- Email: [email protected]
Regional differences in cerebral vasomotor control by nitric oxide
BrainResearchBulletin,Vol.38, No. 4. pp. 365 369. 1995 Copyright ~', 1995ElsevierScienceInc. Printedin the USA.All rightsreserved 0361-9230/95$9.50 + .00
Pergamon 0361-9230(95)02001-8
Regional Differences in Cerebral Vasomotor Control by Nitric Oxide YASUKAZU KAJITA,* MASAKAZU TAKAYASU,.1 YOSHIO SUZUKI* MASATO SHIBUYA* MASAYOSHI MORI,* HIROFUMI OYAMA,* KENICHIRO SUGITA* AND HIROYOSHI HIDAKA1-
Departments of *Neurosurgery and ?~Pharmacology,Nagoya University School of Medicine, Nagoya 466, Japan [Received 14 February 1995; Accepted 1 June 1995] itors generally decrease regional cerebral blood flow (rCBF) by blocking the basal release of NO, but the amount of rCBF reduction may differ among brain regions, depending upon the importance of L-arginine-NO pathway in regulating each regional circulation. Kovach et al., for example, have shown that Nr-nitroL-arginine caused the most striking decrease in rCBF in the cerebellum, pituitary, and medulla oblongata, and caused no decrease in rCBF of cerebral cortex or white matter [12]. In a previous study using cerebral angiography in dogs, we showed that vasodilation in response to intravenous injection of vasopressin, a stimulant of nitric oxide release, is greatest in the basilar artery and much less in the middle cerebral artery [20]. These data suggest that the role of the L-arginine-NO pathway in regulating cerebral circulation may be more pronounced in the posterior circulation than the anterior circulation. To test this hypothesis, we studied regional differences in the effects of a NO synthesis inhibitor, Nr-monomethyl-L-arginine, and a precursor of NO, L-arginine, on cerebral vessels from the anterior and the posterior circulation in vivo in the macrocirculation and in vitro in the microcirculation, using well-established methods. To study the responses of larger arteries, the diameter of the middle and anterior cerebral and the basilar arteries were evaluated angiographically in dogs. Microscopic evaluation of responses by the microcirculation was performed on isolated intracerebral arterioles branching from the middle cerebral and basilar arteries in rats.
ABSTRACT: Regional differences in the role of nitric oxide in cerebral vasomotor control were investigated with a nitric oxide synthesis inhibitor, NG-monomethyI-L-arginine, or a precursor of nitric oxide, L-arginine using both dog cerebral angiography for the larger artery study and rat isolated arterioles for the microcirculation study. Ne-monomathyI-L-arginine (10 pmol) constricted the dog cerebral artedes, by 15.6%, 17.5%, and 27.3% in the middle cerebral, antedor cerebral, and basilar artedes, respectively. The greater constriction of the basilar artery did not reach statistical significance. However, L-arginine (100 pmol) produced significantly greater dilation of basilar artedes than the middle cerebral or anterior cerebral (31.3% vs. 16.7% or 13.1%). NG-monomethyI-L-arginine at 10 -a M constricted rat arterioles originating from basilar arteries significantly more than the middle cerebral arteries (23% vs. 14%). L-arginine at 10 -a M dilated rat arterioles from basilar arteries significantly more than from the middle cerebral artery (24 vs. 11%). These findings suggest that the roles of nitric oxide in vasomotor control differs by region in the brain, and it may be greater in vessels of the postedor than of the antedor circulation. KEY WORDS: Arterioles, Cerebral artery, L-Arginine, NG-Mono methyl-L-arginine, Nitric oxide, Regional difference.
INTRODUCTION Nitric oxide (NO), synthesized from L-arginine by NO synthase (NOS), is a potent vasodilator that plays a substantial role in the regulation of cerebral circulation. NO was originally identified in vascular endothelial cells as a principal endothelium-derived relaxing factor (EDRF) [13]. Recent studies have shown that NOS also is expressed in neurons and in the perivascular fibers that innervate cerebral blood vessels, suggesting roles for NO both as a neurotransmitter and a regulator of regional cerebral blood flow control [3,13]. NOS is heterogenously expressed in brain parenchyma and in its perivascular tissue. NOS-expressing neurons are densest in the cerebellum, hypothalamus, and pituitary, and are sparser in cerebral cortex [3]. Recent studies have suggested that basal release of NO is responsible for the maintenance of normal cerebral blood flow, and that induced release may contribute to the vasodilation initiated by increased local neuronal activity or metabolism [4,6,9,18]. NO synthesis inhib-
METHODS Angiography of Cerebral Arteries in Dogs Cerebral angiography was performed on dogs as previously described in detail [10,16,20]. Briefly, mature mongrel dogs of either sex, weighing 8 to 20 kg, were anesthetized with intravenous pentobarbital (20 to 30 mg/kg), and ventilated through an endotracheal tube with room air by a respirator (model B2, Igarashi Ika Kogyo Co., Tokyo, Japan). The ventilation rate (15 cycles per min) and tidal volume (20 ml/kg) were adjusted to maintain arterial blood gases and pH within normal physiological limits. A catheter was placed in the left femoral artery to monitor mean arterial blood pressure and heart rate. Internal diameters of cerebral arteries were determined by vertebral angiography. A catheter was inserted directly into the right
t Requests for reprints should be addressed to Masakazu Takayasu, Department of Neurosurgery, Nagoya University School of Medicine, 65 Tsuruma-cho, Showa-ku, Nagoya 466, Japan. 365
366
vertebral artery just proximal to the transverse foramen of the C6 cervical vertebra. Angiograms were performed through this catheter using 3.0 ml of 65% iothalamate meglumine and were radiographed periodically at a fixed magnification following cisternal injection of 0.1 to 10 #mol NG-monomethyl-L-arginine (L-NMMA) or 1 to 100 #mol L-arginine, dissolved in 2.0 ml of physiological saline. Injection of solutions into the cistema magna was done following withdrawal of an equal volume of cerebrospinal fluid to maintain constant intracranial pressure. The animals were positioned with heads tilted, to increase the contact of the cisternal injections with the major arteries of the brain. Arterial diameters were measured before and after injection of drugs using a computerized image analysis system (IMAGE 1.27, Macintosh 2cx). Portions of the middle cerebral artery (MCA) (the proximal portion of the M-1 segment), the anterior cerebral artery (ACA) (the proximal portion of the A-2 segment), and the middle third of the basilar artery (BA) were measured.
KAJITA ET AL.
