INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

MURINE MODELS provide unique opportunities to study the underlying mechanisms of the cardiac dysfunction that follows brief periods of ischemia and reperfusion. This so-called phenomenon of myocardial stunning (2) has been partially ascribed to a transient intracellular calcium (Ca²⁺_i) overload during early reperfusion and to a decrease in the responsiveness of the myofilaments to Ca²⁺_i (4, 7, 17). It has been emphasized that activation of a Ca²⁺-calmodulin complex may play an important role in the myocardial injury induced by ischemia and reperfusion (6). Experimental studies have shown that calmodulin antagonists protect the heart against some of the mechanical and metabolic consequences of acute ischemia (1, 6, 13). Genetic modification of Ca²⁺_i- or calmodulin-dependent mechanisms could promote further understanding of their role in the pathogenesis of myocardial ischemia and reperfusion injury (24).

The coronary-perfused heart provides an excellent physiological model to study myocardial stunning. Techniques, however, to describe mouse cardiac physiology are lacking. Grupp et al. (9) have adapted the Langendorff technique to a working mouse heart preparation to compare and quantify the cardiovascular and contractile performance of control and diseased mice hearts. However, changes in vascular resistance, concordant changes in perfusate flow and temperature, and changes in pre- and afterload confound comparison of cardiac muscle mechanics made before and after ischemia and reperfusion in the working heart.

Accordingly, we have developed a technique to perfuse the isovolumic mouse heart at constant flow to characterize myocardial function independently of extramyocardial factors. We have developed a method for loading the bioluminescent Ca²⁺_i indicator aequorin into the isolated mouse heart, which allows us to measure Ca²⁺_i and left ventricular (LV) pressure on a beat-to-beat basis (10). The purpose of this study, then, is to describe a mouse model for investigating changes in Ca²⁺_i and myocardial function during ischemia and reperfusion and to determine the protective effects of the calmodulin antagonist N-(6-aminohexyl)-5-chloro-1-naphthalenesulfonamide (W7) on Ca²⁺_i overload and subsequent stunning.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Preparation of coronary-perfused hearts. Hearts were excised from adult Swiss-Webster mice of either sex that had been anesthetized, weighed, and heparinized (500 U/100 g body wt). Each heart was immediately placed in a preweighed beaker with ice-cold buffer solution. The aorta was slipped over a 20-gauge Luer stub adapter with a stainless steel shaft (Small Parts, Miami Lakes, FL) through which physiological buffer dispensed at a flow rate of 1 ml/min. An incision was made at the root of the pulmonary artery to drain coronary effluent. A constant-flow pump (Masterflex model 7016-20, Cole-Parmer Instruments, Chicago, IL) provided coronary perfusion at a rate of ~3 ml/min. Initial coronary perfusate was composed of (in mM) 118 NaCl, 4.7 KCl, 1.2 KH₂PO₄, 1.5 CaCl₂, 1.2 MgCl₂, 23 NaHCO₃, 10.0 dextrose, and 0.3 EDTA, gassed with 95% O₂-5% CO₂ and adjusted to a pH of 7.4. As shown in Fig. 1, the cannula was connected to a 1-ml syringe shaft that served as a bubble trap and damper immediately upstream of the aortic cannula. To further dampen pump-induced flow oscillations, we connected 100 cm of 0.02-in.-diameter tubing between the pump and the bubble trap. The perfusate temperature was maintained at 30°C to optimize the quality of the aequorin light signals. Coronary perfusion pressure was measured via a Statham P23b transducer (Gould, Cleveland, OH) connected to a sidearm. One platinum pacemaker wire was positioned into the right ventricle, and the other pacing lead to the cannula shaft. The pacing rate was initially set just above the intrinsic heart rate by a Grass model S88 stimulator (Grass Instrument, Quincy, MA).

View larger version (35K): [in this window] [in a new window]   Fig. 1.   Diagram of experimental apparatus. Oxygenated perfusate is pumped through a damper and bubble trap retrogradely to the coronary arteries of the mouse heart. A warming jacket around the bubble trap and the organ bath maintains constant temperature of the perfusate and heart. The bottom of the organ bath is connected to a photomultiplier tube (PMT) via an ellipsoidal reflector. LV, left ventricle.

