Modulation of Atrioventricular Nodal Function by Metabolic and Allosteric Regulators of Endogenous Adenosine in Guinea Pig Heart
Background There has been increasing interest in the development of agents that utilize endogenous adenosine to exert their actions. We tested the hypothesis that substances that either potentiate the activity (allosteric enhancers) or increase the interstitial concentration (inhibitors of metabolism) of endogenous adenosine may cause event (tachycardia)-specific depression of AV nodal conduction.
Methods and Results The frequency-dependent effects of iodotubercidin (ITU, an inhibitor of adenosine kinase), erythro-9-(2-hydroxy-3-nonyl)adenine (EHNA, an inhibitor of adenosine deaminase), draflazine (a nucleoside transport blocker), and PD81,723 (an allosteric enhancer of the A1 adenosine receptor binding) on the stimulus-to–His bundle (SH) interval, a measure of AV nodal conduction, were determined in guinea pig hearts and compared with those of adenosine and diltiazem. All drugs depressed AV nodal conduction in a frequency-dependent manner. The ratios of SH interval prolongations at fast to slow pacing rates for draflazine, ITU+EHNA, PD81,723, adenosine, and diltiazem were 17.5±3.4, 11.1±5.0, 3.5±0.9, 10.1±2.8, and 8.3±3.5, respectively. Coincident with the prolongation of the SH interval at rapid pacing rates, draflazine and ITU+EHNA increased the epicardial fluid adenosine concentrations by 2.2- and 2.6-fold, respectively. In contrast, epicardial transudate levels of adenosine do not change in the presence of PD81,723. The AV nodal effects of draflazine, ITU, EHNA, and PD81,723 were reversed by the A1 adenosine receptor antagonist 8-cyclopentyltheophylline and adenosine deaminase, implicating endogenous adenosine acting at the A1 adenosine receptor.
Conclusions Adenosine-regulating agents that act in an event- and site-specific manner represent a novel drug design strategy that may potentially be valuable for the long-term treatment of supraventricular arrhythmias and control of ventricular rate during atrial fibrillation or flutter.
The ideal drug used to treat reentrant SVTs involving the AV node should have little or no effect on AV nodal conduction during normal sinus rhythm but should markedly depress AV nodal transmission during episodes of tachycardia. Adenosine, a short-acting and potent depressant of AV nodal conduction1 2 3 4 that exacerbates the modulatory effect of atrial rate5 6 7 on AV nodal conduction delay,8 9 10 11 12 has become a drug of choice for the acute management of SVTs.13 14 15 Unfortunately, however, systemic administration of adenosine or its analogues activates not only cardiac A1 adenosine receptors, which mediate the antiarrhythmic effects of the nucleoside, but also other adenosine receptor subtypes throughout the body. This widespread activation of adenosine receptors produces undesirable effects.
In light of recent studies,16 17 18 19 20 adenosine-regulating agents that either enhance the effects or increase the interstitial concentration of endogenous adenosine may have the potential to avoid such problems but still exert depressant effects on AV nodal conduction. We have previously shown16 17 18 that the A1 adenosine receptor–mediated negative dromotropic effect can be selectively potentiated by an allosteric enhancer, PD81,723, and an inhibitor of nucleoside transport, draflazine. However, it has yet to be determined whether these agents can cause frequency-dependent depression of AV nodal conduction, ie, have minimal to no effect during normal heart rates but promote robust depression of AV nodal conduction during tachycardias. Likewise, the cause-and-effect relation between the adenosine-regulating agents and endogenous adenosine acting at the A1 adenosine receptor has yet to be conclusively demonstrated. To date, only dipyridamole (a nucleoside uptake blocker) has been shown to depress AV nodal conduction in a frequency-dependent manner,19 20 an effect associated with an increase in coronary sinus adenosine levels.20 Regardless, no prior study has investigated how specific elements of the cardiac adenosine system can be pharmacologically modulated to either increase the activity of endogenous adenosine (allosteric enhancers) or raise its concentration (inhibition of adenosine metabolism) to levels sufficient to exert antiarrhythmic effects.
Therefore, we tested the hypothesis that (1) ITU (an inhibitor of adenosine kinase), EHNA (an inhibitor of adenosine deaminase), and draflazine (an inhibitor of nucleoside transport) and (2) PD81,723 (an allosteric enhancer of A1 adenosine receptor binding) will depress AV nodal conduction in a frequency-dependent manner by regulating the concentration and activity of endogenous adenosine, respectively. Adenosine deaminase, the A1 adenosine receptor antagonist CPT, and measurement of interstitial adenosine levels were used to establish a causal relation between endogenous adenosine and frequency-dependent modulation of AV nodal conduction as the underlying mechanism of action of adenosine-regulating agents. The targets of the cardiac adenosine system that provide the rationale for our experiments are illustrated in Fig 1⇓.
ITU, EHNA, CPT, and diltiazem were purchased from Research Biochemicals Inc. ADA was purchased from Sigma Chemical Co. PD81,723 and draflazine were gifts from Discovery Therapeutics Inc and Janssen Research Foundation, respectively.
