INTRODUCTION

Top Abstract Introduction Methods Results Discussion Appendix References

CHRONIC PRESSURE OVERLOAD leads to cardiac hypertrophy and dysfunction, which is characterized in part by impaired contractility (8, 10, 11, 18), relaxation (8, 10, 11, 18), cytosolic calcium concentration ([Ca²⁺]_c) handling (9, 18, 25, 28), and -adrenergic responsiveness (2, 3, 7, 51). In the failing heart, the -adrenergic receptors, stimulatory G proteins, and adenylate cyclase are reduced in number and activity (7, 15, 33). The progression to heart failure due to chronic hypertension is also accompanied by the decreased number and function of several other cellular regulatory elements such as sarcoplasmic reticulum (SR), SR Ca²⁺-ATPase and its mRNA, mitochondria, myosin isoforms, and cytoskeletal microtubules (15, 25, 29, 35, 39).

Increased cardiac microtubule assembly has been described in acute pressure overload and can result in a functional defect due to increased intracellular mechanical load (21, 26, 45). However, animal models of chronic pressure overload and human studies on failing heart muscle have suggested that microtubule disruption may be a feature of prolonged cardiac stress (5, 10, 35, 39, 47). A loss of microtubules may have deleterious effects on cellular function in general (1), but more specific to -adrenergic responsiveness are studies suggesting that microtubules may be important in stabilizing the stimulatory G protein and thereby augmenting -adrenergic signal transduction in nonmuscle cells (12, 20, 34, 37, 38, 48, 49). Therefore, altered cytoskeletal microtubule assembly may play a role in modulating cardiomyocyte -adrenergic responsiveness and [Ca²⁺]_c dynamics, in addition to their influence on mechanical properties.

This paper describes the influence of microtubule assembly on the -adrenergic responsiveness of [Ca²⁺]_c and shortening/relaxation dynamics of cardiomyocytes isolated from abdominal aortic-constricted rats that had cardiac hypertrophy 30 wk after surgery. We tested the hypotheses that cardiomyocyte microtubule assembly and -adrenergic responsiveness were reduced in this model and that the defect in -adrenergic-mediated systolic and diastolic functions could be reversed with an increase in microtubule assembly. The results of these studies demonstrate that microtubule assembly and -adrenergic responsiveness of isolated cardiomyocytes are reduced in this abdominal aortic constricted rat model and that -adrenergic responsiveness of cardiomyocyte contractile function and [Ca²⁺]_c handling could be partially restored with treatment by taxol, a microtubule-enhancing agent.

	METHODS

Top Abstract Introduction Methods Results Discussion Appendix References

Animal preparation. Male Sprague-Dawley rats were housed in a 12-h light/12-h dark cycle, given standard rat chow and water ad libidum, and randomly assigned to two blood pressure (BP) groups: normotensive (NT, n = 10) and hypertensive (HT, n = 13). Animals in the HT group received a ligature of 0.8-mm diameter around the abdominal aorta between the renal arteries, and the NT animals underwent a sham operation. After 30 wk of recovery, animals were killed for left ventricular (LV) cardiomyocyte isolation. Body weights and kidney weights were recorded for all rats at the time they were killed. Animal care and use were conducted under the guidelines accepted by the American Physiological Society and received prior approval from the Institutional Animal Care and Use Committee at the University of Colorado, Boulder Campus.

Echocardiography. Echocardiography was used to noninvasively evaluate LV size and function in the rat. LV dimensions were measured for NT and HT rat hearts. In preparation for echocardiography, rats were sedated with tribromoethanol (Avertin 160-220 mg/kg ip) and ketamine (45-90 mg/kg ip), each rat's chest was shaved, and the rat was positioned prone on an acoustic gel "standoff" pad with electrocardiogram leads attached to the extremities and tail in a standard fashion. Echocardiography was performed using a Vingmed CFM800 echocardiography machine (Vingmed, Horton, Norway) with a pediatric 7.5-MHz wide-band annular array transducer operating at frequencies between 9 and 11 MHz. Two-dimensional and M-mode images were obtained in the parasternal long- and short-axis orientations. Spectral Doppler flow patterns were obtained across all four intracardiac valves. Two-dimensional and Doppler images were obtained at 48 frames/s with a sampling rate of 200 per second for M-mode and Doppler. The digital data for all images and Doppler were immediately transferred and stored on a Macintosh computer connected to the Vingmed System. Echocardiographic measurements of chamber size and wall thickness in systole and diastole were made off-line from the stored data utilizing EchoPac (Vingmed).

