INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

DEPRESSED LEFT VENTRICULAR (LV) function after myocardial infarction (MI) is associated with the development of myocardial fibrosis (increased myofibrillar collagen), abnormal myocardial stiffness, and myocardial contractile dysfunction (12, 29, 30). Despite considerable research on these three factors, it remains controversial as to the level of influence that each plays in the development of LV dysfunction. In particular, little is known about the effects of myocardial fibrosis on myocardial stiffness and papillary muscle function. Although myocardial fibrosis and stiffness have long been associated with cardiac failure due to various causes, there are no prior studies that have established the nature of the relationship between myocardial fibrosis, myocardial stiffness, and myocardial dysfunction.

The renin-angiotensin system is a major determinant in the development of LV remodeling. Recently, it has been shown that angiotensin II type 1 (AT₁) receptor blockade reduces myocardial hypertrophy, decreases myocardial fibrosis, and attenuates LV remodeling to the same degree as angiotensin-converting enzyme inhibition in the rat ischemic heart failure model (26). Previous studies suggest that after MI, myocardial systolic function, but not passive stiffness, may be improved with chronic angiotensin-converting enzyme inhibition (12). However, it is not known whether either active or passive myocardial function improves in association with the known favorable alterations in LV remodeling with AT₁ receptor blockade.

The purpose of this study was to determine the effects of angiotensin II receptor blockade on LV hemodynamics, myocardial fibrosis, myocardial stiffness, and papillary muscle function in rats with heart failure after MI. In particular, we sought to determine whether prevention of myocardial fibrosis by early treatment with AT₁ receptor blockade would be associated with improvement in myocardial contractile function and passive stiffness.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Rat model of myocardial infarction. Myocardial infarction was produced in male Sprague-Dawley rats (8-10 wk old) using left coronary artery ligation as previously described (12, 23, 24). In brief, animals were anesthetized with intramuscular 2.4 mg/kg acepromazine maleate and 125 mg/kg ketamine HCl, and a left thoracotomy was performed under sterile conditions. The heart was expressed from the chest cavity, and a 7-0 synthetic ligature was placed around the proximal left coronary artery. The heart was returned to the chest, and the thorax was closed. With this method, 45% of infarcted rats die within the first 24 h. Sham-operated rats [animals in which coronary artery ligation was attempted but in which there was no electrocardiogram (ECG) evidence of infarction, normal hemodynamics, and no scar at autopsy] served as controls. One day after surgery, the rats were randomized to treatment or placebo according to the presence or absence of ECG evidence of MI (23, 24). Randomization by ECG is not intended to confirm the presence or absence of MI but to ensure balanced randomization during treatment. Rats randomized to treatment received losartan in a dose of 2 g/l of drinking water (18, 23). All rats were fed standard rat food, given water ad libitum, and housed in a single room of the animal facility with a 12:12-h light-dark cycle and independent ventilation, temperature, and humidity control. Subsequent experiments were performed 8 wk after surgery, with animals receiving treatment up to the day of experimentation. Final classification of rats as either MI (presence of anterior LV scar) or sham (no LV scar) occurred on the day of experimentation. The study was terminated after 10 rats in each of 4 groups (sham, sham-losartan, MI, and MI-losartan) had been randomized and successfully studied. The study was performed in an American Association for Accreditation of Laboratory Animal Care-accredited facility with approval from the animal use committees of the Tucson Veterans Affairs Medical Center and the University of Arizona.

LV hemodynamics. Eight weeks after surgery, rats were anesthetized with thiobutabarbitol (100 mg/kg ip). A 1-mm micromanometer-tipped catheter (Millar Instruments, Houston, TX) was inserted into the right carotid artery. The catheter was advanced into the aorta and then into the LV under constant pressure monitoring. The zero-pressure baseline was obtained by placing the pressure sensor in 37°C saline before measurements. After a period of stabilization, LV pressures were recorded and digitized at 1,000 Hz, using an IBM 486 PC equipped with an analog-to-digital converter and customized software. From these data, heart rate and the first derivative of LV pressure with respect to time (dP/dt) were derived, according to previously described methods (12, 23, 24). Phasic aortic pressure was measured, and the electronic mean was determined after withdrawal of the LV catheter into the aortic root.