Isolated Perfused Heart Preparation
Guinea pigs were anesthetized with methoxyflurane and killed by cervical dislocation. Hearts were quickly excised, rinsed in ice-cold modified Krebs-Henseleit solution,8 21 perfused according to the Langendorff technique,22 and instrumented for atrial pacing and recording of the His bundle electrogram as previously described.8 Hearts were allowed to equilibrate for 30 to 45 minutes after placement of electrodes.
Measurement of AVNCT
We have previously shown that adenosine has no effect on the stimulus-to-atrium interval; hence, the SH interval was used as a measure of the effect of adenosine on AV nodal conduction.3 During all experimental protocols, the SH interval was measured in real time (beat to beat) from the His bundle electrogram with an automated, on-line data acquisition program.23 Unless otherwise noted, if any of the interventions caused second-degree AV block, the longest SH interval before the onset of AV block was considered to be the maximal response.
Collection of Epicardial Transudate and Adenosine Assay
Epicardial fluid samples were collected by use of porous disks (Magna nylon membrane filters, MSI) placed on the epicardial surface as described previously.24 The concentration of adenosine in the samples was determined by reversed-phase high-performance liquid chromatography according to a previously published isocratic method.24
Multiple-step stimulation protocol
In this stimulation protocol, atrial pacing cycle length was decreased in 50-ms steps beginning at 400 ms every 60 seconds until an atrial pacing cycle length of 200 ms was reached or second-degree AV block occurred. The effect of progressively faster rates of atrial pacing in the absence and presence of a drug on SH interval were automatically recorded and stored for later analysis.
Simulated atrial tachycardia protocol
In this computer-assisted stimulation protocol, the ACL was abruptly decreased from 300 ms to a cycle length 10% longer than the WCL for a period of 60 seconds, followed by a single-step increase to the original atrial pacing cycle length. The duration of the fast pacing was chosen on the basis of our previous studies, which have shown that in the absence of “AV nodal blocking” agents, a steady-state AV node conduction time is achieved within 60 seconds after the beginning of rapid pacing.8 The ratio of SH interval prolongations at fast and slow rates of atrial pacing (ΔSHfast/ΔSHslow), referred to as the frequency-dependent ratio, was used to determine the rate-dependent effects of drugs on AV nodal conduction (for additional details, see Fig 2⇓). In each heart, a maximum of three simulated atrial tachycardia protocols were performed: (1) control, (2) in the presence of drug(s), and (3) drug(s) plus antagonist. The WCL was determined as described previously.25
Effects of the drugs on AV nodal conduction at a fixed atrial pacing cycle length
The negative dromotropic effect of adenosine-regulating agents was studied in hearts paced at an ACL of 250 ms unless otherwise stated. The effects of the drugs on the SH interval were measured at steady state.
Effects of draflazine and ITU on AV nodal conduction
The time-dependent negative dromotropic effects of draflazine and ITU were studied in the absence and presence of ADA in eight guinea pig hearts. After the baseline recording of the SH interval (5 to 10 minutes), draflazine 0.2 μmol/L or ITU 1 μmol/L was infused until a steady-state prolongation of the SH interval was reached (typically 10 to 15 minutes). Subsequently, ADA 5 U/mL was added to the perfusate for a period of 5 to 10 minutes, and the measurements were repeated.
Effect of PD81,723 on the negative dromotropic effect of adenosine
In this series of experiments, the effect of PD81,723 on the SH interval prolongation caused by bolus injections of adenosine into the perfusate line was determined in the absence and presence of the A1 adenosine receptor antagonist CPT. The hearts (n=4) were paced throughout the experiments at an ACL of 300 ms. After a 2-minute baseline recording of the SH interval, a bolus of adenosine (100 μL of a 200 μmol/L adenosine stock solution) was injected into the perfusion line, and the maximal SH interval prolongation was determined. After a 10-minute washout period, PD81,723 at 5 μmol/L was infused for 15 minutes, and the bolus injection of adenosine and the SH measurement were repeated. After a second 10-minute washout period and still in the presence of PD81,723, CPT 1 μmol/L was infused for 15 minutes, and the above injection and measurement protocol was repeated
Effect of PD81,723 on the negative dromotropic effect of draflazine
In this series of experiments, the effect of PD81,723 on the time-dependent negative dromotropic effect of draflazine was studied in the absence and presence of ADA in four guinea pig hearts. After the baseline recording of the SH interval for a period of 5 to 10 minutes, an infusion of draflazine 0.1 μmol/L was started and continued until a steady-state prolongation of the SH interval was reached. Still in the presence of draflazine, PD81,723 5 μmol/L was added to the perfusate and continued until a new steady-state prolongation of the SH interval was achieved. Subsequently, ADA 5 U/mL was added to the perfusate for 5 to 10 minutes, and the measurements were repeated.
