Author + information
- Received June 23, 1998
- Revision received August 6, 1998
- Accepted September 4, 1998
- Published online January 1, 1999.
- Michael R Zile, MD, FACC∗,* (, )
- Masaaki Koide, MD∗,
- Hiroshi Sato, MD∗,
- Yoshiki Ishiguro, MD‡,
- Chester H Conrad, MD, PhD†,
- J.Michael Buckley, MS∗,
- James P Morgan, MD, PhD, FACC‡ and
- George Cooper IV, MD∗
- ↵*Address for correspondence: Dr. Michael R. Zile, Division of Cardiology, Medical University of South Carolina, 171 Ashley Avenue, Charleston, South Carolina 29425-5799
Objectives. We sought to determine whether the ameliorative effects of microtubule depolymerization on cellular contractile dysfunction in pressure overload cardiac hypertrophy apply at the tissue level.
Background. A selective and persistent increase in microtubule density causes decreased contractile function of cardiocytes from cats with hypertrophy produced by chronic right ventricular (RV) pressure overloading. Microtubule depolymerization by colchicine normalizes contractility in these isolated cardiocytes. However, whether these changes in cellular function might contribute to changes in function at the more highly integrated and complex cardiac tissue level was unknown.
Methods. Accordingly, RV papillary muscles were isolated from 25 cats with RV pressure overload hypertrophy induced by pulmonary artery banding (PAB) for 4 weeks and 25 control cats. Contractile state was measured using physiologically sequenced contractions before and 90 min after treatment with 10−5mol/liter colchicine.
Results. The PAB significantly increased RV systolic pressure and the RV weight/body weight ratio in PAB; it significantly decreased developed tension from 59 ± 3 mN/mm2in control to 25 ± 4 mN/mm2in PAB, shortening extent from 0.21 ± 0.01 muscle lengths (ML) in control to 0.12 ± 0.01 ML in PAB, and shortening rate from 1.12 ± 0.07 ML/s in control to 0.55 ± 0.03 ML/s in PAB. Indirect immunofluorescence confocal microscopy showed that PAB muscles had a selective increase in microtubule density and that colchicine caused complete microtubule depolymerization in both control and PAB papillary muscles. Microtubule depolymerization normalized myocardial contractility in papillary muscles of PAB cats but did not alter contractility in control muscles.
Conclusions. Excess microtubule density, therefore, is equally important to both cellular and to myocardial contractile dysfunction caused by chronic, severe pressure-overload cardiac hypertrophy.
A cardinal molecular etiology of the contractile defects characteristic of chronic pressure-overload hypertrophy of a degree relevant to human pathophysiology is a selective and persistent increase in the microtubule component of the extramyofilament cytoskeleton (1–6). In earlier work, we found that right ventricular (RV) pressure-overload hypertrophy (POH) causes a decreased contractile state when measured in isolated RV papillary muscle preparations (7), and that cardiac muscle cells, or cardiocytes, isolated from RVPOH tissue exhibit an equivalent degree of contractile dysfunction (8). In more recent work, we found that RV and left ventricular (LV) POH cause increased microtubule network density and up-regulation of α-tubulin and β-tubulin, the major microtubule proteins, on the mRNA and protein levels (1–4,6). Importantly, microtubule depolymerization in hypertrophied cardiocytes changed the cardiocyte contractile state, characterized in terms of sarcomere and cellular mechanics, from clearly depressed to entirely normal (1–4,6)by removing a viscous load from the shortening sarcomeres (5). To date, however, the role of excess microtubules in causing contractile dysfunction has been examined only in isolated cardiocytes. The next logical step in establishing the pathophysiological significance of this cytoskeletal alteration would be to show that correction of the cellularcontractile state by microtubule depolymerization corrects the myocardialcontractile state when examined at the more highly integrated and complex level of muscle tissue.