LV cardiomyocyte isolation. LV cardiomyocytes were obtained from the LV septal and free wall using methods previously described in detail (28). All chemicals and reagents were obtained from Sigma (St. Louis, MO) except where noted. In brief, animals were heparinized (250 U ip) and then anesthetized with pentobarbital sodium (35 mg/kg ip) (Abbott, North Chicago, IL). Hearts were rapidly excised and placed in ice-cold saline. The aorta was then cannulated, and the heart was retrogradely perfused using a modified Langendorff perfusion apparatus that delivered three different solutions. The first solution was a bicarbonate-based modified Krebs-Henseleit buffer, the second solution was a nominal Ca²⁺-containing solution, and the third solution contained collagenase (Worthington, Freehold, NJ) and hyaluronidase. All solutions were maintained at pH 7.4 and 37°C and were bubbled with 95% O₂-5% CO₂ gas. The atria and right ventricular free wall were removed, leaving the LV free wall and septum, which were minced and placed in a collagenase and hyaluronidase solution. Cardiomyocyte isolation continued with mechanical agitation. Isolated cardiomyocytes were suspended in bicarbonate-based medium 199, seeded onto laminin-coated 2-cm diameter glass coverslips, as well as three laminin-coated 4-cm diameter plates, and placed in an incubator at 37°C and 5% CO₂.

One coverslip was placed under a microscope, and images of all individual cardiomyocytes were recorded onto videotape. These video images were examined for visual length, width, and area using NIH Image 1.41 video frame-grabbing software. The remaining coverslips and plates were equally divided into three tubulin subgroups. Exposure to 1 µM colchicine produced a microtubule depolymerization subgroup (Col), exposure to 10 µM taxol produced a microtubule hyperpolymerization subgroup (Tax), and an equivalent volume of vehicle was used to produce a control subgroup (Ctrl). The Col and Ctrl subgroups were allowed to incubate for an additional 2 h, and the Tax subgroup incubated for an additional 4 h. Cardiomyocytes seeded onto the large plates were then used for assay of cardiac cardiomyocyte tubulin polymerization by Western blot, and those seeded on coverslips were used for experiments designed to characterize the [Ca²⁺]_c and shortening dynamics of the cardiomyocytes.

Western blot analysis of microtubule fraction. After incubation, medium was removed from the plates, and cardiomyocytes were harvested by brushing them off plates with a rubber policeman and placing them into 1 ml of microtubule stabilization buffer. Stabilizing buffer (32) contained (in mM) 10 sodium phosphate, 0.5 MgCl₂, 0.5 GTP, 0.5 ethylenebis(oxyethylenenitrilo)tetraacetic acid, and 100 U/ml aprotinin (Trasylol), which were added to 50 ml of glycerol and 5 ml dimethyl sulfoxide and then brought to a 100-ml volume with water and pH of 6.95. Protease inhibitors (10 µM benzamidine, 1 mM phenylmethylsulfonyl fluoride, 1 mM o-phenanthroline, 10 mM aprotinin, 10 mM leupeptin, and 10 mM pepstatin A) were added to the stabilization buffer just before homogenization to further prevent artificial depolymerization of microtubules (13). Tissue was then homogenized at room temperature, and the homogenate was centrifuged at 100,000 g for 15 min at 25°C. The supernatant was removed and was considered to hold the free tubulin fraction of the cardiomyocytes.

