INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

ADENOSINE A₂ receptors are regarded as coupling to stimulation of adenylate cyclase activity and the production of cAMP (8) and to mediate the dilator response of cerebral arterioles to adenosine (6, 24). Adenosine level increases in the brain with moderate hypotension, and its increment is related to the regulation of cerebral blood flow (CBF) (33). The adenosine A₂ receptors are further classified as A_2A and A_2B subtypes based on the receptor affinity (2). Cloning techniques have confirmed the existence of distinct A_2A and A_2B subtypes of A₂ receptor (29). Although the precise physiological functions of the A_2B receptors remain undefined, roles for the A_2B receptors have been suggested in the regulation of neuroglia functions (7), myocardial contractility (18), vascular smooth muscle tone (20), and intestinal chloride secretion (30).

Because of the lack of specific agonists or antagonists for the A_2B-adenosine receptor, little is known about the physiological role for the A_2B receptors in cerebral autoregulation. Brackett and Daly (1) measured adenosine-evoked accumulations of cAMP in cultured NIH 3T3 fibroblasts showing the A_2B-subtype receptor. In this cell line, 5'-N-ethylcarboxamido-adenosine (NECA) was highly selective for the A_2B receptors, whereas 2-p-(2-carboxyethyl)-phenethylamino-5'-N-ethylcarboxamido adenosine (CGS-21680) (19), which is highly selective for A_2A-subtype receptors in the pheochromocytoma PC12 membranes, has a much lower potency in the fibroblasts. CGS-21680 was demonstrated to be selective for the A_2A- compared with the A_2B-subtype receptors in the rat hippocampus and striatum (19). Of the adenosine receptor antagonists, alloxazine has been demonstrated to be selective for the A_2B-subtype receptor with a selectivity of about ninefold for the NIH 3T3 fibroblasts over the PC12 A_2A receptor (1), whereas ZM-241385 is highly selective for the A_2A subtype (24). According to the functional characterization of adenosine receptors in the vascular beds, NECA was also demonstrated to be a relatively selective agonist for the adenosine A_2B receptors in the guinea pig aorta (10) and the rat mesenteric artery (28).

Recently, Hong et al. (13) demonstrated that when the cortical surface is suffused with cAMP, the release of adenosine is increased in the artificial cerebral spinal fluid (CSF), and they suggested that the cAMP-adenosine pathway as a viable metabolic mechanism is implicated in the production of adenosine in the rat pial artery and contributes to the regulation of vasodilation in response to hypotension.

On the other hand, Martin and Potts (21) documented that the rat renal artery contains A_2B-adenosine receptors that are located on the endothelium, and the A_2B receptors cause release of nitric oxide. Moreover, nitric oxide and nitric oxide synthase expression have been evidenced in the cAMP-induced pial artery vasodilation (26). However, to our knowledge, the role of the adenosine A_2B-receptor subtype has not been clearly characterized in the pial arteries in relation to a physiological role such as the vasodilator response to acute hypotension and CBF autoregulation.

We therefore aimed to examine the effects of L-NAME on the vasodilation and nitrite/nitrate formation evoked by adenosine and NECA (adenosine A_2B-receptor predominant agonist), respectively, in comparison with those elicited by CGS-21680 (adenosine A_2A-receptor predominant agonist). Thereafter, we determined whether the lower limit of CBF autoregulation was shifted or not under pretreatment with ZM-241385 (A_2A-receptor antagonist), alloxazine (A_2B-receptor antagonist), and L-NAME (nitric oxide synthase inhibitor).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Preparation of animals. Male Sprague-Dawley rats (250-320 g) were anesthetized with urethan (1 g/kg ip) and placed on a heating pad (Homeothermic Blanket System, Harvard Apparatus, South Natick, MA) to maintain a constant rectal temperature (37 ± 0.5°C). After a tracheostomy, the rat was mechanically ventilated with room air by a respirator (model 683, Harvard Apparatus) after immobilization with 5 mg/kg gallamine triethiodide. The mean PCO₂ was monitored with an end-tidal CO₂ analyzer (CapStar-100, IITC Life Science, Woodland Hills, CA). Catheters were placed in a carotid artery for measurement of systemic arterial blood pressure (Statham P23D pressure transducer, Gould, Cleveland, OH) and in a femoral artery for withdrawing and sampling arterial blood. The blood was collected before and after installation of a cranial window for blood gas and pH determination (STAT Profile 3, Nova Biomedicals, Boston, MA). The mean arterial blood gases and pH were within normal limits as shown in Table 1.

