The first published observation that intravascular adenosine causes a temporary heart block dates back to 1927, when Drury and Szent-Gyorgyi from the University of Cambridge, United Kingdom, injected extracts from cardiac tissues intravenously into a whole animal. They noticed a transient decrease of the cardiac rhythm and slowing of the heart rate [1]. Following several purification steps, the authors were able to identify the biologically active compound of the extract as an “adenine compound” [1]. It took almost 50 years from these early discoveries of the heart-rate-slowing effects of “adenine compounds” [1] to the clinical use of adenosine in treating patients with supraventricular tachycardia [2]. As of today, intravenous adenosine has remained a mainstay therapy for diagnosing or treating supraventricular arrhythmias [3].

Adenosine mediates its signaling effects through 4 adenosine receptors (ARs): A1AR, A2AR, A2BAR and the A3AR [4]. Previous studies implicated the A1AR in the heart-rate slowing effects of adenosine. This is based on direct and indirect effects associated with the activation of cardiac A1ARs. Direct A1AR signaling effects are thought to lead to a hyperpolarization of sinus node cells as well as cells of the arterioventricular node (AV node) primarily by inducing a potassium current through an inward rectifier potassium channel (I_KAdo) [5], [6], [7]. Indirect effects of adenosine signaling on the heart rate may involve the ability of the A1AR to induce an “anti-adrenergic” state by opposing the effect of sympathetic nervous activation and β₁-stimulation by lowering intracellular cAMP levels [8]. Based on these finding it was concluded that the A1AR causes the adenosine induced bradycardia. However, these studies are limited to observations based on pharmacologic or genetic in vitro studies in a Langendorff apparatus [9]. At present, comparative genetic in vivo studies have yet to confirm the selective role of the A1AR in adenosine-mediated bradycardia.

To elucidate the contribution of individual ARs to adenosine-induced bradycardia in vivo, we utilized mice with specific deletions of each individual AR. In addition, we confirmed our findings utilizing pharmacological approaches in wild-type or gene-targeted mice for ARs. Consistent with previous ex vivo studies, we found a selective role for the A1AR in adenosine-mediated bradycardia in vivo.

Extracellular adenosine represents an endogenous distress signal, particularly generated under injurious conditions such as hypoxia or ischemia [21], [22]. Under such circumstances, extracellular adenosine signaling plays a critical role in attenuating hypoxia-induced tissue injury and inflammation [4], [23]. It elicits its signaling effects through any of four ARs, which are highly conserved among vertebrates [4]. In addition to its role in hypoxia-adaptation [18], [24], [25] and attenuation of inflammation [26], [27], [28], adenosine is known to cause significant slowing of the heart rate, when injected intravenously. In this study, we investigated the contribution of different ARs to the heart-rate slowing effects of vascular adenosine utilizing genetic and pharmacological in vivo approaches. We found that adenosine caused a dose dependent decrease in heart rate in wild type mice, which was completely abolished in gene-targeted mice for the A1AR or following pre-treatment with the A1AR antagonist DPCPX. In contrast, studies in A2AAR^−/−, A2BAR^−/− or A3AR^−/− mice demonstrated heart-rate slowing following vascular adenosine injection, suggesting a high degree of selectivity for the A1AR-mediated role in slowing of the heart-rate.

The present studies closely relate to studies of central nervous system (CNS) adenosine signaling in the regulation of blood pressure, temperature and sleep-awake cycle. Of particular interest has been the non-specific adenosine receptor antagonist caffeine. As such, caffeine is a CNS stimulant, having the effect of temporarily warding off drowsiness, restoring alertness and elevating blood pressure and heart-rate. In fact, caffeine is the world's most widely consumed psychoactive substance. The question through which adenosine receptors the stimulating effects of caffeine are mediated has been the center of several research studies. First evidence comes from studies in gene-targeted mice for the A2AAR [13]. The authors found that A2AAR^−/− mice showed attenuated exploratory activity, whereas caffeine, which normally stimulates exploratory behavior, became a depressant of exploratory activity. In addition, these mice had higher blood pressure and increased heart-rate [13]. A second study in genetic models confirmed that caffeine mediated arousal involves the A2A and not the A1AR [29]. Only recently, a very elegant and complete study from the research group of Bertil Fredholm extended these findings towards the physiological contributions of A1 and A2A receptors in the regulation of heart rate, body temperature, and locomotion as revealed using knockout mice and caffeine [30]. Here, the authors studied heart rate, body temperature, locomotor activity, and oxygen consumption in awake mice lacking one or both of the A1AR or A2AAR using telemetry and respirometry, before and after caffeine administration. When compared with wild-type littermates, HR was higher in male A1AR^−/− mice but lower in A2AAR^−/− mice. A single dose of an unselective beta-blocker abolished the HR differences between these genotypes, thus indicating that the studied differences involve central nervous system mediated alterations of heart rate and blood pressure. Taken together, these studies indicate that the A2AAR plays an important role in the modulation of oxygen comsumption, locomoter activity by acute and chronic caffeine administration. There is also evidence for effects of higher doses of caffeine being independent of both the A1AR and the A2AAR [30].