The remaining pellet containing microtubules was resuspended in 1 ml of depolymerizing buffer (32) and contained 250 mM sucrose, 0.5 mM GTP, 0.5 mM MgCl₂, 100 U/ml Trasylol, and 10 mM sodium phosphate, pH of 6.95, and homogenized at room temperature. The homogenate was then kept on ice for 1 h to depolymerize microtubules and then centrifuged at 100,000 g for 15 min at 4°C. The supernatant contained the fraction of tubulin existing as microtubules. The final pellet was extracted with 10% SDS, 100 mM -mercaptoethanol, in Tris buffer and analyzed by Western blot with anti-tubulin antibodies to ensure all the tubulin had been extracted from the tissue pellet. The protein concentration of each fraction was analyzed by DC protein assay (Bio-Rad, Hercules, CA) to normalize sample loading.

Free tubulin and microtubule fractions were analyzed by electrophoretic separation on 10% SDS-PAGE gel and transferred onto nitrocellulose membranes. Membranes were probed with a mouse-tubulin monoclonal antibody (Amersham, Arlington Heights, IL) and visualized with horseradish peroxidase-conjugated anti-mouse secondary antibody (Sigma) using enhanced chemiluminescence (Amersham). Densitometry was performed to semiquantify protein content in each lane from the integrated absorption of each band. The microtubule fraction was calculated as density of the polymerized lane divided by the sum of densities of the free and polymerized lanes. Figure 1 illustrates typical results of the Western blot analysis of microtubule fractions for the two BP groups and for each of the three tubulin subgroups.

View larger version (41K): [in this window] [in a new window]   Fig. 1.   Cardiomyocyte microtubule content. Western blots demonstrated that microtubule fraction (MT) was reduced and free tubulin fraction (F) increased in hypertensive (HT) group compared with normotensive (NT) group. Agents colchicine (Col) and taxol (Tax), respectively, decreased and increased microtubule assembly in both groups compared with controls (Ctrl).

Experimental protocols. Fura 2-AM (Molecular Probes, Eugene, OR) was introduced into the incubation media bathing the glass coverslips at a concentration of 2 µM. After an additional 5 min in the incubator, each coverslip was removed from the media and used to form the bottom plate of a custom flow-through chamber. The chamber was placed on the stage of an inverted microscope (Nikon Diaphot) fitted with a ×40 oil immersion objective. Superfusion of a Tyrode solution (in mM: 140 NaCl, 6 KCl, 1 MgCl₂, 2 CaCl₂, 10 glucose, 2 pyruvate, and 5 HEPES, pH = 7.4) was maintained at 29°C. Cardiomyocytes were electrically paced via field stimulation using platinum electrodes with a stimulus duration of 0.5 ms, a voltage of 1.5 times their threshold of stimulation, and a pacing frequency of 0.5 Hz (Grass Instruments, Boston, MA).

After a cardiomyocyte was identified for study, pacing was ceased for 2 min to reduce any possible discrepancies due to differential times to identification. Continuous electrical pacing began again, and cardiomyocyte [Ca²⁺]_c and shortening dynamics were recorded at exactly 1 min after the exposure to each of the seven successively increasing concentrations of isoproterenol (Iso): namely, baseline (0 nM), 10 nM, 30 nM, 100 nM, 300 nM, 1 µM, and 10 µM Iso, which were prepared in the Tyrode solution. Maximal -adrenergic response was achieved at 300 nM Iso, and higher doses did not result in further increases in cell function.

Figure 2 illustrates representative fluorescence ratio (R) and cardiomyocyte length transients for the two BP groups at baseline (0 nM Iso). Superimposed on these transients are depictions of the measures used to characterize the [Ca²⁺]_c and shortening/relaxation dynamics.

