Cell Isolation
Rat basilar smooth muscle cells were isolated as previously described. Briefly, Sprague-Dawley female rats were anesthetized with Metofane and decapitated. The basilar arteries were immediately removed to a medium consisting of (mM): 130 NaCl, 5 KCl, 0.8 CaCl2, 1.3 MgCl2, 5 glucose, 10 N-2-hydroxyethylpiperazine- N'-2-ethanesulfonic acid (HEPES), penicillin (100 units/mL), and streptomycin (0.1 g/L). Subsequently, the arteries were cut into 0.2 mm rings and incubated at room temperature in a medium containing collagenase, elastase, hyaluronidase, and deoxyribonuclease. After one hour the rings were triturated gently, plated on glass cover slips and stored at 4oC for 2 hours for recovery.

[Ca2+]i Microfluorimetry
The buffer solution for [Ca2+]i measurement was (mM): 145 NaCl, 2 CaCl2, 3 KCl, 1 MgCl2, 10 HEPES, 10 glucose and the pH was adjusted to 7.4 with NaOH. Cells were loaded with the fluorescence indicator Fura-2/AM (3 :M) for 30 min at room temperature in the extracellular buffer solution. After loading, ells were perfused for 20 min before the experiment to allow de-esterification of the dye.

Digital [Ca2+]i imaging was performed by video microfluorimetry using an intensified CCD camera (Hamamatsu, Bridgewater, NJ) coupled to a Nikon Diaphot microscope (40x Fluor objective, Nikon Inc. New York) and software (Universal Imaging Corp. West Chester, PA) on a 486 personal computer. Sample illumination was supplied by a 150 W-Xenon arc lamp, and excitation wavelengths were selected by computer control of a filter wheel. Fluorescence imaging was obtained with alternating excitation wavelengths of 340 and 380 nm, and an emission wavelength of 510 nm through the CCD camera.

Data Analysis
Data are expressed as mean " SD. Statistical differences between the control and other groups were compared using one-way analysis of variance (ANOVA) and then Tukey-Kramer multiple comparisons procedure (95% lower and upper confidence interval) if significant variance was found, and a value of P <0.05 was considered statistically significant.

General Effects of Nucleotides
Application of ATP, UTP, TTP (100 nM - 1 mM), and at higher concentrations CTP, GTP or ITP (0.1 -1 mM), produced a transient [Ca2+]i peak followed by a steady-state [Ca2+]i plateau which decreased slowly to a level slightly above the resting level after 20 min application in rat basilar smooth muscle cells. In Ca2+-free buffer solution, ATP, UTP and TTP produced a transient [Ca2+]i peak which returned to the baseline rapidly without a plateau phase which suggests that the [Ca2+]i peak was Ca2+ released from intracellular stores and the [Ca2+]i plateau was Ca2+ which entered from the extracellular space. The effect of these nucleotides on [Ca2+]i was dose-dependent, completely reversible by washout and repeatable.

Effect of Different Nucleotides
We tested the effect of different nucleotides to raise [Ca2+]i in cells. The maximum responses to ATPgS, UTP, UDP, ATP, TTP and ITP were obtained and the ED50 values for these nucleotides were calculated from nonlinear regression. The -log10 ED50 values were ATPgS 6.5, UTP 6.1, UDP 5.6, ATP 5.8, TTP 5.4 and ITP 4.6, respectively. Since the responses to CTP and GTP failed to reach the maximum and since UMP and uridine failed to produce marked effect, their ED50 values were not calculated. The potency of nucleotides to raise [Ca2+]i was ATP(S (selective agonist for P2u receptors), UTP/UDP (selective agonists for P2u receptors) ³ ATP (selective agonist for P2 receptors) » TTP. GTP, CTP, UMP and uridine failed to produce marked response at concentration below 1 mM. Since UTP, UDP and ATP(S are selective agonists for P2u but not for P2x or P2y receptors and since UTP, UDP and ATP(S were equally (or more) potent to ATP, these studies indicate the predominance of P2u receptors in rat basilar smooth muscle cells.

