The possible mechanism by which interleukin-1β (IL-1β) affects β-adrenergic responsiveness of L-type Ca2+ current (I Ca,L) was examined in adult rat ventricular myocytes by use of whole cell patch-clamp techniques. In the presence of isoproterenol (Iso), exposure for 3 min to IL-1β suppressed the Iso-activatedI Ca,L. In the presence of IL-1β, the response ofI Ca,L to Iso was decreased, and the EC50 for Iso stimulation was increased. However, IL-1β had no effect on [3H]CGP-12177 binding, displacement of [3H]CGP-12177 binding by Iso, or on basal and Iso-enhanced cAMP content. WhenI Ca,L was activated by extracellular application of forskolin or 8-(4-chlorophenylthio)-cAMP, a membrane-permeable cAMP analog, or by intracellular dialysis with cAMP, IL-1β had little effect onI Ca,L. In contrast, in the presence of cAMP, IL-1β still suppressed the Iso-enhancedI Ca,L. These results show that the IL-1β-induced decrease in β-adrenergic responsiveness ofI Ca,L does not result from inhibition of β-adrenoceptor binding, adenylyl cyclase activity, or cAMP-mediated pathways, suggesting a cAMP-independent mechanism.
- calcium channel
- signal transduction
- cardiac myocytes
interleukin-1β (IL-1β), a 17-kDa proinflammatory cytokine, has been closely associated with immune- and injury-mediated changes in cardiovascular function (8, 9, 25, 33). Marked increases in plasma IL-1β concentration are observed during the cardiac dysfunction associated with myocardial infarction (13, 29), ischemia-reperfusion (11, 15), myocarditis (6), acute septic cardiomyopathy (27), and allograft rejection (16). Studies using PCR techniques show that mRNAs for IL-1β and its receptor are expressed in endomyocardium of patients with inflammatory myocarditis (18) and dilated cardiomyopathy (32) and in hearts with acute viral myocarditis (24). The enhanced expression of IL-1β mRNAs was shown to occur primarily in ventricular myocytes (24). Moreover, these cardiac disorders are associated with an increased sympathetic nervous system activity (22) and altered adrenergic responsiveness of myocardial function (1, 6, 12, 14). The interaction between the increased IL-1β and the enhanced sympathetic tone under these pathophysiological conditions remains unclear.
The direct autocrine and/or paracrine effects of IL-1β on ventricular cell function include decreases in basal L-type Ca2+ channel current (I Ca,L) (26) and contractility (10, 20, 36). In addition, in neonatal rat cardiac myocytes, IL-1β decreases the β-adrenergic responsiveness of contractility by reducing isoproterenol (Iso)-enhanced cAMP levels after a 72-h exposure (14). Studies in adult guinea pig ventricular myocytes suggest that preincubation with IL-1β for 1–5 h inhibits the β-adrenergic control ofI Ca,L via activation of nitric oxide synthase (NOS) (31). These data demonstrate a delayed effect of IL-1β on myocardial β-adrenergic responsiveness. However, available data have not addressed the possibility that IL-1β has an acute effect on adult ventricular myocyte responsiveness to β-adrenergic stimulation.
In the present study we have examined the acute effect of IL-1β on β-adrenergic receptor binding and on intracellular cAMP content andI Ca,L in adult ventricular myocytes. We show that IL-1β decreases the β-adrenergic responsiveness ofI Ca,L primarily via a cAMP-independent pathway.
Single adult ventricular myocytes were isolated from the hearts of male Sprague-Dawley rats (250–300 g) with use of protocols described previously (26). Briefly, hearts were rapidly excised and perfused at 37°C via the aorta with an oxygenated control buffer solution consisting of (in mM) 110 NaCl, 3.8 KCl, 1.2 KH2PO4, 1.2 MgSO4, 25 NaHCO3, 0.2 CaCl2, and 11 glucose (pH 7.4 in 95% O2-5% CO2 at 37°C) and for another 5 min with Ca2+-free buffer solution. Hearts were then perfused for 20 min with a buffer solution containing 25 μM CaCl2 plus 0.5 mg/ml collagenase. The ventricles were removed, minced, rinsed with control buffer solution, and shaken in a water bath at 37°C for two to three periods of 10 min each. Isolated ventricular myocytes were then plated into 60-mm culture dishes (Falcon) containing antibiotic-free, bicarbonate-buffered culture medium 199 (60%; GIBCO, Grand Island, NY) with 36% Earle’s balanced salt solution composed of (mM) 116 NaCl, 4.7 KCl, 0.9 NaH2PO4, 0.8 MgSO4, 26 NaHCO3, and 5.6 glucose and 4% fetal bovine serum (GIBCO; pH 7.4 in 5% CO2-95% air at 37°C).
