INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE MAJOR DETERMINANTS OF cardiac myocyte contractility are 1) Ca²⁺ influx, 2) Ca²⁺ efflux and sequestration, 3) sarcoplasmic reticulum (SR) Ca²⁺ content, 4) myofilament Ca²⁺ sensitivity, and 5) myosin heavy chain isoenzyme distribution pattern. With respect to contractile dysfunction after myocardial infarction (MI), previous studies by our laboratory (4, 31-34, 37, 38) and others (8, 9, 13-18) have identified several steps in excitation-contraction coupling that were altered in myocytes that have survived MI. With the focus on contraction abnormalities, surviving myocytes isolated from hearts post-MI demonstrated an increase (16, 33), decrease (4, 13, 15, 18, 33), or no change (1, 4, 16, 33) in myocyte contraction amplitudes. The discrepancies in results reported by different investigators may be because of differences in animal species, infarct sizes, experimental conditions, and extent of left ventricular remolding post-MI. In a previous study, our laboratory characterized contraction abnormalities in myocytes isolated from rat hearts 3 wk post-MI (33). To summarize, under conditions that preferentially favored Ca²⁺ efflux over influx {low extracellular Ca²⁺ concentration ([Ca²⁺]_o)}, MI myocytes shortened more than those isolated from rats that had received sham operations. Conversely, under conditions that favored Ca²⁺ influx (high [Ca²⁺]_o), sham myocytes shortened to a greater extent than MI myocytes. At intermediate [Ca²⁺]_o, differences in steady-state contractile amplitudes between sham and MI myocytes were no longer significant. This abnormal pattern of contractile behavior observed in MI myocytes is consistent with our hypothesis that both Ca²⁺ influx and efflux pathways were subnormal in MI myocytes and that they contributed to abnormal cellular contractile behavior. Of the major Ca²⁺ influx and efflux pathways in our 3-wk rat MI model, only Na⁺/Ca²⁺ exchange (NCX1) (37), but not L-type Ca²⁺ current (32), was depressed in MI myocytes. Because we have recently demonstrated that it was possible to both up- and downregulate NCX1 in adult rat myocytes (24, 36), the present study was undertaken to test the hypothesis that overexpression of rat NCX1 in MI myocytes would restore contractile abnormalities toward normal.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animal preparation. To induce MI, the left main coronary artery of each anesthetized (2% isoflurane, 98% O₂), intubated, and ventilated male Sprague-Dawley rat (weight ~300 g) was ligated 3-5 mm distal to its origin from the ascending aorta (4, 30-35, 37, 38). Sham operation was identical, except that the coronary artery was not occluded. All surviving rats received rat chow and water at libitum and were maintained on a 12:12-h light-dark cycle. Survivors typically had 36 ± 3% of myocardium infarcted as determined histologically (4). In addition, despite no overt signs of congestive heart failure (edema, ascites) in MI rats, infarcted hearts had 20% lower left ventricular systolic pressure at both 1 and 3 wk after infarction (4). Three weeks after MI, survivors and sham-operated rats were anesthetized with pentobarbital sodium (35 mg/kg body wt ip), and their hearts were excised for myocyte isolation. Our previous studies indicate that, 3 wk after MI, myocyte adaptation included cellular hypertrophy as reflected by an ~10% increase in cell length (4, 33) and by a 13-15% increase in whole cell capacitance (32, 37), reduced dihydropyridine binding sites (32), altered intracellular Ca²⁺ concentration ([Ca²⁺]_i) transients (4, 31, 34) and contractile activities (4, 33), decreased SR Ca²⁺ uptake and sarco(endo)plasmic reticulum Ca²⁺-ATPase (SERCA2) expression (34), depressed NCX1 currents and SR Ca²⁺ contents (37), and attenuated response to -adrenergic agonists (31, 32).

