Endogenous adenosine-the product of degradation of ATP-acts through several receptor subtypes: A1, A2A, A2B and A3 (1). Among them the adenosine A1 receptors have a variety of important functions both in physiology and pathophysiology. They are related to: regulation of sleep, arousal, vasodilatation and neuroprotection in hypoxia and ischaemia (2). Several published findings demonstrate that central A1 receptors contribute to the control of breathing. Their activation in the brainstem slices from neonatal rats depresses respiratory activity. This corresponds to a decrease in the amplitude of the integrated activity of inspiratory neurones and prolongation of the respiratory cycle, primarily by lengthening inspiratory time (3). In anaesthetized rats and cats, adenosine or A1 receptor analogues applied to the cerebral ventricles depress pulmonary ventilation decreasing both the frequency of breathing and tidal volume (4 - 6). Furthermore, the release of adenosine and stimulation of A1 receptors in the central nervous system during hypoxia have been suggested to suppress breathing (7). These results indicate that a central adenosine action is to diminish respiratory function.

In the peripheral nervous system of rats, A1 receptors have been identified on vagal afferent nerves and on the neurones of the nodose ganglia (8). In this species adenosine administered systemically elicits a delayed apnoea, followed by breathing of increased tidal volume, bradycardia and hypotension. The respiratory arrest was ascribed to the activation of pulmonary C-fibres, since perineural blockade of the cervical vagi with capsaicin prevented the apnoea (9). Furthermore, right atrial injection of adenosine excites vagal C-fibres via A1 receptors in the rat lung (10), whereas given via an intralingual route it stimulates carotid body chemoreceptor discharges in cats. Unfortunately respiratory parameters were not recorded at the same time (11). In vagotomized and glomectomized cats adenosine injected systemically acts as a central respiratory depressant (6). An intravenous adenosine challenge in humans induces dyspnoea and increased ventilation and heart rate, and these effects were attributed to the excitation of vagal C-fibres (12). Conflicting evidence was reported in an earlier study in lung-denervated humans, showing that adenosine infusion augmented minute ventilation via stimulation of peripheral chemoreceptors (13).

The contribution of peripheral adenosine A1 receptors to respiratory reflexes is far from clear. Presumably, excitation of receptors located on vagal afferents and nodose ganglia might have a modulatory effect on respiratory pattern.

Therefore, the present experiments were performed to determine if the activation of peripheral adenosine A1 receptors with the selective agonist N6-cyclopentyladenosine (CPA) would contribute to the respiratory pattern and cardiovascular effects, and to what extent the responses depend on lung and nodosal vagal afferentations. We measured the ventilatory effects of CPA in neurally intact rats and after the elimination of vagal input from below and above the nodose ganglia. To verify if post-CPA decreased ventilation is caused by excitation of A1 receptors, we examined the effect of A1 blockade with DPCPX on the respiratory and cardiovascular responses.


MATERIALS AND METHODS
Animals and surgical procedures

Ethical approval for the experimental procedures used in this study was obtained from the local committee. All animal procedures were in accordance with the NIH Guide for the Care and Use of Laboratory Animals. 47 adult male Wistar rats weighing 200-300 g were anaesthetized with an intraperitoneal (i.p.) injection of 600 mg kg-1 urethane (Sigma) and 120 mg kg-1 alpha-chloralose (Fluka AG). The animals were placed in the supine position and they breathed spontaneously with room air. The trachea was exposed in the neck, sectioned below the larynx and cannulated. Catheters were inserted into the femoral vein for drug administration and supplemental doses of anaesthetic and to the femoral artery for blood pressure monitoring. Rectal temperature was maintained close to 36 - 38°C by a heating pad.

To investigate whether peripheral input from the lung vagi and nodose ganglia would contribute to the effects of CPA, two neurotomies were carried out. In the first series of experiments midcervical vagi were bluntly dissected and prepared for bilateral vagotomy, before measuring the respiratory variables in neurally intact rats. In the second experimental series, the rostral parts of the midcervical vagi were separated from the superior cervical ganglia. The nodose ganglia were bluntly dissected from the surrounding tissue; their blood supply was preserved intact. The supranodose vagi were prepared for section performed 2 mm distal from the rostral poles of the ganglia, before measuring the respiratory variables in neurally intact rats. In the third experimental group, the cervical vagi were sectioned at the beginning of the experiment, before CPA injection and blockade with DPCPX.