LV function. A left atrial incision was made to expose the mitral annulus, through which a tiny balloon was passed into the LV. Figure 2 illustrates the construction of an intraventricular balloon-catheter system sufficiently small, yet capable of detecting chamber dynamics at fast heart rates. Briefly, a small square of stretched polyethylene film (Ling Products, Neenah, WI) was enveloped around the tip of a 9-mm-long, narrow-gauge stainless steel tube fixed at one end to a short piece of PE-50 tubing and secured to form a balloon sized such that inflation with 0.03 ml of water was achieved with minimum resistance. The balloon was overinflated to a volume of 0.1 ml and then completely deflated before insertion into the LV. This resulted in a compliant balloon that fills the mouse LV at low pressure. The distal end of the PE-50 catheter-balloon system was connected via a larger catheter (PE-90 tubing, 10 cm long) to another Statham P23b transducer to record intraventricular pressure. The heart so instrumented was then positioned into a small beaker that served as an organ bath. Thebesian venous return drained through the mitral annulus.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Diagram of intraventricular balloon-catheter design for mouse heart. A: small square of polyethylene film (0.012 mm thick) is stretched to decrease thickness without tearing film. B: short, narrow-gauge steel tube is affixed to PE-50 tube. C: film is secured to end of steel tube to form a small balloon. The system is affixed to a PE-90 tube that is connected to LV pressure transducer. The design results in a small compliant balloon that can be easily inserted into the LV chamber through the mitral annulus and provides excellent dynamic response to LV pressure at high heart rates.

Procedure for aequorin loading. We developed a technique for loading aequorin into the isolated mouse heart on the basis of the macroinjection approach previously described in detail (14). Briefly, after 20 min of stable perfusion with the initial coronary perfusate, the buffer solution was replaced by (in mM) 118 NaCl, 4.7 KCl, 1.2 KH₂PO₄, 0.5 CaCl₂, 0.6 MgCl₂, 23 NaHCO₃, 20.0 dextrose, 5.0 pyruvate, and 0.3 EDTA. It has been our observation, in the mouse heart, that aequorin light signals are difficult to obtain unless perfusate Ca²⁺ and Mg²⁺ are reduced. Pacing was discontinued and within 15 min the heart became quiescent. Subsequently, the heart was raised out of the organ bath, and the aequorin loading solution was injected into the interstitium of the epicardium just beneath the epimysium, with a low-resistance glass micropipette. One to three microliters of the aequorin loading solution were injected within 90 s into a localized region of ~2 mm² at the apex of the heart while perfusion was maintained. Care was taken to avoid excessive injection, which could produce dissection and retention of an isolated pocket of aequorin within the tissue. The heart was repositioned into the organ bath such that the aequorin-loaded region was ~2 mm from the bottom of the bath. The Ca²⁺ and Mg²⁺ concentrations of the coronary perfusate were slowly increased to 2.5 mM Ca²⁺ and 1.2 mM Mg²⁺ in a stepwise fashion every 10 min over a period of 60 min for prevention of the Ca²⁺ paradox (12). Pacing was recommenced at 6 Hz.

Light collection and conversion to estimate Ca²⁺_i. Our light collection system from the surface of the mouse heart is the same as previously described (14). Briefly, a part of the coronary perfusion system was positioned in a light-tight box and connected to a photomultiplier tube (9635QA, Thorn EMI, Fairfield, NJ) via an ellipsoid reflector. The bottom of the organ bath was positioned above the reflector (Fig. 1). Aequorin light signals were recorded from the photomultiplier as anodal current. At the end of each experiment, the detergent Triton X-100 (Sigma Chemical, St. Louis, MO) was injected into the coronary perfusate, which quickly permeabilized the myocardial cell membranes and exposed the remaining aequorin to saturated Ca²⁺. This procedure resulted in a rapid burst of light, the integral of which approximated the maximum light (L_max) against which light signals of interest (L) provided the fractional luminescence (L/L_max). The cumulative integrals of each light transient between recording L and L_max were added to L_max to correct for aequorin consumption during the course of the experiment. Fractional luminescence was referred to a calibration equation at 30°C to estimate Ca²⁺_i (14).