Signal Transduction Pathways of P2u Receptors
Since UTP was one of the most potent agonists among these nucleotides, we used UTP as the agonist to study signal transduction pathways in rat basilar smooth muscle cells. The [Ca2+]i responses to UTP (100 :M) in cells pretreated with pertussis toxin (1200 ng/ml) for 10 hrs were compared with those of controls. Pretreatment with pertussis toxin markedly reduced the [Ca2+]i responses to UTP. The phospholipase C (PLC) inhibitor 2-nitro-4-carboxyphenyl N,N-diphenylcarbamate (NCDC) (10 :M) was tested and found, after 15 min incubation with cells, to abolish the [Ca2+]i responses to UTP. Since activation of phospholipase C produces inositol 1,4,5-triphosphate (IP3) which releases Ca2+ from internal stores, we tested the effect of UTP on thapsigargin-sensitive Ca2+ stores. In the absence of external Ca2+, thapsigargin (100 nM), a well-known Ca2+ pump inhibitor, produced a transient [Ca2+]i rise (leak) which lasted about 5-7 min before falling to a level similar to the resting level. The following application of UTP (100 :M), in the presence of thapsigargin, failed to induce any marked [Ca2+]i elevation.

The pathways for Ca2+ entry activated by UTP were studied by using the receptor-operated Ca2+ influx blocker SK&F96365 and the voltage-dependent Ca2+ channel blocker verapamil. SK&F 96365 (5 :M) markedly and verapamil (1 :M) partially reduced the [Ca2+]i plateau phase induced by UTP without significant effect on the [Ca2+]i peak. Lanthanum (100 mM), an inorganic Ca2+ channel blocker, abolished both peak and plateau responses induced by UTP.

P2-purinoceptor Antagonists [Ca2+]i response to UTP was further examined by using P2 receptor antagonists. Suramin, a selective P2 receptor antagonist, and PPADS, a selective P2x receptor antagonist15 were incubated with cells for 5 min before UTP (100 :M) was applied. PPADS at 10 :M failed to reduce the effect of UTP, the concentration which inhibited the effect of ATP on P2x receptors15. Pre-incubation with suramin (1 :M) significantly reduced both the peak and plateau responses induced by UTP.

Summary
We have demonstrated that UTP, ATP, TTP and, at higher concentrations, ITP, GTP and CTP increase [Ca2+]i in rat basilar smooth muscle cells. Selective P2u receptor agonists, ATP(S, UTP and UDP, produced [Ca2+]i responses similar to that of ATP, consistent with our previous report that P2u receptors mediate effects of nucleotides in rat basilar artery smooth muscle cells. Using pertussis toxin, NCDC and thapsigargin, we have shown that the P2u receptor in these cells was coupled to G protein and, by activation of PLC, released Ca2+ from intracellular thapsigargin-sensitive stores. A similar mechanism of Ca2+ mobilization by activation of P2u receptors has been reported previously in other tissues. The partial retention of the [Ca2+]i response to UTP after 10 hrs treatment with a high concentration of pertussis toxin (1200 ng/ml) indicates that other G-proteins which are not sensitive to pertussis toxin may also be involved in the UTP induced [Ca2+]i response. It is established that PLC can be activated either by an "-subunit of the pertussis toxin-insensitive Gq/GII family or the $(-dimer of the pertussis toxin-sensitive Gi216. NCDC at 10 :M completely abolished the [Ca2+]i response to UTP indicating UTP-induced intracellular Ca2+ mobilization was mediated by activation of PLC in these cells.

Furthermore, we have demonstrated that 1 :M suramin inhibited the [Ca2+]i response to UTP, consistent with our previous founding that suramin reduced the effect of ATP-containing erythrocyte lysate on [Ca2+]i in rat basilar smooth muscle cells and contraction in dog basilar artery. PPADS, a selective antagonist for P2x receptors, failed to significantly reduce the effect of UTP at concentrations that blocked the effect of P2x receptors in other tissues.

