INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

CARDIAC HYPERTROPHY is an initial adaptive response to several types of cardiovascular stress and can precede the decompensatory phase of heart failure (42). Many different animal models have been developed and studied to determine the responses of cardiomyocytes to chronic insults (see Refs. 3 and 25 for reviews). Cardiac hypertrophy induced by pressure overload in response to aortic constriction in rats for various periods has yielded much information about Ca and myofilament changes (e.g., Refs. 2, 4, 15, 17, 20, 29, 30, 34, 40, 41, 46, 51). Nevertheless, hypertrophic responses of individual cells are still far from clearly understood.

Studies of pressure-overload hypertrophy, the model used here, have concentrated on how and whether Ca homeostasis and cellular function are altered. In the rat pressure-overload model, there is a switch from the fast isoform of myosin heavy chain (-MHC) to the slow isoform (-MHC) (19), but in skinned fibers, no alteration in myofilament Ca sensitivity has been shown (30, 37). However, reports of other changes in Ca homeostasis during the progression of hypertrophy are controversial and conflicting. Most studies have reported that individual contractions in hypertrophy are both reduced in amplitude and prolonged in time (e.g., 35, 52). A slower relaxation in hypertrophy may be due to prolongation of the Ca transient by reduced sarcoplasmic reticulum (SR) Ca uptake (3, 17, 20, 30, 32, 34, 51). There has also been speculation that downregulation of the SR Ca-adenosinetriphosphatase (Ca-ATPase) is responsible for the transition from compensated cardiac hypertrophy to decompensated heart failure. Reduced Na/Ca exchange and sarcolemmal Ca-ATPase activities have been measured in hypertrophy (1, 24), but in hypertrophic rat and failing human heart tissue, expression of the Na/Ca exchanger has been reported to increase (21, 40, 49).

Although there is general agreement on some findings, a wide array of models have been used to study various different aspects of hypertrophy and heart failure. In some cases, authors have examined changes in a rather limited number of specific biochemical markers, without much data from intact cells. Using a well-developed rat abdominal aortic banding model (18, 19, 43), we here combine expression studies in homogenates with functional studies in isolated ventricular myocytes, in an attempt to characterize cellular Ca regulation in cardiac hypertrophy more comprehensively.

To examine how chronic pressure overload-induced hypertrophy alters contractility, Ca fluxes during relaxation, and excitation-contraction (E-C) coupling in isolated cardiac myocytes, we recorded contractions as well as cellular Ca transients using indo 1 fluorescence. In field-stimulated myocytes, we assessed the contributions of the four mechanisms responsible for removing Ca from the cytosol during cytosolic free Ca concentration ([Ca]_i) decline and relaxation (i.e., the SR Ca-ATPase, Na/Ca exchange, mitochondrial Ca uniporter, and sarcolemmal Ca-ATPase; Refs. 5 and 8). In myocytes under voltage clamp, we examined the relationship between Ca current (I_Ca) and contraction while SR Ca content (assessed by caffeine-induced contractures) was controlled. In cells from banded rats we found little change in the competition among Ca transport systems during relaxation (despite reduced levels of SR Ca-ATPase activity in homogenates). We also found reduced cell shortening for a comparable Ca transient, no change in SR Ca content, and a reduced efficacy of E-C coupling, which was only apparent at 0.5 mM extracellular Ca concentration ([Ca]_o). Some of this work has appeared in abstract form (39).

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Induction of pressure-overload hypertrophy. The banding procedure was as described previously (19). Sprague-Dawley rats (175- to 200-g male; Harlan Sprague Dawley, Indianapolis, IN) were anesthetized by ketamine (60-90 mg/kg im) and xylazine (1-2 mg/kg im). The aorta was dissected and a suprarenal abdominal constriction was applied (closure equivalent to 25-gauge needle) using a hemoclip. Sham-operated animals underwent the same surgical procedure without hemoclip placement.

