Selective attenuation by adenosine of arrhythmogenic action of isoproterenol on ventricular myocytes

Transcription

1 Am J Physiol Heart Circ Physiol 281: H2789 H2795, Selective attenuation by adenosine of arrhythmogenic action of isoproterenol on ventricular myocytes YEJIA SONG, 1 JOHN C. SHRYOCK, 1 HARM J. KNOT, 2 AND LUIZ BELARDINELLI 3 Departments of 1 Medicine and 2 Pharmacology, University of Florida, Gainesville, Florida 32610; and 3 CV Therapeutics, Palo Alto, California Received 11 September 2000; accepted in final form 8 February 2001 Song, Yejia, John C. Shryock, Harm J. Knot, and Luiz Belardinelli. Selective attenuation by adenosine of arrhythmogenic action of isoproterenol on ventricular myocytes. Am J Physiol Heart Circ Physiol 281: H2789 H2795, We examined whether adenosine equally attenuated the stimulatory effects of isoproterenol on arrhythmic activity and twitch shortening of guinea pig isolated ventricular myocytes. Transmembrane voltages and whole cell currents were recorded with patch electrodes, and cell twitch shortening was measured using a video-motion detector. Isoproterenol increased the action potential duration at 50% repolarization (APD 50), L-type Ca 2 current [I Ca(L)], and cell twitch shortening and induced delayed afterdepolarizations (DAD), transient inward current (I Ti), and aftercontractions. Adenosine attenuated the arrhythmogenic actions of isoproterenol more than it attenuated the effects of isoproterenol on APD 50, I Ca(L), or twitch shortening. Adenosine ( mol/l) decreased the amplitude of DADs by 30 6% to 92 5% but attenuated isoproterenol-induced prolongation of the APD 50 by only 14 4% to 59 4% and had no effect on the voltage of action potential plateau. Adenosine (30 mol/l) inhibited I Ti by 91 4% but decreased isoproterenol-stimulated I Ca(L) by only 30 12%. Isoproterenol-induced aftercontractions were abolished by adenosine (10 mol/l), whereas the amplitude of twitch shortening was not reduced. The effects of adenosine on twitch shortenings and aftercontractions were mimicked by the A 1-adenosine receptor agonist CPA (N 6 -cyclopentyladenosine) and by ryanodine. In conclusion, adenosine antagonized the proarrhythmic effect of -adrenergic stimulation on ventricular myocytes without reducing cell twitch shortening. electrophysiology; arrhythmias; twitch shortening; heart ACTIVATION of -adrenoceptors by catecholamines increases camp formation, L-type Ca 2 current [I Ca(L) ], and contractility of ventricular myocytes (10). Stimulation by isoproterenol of I Ca(L) in guinea pig ventricular myocytes is associated with a prolongation of the action potential duration (APD) and a positive shift of the voltage of the action potential plateau (vapp)(1). By increasing camp formation and I Ca(L), -adrenergic stimulation may also induce an arrhythmogenic-triggered activity, including the transient inward current (I Ti ) and delayed afterdepolarizations (DAD) (1, 21, 26). Address for reprint requests and other correspondence: L. Belardinelli, CV Therapeutics, 3172 Porter Dr., Palo Alto, CA ( luizb@cvt.com). Adenosine attenuates isoproterenol-induced increases of camp and I Ca(L), and thereby the increases of the APD, contractility, and triggered activity in ventricular myocytes (1, 6, 12, 26). These effects of adenosine are mimicked and blocked, respectively, by selective agonists (e.g., N 6 -cyclopentyladenosine, CPA) (26) and antagonists (e.g., 8-cyclopentyl-1,3-dipropylxanthine, CPX) (4) of A 1 -adenosine receptors (A 1 AdoRs). Thus adenosine, by activating A 1 AdoRs, can antagonize both the positive inotropic and proarrhythmic effects of -adrenoceptor agonists on ventricular myocardium (2, 3). Consistent with the anti- -adrenergic actions of adenosine, an inhibition by adenosine of isoproterenol-facilitated ventricular tachycardia has been demonstrated in patients (19). However, when applied at therapeutic or physiological concentrations, adenosine did not significantly attenuate isoproterenol-stimulated contractility in vivo, and thus the physiological significance of anti- -adrenergic actions of adenosine was questioned (16, 23, 24). Because these apparently conflicting results were obtained by measurement of different responses (electric activity versus contractility), it is possible that the -adrenoceptor-mediated arrhythmic and inotropic effects of catecholamines are not equally antagonized by adenosine. This study quantitatively compared the effects of adenosine