Statistical Analysis Vessel diameters at each agonist dose were expressed as a percentage of the control vessel diameter. Vasoconstriction and vasodilation were expressed as the percentage change in diameter from control vessel diameter, and the data are reported as means _ SD. The significance of differences among a set of three or more samples was evaluated by one-way analysis of variance (ANOVA) with the Fisher Protected Least Significant Difference (PLSD) multiple range test as a postANOVA test. The significance of differences between two unpaired samples, such as maximum responses in arterioles from the regions of the MCA and the BA, was evaluated by an unpaired t-test• Linear regression analysis and ANOVA were performed to examine the correlation between vessel diameter and responsiveness to L-NMMA or Larginine. We considered differences significant at p < 0•05. RESULTS
Microscopic Evaluation of Isolated lntracerebral Arterioles in Rats Intracerebral arterioles were isolated and cannulated in an organ bath (composed as described below), and changes in vessel diameter in response to the extraluminal administration of agents were measured, as previously described in detail 15,21,22]. Briefly, penetrating branching intracerebral arterioles, 10 to 70 /~m in diameter, were isolated surgically at 4°C from the first (M1) portion of a middle cerebral artery or the middle third of the basilar artery, from the brains of pentobarbital-anesthetizedmale Sprague-Dawley rats weighing 300-400 g. Vessels were transferred to a temperature-controlled chamber on the stage of an Olympus inverted microscope and were cannulated with glass pipettes at 25°C. The inner diameter of the vessel was determined using a video micro-scaler (FOR.A Model IV-550, Tokyo, Japan). After cannulation, transmural pressure was set and maintained throughout the experiments at 60 mm Hg via the cannulating pipette• Bath temperature was brought to 37.5°C, and the vessels were allowed to equilibrate for 45 min at an extraluminal bath pH of 7.3. During the equilibration period of 30 min, vessels developed spontaneous tone, contracting to about 70% of their maximum passive diameter. Vessel responsiveness was then assessed by changing extraluminal pH from 7.3 to 6.8 and from 7.3 to 7.6, The physiological salt solution (PSS) used in this preparation was a modified Ringer's, composed of 144 mM NaCI, 3.0 mM KCI, 2.5 mM CaCI2, 1.5 mM MgSO4, 5.0 mM glucose, 2,0 mM pyruvate, 0.02 mM ethylenediaminetetraacetic acid (EDTA), 2.0 mM 3-(N-morpholino)propanesulfonic acid (MOPS), 1.21 mM NaHzPO4, and 0.9-1.0 g/100 ml bovine serum albumin. The extraluminal solution contained no albumin in our experimental setup to simulate in vivo conditions [5]. Solution pH was maintained at 7.3 except when vessel responsiveness was assessed by pH changes. Control vessel diameter was defined as the diameter to which the vessels spontaneously contracted during the equilibration period at pH 7.3. L-NMMA and L-arginine were dissolved in PSS, and the pH was adjusted to 7.3 with HCI or NaOH as necessary. Dose-response curves to extraluminal L-NMMA and L-arginine were determined by sequential addition of solutions of increasing concentration to the organ bath at constant pH 7.3. The vessel diameter was allowed to stabilize for 5 min between each change in bath solution. The nitric oxide synthesis inhibitor [NG-monomethyl-L-argi nine; (L-NMMA)] was obtained from Calbiochem (La Jolla, CA, USA). L-arginine and all other chemicals were reagent grade.
Angiographic Study of Cerebral Arteries in Dogs Intracisternal injection of L-NMMA (0.1, 1, 10 #mol) produced a significant decrease in the internal diameters of the MCA, the ACA, and the BA on angiogram (Fig. 1). L-NMMA at 10 #mol constricted these arterial diameters by 15.6 + 11.1% (mean ___SD, MCA, n = 5), 17.5 +_ 4.0% (ACA, n = 3) and 27.3 _+ 7.0% (BA, n -- 5) at 15-60 min following intracisternal in-
°].-
L-arginine
3O
20.
I._
Q q)
•
10.
E
.ca
MCA
[] ACA []
BA
0I
10
100
0.1
1
10
Dose (ixmol)
Im
0 O~
C
¢= i-
-10-
¢0 -20°
.L
L-NMMA FIG. 1. Changes in diameter of the middle cerebral artery (MCA), the anterior cerebral artery (ACA), and the basilar artery (BA) in dog cerebral angiograms, in response to intracisternal injection of NC'-monomethyl-L-arginine(L-NMMA)(0.1, 1, or 10/~mol) or L-arginine(1, 10, or 100 #tool). Asterisks indicate significant differences (p < 0.05) in response between each pair of arteries.
V A S O M O T O R C O N T R O L BY NITRIC OXIDE
L-arginine 30.
20, ~r
it
it
Rat Arterioles -" MCA BA
G 10
9
8
7
6
5
-
,
-
•
-
4 g
3
Dose (-log M)
•
cO
367
that branched from the M C A and the BA, respectively. The responses of arterioles to L-NMMA were significantly greater in vessels originating from the BA than from the MCA (10 910 3 M) (Fig. 2). L-arginine caused a dose-dependent vasodilation with maximal increases in diameter of 11.1 _+ 5.4% (n - 5) and 24.3 _+ 11.3% in = 5) at 10 ~ M in arterioles from the M C A and the BA, respectively (Fig. 2). The responses of arterioles to L-arginine again were significantly larger in vessels from the BA than from the M C A (10 ~-10 :~M). Baseline characteristics of arterioles originating from the MCA and the BA that were studied were compared (Table 1). The control arteriolar diameter was 57.6 _+ 8.8 (54.6-81.8) # m of arterioles from the MCA region and 23.6 _+ 5.0 (13.0-29.7) #m of arterioles from the BA region. There were significant differences between the diameters of arterioles from these regions. However, there was no significant correlation between control vessel diameter and responsiveness to L-NMMA or L-arginine among arterioles from the M C A region or among those from the BA region, with p-values ranging from 0.20 to 0.74. There was no significant difference in responsiveness to pH changes between arterioles from the MCA and the BA.