Consistent with our studies indicating a role of the A1AR in mediating acute responses to adenosine bolus injections, other studies from Bertil Fredholm's group indicated that the A1AR is involved in the regulation of heart rate, body temperature and locomotor activity, but the magnitude of the involvement is different in males and females [31]. Female mice had higher heart rate, body temperature and locomotion, both during daytime and during the night. Awake A1AR^−/− mice had a slightly elevated heart rate, and this was more clear-cut in males. Heart rate was also higher in Langendorff-perfused denervated A1AR^−/− hearts. Futhermore the bradycardic response to the adenosine receptor agonist 2-chloroadenosine is absent in isolated A1AR^−/− hearts, which was demonstrated in very elegant studies from the research group of John Headrick [9]. At present it remains somewhat unclear how the central and peripheral mechanisms of adenosine receptor dependent regulation of blood pressure and heart rate act in concert.

While there is some overlap between these studies of Bertil Fredholm's group [30], [31] and our present data, the main differences between both studies are the mechanisms of adenosine receptor mediated alterations in blood pressure, heart rate and locomotion. While the studies above mainly focus on the consequences of caffeine – a CNS stimulant that works through inhibition of CNS adenosine receptors [30] or at baseline heart rate levels without adenosine stimulation [31] – the present studies address the role of peripheral adenosine receptors located in the heart in mediating adenosine-dependent slowing of the hear-rate following intravascular adenosine injection. In fact, intravascular injection of adenosine is performed on a routine basis in patients during the treatment of supraventricular tachycardia [3]. As such, the present studies were performed in an in vivo setting following a time interval of 30 s after adenosine receptor activation, while the studies of Bertil Fredholm's group examined non-anesthetized mice over several days [30], [31], or ex vivo Langendorf preparations [31].

Previous studies had shown that the A1AR can influences the ion current in cardiac pacemakers, causing slower atrioventricular nodal conduction, causing bradycardia as seen in this study [6], [7]. The A1AR promoter is highly active in the atrium [32], leading to high A1AR mRNA levels [31]. In fact, expressional levels of the A1AR are higher in the atrium as compared to the ventricle [33], [34]. Furthermore, activation of A1AR by adenosine causes an inhibition of adenylate cyclase, which results in decreased intracellular levels of cAMP [35]. Thus, A1AR activation counteracts the effect of the β₁-adrenoceptor, which activates the adenylate cylcase and increases cAMP levels [8]. In fact, in transgenic mice, overexpressing the A1AR, it was shown that the β-adrenoceptor sensitivity/reactivity was impaired, despite the fact, that there is an enhanced β-adrenoceptor density in these mice [36], [37]. Moreover, isolated perfused hearts from A1AR overexpressing mice had a lower resting heart rate when compared to control animals [38]. Similarly, previous studies hypothesized that the activation of the A1AR may also be a potential cause for bradycardia in patients following cardiac transplantation. In fact, these studies showed that administration of the unspecific adenosine receptor antagonist theophylline leads normalization of the heart rate in cardiac transplant recepients [39]. However, as of to date it seems unclear whether the A1AR has an influence on the heart rate under physiological conditions in vivo. There are reports that A1AR deficient mice have similar heart-rate as wild type controls [20]. In contrast, when hearts from A1AR^−/− mice are isolated, the heart rate measurements reveal a higher heart-rate than in wild-type mice [31]. In the current study in anesthetized mice, we did not see a difference in baseline heart-rate when comparing A1AR^−/− mice to littermate controls. However, since anesthesia sufficiently blocks the influence of sympathetic nervous system on the heart, we cannot draw conclusion on the HR in the alert state. Other reports stated that male A1AR^−/− mice have a higher heart rate than wild type littermates, suggesting that gender might influence A1AR function of regulating HR [30].

Other ARs have also been suggested to influence the heart rate in mice. A2AAR^−/− deficient mice have a higher blood pressure and heart rate when compared to wild type mice [13]. In the present study the response to intravascular adenosine of A2AAR^−/− mice was not altered when compared to that of littermate controls. These findings suggest that the A2AAR does not influence the cardiac conductance system directly in response to systemic adenosine. It seems plausible that A2AAR, rather, influences HR via central nervous system mediated mechanisms [40].

Systemic adenosine, as used in present studies could simultaneously influence two important variables of the blood pressure: cardiac output and vascular resistance. In the present study the intravascular administration of adenosine leads to a decrease in the mean arterial pressure, simultaneously with the observed slowing of the heart-rate. It is important to point out that from the current data, it is difficult to conclude which AR might mediate the decrease in MAP (adenosine-mediated vasodilatation). Both variables (HR and MAP) were altered simultaneously in the experimental protocol. Various studies have assessed the influence of ARs on arterioles and different adenosine receptors might be involved in arteriolar dilatations, however there is evidence that the A2AAR might play the most important role [41], [42], [43], [44].

Taken together, the present studies reveal a critical role of the A1AR to adenosine bolus injection induced bradycardia in vivo. A2AAR, A2BAR and A3AR [7] do not influence the response to intravascular adenosine.