It has been established that P2u (or P2y217) receptors are G-protein coupled receptors, and that signal transduction is mediated by the receptor-G protein-PLC-IP3 system, which mobilizes internal Ca2+ stores. Store depletion triggers Ca2+ entry via voltage-independent Ca2+ pathways, possibly by diffusible chemical messengers. Ca2+ entry through voltage-dependent Ca2+ channels may also partially contribute to UTP-induced [Ca2+]i elevation. We have demonstrated that both receptor-operated Ca2+ influx inhibitor SK&F 96365 and voltage-dependent Ca2+ channel blocker verapamil markedly reduced the Ca2+ entry induced by UTP, consistent with other reports using peripheral vasculatures. SK&F 96365 but not verapamil significantly reduced the Ca2+ entry induced by erythrocyte lysate in cultured endothelial cells in our previous study. Lanthanum completely abolished both peak and plateau [Ca2+]i responses to UTP in this study, consistent with our previous results that lanthanum abolished [Ca2+]i response to erythrocyte lysate. The mechanism by which lanthanum blocks both peak and plateau [Ca2+]i responses is not clear but may be explained by its multiple actions against Ca2+ entry via membrane Ca2+ channels, Na+/Ca2+ exchangers, and Ca2+-ATPase pumps. Cytosolic Ca2+ binds to calmodulin, and the Ca2+-calmodulin complex subsequently activates myosin light chain kinase which phosphorylates the light chain of myosin thus activating the Mg2+-adenosine triphosphatase activity, which promotes the formation of myosin-actin cross-bridges. Several recent investigations indicated that ATP and UTP elevate [Ca2+]i and contract cerebral arteries and that these effects of ATP and UTP were prevented by P2 receptor antagonists.

Cytoplasmic ATP may be released following cell lysis or selective permeabilization of the plasma membrane. This permeabilization can occur in various types of cells, including the smooth muscle cells, in the absence of irreversible damage, such as hypoxia. Exocytosis of secretory granules such as platelet dense bodies (> 600 mM ATP) also contributes to the presence of extracellular ATP, as well as other nucleotides such as ADP (" 400 mM), GTP and UTP. After aneurysmal SAH, ATP, which could be released from blood clot, would contact major cerebral arteries from the adventitia side, activate P2x and P2u receptors in smooth muscle cells and may produce vasospasm. Since a high level of ATP, ADP and UTP exists in blood cells, these blood cells may play important roles in the pathogenesis of vasospasm. ATP and ADP produce contraction by activation of P2 receptors or by releasing vasocontractile agents such as eicosanoids from cerebral endothelial cells.

The ability of UTP, UDP, ATP and TTP to elevate [Ca2+]i indicates that multiple nucleotides may have vasoactive effects in cerebral vasculature. The vasoconstrictive properties of UTP and its possible role in chronic cerebral vasospasm have been investigated. UTP induced long-lasting contractions of isolated human brain arteries up to 20-24 hrs. Both UTP and UDP produced vasoconstriction of canine cerebral arteries and intracisternal application of UTP produced cerebral vasospasm in dogs. The vasoconstrictive effect of different nucleotides was tested in rabbit basilar artery. The rank order of potency of the pyrimidine nucleotides was UTP = UDP >> UMP =CTP; that of the purine nucleotides was ATP(S > AMP-PNP > ATP > ADP > 2-methylthio-ATP = ",$-methylene-ATP = $,(-methylthio-ATP23. UTP produced much stronger contractions than ATP in canine basilar arteries. We have obtained a similar effect for nucleotides to raise [Ca2+]i in rat basilar smooth muscle cells in this study. Since UTP, UDP and ATP(S are all selective agonists for P2u receptors, these studies indicate a predominance of P2u receptors in cerebral artery. ATP, however, may be a primary candidate for eliciting vascular responses such as vasospasm following SAH due to its abundance relative to the other nucleotides inside all cells and especially in red blood cells and platelets. ATP produced endothelium-dependent contraction by releasing endothelium-derived contracting factors possibly thromboxane A2 and endothelium-independent relaxation by activation of P1 receptors in canine basilar artery. The endothelium-independent and endothelium-dependent contractions induced by ATP were mediated by P2x and P2y receptors in canine basilar artery respectively. In cannulated rabbit basilar artery, application of ATP or UTP from adventitia side caused contraction but not relaxation. Application of ATP directly into canine basilar artery in the canine model of vasospasm, in an attempt to relieve vasospasm, further aggravated vasospasm accompanied by disruption of endothelium. Adenine nucleotides were detected in erythrocyte lysate and were suggested to mediate the effect of erythrocyte lysate on [Ca2+]i and contraction in cerebral arteries.