Myocyte isolation. After the hemodynamic measurements, myocytes were isolated from the left ventricular free wall as previously described (e.g., Refs. 8 and 18). The heart was excised, mounted on a Langendorff perfusion apparatus, and perfused with nominally Ca-free, N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES)-buffered Tyrode solution containing 1.5-2 mg/ml collagenase (type II, Worthington) until it started to become flaccid (~10 min), after which the left ventricular free wall tissue was separated, transferred to a flask containing fresh enzyme, and incubated for a further 10-20 min. Free cells were separated, and the remaining undissociated tissue was reincubated (2 or 3 times if necessary) until most of the tissue had been dissociated. The resultant cell suspension was rinsed several times, with [Ca]_o gradually increased to 1 mM. Myocytes were then plated onto laminin-pretreated glass-bottomed Plexiglas superfusion chambers.

Measurement of Ca transients and cell shortening. In field stimulation experiments, myocytes were loaded with membrane-permeant indo 1-acetoxymethyl ester by a 20-min incubation at 22°C (concentration of indo 1 = 10 µM) followed by washing for 30 min to allow deesterification. Cells from both banded and sham animals were loaded similarly using the same technique, and this degree of loading does not affect contraction parameters significantly (5). Ca-dependent fluorescence was recorded using a microscope-based fluorescence system (Photon Technology International, Monmouth Junction, NJ). Fluorescence emitted at 405 (F₄₀₅) and 485 nm (F₄₈₅) was recorded from a field restricted to one cell (excitation at 365 nm). Autofluorescent backgrounds, measured on comparably sized cells from the same heart, were subtracted from each signal before the ratio F₄₀₅/F₄₈₅ was obtained. This ratio (R) was converted to [Ca]_i, using minimum and maximum R values and the apparent dissociation constant (K_d) which was determined in separate calibration runs (6). System background fluorescence was negligible.

Field stimulation solutions and protocols. Myocytes were continuously superfused with Tyrode solution at 22°C and a flow rate of 2-5 ml/min. The basic (normal) Tyrode solution (NT) contained (in mM) 140 NaCl, 10 glucose, 5 HEPES, 6 KCl, 1 MgCl₂, and 1 CaCl₂, adjusted to pH 7.4 at 22°C with NaOH. Steady-state twitch Ca transients were evoked by field stimulation at 0.5 Hz with platinum electrodes.

Analysis of contributions of Ca removal systems. During twitch relaxation, all of the transport systems removing Ca from the cytosol operate simultaneously and are interdependent because of their common influence and dependence on [Ca]_i. We analyzed separately the [Ca]_i dependence of Ca transport by these systems to further evaluate their relative contributions to Ca removal during twitch relaxation as described by Bassani et al. (5). This analysis uses Ca transient amplitudes ([Ca]_i), diastolic [Ca]_i, and time constants () of [Ca]_i decline during twitches, as well as during caffeine-induced contractures in both NT and 0 Na-0 Ca. The analysis was performed for each group (sham or banded) using the mean values of these parameters for that group (see Table 3). The free [Ca]_i decline during the caffeine-induced contracture in 0 Na-0 Ca was first converted to total Ca using known passive cytosolic and indo 1 binding constants from rabbit ventricular myocytes (26) and then differentiated with respect to time. The dependence of this d[Ca]_total/dt on [Ca]_i was fit to a Hill equation, V_max/(1 + K_d/[Ca]_i)^n_H, where d[Ca]_total/dt is the first derivative of total Ca concentration with respect to time, V_max is the maximum Ca transport velocity, K_d is apparent affinity, and n_H is the Hill coefficient. This fit describes empirically how the lumped slow Ca removal systems (mitochondrial Ca uniport and sarcolemmal Ca-ATPase) depend on [Ca]_i. To determine how Na/Ca exchange flux depends on [Ca]_i, we similarly fit the caffeine-NT transient data to a two-term Hill equation (as this transient is governed by both the slow removal processes and Na/Ca exchange), with one term constrained to use the parameters just estimated from the caffeine-induced contracture in 0 Na-0 Ca. Next, the twitch d[Ca]_total/dt was fit to a three-term Hill equation, with the first two terms being constrained by the slow process and Na/Ca exchange removal models as determined above. This fit predicts how SR reuptake flux depends on [Ca]_i. Finally, using the mean twitch [Ca]_i parameters (diastolic [Ca]_i, [Ca]_i, and  of [Ca]_i decline) as the driving function, we calculated the flux due to each removal process with all operating simultaneously. The resulting fluxes were then integrated numerically to get cumulative Ca transport. This provides information about how these systems interact independently but simultaneously with [Ca]_i (5).