This rather straightforward goal is made more compelling by the fact that the results of such a study are neither obvious nor predictable on the basis of current data. That is, myocardium is a complex tissue composed of cardiocytes admixed with endothelial, neural, vascular, and interstitial cells. The myocardial contractile deficits of chronic POH may be due to changes intrinsic or extrinsic to the cardiocyte. Extracardiocyte factors such as loading conditions, extent of heterogeneity, and neurohumoral factors may alter myocardial function; and cardiocyte interactions with a denser connective tissue network, decreased capillary density, and/or increased diffusion distances may alter myocardial contractile state independent of changes in cardiocyte properties. Intracardiocyte factors, in addition to the extramyofilament cytoskeleton, such as calcium homeostasis, energetics, and myosin ATPase isotypes, may also influence myocardial function. Thus, it is entirely possible that, despite the dramatic amelioration of contractile dysfunction seen at the level of isolated cardiocytes, microtubule depolymerization might well not correct the myocardial contractile dysfunction resulting from severe chronic pressure-overload cardiac hypertrophy.
It is also possible that myocardial contractile defects might be only incompletely corrected by microtubule depolymerization, as in previous experiments most of the studies were performed in freely contracting, completely unloaded cells (1–4,6). Under such conditions, cardiocytes perform no external work, consume little energy, and may exhibit preserved shortening even under marginal conditions. To determine maximum work capacity, contractile state should be assessed under physiological conditions wherein cardiocytes shorten against known, measurable, and variable loads. Because it is possible that some level of contractile dysfunction might be present but unrecognized in freely shortening, unloaded cardiocytes, the completeness of any correction of myocardial contractile state can only be determined in muscle preparations loaded in a physiological manner.
Thus, whether acute microtubule depolymerization would correct myocardial as opposed to cellular contractile dysfunction remained an important but unanswered question. Because our central goal was to determine whether this cytoskeletal abnormality had a role in clinical heart disease, it was our intent to go from the cell, to tissue, to the intact heart in animal models and then to clinical studies of patients with POH. This study, which constitutes the step from cell to tissue, was designed to address the question of whether microtubule depolymerization normalizes myocardial as opposed to cellular contractile dysfunction in severe pressure-overload hypertrophy.
Right ventricular papillary muscles were isolated from cats with RVPOH after pulmonary artery banding (PAB). Muscle function was assessed using physiologically sequenced contractions at baseline and after colchicine treatment; colchicine causes microtubule depolymerization by binding to β-tubulin and preventing αβ-tubulin polymerization into microtubules (9). As the half-life of microtubules is ≈30 min, colchicine causes a reduction in the number of microtubules over 30 to 90 min.
If excess microtubules cause contractile dysfunction in hypertrophied myocardium, the same contractile dysfunction should appear in normal myocardium if excess microtubule polymerization is induced by a nonhemodynamic means, such as treatment with taxol, an agent that increases microtubule network density by lowering the critical concentration of αβ-tubulin heterodimers required to form microtubules (2). Thus, RV papillary muscles were isolated from normal cats, and myocardial function was assessed at baseline and after treatment with taxol.
Pulmonary artery banding
Right ventricular POH was induced as before (7)by partially occluding the pulmonary artery with a 2.9-mm internal diameter band. Twenty-five adult cats of random sex weighing 2.6 to 3.9 kg underwent PAB and then recovered for 4 weeks. Twenty-five normal adult cats of random sex served as controls.