Isolated papillary muscle function. Papillary muscle function was assessed using previously published methods (12, 16). After LV hemodynamic measurements were completed, the rat was rapidly killed by removing the heart to an oxygenated dissecting bath. The LV was opened from the septum, and the noninfarcted posterior papillary muscle was dissected free from the LV wall using a dissecting microscope. The ends of the muscle were grasped with spring clips and suspended vertically from an isometric force transducer (Metrigram, Gould Instruments, Cleveland, OH) in a tissue bath containing modified Krebs-Henseleit solution, which was composed of (in mM) 120 NaCl, 5.9 KCl, 5.5 dextrose, 25 NaHCO₃, 1.2 NaH₂PO₄, 1.2 MgCl₂, and 1.2 CaCl₂. The bath was maintained at a constant temperature of 30°C and bubbled with 95% O₂-5% CO₂. The muscle was allowed to equilibrate for 15 min without any electrical stimulation. The muscle was then stimulated (S44, Grass Instruments, Quincy, MA) to contract isometrically at 0.33 Hz by use of field stimulation delivered through a a pair of platinum electrodes placed parallel to the muscle. Five-millisecond square-wave pulses at a voltage of ~10% above threshold were used. The muscle was allowed 60 min of equilibration while being stimulated. The muscle was stretched to the length at which maximal tension development occurred (L_max). Muscle length was measured at L_max with the aid of a calibrated microscope (M101 AT, Gaertner Scientific, Chicago, IL) with direct real-time video feed into an IBM AT computer. Tension was recorded on a physiological recorder (Gould Recording), and digitized data points recorded at 500 Hz were stored on line on an IBM microcomputer with customized software. All measurements were normalized to papillary muscle cross-sectional area (CSA) determined at L_max. With the assumption of cylindrical geometry and specific gravity of 1.05, papillary muscle CSA was calculated as From the digitized tension signals, maximum developed tension (DT), maximum rate of tension development (+dT/dt), time to peak tension development (TPT), and maximum rate of tension fall (dT/dt) were derived.

Myocardial stiffness. After the isometric papillary muscle function study was completed, myocardial stiffness was measured by passively stretching the muscle at a constant strain rate to 1.4 times its initial length. A slow rate of stretch was chosen to avoid viscous effects. The muscle was initially conditioned by two successive stretches. Natural strain  (ln L/L_o), where L is the instantaneous length and L_o is length at 0.1 g/mm², was determined at each digitized data point on the tension-time curve with the aid of a microscope with direct real-time video link to an IBM microcomputer with a specialized computer software imaging program. By measuring L_o at the beginning of each stretch and knowing the rate of strain, we calculated muscle length (L) at any given time from the video-microscope image, in which the relative displacement of two predetermined markers was electronically measured. Approximately 14-16 tension-natural strain data pairs were recorded for each full-stretch cycle. Three separate stretches (up to 1.4L_o) of each muscle were done, and the results were averaged. Myocardial stiffness was calculated using a modification of previously described methods (12, 23, 24, 27). Natural strain was plotted against stress (tension normalized to CSA), and the curve was fitted to the following exponential equation

(1)

where  is instantaneous stress (indicated as a function of strain), _o is stress at L_o, e is the base of the natural logarithm, K_m is the muscle stiffness constant, and  is natural strain. Taking the logarithm of both sides of Eq. 1 gives

(2)

Therefore, K_m is measured by plotting ln  vs.  in Eq. 2, where K_m is the slope of this line. Next, differentiating Eq. 1 yields

(3)

where E is myocardial stiffness and () is myocardial stress as a function of strain. The term d/d, the elastic stiffness (27), can now be calculated from Eq. 3 and plotted vs. corresponding myocardial strain.