~ .10. DISCUSSION .20-
-30
L-NMMA FIG. 2. Dose-response curves of isolated rat arterioles originating from the middle cerebral artery (MCA) and the basilar artery (BA), in response to extraluminal application of NG-monomethyl-L-arginine (L-NMMA) or L-arginine (10 ') M to 10 3 M). Asterisks indicate significant differences Lo < 0.05) in response between arterioles from the MCA and from the BA.
jection. The greater decrease in diameter of the BA than of the MCA or ACA did not reach statistical significance. L-arginine (1, 10, 100 #mol) caused significant vasodilation of all the arteries (Fig. 1). L-arginine, 100 #mol dilated the measured arterial diameters by 16.7 ___ 6.7% (MCA, n = 4), 13.1 _+ 4.3% (ACA, n = 4) and 31.3 _ 8,1% (BA, n = 5) at 1 5 - 6 0 min following intracisternal injection. The BA dilation in response to L-arginine was significantly greater than that of the M C A or ACA. Control diameters of the MCA, ACA, and BA were 1.18 _ 0.20, 0.74 _+ 0.16, and 0.56 _+ 0.11 mm, respectively. These diameters are significantly different from each other. However, there was no significant correlation between control vessel diameter and responsiveness to L-NMMA or L-arginine among different MCAs, ACAs, or BAs, with p-values ranging from 0.23 to 0.72. No significant change in mean arterial blood pressure occurred following intracisternal injections of L-NMMA or Larginine.
This study demonstrated regional differences in responsiveness of cerebral vessels to NO. A nitric oxide synthesis inhibitor, L-NMMA produced vasoconstriction in both larger cerebral arteries and smaller intracerebral arterioles. The vasoconstriction observed in vessels originating from the BA was greater than that in those from the MCA or ACA region. L-arginine, a precursor of NO, produced vasodilation in both larger cerebral arteries and smaller intracerebral arterioles. Again, the vasodilation was significantly greater in vessels from the BA than in those from the MCA or ACA. L-NMMA produced vasoconstriction in both large arteries and small arterioles provides evidence that a basal release of NO contributes to the maintenance of cerebral vascular resistance in these vessels. This study indicates that NO-mediated regulation of the resting vascular tone may be more pronounced in vessels of the posterior circulation than in those of the anterior circulation. These findings are consistent with those of a cerebral blood flow study using radiolabelled microspheres by Kovach et al. [12], that showed a significant decrease in rCBF in the cerebellum and brainstem following administration of a NO synthesis inhibitor, whereas there was no decrease in the rCBF in cerebral cortex and white matter. In a previous study using cerebral angiography in dogs, we showed that vasodilation in response to intravenous injection of vasopressin, a stimulant of nitric oxide TABLE 1 BASELINE CHARACTERISTICS OF ARTERIOLES IN THIS STUDY
pH Response (%) Region of Arterioles
BA MCA
n
Diameter (,um)
pH 6.8
pH 7.6
10 10
23.6 _+ 5.0 57.6 _+ 8.8
13.4 __+2.0 13.9 _+ 5.4
13.0 _+ 2.8 13.2 _+ 2.0
0.0001 *
0.64
0.89
Microscopic Study of Isolated Cerebral Arterioles From Rats
ANOVA: p value
Extraluminal t,-NMMA produced a dose-dependent vasoconstriction with maximal decreases in diameter of 14.2 ± 2.7% (n = 5) and 23.0 _+ 7.4% (n = 5) at 10 4 M in isolated arterioles
Values are those prior to application of drugs. Data are means _+ SD. BA, the basilar artery; MCA, the middle cerebral artery; ANOVA, analysis of variance; *significant difference.
368
release, is greatest in the BA and much less in the MCA, that is, similar to the results of this study [20]. All of these findings suggest that vessels from the posterior circulation may be subject to greater basal vasomotor control by NO-mediated mechanisms than those from the anterior circulation. L-arginine is a precursor of nitric oxide [17]. The effects of exogenous L-arginine on the vessels, however, have been complex, and inconsistent results have been documented in the literature [4,6,1 !,12,14,18]. If L-arginine is present in suboptimal amounts in the NO-producing cells, t=arginine may be rate-limiting, and exogenous u-arginine may produce vasodilation by supplying more substrate for NO production. On the other hand, exogenous L-arginine is not rate-limiting for NOS when l=arginine is present in optimal amounts in the cells. Most of these studies have been performed in vivo, with L-arginine supplied intravenously or topically. In vivo, exogenous L-arginine may not reach NOS containing cells freely in a certain condition, because glial cells tend to sequester this amino acid and endothelial cells express selective uptake and transport mechanisms [ 1,2]. Few in vitro studies have examined the effects of L-arginine except for the one by Kimura et al. [11]. They used a similar method to ours and observed a tendency for vasodilation by L-arginine, which, however, did not reach statistical significance. This study showed that exogenous L-arginine caused vasodilatation of rat cerebral arterioles in vitro, as well as of canine large arteries in vivo, suggesting that t.-arginine could be rate-limiting factor for NO production under normal physiological conditions. It is not clear why vessels in the region of the basilar artery appear to show a more pronounced vasomotor response to NO than do those from the region of the middle or anterior cerebral artery. We wondered whether the different average diameters of vessels in these regions might influence the magnitude of NOrelated vasomotor response. Arterioles from the BA region are smaller in diameter than those from the MCA region, and smaller arterioles may be more responsive than larger arterioles. However, this explanation appears unlikely, because there was no difference in response to pH 6.8 acidosis or to pH 7.6 alkalosis between arterioles originating from the BA and those from the MCA, and because there was no correlation between the vessel diameter and responsiveness of arterioles within each these regions. Moreover, the associations were reversed in the larger cerebral arteries in dogs, where the BA was more responsive to L-arginine than the M C A or the ACA, whereas it was larger in diameter than both the MCA and the ACA. Again, there was no correlation between the vessel diameter and responsiveness of arteries within each these regions. In angiographic study of dog cerebral arteries, concentration gradient of the drug should be considered because drugs were administered from the cisterna magna, but it may not be a main factor. Using the same mothod, we demonstrated that the effectiveness of pituitary adenylate cyclase-activating polypeptide were in the reverse order, that is, it was more effective to the ACA and MCA than the BA [19]. Therefore, the vessels from the BA region either produce more NO or may be more sensitive to NO than those from the MCA or the ACA region. Immunohistochemical studies have demonstrated that nitric oxide synthase-containing nerve fibers associated with cerebral arteries are abundant and dense in the proximal anterior and middle cerebral arteries and are less numerous in the caudal circle of Willis, including the basilar artery [8,15]. These apparently discrepant findings may suggest that basal or stimulated release of NO through vascular endothelium may be greater in the basilar artery region than those from the middle or anterior cerebral artery region, whereas the relative contribution of NO-related neurogenic control would be the reverse. Another possible expla-
KAJITA ET AL.