Ventricular myocytes were placed on the heated stage of an inverted microscope (Nikon Diaphot) and perfused with a normal Tyrode solution consisting of (in mM) 145 NaCl, 5.4 KCl, 0.8 MgCl2, 1.0 CaCl2, 5.6 glucose, 5.8 HEPES, and 4.2 Tris base (pH 7.4 at 37°C). Cells were patch clamped in the whole cell configuration by conventional techniques (17) with use of a patch-clamp amplifier (Axopatch 200A, Axon Instruments, Foster City, CA), as previously described (26). Briefly, patch electrodes were filled with a pipette solution; tip resistance was 2–5 MΩ. Recorded currents were filtered at 1–2 kHz through a four-pole low-pass Bessel filter and sampled at 5 kHz with a PC/AT computer using pCLAMP 6.03 software (Axon Instruments) through an Axon Digidata 2000A acquisition system.
Measurement ofI Ca,L has been described in our previous studies (26). The pipette solution for experiments measuringI Ca,L consisted of (in mM) 100 CsOH, 70 aspartic acid, 11 CsCl, 15 tetraethylammonium chloride, 2 MgCl2, 5 MgATP, 10 EGTA, 0.1 CaCl2, 5 pyruvic acid, 5.6 glucose, 5 Tris2- phosphocreatine, 0.4 Li4GTP, and 10 HEPES-Tris base (pH 7.2 at 37°C). Myocytes were voltage clamped at −70 mV when the normal Tyrode solution was switched to an external solution consisting of (in mM) 140N-methyl-d-glucamine chloride, 2 CaCl2, 0.8 MgCl2, 2 4-aminopyridine, and 10 HEPES-Tris base (pH 7.40 at 37°C). These conditions eliminated most membrane currents associated with Na+ and K+. After formation of the whole cell configuration,I Ca,L was elicited by a single 250-ms pulse to +10 mV from the holding potential once every 15 s. The peak current-voltage relationship ofI Ca,L was constructed by applying 250-ms voltage pulses to potentials between −60 and +70 mV in 10-mV increments from the holding potential of −70 mV at 0.1 Hz. The magnitude ofI Ca,L was defined by the difference between the peak current and that at the end of the 250-ms pulse. All experiments were carried out at 37°C.
β-Adrenoceptor binding was performed using partially purified membranes that were prepared at 4°C. Ventricular muscle from 8–10 rats was pooled, suspended in 50 mM Tris, 2 mM MgSO4, 0.1 mM dithiothreitol, and 0.1 mM phenylmethylsulfonyl fluoride (pH 7.4), and initially homogenized for 30 s using a Polytron at a setting of 6. Further homogenization was achieved using seven strokes of a Dounce homogenizer. The homogenate was centrifuged at 800g for 20 min, and the supernatant was subsequently centrifuged at 2,500 gfor 20 min. The resultant supernatant was subjected to two sequential centrifugations at 30,000 g for 20 min with use of the homogenizing buffer to wash the pellet between centrifugations. The final pellet, a partially purified membrane preparation, was resuspended in a reaction solution (50 mM Tris and 2 mM MgSO4, pH 7.4) and stored at −80°C. Protein concentrations were determined by the method of Bradford (3), with BSA as the standard.
Radioligand binding to membrane preparations.
Binding assays with [3H]CGP-12177 (42.5 Ci/mmol; New England Nuclear Research Products, Boston, MA) were performed in polypropylene tubes containing reaction solutions in the presence or absence of 4 ng/ml IL-1β. In saturation experiments, final concentrations of the β-adrenergic antagonist [3H]CGP-12177 ranged from ∼0.02 to 10 nM. In competition experiments, increasing levels of nonlabeled Iso were added to reaction solutions containing 1.0 nM [3H]CGP-12177 in the presence or absence of 100 μM 5′-guanylyl imidodiphosphate guanosine (GppNHp). Nadolol (10 μM final concentration) was added to a parallel set of tubes to estimate nonspecific binding in all experiments. The reaction was initiated by addition of membrane protein to assay tubes, and the contents were incubated at 37°C for 30 min. Bound and free [3H]CGP-12177 were separated by filtration through GF/C filters, which were washed three times with ice-cold reaction solution. Filters were then immersed in scintillation fluid, and the retained radioactivity was determined by liquid scintillation spectrometry. Data were analyzed using a microcomputer version of LIGAND (28).