Frequency-dependent effects of drugs on AV nodal conduction
Relationship between ACL and negative dromotropic effect of adenosine, diltiazem, and adenosine-regulating agents
In this series of experiments, the influences of progressively faster rates of atrial pacing on the SH interval prolongation caused by five different drugs were determined in 15 hearts. The multiple-step stimulation protocol was carried out (1) during control conditions (absence of drug), (2) in the presence of drug, and (3) in the presence of drug and CPT 1 μmol/L. The five drugs studied were (1) adenosine 2 μmol/L, (2) diltiazem 0.3 μmol/L, (3) draflazine 0.2 μmol/L, (4) ITU 1 μmol/L plus EHNA 5 μmol/L, and (5) PD81,723 5 μmol/L. Only one drug was used in each heart. The pacing protocol at control conditions was followed by at least a 5-minute recovery period during which the hearts were paced at an ACL of 400 ms before the drug infusion was started. Thereafter, the drug was infused into the perfusate line until steady-state SH interval prolongation was achieved (ie, 10 to 15 minutes), and the pacing protocol was repeated. After the second stimulation protocol and recovery period and still in the presence of drug, CPT 1 μmol/L was added to the perfusate and continued for 10 minutes before the third pacing protocol was initiated.
Effect of adenosine, diltiazem, and adenosine-regulating agents on AV nodal conduction during simulated atrial tachycardia
In this series of experiments carried out in 23 isolated guinea pig perfused hearts, the effects of an abrupt increase in atrial pacing rate (see the simulated atrial tachycardia protocol above) on the SH interval prolongation caused by adenosine 2 μmol/L, diltiazem 0.3 μmol/L, draflazine 0.2 μmol/L, ITU 1 μmol/L plus EHNA 5 μmol/L, and PD81,723 5 μmol/L were determined. The SH interval was measured during slow and fast rates of atrial pacing in the absence of drug (control), during steady-state drug infusion, and in the presence of drug plus CPT 5 μmol/L. The frequency-dependent effects of the drugs were quantified by calculation of the frequency-dependent ratio as described above. In any given heart, only one drug was studied.
Effect of inhibitors of adenosine metabolism on epicardial adenosine concentration during simulated atrial tachycardia
In this series of experiments, the epicardial fluid adenosine levels were measured during the slow and fast pacing periods of the simulated atrial tachycardia protocol in the absence of drug (control) and in the presence of either 1 μmol/L ITU plus 5 μmol/L EHNA or 0.2 μmol/L draflazine. After collection of the control samples, ITU+EHNA or draflazine was infused until a steady state was reached. At that time and once more after a 30-minute washout of the drugs, the collection of the epicardial fluid samples was repeated.
The two-tailed Student's t test was used to analyze paired data. Repeated-measures ANOVA followed by Student-Newman-Keuls testing was used to analyze multiple comparisons among control and interventions. Differences between or among group means were considered significant at a level of P<.05.
Draflazine and ITU Depress AV Nodal Conduction by an Endogenous Adenosine-Mediated Mechanism
Evidence that inhibitors of adenosine metabolism depress AV nodal conduction by a mechanism involving the release of endogenous adenosine is shown in Figs 3⇓ and 4⇓. In a typical example depicted in Fig 3A⇓, the nucleoside transport inhibitor draflazine 0.2 μmol/L increased the SH interval from a baseline value of 44.0 ms to a maximal steady-state value of 56.0 ms at 20 minutes. The addition of ADA 5 U/mL promptly and completely reversed the draflazine-induced 12.0-ms SH interval prolongation. On washout of ADA in the continued presence of draflazine, the SH interval relengthened. Fig 3B⇓ shows the results of four identical experiments. Whereas draflazine prolonged the control SH interval significantly, by 50% (45.0±1.2 to 67.3±8.6 ms), the addition of ADA almost completely reversed the negative dromotropic effect of draflazine (ie, antagonized 91% of the draflazine-induced SH interval prolongation).
Results similar to those of draflazine were obtained with the inhibitor of adenosine kinase ITU. In the typical example depicted in Fig 4A⇑, ITU 1 μmol/L increased the SH interval from a baseline value of 47.0 ms to a maximal steady-state value of 71.0 ms. The addition of ADA 5 U/mL at 15 minutes promptly and almost completely reversed the ITU-induced 24-ms SH interval prolongation. Shown in Fig 4B⇑ are the results of four identical experiments. Whereas ITU significantly prolonged the control SH interval from 49.5±2.8 ms to a maximal steady-state value of 71.5±5.1 ms (ie, an increase of 44%), the addition of ADA reversed the ITU-induced SH interval prolongation by ≈89%.
