INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE FUNCTIONAL ABNORMALITIES of the cardiomyocyte underlying the transition from compensated hypertrophy to failure are unclear. In patients, virtually all data pertaining to this question have been obtained in myocardium from explanted, end-stage failing hearts after treatment with multiple drugs. Thus, for example, there is considerable documentation of a major defect in excitation-contraction coupling related to impaired calcium pumping in end-stage myocardium, but the role this plays in the transition to failure is unclear. Similarly, it has been known for many years that cross-bridge cycling, assessed by myofibrillar ATPase activity, is also depressed in end-stage myocardium. A number of other potential mechanisms underlying the transition to failure have been proposed, including altered adrenergic responsiveness, increased microtubule production, and energy deficits, but once again little is known about their functional significance during the transition to failure.

The present study, performed in "skinned" strips of left ventricular (LV) papillary muscle, was designed to focus on the alterations in cross-bridge kinetics and myofilament viscoelastic properties during the transition from compensated hypertrophy to heart failure in the Dahl salt-sensitive (DS) rat. The DS rat (4) provides a well-characterized model of cardiac hypertrophy and the transition to failure. We found that there were no major abnormalities in cross-bridge kinetics and myofilament viscoelastic properties in compensated hypertrophy, but marked changes in failing hearts. Thus in this model depression of the contractile machinery is associated with the transition to heart failure.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animal model. Procedures were approved by the University of Vermont Institutional Animal Care and Use Committee. At 8 wk of age, male DS rats (Taconic Labs, supported by Merck) were randomly divided into populations that received either high-salt (HS, 8% NaCl) or low-salt diet (LS, 0.3% NaCl) (Harlem Teklad). The diet was started when the animals were 2 wk older than those in our previous study in an effort to reduce the high late mortality we observed (~80%) in HS animals (13). Each group (HS, LS) was separated into subgroups that were euthanized at either 6 or 12 wk after diet initiation. Thus there were four experimental groups divided by diet and duration (in wk): HS-6 (n = 8) or HS-12 (n = 10) wk and LS-6 (n = 8) or LS-12 (n = 12) wk. n is the number of strips, and when more than one strip per animal was tested the mean values were calculated and used.

Echocardiography. After 4, 6, 9, and 12 wk of the diet, the rats were briefly (~5 min) anesthetized with isoflurane (1%) and were lightly secured in the supine position on a warming pad. Echocardiography was performed using a Sequoia system (Acuson; Mountain View, CA) with a 15L8 linear array transducer (15 MHz). Briefly, in the two-dimensional mode, parasternal long- and short-axis views were imaged. This allowed positioning of the M-mode cursor perpendicular to the ventricular septum and LV posterior wall. Subsequently, M-mode images were obtained. Two-dimensional long-axis and short-axis views of the LV were recorded, with the short-axis at the level of the tips of the papillary muscles. After gain settings were optimized, M-mode tracings were recorded at the same level, and posterior wall thickness at end diastole (PWT) was measured. LV internal dimensions [end-diastolic diameter (LVEDd) and end-systolic diameter (LVESd)] were measured using the leading edge-to-leading edge convention. LVEDd measurements were made at the time of maximal diastolic dimension, whereas LVESd was measured at the point of greatest anterior systolic excursion of the posterior wall. Values from all measured beats in each animal were averaged. LV percent fractional shortening (%FS) was calculated as [(LVEDd  LVESd)/LVEDd] × 100. One observer with no knowledge of the animal's study group analyzed the echocardiograhic images.

Dissection and strip preparation. After 6 or 12 wk of the diet, the rats were anesthetized with isoflurane (1%), and the heart was rapidly removed and immediately placed in a standard Krebs solution containing 30 mM 2,3-butanedione monoxime (BDM) and gassed with 95% O₂-5% CO₂. Subsequent procedures have been published in detail previously (2). Briefly, LV papillary muscles were dissected in BDM-Krebs solution to yield thin strips (diameter  0.125 mm; length  1.0 mm) with parallel fiber orientation. The strips were transferred to a vessel containing relaxing solution composed of 5 mmol/l MgATP, 40 mmol/l phosphocreatine, 240 U/ml creatine kinase, 1 mmol/l free Mg²⁺, 0.11 mmol/l CaCl₂, 5 mmol/l EGTA, and 20 mmol/l N,N-bis (2-hydroxyethyl)-2-aminoethanesulfonic acid buffer (pH 7.0), ionic strength 190 mmol/l with added sodium methane sulfonate. The strips were then demembranated (skinned) by the addition of Triton X-100 to a final concentration of 1% wt/vol and 50% glycerol wt/vol (with 10 µg/ml leupeptin, a protease inhibitor), and incubated overnight at 4°C. Strips were then transferred to a vessel with a solution of the same composition except for Triton and stored at 20°C until used within 1 wk.