Morphology, morphometry, and infarct size. Immediately after harvest of the posterior papillary muscle, the left ventricle and septum were separated from the right ventricle (RV free wall), weighed, and immersed in 10% Formalin. Twenty-four hours later, the left ventricle was sliced into four coronal sections from apex to base and embedded in paraffin. One thin (4 µm) section was obtained from each piece, mounted onto slides, and stained with myofibrillar collagen-specific picosirius red. Myocardial collagen (percent fibrosis) was determined by quantitative morphometry of the sirius red-stained slides (9). The percent fibrosis determined by the morphometric approach is correlated with hydroxyproline concentration of the LV (22). Each section was projected using a microscope (Leica) at ×400 magnification, and the red-staining (in direct light) collagen fibers were digitized using a software system (Global Labs) based on their color emission (9, 22). The percent area of fibrosis was calculated as the sum of all connective tissue areas (minus infarcted and perivascular areas) divided by the area occupied by all tissue (muscle plus all connective tissue) in the visual field. Five fields were evaluated in each slice, and their percent areas of fibrosis were averaged. Preliminary work from our laboratory showed that this number of fields was an adequate sampling size for determination of myocardial fibrosis. The average value of each slice was then added to that of the other slices and averaged to obtain the value reported for each rat. The papillary muscle was also immersion fixed at the conclusion of measurements of isometric and passive mechanical function, processed, and stained as described above, and percent fibrosis was calculated from the average of four thin transverse sections cut along the long axis of the muscle. Myocardial infarct size was determined using techniques described previously (12, 23, 24). Briefly, four thin transverse sections of the LV, obtained along the apex-base axis, were stained with trichrome. The LV endocardial and epicardial perimeters were traced using the Global Lab video-image analyzer system, and the arc length of the infarcted region was determined for each section. Infarct size is given as the mean percentage of epicardial and endocardial circumferences occupied by scar tissue for the four sections.

Statistical analysis. All results are expressed as means ± SD. Statistical comparisons between the sham-operated control and MI rats were carried out using two-way ANOVA. This analysis employed a 2 × 2 matrix, in which sham/MI (yes/no) represents one dimension and losartan/placebo treatment (yes/no) represents the other. Interactions were tested using two-way ANOVA, whereas intergroup differences were tested using the Student-Newman-Keuls procedure. Level of significance was taken at P  0.05.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Surgical mortality, body and heart weights, and infarct size. Mortality in the first 24 h after coronary ligation was 46%. One rat died before instrumentation. A total of 40 rats were randomized to placebo or treatment with losartan, with 10 rats in each of the four groups: sham, sham-losartan, MI, and MI-losartan. Body weights, RV and LV weights, and LV-to-body weight ratios are shown in Table 1. Body weights were significantly lower (P = 0.05) in losartan-treated animals, an effect that was independent of MI. However, intergroup comparisons revealed no significant differences in body weight. LV weight was decreased in losartan-treated animals compared with sham controls. When intergroup comparisons were made, LV weight in rats with MI that received losartan decreased (P < 0.05) by 20% when compared with untreated MI rats. RV weight was also decreased (P = 0.02) in losartan-treated rats. Rats with MI that received losartan had a 29% decrease (P < 0.05) in RV weight compared with untreated MI rats. The LV-to-body weight ratio was decreased (P = 0.02) in losartan-treated rats. The ratio was decreased (P < 0.05) ~19% in losartan-treated MI rats compared with untreated MI controls. Average MI size was 35.7 ± 11.3% in untreated rats and 37.4 ± 14.7% in losartan-treated rats (P = NS).

In vivo heart rate, aortic blood pressure, and LV hemodynamics. Data are presented in Table 2. There were no effects of infarction or treatment on heart rate, nor were there any statistically significant differences in heart rate among the four groups of animals. Losartan treatment decreased (P = 0.01) mean arterial pressure (MAP) to approximately the same degree in sham and MI rats. In MI rats treated with losartan, MAP was decreased (P < 0.05) compared with untreated MI rats. Changes in LV systolic pressure were statistically similar to those observed for MAP, i.e., the decrease in peak LV systolic pressure was largely due to losartan treatment. LV end-diastolic pressure (LVEDP) was increased (P = 0.01) in MI rats and was decreased (P = 0.01) in losartan-treated rats. The magnitude of the decrease of LVEDP in losartan-treated rats depended on the presence or absence of MI; in infarcted rats, the decrease in LVEDP was 25% greater than that in sham rats. Maximum positive LV dP/dt was decreased in MI and losartan-treated rats (P = 0.01 and P = 0.01, respectively). Compared with sham rats, LV dP/dt was decreased in untreated and losartan-treated MI rats (P < 0.05, MI vs. sham and MI-losartan vs. sham).