Radioligand binding to ventricular myocytes.
Effects of IL-1β on β-adrenoceptor binding were also monitored in intact ventricular myocytes (0.5 × 106 cells/ml normal Tyrode solution containing 0.5 nM [3H]CGP-12177) with and without 4 ng/ml IL-1β. Iso (100 nM) was included in some tubes, and nonspecific binding was determined using 10−5 M nadolol. After 30 min of incubation at 37°C, bound and free [3H]CGP-12177 were separated by filtration through GF/C filters, which were washed three times with ice-cold Tyrode solution. Radioactivity was determined as described above. Data were analyzed using a microcomputer version of LIGAND (28).
Chemicals and solutions.
Most reagents were purchased from Sigma Chemical (St. Louis, MO). Nucleotides were directly added to pipette or bath solutions. The stock solution of human recombinant IL-1β [106 U/ml (Promega, Madison, WI) and 5 μg/ml (R & D Systems, Minneapolis, MN)] was made in the normal Tyrode solution containing 0.1% BSA. The effect onI Ca,L of the final concentration of 1,000 U/ml IL-1β (or 4 ng/ml) from Promega is equivalent to that of 5 ng/ml IL-1β from R & D Systems.
Values are means ± SE. Statistical significance was evaluated by the two-tailed Student’s t-test or, when more than two conditions are compared, by one-way ANOVA with Duncan’s multiple range test. Differences withP < 0.05 were considered significant.
Suppression of β-adrenergic responsiveness of ICa,L by IL-1β in adult rat ventricular myocytes.
Initial experiments were designed to examine the effect of IL-1β onI Ca,L during stimulation by Iso. In Fig.1 A, exposure of a myocyte to 50 nM Iso caused an 80% increase inI Ca,L that was further enhanced (∼60%) by subsequent exposure to 1 μM Iso. In the presence of 1 μM ISO a 3-min exposure to 0.4 ng/ml IL-1β resulted in an ∼30% decrease in peakI Ca,L. TheI Ca,L recovered after removal of IL-1β and return to 50 nM Iso. Similar experiments show that, in the presence of 1 μM Iso, 0.4 and 4 ng/ml, IL-1β decreased Iso-activatedI Ca,L by 33.2 ± 4.7% (n = 6) and 43.1 ± 1.4% (n = 5; see also Fig.6 A), respectively. The 3-min exposure duration was chosen because, during this period of time, desensitization of β-adrenergic responsiveness was minimal (∼7%; Fig. 1 B). Figure1 B shows the temporal change inI Ca,L in the presence and absence of IL-1β. The result indicates that the IL-1β-induced decrease in Iso-activatedI Ca,L was greater than that observed in the continued presence of Iso.
Figure 2 demonstrates the effect of IL-1β on β-adrenergic responsiveness ofI Ca,L in adult rat ventricular myocytes. Figure 2 Ashows results from a myocyte that was pretreated with 0.4 ng/ml IL-1β for 3 min before exposure to increasing concentrations of Iso (1 nM–1 μM). In the presence of 0.4 ng/ml IL-1β, 1 μM Iso induced a 58% increase inI Ca,L, a level less than that observed in the absence of IL-1β. The maximal stimulation ofI Ca,L by 1 μM Iso was 192 ± 8% (n = 35) and 150 ± 3% (n = 24) of control in the absence and presence of 0.4 ng/ml IL-1β, respectively. Figure 2 B shows the concentration-dependent effects of Iso onI Ca,L by plotting relative peakI Ca,L (as measured at +10 mV) to the value at 1 μM Iso vs. Iso concentrations in the absence and presence of 0.4 ng/ml IL-1β. IL-1β caused a rightward shift in the dose-response curve, with the EC50 for Iso being increased from 50.5 to 156 nM. Thus IL-1β reduced the maximal stimulation (efficacy) and the potency of Iso.
Effect of IL-1β on cardiac β-adrenergic receptor binding.
