Stimulation of α²Adrenoceptors Dilates the Rat Middle Cerebral Artery

The experimental protocol was approved by the Animal Protocol Review Committee at Baylor College of Medicine. Experiments were conducted on approximately 98 MCAs from 49 male Long Evans rats weighing between 270 and 350 g. All rats were housed in the animal holding facility and were allowed free access to food and water before their use.

Rats were placed in a methyl methacrylate polymer anesthetic chamber with 0.25 ml isoflurane and carefully watched. Each rat was decapitated on loss of the righting reflex and lack of a response to tail pinch. The brain was immediately removed and placed in cold physiologic saline solution (PSS; see later for composition). Both the left and right MCAs were carefully removed beginning at the circle of Willis and continuing lateral and dorsal for approximately 5 or 6 mm. A section of the harvested MCA approximately 1 mm long and lying between branch points was mounted as described later.

We used a relatively new technique for studying isolated vessels. This method may be more physiologically relevant than other in vitro methods because the MCAs were pressurized and luminally perfused. [14]The MCAs were placed in an arteriograph (Living Systems, Burlington, VT) and a micropipet was inserted into each end of the MCA segment. [14,15]The ends of the MCAs were secured to the micropipets with 9-O nylon sutures. After securing the proximal end of each MCA, the lumen was gently perfused with PSS to remove blood and other contents from the lumen before securing the distal end. The arteriograph consisted of two independent chambers: one for each MCA harvested from a single rat. Each MCA was bathed in PSS, which was continually circulated from a reservoir where the solution was equilibrated with a gas consisting of 20% Oxygen²and 5% CO²in a balance of Nitrogen²[14](Figure 1). The PSS was heated before entering the chamber to maintain a bath temperature of 37 degrees C.

Luminal or transmural pressure was maintained at 85 mmHg by raising reservoirs, connected to the micropipets, to the appropriate height above each MCA (Figure 1); luminal perfusion was adjusted to 100 micro liter/min by setting the two reservoirs at different heights. With micropipets equally matched for resistance, the transmural pressure was midway between the pressures generated by the PSS in the two reservoirs; i.e., the sum of Delta P/2 and the pressure generated by the output reservoir (Figure 1). The transmural pressure and the luminal perfusion could be controlled independently. Pressure transducers located between the micropipets and the reservoirs provided a measure of perfusion pressure (Figure 1). A flow meter (#11, Gilmont Instruments, Barrington, IL) connected to the tubing leading to the output reservoir measured luminal flow. The luminal perfusate was gassed in the reservoir. In addition, the luminal perfusate traveled through gas-permeable tubing (Silastic) in the bath before perfusing the lumen of the MCAs to ensure that it was properly gassed and equilibrated to 37 degrees C. In one study, the MCAs were pressurized without having a luminal flow by occluding the outflow tubing. Samples of PSS from the bath were analyzed for partial pressure of oxygen, partial pressure of carbon dioxide, and pH using a Corning model 178 analyzer (Medfield, NY).

Vessels were magnified using an inverted microscope equipped with a video camera and monitor. Outside diameters of the MCAs were measured directly from the video monitor.

Middle cerebral arteries were allowed at least 1 h to stabilize before experiments were conducted. During this time, the diameters decreased to approximately 75% of the initial diameter after pressurization. This development of spontaneous tone was indicative of a healthy artery. For dose-response relaxations, the MCAs were preconstricted by the abluminal addition 10 sup -6 M serotonin. Although dexmedetomidine is the agonist most commonly used for alpha²adrenoceptor studies in the anesthesia literature, we choose UK14,304 as the alpha²agonist for the current studies. We wanted to be consistent with previous studies in peripheral vessels that showed that stimulation of alpha²adrenoceptors by UK14,304 elicited an endothelium-mediated dilation. [10-12,16-18]UK14,304 is a recognized selective agonist for alpha²adrenoceptors having an 1,800-fold greater affinity for the alpha²adrenoceptor than for the alpha¹adrenoceptor. [19]Rauwolscine and idazoxan, which are selective antagonists important in the pharmacologic definition of alpha²adrenoceptors, [19]were used to determine the selectivity of UK14,304. UK14,304 does have some affinity for imidazoline receptors; the selectivity for the of alpha²adrenoceptors is 100-fold greater than for imidazoline receptors. [20].