on isoproterenol-induced increases of the duration and plateau of action potential, I Ca(L), and cell twitch shortening with the effects of adenosine on isoproterenol-stimulated DAD, I Ti, and aftercontractions of guinea pig isolated ventricular myocytes. The goal of the present study was to determine whether adenosine could differentially attenuate the stimulatory effects of isoproterenol on arrhythmic activity and cell twitch shortening. METHODS Isolation of ventricular myocytes. Use of animals in the present study was in accordance with the Guide for the Care and Use of Laboratory Animals (National Institutes of Health Publication No , 1985) and was approved by the Institutional Animal Care and Use Committee of the University of Florida. Single ventricular myocytes were isolated 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. Section 1734 solely to indicate this fact /01 $5.00 Copyright 2001 the American Physiological Society H2789

2 H2790 from the hearts of adult Harlan guinea pigs of either sex, as described previously (26). Measurements of transmembrane potential and current. Myocytes were placed into a recording chamber and superfused with Tyrode solution containing (in mmol/l) 140 NaCl, 4.6 KCl, 1.8 CaCl 2, 1.1 MgSO 4, 10 glucose, and 5 HEPES (ph 7.4, 35 C). Drugs were applied via the superfusate. The effect of a drug in the presence of isoproterenol was determined by simultaneous application of the drug with isoproterenol. Each drug treatment usually took about 2 min. Measurements were made when the response to a drug had reached a maximum. Transmembrane voltages and currents were measured using glass microelectrodes filled with solution containing (in mmol/l) 120 potassium aspartate, 20 KCl, 1 MgCl 2,4Na 2ATP, 0.1 Na 3GTP, 10 glucose, and 10 HEPES (ph 7.2). Microelectrode resistance was 1 3 M. An Axopatch-200 amplifier, a DigiData-1200A interface, and pclamp6 software (Axon Instruments; Foster City, CA) were used to perform electrophysiological measurements. The electrode capacitance, whole cell capacitance, and series resistance were maximally compensated. For measurements of I Ca(L) and I Ti, cells were voltage clamped at 40 mv to inactivate the fast Na channels. A 500-ms depolarizing pulse to 10 mv was applied at a frequency of 0.5 Hz to elicit I Ca(L) and I Ti. In some experiments, Cs was added to both Tyrode and pipette solutions to replace K on an equimolar basis to reduce contamination of I Ca(L) with K current. In other experiments, the regular (i.e., K containing) Tyrode and pipette solutions were used. This was done to exclude a possible interference of Cs with intracellular excitation-contraction coupling (20) and to facilitate the interpretation of results of experiments to measure membrane potential and twitch shortening in which K - containing solutions were used. The amplitude of I Ca(L) was measured from the zero current to the maximal inward current, and the amplitude of I Ti was measured from the holding current to the peak inward I Ti. Values of the amplitude of I Ca(L) and I Ti were normalized by relation to the whole cell capacitance (33 2 pf, read from the capacitance meter of the amplifier) and expressed as pico ampere per pico farad. For recording of action potentials and DADs, a 5-ms depolarizing pulse was applied at a frequency of 0.5 to 1 Hz. APD 50 was measured, vapp was determined at 50 ms after the upstroke of the action potential, and the amplitude of DAD was measured from the maximal diastolic potential to the peak of the deflection. Measurement of cell twitch shortening. The amplitude of unloaded contraction of myocytes was used as an index of cell contractility (27). In the text, twitch shortening indicates a normal cell contraction, whereas an aftercontraction denotes a small contraction that occurs during diastole and is triggered by events after the preceding normal contraction. The procedures described above to trigger action potentials and I Ca(L) were used to induce cell twitch shortenings. Movement of the cell edge image across a raster line of the monitor was analyzed using a video motion detector (Crescent Electronics; Logan, UT). The amplitude of twitch shortening and aftercontraction was measured from the maximal cell relaxation to the peak contraction and calculated as an average of 10 consecutive beats. Statistical analysis. Data are expressed as means SE, and n indicates the number of cells studied. Percentage inhibition by adenosine of the