Signal recording. After passing though a current discriminator, the light signal was filtered through an analog low-pass window (6 dB at 25 Hz; 8-pole bessel, UAF-41, Burr Brown, Tucson, AZ). The light signal, LV isovolumic pressure, and coronary perfusion pressure were then simultaneously recorded on a magnetic tape recorder (model HR-D8600, JVC, Yokohama, Japan), a four-channel strip recorder (model 35-V704-10, Gould), and a digital oscilloscope (model 4094A, Nicolet Instruments, Madison, WI). Digitized LV pressure and light signals were stored every 0.2 ms on a disk recorder (model XF-44, Nicolet) connected to the oscilloscope. LV pressure recordings were analyzed with regard to developed pressure, end-diastolic pressure, peak positive and peak negative pressure derivatives, and time to 90% pressure decline. Aequorin light signals were analyzed with regard to diastolic and peak systolic Ca²⁺_i.

Drug treatment. After a 15-min equilibrium period, baseline conditions were recorded. Subsequently, the calmodulin antagonist W7 (Sigma) was added to the perfusate to achieve an end concentration of 10 µM. This concentration was selected on the basis of a study of W7 in the rat heart (13). After an additional 15-min equilibrium period with W7 in 13 hearts or control perfusate in 10 hearts, preischemia conditions were recorded.

Ischemia and reperfusion protocol. The heart was then subjected to no-flow global ischemia for 15 min. At the onset of ischemia, the organ bath was evacuated of its oxygenated solution and refilled with nitrogen-saturated perfusate maintained at 30°C. Pacing at 6 Hz was maintained during ischemia. Mean ischemic Ca²⁺_i was calculated as the mean Ca²⁺_i recorded from the 2nd minute through the 14th minute of ischemia. Contracture was defined as an abrupt and sustained rise in intraventricular pressure above 4 mmHg. Contracture time was measured as the time from the onset of ischemia to the onset of contracture. At the end of 15 min of ischemia, the nitrogen-saturated bath was replaced by the original bath maintained at 30°C. Flow was recommenced. Mean Ca²⁺_i during early reflow was calculated as the mean of the peaks of Ca²⁺_i recorded during the first minute of reperfusion. After 20 min of reperfusion, Ca²⁺_i and functional parameters were again measured.

Statistical analysis. Observations before and after drug administration were statistically compared using analysis of variance. When an overall significance was observed, multiple comparisons were performed to determine which comparisons were significant. A value of P < 0.05 was considered significant. Data are expressed as means ± SE.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Baseline conditions and effect of W7. Figure 3 illustrates simultaneous recordings of aequorin light signals and LV pressure in a mouse heart at baseline and during perfusion with the calmodulin antagonist W7. Analyses of the aequorin light signals and pressure waveforms demonstrated no effect of W7 on Ca²⁺_i or LV function at baseline. Table 1 summarizes these results.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Aequorin light signals (top) and isovolumic LV pressure (LVP) waveforms (bottom) from a coronary-perfused mouse heart at baseline conditions (solid lines) and after perfusion with the calmodulin antagonist N-(6-aminohexyl)-5-chloro-1-naphthalenesulfonamide (W7) (dashed lines). In this example, the heart was perfused with 2.5 mM Ca2+ at 30°C and paced at 6 Hz. Estimated peak intracellular Ca2+ (Ca2+i) was 0.75 µM and diastolic Ca2+i was 0.33 µM. Aequorin light signals and pressure waveforms are superimposable, demonstrating that W7 had no effect on Ca2+i and myocardial function at baseline.

Effects of ischemia and reperfusion. Figure 4 illustrates continuous simultaneously recorded aequorin light signals, LV pressure, and coronary perfusion pressure at baseline, during ischemia, and during reperfusion in a control heart and a heart perfused with W7. When coronary flow was interrupted, there was an abrupt fall in LV pressure and an increase in diastolic and peak light. Table 2 summarizes the time to ischemic contracture, the mean Ca²⁺_i during ischemia, and the mean of Ca²⁺_i during early reflow in control hearts and hearts perfused with W7. In control hearts, ischemic contracture occurred within 686 s. W7 either abolished or significantly delayed the onset of contracture to 825 s (Fig. 4 and Table 2). W7 significantly lowered the rise in Ca²⁺_i during ischemia from 0.66 to 0.48 µM. During early reperfusion, there was a marked increase in peak Ca²⁺_i to 1.29 µM in control hearts. W7 significantly lowered this increase to 0.88 µM.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Simultaneous recordings of Ca2+i, isovolumic LVP, and coronary perfusion pressure (CPP) during ischemia and reperfusion in a control heart (A) and a heart perfused with the calmodulin antagonist W7 (B).