There are challenges to the hypothesis that UTP and ATP mediate the smooth muscle contraction observed in vasospasm. Extracellular ectonucleotidases rapidly metabolize ATP and other nucleotides. In order to elicit vasoconstriction, the release of UTP or ATP must overcome degradation by extracellular nucleotides hydrolyzing enzymes. However, subsequent metabolism of ATP to di- and monophosphates and adenosine by ectonucleotidases may enhance the effects of ATP. Future studies determining the time course of the concentrations of nucleotides in cerebrospinal fluid and blood clot during vasospasm may provide answers to these questions.

Bevan JA, Bevan RD. Arterial wall changes in chronic cerebrovasospasm: in vitro and in vivo pharmacological evidence. Ann Rev Pharmacol Toxicol 1988;28:311-329.

Aoki T, Takenaka K, Suzuki S, Kassell N, Sagher O, Lee K. The role of hemolysate in the facilitation of oxyhemoglobin-induced contraction in rabbit basilar arteries. J Neurosurg 1994;81:261-266.

Zhang H, Weir B, Marton L, Macdonald RL, Bindokas V, Miller R, Brorson J. Mechanisms of hemolysate-induced calcium elevation in cerebral smooth muscle cells. Am J Physiol 1995;269:H1874-1890.

Sima B, Macdonald RL, Marton LS, Weir B, Zhang J. Effect of P2-purinoceptor antagonists on hemolysate- and ATP-induced contractions of dog basilar artery in vitro. Neurosurgery 1996;39:815-822.

Dubyak GR, El-Moatassim C. Signal transduction via P2-purinergic receptors for extracellular ATP and other nucleotides. Am J Physiol 1993;265:C577-C606.

Zhang H, Weir B, Macdonald RL, Marton L, Solenski N, Kwan J, Lee K. Mechanisms of [Ca2+]i elevation induced by erythrocyte components in endothelial cells. J Pharmacol Exp Ther 1996;277:1501-1509.

Yuan Y, Granger HJ, Zawieja DC, DeFily DV, Chilian WM. Histamine increases venular permeability via a phospholipase C-NO synthase-guanylate cyclase cascade. Am J Physiol 1993;264:H1734-H1739.

Macdonald RL, Wang X, Zhang J, Marton L. Molecular changes with subarachnoid hemorrhage and vasospasm. In: The Molecular Biology of Neurosurgical Disease (Raffel C, Harsh G, eds), Baltimore, Williams & Wilkins, 1997, pp 278-293.

Barnard EA, Burnstock G, Webb TE. G protein-coupled receptors for ATP and other nucleotides: a new receptor family. Trends Pharmacol Sci 1994;15:67-70.

Shirasawa Y, White RP, Robertson JT. Mechanisms of the contractile effect induced by uridine 5-triphosphate in canine cerebral arteries. Stroke 1983;14:347-354.

Von Khgelgen I, Starke K. Evidence for two separate vasoconstriction-mediating nucleotide receptors, both distinct from the P2x-receptors, in rabbit basilar artery: a receptor for pyrimidine nucleotides and a receptor for purine nucleotides. Naunyn-Schmiedeberg=s Arch Pharmacol 1990;341:538-546.

Ikeuchi Y, Nishizaki T. ATP activates the potassium channel and enhances cytosolic Ca2+ release via a P2y purinoceptor linked to pertussis toxin-insensitive G-protein in brain artery endothelial cells. Biochem Biophys Res Commun 1995;215:1022-1028.

Ralevic V, Burnstock G. Roles of P2-purinoceptors in the cardiovascular system. Circulation 1991;84:1-14.

Urquilla PR. Prolonged contraction of isolated human and canine cerebral arteries induced by uridine 5'-triphosphate. Stroke 1978;9:133-136.

Haciyakupo™lu S, Kaya M, Cetinalp E, Yucesoy A. Effect of prostacyclin and adenosine triphosphate on vasospasm of canine basilar artery. Surg Neurol 1985;24:126-140.