effects of isoproterenol was calculated using the formula [(isoproterenol adenosine)/ (isoproterenol control)] 100, where isoproterenol, adenosine, and control indicate measurements obtained in the presence of isoproterenol alone or isoproterenol plus adenosine and in the absence of drugs, respectively. The paired Student s t-test was used for statistical analysis of paired data, and the one-way repeated measures ANOVA followed by Student-Newman-Keuls test was applied for multiple comparisons. Differences between means were considered statistically significant at P RESULTS Effects of isoproterenol and adenosine on action potential and DAD. In the absence of drugs, the resting membrane potential of myocytes (n 32 cells) was 84 1 mv. The APD 50 and vapp were 217 7ms and 37 1 mv, respectively. Neither spontaneous activity nor triggered activity was observed (Fig. 1A). Isoproterenol (25 nmol/l) caused a prolongation of the Fig. 1. Effects of isoproterenol (Iso) and adenosine (Ado) on action potential duration at 50% repolarization (APD 50), voltage of the action potential plateau (vapp), and delayed afterdepolarization (DAD). A: action potentials of a myocyte recorded in the presence of no drug (control, trace a), Iso (25 nmol/l, trace b) alone, and Iso Ado (10 mol/l, trace c). Arrow indicates the APP. Note the DAD in the presence of Iso. B: summary of the effects of 25 nmol/l Iso and 10 mol/l Ado on APD 50 and DADs. Each bar represents data obtained from 10 myocytes in the absence (control) and presence of Iso alone or Iso Ado. *P C: summary of the effects of 25 nmol/l Iso and 10 mol/l Ado on vapp. Each bar represents data obtained from 10 myocytes in the absence and presence of drugs as indicated. *P 0.05 vs. control. D: percentage inhibition of Iso (25 nmol/l)-induced changes in APD 50, DAD, and vapp by various concentrations of Ado. Each point represents data collected from 8 to 10 myocytes. Differences between percent inhibitions of APD 50 and DADs are statistically significant at each concentration of Ado.

3 H2791 APD 50 from 217 7to ms (n 32 cells, P 0.05) and a positive shift of the vapp by 8 1mV(n 32 cells, P 0.05, Fig. 1A). Isoproterenol also induced DAD with amplitudes of mv (n 32 cells, Fig. 1A). Only a single DAD after each action potential was observed. In the presence of isoproterenol, adenosine caused a greater reduction of the amplitude of DAD than of either the APD 50 or vapp (Fig. 1A). The amplitude of DAD was significantly reduced by 83 6% by 10 mol/l adenosine from to mv (n 10 cells, P 0.05) (Fig. 1B). In contrast, the isoproterenol-induced prolongation of APD 50 was modestly reduced by 43 4%, from to ms (n 10 cells, P 0.05) (Fig. 1B), and the vapp was not affected by adenosine (Fig. 1C). Overall, adenosine at 0.1, 0.3, 1, 3, 10, 30, and 100 mol/l decreased the amplitude of isoproterenol-induced DADs by 30 6to 92 5% but attenuated isoproterenol-induced prolongation of APD 50 by only 14 4to59 4% and had no significant effect on the vapp (Fig. 1D). The effects of adenosine were reversible on washout of the drug (not shown). The difference between the percent inhibitions by adenosine of APD 50 and DAD was statistically significant at each concentration of adenosine tested (Fig. 1D). Effects of isoproterenol and adenosine on I Ca(L) and I Ti. In the absence of drugs, the amplitude of I Ca(L) was 29 2 pa/pf (n 8 cells), and no I Ti was observed (Fig. 2). Isoproterenol (25 nmol/l) increased the amplitude of I Ca(L) to pa/pf (P 0.05) and induced I Ti with an amplitude of pa/pf (Fig. 2). Adenosine (30 mol/l) inhibited isoproterenol-induced I Ti by 91 4% (from to pa/pf, P 0.05) but reduced isoproterenol-stimulated I Ca(L) by only 30 12% (from to 64 8 pa/pf, P 0.05, Fig. 2). The inhibitory effect of adenosine on I Ti was significantly greater than the effect of adenosine on I Ca(L) (P 0.05). The effects of adenosine were reversed after washout of the drug (not shown). Increases of I Ca(L) and I Ti contribute to isoproterenolinduced prolongation of APD 50 and positive shift of vapp and to isoproterenol-stimulated triggered activity, respectively (1, 12, 26). Therefore, the lower sensitivity of I Ca(L) compared with I Ti to inhibition by adenosine explains why the effects of -adrenergic stimulation on action potentials and triggered activity are differentially attenuated by adenosine. Effects of isoproterenol and adenosine on twitch shortening and aftercontraction. Isoproterenol (25 nmol/l) increased the