Apparatus and measurements

Tidal volume signals were recorded with a pneumotachograph head attached to the tracheal cannula linked to a Research Pneumotach System (RSS 100 HR, Hans Rudolph inc.) and a computerized recording system (Windows software version 3.07.02, KORR Medical Technologies Inc.) for measuring and recording respiratory frequency (f), tidal volume (VT), respiratory minute volume (VE), inspiratory (TI) and expiratory (TE) times. Arterial blood pressure was measured with a BP-2 blood pressure monitor (Columbus Instruments). The recordings were registered on an Omnilight 8 M 36 apparatus (Honeywell).

Drugs

N6-cyclopentyladenosine (CPA, Sigma) at a dose of 5 µg kg-1 dissolved in 0.9% saline was injected as a bolus to the femoral vein. 1, 2- dipropyl-8-cyclopentylxanthine (DPCPX, Sigma) at a dose of 100 µg kg-1 (14), dissolved in dimethyl sulphoxide (DMSO) and subsequently diluted in saline (containing no more than 0.01% DMSO), was injected intravenously. Each drug bolus was immediately flushed with an 0.2 ml aliquot of saline.

Treatment schedule and groups

The respiratory effects of the stimulation of A1 receptors were tested using single CPA boluses in the following experimental designs: (1) in neurally intact rats and then midcervically vagotomized (n = 9); (2) in neurally intact rats and after section of the supranodose vagi (n = 14); (3) before and after the blockade of A1 receptors with the antagonist DPCPX, in rats with transected midcervical vagi (n = 9).

The baselines of each individual value of VT, minute ventilation (VE) and respiratory rate (f) were determined by averaging the measured variables for five respiratory cycles before injection. The ventilatory parameters were derived from the integrated pneumotachograph signal just before CPA injection, at the immediate post-CPA phase, and 30 s, 60 s and 2 min after the challenge.

Data analysis

All experimental data were analyzed by repeated measures two-way ANOVA with time (pre-challenge and defined time points after challenge) and either innervation status (neurally intact and vagi cut, or neurally intact and nodose vagi cut) or DPCPX pretreatment (yes or no) as a between condition factor. Differences between individual time points and experimental situations were evaluated by planned contrast analysis and Tuckey's post-hoc test. In all cases, P<0.05 was considered significant. All results shown are means ± SEM


RESULTS
Effect of midcervical vagotomy on CPA-induced pattern of respiration

The dose of the drug used in our study was derived from preliminary dose-response experiments. Out of three doses tested (ranging from 1-7.5 µg kg-1) 5 µg kg-1 proved to be most effective and reasonably well tolerated.

In the neurally intact rats, intravenous CPA challenge produced uniform cardiorespiratory effects, mainly comprising significant decreases in VT, respiratory rate and VE, and a significant drop in arterial blood pressure concomitant with bradycardia.

The injection of an equal volume (0.2 ml) of the vehicle (0.9% saline or DMSO-saline solution) resulted in no respiratory change irrespective of the neural state (not shown).

As shown in Table 1, CPA promptly evoked a significant stable decrease in VT from the immediate post-injection phase to the later time points, and this applied to neurally intact and vagotomized rats (ANOVA, P<10-4). CPA concomitantly produced a significant reduction in respiratory rate both before and after midcervical vagotomy (ANOVA, P<10-4). Vagotomy by itself produced significant increase in mean control value of VT (P<0.001). Minute ventilation was decreased due to CPA (ANOVA, P<0.001). The decrease in VE appeared in both neural states after CPA injection and persisted at 2 min post-challenge. The largest reduction by 29 % was present at 1 min in the intact rats (P<0.01) and of 35% at 30 s after midcervical vagotomy (P<0.001).