nation is that there might be a difference in sensitivity to nitric oxide or in the activity of NOS in the different regions of the cerebral vasculature. The clinical significance of these results could be imagined. Because the vessels in the posterior circulation seem to be controlled in a more dilated state by NO, there may be less room for further vasodilation in these vessels. This could explain the fact that the cause of the transient ischemic attacks (TIA) in the posterior circulation is more often hemodynamic rather than embolic; the latter is more common in the TIA in the anterior circulation. The role of NO in the regulation of small cerebral arteriolar tone remains controversial. Faraci has reported, in an in vivo study, that NO might have more influence on the regulation of basal tone in the larger cerebral arteries than in cerebral arterioles [7]. In this study, I=NMMA produced a significant vasoconstriction of rat intracerebral arterioles originating from the MCA and the BA, by as much as 14% and 23%, respectively, and 1=arginine caused significant vasodilation of arterioles from the MCA and the BA regions by as much as 11% and 24%, respectively. The magnitudes of L-NMMA-induced vasoconstriction and of L-arginine-induced vasodilation were similar in dog large cerebral arteries in vivo and in rat smaller intracerebral arterioles in vitro. The observed regional heterogeneity of NO-related vasomotor responses between the anterior and the posterior circulation thus may be greater than the longitudinal heterogeneity between large arteries and smaller arterioles. In conclusion, we have shown that the role of the L-argininenitric oxide pathway may differ by region in the cerebral circulation, and that it may be greater in vessels of the posterior circulation than of the anterior circulation.
REFERENCES 1. Aoki, E.; Semba, R.; Mikoshiba, K.; Kashiwamata, S. Predominant localization in glial cells of free L-arginine: Immunocytochemical evidence. Brain Res. 547:190-192; 1991. 2. Bogle, R. G.; Moncada, S.; Pearson, J.; Mann, (3. E. Identification of inhibitors of nitric oxide synthase that do not interact with endothelial cell l~-arginine transporter. Br. J. Pharmacol. 105:768-770; 1992. 3. Bredt, D. S.; Hwang, P. M.; Snyder, S. H. Localization of nitric oxide synthase indicating a neural role for nitric oxide. Nature 347:768-770; 1990. 4. Busija, D. W.; Leffler, C. W.; Wagerle, L. C. Mono-l=arginine-containing compounds dilate piglet pial arterioles via an endotheliumderived relaxing factor-like substance. Circ. Res. 67:1374-1380; 1990. 5. Dacey, R. (3., Jr,; Duling, B. R. A study of rat intracerebral arterioles: Methods, morphology, and reactivity. Am. J. Physiol. 243:H592H606; 1982. 6. Faraci, F. M. Role of nitric oxide in regulation of basilar artery tone in vivo. Am. J. Physiol. 259:H1216-H1221; 1990. 7. Faraci, F. M. Role of endothelium-derived relaxing factor in cerebral circulation: Large arteries vs. microcirculation. Am. J. Physiol. 261 :H1038-HI042; 1991. 8. ladecola, C. Regulation of the cerebral microcirculation during neural activity: Is nitric oxide the missing link? T.I.N.S. 16:206-215; 1993. 9. ladecola, C.; Beitz, A. J.; Renno, W. Nitric oxide synthase-containing neural processes on large cerebral arteries and cerebral microvessels. Brain Res. 606:148-155; 1993. 10. Kajita, Y.; Suzuki, Y.; Oyama, H.; Tanazawa, T.; Takayasu, M.; Shibuya, M.; Sugita, K. Combined effect of t=arginine and superoxide dismutase on the spastic basilar artery after subarachnoid hemorrhage in dogs. J. Neurosurg. 80:476-483; 1994. I I. Kimura, M.; Dietrich, H. H.; Dacey, R. G., Jr. Nitric oxide regulates cerebral arteriolar tone in rats. Stroke 25:2227-2234; 1994.
V A S O M O T O R C O N T R O L BY NITRIC O X I D E
12. Kovach, A. G. B.; Szabo, C.; Benyo, Z. Effects of N~Lnitro-u-argi nine and L-arginine on regional cerebral blood flow in the cat. J. Physiol. 449:183-196; 1992. 13. Moncada, S.; Palmer, R. M. J.; Higgs, E. A. Nitric oxide: Physiology, pathophysiology, and pharmacology. Pharmacol. Rev. 43:109142; 1991. 14. Morikawa, E.; Huang, Z.; Moskowitz, M. A. L-Arginine dilates rat pial arterioles by nitric oxide-dependent mechanisms and increases blood flow during focal cerebral ischemia. Br. J. Pharmacol. 107:905-907; 1992. 15. Nozaki, K.; Moskowitz, M. A.; Maynard, K. I.; Koketsu, N.; Dawson, T. M.; Bredt, D. S.; Snyder, S. H. Possible origins and distribution of immunoreactive nitric oxide synthase-containing nerve fibers in cerebral arteries. J. Cereb. Blood Flow Metab. 13:70-79; 1993. 16. Oyama, H.; Suzuki, Y.; Satoh, S.; Kajita, Y.; Takayasu, M.; Shibuya, M.; Sugita, K. Role of nitric oxide in the cerebral vasodilatory response to vasopressin and oxytocin in dogs. J. Cereb. Blood Flow Metab. 13:285-290; 1993.
369
17. Palmer, R. M. J.; Ashton, D. S.; Moncada, S. Vascular endothelial cells synthesize nitric oxide from u-arginine. Nature 333:664-666; 1990. 18. Rosenblum, W, I.; Nishimura, H.; Nelson, G. H. Endothelium-dependent L-arg-and u-NMMA-sensitive mechanism regulate tone of brain microvessels. Am. J. Physiol. 259:H1396-1401; 1990. 19. Seki, Y.; Suzuki, Y.; Baskaya, M. K.; Kano, T.; Saito, K.; Takayasu, M.; Shibuya, M.; Sugita, K. The effects of pituitary adenylate cyclase-activating polypeptide on cerebral arteries and vertebral artery blood flow in anesthetized dogs. Eur. J. Pharmacol. (in press). 20. Suzuki, Y.; Satoh, S.; Oyama, H.; Takayasu, M.; Shibuya, M. Regional differences in the vasodilator response to vasopressin in canine cerebral arteries in vivo. Stroke 24:1049-1054; 1993. 21. Takayasu, M.; Bassett, J. E.; Dacey, R. G., Jr. Effects of calcium antagonists on intracerebral penetrating arterioles in rats. J. Neurosurg. 69:104-109 1988. 22. Takayasu, M.; Kajita, Y.; Suzuki, Y.; Shibuya, M.; Sugita, K.; Ishikawa, T.; Hidaka, H. Triphasic response of rat intracerebral arterioles to increasing concentration of vasopressin in vitro. J. Cereb. Blood Flow Metab. 13:303-309; 1993.