Myocyte isolation and culture. Cardiac myocytes were isolated from the septum and left ventricular free wall of rat hearts by successive perfusion with collagenase and hyaluronidase (5). Freshly isolated myocytes were seeded on laminin-coated coverslips (6), and a portion was used within 2 h of isolation for contractility measurements. The remaining myocytes were cultured in modified, serum-free medium 199 ([Ca²⁺]_o = 1.8 mM) as described in detail previously (23, 24, 36). After 2 h, media were changed to remove nonadherent myocytes. The myocytes were incubated for an additional 3-4 h before initiation of pacing (1-Hz, 5-ms pulses of alternating polarity, field strength of 4 V/cm) as described previously (23, 24, 36). Culture media were changed daily. Our laboratory (23) and others (2) have previously demonstrated that continuous pacing of adult rat cardiac myocytes in culture maintained normal contractile function for at least 72 h.

Adenoviral infection of cardiac myocytes. Recombinant, replication-deficient adenovirus containing both green fluorescent protein (GFP) and rat NCX1 [each under a separate cytomegalovirus promoter] (Adv-GFP-NCX1) was constructed as described by our laboratory previously (36). Two hours after isolation, myocytes seeded in four-well trays (Nuclone) were infected with either Adv-GFP-NCX1 or adenovirus expressing GFP alone (Adv-GFP) at a multiplicity of infection of 1 for 3 h. Medium was then changed, and myocytes were studied after 6, 24, and 48 h in continued pacing culture. Over 95% of myocytes fluoresced green (excitation 478 nm, emission 535 nm) within 6 h, indicating successful adenoviral infection and GFP expression. For brevity, MI myocytes infected with Adv-GFP and Adv-GFP-NCX1 are referred to as MI-GFP and MI-NCX1 myocytes, respectively.

Myocyte shortening measurements. Cell contraction dynamics were measured in myocytes incubated in HEPES-buffered (20 mM, pH 7.4) medium 199 (37°C), containing either 0.6, 1.8, or 5.0 mM [Ca²⁺]_o, by using a charged-coupled device video camera (Ionoptix; Milton, MA) and edge-detection software (Ionoptix) as described previously (23, 24, 30, 33, 36). For calibration of pixels vs. micrometer, a high-resolution test target (model 22-8635, Ealing Electro-Optics; Natick, MA) was used.

[Ca²+]_i transient measurements. Myocytes were exposed to 0.67 µM of fura 2-AM for 15 min at 37°C (6). Fura 2 epifluorescence (excitation 360 and 380 ± 10-nm band pass; emission 510 ± 18-nm band pass) from myocytes (37°C) was measured with our quantitative fluorescence microscopy system (Olympus DApo ultraviolet ×40/1.30 numerical aperture objective) described previously (23, 24, 31, 34, 36). Background and cellular autofluorescence measured in MI-GFP or MI-NCX1 myocytes not loaded with fura 2 accounted for <10% of the fura 2 signal (36). Intracellular fura 2 fluorescence was calibrated daily for each batch of myocytes, as previously described (6). [Ca²⁺]_i transient data were derived from fura 2 signals, using 224 nM as the Ca²⁺-fura 2 dissociation constant, and analyzed with custom-written software (Ionoptix).

NCX1 and calsequestrin immunoblotting. Cultured myocytes were rinsed three times with ice-cold phosphate-buffered saline, then scraped into ice-cold lysis buffer as described previously (23, 24, 35, 36). The cell lysate was snap frozen with dry ice ethanol and stored at 80°C. Myocyte lysates (16 µg each lane) in SDS sample buffer containing 10 mM N-ethylmaleimide were applied to 7.5% polyacrylamide gel, and proteins were separated by electrophoresis (23, 24, 35, 36). After protein transfer onto Immunblot polyvinylidene difluoride membranes, NCX1 was detected with rabbit anti-NCX1 antibody (1:1,000 dilution; 11-13 Swant, Bellinzona, Switzerland), and donkey anti-rabbit antibody (1:2,000; Amersham, Buckinghamshire, UK) was used as the secondary antibody. Under nonreducing conditions, NCX1 was detected as a single band of apparent molecular mass of 160 kDa (23, 24, 35, 36). For calsequestrin immunoblotting, membranes stripped of NCX1 antibodies were sequentially exposed to rabbit anti-calsequestrin antibody (1:2,500; Swant) and donkey anti-rabbit IgG (1:5,000; Amersham). Our laboratory has previously used 11-13 and anti-calsequestrin antibodies to successfully detect NCX1 and calsequestrin, respectively (23, 24, 35, 36). Immunoreactive proteins were detected with the enhanced chemiluminescense-Western blotting system (Amersham). We quantified protein band-signal intensities by scanning autoradiograms of the blots with a phosphorimager (Molecular Dynamics, Sunnyvale, CA).