Papillary muscle mechanics (myocardial function). Papillary muscle weights (PMW) were 7.45 ± 2.97, 5.77 ± 1.11, 6.54 ± 1.73, and 5.38 ± 1.73 mg in sham, sham-losartan, MI, and MI-losartan rats, respectively. Papillary CSA values were 1.20 ± 0.28, 1.20 ± 0.21, 1.40 ± 0.40, and 1.37 ± 0.41 mm² in sham, sham-losartan, MI, and MI-losartan rats, respectively. There were no significant effects of MI or treatment on PMW and papillary CSA and no intergroup differences with respect to these two variables.

Myocardial fibrosis and passive myocardial stiffness. The effects of MI and treatment with losartan on noninfarcted papillary muscle percent fibrosis (myofibrillar collagen) are presented in Figs. 1 and 2. The values for percent fibrosis were 1.84 ± 0.94, 1.49 ± 0.67, 3.12 ± 1.12, and 1.69 ± 0.53% for sham, sham-losartan, MI, and MI-losartan rats, respectively. Percent fibrosis was increased (P = 0.01) after MI and was decreased (P = 0.03) with losartan treatment. Compared with sham rats, infarcted rats had a 70% increase (P < 0.05) in percent fibrosis. Losartan decreased (P < 0.05) percent fibrosis in MI rats by 46%. There was no significant difference in percent fibrosis between MI rats treated with losartan and sham rats. The reduction in percent fibrosis in MI rats treated with losartan was 24% greater than that observed in sham-treated rats (P = 0.05, for interaction). The values for percent fibrosis in LV coronal sections were closely correlated with papillary muscle percent fibrosis in the four treatment groups; the values, reported as the average of three coronal sections, were 1.39 ± 0.42, 1.07 ± 0.45, 2.98 ± 1.01, and 1.39 ± 0.56% for sham, sham-losartan, MI, and MI-losartan rats, respectively. The results of statistical comparisons of percent fibrosis of the coronal sections paralleled those for papillary muscle. There were significant effects of MI (P = 0.03) and treatment (P = 0.01) on global LV percent fibrosis, with a significant (P = 0.05) interactive effect of losartan and MI on this variable.

View larger version (122K): [in this window] [in a new window]   Fig. 1.   Photomicrograph of myocardial fibrosis in sham (A), myocardial infarction (MI; B), and MI rats treated with losartan (C). Fibrosis is detected by myofibrillar collagen-specific picosirius red. Note red staining for extracellular matrix around myocyte. An increase in amount of red staining in MI rats and a decrease resulting from losartan treatment are shown.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Data are percent fibrosis of papillary muscles in sham, sham-losartan, MI, and MI-losartan rats. Fibrosis is increased after MI and normalized with losartan treatment. There is no effect of losartan on fibrosis in sham rats. Tx, treatment with losartan. Data are means ± SD; n = 10 rats in each group.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Myocardial stiffness (E) vs. strain for isolated papillary muscles in sham, MI, and MI-losartan rats. Myocardial stiffness-strain curve for sham rats treated with losartan is not displayed, since it is virtually identical to that of untreated sham rats. There are no significant differences in myocardial stiffness among 3 groups at strain values <0.10; however, at all strain values >0.10, myocardial stiffness is increased (* P < 0.05) with MI compared with sham. Losartan treatment normalizes myocardial stiffness in MI rats at all strains >0.15 ( P < 0.05, MI-losartan vs. MI). Data are means ± SE; n = 6 in each group except for MI-losartan, where n = 5. Km, muscle stiffness constant.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

Numerous investigations have documented the beneficial effects of chronic AT₁ receptor blockade in experimental heart failure due to MI. In the rat postinfarction model, for example, it has been shown that AT₁ receptor blockade decreases peripheral vasoconstriction and reduces LVEDP (23), attenuates LV remodeling (25), and is equivalent to angiotensin-converting enzyme inhibition in improving survival in rats (19). Despite these important findings, there are no data on the effects of AT₁ receptor blockade on myocardial systolic and diastolic function or on the relationship between improvements in LV remodeling and prevention of myocardial fibrosis and abnormal passive myocardial stiffness. We report that chronic therapy with losartan, a potent AT₁ receptor blocker, prevents LV hypertrophy and an increase in myofibrillar collagen content (myocardial fibrosis). The prevention of myocardial fibrosis is associated with attenuation of abnormal passive myocardial stiffness.