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Representative examples of cytosolic Ca2+ concentration ([Ca2+]c) and shortening/relaxation transients at baseline. A: [Ca2+]c transients were fit with a double-exponential function (see APPENDIX) that is superimposed here on the original data. Variables describing double-exponential function are depicted. TTPR, time to peak ratio (R); Rpeak, peak R; Rrest, resting R; Rdiff, peak minus resting R; kfall, rate of ratio fall; krise, rate of ratio rise. B: shortening transients were analyzed for resting length, maximal shortening, maximal velocities of shortening and relaxation, and time to peak shortening (TTPS). Other characteristics used for comparisons were maximal shortening as a percentage of resting length, maximal shortening rate (MSR = maximal shortening velocity/maximal shortening), maximal relaxation rate (MRR = maximal relaxation velocity/maximal shortening), and times to 25, 50, 75, and 90% recovery.

Measurements of [Ca²⁺]_c dynamics. Fluorescence of fura 2 was induced with a fluorescence system (IonOptix, Milton, MA) fitted with optical filters of wavelength 1 = 400 nm and wavelength 2 = 360 nm (the isosbestic wavelength). This choice of filters takes advantage of a linear relationship between [Ca²⁺]_c and R when an excitation wavelength over 390 nm is used (42). Fluorescence intensities were recorded as photon counting rates using a personal computer. The value for cardiomyocyte fluorescence background was determined for each cell by superfusion of calcium-free Tyrode + 1 µM digitonin for 4 min, which released cytosolic fura 2, and the subsequent measure of fluorescence with calcium-free Tyrode as superfusate. Background and compartmentalization of fura 2 into organelles or into areas inaccessible to [Ca²⁺]_c were therefore incorporated into the calculation of R between wavelength 1 and wavelength 2 (17). The recorded cardiomyocyte R transients were analyzed to determine the following characteristics: resting R (R_rest), peak R (R_peak), peak minus resting R (R_diff), two exponential rate constants (k_rise and k_fall) determined by nonlinear least-squares fitting of a double-exponential function to the recorded transient, and the time to peak R (TTPR) determined from the exponential rate constants as (ln k_rise  ln k_fall)/(k_rise  k_fall) (see APPENDIX).

	RESULTS

Top Abstract Introduction Methods Results Discussion Appendix References

Animal model. The animal morphology data collectively suggest that stenosis of the abdominal aorta between the renal arteries successfully produced a model of renovascular hypertension, which resulted in cardiac hypertrophy like that described by others (6, 22). Table 1 presents the results of animal organ morphologies and cardiomyocyte dimensions. There was no statistically significant difference for the body mass of the NT and HT groups. Right kidney mass was also not different, but left kidney mass was significantly smaller in the HT group without any visual signs of necrosis, thereby providing one noncardiac-related indicator of a successful abdominal aorta stenosis.

Although LV mass was not measured to demonstrate hypertrophy, there was an ~20% increase in anterior and posterior LV wall thicknesses at diastole in the HT group, and the LV lumen diameter was smaller in the HT group (Table 1). In addition, isolated cardiomyocyte width and area were also greater in the HT group by ~7 and ~11%, respectively, whereas cardiomyocyte length was not found to be different between the groups.

Western blot analysis of microtubule fraction. Western blot analysis of free and microtubule fractions showed that the microtubule fraction was found to be consistently decreased in the HT group compared with the NT group (Fig. 1). Chemical treatment with colchicine or taxol caused either a decrease or increase of microtubules, respectively. Table 1 provides the tabular results of the spot densitometry for the NT and HT groups. The present results in a rat model 30-wk postabdominal aortic constriction suggest that the microtubule assembly after chronic left heart pressure overload in the rat with renovascular hypertension is decreased, which would be consistent with findings from others (35, 36, 39).