Statistics. All results are expressed as means ± SE. In experiments in which maximal contraction amplitudes were measured as functions of experimental group (e.g., sham vs. MI), [Ca²⁺]_o, and days in culture, three-way ANOVA was performed to determine significance of difference. A linear model-fitted standard least-squares analysis (JMP version 4, SAS Institutes, Cary, NC) was used. For analysis of a parameter (e.g., maximal shortening velocity) as a function of a group (MI-GFP vs. MI-NCX1) and [Ca²⁺]_o, two-way ANOVA was used to determine statistical significance. For comparison of NCX1 and calsequestrin abundance between MI-GFP and MI-NCX1 myocytes, Student's paired t-test was used. In all analyses, P  0.05 was taken to be statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effects of continual pacing culture on sham and MI myocyte contractility. We have previously characterized contraction abnormalities in 3-wk MI myocytes (4, 33). To summarize, compared with sham myocytes, steady-state contraction amplitudes of MI myocytes were higher at 0.6, similar at 1.8, and lower at 5.0 mM [Ca²⁺]_o. In the present series of experiments, differences in contraction amplitudes between sham and MI myocytes were confirmed in freshly isolated (day 0) cells (Table 1). In addition, with continual pacing culture (23, 24, 36), both sham and MI myocytes preserved their contractile performance at 48 h of culture (Table 1) in contrast to adult rat myocytes cultured under quiescent conditions that exhibited progressive deterioration in cell shortening (2, 29). More importantly, at 48 h of continual pacing culture, MI myocytes maintained their phenotypic differences from sham myocytes, i.e., they shortened less at 5 mM [Ca²⁺]_o and more at 0.6 mM [Ca²⁺]_o (Table 1). This conclusion is supported by three-way ANOVA {group (sham vs. MI), [Ca²⁺]_o, day in culture} that indicated significant group (P < 0.05), [Ca²⁺]_o (P < 0.0001), and group × [Ca²⁺]_o interaction (P < 0.0001) effects, indicating that the magnitude and/or direction of the effects of [Ca²⁺]_o on cell shortening was different between sham and MI myocytes. Our important observation that MI myocyte shortening characteristics were different than those of sham myocytes even after 2 days of culture indicates that, for the first time, there is an in vitro-cultured MI myocyte model system that allows gene manipulation.

Effects of adenoviral infection on cultured Sham and MI myocyte contractility. We have previously demonstrated the feasibility of upregulating NCX1 in adult rat myocytes by adenovirus-mediated gene transfer (36). In the present study, it was first necessary to demonstrate that adenoviral infection did not affect contractile amplitudes in MI myocytes. Compared with uninfected MI myocytes (open symbols, Fig. 1), adenovirus-infected MI myocytes (solid symbols, Fig. 1) at 6 (day 0), 24, and 48 h of continual pacing culture had similar cell-shortening amplitudes, regardless of whether cell shortening was measured at 0.6, 1.8, or 5.0 mM [Ca²⁺]_o (Fig. 1). Three-way ANOVA (group, [Ca²⁺]_o, and day) indicated insignificant group (Adv-GFP vs. +Adv-GFP, P > 0.8) and day (P > 0.1) but significant Ca²⁺ concentration (P < 0.0001) effects. In addition, the group × [Ca²⁺]_o × day interaction effect was also insignificant (P > 0.8).