UK14,304 was cumulatively added to the extraluminal bath at the appropriate concentration. After each dose of UK14,304, an interval of 5 min was allowed before the diameter of the vessel was measured. Antagonists were added 30 min before any study was performed; pertussis toxin (100 ng/ml) was added 2 h before the UK14,304 dose-response curve. [11]In studies where the alpha adrenoceptor antagonists, phentolamine and rauwolscine, were used, the concentration of serotonin applied was 3 x 10 sup -6 M or 10 sup -5 M. Because phentolamine and rauwolscine are weak serotonin antagonists, the concentration of serotonin had to be increased to obtain the same degree of contraction as in the absence of these antagonists. In other studies, the diameters of the MCAs were already decreased by the addition of N-nitro-L-arginine methyl ester (L-NAME) or removal of the endothelium before the addition of serotonin. The addition of 10 sup -6 M serotonin produced even more constriction in the artery. We previously demonstrated that this added constriction did not affect the response to UK14,304. [21].

Two UK14,304 dose-response curves were obtained for each MCA. In some MCAs, control dose-response curves were followed by a dose-response after the experimental condition (endothelium removal, blocker, etc.). In other MCAs, which served as time controls, two consecutive dose-response curves were run. In studies where the experimental condition could be reversed (i.e., with or without luminal flow), the control and experimental curves were randomized as to the order for each MCA.

After the first dose-response curve, the bath was washed with fresh PSS and the MCA was allowed 30 min to recover before performing the second dose-response curve. Although two MCAs were harvested from each rat, each MCA was used in a different experimental protocol.

In one study, the endothelium was removed by forcing 8 ml air through the lumen of the vessels. [22]Care was taken to ensure that pressure did not exceed 85 mmHg during this process. Removal of the endothelium was confirmed by the absence of a dilation to the addition of 10 sup -5 M acetylcholine.

UK14,304 (5-bromo-N-(4,5-dihydro-1H-imidazol-2-yl)-6-quinoxalinamine), prazosin, idazoxan, rauwolscine, and S-nitroso-N-acetylpenicillamine were obtained from Research Biochemicals (Natick, MA). L-NAME [N^G-nitro-L-arginine methyl ester], L- and D-arginine, phentolamine, serotonin, and pertussis toxin were obtained from Sigma Chemical (St. Louis, MO). Reagents were dissolved in either PSS or distilled water with the following exceptions: 0.25 mg UK14,304 was dissolved in dimethyl sulfoxide/distilled water (0.25 ml/11.15 ml), 8.8 mg S-nitroso-N-acetylpenicillamine was dissolved in dimethyl sulfoxide/water (0.53 ml/2.5 ml), and 21 mg prazosin was dissolved in 5 ml methanol. Dimethyl sulfoxide alone did not change the diameter of the MCAs at the above concentrations for UK14,304 and S-nitroso-N-acetylpenicillamine. In the prazosin studies, the methanol vehicle was added to the control group. Pertussis toxin was in solution (50 micro gram/ml in 50% glycerol, 50 mM Tris, 10 mM glycine, 0.5 M NaCl, pH 7.5) when purchased and was added directly to the bath.

The PSS comprised the following (in mM) [15]: NaCl 119, NaHCO³24, KCl 4.7, KH²PO⁴1.18, MgSO⁴1.17, CaCl²1.6, glucose 5.5, and EDTA 0.026.

Data are expressed as the mean+/-the standard error of the mean with n representing the number of MCAs. Values for the MCAs presented represent the diameters before the addition of serotonin. For comparison of responses to UK14,304, a repeated-measures analysis of variance was used followed by the Fisher test for multiple comparisons when appropriate. The acceptable level of significance was defined as P < 0.05.