[Ca²⁺]_c dynamics. In general, [Ca²⁺]_c dynamics of the HT group were subtly different from the NT group. TTPR was significantly higher and R_rest lower in the HT + Ctrl subgroup relative to the NT + Ctrl subgroup (Fig. 3). More pertinent to this study, however, was the significantly impaired -adrenergic responsiveness of the HT group. The significant BP × Iso interaction for R_rest indicates that the HT + Ctrl subgroup did not appreciably raise resting [Ca²⁺]_c in response to Iso compared with the NT + Ctrl subgroup (Table 2). At 300 nM Iso, all [Ca²⁺]_c transient characteristics (except k_rise) of the HT group tended to be different from those of the NT group, therefore indicating a decreased response of the HT group to Iso. These later differences in R_diff, TTPR, and k_fall are consistent with previous reports of lower peak [Ca²⁺]_c, lower rate of [Ca²⁺]_c rise, and lower rate of SR Ca²⁺ reuptake found with cardiac hypertrophy due to hypertension in general (9, 18, 25, 28) and in response to -adrenergic stimulation (3, 51).

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Response of [Ca2+]c variables to isoproterenol (Iso). A: -adrenergic stimulation increased resting [Ca2+]c in NT group more than in HT group. An increase in microtubule assembly in HT + taxol (Tax) subgroup tended to restore normal -adrenergic response. B: TTPR increased slightly with -adrenergic stimulation in HT group but not in NT group. HT + Tax subgroup showed a response indistinguishable from NT group. C: kfall increased with -adrenergic stimulation but more so in the NT group than in HT group (see Table 2). Microtubule assembly appeared to partially restore kfall of HT + Tax group at high Iso concentrations. * Significantly different from NT + Ctrl (P < 0.05).

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Response of shortening/relaxation variables to Iso. A: maximal shortening increased with -adrenergic stimulation but was not found to have a different -adrenergic responsiveness between blood pressure groups. B: TTPS tended to decrease with -adrenergic stimulation but more so in NT group than in HT group. Microtubule assembly in HT + Tax subgroup tended to recover -adrenergic responsiveness of this variable. C: -adrenergic stimulation increased MRR but more so in NT group than in HT group. State of microtubule assembly was found to be directly proportional to -adrenergic responsiveness of this variable. * Significantly different from NT + Ctrl (P < 0.05).

Relationships between [Ca²⁺]_c and shortening/relaxation dynamics. Our main observation, that depressed cardiomyocyte -adrenergic responsiveness in the HT group was improved by a taxol-induced increase in microtubule assembly, may be best explained through visualizing the effects of -adrenergic stimulation on the relationships between [Ca²⁺]_c and shortening/relaxation dynamics, as illustrated in Fig. 5. Figure 5A graphically depicts the relationship between systolic variables for [Ca²⁺]_c and shortening/relaxation, namely TTPR and TTPS, for the NT group at baseline and 300 nM Iso conditions. All subgroups of the NT groups were found to follow a similar pattern with -adrenergic stimulation, i.e., TTPS decreased without a significant change in TTPR. Figure 5B depicts the same relationship for the HT group, which followed a different pattern, i.e., TTPS decreased and TTPR increased with -adrenergic stimulation. However, the HT + Tax subgroup partially recovered the normal -adrenergic response of no change in TTPR with -adrenergic stimulation. Diastolic variables for [Ca²⁺]_c and shortening/relaxation, namely k_fall and MRR, for the NT group are compared in Fig. 5C. Note that the slopes of the relationships between baseline and 300 nM Iso were similar for all groups and subgroups. However, the NT group and the HT + Tax subgroup moved further along this common relationship in response to Iso compared with the HT + Col and HT + Ctrl subgroups.

View larger version (27K): [in this window] [in a new window]   Fig. 5.   Effect of Iso on relationships between [Ca2+]c and shortening/relaxation characteristics. A: relationships between systolic variables TTPR and TTPS for NT tubulin subgroups respond similarly to Iso. B: relationships between TTPR and TTPS for HT + Col and HT + Ctrl subgroups respond differently to Iso compared with NT, but HT + Tax tended to recover normal Iso response. C: relationships between diastolic variables kfall and MRR for NT tubulin subgroups respond similarly to Iso. D: relationships between kfall and MRR for HT + Col and HT + Ctrl subgroups respond to Iso similar to NT, but maximal response is blunted. HT + Tax subgroup partially recovered normal -adrenergic response.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion Appendix References

This study demonstrated that microtubule assembly and -adrenergic responsiveness of isolated cardiomyocytes were depressed in a rat model of cardiac hypertrophy 30 wk after abdominal aortic constriction. The unique finding of this study was the partial reversal of the depressed cardiomyocyte -adrenergic responsiveness of [Ca²⁺]_c and shortening/relaxation dynamics by a taxol-induced increase in microtubule assembly in this model.