amplitude of twitch shortening from to m (n 8 cells, P 0.05) and induced aftercontractions with amplitudes of m in current-clamped myocytes. DADs (Fig. 3A, top) occurred concomitantly with the aftercontractions (Fig. 3A, bottom). The amplitude of a twitch shortening after an aftercontraction was often smaller than that without a preceding aftercontraction and was inversely correlated with the amplitude of the aftercontraction, the interval between the twitch shortening, and the preceding aftercontraction (not shown). In the presence of both isoproterenol and adenosine (10 mol/l), the Fig. 2. Effects of Iso and Ado on L-type Ca 2 current [I Ca(L)] and transient inward current (I Ti). A: I Ca(L) and I Ti recorded in the presence of no drug (control, trace a), 25 nmol/l Iso (trace b), and Iso 30 mol/l Ado (trace c). K in all solutions was replaced by Cs. B: summary of 8 experiments to measure the amplitude of I Ca(L) and I Ti in the absence of drugs (control), in the presence of Iso (25 nmol/l), or Iso Ado (30 mol/l). *P aftercontractions and DADs were completely suppressed, whereas the amplitude of twitch shortening was not reduced ( m, P 0.05 vs. isoproterenol alone, Fig. 3B). Isoproterenol (25 nmol/l) increased I Ca(L) and induced I Ti in voltage-clamped myocytes (Fig. 4A, top). The increases of I Ca(L) and I Ti corresponded temporally to the increase of twitch shortening and the occurrence of aftercontractions, respectively (Fig. 4A, bottom). Adenosine (10 mol/l) completely inhibited isoproterenol-stimulated I Ti and aftercontractions but had only a minor effect on I Ca(L) and twitch shortenings (Fig. 4A). In a total of six cells, isoproterenol alone increased the amplitude of twitch shortening from to m (P 0.05) and induced aftercontractions with amplitude of m. In the presence of adenosine, the amplitude of twitch shortening was not significantly changed ( m, P 0.05 vs. isoproterenol alone), whereas the aftercontractions were completely abolished (Fig. 4B). The actions of adenosine and ryanodine were similar. We speculated that the effect of adenosine to reduce aftercontractions in the presence of isoproterenol

4 H2792 might involve an inhibition of Ca 2 release from the sarcoplasmic reticulum (SR) during diastole. To test this hypothesis, we examined the effect of ryanodine, an inhibitor of the SR Ca 2 -release channels (7) on cell twitch shortening and aftercontractions. Ryanodine alone (100 nmol/l) caused a 26 3% decrease of the amplitude of twitch shortening from to m (n 5 cells, P 0.05). In the presence of isoproterenol, ryanodine (100 nmol/l) mimicked the actions of adenosine on cell twitch shortening and aftercontractions. In six experiments, such as those shown in Fig. 5, isoproterenol (25 nmol/l) induced aftercontractions of current-clamped myocytes and increased the amplitude of twitch shortening from to m(p 0.05). Addition of ryanodine in the presence of isoproterenol reduced the amplitude of aftercontractions from to m (P 0.05) but did not reduced the amplitude of twitch shortening ( vs m, P 0.05). Role of A 1 AdoR. To establish the role of A 1 AdoRs in the differential anti- -adrenergic actions of adenosine, we replaced adenosine with the selective A 1 AdoR agonist CPA. In a total of six cells (Fig. 6), isoproterenol (25 nmol/l) increased the amplitude of I Ca(L) from 53 7to pa/pf (P 0.05) and induced I Ti with an amplitude of 18 3 pa/pf. CPA (100 nmol/l) had no significant effect on isoproterenol-stimulated I Ca(L) Fig. 4. Effects of Iso and Ado on membrane currents, cell twitch shortenings, and aftercontractions in voltage-clamp experiments. A: currents (top) and twitch shortenings (bottom) recorded simultaneously from a myocyte superfused with K -containing solutions. Current a, control, no drug. Current b, in the presence of 25 nmol/l Iso. Arrows point to I Ti (top) and aftercontractions (bottom). Current c, in the presence of Iso and Ado (10 mol/l). B: summary of the amplitude of twitch shortenings and aftercontractions of 6 myocytes measured in the absence (control) and presence of Iso (25 nmol/l) or Iso Ado (10 mol/l). *P 0.05 vs. control. Fig. 3. Effects of Iso and Ado on action potentials, DADs, cell twitch shortening, and aftercontractions in current-clamp experiments. A: action potentials (top) and twitch shortenings (bottom) recorded simultaneously from a myocyte. Trace a, control, no drug. Trace b, in the presence of 25 nmol/l Iso. Arrows indicate DADs (top) and aftercontractions (bottom). Trace c, in the presence of Iso and Ado (10 mol). B: summary of the amplitude of twitch shortenings and aftercontractions of 8 myocytes recorded in the absence (control) and presence of Iso (25 nmol/l) or Iso Ado (10 mol/l). *P 0.05 vs. control. ( vs pa/pf, P 0.05) but markedly inhibited the I Ti by 80 5% (from 18 3to4 1 pa/pf, P 0.05). In four of these cells (Fig. 6A), the selective A 1 AdoR antagonist CPX was applied, and the effects of CPA were reversed by CPX. In these experiments, isoproterenol (25 nmol/l) increased the amplitude of I Ca(L) from 47 8to pa/pf (n 4 cells, P 0.05) and induced I Ti with an amplitude of 15 2 pa/pf. In the presence of both isoproterenol and CPA (100 nmol/l), the amplitude of I Ca(L) was not affected ( pa/pf, P 0.05), but the I Ti was reduced to 2 1 pa/pf (P 0.05). After addition of CPX (100 nmol/l) in the continuous presence of isoproterenol and CPA, the amplitude of I Ca(L) remained unchanged ( pa/pf, P 0.05), whereas the I Ti was increased to 9 1 pa/pf (P 0.05). In a group of current-clamped myocytes (n 4 cells), isoproterenol (25 nmol/l) increased the amplitude of twitch shortenings from to m (P 0.05) and induced aftercontractions with amplitudes of m. The isoproterenol-induced aftercontractions were completely inhibited by CPA (100 nmol/l). However, the amplitude of twitch shortenings in the presence of both isoproterenol and CPA was not reduced ( m, P 0.05 vs. isoproterenol alone).

5 H2793 Fig. 5. Effects of Iso and ryanodine (Rya) on cell twitch shortening and aftercontractions. Action potentials (top) and twitch shortenings (bottom) were recorded simultaneously from a current-clamped myocyte exposed to no drug (A), 25 nmol/l Iso (B), and Iso 100 nmol/l Rya (C). Filled circles indicate DADs and associated aftercontractions, whereas arrowheads point to a spontaneous action potential and an associated twitch shortening. The inhibitory effect of CPA on aftercontractions was antagonized by CPX. After application of CPX (100 nmol/l) in the presence of isoproterenol and CPA, aftercontractions reappeared with amplitudes of m, whereas the amplitude of twitch shortenings was m (P 0.05). These data indicate that Fig. 6. Effects of Iso, N 6 -cyclopentyladenosine (CPA), and 8-cyclopentyl-1,3-dipropylxanthine (CPX) on I Ca(L) and I Ti. A: currents recorded from a myocyte superfused with K -containing solutions. Trace a,noi Ti was observed in the absence of drug. Trace b, 25 nmol/l of Iso increased the amplitude of I Ca(L) and induced I Ti. Trace c, 100 nmol/l of Iso CPA caused a small attenuation of I Ca(L) but a complete inhibition of I Ti. Trace d, the effects of CPA were partially reversed by CPX (100 nmol/l). B: summary of the results obtained from 6 myocytes in the absence of drug (control) and in the presence of Iso (25 nmol/l) or Iso CPA (100 nmol/l). NS, no significant difference. *P the differential anti- -adrenergic actions of adenosine are mediated by the A 1 AdoR. Roles of 1 - and 2 -adrenoceptors. A possible explanation for the differential anti- -adrenergic actions of adenosine is that the stimulatory effects of isoproterenol are mediated by both 1 - and 2 -adrenoceptors and that 2 -adrenoceptor-mediated effects are insensitive to inhibition by adenosine. This hypothesis was examined by determining the roles of 1 - and 2 -adrenoceptors in stimulation by isoproterenol of I Ca(L) using highly selective antagonists for 1 (CGP-20712A)- and 2 (ICI-118,551)-adrenoceptors. The inhibitory constant values for CGP-20712A and ICI at the high-affinity binding sites of ventricular membranes are nm and nm, respectively (15). CGP-20712A at 300 nm and ICI at 50 nm occupied 100% of their high-affinity binding sites (15). Therefore, we compared the effects of 250 nmol/l CGP A and 100 nmol/l ICI-118,551 on isoproterenolstimulated I Ca(L). In eight myocytes, isoproterenol (25 nmol/l) increased I Ca(L) from 30 3to70 11 pa/pf (P 0.05). Addition of CGP-20712A (250 nmol/l) caused a 96 2% inhibition of isoproterenol-stimulated I Ca(L), reducing the I Ca(L) to 32 3 pa/pf (P 0.05 vs. isoproterenol alone). In another series of experiments, the effect of the 2 -adrenoceptor antagonist ICI-118,551 was studied. Isoproterenol alone (25 nmol/l) increased I Ca(L) from 27 4 to pa/pf (n 6 cells, P 0.05). ICI-118,551 (100 nmol/l) decreased the amplitude of I Ca(L) by only 8 4% to pa/pf (P 0.05 vs. isoproterenol alone). Thus the effect of CGP-20712A to attenuate the increase of I Ca(L) caused by isoproterenol was significantly greater than the effect of ICI-118,551 (P 0.05). These results suggest that the stimulatory effect of isoproterenol on I Ca(L) of guinea pig ventricular myocytes is primarily mediated by 1 -adrenoceptors. DISCUSSION The results of our experiments demonstrated that inhibition by adenosine of isoproterenol-stimulated arrhythmic activity is significantly greater than inhibition by adenosine of isoproterenol-induced increases of APD 50 (Fig. 1), vapp (Figs. 1), I Ca(L) (Fig. 2), and cell twitch shortening (Figs. 3 and 4). The effects of aden-