Determination of I_Ca-contraction relationship. In these experiments cell shortening and I_Ca were measured simultaneously during test pulses to different membrane potentials (E_m). To ensure comparable SR Ca loading, the test pulses were preceded by a conditioning train (38). I_Ca was measured in whole cell voltage clamp, using an Axopatch 1B patch-clamp amplifier (Axon Instruments, Burlingame, CA) and patch electrodes with 1- to 3-M resistances (glass type 1B150F-6, World Precision Instruments, Sarasota, FL) containing (in mM) 140 CsCl, 5 MgATP, 5 HEPES, and 50 µM EGTA, pH 7.1 with CsOH at 22°C. Myocytes were superfused with NT containing either 0.5 or 1 mM [Ca]_o, and E_m was held at 90 mV. Five to eight conditioning pulses to 0 mV (400 ms, 0.5 Hz) were applied. The test pulse followed 2 s after the conditioning train and just after a 400-ms prepulse to 40 mV to inactivate Na current (I_Na). Test pulses were 300-ms depolarizations to E_m between 40 and +30 mV (in 10-mV increments). I_Ca magnitude was measured as the difference between the peak and final current during this step. For measurement of the steady-state SR Ca load, additional conditioning trains were given where the test pulse was replaced by rapid application of 10 mM caffeine [at holding potential (E_h) of 90 mV]. The conditioned current-voltage relations and caffeine-induced contracture were measured first at 0.5 mM and then at 1 mM [Ca]_o. Data from cells showing significant rundown were discarded.

Homogenization and Ca-ATPase assays. Ventricular tissue was put into 5 ml cold KCl buffer (see below) with protease inhibitors (75 nM aprotinin and 1 µM leupeptin) and homogenized by three 30-s bursts at 67% maximum setting with a Polytron (Brinkmann Instruments, Westbury, NY). Aliquots of this homogenate were used for protein, ryanodine binding, Na/Ca exchange, and Ca-ATPase assays. Protein concentration was measured by the Lowry method. The KCl buffer contained 140 mM KCl and 10 mM HEPES (pH 7.2). The Ca-ATPase inhibitor thapsigargin (Calbiochem-Novabiochem; Ref. 29) was used to provide a specific assay of Ca-ATPase activity (37°C, pH 6.9). The reaction was initiated by the addition of 3 mM NaATP to buffer containing 3 mM oxalate, 1.87 mM MgCl₂, 5 mM EGTA, and 5.14 mM CaCl₂, which contained ~20 µM free [Ca] and 0.1 mM free [Mg]. Ouabain and azide were added separately, to give final concentrations of 1 and 10 mM, respectively. Either 5 µM thapsigargin in dimethyl sulfoxide (DMSO) (<2%) or DMSO alone (control) were added to the protein sample immediately before its addition to the reaction mixture. After a 3-min incubation the reaction was stopped by the addition of acid molybdate. Malachite green (293 µg/ml final) was added, and phosphate produced was measured by absorbance at 650 nm using a phosphate standard curve.

Na/Ca exchange. Na/Ca exchange activity was assessed in ventricular homogenates using a modification of the method of Reeves and Sutko (45). Homogenate aliquots were diluted 1:2 into a buffer solution (140 mM NaCl, 10 mM HEPES, pH 7.2) to load the vesicles passively with Na (final concentrations 46.7 mM KCl, 93.3 mM NaCl) and then treated with either 20 µM digitonin (to selectively permeabilize the sarcolemma and prevent Na/Ca exchange) or 0.2% ethanol (vehicle). The Na-loaded homogenate (3 µl) was added to 97 µl of a solution containing 25 µM ⁴⁵CaCl₂, 10 mM HEPES (pH 7.2), and either 140 mM KCl (Na free) or 46.7 mM KCl-93.3 mM NaCl (control incubate without Na gradient). Na-dependent Ca uptake via Na/Ca exchange should only occur in Na-loaded vesicles diluted into Na-free solution without digitonin. After 10 s, Ca uptake was stopped by addition of 3 ml of an ice-cold solution containing 1 mM EGTA, 200 mM KCl and 3-(N-morpholino)propanesulfonic acid · Tris (pH 7.4). Membranes were collected by vacuum filtration onto glass-fiber filters (Whatman, GF/C), with the ⁴⁵Ca content determined by liquid-scintillation spectroscopy.