The abdominal aortic constriction rat model produced here developed morphological characteristics consistent with cardiac hypertrophy due to hypertension. A reduced mass of the left kidney, which was distal to aortic stenosis, serves as an important noncardiac indication of surgical success. Although not tested, this finding suggests that the renin-angiotensin system probably played a role in the development of cardiac hypertrophy (6, 22). For this reason, this model may differ from other models of pressure overload using thoracic or pulmonary aortic constrictions (3, 21, 26, 45). We found significant increases in anterior and posterior LV wall thicknesses and a decrease in the LV lumen diameter, which are consistent with cardiac remodeling that would recover normal wall stress in response to increased afterload (8). Cardiac hypertrophy was also reflected in the width and area of the isolated LV cardiomyocytes, thereby indicating that morphological remodeling was pervasive to the cellular level (41, 50). These data collectively indicate that the hearts of these male Sprague-Dawley rats at 30 wk postabdominal aortic constriction were in a state of cardiac hypertrophy due to hypertension. This is in contrast to the Fischer 344 strain that may reach heart failure by 20 wk of comparable cardiac stress (8, 15).

Cardiomyocyte microtubule fraction was reduced in this 30-wk abdominal aortic-constricted rat model. This result is in stark contrast to the increase in microtubules observed with an 8-wk pressure overload left ventricle rat model (21) and with a 2-wk model of feline right heart pressure overload (45). However, the data are consistent with the possibility that microtubule assembly after pressure overload may follow a pattern of acute increase followed by chronic decrease. This specific pattern of microtubule assembly has been reported in the hypertensive rat model by others (35, 36, 39) and is a pattern similar to that seen for other intracellular proteins and mRNA during the progression of chronic hypertension to heart failure (15, 25, 29). The activation of angiotensin II may have further affected microtubule assembly due to activation of mitogen-activated protein kinase (27) and the subsequent destabilizing effect of mitogen-activated protein kinase phosphorylation on microtubules (16). We therefore conclude that our finding of decreased cardiomyocyte microtubule fraction in this 30-wk abdominal aortic-constricted rat model is characteristic of a later stage of cardiac hypertrophy due to renovascular hypertension.

The resting [Ca²⁺]_c of the NT group increased with Iso (Fig. 3A), thereby indicating that the net rate of calcium influx into NT cardiomyocytes became greater than the net rate of calcium efflux after -adrenergic stimulation. -Adrenergic stimulation normally leads to an increase in calcium current by modulation of the L-type channel, either by cAMP-mediated phosphorylation or by direct action of the stimulatory G protein (4, 44), but it is not known to affect efflux mechanisms, namely the Na⁺/Ca²⁺ exchanger and the sarcolemmal Ca²⁺-ATPase. Therefore, calcium efflux by Na⁺/Ca²⁺ exchanger and sarcolemmal Ca²⁺-ATPase did not match the -adrenergic-induced increase in calcium current and constitutes the normal response of cardiomyocytes in our experimental setting. In contrast, the resting [Ca²⁺]_c of the HT group did not increase with Iso as much (Fig. 3A). Because -adrenergic stimulation elicits a blunted increase in calcium current of HT cardiomyocytes (40, 51), and calcium efflux mechanisms have been shown to be upregulated in cardiac hypertrophy (14, 24, 31), we conclude that calcium efflux by Na⁺/Ca²⁺ exchanger and sarcolemmal Ca²⁺-ATPase was better able to maintain intracellular calcium homeostasis in response to the impaired Iso-induced increases in calcium current in the HT group.