Measurement of vessel diameter. Pial microvessels were visualized through an implanted closed cranial window. The head was fixed in a prone position with a stereotaxic apparatus (Stoelting, Wood Dale, IL), and a square-shape (5 × 5 mm) burr hole was made over the right parietal cortex. The dura was resected with caution. Pial precapillary microvessels were visualized through the cranial window, where prewarmed artificial CSF saturated with a gas mixture of 95% O₂-5% CO₂ was constantly suffused over the cortical surface at 0.3 ml/min. Cerebral microvessels were allowed to equilibrate for 60-90 min after the installation of the cranial window. The image of the pial microvessels was captured with a charge-coupled device video camera (VDC 3900, Sanyo, Japan) through a stereomicroscope (model SMZ-2T, Nikon, Japan). It was fed to a television monitor for direct observation, and the caliber was measured using a width analyzer (C3161, Hamamatsu Photonics, Hamamatsu, Japan). The composition of the artificial CSF was as follows (in mmol/l): 132 NaCl, 2.9 KCl, 1.4 MgCl₂, 24.6 NaHCO₃, 1.2 CaCl₂, 6.7 urea, and 3.7 D-glucose (pH 7.4). The intracranial pressure was maintained constant at 5-6 mmHg throughout the experiment by adjusting the height of the free end of the plastic tubing, which was connected to the outlet of the window. Only one arteriole was observed under the window in each rat.

Protocol for in vivo experiments. To determine the vasodilating responses of the resting pial arteries to adenosine (0.01-10 µmol/l), CGS-21680 (0.01-10 µmol/l), and NECA (0.01-1 µmol/l), the cortical surface was suffused with artificial CSF containing increasing concentrations of each agent. The inhibitory effects of L-NAME (10 µmol/l), ZM-241385 (1 µmol/l), and alloxazine (1 µmol/l) were determined on the vasodilation induced by adenosine or NECA, respectively. The antagonists were applied 30 min before and during suffusion of either adenosine or NECA.

Measurement of lower limit of CBF. The animal's head was fixed in a stereotaxic instrument, and the animals spontaneously breathed room air. After a craniotomy was performed, the CBF to the pial artery over the parietal cortical surface was measured by using laser-Doppler flowmetry (Laserflo BPM², Vasamedics, St. Paul, MN) equipped with a 1-mm- diameter needle probe (model P-433-5 Needle probe, Vasamedics, St. Paul, MN). After careful section of the dura mater, the probe was placed lateral to, but near the pial artery, in the open window and advanced into the CSF ~0.2-mm above the surface of the cortex. After measurement of two baselines at intervals of 10 min, the prewarmed artificial CSF saturated with a gas mixture of 95% O₂-5% CO₂ (37°C) containing antagonists was applied locally to the open window with a bolus volume of 100 µl for three times every 10 min. The baseline CBF levels were not altered during application of the antagonists by the concentrations used in this experiment. The antagonist concentration was expressed as micromoles per liter. The laser-Doppler flowmetry outputs were regarded as arbitrary units, and the changes in CBF were expressed as a percentage of the baseline CBF.

Nitrite/nitrate determination. Nitric oxide production, as assessed by nitrite/nitrate concentrations in the artificial CSF, was determined by using the Griess reagent (Promega, Madison, WI). Briefly, aliquots (3 ml) of artificial CSF drained from the cortical surface were collected for 5 min during the suffusion containing each concentration of adenosine or NECA in the absence and the presence of inhibitors. They were concentrated and dried using a centrifugal vaporizer (Eyela, CVE 200D, Japan). The samples were dissolved in the 300 µl of artificial CSF and incubated at room temperature for 15 min. Thereafter, 50 µl of each sample were added to the wells with 50 µl of the sulfanilamide solution and allowed to react for 5-10 min at room temperature while protected from light. After adding 50 µl of 0.1% N-1-naphthylethylenediamine dihydrochloride solution to all wells, we analyzed total amounts of NO₂/NO₃ by measuring absorbance at 540 nm by Power Wave ×340 (Bio-Tek Instruments). All samples were run in duplicates or triplicates.