PD81,723 Potentiates the Negative Dromotropic Effect of Adenosine and Draflazine
In the first series of experiments carried out in guinea pig isolated hearts paced at an ACL of 300 ms, PD81,723 at 5 μmol/L markedly potentiated the negative dromotropic effect of adenosine bolus (100-μL bolus of 200 μmol/L solution) by a mechanism involving the A1 adenosine receptor (Fig 5⇓). As shown in the typical example of Fig 5A⇓, a bolus of adenosine in the absence of PD81,723 (control) increased the SH interval by 13.5 ms, from 51.0 ms to a maximum value of 64.5 ms. Whereas an infusion of PD81,723 5 μmol/L increased the baseline SH interval by only 2 ms (51.0 to 53.0 ms), it increased the maximum SH interval caused by bolus adenosine by 17 ms (64.5 to 81.5 ms), which corresponds to an 8.5-fold increase in SH interval prolongation. The addition of the A1 adenosine receptor antagonist CPT 1 μmol/L completely reversed the SH interval prolongation caused by bolus injections of adenosine as well as the allosteric effects of PD81,723. Similar results from each of four guinea pig hearts are summarized in Fig 5B⇓. Although PD81,723 at 5 μmol/L increased the baseline SH interval by only 1.9 ms (50.8±0.7 to 52.7±0.3 ms; P=.21), it markedly potentiated the maximal negative dromotropic effect of bolus injections of adenosine. That is, in hearts pretreated with PD81,723, boluses of adenosine prolonged the SH interval by 59% (52.7±0.3 to 83.8±1.9 ms), compared with the 33% increase above baseline values (ie, 50.8±0.7 to 67.5±2.3 ms) before treatment with PD81,723. The A1 adenosine receptor antagonist CPT 1 μmol/L completely reversed the effects of PD81,723 and adenosine on AV nodal conduction.
In a second series of experiments carried out in guinea pig isolated hearts paced at an ACL of 250 ms, PD81,723 at 5 μmol/L potentiated the negative dromotropic effect of draflazine (Fig 6⇓). In the typical example shown in Fig 6A⇓, draflazine 0.1 μmol/L gradually increased the SH interval by 5.5 ms, from a baseline value of 35.5 ms to a maximal steady-state value of 41.0 ms. The addition of PD81,723 caused the SH interval to prolong an additional 10.5 ms to a new steady-state value of 51.5 ms. ADA 5 U/mL immediately and completely reversed the negative dromotropic effects of both draflazine and PD81,723. Similar results were obtained in experiments carried out in four separate hearts (Fig 6B⇓). Compared with draflazine, which increased the maximal steady-state SH interval only 12% above baseline values (35.2±0.2 to 39.3±1.2 ms), pretreatment of hearts with PD81,723 further prolonged the maximal steady-state SH interval caused by draflazine to 32% above baseline values (35.2±0.2 to 46.3±3.8 ms). The addition of ADA 5 U/mL almost completely reversed (ie, antagonized 86% of maximal SH interval prolongation) the negative dromotropic effects of draflazine plus PD81,723.
Frequency-Dependent Effects of Drugs on AV Nodal Conduction
Fast pacing exacerbates the AV nodal depressant effects of adenosine and diltiazem
The prolongations of the SH interval caused by 2 μmol/L adenosine (Fig 7A⇓) and 0.3 μmol/L diltiazem (Fig 7B⇓) became progressively greater at incrementally faster rates of atrial pacing. For example, at an ACL of 400 ms, adenosine increased the SH interval above control values by 4.5±1.6 ms, whereas at a cycle length of 200 ms, the increase in the SH interval caused by the same concentration of adenosine was 9.0 ms (ΔSH, 23.5±4.3 to ΔSH, 32.5±3.7 ms). Likewise, at an ACL of 400 ms, diltiazem increased the SH interval above control values by 4.7±1.0 ms, whereas at a cycle length of 200 ms, the increase in SH interval caused by the same concentration of diltiazem was 10.0 ms (ΔSH, 21.0±2.8 to ΔSH, 31.0±2.2 ms) above control values. Both adenosine and diltiazem caused an upward and rightward shift of the ACL versus SH interval relation (Fig 7⇓). Unlike adenosine, the negative dromotropic effects of diltiazem were not reversed by CPT 1 μmol/L.
In a second series of experiments, the depressant effects of adenosine and diltiazem on AV nodal conduction were determined during the simulated atrial tachycardia protocol (Table⇓). During slow rates of atrial pacing (ACL=300 ms), both adenosine 2 μmol/L and diltiazem 0.3 μmol/L caused small but significant (P<.05) increases in the SH interval. Adenosine and diltiazem prolonged the SH interval above control values by 12% and 10%, respectively. In contrast, during rapid heart rates, adenosine and diltiazem increased the SH interval by 67% and 80%, respectively. Thus, both drugs prolonged the SH interval above control values to a much greater extent (about 10-fold) during fast pacing. The negative dromotropic actions of adenosine but not those of diltiazem were antagonized by CPT 1 μmol/L.