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Effects of adenovirus infection on contractility in myocardial infarction (MI) rat myocytes. Left ventricular and septal myocytes isolated from 3-wk post-MI rats were infected with recombinant adenovirus expressing green fluorescent protein (filled symbols) and cultured at 1.8 mM extracellular Ca2+ concentration ([Ca2+]o) for 48 h under continuous pacing (1 Hz) conditions (23, 24, 36). MI myocytes not infected with adenovirus (open symbols) served as controls. Myocyte contraction was measured 6 (day 0), 24, and 48 h after adenovirus infection. Myocytes were paced (1 Hz) to contract at 37°C and an [Ca2+]o of 0.6 (circles), 1.8 (triangles), and 5.0 (squares) mM. Values are means ± SE of n = 15-51 myocytes. Error bars are not shown if they fall within the boundaries of the symbol. Statistical results are given in RESULTS.

Effects of Adv-GFP-NCX1 infection on NCX1 abundance in MI myocytes. Forty-eight hours after infection with either Adv-GFP or Adv-GFP-NCX1, MI myocyte lysates were collected to analyze for NCX1 abundance by immunoblotting (23, 24, 35, 36). Compared with MI-GFP myocytes, MI-NCX1 myocytes had significantly (P < 0.05) more NCX1 protein (45.1 ± 11.7 vs. 30.2 ± 6.6 arbitrary units; n = 5) (Fig. 2). By contrast, there were no differences (P < 0.17) in calsequestrin amounts between MI-GFP (22.6 ± 1.6 arbitrary units) and MI-NCX1 (24.2 ± 1.0 arbitrary units) myocytes (Fig. 2).

View larger version (11K): [in this window] [in a new window]   Fig. 2.   Immunoblots of Na+/Ca2+ exchanger and calsequestrin. Proteins in MI myocyte homogenates (16 µg/lane) were separated by gel electrophoresis and transferred to immunoblot polyvinylidene difluoride membranes, and Na+/Ca2+ exchanger and calsequestrin were identified by immunoblotting as described in METHODS. Lanes 1 and 2 are paired samples from the same isolated myocyte preparation, whereas lanes 3 and 4 are paired samples from another myocyte preparation. GFP, green fluorescent protein; NCX1, adenovirus expressing both GFP and Na+/Ca2+ exchanger. Composite results are presented in RESULTS. Numbers on left refer to apparent molecular mass.

Effects of NCX1 overexpression on MI myocyte contractile function. After 48 h of adenoviral infection, overexpression of NCX1 reduced contraction amplitudes in MI myocytes stimulated at 0.6 mM [Ca²⁺]_o (Fig. 3, A and B; Table 3). By contrast, at 5.0 mM [Ca²⁺]_o, MI-NCX1 myocytes contracted more than MI-GFP myocytes (Fig. 3, E and F; Table 3). At 1.8 mM [Ca²⁺]_o, there were no differences in contraction amplitudes between MI-NCX1 and MI-GFP myocytes (Fig. 3, C and D; Table 3). These conclusions are supported by highly significant group × [Ca²⁺]_o interaction effects (P < 0.0001). Thus NCX1 overexpression ameliorated the abnormal contraction pattern observed in MI myocytes. Indeed, when one compares contraction amplitudes between Adv-GFP-sham (Table 2) and MI-NCX1 myocytes (Table 3) measured after 48 h of adenoviral infection, there were no significant differences detected (group effect, P > 0.8; group × [Ca²⁺]_o interaction effect, P > 0.05).

View larger version (32K): [in this window] [in a new window]   Fig. 3.   Overexpression of Na+/Ca2+ exchanger rescues contractile abnormalities in adult rat myocytes isolated from 3-wk postinfarction hearts. Myocytes isolated from 3-wk postinfarction rat hearts were infected with recombinant adenovirus expressing either GFP (MI-GFP) or both GFP and Na+/Ca2+ exchanger (MI-NCX1) and then cultured for 48 h under continuous pacing (1 Hz) conditions. For contraction studies, cultured myocytes were paced to contract at 37°C and a [Ca2+]o of 0.6, 1.8, and 5.0 mM. Shown are steady-state paced twitches from MI-GFP (A, C, E) and MI-NCX1 (B, D, F) myocytes. Results are summarized in Table 3.