Table 1. Changes in tidal volume (VT), respiratory rate and minute ventilation (VE) after intravenous CPA challenge in neurally intact and midcervically vagotomized rats (n = 9).

Two-way ANOVA showed: significant effect of time (P<10-4) and time x vagotomy interaction (P<0.01) but no effect of vagotomy (P=0.2) on VT; significant effect of time (P<10-4) but neither vagotomy (P=0.13) nor time x vagotomy interaction effect (P=0.2) on respiratory rate; a significant effect of time (P<10-5) and time x vagotomy interaction (P=0.04) but no effect of vagotomy (P=0.7) on VE. All values are means ± SEM* P<0.05, **P<0.01,***P<0.001 vs. the respective pre-CPA value. #P<0.05, ##P<0.01 vs. the corresponding pre-vagotomy value.

Ventilatory response to CPA after supranodose vagotomy

To determine, whether the nodose ganglia contribute to the response to the CPA challenge, we tested the effect of section of the supranodose vagi on the respiratory effects of the drug in intact rats. Cutting the central vagal trunk above the nodose ganglia increased baseline tidal volume (P<0.01) and lowered the frequency of breathing (P<0.001). In the current experiments this neurotomy showed a significant interactive effect with CPA challenge on the rate of breathing (ANOVA, P = 10-5), but not on tidal volume (ANOVA, P = 0.25). As shown in Fig.1, following nodosectomy CPA still evoked depression of tidal volume but the decrease in frequency of breathing was eliminated compared with the intact state. Minute ventilation was decreased in both neural states, being apparently more profound in the intact animals (Fig. 1C). In these latter, mean values of VE after CPA challenge were lowered by 29 % (P<0.001) and scarcely reduced by 12 % (P<0.05) after supranodosal vagotomy, compared with the respective baseline values.

Fig. 1. Effects of i.v. administration of CPA and section of supranodose vagi on VT (A), respiratory rate (B) and minute ventilation (C) in the intact rats (n = 12). ***P<0.001, ** P<0.01, *P<0.05 versus the respective pre-CPA baseline value. ###P<0.001, ## P<0.01, #P<0.05 versus the corresponding pre-NG section value.

Ventilatory response to CPA after DPCPX pre-treatment

To test whether CPA-induced effects are specific for excitation of A1 receptors, we used the highly selective A1 antagonist DPCPX in rats subjected to midcervical vagotomy. We examined the response to i.v. CPA challenge in the control conditions and 2 min after pre-treatment with DPCPX. As shown in Fig. 2, the DPCPX-mediated blockade of A1 receptor prevented the CPA-evoked decreases in tidal volume and respiratory rate. Significant inhibition of minute ventilation produced by CPA (from the baseline of 220.8 ± 26 to 157.7 ± 20 ml min-1) occurring before the blockade (P<0.001) was eliminated after inactivation of A1 adenosine receptors with DPCPX. The mean baseline value of VE in rats pretreated by DPCPX was 211.8 ± 15 ml min-1 and remained at 214.1 ± 14 ml min-1 after CPA challenge (P = 0.9).

Fig. 2. Effect of adenosine A1 receptor blockade on post-CPA changes in tidal volume (A), breathing rate (B) and minute ventilation (C) in midcervically vagotomized rats (n = 9). Note that DPCPX treatment prevented the decreases in VT, VE and respiratory rate. ***P<0.001, ** P<0.01, *P<0.05 versus the respective pre-CPA baseline value. ###P<0.001, ## P<0.01, #P<0.05 versus the corresponding post-DPCPX value.

Cardiovascular effects

Fig. 3 and Table 2 include data on cardiovascular effects of CPA in all experimental conditions. Mean arterial blood pressure (MAP) displayed a long-lasting fall (ANOVA, P<10-6) after a CPA challenge, attaining the lowest level at 30 s post-challenge, declining from the pre-drug values of 106.7 ± 4.5 to 41.4 ± 3.3 mmHg (P<0.001) in the intact rats, and from 109.8 ± 10 to 44.0 ± 4.5 mmHg treated by midcervical vagotomy (P<0.001). Two minutes after the CPA injection, MAP began to return to the baseline, reaching it at 15 min post-drug.