Hemodynamic studies and papillary muscle isolation
At final study, the cats were anesthetized, right heart pressures were obtained, and arterial pressure was monitored. Arteriovenous oxygen content was used as a measure of cardiac output (7). The heart was then perfused with a Krebs-Henseleit cardioplegia solution consisting of (mmol/liter): NaCl 98.0; KCl 4.7; MgSO41.2; KH2PO41.1; NaHCO324.0; NaAc 20.0; CaCl22.5; glucose 11.2; 2,3-butanedione monoxime (BMD) 30.0; and insulin 10 U/l. This solution, continuously bubbled with 95% O2–5% CO2at room temperature and pH 7.38, was first infused as a 50-ml bolus over 1 min, causing cardiac standstill, and then at a rate of 5 to 10 ml/min while 1 to 3 RV papillary muscles were excised. A 6-0 silk suture was tied to the top of each muscle at the chorda–muscle junction, and the base was attached to a spring clip. The muscles were placed in oxygenated cardioplegia solution for 30 min and then in this cardioplegia solution without BDM for a 15-min washout period before electrical stimulation was begun. Once the papillary muscle was placed in the isolated muscle chamber, and throughout the subsequent study, the temperature was held at a constant 29°C.
Papillary muscle servo control system
After the washout period, the muscle was electrically stimulated by parallel platinum electrodes delivering 5-ms pulses at a voltage 10% over threshold. The silk suture on the upper end of the papillary muscle was attached to a Cambridge 300 B Servo Control System, and the lower clip was attached to a semiconductor strain gauge transducer (DSC-3; Kistler-Morse). A digital computer with an analog–digital interface controlled either tension or length of the preparation. Tension and length data were sampled at a rate of 1 kHz and stored for later analysis. The precision of the force and length settings was 5 mg and 2 μm, respectively. The stepped response of the system to an imposed length change was 95% complete in 2 ms. Equipment compliance was <1.0 μm/mN.
Measurements of papillary muscle mechanics
At each study point, papillary muscle function was assessed defining three relationships: active force versus muscle length; shortening velocity versus force; and shortening extent versus force. Muscle length was expressed as a percent of the length at peak active tension (Lmax), and muscle load was normalized to the cross-sectional area of the muscle (mN/mm2). At the end of each study, muscle length (ML) was measured at the Lmaxpreload. Muscle cross-sectional area was determined assuming a uniform cross section, from muscle length at Lmax, muscle dry weight, a wet-to-dry weight ratio of 4:1, and a specific gravity of 1.0. Muscles with cross-sectional areas <0.5 mm2or >1.5 mm2were excluded from analysis, as previous studies have shown that if cross-sectional area is <1.5 mm2, there is no central core hypoxia (10). There was no significant difference between control versus experimental groups either in muscle length (6.0 ± 1.0 mm in normal vs. 6.2 ± 1.0 mm in PAB) or in muscle cross-sectional area (1.0 ± 0.2 mm2in normal vs. 1.2 ± 0.3 mm2in PAB).
Protocols used to assess myocardial function
Each papillary muscle was allowed to equilibrate by contracting isotonically at a 0.5-g preload for 120 min; at 15-min intervals the muscle was gradually stretched to Lmax, and isotonic contractions at a 0.5-g preload and isometric contractions at the Lmaxpreload were studied. Mechanical equilibrium was said to have occurred when muscle length at Lmax, shortening extent during isotonic contraction, and active tension during isometric contraction were constant during three consecutive measurements at 15-min intervals. Baseline values were then recorded, Lmaxwas determined at baseline, and isotonic, isometric, and graded physiologically sequenced contractions were studied.
After the baseline measurements, colchicine was added to the buffer to a final concentration of 10−5mol/liter. This concentration, which was higher then that used in our isolated cardiocyte studies, was chosen to ensure diffusion throughout the muscle. At 15-min intervals for a total of 90 min, Lmaxwas determined, and isotonic, isometric, and graded physiologically sequenced contractions were studied. This protocol of 120 min of equilibrium followed by 90 min of colchicine (Protocol A) was followed for 10 control muscles from normal cats and for 10 hypertrophied muscles from PAB cats. To show that the muscles were stable during the time required for these studies, two normal and two PAB muscles were examined over a 300-min period without colchicine (Protocol B).