Pergamon 0361-9230(95)02001-8
Regional Differences in Cerebral Vasomotor Control by Nitric Oxide YASUKAZU KAJITA,* MASAKAZU TAKAYASU,.1 YOSHIO SUZUKI* MASATO SHIBUYA* MASAYOSHI MORI,* HIROFUMI OYAMA,* KENICHIRO SUGITA* AND HIROYOSHI HIDAKA1-
Departments of *Neurosurgery and ?~Pharmacology,Nagoya University School of Medicine, Nagoya 466, Japan [Received 14 February 1995; Accepted 1 June 1995] itors generally decrease regional cerebral blood flow (rCBF) by blocking the basal release of NO, but the amount of rCBF reduction may differ among brain regions, depending upon the importance of L-arginine-NO pathway in regulating each regional circulation. Kovach et al., for example, have shown that Nr-nitroL-arginine caused the most striking decrease in rCBF in the cerebellum, pituitary, and medulla oblongata, and caused no decrease in rCBF of cerebral cortex or white matter [12]. In a previous study using cerebral angiography in dogs, we showed that vasodilation in response to intravenous injection of vasopressin, a stimulant of nitric oxide release, is greatest in the basilar artery and much less in the middle cerebral artery [20]. These data suggest that the role of the L-arginine-NO pathway in regulating cerebral circulation may be more pronounced in the posterior circulation than the anterior circulation. To test this hypothesis, we studied regional differences in the effects of a NO synthesis inhibitor, Nr-monomethyl-L-arginine, and a precursor of NO, L-arginine, on cerebral vessels from the anterior and the posterior circulation in vivo in the macrocirculation and in vitro in the microcirculation, using well-established methods. To study the responses of larger arteries, the diameter of the middle and anterior cerebral and the basilar arteries were evaluated angiographically in dogs. Microscopic evaluation of responses by the microcirculation was performed on isolated intracerebral arterioles branching from the middle cerebral and basilar arteries in rats.
ABSTRACT: Regional differences in the role of nitric oxide in cerebral vasomotor control were investigated with a nitric oxide synthesis inhibitor, NG-monomethyI-L-arginine, or a precursor of nitric oxide, L-arginine using both dog cerebral angiography for the larger artery study and rat isolated arterioles for the microcirculation study. Ne-monomathyI-L-arginine (10 pmol) constricted the dog cerebral artedes, by 15.6%, 17.5%, and 27.3% in the middle cerebral, antedor cerebral, and basilar artedes, respectively. The greater constriction of the basilar artery did not reach statistical significance. However, L-arginine (100 pmol) produced significantly greater dilation of basilar artedes than the middle cerebral or anterior cerebral (31.3% vs. 16.7% or 13.1%). NG-monomethyI-L-arginine at 10 -a M constricted rat arterioles originating from basilar arteries significantly more than the middle cerebral arteries (23% vs. 14%). L-arginine at 10 -a M dilated rat arterioles from basilar arteries significantly more than from the middle cerebral artery (24 vs. 11%). These findings suggest that the roles of nitric oxide in vasomotor control differs by region in the brain, and it may be greater in vessels of the postedor than of the antedor circulation. KEY WORDS: Arterioles, Cerebral artery, L-Arginine, NG-Mono methyl-L-arginine, Nitric oxide, Regional difference.
INTRODUCTION Nitric oxide (NO), synthesized from L-arginine by NO synthase (NOS), is a potent vasodilator that plays a substantial role in the regulation of cerebral circulation. NO was originally identified in vascular endothelial cells as a principal endothelium-derived relaxing factor (EDRF) [13]. Recent studies have shown that NOS also is expressed in neurons and in the perivascular fibers that innervate cerebral blood vessels, suggesting roles for NO both as a neurotransmitter and a regulator of regional cerebral blood flow control [3,13]. NOS is heterogenously expressed in brain parenchyma and in its perivascular tissue. NOS-expressing neurons are densest in the cerebellum, hypothalamus, and pituitary, and are sparser in cerebral cortex [3]. Recent studies have suggested that basal release of NO is responsible for the maintenance of normal cerebral blood flow, and that induced release may contribute to the vasodilation initiated by increased local neuronal activity or metabolism [4,6,9,18]. NO synthesis inhib-
METHODS Angiography of Cerebral Arteries in Dogs Cerebral angiography was performed on dogs as previously described in detail [10,16,20]. Briefly, mature mongrel dogs of either sex, weighing 8 to 20 kg, were anesthetized with intravenous pentobarbital (20 to 30 mg/kg), and ventilated through an endotracheal tube with room air by a respirator (model B2, Igarashi Ika Kogyo Co., Tokyo, Japan). The ventilation rate (15 cycles per min) and tidal volume (20 ml/kg) were adjusted to maintain arterial blood gases and pH within normal physiological limits. A catheter was placed in the left femoral artery to monitor mean arterial blood pressure and heart rate. Internal diameters of cerebral arteries were determined by vertebral angiography. A catheter was inserted directly into the right
t Requests for reprints should be addressed to Masakazu Takayasu, Department of Neurosurgery, Nagoya University School of Medicine, 65 Tsuruma-cho, Showa-ku, Nagoya 466, Japan. 365
366
vertebral artery just proximal to the transverse foramen of the C6 cervical vertebra. Angiograms were performed through this catheter using 3.0 ml of 65% iothalamate meglumine and were radiographed periodically at a fixed magnification following cisternal injection of 0.1 to 10 #mol NG-monomethyl-L-arginine (L-NMMA) or 1 to 100 #mol L-arginine, dissolved in 2.0 ml of physiological saline. Injection of solutions into the cistema magna was done following withdrawal of an equal volume of cerebrospinal fluid to maintain constant intracranial pressure. The animals were positioned with heads tilted, to increase the contact of the cisternal injections with the major arteries of the brain. Arterial diameters were measured before and after injection of drugs using a computerized image analysis system (IMAGE 1.27, Macintosh 2cx). Portions of the middle cerebral artery (MCA) (the proximal portion of the M-1 segment), the anterior cerebral artery (ACA) (the proximal portion of the A-2 segment), and the middle third of the basilar artery (BA) were measured.
KAJITA ET AL.
