricle as a hallmark of hypertrophic cardiomyopathy. Recently, Geisterfer-Lowrance et al. (9) generated a mouse model simulation of FHC by introducing an Arg⁴⁰³  Gln mutation into the -cardiac MHC gene, which is highly homologous to human -cardiac MHC with 92% identity overall. They (9) demonstrated that cardiac dysfunction, i.e., lower cardiac output and abnormal LV relaxation, preceded histopathological changes such as ventricular hypertrophy with fibrosis, indicating that altered mechanical properties caused abnormal cardiac function. In the present study, isolation of single cardiac myocytes, which provides intrinsic myocyte function independent of systemic hemodynamic and neurohormonal effects, revealed that both contraction and relaxation properties of -MHC^403/+ myocytes were significantly impaired. Marian et al. (22) recently demonstrated that human myocytes expressing mutant cardiac troponin T (Arg⁹²  Gln), known to cause hypertrophic cardiomyopathy in humans, demonstrated impaired contractility.

Interestingly, we noted that many LV -MHC^403/+ myocytes were shorter and wider, with marked misalignment of myofibrils, compared with WT myocytes. However, sarcomere lengths were not altered. Not all myopathic myocytes demonstrate these abnormalities, e.g., the myocytes from cardiomyopathic Syrian hamsters are actually longer and wider (14, 15). One possible difference between their results and ours could be due to the amount of Ca²⁺ in solution. Therefore, to determine whether Ca²⁺ in the isolation solution was responsible for the alteration of cell morphology, the myocytes in the present study were also isolated in Ca²⁺-free solution, as shown in Fig. 2. Under these conditions, we confirmed that -MHC^403/+ myocytes, which were isolated without Ca²⁺ in the solution, were also shorter and wider. Another consideration is that the isolation procedures caused the disarray in myocyte architecture. However, isolation procedures were identical for both groups. Furthermore, we also observed disorientation of myofibrils in situ in perfusion-fixed myocardium without myocyte isolation (Fig. 1). Accordingly, the presence or absence of Ca²⁺ in the isolation solution or the effects of enzymatic myocyte isolation was not responsible for the altered morphology. Another important morphological aspect of the LV -MHC^403/+ myocytes is an irregular cell outline with apparently more cell junctional areas (Fig. 1). More importantly, we also noted that many -MHC^403/+ myocytes showed less well-aligned myofibrils within the cells, which may be responsible for the primary functional defects.

Because the population of shorter and wider cells was increased in -MHC^403/+, it was important to determine whether this alone was responsible for the depressed contractile (dL/dt) and relaxation function (+dL/dt; TR 50% and TR 70%). When myocytes of similar length, generally characterized by aligned myofibrils, were compared (Table 2), it was found that the -MHC^403/+ myocytes also demonstrated impaired contraction and relaxation function, albeit less so than the shorter myocytes. Thus the shorter and wider myocytes with greater misalignment of myofibrils demonstrated more severe expression of the myopathy. We then wanted to determine whether the difference in end-diastolic length resulted in differences in the extent of shortening. When these two variables were correlated, it was found that end-diastolic length correlated significantly with the extent of shortening of myocytes, but there was no correlation with percent contraction of the myocytes (Fig. 7). This indicates that the extent of shortening was preserved regardless of myocyte length in -MHC^403/+ mice. However, the rates of contraction and relengthening as well as the times for 50 and 70% recovery were depressed in LV myocytes from -MHC^403/+ mice.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Relationships between diastolic length and extent shortening (left) and %contraction (right) in WT and -MHC403/+ mice. Extent of shortening (y = 0.04x + 0.06, r = 0.60), but not %contraction (y = 0.006x + 3.8, r = 0.02), was correlated with diastolic length, regardless of myocyte length in both WT and -MHC403/+ mice. 403L, longer than the mean end-diastolic length (>120 µm); 403S, shorter than the mean end-diastolic length (<120 µm).

In addition to distinct abnormalities of morphology and mechanical function of -MHC^403/+ myocytes, we utilized an indirect measurement of intracellular Ca²⁺ to understand the underlying mechanism of contractile and relaxation dysfunction. In an earlier study, Sen et al. (25), using the Syrian hamster (Bio 14.6 strain) as a model of the hereditary cardiomyopathy, observed distinct abnormalities of contractile function in cardiomyopathic cells, i.e., decreases in amplitude and velocity of contraction but not in diastolic relaxation velocity. They (25) also noted that mean cytosolic Ca²⁺ was significantly higher and Ca²⁺ uptake was significantly increased in cardiomyopathic cells compared with normal control cells, suggesting that abnormal myocardial Ca²⁺ homeostasis has a possible role in contractile dysfunction. This has been noted by others (14, 15) in hereditary cardiomyopathy. In the present study, we observed delayed sequestration of Ca²⁺ into the sarcoplasmic reticulum during relaxation, although the levels of diastolic free Ca²⁺ concentration were similar between WT and -MHC^403/+ mice. Several other studies of pressure-overload hypertrophy have identified delayed relaxation in isolated papillary muscles and isolated myocytes using human (11) and various other models such as ferret (12), rat (13, 27), guinea pig (17), and cat (1-3). Furthermore, prior studies also demonstrated that the mechanism responsible for the diastolic dysfunction in pressure-overload hypertrophy involves impaired Ca²⁺ reuptake due to a decrease in SERCA (4, 5, 17-19, 21). In addition, transgenic mice with cardiac-specific overexpressed phospholamban exhibit decreases in baseline contractile and relaxation function that are associated with decreases in the amplitude of the Ca²⁺ transient and prolongation of Ca²⁺ reuptake in isolated myocytes (16). One possibility for the impaired reuptake of Ca²⁺ with -MHC^430/+ mice involved decreased SERCA levels. However, we found no change in message levels of SERCA in -MHC^430/+ myocytes compared with those in WT myocytes. Interestingly, prior studies by Sweeney et al. (26) and Lankford et al. (20) demonstrated that replacement of Arg-403 with Gln decreased the rate of transition within the actin-myosin cross-bridge cycle, depressed myosin ATPase activity, and displayed abnormal force-velocity relationships. Thus, although we do not know the exact mechanism for the alteration of Ca²⁺ reuptake, it is conceivable that the disorientation of myofibrils alters cross-bridge kinetics, which could result in not only contractile but also relaxation dysfunction.

In conclusion, -MHC^403/+ myocytes demonstrated impaired contraction and relaxation. The impaired function in -MHC^403/+ myocytes may be due, at least in part, to distinct abnormalities of morphology, i.e., marked misalignment of myofibrils. Potentially, these myocytes with abnormal morphology are those that exhibit a more severe phenotype of the Arg⁴⁰³  Gln missense mutation. The altered cardiac dysfunction of -MHC^403/+ myocytes is directly due to defective myocyte function rather than to secondary changes in global cardiac function and/or loading conditions. Furthermore, alterations in the rate of contraction and relaxation in the face of preserved percent contraction, in combination with preserved ejection fraction (unpublished data), support the concept that the reduced rate of contraction and relaxation in isolated myocytes is an early finding that occurs before decreased global systolic function.