The resting outside diameter for MCAs in one study was 239+/- 13 micro meter (n = 8); the resting diameter in the remainder of the studies did not differ statistically from the above value. The abluminal pH level was 7.332+/-0.004 (n = 25), partial pressure of carbon dioxide was 35.0+/-0.5 mmHg (n = 26), and partial pressure of oxygen was 144+/-1 mmHg (n = 25). In preliminary studies, PSS was collected from the input micropipet to determine if the luminal perfusate was correctly gassed (see Arteriograph and Mounting the MCAs in Methods and Materials for the technique of gassing the luminal PSS). There were no significant differences in gasses or pH level between the luminal and abluminal PSS. The mean input and output pressures (n = 25) across the two micropipets and MCAs were 87.1+/-0.2 and 82.9+/-0.2 mmHg, respectively. When the MCAs were preconstricted with 10 sup -6 M serotonin, the vessels constricted approximately 20%. MCAs constricted with serotonin without the addition of UK14,304 were stable (i.e., no significant change in diameter, n = 28) over 30 min. The time required for the UK14,304 dose-response curve was 25 min.

In the first study, we determined the response of abluminally applied UK14,304 in MCAs with flow (100 micro liter/min) and without flow through the lumen (Figure 2(A)). Luminal flow was stopped by occluding the outflow tubing distal to the micropipet and pressure transducer. Regardless of whether or not there was luminal flow, transmural pressure was maintained at 85 mmHg. Two dose-response curves were conducted on each MCA. In half of the arteries (four) the dose-response curve was conducted with flow through the lumen followed by the dose-response curve without flow; the order was reversed for the remaining four MCAs. The response to UK14,304 with "flow" or "no flow" was independent of the order for the concentration-response curves. Therefore, the data were combined for the graphs in Figure 2(A). The mean dilation (% diameter change) at a concentration of 10 sup -4 M UK14,304 in MCAs with and without flow was 11.4+/-1.6% and 9.7+/-1.6% (n = 8 MCAs), respectively. There was no significant difference in the dilations produced by UK14,304 in MCAs with flow or without flow. For the remainder of the studies described later, MCAs were maintained with a luminal flow of 100 micro liter/min.

(Figure 2)(B) shows the mean changes for two consecutive UK14,304 dose-response curves conducted on six MCAs. This study demonstrated that repeated dose-response curves could be conducted on a single MCA without having the response change from the first to second curve. Consequently, in the remaining studies, except where noted, we compared the second dose-response curve (experimental condition) to the first curve, which served as a control. Throughout the remaining studies, time controls were conducted at intervals to ensure that this relationship between the first and second curves remained constant.

(Figure 3) shows the results of studies that determined the receptor type stimulated by the agonist, UK14,304. Prazosin (10 sup -6 M), a selective alpha¹antagonist, was without effect (n = 6 MCAs, Figure 3(A)) while phentolamine (10 sup -6 M), a nonselective alpha antagonist, completely abolished the dilation produced by the abluminal application of UK14,304 (n = 8 MCAs, Figure 3(B)). Idazoxan (10 sup -6 M), a selective alpha²antagonist, significantly attenuated the dilation (n = 7 MCAs, Figure 3(C)). Rauwolscine, a selective alpha²antagonist, abolished the dilation at a concentration of 10 sup -5 M (n = 5 MCAs, Figure 3(D)), but at a concentration of 10 sup -6 M (n = 6 MCAs) only affected the dilations at the greatest UK14,304 concentrations (data not shown). Addition of idazoxan produced a 7% decrease in the resting diameter of the MCAs (P = 0.02); no other antagonist had an effect.

(Figure 4) shows the response to UK14,304 in control MCAs and in MCAs after the endothelium was removed by forcing air through the lumen of the vessels. [22]Integrity of the endothelium was evaluated with acetylcholine (10 sup -5 M), which produced a dilation of 24+/-5% (n = 5). After removal of the endothelium, the dilation elicited by acetylcholine and UK14,304 was abolished (n = 7 MCAs).