Ryanodine binding assay. Ryanodine binding was measured as described by Bers and Stiffel (12). Homogenate (1-2 mg/ml) was incubated in 1 M NaCl, 20 mM HEPES, 25 mM Tris, 5 mM AMP, 0.5 mM CaCl₂, and 1-100 nM [³H]ryanodine (New England Nuclear) at pH 7.4 for 90 min at 37°C. Unlabeled ryanodine (17 µM) was used to displace [³H]ryanodine, allowing measurement of specific binding. The reaction was terminated by vacuum filtration through Whatman GF/B filters using a Brandel cell harvester. Filters were washed three times with 3 ml distilled water to eliminate excess [³H]ryanodine. Membrane-bound [³H]ryanodine on the filter was estimated by -scintillation spectroscopy. The data on Scatchard plots were fit by linear regression.

RNA isolation and Northern blotting. Total RNA was extracted from left ventricular tissue using the guanidinium thiocyanate method (16). Fifteen micrograms of total RNA were used for Northern blotting analysis as described by Qi et al. (43). The following cDNA probes were used for Northern blot analysis: 1) SR Ca-ATPase [2.3-kilobase (kb) cDNA fragment of the rat cardiac SERCA2, kindly provided by Dr. Wolfgang Dillmann, University of California, San Diego]; 2) Na/Ca exchanger (1.5-kb cDNA of guinea pig cardiac Na/Ca exchanger from Dr. Kenneth Philipson, University of California, Los Angeles); 3) phospholamban (710-bp cDNA fragment of the rat cardiac phospholamban provided by Dr. Huaping He, University of California, San Diego); 4) ryanodine receptor (580-bp cDNA of rabbit cardiac ryanodine receptor from Dr. Andrew Marks, Mount Sinai Medical Center, New York, NY); 5) calsequestrin (1.4-kb cDNA of canine cardiac calsequestrin from Dr. Larry Jones, Indiana University, Indianapolis, IN); 6) atrial natriuretic factor (ANF; 0.8-kb cDNA of rat ANF from Dr. Tadashi Inagami, Vanderbilt University Medical Center, Nashville, TN); and 7) glyceraldehyde 3-phosphate dehydrogenase (GAPDH; 750-bp cDNA from human fetal liver obtained from American Type Culture Collection). The cDNA probes were radiolabeled with [³²P]dCTP and hybridized as previously described (19). The amounts of Ca transporter mRNAs and ANF mRNA were quantified by autoradiography at 80°C and laser densitometry and are expressed relative to the amounts of GAPDH mRNA.

Reagents. Unless otherwise stated experimental reagents used were of analytical grade and were supplied by Sigma (St. Louis, MO).

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Hemodynamic parameters. Table 1 shows the hemodynamic data obtained from the animals used in this study. Aortic banding for 16 wk resulted in significant cardiac hypertrophy, increased systemic blood pressures, and elevated left ventricular end-diastolic pressure (LVEDP), as we have reported previously with this hypertrophic model (18, 43). Heart weight or heart weight-to-body weight ratio increased by ~25%. This probably underestimates the degree of left ventricular hypertrophy. The left ventricle could not be weighed separately in most hearts because the cell isolation procedure precluded removal of atria or the right ventricle before weighing. In parallel experiments where we measured both heart weight and left ventricular weight, this 25% cardiac hypertrophy corresponds to ~43% left ventricular hypertrophy (43). The elevated arterial pressures indicate that this is a hypertensive hypertrophy model.