In attempts to determine the mechanism of the IL-1β-induced decrease in β-adrenergic responsiveness, we first examined whether IL-1β affects β-adrenoceptor binding. Figure 3shows representative Scatchard plots from saturation binding assays with use of [3H]CGP-12177, a β-adrenergic antagonist, in partially purified membranes prepared from rat ventricular myocardium. Results from repeated assays were analyzed by the nonlinear curve-fitting program LIGAND (28) and consistently showed the best fit to be a single population of high-affinity binding sites with a dissociation constant (K d) of 0.29 ± 0.03 nM and a maximal binding site density (Bmax) of 54.2 ± 12.6 fmol/mg protein (n = 3), values similar to those reported by others (5). Figure 3 also shows that IL-1β (4 ng/ml) had no effect onK d (0.36 ± 0.04 nM, n = 3) or Bmax (56.6 ± 12.0 fmol/mg protein, n = 3).
Figure 4 shows results from experiments examining competitive displacement of [3H]CGP-12177 binding by increasing concentrations of Iso. Figure4 A shows that, in the absence of guanine nucleotides, Iso competition for [3H]CGP-12177 binding was best characterized by a two-binding-site model with relative affinities (K i) of 5.8 ± 1.8 × 10−8 M (representing 80.3 ± 3.0% of total specific [3H]CGP-12177 binding) and 1.4 ± 0.5 × 10−5 M (19.6 ± 2.8% of total specific binding). Analysis by LIGAND always suggested that a two-site model was a better fit than a single-site model; however, in two of the six experiments, the Fvalue did not indicate significant differences between the two models. The presence of 4 ng/ml IL-1β did not significantly alter the effect of Iso on [3H]CGP-12177 binding; observed K ivalues were 6.9 ± 2.7 × 10−8 M (77.3 ± 4.3% of total specific binding) and 1.0 ± 0.4 × 10−5 M (22.7 ± 4.3% of total specific binding). Figure 4 Bshows that, in the presence of 100 μM GppNHp, Iso antagonized [3H]CGP-12177 binding from a single population of binding sites withK i values of 0.71 ± 0.31 × 10−6 M in control conditions (n = 3) and 0.82 ± 0.23 10−6 M in the presence of 4 ng/ml IL-1β (n = 3). LIGAND analysis demonstrated a single-site model to be the best fit in all experiments. These findings were supported by data obtained from binding studies with intact ventricular myocytes. Table1 shows that 4 ng/ml IL-1β did not alter specific [3H]CGP-12177 binding (0.5 nM final concentration) to adult rat ventricular myocytes in the presence or absence of 0.1 μM Iso. In summary, results of the binding studies indicated that IL-1β does not affect agonist binding to β-adrenoceptors.
Effects of IL-1β on cAMP levels.
Studies in neonatal rat cardiac myocytes have shown that cytokines decrease β-adrenergic responsiveness by suppressing the Iso-induced increase in cell cAMP concentration (14). We determined whether this is the case in adult rat ventricular myocytes by examining cell cAMP levels in response to Iso in the absence and presence of IL-1β. The basal intracellular cAMP concentration in adult rat ventricular myocytes was 4.73 ± 0.33 pmol/105 cells (n = 13) or 8.03 ± 0.56 pmol/mg protein, a value comparable to that reported by other investigators (37). In Fig. 5, a 5-min incubation in 0.1 μM Iso approximately doubled the intracellular cAMP concentration. Incubation for 5 or 10 min with 5 ng/ml IL-1β had no effect on basal or Iso-enhanced cAMP levels. These results suggest that the IL-1β-induced decrease in β-adrenergic responsiveness did not result from alterations in Iso-stimulated intracellular cAMP accumulation.
Effects of IL-1β on the cAMP-dependent activation of ICa,L.
Because IL-1β had no effect on Iso-enhanced cAMP content, we then determined whether IL-1β altersI Ca,L activated by forskolin (Fsk). Figure6 A shows results from a control experiment in which 4 ng/ml IL-1β attenuated Iso-enhancedI Ca,L by 42%. In Fig. 6 B, 1 μM Fsk caused an ∼77% increase inI Ca,L, a level similar to that induced by 1 μM Iso. However, in contrast to the effect observed with Iso, a 3-min exposure to 4 ng/ml IL-1β did not significantly affect peakI Ca,L in the presence of Fsk. A subsequent exposure to IL-1β also had no effect onI Ca,L, further supporting the lack of its effect on Fsk-enhancedI Ca,L (Fig.6 B). Average peakI Ca,L in the Fsk and Fsk + IL-1β conditions were 167 ± 5% (n = 5) and 168 ± 5% (n = 5) of control, respectively. These results showed that IL-1β has no effect on Fsk-activated adenylyl cyclase activity or downstream effects of cAMP.