Inhibition of nitric oxide synthase with 10 sup -5 M L-NAME completely abolished any dilation produced by UK14,304 (Figure 5). The inhibition was reversed with the addition of 10 sup -3 M L-arginine, the substrate for nitric oxide synthase, but not by the inactive isomer, D-arginine (10 sup -3 M). Two studies are shown in Figure 5. The first study using eight MCAs is represented by the filled symbols; the second study using eight MCAs is represented by the open symbols. For the studies using arginine, the concentration-response curves were conducted in the presence of L-NAME in four randomly selected MCAs using 10 sup -3 M L-arginine followed by 10 sup -3 M D-arginine. In the remaining MCAs, D-arginine curves in the presence of L-NAME were obtained first followed by L-arginine. The responses to the arginine isomers were independent of the order in which the curve was obtained; i.e., the response of the MCAs to UK14,304 in the presence of L-NAME and L-arginine was independent of whether the concentration-response curve was obtained before or after the curve with D-arginine. The same held true for D-arginine. Therefore, the data were combined for all L-arginine curves and combined for all D-arginine curves.

Pertussis toxin (100 ng/ml) completely abolished the dilation produced by UK14,304 (Figure 6, n = 6 MCAs). For this study we show the time controls (open symbols), because 2 h elapsed between the control curve and the curve in the presence of pertussis toxin.

Because removal of the endothelium, addition of L-NAME, or addition of pertussis toxin completely blocked the UK14,304-induced dilation, it was important to determine if the vascular smooth muscle could still respond to elevations in nitric oxide in the above experimental paradigms. The response of the smooth muscle to additions of S-nitroso-N-acetylpenicillamine (10 sup -4 M), an exogenous nitric oxide donor, was not altered by the experimental perturbations (Figure 7).

We report four new findings regarding alpha²adrenoceptors and cerebral arteries. The first finding involves an observation: (1) stimulation of alpha²adrenoceptors on preconstricted rat middle cerebral arteries produced a dilation. The next three findings involve aspects of the mechanism of dilation: (2) the dilation required intact endothelium, (3) the dilation involved nitric oxide, and (4) a pertussis toxin-sensitive G protein was part of the signal transduction process.

In rat MCAs preconstricted with serotonin, the addition of UK14,304 to the extraluminal bath produced a dose-dependent dilation that was not influenced by the absence or presence of flow through the lumen (100 micro liter/min; Figure 2(A)). The importance of conducting the dose-response curve during the different flow states involves shear forces produced by the flow of the PSS through the lumen of the vessel. Changes in shear stress produced by luminal flow affects the contractile state of arteries by endothelial and nonendothelial mechanisms. [23]Changes in the contractile state could alter the dilations produced by stimulation of various receptor systems; however, in the case of alpha²adrenoceptors, flow (and thus shear stress) did not appear to alter the dose-dependent dilation.

Pharmacologic studies reveal that the UK14,304-mediated dilations were elicited by stimulation of alpha²adrenoceptors (Figure 3). We can rule out the involvement of alpha¹adrenoceptors because the dilations produced by UK14,304 were not affected by prazosin, a selective alpha¹antagonist (Figure 3(A)), and were respectively blocked and attenuated by rauwolscine (Figure 3(D)) and idazoxan (Figure 3(C)), selective alpha²antagonists. Although UK14,304 is a selective alpha²agonist, it does have imidazoline binding properties and could possibly dilate the MCAs through an imidazoline receptor. [24,25]The dilations produced by UK14,304 were blocked by phentolamine, a nonselective alpha-antagonist, and by idazoxan; both of these compounds also can antagonize imidazoline receptors. However, the involvement of imidazoline receptors can be ruled out because the dilations were completely blocked by rauwolscine, a selective alpha²antagonist without imidazoline effects. [26]Furthermore, the selective imidazoline agonist, moxonidine, [25]and the proposed endogenous imidazoline agonist, agmatine, [27]both produced a small contraction in the rat MCA; no dilations were observed (unpublished). Taken together, the results of our pharmacologic studies indicate that the vasodilatory agonist, UK14,304, stimulated alpha²adrenoceptors.

Dilations mediated by alpha²adrenoceptors have been reported for canine and porcine coronary arteries and canine femoral arteries. [8-11]Conversely, cerebral arteries from dogs and cats constrict when exposed to norepinephrine apparently through alpha²adrenoceptors. [7]Comparisons of various studies of cerebrovascular effects of alpha²adrenoceptor stimulation are complicated by the different species used, [7]different pharmacologic agents, which may have slightly different actions, and the different vessels used.

