Journal of Neurochemiswy Raven Press, Ltd., New York 0 1992 International Society for Neurochernistry

Receptor-Linked Hydrolysis of Phosphoinositides and Production of Prostacyclin in Cerebral Endothelial Cells Jian Xu, Zhi-Xiang Qu, *Steven A. Moore, TChung Y . Hsu, and Edward L. Hogan Department of Neurology, Medical University of South Carolina, Charleston, South Carolina; *Department of Pathology, University of Iowa, Iowa City, Iowa: and tDivision of Restorative Neurology, Baylor College of Medicine, Houston, Texas, U.S.A.

Abstract: The receptor agonist-mediated hydrolysis of phosphoinositides and production of prostacyclin were studied in murine cerebral endothelial cells (MCEC). Of 1 1 neurotransmitters and neuromodulators examined, carbachol, noradrenaline (NE), bradykinin, and thrombin significantly increased 3H-inositol phosphate accumulation in the presence of LiCl (20 mM). The maximal stimulation of [3H]inositol monophosphate ( ['H]IPI ) reached approximately 1 1, 1 1, seven, and four times the basal levels for carbachol, NE, bradykinin, and thrombin, respectively The ECso values of IPI accumulation for carbachol and NE were 34 and 0.16 pM, respectively. The muscarinic antagonists, atropine and pirenzepine, blocked the carbachol-induced IP, accumulation with Kivalues of 0.3 and 30 nM. respectively. The adrenergic antagonist, prazosin, blocked NE-induced IPI accumulation

with a Ki of 0.1 nM. The calcium ionophore A23 187, histamine, glutamate, vasopressin, serotonin, platelet activating factor, and substance P did not stimulate IPI accumulation. A23 187, bradykinin, and thrombin stimulated prostacyclin release to approximately four, four, and two times the basal levels, respectively, whereas carbachol and NE had little effect upon prostacyclin release. These results suggest that the activation of phospholipase C and of phospholipase Az in MCEC are regulated separately. Key Words: BradykininCalcium ionophore-Carbachol-Norepinephrine-Thrombin-Endothelial cells-Prostacyclin. Xu J. et al. Receptorlinked hydrolysis of phosphoinositides and production of prostacyclin in cerebral endothelial cells. J. Neurochem. 58, 1930-1935 (1992).

The cerebral endothelium plays an important role in the maintenance of the blood-brain barrier, where its plasma membrane forms the interface between blood-borne humoral factors and brain parenchyma (Abbott and Revest, 1991). Modulation of brain endothelial second messengers has been associated with changes in blood-brain barrier permeability (Jo6 and Klatzo, 1989). These cells are also known to produce a variety of compounds that mediate endothelium-dependent relaxation and endothelium-dependent constriction (Kontos et al., 1990; Yoshimoto et al., 1990). It is, therefore, important to examine agonist-mediated events in the brain endothelium. Two such events are the hydrolysis of polyphosphoinositidesand prostaglandin production. In several different cell types, a variety of stimuli are known to ac-

tivate phospholipase C (PLC) which catalyzes hydrolysis of phosphatidylinositol 4,5-bisphosphate to form inositol 1,4,5-trisphosphate (IP,) and diacylglycerol (DG). IP3 release of intracellular calcium and DG activation of protein kinase C are early steps in the sequence of some postreceptor events (Berridge and Irvine, 1984; Chuang, 1989) that may mediate changes in endothelial cell (EC) permeability (Abbott and Revest, 199 1). Another receptor-mediated event is release of free fatty acids, including arachidonic acid, mainly through the phospholipase A2 (PLA2)pathway. In EC, prostacyclin (PG12)is the major arachidonic acid metabolite (Salzman et al., 1980; Gerritson and Printz, 1981; Moore et ai., 1988). In some EC, the production of PGI2, a vasodilator and inhibitor of platelet aggregation, has been demonstrated in response to numerous

Received June 4, 1991; revised manuscript received October 1, 1991;-acceptedOctober 24, I99 1. Address correspondence and reprint requests to Dr. E. L. Hogan at Department of Neurology, Medical University of South Carolina, 171 Ashley Ave. Charleston, SC 29425, [J.S.A. or to Dr. C. Y. Hsu at Baylor College of Medicine, Division of Restorative Neurology, One Baylor Plaza, Room S8 15, Houston, TX 77030, U.S.A.