Fig. 3. Effects of CPA on mean arterial blood pressure (MAP) in (A) neurally intact and midcervically vagotomized rats (n = 7), (B) intact and treated by supranodosal vagotomy (n = 12) and (C) midcervically vagotomized prior to and after DPCPX blockade (n = 8). Note that DPCPX treatment abolished CPA-evoked decrease in blood pressure. ***P<0.001, ** P<0.01, *P<0.05 versus the respective pre-CPA baseline value. ###P< 0.001, ## P< 0.01, versus the corresponding post-NG section and post-DPCPX value.

Table 2. Effects of CPA on heart rate (HR) in (a) neurally intact and midcervically vagotomized rats (n = 7), (b) intact and treated by supranodosal vagotomy (n = 14) and (c) midcervically vagotomized prior to and after DPCPX blockade (n = 8).

Two-way ANOVA showed: significant effect of time (P<10-6) but not of vagotomy (P=0.9) nor a time x vagotomy interaction (P=0.9) on HR in intact and vagotomized rats; and significant effect of time (P<10-6) and vagotomy (P<0.05) but no time x vagotomy interaction effect (P=0.2) in the intact and NG cut rats; significant effect of time (P<10-6), time x vagotomy interaction (P<10-6) and effect of blockade (P<0.001) on HR prior to and after DPCPX. All values are means ± SEM* P<0.05, **P<0.01,***P<0.001 vs. the respective pre-CPA value. #P<0.05, ###P<0.001 vs. the corresponding pre-vagotomy or pre-DPCPX value.

Bradycardia appeared immediately after CPA injection, both before and after the section of midcervical vagi, showing the maximal fall between 30 and 60 s. At this latter time point heart rate started to increase slowly, returning close to the control values within 15 min. As displayed in Table 2 and Fig. 3, CPA evoked similar decreases in blood pressure and heart rate in rats subjected to supranodose vagotomy.

Blockade with DPCPX significantly affected blood pressure response (ANOVA, P = 0.0004). As indicated in Table 2 and Fig. 3 DPCPX prevented the hypotension and bradycardia that formerly occurred after CPA injection.


DISCUSSION
The present study has shown that in anaesthetized rats the predominant effect of an intravenous injection of CPA was a short-lived depression of respiratory rate and tidal volume and prolonged fall in blood pressure and bradycardia. The same pattern of response was reported by Hedner et al. (4) after intracerebroventricular challenge with N6 (L-2-phenylisopropyl) adenosine (PIA) in rats.

Of the neural conditions tested in the current experiments midcervical vagal deafferentation of the lungs had no influence on the respiratory changes evoked by CPA. This observation is in general agreement with the results obtained in vagotomized and glomectomized cats, which responded with the same pattern either to the intracerebroventricular or intravenous injection of PIA (15).

The selective adenosine A1 receptor agonist-CPA applied by us has not been largely used in the experimental animals. Scant reports have evidenced the decrease in VT on intramuscular injection in awake rhesus monkeys (16) and no visible respiratory effects on intravenous infusion of CPA in fetal sheep (17). The present study revealed an immediate post-CPA decrease in VT and f both before and after midcervical vagotomy. This result is in contrast to the report on i.v. adenosine in rats, which evoked a short apnoea as the only respiratory response, effectively abolished by the blockade of A1 receptors and pulmonary C-fibres (9). Yet, the occurrence of the respiratory depression following midcervical vagotomy in our experiments infers that CPA might have affected the volume and timing components of the breathing pattern through the afferent neurons of the nodose ganglia.

This is the first study assessing how much the nodose ganglia contribute to cardiorespiratory effects of adenosine A1 agonist. In rats subjected to nodose ganglionectomy, CPA administration did not affect the respiratory rate, the decline in minute ventilation was significantly lessened, compared to the intact state, while the drop in tidal volume and cardiovascular effects were still preserved.