To then prove that the colchicine effect was specific, two papillary muscles were isolated from each of three PAB cats. In the first cat, the first papillary muscle was allowed to equilibrate for 120 min and then treated with colchicine for 180 min; the second papillary muscle was allowed to equilibrate for 120 min and then not treated with colchicine and was reassessed throughout the subsequent 180 min (Protocol C). In the second catthis order was reversed: the first papillary muscle was allowed to equilibrate and then not treated with colchicine; the second papillary muscle was allowed to equilibrate and then treated with colchicine. In the third cat, the first muscle was allowed to equilibrate for 210 min and then treated with colchicine for 90 min; the second papillary muscle was allowed to equilibrate for 120 min but then not treated with colchicine and followed for a total of 300 min (Protocol D). By studying two muscles from an individual animal, interanimal variation was avoided, and the effects of colchicine could be more specifically determined.
After baseline measurements, taxol was added to the buffer to achieve a final concentration of 10−5mol/liter. At 15-min intervals for a total of 4 h, Lmaxwas determined, and isotonic, isometric, and graded physiologically sequenced contractions were then studied.
In addition to demonstrating the effects on myocardial mechanics of interventions designed to have selective effects on microtubules, the relative amounts of free and polymerized β-tubulin in papillary muscles were characterized by immunoblotting, and the appearance and density of the cardiocyte microtubule network were visualized by indirect immunofluorescence microscopy.
Immunoblot analysis, using a monoclonal β-tubulin antibody (Amersham), was performed on papillary muscles isolated from the RVs of normal and PAB cats using previously described methods (1–3). Papillary muscles were processed immediately after isolation or after being treated with colchicine (10−5mol/liter for 90 min) or taxol (10−5mol/liter for 4 h).
Indirect immunofluorescence confocal microscopy
The microtubule network was characterized in papillary muscles isolated from normal and PAB cats in the baseline state after treatment with colchicine. In addition, the microtubule network was assessed in the RV free-wall tissue of normal and PAB cats to ensure that papillary muscles were representative of the myocardium as a whole. Of particular importance, micrographs from the papillary muscles were taken from the geometric center of the papillary muscle to ensure that colchicine-induced microtubule depolymerization had occurred throughout the entire muscle.
Micrographs were obtained in papillary muscles that were immersion-fixed and in myocardial tissue, which was perfusion-fixed with a periodate-lysine-paraformaldehyde fixative. After blocking with 10% horse serum in 0.1 mol/liter glycine, the sections were incubated with the monoclonal β-tubulin antibody and then with fluorescein isothiocyanate-labeled donkey anti-mouse immunoglobulin. Optical sections (≈0.7 μm) were acquired by confocal laser scanning microscopy.
Intracellular calcium transients
Intracellular calcium transients were measured in normal and PAB RV papillary muscles with the Ca2+-regulated bioluminescent indicator aequorin using previously described methods (11). Aequorin solution (1 mg/ml; 1.0 to 1.5 μl) was pressure-injected through a glass micropipette just beneath the endocardium of the muscles. One hour after aequorin loading, when Ca2+transients and resting and developed tension were stable, the muscles were stretched to Lmaxand stimulated to contract isometrically at 4-s intervals.
Measurements of cAMP
Adenosine 3′:5′-cyclic monophosphate (cAMP) was measured by the manufacturer’s protocol using a cAMP [I125] assay (Amersham) in normal and PAB papillary muscles under 1 of 6 protocols. In Protocol 1, cAMP was measured in normal and PAB muscles immediately after isolation. In Protocol 2, cAMP was measured in normal and PAB muscles immediately after 60-min exposure to 10−5mol/liter colchicine. In Protocol 3, cAMP was measured in normal and PAB muscles after 30-min exposure to 10−5mol/liter colchicine followed by 2-min concurrent exposure to 3-isobutyl-1-methylxanthine (IBMX). In Protocol 4, normal and PAB muscles were kept in Krebs-Henseleit buffer for 30 min; IBMX was then added for 2 min, and cAMP levels were measured. In Protocol 5, cAMP was measured in normal and PAB muscles after 30-min exposure to 10−5mol/liter colchicine. In Protocol 6, normal and PAB muscles were kept in Krebs-Henseleit buffer for 30 min, and cAMP levels were measured.