Drugs. Adenosine and L-NAME were purchased from Sigma Chemical (St. Louis, MO). 2-p-(2-Carboxyethyl)-phenethylamino-5'-N-ethylcarboxamido adenosine (CGS-21680), NECA, and alloxazine (benzo[g]pteridine-2,4(1H,3H)-dione) were from Research Biochemicals International (Natick, MS). ZM-241385 (4-2-[7-amino-2-(2-furyl)[1,2,4]-triazolo[2,3-a]][1,3,5]· triazin-5-yl amino]ethyl)phenol) was obtained from Tocris Cookson (Bristol, UK) and dissolved in dimethyl sulfoxide to make a stock solution of 10 mmol/l. Rp-adenosine-3',5'-cyclic monophosphorothioate (Rp-cAMPS) and Rp-8-bromoguanosine 3',5'-cyclic monophosphorothioate sodium salt (Rp-8-BrcGMPS) were obtained from Biomol Research.

Statistical analysis. All data are expressed as means ± SE. A Student's t-test was used for analyzing values between the data of vehicle and inhibitor-treated groups (EC₂₅ and lower limit of CBF autoregulation). Two-way repeated-measures ANOVA was used for the comparison of time-dependent arterial diameter changes in response to agonists between inhibitor-treated and untreated groups. ANOVA for repeated measurement followed by Dunnett's method was used for statistical analysis of CBF between baseline value and the value at each blood pressure level. P < 0.05 was accepted as statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Under control conditions, mean arterial blood pressure was within the normal ranges (Table 1), and the baseline pial arterial diameters were 37.3 ± 4.6 µm (n = 101 rats).

Effects of L-NAME, ZM-241385, and alloxazine. Adenosine (0.01-10 µmol/l), CGS-21680 (0.01-10 µmol/l), and NECA (0.01-1 µmol/l) exerted concentration-dependent vasodilations of the pial arterioles. The EC₂₅ values for adenosine and NECA were significantly increased by pretreatment with L-NAME from 0.10 ± 0.02 to 1.87 ± 0.07 µmol/l (n = 7; mean dose ratio, 18.7; P < 0.001) and from 0.02 ± 0.01 to 0.33 ± 0.07 µmol/l (n = 6; mean dose ratio, 16.5; P < 0.05), respectively, with significantly decreased maximum dilation (P < 0.05). Suffusion with artificial CSF containing 10 µmol/l NECA caused a slight fall in arterial blood pressure. Thus the concentration of NECA used to elicit vasodilation was between 0.01 and 1 µmol/l for vasodilation. However, CGS-21680-induced vasodilation was little influenced by L-NAME (Fig. 1). The baseline pial arterial diameters were not changed on suffusion of 10 µmol/l L-NAME from 30 min before experiment.

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Percent changes in diameter of rat pial arteries in response to adenosine (0.01-10 µmol/l, n = 7, A), CGS-21680 (0.01-10 µmol/l, n = 5, B), and 5'-N-ethylcarboxamidoadenosine (NECA, 0.01-1 µmol/l, n = 6, C) in absence and presence of N-nitro-L-arginine methyl ester (L-NAME, 10 µmol/l). Cortical surface was suffused with artificial cerebrospinal fluid containing agonist and antagonist 30 min before and during experiment.