SR Ca-ATPase, Na/Ca exchange, ryanodine binding, and mRNA in homogenates. Previous studies in this same 16-wk banding model have shown that the SR Ca-ATPase is downregulated by 76% at the mRNA level and 34% at the protein level; further, thapsigargin-sensitive, oxalate- and ATP-supported ⁴⁵Ca uptake is reduced by 27-50% (41, 43). Consistent with these results, we find here that the thapsigargin-sensitive Ca-stimulated ATPase activity in ventricular homogenates was decreased by 34%, comparable to reductions in mRNA (see Table 2). However, despite a depression of Na/Ca exchange mRNA in the banded group, we could not detect a difference in sarcolemmal Na-dependent ⁴⁵Ca uptake activity in homogenate (Table 2). Because a major focus in the present study is on E-C coupling, we also measured the number of ryanodine receptors using radioligand binding. As shown in Fig. 1, there was no significant difference in ryanodine binding between sham and banded animals. The Scatchard plot (inset) shows that maximal binding (B_max) was ~400 fmol/mg in both and there was no apparent difference in affinity under these assay conditions (K_d = 3.3 nM). Figure 2 shows sample Northern blots for expression levels of mRNAs for several proteins. Quantitative analysis showed no significant change in ryanodine receptor or calsequestrin message but a reduction in phospholamban mRNA that roughly paralleled the decrease in SR Ca-ATPase activity and mRNA levels (Table 2). ANF mRNA was greatly increased in the banded group, excluding the possibility that there was simply a generalized reduction in mRNA levels compared with GAPDH.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Specific ryanodine binding in ventricular homogenates from banded and sham rats. Membrane-bound [3H]ryanodine was quantified as a function of free ryanodine concentration ([ryanodine]). There was no significant difference between sham () and banded groups () at any [ryanodine]. Data were fit for a single class of binding sites by Scatchard plot (inset). Maximal binding (Bmax) was 377 fmol/mg in sham and 408 fmol/mg in banded, and apparent affinity [dissociation constant (Kd)] was 3.3 nM in both sham and banded groups (r2 = 0.95 for both linear regressions). Bmax values determined from individual experiments were more variable but not different between groups (sham 416 ± 97 fmol/mg; banded 417 ± 109 fmol/mg).

View larger version (64K): [in this window] [in a new window]   Fig. 2.   Northern blot analysis of mRNA levels for Ca transport proteins, atrial natriuretic factor (ANF), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH). Fifteen micrograms of total RNA were loaded onto each lane, S1-S3 are samples from 3 sham rats, and B1-B3 are from 3 banded rats. Quantified data from this type of experiment are shown in Table 2. SERCA2, sarcoplasmic reticulum (SR) Ca-ATPase; Na/Ca X, Na/Ca exchange; RyR, ryanodine receptor; CSQ, calsequestrin; PLB, phospholamban.

Contractions and Ca transients in field-stimulated myocytes. Isolated ventricular myocytes in this aortic banding model are hypertrophied compared with the sham myocytes (in length by ~10%, width by 5%, and volume by ~28%; Ref. 18). Figure 3 shows, for a typical myocyte from a banded animal, Ca transients and contractions associated with steady-state twitches followed by caffeine-induced contractures in either NT or 0 Na-0 Ca solution. The three types of contractions shown provide sufficient data to evaluate Ca transport via the SR Ca-ATPase, Na/Ca exchange, and slow systems (mitochondrial Ca uniport and sarcolemmal Ca-ATPase) in ventricular myocytes (5). The amplitudes and time constants () of decline of the corresponding types of contractions and Ca transients were pooled across cells, and statistical comparisons appear in Table 3.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Twitch- and caffeine-induced contractions and Ca transients in a representative cell (banded animal). Twitches were at a steady state at 0.5-Hz stimulation, at which the Ca load in the SR was relatively constant. Stimulation was interrupted, and after a 4-s delay, caffeine (10 mM) was rapidly applied. Records with caffeine in normal Tyrode (NT, containing Na) solution and in Na- and Ca-free (0 Na-0 Ca) solution are superimposed. Top: fluorescence-based Ca transients. Bottom: contraction, recorded simultaneously as % resting cell length (RCL). [Ca]i, cytosolic free Ca concentration.