6 H2794 osine were mimicked by CPA and thus were mediated by the A 1 AdoR (Fig. 6). This conclusion is supported by previous reports that the stimulatory effect of isoproterenol on myocardial contractility is not inhibited by adenosine and A 1 AdoR agonists in vivo (16, 23, 24) and is maintained in mouse hearts overexpressing A 1 AdoRs (9). Stimulation of I Ca(L) is the major ionic mechanism for the positive inotropic effect of -adrenoceptor agonists (10, 22) and is associated with an elevation of the plateau and a prolongation of the duration of the ventricular myocyte action potential (1, 12). Therefore, in this study, the effect of isoproterenol on cell contraction was assessed not only by measurement of the amplitude of cell twitch shortening but also by measurements of the amplitude of I Ca(L) and the duration and plateau voltage of action potentials. Similarly, activation of I Ti underlies the induction of DADs (17) and is associated with aftercontractions (14). Hence, the effect of isoproterenol to initiate arrhythmias was evaluated by measurements of both electrical (I Ti, DADs) and mechanical (aftercontractions) activities. The results of our experiments to measure membrane potentials (action potential and afterdepolarization), membrane currents [I Ca(L) and I Ti ], and contractility (cell twitch shortening and aftercontraction) are qualitatively similar, indicating a close link between these electrical and mechanical parameters. Adenosine inhibited the aftercontractions but did not decrease the amplitude of cell twitch shortening in the presence of isoproterenol. Moreover, the amplitude of twitch shortening tended to increase after the aftercontractions were inhibited by adenosine in currentclamped myocytes (Fig. 3A), although this effect was not statistically significant. This apparent paradox is consistent with the observations (Fig. 5) that an aftercontraction may decrease the amplitude of the following twitch shortening. The intracellular ionic basis for induction of DADs and aftercontractions is the oscillatory release of Ca 2 from the SR (13). In the present study, the oscillatory release of Ca 2 was likely to be caused by -adrenergic stimulation (29). Although Ca 2 uptake by the SR is also enhanced at the same time (8), frequent oscillatory releases of Ca 2 induced by -adrenergic stimulation may decrease the SR Ca 2 content, the amount of Ca 2 released during an action potential, and the amplitude of cell twitch shortening (5, 28). We observed that ryanodine, which was reported to block SR Ca 2 -release channels (7), reduced the frequency and amplitude of aftercontractions without decreasing the twitch shortening. It is possible that adenosine, by antagonizing the -adrenergic stimulation, can also reduce oscillatory release of Ca 2 from the SR in the presence of isoproterenol. Thus we speculate that a reduction of diastolic Ca 2 release may underlie the actions of adenosine to attenuate isoproterenol-induced DADs and aftercontractions. The observation that an inhibition by adenosine of DADs and aftercontractions (i.e., inhibition of oscillatory releases of Ca 2 from the SR) sometimes coincided with an increase of cell twitch shortening (Fig. 3A) is not in conflict with the inhibition by adenosine of isoproterenol-stimulated cell twitch shortening reported previously (1, 25). In the absence of DADs and aftercontractions, attenuation by adenosine of isoproterenolenhanced twitch shortenings is expected to be more manifest. When aftercontractions are present, twitch shortenings are reduced. When adenosine reduces arrhythmic activity, it may allow increased loading of Ca 2 into the SR and greater twitch shortenings. The negative effect of aftercontractions on the amplitude of twitch shortenings was less noticeable in voltageclamped myocytes (Fig. 4) than in current-clamped myocytes (Fig. 3). This finding was consistent with the