The concentration-dependent vasodilation evoked by adenosine (0.01-10 µmol/l) was significantly inhibited not only by 1 µmol/l of ZM-241385 but also by 1 µmol/l of alloxazine with significantly increased EC₂₅ values and decreased dilatation magnitude at 10 µmol/l of adenosine. Otherwise, the NECA-induced vasodilation was significantly suppressed by alloxazine (1 µmol/l) with significantly increased EC₂₅ value (from 0.02 ± 0.01 to 0.60 ± 0.04 µmol/l; n = 4; mean dose ratio, 30; P < 0.001). However, pretreatment with ZM-241385 (1 µmol/l) caused a decrease in dilatation magnitude at 1 µmol/l of NECA with little change in EC₂₅ value (Fig. 2).

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Percent changes in diameter of rat pial arteries in response to adenosine (0.01-10 µmol/l, n = 7, A) and NECA (0.01-1 µmol/l, n = 7, B) in absence and presence of 1 µmol/l of ZM-241385 and 1 µmol/l of alloxazine, respectively. Cortical surface was suffused with artificial cerebrospinal fluid containing agonist and antagonist 30 min before and during experiment.

Release of nitrite and nitrate. The adenosine (0.01-10 µmol/l)- and NECA (0.01-1 µmol/l)-stimulated nitrite/nitrate levels released from suffused artificial CSF were increased in a concentration-dependent manner, and they were markedly inhibited by pretreatment with L-NAME (10 µmol/l) (Fig. 3). In contrast, CGS-21680 caused increase in nitrite/nitrate releases, which were not significantly influenced by L-NAME (Fig. 3).

View larger version (39K): [in this window] [in a new window]   Fig. 3.   Effects of L-NAME (10 µmol/l) on adenosine (0.01-10 µmol/l, n = 11, A)-, NECA (0.01-1 µmol/l, n = 13, B)-, and CGS-21680 (0.01-10 µmol/l, n = 4, C)-stimulated releases of nitrite/nitrate from artificial cerebrospinal fluid suffusing over cortical surface. * P < 0.05; ** P < 0.01; *** P < 0.001 vs. basal value. # P < 0.05; ## P < 0.01 vs. corresponding agonist effect.

NECA (0.01 and 1 µmol/l)-stimulated nitrite/nitrate releases were significantly suppressed by alloxazine (1 µmol/l, P < 0.05) but not by ZM-241385 (1 µmol/l), suggesting that adenosine-stimulated nitrite/nitrate release was mediated via activation of A_2B-subtype receptors (P < 0.05) (Fig. 4).

View larger version (25K): [in this window] [in a new window]   Fig. 4.   Effects of ZM-241385 (1 µmol/l, n = 4) and alloxazine (1 µmol/l, n = 10) on NECA (0.01 and 1 µmol/l)-stimulated releases of nitrite/nitrate from artificial cerebrospinal fluid suffusing over cortical surface. * P < 0.05 vs. NECA alone.

Effect of Rp-cAMPS and Rp-8-BrcGMPS. To assess the signal transduction mechanisms underlying CGS-21680- and NECA-induced vasodilation, we examined the effects of Rp-cAMPS and Rp-8-BrcGMPS on vasodilation. As shown in Fig. 5, CGS-21680 (10 µmol/l)-induced vasodilation was significantly inhibited by Rp-cAMPS (10 µmol/l, P < 0.01) but not by Rp-8-BrcGMPS (10 µmol/l). In contrast, NECA (1 µmol/l)-induced vasodilation was not inhibited by Rp-cAMPS (10 µmol/l) but by Rp-8-BrcGMPS (10 µmol/l, P < 0.001).

View larger version (46K): [in this window] [in a new window]   Fig. 5.   Percent changes in diameter of rat pial arteries in response to NECA (1 µmol/l) and CGS-21680 (10 µmol/l) in absence and presence of Rp-cAMPS (10 µmol/l) and Rp-8-BrcGMPS (10 µmol/l), respectively. Cortical surface was suffused with artificial cerebrospinal fluid containing agonist and antagonist 30 min before and during experiment. Values are means ± SE from 4 to 6 experiments. * P < 0.01; ** P < 0.001 vs. corresponding agonist effect.