Fast pacing exacerbates the AV nodal depressant effects of draflazine and ITU+EHNA
In the experiments using the multiple-step stimulation protocol, draflazine 0.2 μmol/L and ITU 1 μmol/L plus EHNA 5 μmol/L significantly increased the SH interval at all atrial pacing cycle lengths. Draflazine (Fig 8A⇓) and ITU+EHNA (Fig 8B⇓) caused an upward and rightward shift of the curves of ACL versus SH interval. As illustrated in Fig 8⇓, the magnitude of the difference between the control SH interval values and those in the presence of either draflazine or ITU+EHNA was greater at shorter than at longer ACLs. For example, at an ACL of 400 ms, draflazine increased the control SH interval by 2.7±0.9 ms, whereas at an ACL of 200 ms, it increased the SH interval by 7.3 ms (ΔSH, 21.2±2.5 to ΔSH, 28.5±5.6 ms). Similarly, at a cycle length of 400 ms, ITU+EHNA increased the control SH interval by 6.5±0.8 ms, whereas at an ACL of 200 ms, they increased the SH interval by 30.8 ms (ΔSH, 20.2±2.5 to ΔSH, 51.0±7.0 ms). The differences in SH interval between control and drug (draflazine or ITU+EHNA) became significant at ACLs <300 ms. CPT 1 μmol/L almost completely reversed the shift that draflazine and ITU+EHNA caused in the relation of ACL versus SH interval (Fig 8⇓). There was no significant difference between the control SH interval values and those in the presence of the adenosine antagonist.
In a second series of experiments using the simulated atrial tachycardia protocol, draflazine 0.2 μmol/L and ITU 1 μmol/L plus EHNA 5 μmol/L caused a significantly greater increase in the SH interval at faster than at slower rates of atrial pacing (Table⇑). At an ACL of 300 ms, draflazine prolonged the SH interval by 6%, whereas at the end of 1 minute of pacing at an ACL 10% longer than the WCL, it prolonged the SH interval by 66%. Similarly, ITU+EHNA prolonged the SH interval by 11% at a slow rate of pacing, whereas at the end of 1 minute of fast atrial pacing, they prolonged the SH interval by 68%. The addition of CPT 1 μmol/L to the perfusate abolished 73% and 95% of the negative dromotropic effects of draflazine and 98% and 95% of those of ITU+EHNA during slow and fast pacing rates, respectively.
Fast pacing exacerbates the AV nodal depressant effect of PD81,723
In the experiments using the multiple-step stimulation protocol, PD81,723 5 μmol/L caused an upward and rightward shift of the curves of ACL versus SH interval (Fig 9⇓). The differences in SH interval between control and PD81,723 became significant at ACLs ≤250 ms. Most importantly, however, the magnitude of the difference between the control SH interval values and those in the presence of PD81,723 was greater at shorter than at longer ACLs. For example, at an ACL of 400 ms, PD81,723 increased the control SH interval by only 0.5±0.4 ms, whereas at an ACL of 200 ms, it increased the SH interval by 4.7 ms (ΔSH, 10.5±1.5 to ΔSH, 15.2±1.8 ms). CPT 1 μmol/L completely reversed the effect of PD81,723.
The experiments using the simulated atrial tachycardia protocol confirmed that the SH interval prolongation caused by PD81,723 at 5 μmol/L was significantly greater at faster than at slower rates of atrial pacing (Table⇑). At an ACL of 300 ms, PD81,723 prolonged the SH interval by 6%, whereas at the end of 1 minute of pacing at an ACL 10% longer than the WCL, it prolonged the SH interval by 17%. The addition of ADA 5 U/mL to the perfusate abolished 71% and 100% of the negative dromotropic effect of PD81,723 during slow and fast pacing rates, respectively.
Quantification of frequency-dependent AV nodal effect of drugs (frequency-dependent ratio)
The effects of drugs on the SH interval during the simulated atrial tachycardia protocol are summarized in the Table⇑. It should be noted that the drugs (ie, draflazine and ITU+EHNA) whose negative dromotropic action depends on the elevation of endogenous adenosine levels caused frequency-dependent depression of AV nodal conduction of approximately the same magnitude as that caused by diltiazem and adenosine.
Effect of Inhibitors of Adenosine Metabolism and Nucleoside Transport on Epicardial Fluid Adenosine Levels
The effects of draflazine 0.2 μmol/L or ITU 1 μmol/L plus EHNA 5 μmol/L on epicardial fluid adenosine levels were determined during the slow and fast phases of the simulated atrial tachycardia protocol (Fig 10⇓). Coincident with prolongation of the SH interval, draflazine and ITU+EHNA increased epicardial adenosine levels significantly more at fast than at slow heart rates. In the control hearts, rapid atrial pacing did not cause any significant changes in the epicardial adenosine concentration (0.61±0.10 to 0.78±0.12 μmol/L, P=.26). Conversely, pretreatment of hearts with either ITU+EHNA or draflazine increased the epicardial adenosine concentration during rapid atrial pacing significantly, by 2.56-fold (0.86±0.21 to 2.20±0.47 μmol/L) and 2.21-fold (0.91±0.17 to 2.01±0.37 μmol/L), respectively. After washout of ITU+EHNA and draflazine, the concentrations of epicardial adenosine during slow and rapid pacing returned to control values. Note that we have previously shown that the negative dromotropic effect of the A1 adenosine receptor allosteric enhancer PD81,723 is not accompanied by an increase in endogenous adenosine levels.17 18
The major finding of this study was that specific components of the cardiac adenosine system (see Fig 1⇑) can be pharmacologically modulated to regulate the concentration and/or activity of endogenous adenosine in a manner that promotes frequency-dependent depression of AV nodal conduction. Two different but complementary pharmacological approaches to augment the AV nodal effects of endogenous adenosine were tested: (1) The degradation of adenosine was reduced by use of specific inhibitors of adenosine deaminase, adenosine kinase, and nucleoside uptake, and (2) the action of endogenous adenosine was allosterically enhanced by PD81,723. Each approach independently provided compelling and mutually supportive evidence that these “adenosine-regulating agents” depress AV nodal conduction during fast heart rates via a mechanism involving activation of A1 adenosine receptors by endogenous adenosine. These results also validate the value of using adenosine-regulating agents as probes to elucidate the role of endogenously released adenosine as a modulator of cellular and organ function under physiological and pathophysiological conditions.