Effect of AT₁ receptor blockade on cardiac weight and LV hemodynamics. Losartan decreased body weight, an effect that may be due to inhibition of the dipsinogenic effects of angiotensin II and decreased total body sodium and water (10, 14, 23, 33). RV weight was decreased by AT₁ receptor blockade, which most likely is related to decreased RV afterload (elevated LVEDP) and inhibition of the direct effects of angiotensin II on myocyte growth (1, 3, 11, 17). A significant decrease in LV weight was also seen in losartan-treated rats, both at the ventricular level and in the LV-to-body weight ratio. One explanation is that AT₁ receptor blockade by losartan is virtually complete, inhibiting growth and decreasing afterload independent of degree of neurohumoral activation.

Effect of AT₁ receptor blockade on myocardial contractile function. It was recently shown that a specific angiotensin II receptor blocker could decrease LV volumes and increase ejection fraction after MI in rats (15). However, in that study, MAP was also reduced, and the improvement in LV function may, therefore, have been due to a decrease in LV afterload. Although papillary muscle systolic function, which reflects myocardial contractility, was depressed in untreated MI compared with sham, the effect of losartan depended on the presence or absence of MI. The explanation for this observation is unknown but may be related to increased displacement of tissue angiotensin II from the AT₁ to the AT₂ receptor due to losartan blockade of the AT₁ receptor. In sham rats, increased AT₂ receptor stimulation by angiotensin II, in turn, may increase nitric oxide synthesis and subsequently depress myocardial contractility (32). On the other hand, the positive effect of losartan on myocardial contractility in MI rats may be due to chronic unloading via peripheral vasodilatation, which could produce modest enhancement in intracellular calcium handling (18) or restore calcium sensitivity of myofilaments (13).

Effect of AT₁ receptor blockade on myocardial fibrosis and passive myocardial stiffness. After MI, there was a 67% increase in myofibrillar collagen, as reflected by percent myocardial fibrosis, while treatment with losartan prevented myocardial fibrosis. The magnitude of increase in percent fibrosis after MI and attenuation with AT₁ receptor blockade is consistent with data from an earlier study (25). In contrast, we have previously shown that angiotensin-converting enzyme inhibition with captopril did not reduce myocardial fibrosis, as measured by collagen content (µg hydroxyproline/mg dry LV weight), in rats after MI (12). Differences between the two studies in relation to timing and duration of treatment, rather than type of therapy, may account for the different effect on myocardial fibrosis seen in MI rats treated with an AT₁ receptor blocker in the present study (5, 29).

Myocardial fibrosis and contractile function. It has been postulated that excessive myofibrillar collagen attachments or unfavorable myofibrillar collagen orientation could exert excess loading effects during active as well as passive myocyte displacement (4, 16, 29). Because prevention of myocardial fibrosis was not associated with an improvement in myocardial contractile function, it is plausible that abnormal levels or types of myofibrillar collagen may not significantly affect myocardial contractility. Our study, which was based on measurements of contractile function under isometric conditions, minimized potential mechanical effects secondary to myofibrillar collagen. However, an improvement in myocardial contractile function might have been demonstrated under conditions of isotonic muscle twitch or in isolated contracting myocytes (4).

Limitations. The use of isometric papillary muscle mechanics involves inherent limitations, including central section slippage and core ischemia. We tried to limit slippage due to disruption of the muscle tips by mounting the ends to a special ring system. At least as seen under low microscopic power, disruption and subsequent slippage was minimal and evenly distributed among the treatment groups. Core ischemia was controlled by ensuring that papillary CSA were similar among the treatment groups. The present study does not address potential mechanisms whereby AT₁ receptor blockade improves myocardial function. Although it is possible that upregulation of calcium-handling proteins or myosin heavy chain isoforms might have occurred after treatment, such a demonstration would still not prove whether the effects of AT₁ receptor blockade were direct or indirect (i.e., secondary to improved cardiac loading conditions) or whether the effects were specific for AT₁ receptor blockade. With respect to passive myocardial stiffness, its exact determinants after MI remain unclear. In the future, an experimental design that permits independent analysis of the material properties of the myocyte would be helpful in determining whether alterations in the contractile unit itself after MI contribute to abnormal myocardial stiffness. Finally, it should be pointed out that our study did not address the issue of LV chamber stiffness. It is possible that the contribution of infarct scar to global LV stiffness is important and may be altered by AT₁ receptor blockade.