There was a slight recovery of a normal -adrenergic response of R_rest in the HT + Tax subgroup at high Iso concentration, therefore suggesting that microtubule assembly partially recovered the normal -adrenergic responsiveness of the L-type calcium current in the HT group. Interestingly, the HT + Col subgroup demonstrated a slight recovery of a normal -adrenergic response of R_rest at low Iso concentrations (Fig. 3A). This and similar observations for the HT + Col subgroup with TTPR and MRR at low Iso concentrations raise the possibility that Col and Tax may differentially influence cardiomyocyte function, including -adrenergic responsiveness, through means independent of their influence on microtubule assembly.

Whereas R_rest is indicative of intracellular calcium homeostasis maintained by the sarcolemma, TTPR is more indicative of [Ca²⁺]_c dynamics during a contraction and is principally influenced by the SR. The TTPR of a [Ca²⁺]_c transient represents the relative balance between the rates of [Ca²⁺]_c influx and removal, i.e., an increase in the TTPR would occur if either rates of [Ca²⁺]_c influx or [Ca²⁺]_c removal were decreased (see APPENDIX). Because TTPR for the NT group was not substantially changed by Iso (Figs. 3B and 5A), we conclude that the normal response of calcium regulation to Iso is to preserve the time to peak of the [Ca²⁺]_c dynamics. Yet k_fall increased and k_rise tended to decrease with Iso. Because k_fall and k_rise are analogous to the respective velocities of [Ca²⁺]_c movement divided by the total [Ca²⁺]_c transferred, -adrenergic stimulation must have increased SR calcium reuptake velocity relative to SR calcium content and increased SR calcium content relative to SR calcium release velocity in such a way as to preserve the time to peak [Ca²⁺]_c in the NT group.

TTPR was generally longer in the HT group compared with the NT group (Figs. 3B and 5B), indicating a relative mismatch between the rates of [Ca²⁺]_c influx and removal typical of diastolic dysfunction. L-type calcium current has not been found to be lower in HT (19, 23, 40); however, other factors influencing [Ca²⁺]_c influx during calcium-induced calcium release, such as density and calcium sensitivity of the SR Ca²⁺ release channels, SR calcium content, and conductance of SR calcium to the cytosol have been implicated as being reduced in the HT state (41, 46). In addition, the rate of [Ca²⁺]_c removal by the SR has been reported to be reduced in the HT state (18, 28, 51). Therefore, the relative mismatch in the rates of [Ca²⁺]_c influx and removal in the HT group, as indicated by the generally longer TTPR, may be due to multiple impaired calcium regulatory mechanisms. Taxol treatment of the HT group partially recovered a normal TTPR, particularly at high Iso concentrations (Figs. 3B and 5B), therefore indicating that either a direct effect of Tax or its influence on microtubule assembly in the HT group enhanced the -adrenergic-mediated changes in calcium regulation, such as increased calcium current and/or increased rate of [Ca²⁺]_c removal by the SR (Fig. 3C).

The rate of [Ca²⁺]_c removal, k_fall, increased with Iso in both groups as would be expected via cAMP-mediated phosphorylation of phospholamban and its subsequent decreased inhibition of the SR Ca²⁺-ATPase (4). However, the effect of maximal Iso on k_fall of the HT group was less than that on the NT group (Table 2), which is consistent with previous findings of decreased -adrenergic responsiveness in HT cardiomyocytes (51). Iso also induced a decrease in k_rise of the HT group similar to that of the NT group. Therefore, our data support the idea that the principal defect in the -adrenergic response in the HT group is a blunted -adrenergic increase in the velocity of SR calcium reuptake.