We can therefore surmise that at least one component of respiratory response- the frequency of breathing is mediated by receptors located on the vagal nodose ganglia. This conclusion is reinforced by the visualization of A1- radioligand binding reported in the neurones of rat nodose ganglia (8, 18).

Considering that CPA showed average transport through the blood-brain barrier (19), its ability to decrease VT following the ablation of supranodosal pathway might imply the central origin of the response. It is further supported by the observation in cats, that adenosine given intracellularly to the neurons of the ventral respiratory group depresses their activity, which is reversed by blockade of A1 receptors with DPCPX (7). Adenosine analogs -CPA and L-PIA showed a similar depressant action on the inspiratory drive in the fetal sheep and neonatal rats, in these latter were blocked by the selective A1 receptor antagonist (20, 21).

In the current experiments DPCPX, the selective A1 antagonist eliminated completely the depression of breathing, fall in mean arterial blood pressure and heart rate caused by CPA, showing that adenosine A1 receptors contribute to the cardiorespiratory response. Since DPCPX crosses easily blood-brain barrier (22), we can speculate that the decrease in tidal volume described above, might have resulted from stimulation of the central A1 receptors. These were described to be widespread throughout the whole brain (2, 23).

The mechanism of adenosine-evoked cardiovascular response has been the subject of a considerable investigation. Intravenous or intracerebroventricular injection of adenosine and A1 agonists (L-PIA and CHA) has been shown to produce consistent vasodilatation, hypotension and bradycardia attributed to the release of other neurotransmitters (4, 24). Our results are in line with previous findings and more recent report on the effects of intraperitoneal CPA-evoked prolonged fall in blood pressure and heart rate, with a clear contribution of peripheral A1 receptors in rats (25).

Since in our experiments the cardiovascular response still occurred after the consecutive transection of the lung vagi and nodose ganglia, it points again to the central action of CPA. However, A1 receptors located in the heart could reasonably be of importance (26).

In conclusion, this study has shown that i.v. adenosine agonist-N6-cyclopentyladenosine challenge depresses ventilation due to the decrease in tidal volume and respiratory rate. The hypoventilation is mediated via excitation of adenosine A1 receptors outside the lung vagi, with contribution those of nodose ganglia to the timing component of the breathing pattern. Predominant effect on tidal volume and cardiovascular inhibition relays on the central A1 neuronal mechanisms.

Acknowledgements: Mrs. Teresa Warnawin is thanked for her excellent technical assistance.

Conflicts of interest statement: None declared.