Ca flux analysis. As we noted above, parameters of twitches and caffeine contractures in NT reflect the operation of multiple Ca removal processes, which simultaneously influence [Ca]_i, but the contributions of the processes can be measured separately (5). Measuring the relative strengths of Ca removal processes within a group (banded or sham) overcomes the limitation of comparing  values for Ca transients of different amplitudes between the groups. To make these comparisons we have to take into account the buffering capacities of sham and banded cells. The assumption we have made is that there are not major differences in passive cytosolic Ca buffering between these two groups. Evidence from a previous study using this model (18) and our results regarding E-C coupling at 1 mM [Ca]_o (see below) suggest that this is a reasonable assumption.

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Integrated Ca flux from cytosol by processes contributing to [Ca]i decline during a normal twitch. Note that Ca transport is on a logarithmic scale. This analysis is based on Ca transients measured during twitches as well as caffeine-induced contracture in NT and 0 Na-0 Ca in myocytes from sham animals. Nos. in parentheses indicate percentages of total Ca flux carried by each system in sham group. Values in banded group were similar to sham values: 97.8% via SR Ca-ATPase (SR), 2% via Na/Ca exchange (Na/Ca X), and <1% via slow systems (Slow; mitochondrial Ca uniport plus sarcolemmal Ca-ATPase).

Maximal SR Ca load. Stimulating at high frequency could increase SR Ca loading toward a maximum and thereby potentiate twitch amplitudes (7). This potentiation could be greater in cells from sham rats, if functional expression of SR Ca-ATPase was reduced in hypertrophy. To test this, we paced cells at 0.5 Hz for 10 s, followed by 5 Hz for 10 s. As soon as diastolic [Ca]_i fell to a stable value, a single twitch was evoked. Figure 5 shows records of Ca transients and contractions elicited in response to this protocol in a representative cell from a banded animal. Cells used in this protocol are a subpopulation of those used in the steady-state twitch and caffeine-induced contracture protocols described above.

View larger version (25K): [in this window] [in a new window]   Fig. 5.   Transient increase in stimulation frequency to 5 Hz in a representative cell from a banded animal. Steady-state twitches were evoked at 0.5 Hz (tick marks at left, top x-axis). Stimulus rate was abruptly increased to 5 Hz (densely spaced tick marks, top x-axis) to load the SR heavily. At ~19 s, stimulus was stopped and a single pulse was given after [Ca]i had declined (at ~21 s).

E-C coupling in voltage-clamped cells. To study E-C coupling in a controlled manner, we voltage-clamped cells from each group so that contractile amplitude could be expressed as a function of I_Ca. The mean membrane capacitance values were significantly higher in banded (354 ± 78 pF, n = 17) than in sham cells (209 ± 23 pF, n = 16; P < 0.05), consistent with cellular hypertrophy in the banded group. On the other hand, the membrane charging time constants (2.2 ± 0.3 ms in sham vs. 2.4 ± 0.2 ms in banded) were not significantly different. Although the speed of voltage clamp is less than ideal in these large myocytes, this constitutes only a systematic limitation that was not different between the groups of cells. Thus comparing peak I_Ca values here is reasonable and the results also agree with our prior study (18) that focused more on I_Ca characteristics.

View larger version (26K): [in this window] [in a new window]   Fig. 6.   Contraction (left) and calcium current (ICa; right) at extracellular Ca concentration ([Ca]o) of 1.0 (A) or 0.5 mM (B). Top: voltage-clamp protocol. Conditioning depolarizations from membrane potential (Em) of 90 to 0 mV were followed by depolarization to 40 mV to inactivate sodium current (INa) and then stepped to test potential (in case shown at 0 mV). Left: conditioning and test contractions in a sham and a banded cell. Right: expanded view of test step and ICa traces. Contraction and ICa were recorded simultaneously during test pulses (indicated by * on contraction traces). Data were obtained from sham and banded cells of comparable size (RCLs: sham, 192 µm and 152 µm for 1 mM and 0.5 mM [Ca]o, respectively; banded, 170 µm and 182 µM for 1 mM and 0.5 mM [Ca]o, respectively).