Sinusoidal length perturbations. Small amplitude sinusoidal length perturbation of the skinned strips (sinusoidal analysis) was used to obtain information about diet-induced alterations in myocardial mechanical properties and cross-bridge kinetics at varying calcium concentrations. Details have been described previously (2, 14). Briefly, after steady-state isometric tension was reached at each calcium concentration, sinusoidal perturbations of 0.125% strip length (amplitude) were applied at 42 discrete frequencies (f, 0.125- 100 Hz). The length and force signals were digitized and normalized to the initial length and cross-sectional area of the muscle strip (Fig. 1A). The complex stiffness modulus obtained from the data provides a measure of the relative amplitudes (dynamic stiffness, kN/m²) and phases of the fractional length and tension sinusoids. The phase shift (_s) corresponds to a time shift (t_s) between the two sinusoids, where t_s = _s/2f. Figure 1B is a representative Nyquist plot of the viscous and elastic moduli at each frequency under activated (pCa 5.0) conditions. Maximal oscillatory work production by the strip corresponds to the complex modulus data point with the lowest viscous modulus in the Nyquist plot. The strip produces maximal oscillatory power at the data point with the maximal product of frequency and work, which occurs at or near the frequency producing the lowest viscous modulus.

View larger version (28K): [in this window] [in a new window]   Fig. 1.   A: representative digitized force (F) and length (L) signals from sinusoidal length perturbation at one frequency. ts is the time shift between the two sinusoids and relates to the phase shift (s) as ts = s/s, where s equals two times the sinusoid frequency (in radians/s). B: representative Nyquist diagram of complex modulus with data plotted from specified frequencies. The vector sum of the viscous (Ev) and elastic (Ee) moduli is the complex modulus. The phase relationship between Ev and Ee is given by . Solid line, curve fit to the data solving for Eq. 1. C: deconvoluted Nyquist diagram with processes A, B, and C. Values are determined by solving the complex modulus equation (Eq. 1, see text for details).

Statistics. Data are reported as means ± SD unless indicated. Two-way analysis of variance (ANOVA) was used to detect differences. A Student-Newman-Keuls test was used for multiple comparisons. A value of P < 0.05 was taken to indicate significance.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Body weight, LV-to-body weight ratio, and mortality. Table 1 shows the effects of diet on body weight (BW), LV-to-BW (LV/BW) ratio, and mortality. Initiation of the HS diet when the animals were 8 wk old rather than 6 wk old resulted in a greatly reduced mortality (20% vs. ~80%) compared with our previous report (13), whereas still causing a marked increase in LV/BW ratio at 6 wk. LV/BW ratio then decreased at 12 wk as the animals progressed toward heart failure. LV/BW ratio was significantly higher in HS-6 and HS-12 compared with age-matched controls. Previous studies from our laboratory as well as others have documented that between the sixth and twelfth week of the HS diet there are significant increases in mortality, wet lung weight, blood urea nitrogen, and creatinine consistent with a transition to heart failure (11, 13). BW of LS-12 rats was slightly larger than HS-12 rats.

Echocardiograhic variables. Echocardiographic assessment was performed after 4, 6, 9, and 12 wk of the diet. LVEDd was significantly larger in the HS-9 group compared with both age-matched LS controls and the previous HS-6 measurement. LVEDd continued to increase such that the HS-12 group had a significantly larger LVEDd than either HS-9 or LS-12 groups (P < 0.01). The LS group maintained a constant LVEDd and LVSDd throughout. LVSDd was significantly greater in the HS-12 compared with that of the HS-9 or the LS-12 group. PWT was increased by HS diet at 6, 9, and 12 wk compared with that of the LS controls. However, the HS-12 group PWT was smaller than that in the HS-9 group, consistent with progressive dilatation without additional hypertrophy. LS-6 %FS was ~40% and remained at about this value during the study. HS-6 %FS was increased compared with LS-6 and, in combination with the increased PWT, indicates compensated hypertrophy. At 9 wk %FS had decreased compared with HS-6 and was similar to the age-matched LS controls. This change in function at 9 wk, along with the increased LV dimension in diastole and systole, suggests that HS-9 hearts are starting to undergo a transition to failure. At the 12 wk time point, %FS in the HS group (32%) was greatly reduced compared with both LS controls and the previous HS time point. In conjunction with major increases in LVEDd and LVSDd (~25%), this indicates LV decompensation.