REFERENCES

1. Bivalacqua TJ, Champion HC, Lambert DG, Kadowitz PJ. Vasodilator response to adenosine and hyperemia are mediated by A1 and A2 receptors in the cat vascular bed. Am J Physiol Regulatory Integrative Comp Physiol 2002; 282: R1696-R1709.
2. Dunwiddie TV, Masino SA. The role and regulation of adenosine in the central nervous system. Annu Rev Neurosci 2001; 24: 31-55.
3. Wang JL, Wu ZH, Wang NQ. Adenosine A1 receptors are involved in the modulation of the rhythmical respiration in neonatal rat brainstem slice in vitro. Acta Physiol Sinica 2005; 57: 91-96.
4. Hedner T, Hedner J, Jonason J. Regulation of breathing in the rat: indications for a role of central adenosine mechanisms. Neurosci Lett 1982; 33: 147-151.
5. Barraco RA, el-Ridi MR, Parizon M. The adenosine analog, 5'-N-ethylcarboxamidoadenosine, exerts mixed agonist action on cardiorespiratory parameters in the intact but not decerebrate rat following microinjections into the nucleus tractus solitarius. Brain Res 1990; 530: 54-72.
6. Eldridge FL, Millhorn DE, Kiley JP. Antagonism by theophylline of respiratory inhibition induced by adenosine. J Appl Physiol 1985; 59: 1428-1433.
7. Schmidt C, Bellingham MC, Richter DW. Adenosinergic modulation of respiratory neurones and hypoxic responses in the anaesthetized cat. J Physiol 1995; 483.3: 769-781.
8. Lawrence AJ, Krstew E, Jarrott B. Complex interactions between nitric oxide and adenosine receptors in the rat isolated nodose ganglion. Eur J Pharmacol 1997; 328: 83-88.
9. Kwong K, Hong JL, Morton RF, Lee LY. Role of pulmonary C fibers in adenosine-induced respiratory inhibition in anesthetized rats. J Appl Physiol 1998; 84: 417-424.
10. Hong JL, Ho CY, Kwong K, Lee LY. Activation of pulmonary C fibres by adenosine in anaesthetized rats: role of adenosine A1 receptors. J Physiol 1998; 508(1): 109-118.
11. McQueen DS, Ribeiro JA. Effect of adenosine on carotid chemoreceptor activity in the cat. Br J Pharmacol 1981; 74: 129-136.
12. Burki NK, Dale WJ, Lee LY. Intravenous adenosine and dyspnea in humans. J Appl Physiol 2005; 98: 180-185.
13. Morgan-Hughes NJ, Corris PA, Healey MD, Dark JH, McComb JM. Cardiovascular and respiratory effects of adenosine in humans after pulmonary denervation. J Appl Physiol 1994; 76: 756-759.
14. Walsh MP, Marshall JM. The role of adenosine in the early respiratory and cardiovascular changes evoked by chronic hypoxia in the rat. J Physiol 2006; 575(1): 277-289.
15. Eldridge FL, Millhorn DE, Kiley JP. Respiratory effects of a long-acting analog of adenosine. Brain Res 1984; 301: 273-280.
16. Howell LL, Morse WH, Spealman RD. Respiratory effects of xanthines and adenosine analogs in rhesus monkeys. J Pharmacol Exp Ther 1990; 254: 786-791.
17. Koos BJ, Maeda T, Jan C. Adenosine A1 and A2A receptors modulate sleep state and breathing in fetal sheep. J Appl Physiol 2001; 91: 343-350.
18. Krstew E, Jarrott B, Lawrence AJ. Autoradiographic visualization of axonal transport of adenosine A1 receptors along the rat vagus nerve and characterization of adenosine A1 receptor binding in the dorsal vagal complex of hypertensive and normotensive rats. Brain Res 1998; 802: 61-68.
19. Schaddelee MP, Voorwinden HL, Groenendaal D, et al. Blood-brain barrier transport of synthetic adenosine A1 receptor agonists in vitro: structure transport relationships. Eur J Pharm Sci 2003; 20: 347-356.
20. Bissonnette JM, Hohimer AR, Knopp SJ. The effect of centrally administered adenosine on fetal breathing movements. Respir Physiol 1991; 84: 273-285.
21. Dong XW, Feldman JL. Modulation of inspiratory drive to phrenic motoneurons by presynaptic adenosine A1 receptors. J Neurosci 1995; 15: 3458-3467.
22. Bisserbe JC, Pascal O, Deckert J, Maziere B. Potential use of DPCPX as probe for in vivo localization of brain A1 adenosine receptors. Brain Res 1992; 599: 6-12.
23. Dixon AK, Gubitz AK, Sirinathsinghji DJ, Richardson PJ, Freeman TC. Tissue distribution of adenosine receptor mRNAs in the rat. Br J Pharmacol 1996; 118(6): 1461-1468.
24. Stella L, Berrino L, Maione S, de Novellis V, Rossi F. Cardiovascular effects of adenosine and its analogs in anaesthetized rats. Life Sci 1993; 53: 755-763.
25. Schindler CW, Karcz-Kubicha M, Thorndike EB, et al. Role of central and peripheral adenosine receptors in the cardiovascular responses to intraperitoneal injections of adenosine A1 and A2 subtype receptor agonists. Br J Pharmacol 2005; 144: 642-650.
26. Shryock JC, Belardinelli L. Adenosine and adenosine receptors in the cardiovascular system: biochemistry, physiology, and pharmacology. Am J Cardiol 1997; 79(12A): 2-10.