Cardiomyocyte shortening/relaxation dynamics of the two BP groups were generally differentially affected by -adrenergic stimulation. One obvious exception, however, was the variable of maximal shortening, which increased similarly in both groups with -adrenergic stimulation (Fig. 4A). Although we would have expected group differences in maximal shortening (21, 26, 45, 50) and its -adrenergic responsiveness (51), they were not observed in this 30-wk cardiac hypertrophy model compared with normal. Nevertheless, -adrenergic stimulation did differentially affect the MSR and MRR of the two BP groups, and all other variables were indeed found to be different between the two groups at maximal -adrenergic stimulation (Table 2). Therefore, there were many indicators of differential -adrenergic responsiveness of the shortening/relaxation dynamics.

The variable TTPS represents the relative balance between the rates of mechanical shortening and relaxation, i.e., an increase in the TTPS would occur if either of the rates of shortening or relaxation were decreased. TTPS was significantly decreased in both groups with -adrenergic stimulation (Figs. 4B and 5, A and B), therefore, either or both rates of mechanical shortening and relaxation must have increased, as would be expected. Microtubule assembly in the HT group by Tax partially recovered the TTPS at high Iso concentrations (Figs. 4B and 5B) and, therefore, must have enhanced the -adrenergic-mediated increase in the rates of mechanical shortening and/or relaxation.

We observed that MRR was increased with -adrenergic stimulation, as observed by others (30), but the response was blunted in the HT group (Fig. 4C). This measure of mechanical relaxation was also reflected in other measures of cardiomyocyte relaxation, such as T₂₅, T₅₀, T₇₅, and T₉₀. The tubulin main effect for the MRR at the 300 nM Iso condition suggests that -adrenergic responsiveness of MRR was directly proportional to the state of microtubule assembly, i.e., the response was increased by Tax and decreased by Col. This result may be due to one or both of the following possibilities: 1) microtubule assembly is directly related to -adrenergic responsiveness of this variable in both BP groups, and 2) microtubule assembly directly correlates with relaxation function through mechanical influences. However, microtubules have been implicated to reduce contractile and relaxation function by increasing intracellular stiffness and/or viscosity (26, 43, 45). Because -adrenergic-mediated relaxation improved with increased microtubules in the present study, we conclude that microtubules augmented relaxation by improving -adrenergic responsiveness more so than opposed relaxation through their mechanical properties. This microtubule-induced improvement in -adrenergic-mediated function was observed for all measures of relaxation, including T₂₅, T₅₀, T₇₅, and T₉₀, as well as for the shortening variable TTPS.

We have described above several examples of Tax-induced improvements in -adrenergic responses in the HT group that were not clearly accompanied by similar responses in the NT group. This may be due to the NT group having an intact -adrenergic reserve and microtubule assembly; therefore, any amplification of the -adrenergic responsiveness to be gained by Tax treatment may have been too small to discern under these experimental conditions. In contrast, Col and Tax treatments of the HT group may independently influence one or several impaired sites along the -adrenergic-signaling pathway: -receptor, G protein, adenylate cyclase, cAMP metabolism, protein kinase A, SR Ca²⁺-ATPase, or phospholamban (7, 15, 33). The data of the present report suggest that the specific mechanisms most influenced by microtubule assembly remain poorly understood, and that future investigations should focus on identifying those sites at which microtubules contribute to restoring intrinsic cardiomyocyte function and -adrenergic responsiveness.

In summary and in conclusion, microtubules were reduced in this rat model of chronic cardiac hypertrophy due to renovascular hypertension. The -adrenergic responses of cardiomyocyte functions were generally blunted in this model. More specifically, the -adrenergic-stimulated increase in resting [Ca²⁺]_c, preservation of time to peak [Ca²⁺]_c, increase in SR calcium reuptake rate, decrease in TTPS subjects, and increase in mechanical relaxation rate were less effective in this model than in normal subjects. However, increased microtubule assembly by taxol partially recovered the normal -adrenergic response for resting [Ca²⁺]_c, time to peak [Ca²⁺]_c, TTPS, and mechanical relaxation rate. Microtubule assembly may therefore play a significant role in determining the -adrenergic response of cardiomyocytes in chronic cardiac hypertrophy. Chemical agents that promote microtubule assembly may be able to partially restore -adrenergic responsiveness toward normal.