Regulation of AV Nodal Function by Inhibitors of Adenosine Metabolism
Modulation of the cardiac adenosine system either by draflazine or ITU+EHNA was found to cause frequency-dependent negative dromotropic effects by a mechanism involving endogenous adenosine acting at the A1 adenosine receptor. Three separate lines of evidence support this conclusion. First, the rate-dependent depression of AV nodal conduction by draflazine and ITU+EHNA was associated with a significant increase in epicardial transudate adenosine concentration during fast pacing rates. This finding is in agreement with previous studies that have shown that draflazine, ITU, and EHNA can increase the concentration of interstitial adenosine24 26 27 by inhibiting nucleoside transporter, adenosine kinase, and adenosine deaminase, respectively.18 24 26 27 28 29 30 31 Second, the A1 adenosine receptor antagonist CPT, which is ≈140-fold more selective for the A1 than the A2 receptor,32 completely antagonized the rate-dependent AV nodal effects of draflazine and ITU+EHNA. Third, ADA completely reversed the AV nodal effects of draflazine and ITU. This effect of ADA rules out the possibility that draflazine or ITU may have direct A1 adenosine receptor agonist activity, a possibility that the results of the experiments using the A1 adenosine receptor antagonist CPT cannot rule out. Taken together, the above data provide overwhelming evidence that the frequency-dependent depressant effect of inhibitors of myocardial adenosine metabolism on AV nodal conduction is mediated by endogenously released adenosine acting at the A1 adenosine receptor.
Regulation of AV Nodal Function by an A1 Adenosine Receptor Allosteric Enhancer
Another major finding of this study was that PD81,723, a 2-amino-3-benzoylthiophene derivative previously shown to allosterically enhance the cardiac effects of adenosine mediated by A1 but not A2 adenosine receptors,16 17 18 increased the activity of endogenous adenosine sufficiently to depress AV nodal conduction during rapid heart rates. Like draflazine and ITU+EHNA, PD81,723 depressed AV nodal conduction in a rate-dependent manner by a mechanism involving endogenous adenosine acting at the A1 adenosine receptor. However, as previously shown, the potentiation of A1 adenosine receptor–mediated negative dromotropic effects of exogenous adenosine or hypoxia by PD81,723 is not accompanied by any further increase in epicardial transudate or venous effluent concentrations of adenosine but rather is due to increased binding of adenosine to the cardiac A1 adenosine receptor.17 18
This difference in the mechanism of action of PD81,723 and inhibitors of adenosine metabolism is important in light of the results from previous investigations, which have demonstrated that the A2 adenosine receptor–mediated coronary vasodilation occurs at much lower adenosine concentrations than those required to elicit A1 adenosine receptor–mediated prolongation of AVNCT.9 Consistent with this finding, we found that ITU+EHNA and draflazine caused only minimal (6% to 20%) prolongation of the SH interval at concentrations previously shown to cause maximal increases in coronary conductance.16 21 22 Unlike PD81,723, which is selective for the A1 adenosine receptor,17 18 inhibitors of adenosine metabolism, by relying on the elevation of endogenous adenosine levels for their effects, are likely to indiscriminately activate all subtypes of adenosine receptors (ie, A1, A2A, A2B, and A3) whose affinity for adenosine is within the range of interstitial adenosine concentration achieved. As a consequence, inhibitors of adenosine metabolism may be more likely to cause hemodynamic side effects than selective A1 adenosine receptor allosteric enhancers.
In summary, these and previous findings from our laboratory16 17 18 indicate that allosteric enhancers of the A1 adenosine receptor can be used to selectively and specifically amplify the AV nodal actions of endogenous adenosine. In addition, our results suggest that modulation of AV nodal function by allosteric enhancers of the A1 adenosine receptor may offer distinct advantages over those of metabolic inhibitors.