Effects of inhibitors on CBF autoregulation. Mean arterial blood pressure was 105.5 ± 4.6 mmHg (n = 101) under resting conditions. CBF to the pial arteries was well preserved despite a decrease in mean arterial blood pressure to ~55 mmHg. When the blood pressure level further decreased, CBF fell steeply depending on the magnitude of the fall in blood pressure thereafter. The lower limit of autoregulation was defined as the blood pressure at which CBF decreased by 10% of its value at resting mean arterial blood pressure. After pretreatment with either 10 µmol/l of alloxazine, an A_2B-receptor antagonist, or 100 µmol/l of L-NAME, a nitric oxide synthase inhibitor, with a bolus volume of 100 µl for three times every 10 min, the lower limit of CBF autoregulation was not altered. By contrast, the lower limit of CBF autoregulation, when assessed under pretreatment with ZM-241385 (1 µmol/l, a selective A_2A-receptor antagonist), significantly shifted to a higher blood pressure level (control, 54.3 ± 4.7 to 67.0 ± 2.1 mmHg, P < 0.05) (Fig. 6).

View larger version (28K): [in this window] [in a new window]   Fig. 6.   Relation of mean arterial blood pressure (MABP) to cerebral blood flow (CBF) in cortical pial arteries under pretreatment with ZM-241385 (1 µmol/l; n = 8, A), alloxazine (10 µmol/l; n = 4, B) and L-NAME (100 µmol/l; n = 7, C) during stepwise hypotension in rats. Arrows, lower limits of autoregulation defined as MABP at which CBF decreased by 10% of its value at resting MABP. # P < 0.05 vs. lower limit value of vehicle group; * P < 0.05 and ** P < 0.01 vs. baseline level.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The purpose of this investigation was to clarify the involvement of nitric oxide in adenosine-induced vasodilation mediated by the A_2B-subtype receptors in rat pial arterioles and its physiological importance in CBF autoregulation. The major findings are as follows: 1) adenosine- and NECA-induced vasodilations were significantly suppressed by L-NAME, whereas CGS-21680-induced vasodilation was not affected by it; 2) NECA-induced vasodilation was significantly inhibited by alloxazine, an A_2B-subtype receptor antagonist, but not by ZM-241385, a selective adenosine A_2A antagonist; 3) the nitrite/nitrate levels released into the suffusate were significantly increased by adenosine and NECA, and their release was largely inhibited by L-NAME and alloxazine, but not by ZM-241385; and 4) the lower limit of CBF autoregulation was little affected under pretreatment with either L-NAME or alloxazine.

Adenosine produces receptor-mediated activation of adenylyl cyclase and dilation of intracerebral (15) and pial (23) arteries. Headrick and Berne (11) reported that adenosine mediates relaxation in the guinea pig aorta by endothelium-dependent and -independent mechanisms, and the receptors involved in these relaxations are characteristic of the A₂-adenosine subtype. Vials and Burnstock (31) further documented that in the guinea pig coronary artery, a major part of vasodilator action of adenosine is directly mediated via A₂ receptors on the smooth muscle, and activation by adenosine of A₂ purinoceptors on endothelial cells induces relaxation via production of nitric oxide.

Recently, we observed that adenosine-induced vasodilation in the rat pial artery was mediated via activation of adenosine A_2A- and A_2B-subtype receptors, but not A₁ subtype. Whereas a part of the vasodilator action of adenosine is mediated directly via adenosine A_2A-subtype receptors (13), in our study increasing concentrations of NECA as well as adenosine elicited an enhanced nitric oxide production in association with vasodilations in a concentration-dependent manner. Moreover, the agonist effects were significantly inhibited by L-NAME, an inhibitor of nitric oxide synthase (27). Li et al. (17) reported an enhancement of nitric oxide release from vascular endothelial cells by adenosine. Our results showing that the stimulatory effects of adenosine and NECA on nitric oxide formation were not mimicked by CGS-21680, a selective A_2A-receptor agonist, indicate that the effect of adenosine and NECA on vasodilation is coupled to nitrite/nitrate production mediated by A_2B- and not by A_2A-subtype receptors. Martin and Potts (21) demonstrated a similar result in that renal artery of rat contains adenosine A_2B receptors that are located exclusively on the endothelium and cause release of nitric oxide. However, Danialou et al. (4) documented a regional difference in the involvement of adenosine receptor subtypes, in that the A₁ receptor is coupled to nitric oxide release in the adenosine-induced vasodilation of the diaphragmatic artery. In the present in vivo study, a question remains unanswered as to the source from which nitric oxide is derived, endothelial cells (9) or smooth muscle cells (22). Moreover, further study is required to show what type of the nitric oxide synthase (constitutive or inducible) is activated upon stimulation of A_2B-subtype receptor (16).