The mean and standard error are shown for each group. Differences between group means at baseline and following colchicine treatment, considered significant at p < 0.05, were determined using a multiway analysis of variance and a Newman-Keuls multiple sample comparison test.
Effects of PAB on myocardial contractile function
The effects of PAB on in vivo measures of pressure, oximetry, and mass are shown in Table 1; the effects on in vitro myocardial contractile functions are shown in Table 2and Figures 1 to 4. ⇓⇓⇓⇓The in vivo data for normal and PAB cats are similar to those in our previous studies (1–5,7,8), in that PAB caused significant increases in RV systolic pressure and mass but did not cause RV failure. Figures 1 and 2, examples of isotonic and isometric contraction from normal and PAB muscles, show that PAB caused marked decreases in both active tension generation and shortening extent. Figure 3shows that PAB caused a marked decrease in active tension at all muscle lengths. Figure 4, Ashows that PAB caused a marked decrease in the entire shortening extent versus force relationship, and Figure 4, Bshows that PAB caused a marked decrease in the entire shortening velocity versus force relationship. Thus, RVPOH causes a marked decrease in myocardial papillary muscle contractility.
Effects of colchicine on myocardial contractile function
Figures 1 to 4show the effects of colchicine treatment on the contractile state of normal and PAB papillary muscles. Colchicine had no significant effect on the contractile state of normal muscles. In these muscles, 90 min after exposure to colchicine, active tension at Lmaxwas 63 ± 5 mN/mm2, shortening extent at 0.5 gof force was 0.22 ± 0.01 muscle lengths (ML), and shortening velocity at 0.5 gof force was 1.24 ± 0.14 ML/s. In contrast, in hypertrophied PAB muscles, colchicine caused a marked increase in contractile state. In PAB muscles, 90 min after exposure to colchicine, active tension at Lmaxwas 55 ± 5 mN/mm2, shortening extent at 0.5 gof force was 0.19 ± 0.01 ML, and shortening velocity at 0.5 gof force was 1.10 ± 0.07 ML/s. Thus, the contractile defects present in hypertrophied papillary muscles from PAB cats were fully normalized by treatment with colchicine.
Specificity of colchicine effects
Studies were done to prove that the improvement in myocardial contractility was specific to colchicine treatment and that time alone did not normalize function. Figure 5shows that the mechanical performance was stable in both normal and hypertrophied papillary muscles. Function reached equilibrium after 90 to 120 min and then remained constant for up to 5 h.
We studied pairs of papillary muscles from three PAB cats. In the first cat, the mechanical performance of both muscles had reached equilibrium by 90 min. The first muscle was then treated with colchicine at 120 min; contractile function was abnormal initially but returned to normal at 90 min of colchicine exposure and remained stable through 300 min (Figure 6, papillary muscle 1). The second muscle was not treated with colchicine; contractile function was abnormal initially and remained abnormal through 300 min (Figure 6, papillary muscle 2). In the second cat, the order of study of the papillary muscle pair was changed, but the results were very similar. In the third cat, mechanical performance of both muscles reached equilibrium by 90 min. The first muscle was then treated with colchicine at 210 min, after 120 min of stable but depressed mechanical performance. After colchicine treatment contractility became normal (Figure 7, papillary muscle 1). The second muscle was not treated with colchicine and showed characteristic persistent contractile deficits (Figure 7, papillary muscle 2). Thus, the mechanical effects of colchicine are specific and independent of the time of treatment.