The in vivo administration of the alpha²adrenoceptor agonist, dexmedetomidine, at doses that reduced the anesthetic requirement, decreased cerebral blood flow in dogs anesthetized with either isoflurane or halothane. [3-5]In pentobarbital-anesthetized cats, clonidine, a less specific and less potent alpha²adrenoceptor agonist, decreased cerebral blood flow when measured with the tracer¹³³Xenon, but had no effect when the tracer⁸⁵Krypton was used. [6]In two of these studies, cerebral blood flow decreased in spite of an increase in mean arterial blood pressure after administration of dexmedetomidine [4,5]whereas in the other two studies blood pressure was unchanged after administration of the dexmedetomidine [3]or clonidine. [6]In the human, clonidine decreased cerebral blood flow. [28,29]Although mean arterial blood pressure decreased in these latter studies, the calculated resistance was reported to increase. The reduction in cerebral blood flow produced by the administration of alpha²adrenoceptor agonists could be due, at least in part, to neural mechanisms secondary to altered activity of the locus ceruleus or the sympathetic nervous system. [6,30]Thus, the change in cerebral blood flow may be more complicated than simply a direct effect on arteries. Furthermore, the above studies used different agonists than that used in the current study, which may have different pharmacologic actions.

Our results support the idea that alpha²agonists dilate rat middle cerebral arteries. The dilation requires intact endothelium and appears to involve nitric oxide. These conclusions are based on the results showing that removal of the endothelium and inhibition of nitric oxide synthase with L-NAME abolished the dilation to UK14,304 (Figure 4and Figure 5). After the inhibition of nitric oxide synthase with L-NAME, the response could be restored by the addition of L-arginine, the precursor for the enzyme nitric oxide synthase, but not the inactive isomer, D-arginine (Figure 5). This reversal of the inhibition provides evidence that L-NAME is acting specifically on a pathway involving L-arginine further implicating nitric oxide as a key factor in the dilation produced by UK14,304.

Our data indicate that pertussis toxin abolished the dilation to UK14,304. Pertussis toxin inactivates specific G proteins by adenosine diphosphate ribosylation of the alpha-subunit of the protein. [31]Pertussis toxin is known to inactivate several classes of G proteins including the Gi and Go. [31]The Go protein is the primary G protein found in the brain but is present in limited amounts outside the central nervous system. [32]It is not known if the Go protein is found in cerebral vessels. However, a well-accepted fact is that alpha²adrenoceptors are coupled to the Gi protein. [31,33]Little is known about the link between Gi protein activation and the resulting relaxation other than it is a key element in the alpha²adrenoceptor-mediated dilations. [11]Further studies will be needed to elucidate the G protein role in alpha²adrenoceptor dilations.

We had originally thought that the dilations mediated by UK14,304 conformed to the "classical model" where the alpha²adrenoceptors were located on the endothelium (see [21]). When stimulated, these alpha²adrenoceptors would activate nitric oxide synthase to synthesize nitric oxide. The nitric oxide formed in the endothelium would diffuse to the vascular smooth muscle and stimulate the synthesis of cyclic 3',5' guanosine monophosphate by activating soluble guanylate cyclase. Cyclic 3',5' guanosine monophosphate would produce the relaxation of the vascular smooth muscle through a number of different processes. In recent studies, [21]we provided evidence that dilations produced by alpha²adrenoceptor stimulation likely works through another mechanism that we have termed the "permissive model." In this model, the basal release of nitric oxide, which occurs even in nonstimulated conditions, permits the vasodilation to occur. Abolishing the basal release of nitric oxide by either removing the endothelium or inhibiting its synthesis with L-NAME would inhibit the vasodilation. The alpha²adrenoceptors appear to be located on the vascular smooth muscle and the direct activation of the receptor by UK14,304 produce the relaxation. In this later model, nitric oxide acts only in a permissive or modulatory role and is not the active vasodilator. [21].

We have shown that stimulating alpha²adrenoceptors in the rat MCA with UK14,304 produced a dose-dependent dilation. This dilation required intact endothelium and involved nitric oxide and a pertussis toxin-sensitive G protein. Whether cerebral arteries in the human have a similar alpha²adrenoceptor-mediated dilatory response will have to be determined.