Statistical Analysis Vessel diameters at each agonist dose were expressed as a percentage of the control vessel diameter. Vasoconstriction and vasodilation were expressed as the percentage change in diameter from control vessel diameter, and the data are reported as means _ SD. The significance of differences among a set of three or more samples was evaluated by one-way analysis of variance (ANOVA) with the Fisher Protected Least Significant Difference (PLSD) multiple range test as a postANOVA test. The significance of differences between two unpaired samples, such as maximum responses in arterioles from the regions of the MCA and the BA, was evaluated by an unpaired t-test• Linear regression analysis and ANOVA were performed to examine the correlation between vessel diameter and responsiveness to L-NMMA or Larginine. We considered differences significant at p < 0•05. RESULTS
Microscopic Evaluation of Isolated lntracerebral Arterioles in Rats Intracerebral arterioles were isolated and cannulated in an organ bath (composed as described below), and changes in vessel diameter in response to the extraluminal administration of agents were measured, as previously described in detail 15,21,22]. Briefly, penetrating branching intracerebral arterioles, 10 to 70 /~m in diameter, were isolated surgically at 4°C from the first (M1) portion of a middle cerebral artery or the middle third of the basilar artery, from the brains of pentobarbital-anesthetizedmale Sprague-Dawley rats weighing 300-400 g. Vessels were transferred to a temperature-controlled chamber on the stage of an Olympus inverted microscope and were cannulated with glass pipettes at 25°C. The inner diameter of the vessel was determined using a video micro-scaler (FOR.A Model IV-550, Tokyo, Japan). After cannulation, transmural pressure was set and maintained throughout the experiments at 60 mm Hg via the cannulating pipette• Bath temperature was brought to 37.5°C, and the vessels were allowed to equilibrate for 45 min at an extraluminal bath pH of 7.3. During the equilibration period of 30 min, vessels developed spontaneous tone, contracting to about 70% of their maximum passive diameter. Vessel responsiveness was then assessed by changing extraluminal pH from 7.3 to 6.8 and from 7.3 to 7.6, The physiological salt solution (PSS) used in this preparation was a modified Ringer's, composed of 144 mM NaCI, 3.0 mM KCI, 2.5 mM CaCI2, 1.5 mM MgSO4, 5.0 mM glucose, 2,0 mM pyruvate, 0.02 mM ethylenediaminetetraacetic acid (EDTA), 2.0 mM 3-(N-morpholino)propanesulfonic acid (MOPS), 1.21 mM NaHzPO4, and 0.9-1.0 g/100 ml bovine serum albumin. The extraluminal solution contained no albumin in our experimental setup to simulate in vivo conditions [5]. Solution pH was maintained at 7.3 except when vessel responsiveness was assessed by pH changes. Control vessel diameter was defined as the diameter to which the vessels spontaneously contracted during the equilibration period at pH 7.3. L-NMMA and L-arginine were dissolved in PSS, and the pH was adjusted to 7.3 with HCI or NaOH as necessary. Dose-response curves to extraluminal L-NMMA and L-arginine were determined by sequential addition of solutions of increasing concentration to the organ bath at constant pH 7.3. The vessel diameter was allowed to stabilize for 5 min between each change in bath solution. The nitric oxide synthesis inhibitor [NG-monomethyl-L-argi nine; (L-NMMA)] was obtained from Calbiochem (La Jolla, CA, USA). L-arginine and all other chemicals were reagent grade.
Angiographic Study of Cerebral Arteries in Dogs Intracisternal injection of L-NMMA (0.1, 1, 10 #mol) produced a significant decrease in the internal diameters of the MCA, the ACA, and the BA on angiogram (Fig. 1). L-NMMA at 10 #mol constricted these arterial diameters by 15.6 + 11.1% (mean ___SD, MCA, n = 5), 17.5 +_ 4.0% (ACA, n = 3) and 27.3 _+ 7.0% (BA, n -- 5) at 15-60 min following intracisternal in-
°].-
L-arginine
3O
20.
I._
Q q)
•
10.
E
.ca
MCA
[] ACA []
BA
0I
10
100
0.1
1
10
Dose (ixmol)
Im
0 O~
C
¢= i-
-10-
¢0 -20°
.L
L-NMMA FIG. 1. Changes in diameter of the middle cerebral artery (MCA), the anterior cerebral artery (ACA), and the basilar artery (BA) in dog cerebral angiograms, in response to intracisternal injection of NC'-monomethyl-L-arginine(L-NMMA)(0.1, 1, or 10/~mol) or L-arginine(1, 10, or 100 #tool). Asterisks indicate significant differences (p < 0.05) in response between each pair of arteries.
V A S O M O T O R C O N T R O L BY NITRIC OXIDE
L-arginine 30.
20, ~r
it
it
Rat Arterioles -" MCA BA
G 10
9
8
7
6
5
-
,
-
•
-
4 g
3
Dose (-log M)
•
cO
367
that branched from the M C A and the BA, respectively. The responses of arterioles to L-NMMA were significantly greater in vessels originating from the BA than from the MCA (10 910 3 M) (Fig. 2). L-arginine caused a dose-dependent vasodilation with maximal increases in diameter of 11.1 _+ 5.4% (n - 5) and 24.3 _+ 11.3% in = 5) at 10 ~ M in arterioles from the M C A and the BA, respectively (Fig. 2). The responses of arterioles to L-arginine again were significantly larger in vessels from the BA than from the M C A (10 ~-10 :~M). Baseline characteristics of arterioles originating from the MCA and the BA that were studied were compared (Table 1). The control arteriolar diameter was 57.6 _+ 8.8 (54.6-81.8) # m of arterioles from the MCA region and 23.6 _+ 5.0 (13.0-29.7) #m of arterioles from the BA region. There were significant differences between the diameters of arterioles from these regions. However, there was no significant correlation between control vessel diameter and responsiveness to L-NMMA or L-arginine among arterioles from the M C A region or among those from the BA region, with p-values ranging from 0.20 to 0.74. There was no significant difference in responsiveness to pH changes between arterioles from the MCA and the BA.