fact that the amplitudes of aftercontraction in these voltage-clamp experiments were small and hence had little effect on twitch shortenings. The most likely reason why adenosine differentially attenuates the inotropic and arrhythmogenic effects of an -adrenoceptor agonist is that adenosine only causes a partial inhibition of camp formation. It appears that in the presence of an -adrenoceptor agonist, a small (25% of maximum) increase of camp formation is sufficient to elicit a maximal inotropic effect (18). The minimal camp formation required for the arrhythmogenic effect of -adrenoceptor agonists may be higher than that for their positive inotropic effect. Thus, in the presence of adenosine, the amount of isoproterenol-stimulated camp may be reduced below the threshold to induce I Ti and DADs but still above the threshold to increase I Ca(L), APD, and contractility. An alternative hypothesis, namely that adenosine may differentially attenuate the stimulatory effects of isoproterenol mediated by the 1 - and 2 -adrenoceptor subtypes, was not supported by our results. It has been reported that both 1 - and 2 -adrenoceptor stimulation increased I Ca(L), but 1 -adrenoceptor stimulation was more likely to induce oscillatory releases of intracellular Ca 2 in rat ventricular myocytes (29). However, our data showed that the stimulatory effect of isoproterenol on I Ca(L) was antagonized by the 1 (CGP-20712A)- but not by the 2 (ICI-118,551)-adrenoceptor blocker. This result suggests that the stimulatory effects of isoproterenol on guinea pig ventricular myocytes are predominantly mediated by the 1 -adrenoceptor. The authors (11) of a previous study using guinea pig ventricular myocytes have also concluded that isoproterenol regulates I Ca(L) solely through an activation of the 1 -adrenoceptor. A limitation of our study is that a mechanism to explain the greater attenuation by adenosine of arrhythmogenic than of positive inotropic actions of catecholamines is not clearly demonstrated, nor is it likely to be easily elucidated by further investigation. A knowledge of effects of adenosine on intracellular camp and Ca 2 homeostasis in various subcellular compartments of the myocytes and on the activity of specific proteins (e.g., ryanodine receptors, Ca 2 transporters and exchangers, and protein phosphatases) is necessary to understand the modulation of cell function by adenosine. This knowledge is currently incom-

7 H2795 plete. However, our results do provide a rationale for investigation of the effect of A 1 AdoR agonists on 1) the state of phosphorylation and activity of the SR Ca 2 - release channels (ryanodine receptors) and 2) the incidence of ventricular arrhythmias in appropriate animal models of heart disease. Ventricular arrhythmias are a cause of sudden death of patients and animals with hypertrophied and failing hearts. Catecholamine levels are elevated in these circumstances. We suggest that the effect of A 1 AdoR agonists to reduce ventricular arrhythmias in animal models of cardiac hypertrophy and failure should be examined. In summary, the significant finding of the present study is that activation by adenosine of A 1 AdoRs can attenuate the proarrhythmic effect of isoproterenol to cause DADs and aftercontractions, without decreasing the contractility of cardiac ventricular myocytes. Further study is needed to elucidate the mechanism by which adenosine reduces DADs and aftercontractions in ventricular myocytes. The therapeutic potential of A 1 AdoR agonists for treatment of catecholamine-induced ventricular arrhythmias should be explored. It is possible that analogs of adenosine may have greater selectivity than adenosine itself to antagonize the proarrhythmic but not the positive inotropic effect of catecholamines. This study was supported by National Heart, Lung, and Blood Institute Grant HL and American Heart Association Grants N (National Center) and T (Maine, New Hampshire, and Vermont Affiliate). REFERENCES 1. Belardinelli L and Isenberg G. Actions of adenosine and isoproterenol on isolated mammalian ventricular myocytes. Circ Res 53: , Belardinelli L, Linden J, and Berne RM. The cardiac effects of adenosine. Prog Cardiovasc Dis 32: 73 97, Belardinelli L, Shryock JC, Song Y, Wang D, and Srinivas M. Ionic basis of the electrophysiological actions of adenosine on cardiomyocytes. FASEB J 9: , Brown LA, Humphrey SM, and Harding SE. The anti-adrenergic effect of adenosine and its blocked by pertussis