Effects of NCX1 overexpression on [Ca²+]_i transients in MI myocytes. As a group, end-diastolic [Ca²⁺]_i levels in MI-NCX1 myocytes paced at 1 Hz were significantly lower than those in MI-GFP myocytes (Table 4; group effect, P = 0.0003) across the range of [Ca²⁺]_o examined (group × [Ca²⁺]_o effect, P > 0.3). Varying [Ca²⁺]_o had no effect on diastolic [Ca²⁺]_i in both groups ([Ca²⁺]_o effect, P > 0.6).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The first major finding of the present study is that myocytes isolated from post-MI hearts, compared with sham myocytes, maintained their phenotypic differences in terms of contractile abnormalities and cell length increases even after 48 h of culture (Table 1). This observation demonstrates, for the first time, an in vitro-cultured MI myocyte model system suitable for short-term manipulation of gene expression. In addition, the differences in contraction amplitudes between sham and MI myocytes were not affected by adenovirus infection (Tables 1 and 2), indicating that it is possible to utilize the highly efficient adenovirus-mediated gene transfer technique in MI myocytes to study the effects on myocyte contractile performance.

NCX1 mediates both Ca²⁺ influx and efflux during a cardiac cycle (24-27, 36). At physiological [Ca²⁺]_o and at rest, it is generally accepted that NCX1 primarily functions in the Ca²⁺ efflux mode (3 Na⁺ in, 1 Ca²⁺ out). During systole, when the resting membrane potential exceeds the equilibrium potential of NCX1, Ca²⁺ influx is favored (27). The extent and duration of Ca²⁺ influx mediated by NCX1 is species-dependent, most likely due to differences in intracellular Na⁺ concentration and action potential morphology among species (26, 27). For example, in rat myocytes, the relatively higher intracellular Na⁺ concentration (compared with rabbit myocytes) would bias the NCX1 to the Ca²⁺ influx mode during the action potential most of the time (26). Alterations in NCX1 function and/or abundance would thus be expected to affect both Ca²⁺ influx and efflux during the cardiac cycle.

In rat myocytes studied 3-9 wk postinfarction, both Na⁺-dependent Ca²⁺ uptake in sarcolemmal vesicles (8) and whole cell NCX1 current (I_NaCa) (37), but not amounts of NCX1 protein (35), were depressed. In post-MI rabbit myocytes, NCX1 activity was reported to be increased (16) or decreased (21). In canine subepicardial cells isolated from 5-day infarcted hearts, I_NaCa was not different compared with noninfarcted hearts (20). The relationship between depressed NCX1 activity and abnormal contractile function in post-MI rat myocytes was suggested from the following observations. First, contractile abnormalities in post-MI rat myocytes (Tables 1 and 2; Ref. 33) mimicked those observed in normal rat myocytes in which NCX1 was downregulated by antisense exposure (24). Specifically, both post-MI and NCX1-downregulated rat myocytes, compared with their respective controls, had higher contraction amplitudes at 0.6 mM [Ca²⁺]_o but lower contractility at 5 mM [Ca²⁺]_o (24, 33). Second, a 6- to 8-wk program of high-intensity sprint training instituted 3 wk after MI was effective in increasing the depressed I_NaCa (35) and ameliorating the contractile abnormalities (30), thereby providing circumstantial evidence that enhancing NCX1 activity might lead to improvement in MI myocyte contractile function. Thus a second major finding of the present study is that increased NCX1 activity in rat MI myocytes by overexpression of NCX1 (Fig. 2) directly led to restoration of contraction abnormalities toward normal (Fig. 3; Table 3). Specifically, the "supranormal" contraction amplitude in MI myocytes at 0.6 mM [Ca²⁺]_o (33) was reduced back toward normal (Fig. 3, A and B; Table 3), and the diminished contraction amplitude at 5 mM [Ca²⁺]_o (4, 33) was enhanced to levels observed in sham myocytes (Fig. 3, E and F; Table 3).