Effects of taxol on myocardial contractile function
Figure 8shows the effects of taxol on the contractile state of normal papillary muscles. After 4 h of exposure to taxol, active tension at Lmaxwas 27 ± 8 mN/mm2, shortening extent at 0.5 gof force was 0.13 ± 0.03 ML, and shortening velocity at 0.5 gof force was 0.60 ± 0.05 ML/s. Thus, taxol caused a contractile defect in normal muscles equivalent to that seen in RVPOH muscles.
The immunoblots in Figure 9show that at baseline hypertrophied muscles had an increase in both free and polymerized β-tubulin when compared to normal papillary muscles. In addition, treatment of either normal of hypertrophied muscles with colchicine caused microtubule depolymerization. Finally, treatment of normal control papillary muscles with taxol increased the amount of polymerized β-tubulin to levels comparable to those seen in hypertrophied myocardium.
The micrographs in Figure 10show that, compared to muscles from normal cats (Panel A), muscles from PAB cats exhibit an increase in the density of the microtubule network (Panel C). Treatment of both normal muscles (Panel B) and PAB muscles (Panel D) with 10−5mol/liter colchicine for 90 min caused nearly complete depolymerization of the microtubules. The findings were similar when the microtubule network was assessed in RV free-wall samples from normal and PAB cats (data not shown). Of note, these data from ventricular tissue and papillary muscles are identical to our data from isolated cardiocytes both before and after colchicine treatment.
Measurements of calcium transients and cAMP
Treating papillary muscles isolated from normal and PAB cats with colchicine did not augment either the intracellular calcium transient or intracellular cAMP levels. Figure 11shows that in normal papillary muscles, colchicine caused a small decrease in peak developed tension, which corresponded to a small decrease in peak activated calcium. Neither the rate of rise nor the rate of decline in intracellular calcium was altered by colchicine. In contrast, in hypertrophied papillary muscles, although the prolonged calcium transient that we have seen before with longer-term hypertrophy and most apparently at supraphysiological [Ca2+]o(11; Figure 1) was not observed, colchicine caused a significant increase in peak developed force but a small decrease in peak activation calcium, and no significant change in either the rate of rise of the rate of decline in intracellular calcium.
In each of the six cAMP protocols used, there were no significant differences in cAMP between normal and hypertrophied muscles. For example, in freshly isolated muscles, cAMP was 1.03 ± 0.02 pmol/mg in normal and 0.93 ± 0.01 pmol/mg in hypertrophied muscles. After 90 min of treatment with 10−5mol/liter colchicine, cAMP was 0.60 ± 0.02 pmol/mg in normal and 0.47 ± 0.01 pmol/mg in hypertrophied muscles. After 10−5mol/liter colchicine for 30 min, then IBMX for 2 min, cAMP was 1.38 ± 0.02 pmol/mg in normal and 0.99 ± 0.02 pmol/mg in hypertrophied muscles.
Our previous studies suggested that a selective and persistent increase in microtubules has an important causative role in the transition from severe but compensated pressure-overload cardiac hypertrophy to decompensated cardiac failure (1–6). However, the role of microtubules in causing contractile dysfunction had been examined only in isolated cardiocytes. The next logical step was to determine whether correction of cardiocyte contractile function in situ by microtubule depolymerization would correct myocardial contractile dysfunction when examined at the more highly integrated and complex level of cardiac tissue. We show here that RVPOH caused a significant decrease in myocardial papillary muscle contractility, which was associated with a marked increase in the density of the microtubule portion of the extramyofilament cytoskeleton. Depolymerization of these microtubules with colchicine resulted in normalization of myocardial papillary muscle contractility.
Furthermore, when excess microtubule polymerization was created in normal myocardium by treatment with taxol, contractile dysfunction developed in these normal papillary muscles that was identical to that seen in chronic pressure-overload hypertrophy. Thus, the abnormality in myocardial contractility seen in severe chronic pressure-overload hypertrophy was caused specifically by an increase in intracellular microtubules, such that correction of cardiocyte contractile function by microtubule depolymerization resulted in complete normalization of the myocardial contractile state.