To further determine whether IL-1β interferes with cAMP-mediated activation ofI Ca,L, we examined the effect of IL-1β onI Ca,L in myocytes internally dialyzed with cAMP or extracellularly perfused with 8-(4-chlorophenylthio)-cAMP (CPT-cAMP), a membrane-permeable analog of cAMP. In Fig.7 A, 10 μM cAMP in the pipette solution almost doubled the peakI Ca,L, a level similar to that induced by 1 μM Iso. Exposure of the myocyte to 4 ng/ml IL-1β had no significant effect on the cAMP-enhancedI Ca,L. Averaged current magnitude in IL-1β was 0.96 ± 0.02 (n = 4) and 0.97 ± 0.03 (n = 4) of the control peakI Ca,L in the presence of 1 and 10 μM cAMP, respectively. Similarly, in a representative experiment in Fig. 7 B, extracellular perfusion of a myocyte with 0.3 mM CPT-cAMP caused a 240% increase in peakI Ca,L. Subsequent exposure to 5 ng/ml IL-1β did not significantly alter the cAMP-activatedI Ca,L. Averaged current magnitude in IL-1β was 0.94 ± 0.02 (n = 4) of the control peakI Ca,L in the presence of CPT-cAMP. These results suggested that the IL-1β-induced suppression of Iso-enhancedI Ca,L is not mediated by inhibition of the cAMP-dependent activation of Ca2+ channels. We then examined whether IL-1β suppresses Iso-enhancedI Ca,L in the presence of cAMP. Figure 8 shows an increase in I Ca,Lof 150% in the presence of 0.1 mM CPT-cAMP that was further enhanced by addition of 1 μM Iso. Under these conditions, exposure to 5 ng/ml IL-1β caused an ∼43% inhibition of Iso-stimulatedI Ca,L. Results from five experiments show that IL-1β reducedI Ca,L by 21.9 ± 5.2% in the presence of CPT-cAMP and Iso. Similarly, in myocytes internally dialyzed with 10 μM cAMP, 5 ng/ml IL-1β decreased Iso-stimulatedI Ca,L by 24.6 ± 6.1% (n = 3). These data further support the suggestion that the IL-1β-induced suppression of β-adrenergic responsiveness ofI Ca,L is mainly mediated by a cAMP-independent mechanism rather than antagonism of the cAMP-induced activation of Ca2+channels.
We previously showed that IL-1β reduces basal peakI Ca,L in rat ventricular myocytes (26). The present study demonstrates that IL-1β decreases the β-adrenergic responsiveness ofI Ca,L by suppressing the maximal effect of Iso and increasing its EC50. IL-1β does not alter basal or Iso-induced cAMP levels, β-adrenoceptor binding, or Fsk-stimulatedI Ca,L and has little, if any, effect on cAMP-activatedI Ca,L. These results suggest that the IL-1β-induced acute inhibition of Iso’s effects on I Ca,Lis mediated primarily at a site other than the β-adrenoceptor-adenylyl cyclase-protein kinase A pathway.
Comparison with findings observed in other cardiac myocytes.
Studies with neonatal rat cardiac myocytes have shown that a 72-h incubation with activated splenocyte-conditioned medium or IL-1β inhibits the β-adrenergic responsiveness of contractility by suppressing the Iso-enhanced cAMP level (14). A consecutive study showed that β-adrenoceptor binding was unaltered; however, Fsk-stimulated cAMP concentrations were enhanced, whereas Iso-stimulated cAMP content was decreased by 7% after a 24-h treatment (6). In contrast, the present study in adult rat ventricular myocytes shows no effect of IL-1β on basal or Iso-enhanced cAMP levels. The discrepancy in these results could be due to numerous factors, such as different developmental stages, duration of incubation with IL-1β, or different cell populations in primary culture of neonatal cardiac myocytes. Developmental differences in the β-adrenergic and Fsk responsiveness ofI Ca,L have been shown in rat (23) and rabbit ventricular myocytes (30).