Modulation of AV Nodal Transmission by Endogenous Adenosine
Substantial evidence indicates that adenosine is released by hypoxic and ischemic myocardium in amounts sufficient to depress AV nodal conduction.1 Conversely, in the normoxic heart, most if not all current evidence suggests that the myocardial interstitial level of adenosine is subthreshold for its negative dromotropic effect. For instance, in the normoxic heart, neither potent and selective A1 adenosine receptor antagonists nor adenosine deaminase significantly shortens AVNCT.33 34 In the present study, the interstitial concentration of adenosine was raised by an increase in cytosolic adenosine with drugs that block the catabolism of adenosine. The concentrations of ITU (1 μmol/L) and EHNA (5 μmol/L) used in our study have previously been shown to maximally inhibit adenosine kinase24 26 and deaminase,26 respectively. Our results confirm those of previous reports that demonstrated that ITU and EHNA significantly increase venous and interstitial dialysate levels of adenosine.24 26 27 28 29 We found that rapid atrial pacing per se did not significantly increase epicardial transudate levels of adenosine. However, this finding alone does not rule out the possibility that in the normoxic isolated heart, production of adenosine is increased during rapid atrial pacing. In fact, our results can be interpreted to indicate that the myocardial production of endogenous adenosine is increased during rapid atrial pacing. The highly efficient conversion of adenosine to AMP mediated by adenosine kinase should be able to keep cytosolic adenosine concentration relatively constant even in the face of slightly elevated adenosine production.26 This interpretation is consistent with the 2.6-fold increase in the epicardial transudate concentration of adenosine above control values in the presence of ITU and EHNA during rapid atrial pacing.
Another line of evidence that suggests that rapid atrial pacing may increase adenosine production was the finding that PD81,723 significantly depressed AV nodal conduction at fast heart rates by a mechanism involving endogenous adenosine acting at the A1 adenosine receptor. PD81,723 is known to potentiate the negative dromotropic effect of exogenous adenosine and hypoxia without causing any further increase in interstitial levels of adenosine.17 18 Hence, the greater AV nodal depression during rapid atrial pacing in hearts treated with PD81,723 strongly suggests that sufficient interstitial adenosine had to be present for this allosteric enhancer to exert its effect. Furthermore, in guinea pig isolated hearts perfused with oxygenated (95% O2) Krebs-Henseleit solution, BW-A1433, a potent A1 adenosine receptor antagonist, significantly shortens the WCL and shifts the relationship between ACL and AVNCT leftward.8 This finding, in conjunction with our observation that the relationship between ACL and AVNCT is shifted rightward by the allosteric enhancer PD81,723, provides the strongest evidence thus far in support of the role of endogenous adenosine as a modulator of rate-dependent AV nodal transmission in the normoxic heart. Whether these observations are specific to the isolated perfused heart preparation remains to be determined. In this regard, it is worth noting that the nucleoside uptake blocker dipyridamole has been shown to significantly prolong the WCL both in guinea pig8 and human heart.19 20
The significantly greater epicardial transudate concentration of adenosine during rapid than during slow pacing rates in hearts treated with inhibitors of adenosine metabolism indicates that myocardial adenosine production can be increased by fast heart rates. It is worth noting that draflazine and ITU+EHNA caused similar increases in epicardial transudate concentration of adenosine and AVNCT during fast atrial pacing and yielded similar frequency-dependent ratios. This finding, together with the reversal by adenosine deaminase of their depressant effects on AV nodal conduction, demonstrates the central role of endogenous adenosine as the mediator of the negative dromotropic actions of inhibitors of adenosine metabolism. The excellent agreement between these two measurements, epicardial adenosine transudate concentration and AVNCT, also supports the validity of using transudate levels of adenosine as an index of interstitial concentrations of this nucleoside.
Potential Therapeutic Implications
The therapeutic value of the adenosine-regulating agents has been tentatively demonstrated in two recent clinical studies in which dipyridamole was found to terminate reentrant SVTs involving the AV node.19 20 In agreement with the present findings, Lerman et al19 and Conti et al20 found that dipyridamole prolonged the AV nodal functional refractory period and WCL. These studies also showed that the effects of dipyridamole were reversed by the nonselective adenosine receptor antagonist theophylline19 20 and that the prolongation of AV nodal refractoriness caused by dipyridamole was associated with significant elevation of coronary sinus (ie, endogenous) adenosine levels.20 When comparing these findings with our results, one should note that the newly developed nucleoside blocker draflazine is more specific than dipyridamole.30 Moreover, as discussed in detail above, the modulation of AV nodal conduction by selective A1 adenosine receptor allosteric enhancers (eg, PD81,723) may offer distinct advantages over the inhibitors of adenosine metabolism. Taken together, the findings that adenosine-regulating agents can raise the concentration or activity of endogenous adenosine to levels that significantly depress AV nodal transmission during rapid atrial pacing (or tachycardia) are interesting and should encourage further development of site- and event-specific drugs for long-term treatment of adenosine-sensitive tachycardias.
Selected Abbreviations and Acronyms
|ACL||=||atrial cycle length|
|AVNCT||=||AV nodal conduction time|
|WCL||=||Wenckebach cycle length|
This work was supported by an Initial Investigator Award of the American Heart Association, Florida Affiliate, to Dr Dennis and National Institutes of Health grant HL-50488 to Drs Belardinelli and Dennis.
- Received April 11, 1996.
- Revision received June 5, 1996.
- Accepted June 7, 1996.
- Copyright © 1996 by American Heart Association
Belardinelli L, Linden J, Berne RM. The cardiac effects of adenosine. Prog Cardiovasc Dis. 1989;22:73-97.