Role of microtubules in contractile dysfunction
Identifying the mechanism or mechanisms that cause the transition from compensated hypertrophy to compensated failure clearly has important clinical implications and has been the subject of intense and extensive investigation. Although the current study was aimed specifically at defining the role that microtubules play in this process, we and others have studied a number of other molecular and cellular mechanisms that have been postulated to contribute to the development of contractile dysfunction in pressure-overload hypertrophy (7,12–14). In the model of RVPOH used in the current study, abnormalities have been identified in myocardial energetics, calcium homeostasis, and the extracellular matrix (12–14). Further, a number of other mechanisms have been identified in RV and LV models of POH that may be pertinent to the current RVPOH model, including changes in the myosin heavy-chain and myosin light-chain isoforms and changes in the β-adrenergic receptor pathway and in other receptor pathways (12–14).
There are, then, a number of molecular, cellular, and extracellular mechanisms, in this and other models of POH, which might have causal roles in the development of myocardial dysfunction. It is for this reason that the results of this study were both unpredictable and surprising. We expected that correcting cardiocyte microtubule hyperpolymerization alone would only partially correct the abnormalities in myocardial contractile state in this model. This prediction was based on the assumption that each of the above mechanisms would have some role in causing contractile dysfunction and the fact that colchicine, acting specifically on microtubule polymerization, would have no effect on the other mechanisms noted earlier. It is thus remarkable that correcting microtubule hyperpolymerization alone caused complete normalization of myocardial contractile state on the tissue level, indicating that changes in microtubules play a predominant role not only in cardiocyte but also in myocardial contractile dysfunction in severe POH.
Nonetheless, although this study shows that microtubules have the predominant role in causing contractile dysfunction in this model or RVPOH, it is clear that microtubule hyperpolymerization is not the only mechanism responsible for the transition from compensated cardiac hypertrophy to decompensated cardiac failure. Instead, the role that microtubules have in the development of contractile dysfunction appears to be specific to POH in adult mammals in which wall stress in not normalized (6). That is, an increase in the microtubule component of the extramyofilament cytoskeleton is not a generalized response to hemodynamic overloading, such that it would be expected that contractile dysfunction in many forms of hemodynamic overloading would be caused by other molecular and cellular abnormalities (1–4). Thus, although all cardiomyopathic states are not related to changes in microtubules, our studies have defined a specific set of clinically relevant circumstances that do produce contractile dysfunction caused by excessive microtubules. These circumstances are not limited to the cat or to the RV, and these findings have been confirmed by others. Cardiocytes isolated from a canine LV POH model showed similar increases in microtubule polymerization and similar decreases in contractile state, which were normalized by depolymerization of the cardiocyte microtubules (6). In a confirmatory study, Ishibashi et al. (15)showed similar results using an adult rodent model of LV POH, as has Wang et al. (16), and increased cardiocyte microtubule density has now been found in end-stage LV failure in humans as well (17).
Specificity of colchicine effects
Our conclusion that microtubules play a pivitol role in the development of contractile dysfunction in POH is based in part on the assertion that colchicine acts specifically through its effects on microtubule depolymerization and not through a nonspecific, microtubule-independent effect on contractile state. Data obtained from the current study and from our previous studies clearly support this hypothesis (1–6). These studies addressed three specific issues: 1) does colchicine affect contractile state independent of its effects of microtubules, 2) does colchicine affect intracellular calcium homeostasis, and 3) does colchicine alter intracellular cAMP?