Studies in adult guinea pig ventricular myocytes have shown that IL-1β does not alter the β-adrenergic responsiveness ofI Ca,L unless the exposure duration of the cytokine is >1 h (31). This IL-1β-induced inhibition of β-adrenergic responsiveness ofI Ca,L was attenuated by replacement ofl-arginine withd-arginine and by incubation with an inhibitor of NOS, suggesting the involvement of the NOS pathway (31). These investigators did not provide information about whether IL-1β affects basal or Iso-enhanced cAMP content; however, they did show that IL-1β has no effect on Fsk-activatedI Ca,L, as indicated by our data (Fig. 6 B). In addition, the present results obtained from adult rat ventricular myocytes show that the IL-1β-induced inhibition of β-adrenergic responsiveness ofI Ca,L occurs after only a couple minutes of cytokine exposure (Figs.1 A and6 A), when cGMP production is not significantly altered (unpublished data). These data suggest that the NOS pathway is not involved in this action. The cause of the discrepancy between these two studies is unclear but could be attributed to differences in species and/or experimental conditions. Species variations in the β-adrenergic and Fsk responsiveness ofI Ca,L in ventricular myocytes have been reported (23, 30).
IL-1β and cAMP.
Studies in vascular smooth cells showed that IL-1β stimulates cAMP but not cGMP production within 1 h (2). The increased cAMP has been suggested to mediate the stimulation of expression of inducible NOS and production of nitrite that causes vasodilatation (2, 35). Similarly, cAMP has been shown to upregulate IL-1β-induced inducible NOS mRNA expression and nitrite production in neonatal rat cardiac myocytes (21). In contrast, a study in decidual cells showed that low concentrations of IL-1β increase the production of cAMP during a 24-h exposure, whereas low concentrations of the cytokine (1 ng/ml) inhibit cAMP production (7). This study suggested that cAMP does not mediate the IL-1β-induced stimulation of prostaglandin production. In addition, in astrocytoma cells, IL-1β induces IL-6 release without altering cAMP formation (4). Our present study showed that IL-1β does not affect the basal or Iso-induced cAMP production in adult rat ventricular myocytes. Therefore, data suggest that the role of cAMP in the signal transduction mechanisms for IL-1β varies among species and cell types.
Our results show that IL-1β has no effect on Iso-stimulated cAMP content, β-adrenoceptor binding, orI Ca,L in the presence of Fsk. The minor inhibitory effect (<5%) of IL-1β in the presence of intracellular or extracellular cAMP suggests that a cAMP-independent mechanism is involved in the IL-1β-induced inhibition of Iso-activatedI Ca,L. This is supported by results showing that, in the presence of cAMP,I Ca,L is further increased by additional exposure to 1 μM Iso, and the Iso-induced increase in I Ca,Lis decreased by addition of IL-1β. It has been suggested that Iso can stimulate I Ca,Lvia a cAMP-independent pathway that involves direct regulation via a G protein in adult rat ventricular myocytes (23). However, because the relative contribution of this direct G protein cAMP-independent effect of Iso on I Ca,Lhas been questioned (19), the role of this proposed pathway for the observed IL-1β-induced suppression of β-adrenergic responsiveness of I Ca,L requires further investigation. We previously showed that IL-1β stimulates the production of ceramide and that ceramide mediates the IL-1β-induced suppression of basalI Ca,L in adult rat ventricular myocytes (34). It is very likely that ceramide may be involved in this cAMP-independent pathway that suppresses the Iso-activatedI Ca,L.
In summary, IL-1β suppresses β-adrenergic responsiveness ofI Ca,L via a cAMP-independent pathway. This pathway may include IL-1β-stimulated ceramide production, which counterbalances, rather than disrupts, the Iso-stimulated cAMP-dependent pathway. This action may play an important role in the reduced myocardial function observed during various cardiac disorders associated with cell injury and immune and inflammatory responses. It is also possible that the IL-1β-induced decrease in β-adrenergic responsiveness plays a cardioprotective role by reducing energy demand during the compensatory phase in cardiac dysfunction.
We thank Meei-Yueh Liu for excellent technical assistance.
Address for reprint requests: S. J. Liu, Dept. of Biopharmaceutical Sciences, University of Arkansas for Medical Sciences, 4301 West Markham St., MS 522–3, Little Rock, AR 72205.
This work was supported in part by grants from the American Heart Association/Arkansas Affiliate, the American Health Assistance Foundation, and the Office of Naval Research.
Present address of W. Zhou: Dept. of Anesthesiology, Baylor College of Medicine, Houston, TX 77030.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
- Copyright © 1999 the American Physiological Society