Stunning. Fifteen minutes of global ischemia followed by 20 min of reperfusion resulted in prolonged ventricular dysfunction or myocardial stunning. Figure 4 and Table 3 summarize the effects of W7 on Ca²⁺_i and LV function during myocardial stunning 20 min after reperfusion. Peak systolic Ca²⁺_i returned to preischemic levels in control as well as in W7-treated hearts. W7, however, significantly reduced elevated diastolic Ca²⁺_i from 0.43 to 0.28 µM. LV developed pressure recovered to only 55% in control hearts, whereas recovery was significantly improved to 72% in hearts perfused with W7. W7 significantly reduced the rise in LV end-diastolic pressure from 11 to 5 mmHg and significantly improved contractility and relaxation during the late recovery period.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The most important result of our study is that it provides direct quantitative evidence, on a beat-to-beat basis in the coronary-perfused mouse heart, that antagonism of calmodulin-regulated increases in Ca²⁺_i during ischemia and reperfusion correlate with improved myocardial functional recovery. We observe in the mouse heart similar qualitative and quantitative changes in Ca²⁺_i during ischemia and reperfusion, as have been reported for other methods and species (4, 17, 18, 20). During ischemia, Ca²⁺_i progressively increases. Early reperfusion causes an initial overload of Ca²⁺_i with huge transients and elevated diastolic Ca²⁺_i. Although Ca²⁺_i transients gradually return to preischemic levels, LV dysfunction persists through 20 min of reperfusion. In hearts treated with W7, however, the rises in Ca²⁺_i during ischemia and on reperfusion are significantly reduced and the extent of functional recovery significantly improved (Fig. 4).

The results of our study concerning the functional effects of short periods of global ischemia followed by reperfusion on LV pressure development and indexes of relaxation are similar to those previously reported (20, 23). Plumier et al. (23) incorporated an apical strain gauge to demonstrate 50% stunning in the mouse heart after 30 min of ischemia and reperfusion. These investigators, however, may not have removed oxygen from the organ bath during the prolonged ischemic period, which could have promoted recovery during reperfusion at 37°C. We emphasize the importance of replacing the organ bath with nitrogen-saturated solution during ischemia to simulate a hypoxic environment. When the organ bath's oxygen status was unchanged, we noted distinct Ca²⁺_i transients through 15 min of ischemia and accelerated recovery of LV function to baseline, suggesting that although coronary flow had ceased, oxygen from the buffer may have diffused into the myocardium (unpublished observations).

After 20 min of reperfusion, myocardial stunning occurs in the mouse heart despite return of peak Ca²⁺_i to baseline values. These findings are similar to those observed in the rat (20) and ferret heart (4, 16) and further demonstrate that stunning does not result from decreased levels of activator Ca²⁺. Rather, our findings show that murine myocardial stunning results from decreased myofilament responsiveness to Ca²⁺_i. Whether this decreased myofilament responsiveness is attributed to a decrease in maximum Ca²⁺-activated force, a decrease in Ca²⁺ sensitivity, or both, is not answered by the present study. Elevated Ca²⁺_i during ischemia and early reperfusion, however, may activate Ca²⁺-dependent proteases, which could partially degrade the contractile proteins in the stunned myocardium (8). Diastolic Ca²⁺_i remained elevated through 20 min of reperfusion. Such abnormalities in diastolic Ca²⁺_i in stunned myocardium have been previously reported (7, 20). Gao et al. (7) suggested a leaky sarcoplasmic reticulum as the underlying mechanism for the increased diastolic Ca²⁺_i.

It is well established that attenuation of harmful increases in Ca²⁺_i during ischemia and reperfusion improve the extent of myocardial functional recovery (17, 21). It has been emphasized that activation of Ca²⁺-calmodulin-regulated pathways may play an important role in the myocardial injury induced by ischemia and reperfusion (6). Observations that calmodulin antagonists confer cardioprotection from ischemic injury have been ascribed to several mechanisms (13). Calmodulin antagonists have been shown to provide protection through interruption of Ca²⁺ -calmodulin- dependent energy expenditure and preservation of high-energy phosphates during ischemia (6). W7 has been thought to antagonize calmodulin-induced activation of the slow Ca²⁺ channel during ischemia and reperfusion (13). W7, however, is not a highly selective calmodulin antagonist. For example, it has been suggested that calmodulin antagonists may inhibit activation of Na⁺/Ca²⁺ exchange through inhibition of the Ca²⁺-calmodulin-dependent protein kinase C (3), thereby preventing harmful Ca²⁺_i overload (15, 22). It has been shown by Tanaka et al. (25) that W7 inhibits protein kinase C. Kusuoka et al. (15) have elucidated the important role of Na⁺/Ca²⁺ exchange in the mechanism of stunned myocardium. In addition, ischemia-induced degenerative changes in membrane phospholipids may be calmodulin dependent (5).