toxin: a comparative study in myocytes isolated from guinea-pig, rat and failing human hearts. Br J Pharmacol 101: , Capogrossi MC, Stern MD, Spurgeon HA, and Lakatta EG. Spontaneous Ca 2 release from the sarcoplasmic reticulum limits Ca 2 -dependent twitch potentiation in individual cardiac myocytes. A mechanism for maximum inotropy in the myocardium. J Gen Physiol 91: , Dobson JG Jr. Reduction by adenosine of the isoproterenolinduced increase in cyclic adenosine 3,5 -menophosphate formation and glycogen phosphorylase activity in rat heart muscle. Circ Res 43: , Ehrlich BE, Kaftan E, Bezprozvannaya S, and Bezprozvanny I. The pharmacology of intracellular Ca 2 -release channels. Trends Pharmacol Sci 15: , Entman ML, Levey GS, and Epstein SE. Mechanism of action of epinephrine and glucagon on the canine heart. Evidence for increase in sarcotubular calcium stores mediated by cyclic 3,5 - AMP. Circ Res 25: , Gauthier NS, Headrick JP, and Matherne GP. Myocardial function in the working mouse heart overexpressing cardiac A 1 adenosine receptors. J Mol Cell Cardiol 30: , Hartzell HC. Regulation of cardiac ion channels by catecholamines, acetylcholine and second messenger systems. Prog Biophys Mol Biol 52: , Hool LC and Harvey RD. Role of 1- and 2-adrenergic receptors in regulation of Cl and Ca 2 channels in guinea pig ventricular myocytes. Am J Physiol Heart Circ Physiol 273: H1669 H1676, Isenberg G and Belardinelli L. Ionic basis for the antagonism between adenosine and isoproterenol on isolated mammalian ventricular myocytes. Circ Res 55: , January CT and Fozzard HA. Delayed afterdepolarizations in heart muscle: mechanisms and relevance. Pharmacol Rev 40: , Kass RS, Lederer WJ, Tsien RW, and Weingart R. Role of calcium ions in transient inward currents and aftercontractions induced by strophanthidin in cardiac Purkinje fibers. J Physiol 281: , Kitagawa Y, Adachi-Akahane S, and Nagao T. Determination of -adrenoceptor subtype on rat isolated ventricular myocytes by use of highly selective -antagonists. Br J Pharmacol 116: , Koglin J, Bohm M, von Scheidt W, Stablein A, and Erdmann E. Antiadrenergic effect of carbachol but not of adenosine on contractility in the intact human ventricle in vivo. J Am Coll Cardiol 23: , Lederer WJ and Tsien RW. Transient inward current underlying arrhythmogenic effects of cardiotonic steroids in Purkinje fibers. J Physiol 263: , Lemoine H, Teng KJ, Slee SJ, and Kaumann AJ. On minimum cyclic AMP formation rates associated with positive inotropic effects mediated through 1-adrenoceptors in kitten myocardium. 1-Specific and nonadrenergic stimulant effects of denopamine. Naunyn Schmiedebergs Arch Pharmacol 339: , Lerman BB, Belardinelli L, West GA, Berne RM, and Di- Marco JP. Adenosine-sensitive ventricular tachycardia: evidence suggesting cyclic AMP-mediated triggered activity. Circulation 74: , Levi AJ, Mitcheson JS, and Hancox JC. The effect of internal sodium and caesium on phasic contraction of patch-clamped rabbit ventricular myocytes. J Physiol 492: 1 19, Priori SG and Corr PB. Mechanisms underlying early and delayed afterdepolarizations induced by catecholamines. Am J Physiol Heart Circ Physiol 258: H1796 H1805, Reuter H and Beeler GW Jr. Calcium current and activation of contraction in ventricular myocardial fibers. Science 163: , Saito D, Mima T, Uchida S, Obayashi N, Marutani M, Mukohara N, Kusachi S, Tsuji T, and Haraoka S. Interaction between adenosine and -adrenoceptors. Jpn Circ J 55: 81 87, Seitelberger R, Schutz W, Schlappack O, and Raberger G. Evidence against the adenosine-catecholamine antagonism under in vivo conditions. Naunyn Schmiedebergs Arch Pharmacol 325: , Song Y and Belardinelli L. Electrophysiological and functional effects of adenosine on ventricular myocytes of various mammalian species. Am J Physiol Cell Physiol 271: C1233 C1243, Song Y, Thedford S, Lerman BB, and Belardinelli L. Adenosine-sensitive afterdepolarizations and triggered activity in guinea pig ventricular myocytes. Circ Res 70: , Steadman BW, Moore KB, Spitzer KW, and Bridge JNB. A video system for measuring motion in contracting heart cells. IEEE Trans Biomed Eng 35: , Valdeolmillos M and Eisner DA. The effects of ryanodine on calcium-overloaded sheep cardiac Purkinje fibers. Circ Res 56: , Xiao RP and Lakatta EG. 1-Adrenoceptor stimulation and 2-adrenoceptor stimulation differ in their effects on contraction, cytosolic Ca 2, and Ca 2 current in single rat ventricular cells. Circ Res 73: , 1993.