View larger version (27K): [in this window] [in a new window]   Fig. 7.   Voltage dependencies of test contraction and ICa amplitudes at 1 (left) and 0.5 mM [Ca]o (right) in cells from sham and banded animals. ICa of each cell was normalized to cell capacitance (in pA/pF; plotted on downward ordinate), and contraction was normalized to individual RCL (plotted on upward ordinate). Voltage-clamp protocol was as shown in Fig. 6. Data are means ± SE; n = 19 cells (1 mM [Ca]o) and 15 cells (0.5 mM [Ca]o). RCLs were 204 ± 8 µm (banded) and 146 ± 7 µm (sham). Absolute amplitude of last conditioning contraction was smaller at 0.5 mM [Ca]o than at 1 mM [Ca]o, respectively, for both sham (23.1 ± 1.2 µm vs. 27.6 ± 0.5 µm) and banded groups (15.1 ± 0.8 µm vs. 24.4 ± 1.0 µm).

SR Ca content and E-C coupling. An important aspect of the interpretation of the E-C coupling results in Figs. 6 and 7 is the state of SR Ca loading when I_Ca and contraction are measured. Caffeine-induced contractures and Ca transients in Fig. 3 and Table 3 suggested that the SR Ca content of cells from sham and banded groups in steady-state field stimulation was the same, as we have also reported previously for similar voltage-clamp pulses in this model (18). In the same cells undergoing E-C coupling protocols summarized in Fig. 7, caffeine was also applied after the same conditioning trains to assess SR Ca content available for release during the test twitches. Figure 8, left, shows conditioning trains and subsequent caffeine-induced contractures in example cells from sham and banded groups. Figure 8, right, shows pooled data for caffeine-induced contracture in both groups and at both [Ca]_o levels.

View larger version (22K): [in this window] [in a new window]   Fig. 8.   Assessment of SR Ca load after conditioning voltage-clamp trains. Left top: voltage-clamp protocol. Cell was depolarized from holding potential (Eh) of 90 mV to 0 mV at 0.5 Hz for 8 pulses of 400-ms duration to achieve steady-state SR Ca loading. Caffeine was applied 2 s after start of final voltage-clamp pulse to measure contraction in %RCL. Also shown are original records of twitches and caffeine-induced contractures (CafC) elicited using this protocol in representative cells from sham (left middle) and banded (left bottom) animals (1 mM [Ca]o). Right: mean CafC amplitudes of cells from sham and banded animals at [Ca]o = 0.5 and 1.0 mM. Data are means ± SE; n = 6 animals/group. CafC in %RCL were not significantly different in sham vs. banded animals (sham 22.7 ± 3.5% vs. banded 17.4 ± 3.9% at 1 mM [Ca]o; sham 15.9 ± 1.5% vs. banded 13.7 ± 2.9 at 0.5 mM [Ca]o).

Relationship between I_Ca and contraction. The way contraction depends on I_Ca (at comparable SR Ca load) is at the core of E-C coupling. In Fig. 9 the data of Fig. 7 are replotted using I_Ca rather than E_m as the independent variable. The direction of the arrows indicates increasingly positive test voltage steps. The steepness of the relationship indicates the gain or efficacy of E-C coupling, i.e., the relationship between a given trigger I_Ca and SR Ca release (as measured by contraction in this study) for a given SR Ca load. At 1 mM [Ca]_o the sham and banded groups follow similar trajectories, with the banded falling just lower than the sham. This small reduction is consistent with the modest reduction of contraction for a comparable [Ca]_i in the banded rats.