~ .10. DISCUSSION .20-
-30
L-NMMA FIG. 2. Dose-response curves of isolated rat arterioles originating from the middle cerebral artery (MCA) and the basilar artery (BA), in response to extraluminal application of NG-monomethyl-L-arginine (L-NMMA) or L-arginine (10 ') M to 10 3 M). Asterisks indicate significant differences Lo < 0.05) in response between arterioles from the MCA and from the BA.
jection. The greater decrease in diameter of the BA than of the MCA or ACA did not reach statistical significance. L-arginine (1, 10, 100 #mol) caused significant vasodilation of all the arteries (Fig. 1). L-arginine, 100 #mol dilated the measured arterial diameters by 16.7 ___ 6.7% (MCA, n = 4), 13.1 _+ 4.3% (ACA, n = 4) and 31.3 _ 8,1% (BA, n = 5) at 1 5 - 6 0 min following intracisternal injection. The BA dilation in response to L-arginine was significantly greater than that of the M C A or ACA. Control diameters of the MCA, ACA, and BA were 1.18 _ 0.20, 0.74 _+ 0.16, and 0.56 _+ 0.11 mm, respectively. These diameters are significantly different from each other. However, there was no significant correlation between control vessel diameter and responsiveness to L-NMMA or L-arginine among different MCAs, ACAs, or BAs, with p-values ranging from 0.23 to 0.72. No significant change in mean arterial blood pressure occurred following intracisternal injections of L-NMMA or Larginine.
This study demonstrated regional differences in responsiveness of cerebral vessels to NO. A nitric oxide synthesis inhibitor, L-NMMA produced vasoconstriction in both larger cerebral arteries and smaller intracerebral arterioles. The vasoconstriction observed in vessels originating from the BA was greater than that in those from the MCA or ACA region. L-arginine, a precursor of NO, produced vasodilation in both larger cerebral arteries and smaller intracerebral arterioles. Again, the vasodilation was significantly greater in vessels from the BA than in those from the MCA or ACA. L-NMMA produced vasoconstriction in both large arteries and small arterioles provides evidence that a basal release of NO contributes to the maintenance of cerebral vascular resistance in these vessels. This study indicates that NO-mediated regulation of the resting vascular tone may be more pronounced in vessels of the posterior circulation than in those of the anterior circulation. These findings are consistent with those of a cerebral blood flow study using radiolabelled microspheres by Kovach et al. [12], that showed a significant decrease in rCBF in the cerebellum and brainstem following administration of a NO synthesis inhibitor, whereas there was no decrease in the rCBF in cerebral cortex and white matter. In a previous study using cerebral angiography in dogs, we showed that vasodilation in response to intravenous injection of vasopressin, a stimulant of nitric oxide TABLE 1 BASELINE CHARACTERISTICS OF ARTERIOLES IN THIS STUDY
pH Response (%) Region of Arterioles
BA MCA
n
Diameter (,um)
pH 6.8
pH 7.6
10 10
23.6 _+ 5.0 57.6 _+ 8.8
13.4 __+2.0 13.9 _+ 5.4
13.0 _+ 2.8 13.2 _+ 2.0
0.0001 *
0.64
0.89
Microscopic Study of Isolated Cerebral Arterioles From Rats
ANOVA: p value
Extraluminal t,-NMMA produced a dose-dependent vasoconstriction with maximal decreases in diameter of 14.2 ± 2.7% (n = 5) and 23.0 _+ 7.4% (n = 5) at 10 4 M in isolated arterioles
Values are those prior to application of drugs. Data are means _+ SD. BA, the basilar artery; MCA, the middle cerebral artery; ANOVA, analysis of variance; *significant difference.
368
release, is greatest in the BA and much less in the MCA, that is, similar to the results of this study [20]. All of these findings suggest that vessels from the posterior circulation may be subject to greater basal vasomotor control by NO-mediated mechanisms than those from the anterior circulation. L-arginine is a precursor of nitric oxide [17]. The effects of exogenous L-arginine on the vessels, however, have been complex, and inconsistent results have been documented in the literature [4,6,1 !,12,14,18]. If L-arginine is present in suboptimal amounts in the NO-producing cells, t=arginine may be rate-limiting, and exogenous u-arginine may produce vasodilation by supplying more substrate for NO production. On the other hand, exogenous L-arginine is not rate-limiting for NOS when l=arginine is present in optimal amounts in the cells. Most of these studies have been performed in vivo, with L-arginine supplied intravenously or topically. In vivo, exogenous L-arginine may not reach NOS containing cells freely in a certain condition, because glial cells tend to sequester this amino acid and endothelial cells express selective uptake and transport mechanisms [ 1,2]. Few in vitro studies have examined the effects of L-arginine except for the one by Kimura et al. [11]. They used a similar method to ours and observed a tendency for vasodilation by L-arginine, which, however, did not reach statistical significance. This study showed that exogenous L-arginine caused vasodilatation of rat cerebral arterioles in vitro, as well as of canine large arteries in vivo, suggesting that t.-arginine could be rate-limiting factor for NO production under normal physiological conditions. It is not clear why vessels in the region of the basilar artery appear to show a more pronounced vasomotor response to NO than do those from the region of the middle or anterior cerebral artery. We wondered whether the different average diameters of vessels in these regions might influence the magnitude of NOrelated vasomotor response. Arterioles from the BA region are smaller in diameter than those from the MCA region, and smaller arterioles may be more responsive than larger arterioles. However, this explanation appears unlikely, because there was no difference in response to pH 6.8 acidosis or to pH 7.6 alkalosis between arterioles originating from the BA and those from the MCA, and because there was no correlation between the vessel diameter and responsiveness of arterioles within each these regions. Moreover, the associations were reversed in the larger cerebral arteries in dogs, where the BA was more responsive to L-arginine than the M C A or the ACA, whereas it was larger in diameter than both the MCA and the ACA. Again, there was no correlation between the vessel diameter and responsiveness of arteries within each these regions. In angiographic study of dog cerebral arteries, concentration gradient of the drug should be considered because drugs were administered from the cisterna magna, but it may not be a main factor. Using the same mothod, we demonstrated that the effectiveness of pituitary adenylate cyclase-activating polypeptide were in the reverse order, that is, it was more effective to the ACA and MCA than the BA [19]. Therefore, the vessels from the BA region either produce more NO or may be more sensitive to NO than those from the MCA or the ACA region. Immunohistochemical studies have demonstrated that nitric oxide synthase-containing nerve fibers associated with cerebral arteries are abundant and dense in the proximal anterior and middle cerebral arteries and are less numerous in the caudal circle of Willis, including the basilar artery [8,15]. These apparently discrepant findings may suggest that basal or stimulated release of NO through vascular endothelium may be greater in the basilar artery region than those from the middle or anterior cerebral artery region, whereas the relative contribution of NO-related neurogenic control would be the reverse. Another possible expla-
KAJITA ET AL.