Our findings of reduced Ca²⁺_i during ischemia and reperfusion, attenuation of ischemic contracture, and improved functional recovery in W7-perfused mouse hearts further underscore a role for calmodulin-regulated mechanisms in the pathogenesis of stunning. Because W7 does not alter LV function and Ca²⁺_i at baseline, cardioprotection cannot be attributed to changes in the inotropic state before ischemia. Ischemic contracture, however, is significantly attenuated by W7, suggesting that the calmodulin antagonist may have promoted preservation of energy stores during ischemia (6). In the presence of W7, the rises in Ca²⁺_i during ischemia and early reperfusion are significantly reduced compared with control hearts, suggesting that the rises in Ca²⁺_i may be mediated by activation of a Ca²⁺-calmodulin complex.

The mouse heart perfusion technique we describe is unique in several ways. First, the design of the balloon facilitates insertion into the LV and allows measurement of chamber dynamics at fast heart rates. The simplicity and reproducibility of the isovolumic preparation increase its utility to systematically assess murine heart muscle physiology. Second, we incorporate a damping system to attenuate detrimental pump-induced pressure oscillations. We observe that the mouse heart tolerates retrograde coronary perfusion above 3 ml/min while perfusion pressure is maintained below 100 mmHg if pump-induced pressure oscillations are minimized. Pump perfusion in the isolated mouse heart is essential because subtle vascular changes can significantly affect flow rate and perfusate temperature, either of which can markedly affect myocardial function. Third, we measure Ca²⁺_i via the macroinjection of the photoprotein aequorin as previously described. Peak systolic Ca²⁺_i in the present study is comparable to that which we have previously reported (10). The diastolic Ca²⁺_i estimates we report in the aequorin-loaded mouse heart are within the range of those previously reported using aequorin in rats (20) and ferrets (4). Moreover, our estimates are similar to values measured using nuclear magnetic resonance spectroscopy in ferret hearts (16) and slightly higher than those reported by others using fluorescent methods (7). To the best of our knowledge, only one other laboratory has published values of Ca²⁺_i in the mouse heart. Ca²⁺_i measured by these investigators in neonatal mouse cardiomyocytes (19) is within the range of our measurements in the intact heart. Ca²⁺_i measurements in the mouse heart, however, should be qualified by the experimental conditions. In the mouse heart, Ca²⁺_i increases steeply as perfusate Ca²⁺ (Ca²⁺_o) is increased through the range 0.5-5.0 mM. Maximal myofilament activation, however, occurs near 2.5 mM Ca²⁺_o (11). Moreover, stimulation rate and temperature have profound effects on diastolic Ca²⁺_i. For example, in one heart perfused at 30°C and paced at 5.5, 6.0, and 6.5 Hz, peak Ca²⁺_i was 0.75, 0.75, and 0.74 µM respectively, whereas diastolic Ca²⁺_i was 0.33, 0.28, and 0.24 µM, respectively. In another heart in which the temperature of the perfusate was increased from 30 to 35°C, diastolic Ca²⁺_i increased from 0.40 to 0.43 µM, but peak Ca²⁺_i remained unchanged. These observations underscore the difficulty in measuring absolute levels of Ca²⁺_i.

We describe a technique in the coronary-perfused mouse heart for simultaneous measurements of LV function and Ca²⁺_i on a beat-to-beat basis at baseline, during ischemia, and after reperfusion. We describe the phenomena of myocardial ischemia and reperfusion in the isolated mouse heart, characterized by a rise in Ca²⁺_i and myocardial contracture during ischemia, Ca²⁺_i overload during early reperfusion, and prolonged ventricular dysfunction, or stunning. Moreover, we demonstrate the effectiveness of the model by showing that harmful increases in Ca²⁺_i during ischemia and reperfusion and the consequent stunning may be regulated by calmodulin. This technique should be useful for exploring how specific gene alterations affect Ca²⁺-regulated mechanisms in the heart.