In the NIH 3T3 fibroblast cell line that displays characteristics of the A_2B-subtype receptor (1), NECA is selective for the A_2B receptors, whereas CGS-21680 (19) which is highly selective for A_2A-subtype receptors, has a much lower potency. Consistent with our results, NECA was demonstrated to be relatively selective for the adenosine A_2B receptors in guinea pig aorta (10) and rat mesenteric artery (28). Of the adenosine receptor antagonists, alloxazine has been demonstrated to be very selective for the A_2B-subtype receptor with a selectivity of about ninefold for the NIH 3T3 fibroblast over the PC12 A_2A receptor (1). More recently, Pilitsis and Kimelberg (25) demonstrated a high selectivity of alloxazine as an A_2B-receptor antagonist in preventing the increase in intracellular calcium by NECA in acutely isolated astrocytes. In our results, alloxazine strongly inhibited not only NECA-induced vasodilation but also its stimulation of nitrite/nitrate production. However, this was not the case with ZM-241385. The observations that NECA-induced vasodilation and nitrite/nitrate formation are compromised by nitric oxide synthase inhibition and by alloxazine have led us to postulate that NECA elicits vasodilation by stimulating the release of nitric oxide mediated via adenosine A_2B receptors.

In our in vivo study, it was not easy to measure the intracellular nucleotides/protein kinase concentrations in the pial arteries. Instead, we demonstrated that NECA-induced vasodilation was significantly inhibited by Rp-8-BrcGMPS, an inhibitor of cGMP protein kinase, but not by Rp-cAMPS, a preferential inhibitor of cAMP-dependent protein kinase in agreement with the result of Dubey et al. (5). By contrast, CGS-21680-induced vasodilation was not suppressed by Rp-8-BrcGMPS but by Rp-cAMPS. Thus it is likely that NECA evokes vasodilation via a pathway that does not involve adenylyl cyclase-protein kinase A but involves a guanylyl cyclase-protein kinase G pathway.

On the other hand, Hong et al. (12, 14) demonstrated that calcitonin gene-related peptide is implicated in the autoregulatory vasodilation of the pial artery in response to hypotension, and its effect is closely related with accumulation of intracellular cAMP. Recently, an involvement of the metabolic cAMP-adenosine pathway was demonstrated as a likely mechanism in the production of adenosine that acted as a regulator of vasodilation in response to hypotension (13). In this study, we predicted that, after pretreatment with L-NAME, the lower limit of autoregulation would shift to higher blood pressure levels. However, L-NAME failed to alter the lower limit of CBF autoregulation in comparison to the pre-L-NAME levels, suggesting that nitric oxide is not involved in the autoregulatory vasodilation following a decrease in mean arterial blood pressure. These results are consistent with the reports of Wang et al. (32) and Buchanan and Phillis (3). Pretreatment with alloxazine did not influence the lower limit of CBF autoregulation. Thus it is unlikely that the vasodilation coupled to nitrite/nitrate formation that is mediated by activation of adenosine A_2B-subtype receptors is involved in the modulation of CBF autoregulation in the pial artery. By contrast, pretreatment with ZM-241385 (a selective A_2A-receptor antagonist) (24) caused a significant shift of the lower limit of CBF autoregulation to higher blood pressure levels, suggesting that activation of adenosine A_2A-, but not A_2B-, subtype receptors is predominantly implicated in the mediation of autoregulatory vasodilation in response to hypotension and CBF autoregulation.

The findings of the present study indicate that adenosine-induced vasodilation via activation of A_2B receptors of rat pial arteries is coupled to the production of nitric oxide; however, these effects contribute little physiologically to the CBF autoregulation.