Isometric tension-calcium relation. There were no significant differences in the absolute relaxed (pCa 8.0) or maximal tension (pCa 5.0) after 6 wk of the diet. The calcium concentrations producing half-maximal relative (percentage of maximal) tension generation (pCa₅₀) were equivalent after 6 wk of the diet (HS 5.28 ± 0.02 vs. LS 5.32 ± 0.02). Figure 2 illustrates the isometric tension-calcium relation in strips from the 12-wk diet groups expressed both as absolute (kN/m²) tension values (Fig. 2A) and as a percentage of maximal tension (Fig. 2B). The regression lines in Fig. 2 represent the data fit to a modified Hill equation. There was a significant increase in pCa₅₀ in the HS-12 group assessed from the pCa-absolute tension relationship (LS 5.14 ± 0.05 vs. HS 5.29 ± 0.03, P < 0.03, Fig. 2A) and from the pCa-relative tension relationship (LS 5.28 ± 0.03 vs. HS 5.39 ± 0.02, P < 0.001, Fig. 2B), indicative of increased calcium sensitivity. The trend toward higher maximal tension in the LS-12 group (Fig. 2A) was not statistically significant.

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Tension-pCa relationship of skinned papillary fibers from Dahl salt-sensitive rats on specified diet for 12 wk. Data are means ± SD. 1 kN/m2 = 1 mN/mm2.

Sinusoidal analysis. In Fig. 3 a Bode plot of the dynamic stiffness-frequency relation from the sinusoidal analysis is illustrated for the 12-wk diet groups. When the muscle is relaxed (pCa 8.0) there is a roughly linear relationship between dynamic stiffness and the frequency of length perturbation (Fig. 3A). With activation (Fig. 3, B and C) there is an increase in the magnitude of dynamic stiffness that reflects an increased number of force-generating cross bridges. In addition, the dynamic stiffness-frequency relation becomes more complex, with relatively low dynamic stiffness at low frequencies that decreases to a minimum as the frequency increases. The nadir of dynamic stiffness is referred to as F_min or the "dip" frequency. At frequencies greater than F_min there is a dramatic increase in stiffness because an increasing number of attached cross bridges are unable to complete the transition to the postforce producing state. There was no difference in the magnitude of minimal dynamic stiffness or F_min in LS-6 and HS-6 groups at pCa 5.0 or 5.5. At pCa 5.0, F_min was significantly lower in the HS-12 group than the LS-12 group (3.30 ± 0.79 vs. 4.98 ± 1.28 Hz, Fig. 3C, P < 0.05). The magnitude of minimal dynamic stiffness was 35% less in the HS-12 group, although this difference did not reach statistical significance (P = 0.084). The difference between LS-12 and HS-12 is accentuated when dynamic stiffness is examined at a submaximal calcium concentration of pCa 5.5 (Fig. 3B), which is closer to normal in vivo systolic calcium levels (6). HS-12, which has greater calcium sensitivity (Fig. 2), displays the characteristic dip frequency observed in maximally activated strips, whereas a clearly defined dip in dynamic stiffness is absent in the plot for LS-12 (Fig. 3B). At pCa 5.5 HS-12 dynamic stiffness was significantly less than LS-12 between 2.6 and 6.0 Hz and is attributed to an increased sensitivity of the cross bridges to length perturbation.

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Dynamic stiffness-frequency relation after 12 wk of diet with varying levels of calcium activation. A: pCa 8.0; B: pCa 5.50; C: pCa 5.0. Lines are fits to complex modulus equation (Eq. 1) using values (see Table 3) for model-defined processes. See RESULTS for description. Data are means ± SE; n = 9 for each group.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Plot of phase-frequency relation after 12 wk of diet at pCa 5.0.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Plot of oscillatory power-frequency relation after 12 wk of diet at pCa 5.0. For ease of presentation the net power produced axis is inverted. , LS; , HS.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

To understand the mechanisms responsible for the transition from compensated cardiac hypertrophy to myocardial failure, it is important to measure functional characteristics during each of these phases. Such information is unavailable from patients, because myocardial tissue is not ordinarily available during the compensated or transition phases. The DS rat model is well suited to address this problem because it undergoes a well-characterized, predictable transition from compensation to failure. Thus 6 wk of the HS diet induces compensated hypertrophy with normal chamber diameter and %FS (Tables 1 and 3).(11, 13) In skinned strips, we found that the isometric tension-calcium relation, dynamic stiffness, and complex modulus model parameters were unchanged at this time (Table 2). After 12 wk of the HS diet, the hearts had undergone a transition to a decompensated state accompanied by LV dilatation and a decrease in %FS (Tables 1 and 3). In HS-12 skinned strips, an increase in calcium sensitivity of isometric tension was observed, with no change in maximal tension (Fig. 2). Sinusoidal analysis indicated that there is a decrease in F_min, a reduced frequency range for oscillatory power output, and a decrease in the apparent rate constant for the work-producing step in the cross-bridge cycle (Table 2). Corresponding to the changes in F_min and power frequency range, the maximum positive and negative phase shifts occur at lower frequencies in HS-12 rats. Thus, whereas there are no significant changes in myofilament viscoelastic properties in the compensated state, the transition to a decompensated state is associated with major viscoelastic alterations involving depression of both cross-bridge cycling and mechanical performance.