Lerman BB, Belardinelli L. Cardiac electrophysiology of adenosine. Circulation. 1991;83:1499-1509.
Clemo HF, Belardinelli L. Effect of adenosine on atrioventricular conduction, I: site and characterization of adenosine action in the guinea pig atrioventricular node. Circ Res. 1986;59:427-436.
Belardinelli L, Lerman BB. Electrophysiological basis for the use of adenosine in the diagnosis and treatment of cardiac arrhythmias. Br Heart J. 1990;63:3-4.
Merideth J, Mendez C, Mueller WJ, Moe GK. Electrical excitability of atrioventricular nodal cells. Circ Res. 1968;23:69-85.
Bilette J, Janse MJ, van Capelle FJL, Anderson RH, Touboul P, Durrer D. Cycle-length-dependent properties of AV nodal activation in rabbit hearts. Am J Physiol. 1976;231:1129-1139.
Loeb JM, DeTarnowski JM, Whitson CC, Warner MR. Atrioventricular nodal accommodation: rate and time-dependent effects. Am J Physiol. 1987;252:H578-H584.
Jenkins JR, Belardinelli L. Atrioventricular nodal accommodation in isolated guinea pig hearts: physiological significance and role of adenosine. Circ Res. 1988;63:97-116.
Belardinelli L, Shryock JC. Does adenosine function as a retaliatory metabolite in the heart? News Physiol Sci. 1992;7:52-56.
Nayebpour M, Billette J, Amellal F, Nattel S. Effects of adenosine on rate-dependent atrioventricular nodal function: potential roles in tachycardia termination and physiological regulation. Circulation. 1993;88:2632-2645.
Lai W-T, Chee-Siong L, Wu S-N. Rate-dependent properties of adenosine-induced negative dromotropism in humans. Circulation. 1994;90:1832-1839.
DiMarco JP, Miles W, Akhtar M, Milstein S, Sharma AD, Platia E, McGovern B, Scheinman MM, Govier WC. Adenosine for paroxysmal supraventricular tachycardia: dose ranging and comparison with verapamil. Ann Intern Med. 1990;113:104-110.
Amoah-Apraku B, Xu J, Pelleg A, Belardinelli L. Selective potentiation by an A1-adenosine selective enhancer of the negative dromotropic action of adenosine in the guinea pig heart. J Pharmacol Exp Ther. 1993;266:611-617.
Kollias-Baker C, Ruble J, Dennis D, Bruns RF, Linden J, Belardinelli L. Allosteric enhancer PD81,723 acts by novel mechanism to potentiate cardiac actions of adenosine. Circ Res. 1994;75:961-971.
Kollias-Baker C, Xu J, Pelleg A, Belardinelli L. Novel approach for enhancing atrioventricular nodal conduction delay mediated by endogenous adenosine. Circ Res. 1994;75:972-980.
Lerman BB, Wesley RC, Belardinelli L. Electrophysiological effects of dipyridamole on atrioventricular nodal conduction and supraventricular tachycardia: role of endogenous adenosine. Circulation. 1989;80:1536-1543.
Conti JB, Curtis AB, Utterback D, Belardinelli L. Endogenous adenosine is an antiarrhythmic agent. Circulation. 1994;91:1761-1769.
Krebs HA, Henseleit K. Untersuchungen u¨ber die Harnstoffbildung im Tierko¨rper. Hoppe Seylers Z Physiol Chem. 1932;210:333-336.
Tietjan CS, Tribble CG, Gidday JM, Phillips CL, Belardinelli L. Interstitial adenosine in guinea pig hearts: an index obtained by epicardial disk. Am J Physiol. 1990;259:H1471-H1476.
Kroll K, Decking UKM, Deikorn K, Schrader J. Rapid turnover of the AMP-adenosine metabolic cycle in the guinea pig heart. Circ Res. 1993;73:846-856.
Wang T, Mentzer RM, Van Wylen DGL. Interstitial adenosine with dipyridamole: effect of adenosine receptor blockade and adenosine deaminase. Am J Physiol. 1992;32:H552-H558.
Van Belle H. Specific metabolically active antiischemic agents: adenosine and nucleoside transport inhibitors. In: Singh B, Dzan V, Vanhoutte P, Woosley R, eds. Cardiovascular Pharmacology and Therapeutics. New York, NY: Churchill-Livingstone Inc; 1993:217-245.
Bruns RF, Lu GH, Pugley TA. Adenosine receptor subtypes: binding studies. In: Becker B, Gerlach E, eds. Topics and Perspectives of Adenosine Research. Berlin, Germany: Springer-Verlag; 1987:59-73.
Belardinelli L, Shryock J, West GA, Clemo HF, DiMarco JP, Berne RM. Effects of adenosine and adenine nucleotides on the atrioventricular node of isolated guinea pig hearts. Circulation. 1984;70:1083-1091.
Dennis D, Jacobson K, Belardinelli L. Evidence of spare A1-adenosine receptors in guinea pig atrioventricular node. Am J Physiol. 1992;262:H661-H671.