View larger version (28K): [in this window] [in a new window]   Fig. 9.   Efficacy of excitation-contraction (E-C) coupling in myocytes from sham () and banded rats () at 1 (left) and 0.5 mM [Ca]o (right). Data from Fig. 7 are plotted to show dependence of peak contraction on ICa amplitude at test pulses. Arrows indicate direction of increasingly positive Em during test pulses. Maximal ICa and contraction occurs at the 0- or +10-mV test pulse. E-C coupling efficacy was reduced in cells from banded animals at 0.5 mM [Ca]o.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Hypertrophic model. Many animal models have been employed to study hypertrophy, hypertension, and heart failure (3, 25). We do not intend to review the merits of different models here. The rat suprarenal aortic constriction model of left ventricular hypertrophy has been well characterized by us (18, 19, 43) and others (e.g., 2, 4, 15, 17, 29, 30, 34, 40, 41, 46). Although models using ascending aortic constriction have also produced valuable results (e.g., 20, 51), these models differ functionally (e.g., with respect to systemic hypertension and cardiac load). The duration and severity of constriction varies among the above 10 studies of abdominal aortic constriction, but the degree of left ventricular hypertrophy (23-62%, mean 42%) is similar to what we find in this model (43%). Here we study relatively long-term (~16 wk) effects and changes in cellular Ca transport and E-C coupling.

Decreased mechanical response. During caffeine-induced contracture in isolated ventricular myocytes the extent of cell shortening was reduced in cells from the banded group despite unchanged Ca transient amplitudes. This suggests a reduced contractile response to Ca in these cells. Although there is clearly a shift in MHC isoform (-MHC increases from 25 to 59%; Ref. 43), skinned fibers from rats made hypertrophic by aortic banding pressure overload have not shown any difference in myofilament Ca sensitivity (30, 37). Other factors, such as increased stiffness, restoring force, or changes in microtubules, could depress unloaded shortening in isolated myocytes (50). Notably, these effects could have much less effect on isometric twitch force or steady-state myofilament Ca sensitivity. Although there is increased collagen in the banded rat hearts (19, 43), this seems like an unlikely explanation for the cellular results here because the myocytes are isolated with extensive collagenase treatment. The caffeine-induced contracture produces a relatively long Ca transient duration, removing most kinetic concerns that complicate myofilament Ca sensitivity conclusions drawn from twitches. If the caffeine-induced increase in myofilament Ca sensitivity was smaller in the banded group, that might explain the reduced caffeine-induced contracture amplitudes. However, the opposite result was reported in ferret pressure overload (9). In addition, even without caffeine, the twitch contractions here were significantly reduced, despite comparable [Ca]_i. Thus there was a significant reduction in cell contraction for a given [Ca]_i. Although our focus here is mainly on Ca transport, this is a potentially important mechanical alteration exhibited at the cellular level.

Ca removal fluxes. Increased Na/Ca exchange has recently been reported in human heart failure and was suggested to be compensatory for the reduced levels of SR Ca-ATPase (21, 49). However, in banded vs. sham rats we did not find any difference in Na/Ca exchange activity either in homogenates or functionally in intact cells (on the basis of the  of [Ca]_i decline or relaxation of caffeine-induced contracture). There was also no difference in the rate of [Ca]_i decline attributed to the combined action of the mitochondrial Ca uniporter and the sarcolemmal Ca-ATPase. These three systems only make a very minor contribution to Ca removal during twitch relaxation in these experiments in adult rats, whereas the SR Ca-ATPase is responsible for 96-98% of Ca removal flux.

E-C coupling. We found no difference in I_Ca characteristics between sham and banded groups in the present study. This is consistent with previous, more detailed studies of I_Ca in this rat hypertrophy model (18, 46) and with the lack of alteration in dihydropyridine receptor density (44). In addition, we found no difference in the number of ryanodine receptors (Fig. 1). Whereas Kim et al. (29) found a reduction in ryanodine binding in a comparable rat model of hypertrophy, our data agree with those of Rannou et al. (44), who found no difference in ryanodine binding in moderate hypertrophy in rat (~40%, comparable to the present study). However, in rat when hypertrophy was more severe (60%), Rannou et al. (44) found a decreased number of ryanodine receptors. Furthermore, they found that reduced ryanodine binding occurred at more moderate levels of hypertrophy in guinea pig or ferret heart, consistent with heart failure results in dogs and humans (13, 53). Thus the amount of ryanodine receptor downregulation may depend on the species and severity of hypertrophy or failure.