We previously reported that compensated hypertrophy in DS rats is associated with a relatively modest increase in the percentage of V3 isomyosin (from 10 to 44%) and a 25% decrease in maximum myofibrillar ATPase (13). With decompensation, V3 isomyosin increases to about 75% and myofibrillar ATPase decreases by 44%. Consequently, it is unlikely that myofilament alterations that occur during the transition to failure are attributable entirely to the isomyosin shift and reduced ATPase, because both are significantly altered in the compensated phase. Furthermore, V3 isomyosin expression per se does not necessarily result in myocardial failure as evidenced by animals rendered hypothyroid who convert to 100% V3 isomyosin (18).

We also previously documented a small change in troponin T isoform distribution and a decrease in troponin T phosphorylation in failing Dahl rats (13). More recently, we reported a decrease in troponin I (TnI) phosphorylation and, using a combination of isolated native thin filaments and skeletal muscle myosin in the in vitro motility assay, a reduction in calcium sensitivity of unloaded velocity in failing DS rats (28). The latter finding indicates that a change in the thin filament contributes to the altered cross-bridge kinetics in the failing state. The decreased phosphorylation of TnI is likely due at least in part to decreased -adrenergic responses. However, this effect may not be a primary cause of the transition to failure, because there is a decreased -adrenergic response independent of -adrenergic receptor density or L-type Ca²⁺ channel function in the compensated heart (3, 12, 21). Thus it may not be possible to ascribe the transition from compensated hypertrophy to failure to a simple alteration in thin filament proteins.

We found striking reductions in the frequency of maximal work and power generation by the myofilaments during the transition to failure. An impairment of force augmentation with increasing contraction frequency and a decrease in the optimal frequency of force production has been demonstrated repeatedly in failing myocardium in vitro and in vivo (5, 20). Because the force-frequency relationship (FFR) is the most important determinant of increased contractile performance during exercise (22), the depression observed in failing myocardium may have major clinical consequences. It has generally been considered that the depressed FFR in failing myocardium results from impairment of both -adrenergic responsiveness (22) and excitation-contraction coupling, the latter related to depressed sarcoplasmic reticulum calcium pumping (8, 26). However, our observation of a decrease in characteristic frequency b and apparent rate constant 2b of the work-producing step in HS-12 strips, as well as the reduced frequency range over which the HS-12 strips produced work and power, suggests that depressed cross-bridge kinetics and alteration of the intrinsic viscoelastic properties of the myofilament may also contribute to FFR depression.

The substantial decrease in F_min after the transition to failure in our study is consistent with similar results in human heart failure studies (9, 24). Whereas F_min reflects the mean cycling rate of the cross bridges (1, 14, 23, 27), it more precisely indicates the frequency where there is maximal resonance between the imposed length change and the frequency-dependent oscillatory force produced by the cross bridges. A reduction in mean cross-bridge cycling rate or rate of maximal work output is consistent with our previous report of decreased myofibrillar ATPase and increased V3 isomyosin (13). It is intriguing that in the compensated state there was no difference in F_min despite a 24% reduction in ATPase in association with an increase in V3 to about 45% of total myosin. F_min may thus be a more sensitive indicator of the complex alterations in contractile proteins and thin filament composition accompanying the transition to myocardial failure. Moreover, changes in ATPase rate may not necessarily result in closely correlated changes in contractile performance. In in vitro motility assay studies (10), pure populations of V1 or V3 myosin produced velocities proportional to their actin-activated ATPase rates. However, mixtures of myosin species resulted in velocities that reflected complex mechanical interactions and did not yield a linear relationship.

We documented an increased calcium sensitivity of isometric tension in HS-12 strips. The increased sensitivity is consistent with previous reports in both human (19, 25, 29) and canine (30) failing myocardium. The observed decrease in characteristic frequency b (apparent rate constant of force production 2b) reported here suggests a mechanism responsible for this result. We postulate that a reduction of 2b leads to an increase in the number of cross bridges in a preforce, weakly bound state that promotes cooperative activation of myosin S1 binding to the thin filament (2, 17). The effect of cross bridges accumulating in this state would be most evident in the submaximal range of calcium activation, where the amplitude of force development depends considerably on the cooperative activation of the thin filament by S1 binding.

In summary, isometric tension generation and sinusoidal analysis of papillary muscle strips from DS rats revealed no differences in the mechanical response at the compensated hypertrophy stage of adaptation but major alterations with the transition to failure. The latter are associated with multiple changes in contractile protein composition and appear to be an important mechanism for the transition to depressed contractile function in this model.