Improvement of contractile function by overexpressing NCX1 in MI myocytes was most likely due to restoration of [Ca²⁺]_i homeostasis during excitation-contraction toward normal. First, end-diastolic [Ca²⁺]_i was significantly lower in MI myocytes overexpressing NCX1 (Table 4), suggesting enhanced forward NCX1 activity and improved diastolic compliance (19). In addition, under conditions that favored Ca²⁺ efflux, lower systolic [Ca²⁺]_i and [Ca²⁺]_i transient amplitudes (Table 4) in MI-NCX1 myocytes are consistent with increased Ca²⁺ efflux via NCX1. Conversely, under conditions that favored Ca²⁺ influx, increased NCX1 would lead to more Ca²⁺ influx during systole, resulting in higher SR Ca²⁺ content and larger [Ca²⁺]_i transient (Table 4) and contraction amplitudes (Table 3) in MI-NCX1 myocytes.

We have previously demonstrated that changing NCX1 amounts in normal adult rat myocytes by overexpression or antisense treatment had no effects on other important parameters in excitation-contraction coupling such as L-type Ca²⁺ current, action potential morphology, SR Ca²⁺ uptake activity, and SERCA2 and calsequestrin amounts (24, 36). In addition, in transgenic murine myocytes in which compensatory responses are more likely to occur, NCX1 overexpression had no effect on peak L-type Ca²⁺ current density (28), phospholamban, and SERCA2 and calsequestrin levels (25). Thus when NCX1 amounts in our short-term culture model were altered by gene transfer, it was unlikely to significantly perturb other important Ca²⁺ homeostatic pathways. The apparent differences in t_1/2 of [Ca²⁺]_i decline between MI-GFP and MI-NCX1 myocytes (Table 4) may well relate to the differences in [Ca²⁺]_i-transient amplitudes and not necessarily to any inherent differences in SR Ca²⁺ transport properties (3).

In failing human myocardium from dilated cardiomyopathy and ischemic cardiomyopathy, NCX1 protein was reported to be either increased (11) or unchanged (11, 12). Whether the activity of NCX1 was altered in human congestive heart failure, compared with that in normal human myocardium, is unclear at present. Recent studies in failing human ventricular myocytes, however, clearly demonstrated that NCX1 (especially in the 3 Na⁺ out-1 Ca²⁺ in mode) contributed importantly to [Ca²⁺]_i transient, contraction, and relaxation (7, 10). Therefore, increases in NCX1 activity in failing human myocardium, if it indeed occurs, may be viewed as an important compensatory mechanism for decreased contractile function in heart failure myocytes.

In this study we focused on the role of NCX1 on contractile abnormalities in post-MI rat myocytes. Our laboratory and others (4, 8, 9, 13-18, 31-34, 37, 38) have shown that there are many steps involved in excitation-contraction coupling that are altered in post-MI myocytes. The importance, or lack thereof, of alterations in these other pathways (e.g., action potential morphology, SR Ca²⁺ uptake, myosin heavy chain isoenzyme distribution, etc.) in causing contractile dysfunction post-MI has not been addressed in the present study. To dissect the importance of each pathway involved in excitation-contraction coupling in causing contractile abnormalities post-MI will require a systematic and comprehensive study, which is well beyond the scope of the present report. Another limitation of the present study is that we increased NCX1 activity by overexpressing NCX1. It is known that despite decreases in NCX1 activity in post-MI rat myocytes (8, 37), the amounts of NCX1 protein were not different between sham and MI myocytes (35).

In summary, we have established an in vitro MI myocyte culture model system in which contractile dysfunction (as compared with sham myocytes) was maintained for at least 48 h after myocyte isolation. Infection with adenovirus did not affect the phenotypic differences between sham and MI myocytes. Contractile and [Ca²⁺]_i-transient abnormalities observed in MI myocytes were restored toward normal by overexpression of NCX1. We speculate that decreased NCX1 activity may play a significant role in contractile dysfunction in postinfarction myocytes.