Data from the current and previous studies (1–6)support the conclusion that alterations in contractile state caused by treating muscles and cardiocytes isolated from cats with POH with colchicine are specific to microtubule depolymerization. In these studies, microtubule depolymerization was accomplished using both chemical and physical means. For example, microtubule depolymerization was induced in hypertrophied cardiocytes by treatment with colchicine (a chemical approach) or treatment with hypothermia (a physical approach). Although both may have had other effects, the one effect that these two very different approaches clearly shared was microtubule depolymerization. Because both approaches normalized the contractile dysfunction present in hypertrophied cardiocytes, it is likely that this outcome was caused by the effect they shared in common (microtubule depolymerization) and was independent of other effects by which they differed. In addition, if colchicine had direct inotropic effects, independent of microtubule depolymerization, then its effects on contractility would have been greater than the effects of hypothermia. Finally, colchicine would have caused significant changes in contractile state in normal muscles or cells. Colchicine had neither of these latter two effects.
Colchicine does not alter calcium homeostasis. We previously showed that colchicine does not increase either resting or peak activated calcium levels in normal or hypertrophied cardiocytes (2). The current study using aequorin-labeled papillary muscles shows that colchicine does not increase the peak activated calcium level or the rate of rise or fall of intracellular calcium.
Colchicine does not alter intracellular cAMP levels. These findings are in concordance with a study performed in isolated leukocytes which showed that treatment with colchicine alone did not significantly affect cAMP levels (18). Colchicine did potentiate the effect of specific drugs on cAMP levels on leukocytes. Thus, colchicine pretreatment caused the action of β-adrenergic agonists, prostaglandin E, and phosphodiesterase inhibitors on cAMP to be markedly increased when compared to the effects of these drugs when leukocytes were not pretreated with colchicine. However, no changes in β-adrenergic tone, prostaglandin, or phosphodiesterase levels would be expected to occur in the current isolated muscle preparation. Therefore, the lack of change in cAMP in response to colchicine in our papillary muscle preparation was an expected outcome.
Chronic RVPOH causes a significant decrease in papillary muscle contractility, which in turn is caused by a significant increase in the microtubule portion of the extramyofilament cardiocyte cytoskeleton. Acute depolymerization of these excessive microtubules with colchicine normalized myocardial papillary muscle contractility. These changes in microtubule density and polymerization state are equally important both to cardiac tissue and to cardiac cellular contractile state. Thus, these specific and persistent cardiocyte microtubule abnormalities have a critical role in the changes in myocardial contractility that occur during severe chronic pressure overload cardiac hypertrophy.
The authors thank Mary Barnes, Sebette Hamill, and Gilberto DeFreytes for technical assistance.
☆ This study was supported by the Department of Veterans Affairs Research Service (G.C., M.R.Z.), by NIH grant P01-HL-48788 (G.C., M.R.Z.), and by the National Aeronautics and Space Administration (M.R.Z.).
- 2,3-butanedione monoxime
- adenosine 3′:5′-cyclic monophosphate
- length for peak active tension
- left ventricular
- muscle length
- pulmonary artery band
- pressure-overload hypertrophy
- right ventricular
- Received June 23, 1998.
- Revision received August 6, 1998.
- Accepted September 4, 1998.
- American College of Cardiology
- Tsutsui H,
- Ishihara K,
- Cooper G
- Tsutsui H,
- Tagawa H,
- Kent R.L,
- et al.
- Tagawa H,
- Rozich J.D,
- Tsutsui H,
- et al.
- Tagawa H,
- Wang N,
- Narishige T,
- Ingber D.E,
- Zile M.R,
- Cooper G
- Tagawa H,
- Koide M,
- Zile M.R,
- Carabello B.A,
- Cooper G
- Cooper G,
- Satava R.M,
- Satava R.M,
- Harrison C.E,
- Coleman H.N
- Mann D.L,
- Urabe Y,
- Kent R.L,
- Vinciguerra S,
- Cooper G
- Coleman H.N
- Gwathmey J.K,
- Morgan J.P
- Ishibashi Y,
- Tsutsui H,
- Yamamoto S,
- et al.
- Wang X,
- Campbell S.E,
- Gerdes A.M
- Hein S,
- Schaper J
- Rudolph S.A,
- Greengard P,
- Malawista S.E