nation is that there might be a difference in sensitivity to nitric oxide or in the activity of NOS in the different regions of the cerebral vasculature. The clinical significance of these results could be imagined. Because the vessels in the posterior circulation seem to be controlled in a more dilated state by NO, there may be less room for further vasodilation in these vessels. This could explain the fact that the cause of the transient ischemic attacks (TIA) in the posterior circulation is more often hemodynamic rather than embolic; the latter is more common in the TIA in the anterior circulation. The role of NO in the regulation of small cerebral arteriolar tone remains controversial. Faraci has reported, in an in vivo study, that NO might have more influence on the regulation of basal tone in the larger cerebral arteries than in cerebral arterioles [7]. In this study, I=NMMA produced a significant vasoconstriction of rat intracerebral arterioles originating from the MCA and the BA, by as much as 14% and 23%, respectively, and 1=arginine caused significant vasodilation of arterioles from the MCA and the BA regions by as much as 11% and 24%, respectively. The magnitudes of L-NMMA-induced vasoconstriction and of L-arginine-induced vasodilation were similar in dog large cerebral arteries in vivo and in rat smaller intracerebral arterioles in vitro. The observed regional heterogeneity of NO-related vasomotor responses between the anterior and the posterior circulation thus may be greater than the longitudinal heterogeneity between large arteries and smaller arterioles. In conclusion, we have shown that the role of the L-argininenitric oxide pathway may differ by region in the cerebral circulation, and that it may be greater in vessels of the posterior circulation than of the anterior circulation.
REFERENCES 1. Aoki, E.; Semba, R.; Mikoshiba, K.; Kashiwamata, S. Predominant localization in glial cells of free L-arginine: Immunocytochemical evidence. Brain Res. 547:190-192; 1991. 2. Bogle, R. G.; Moncada, S.; Pearson, J.; Mann, (3. E. Identification of inhibitors of nitric oxide synthase that do not interact with endothelial cell l~-arginine transporter. Br. J. Pharmacol. 105:768-770; 1992. 3. Bredt, D. S.; Hwang, P. M.; Snyder, S. H. Localization of nitric oxide synthase indicating a neural role for nitric oxide. Nature 347:768-770; 1990. 4. Busija, D. W.; Leffler, C. W.; Wagerle, L. C. Mono-l=arginine-containing compounds dilate piglet pial arterioles via an endotheliumderived relaxing factor-like substance. Circ. Res. 67:1374-1380; 1990. 5. Dacey, R. (3., Jr,; Duling, B. R. A study of rat intracerebral arterioles: Methods, morphology, and reactivity. Am. J. Physiol. 243:H592H606; 1982. 6. Faraci, F. M. Role of nitric oxide in regulation of basilar artery tone in vivo. Am. J. Physiol. 259:H1216-H1221; 1990. 7. Faraci, F. M. Role of endothelium-derived relaxing factor in cerebral circulation: Large arteries vs. microcirculation. Am. J. Physiol. 261 :H1038-HI042; 1991. 8. ladecola, C. Regulation of the cerebral microcirculation during neural activity: Is nitric oxide the missing link? T.I.N.S. 16:206-215; 1993. 9. ladecola, C.; Beitz, A. J.; Renno, W. Nitric oxide synthase-containing neural processes on large cerebral arteries and cerebral microvessels. Brain Res. 606:148-155; 1993. 10. Kajita, Y.; Suzuki, Y.; Oyama, H.; Tanazawa, T.; Takayasu, M.; Shibuya, M.; Sugita, K. Combined effect of t=arginine and superoxide dismutase on the spastic basilar artery after subarachnoid hemorrhage in dogs. J. Neurosurg. 80:476-483; 1994. I I. Kimura, M.; Dietrich, H. H.; Dacey, R. G., Jr. Nitric oxide regulates cerebral arteriolar tone in rats. Stroke 25:2227-2234; 1994.
V A S O M O T O R C O N T R O L BY NITRIC O X I D E
12. Kovach, A. G. B.; Szabo, C.; Benyo, Z. Effects of N~Lnitro-u-argi nine and L-arginine on regional cerebral blood flow in the cat. J. Physiol. 449:183-196; 1992. 13. Moncada, S.; Palmer, R. M. J.; Higgs, E. A. Nitric oxide: Physiology, pathophysiology, and pharmacology. Pharmacol. Rev. 43:109142; 1991. 14. Morikawa, E.; Huang, Z.; Moskowitz, M. A. L-Arginine dilates rat pial arterioles by nitric oxide-dependent mechanisms and increases blood flow during focal cerebral ischemia. Br. J. Pharmacol. 107:905-907; 1992. 15. Nozaki, K.; Moskowitz, M. A.; Maynard, K. I.; Koketsu, N.; Dawson, T. M.; Bredt, D. S.; Snyder, S. H. Possible origins and distribution of immunoreactive nitric oxide synthase-containing nerve fibers in cerebral arteries. J. Cereb. Blood Flow Metab. 13:70-79; 1993. 16. Oyama, H.; Suzuki, Y.; Satoh, S.; Kajita, Y.; Takayasu, M.; Shibuya, M.; Sugita, K. Role of nitric oxide in the cerebral vasodilatory response to vasopressin and oxytocin in dogs. J. Cereb. Blood Flow Metab. 13:285-290; 1993.
369
17. Palmer, R. M. J.; Ashton, D. S.; Moncada, S. Vascular endothelial cells synthesize nitric oxide from u-arginine. Nature 333:664-666; 1990. 18. Rosenblum, W, I.; Nishimura, H.; Nelson, G. H. Endothelium-dependent L-arg-and u-NMMA-sensitive mechanism regulate tone of brain microvessels. Am. J. Physiol. 259:H1396-1401; 1990. 19. Seki, Y.; Suzuki, Y.; Baskaya, M. K.; Kano, T.; Saito, K.; Takayasu, M.; Shibuya, M.; Sugita, K. The effects of pituitary adenylate cyclase-activating polypeptide on cerebral arteries and vertebral artery blood flow in anesthetized dogs. Eur. J. Pharmacol. (in press). 20. Suzuki, Y.; Satoh, S.; Oyama, H.; Takayasu, M.; Shibuya, M. Regional differences in the vasodilator response to vasopressin in canine cerebral arteries in vivo. Stroke 24:1049-1054; 1993. 21. Takayasu, M.; Bassett, J. E.; Dacey, R. G., Jr. Effects of calcium antagonists on intracerebral penetrating arterioles in rats. J. Neurosurg. 69:104-109 1988. 22. Takayasu, M.; Kajita, Y.; Suzuki, Y.; Shibuya, M.; Sugita, K.; Ishikawa, T.; Hidaka, H. Triphasic response of rat intracerebral arterioles to increasing concentration of vasopressin in vitro. J. Cereb. Blood Flow Metab. 13:303-309; 1993.