Abbreviarions used:DG,diacylglycerol;EC, endothelial cell(s);IP, inositol phosphate; IP, , inositol monophosphate; IP,, inositol 1,4bisphosphate; IP3, inositol 1,4,5-trisphosphate;6-keto-PGFI,, 6-ketoprostadandin Fie; MCEC, m u r k cerebral endothelial cell(s); NE, norepinephrine; PGIz,prostacyclin; PI, phosphatidylinositol; PLAz, phospholipase A2 ; PLC, phospholipase C.

I930

RECEPTOR-LINKED PI HYDROLYSIS AND PGI2 PRODUCTION

agonists, including bradykinin, thrombin, ionophore A23 187, histamine, and y-interferon, that are also known to activate phosphoinositide turnover (Forsberg et al., 1987; Resink et al., 1987; DOrlbans-Juste et al., 1989;Mattila and Renkonen, 1989; Garcia et al., 1990; Leszczynski and Ustinov, 1990). EC from vessels of different sizes or from different vascular tissue territories are heterogeneous by morphology, biochemistry, proliferative rate, and response to injury (Fajardo, 1989). Hence, it is not surprising that the rise in inositol phosphates (IPS) and the synthesis of PG12 induced by a number of agonists vary between EC from different sources. For example, stimulation of IP3 and PG12production by bradykinin has been found in bovine aortic EC (Derian and Moskowitz, 1986; Bartha et al., 1989) and in porcine aortic EC (Lambert et al., 1986),but not in human umbilical vein EC (Bartha et al., 1989). Thrombin stimulates only PG12 production in human umbilical vein EC (Jaffe et al., 1987; Bartha et al., 1989). A23187 stimulates PG12production in EC from most sources (Clark et al., 1986; Demolle and Boeynaems, 1988; Moore et al., 1988), but does not universally increase EC production of IPS (Derian and Moskowitz, 1986; Moscat et al., 1988). The studiesjust cited were performed predominantly in large-vessel endothelium, and the participation of IPS and PG12 in the differential responsiveness of cerebral EC to physiologic and pharmacologic agonists has not yet been addressed in detail. To this end, we now report the study of phosphotidylinositol (PI) hydrolysis and PG12 production in cultured murine cerebral EC (MCEC) in response to a variety of stimuli. MATERIALS AND METHODS M 199 Hank's medium, fetal calf serum, and supplements, including vitamins and amino acids, were obtained from GIBCO (Grand Island, N Y , U.S.A.). Bacto-peptone was obtained from DIFCO (Detroit, MI, U.S.A.). my~-[~H]Inositol (1 7.4 Ci/mmol) was from Amersham (Arlington Heights, IL, U.S.A.). All other chemicals and drugs were from Sigma (St. Louis, MO, U.S.A.), except as otherwise noted in the text.

Cell culture MCEC cultures were prepared and characterized as described previously (DeBault et al., 1979, 1981). Briefly, the brains of weanling young adult Swiss-Webster mice were aseptically removed and disrupted in a Dounce homogenizer. Microvessels were collected on either nylon mesh (150 pm) or glass beads and plated onto tissue culture dishes. EC migrating from vessels were pooled to form a culture of proliferating endothelium that was maintained in medium containing 10%fetal bovine serum, EC identity and culture purity were established by light and electron microscopic appearance, thrombomodulin activity, uptake of acetylated low density lipoprotein (DiI-Ac-LDL; Moore et al., 1990), and Griffonia simplicifolia agglutinin cytofluorimetry (Sahagun et al., 1989). Contamination of cultures by smooth muscle and astrocytes was assessed by immunohistochemical staining with anti-a-actin (Tsukada et al., 1987) and anti-glial fibdlary acidic protein (DAKO, Santa Barbara, CA, U.S.A.) antibod-

1931

ies, respectively. The cell cultures contained 90-989'0 EC, with smooth muscle comprising the majority of the contaminating cells. EC (passage 12-20) were plated on six-well or 12-we11 clusters, maintained in M 199 Hank's medium with 10%fetal calf serum, and grown to confluence for study.

Measurement of phosphoinositide hydrolysis in MCEC The hydrolysis of phosphoinositides was expressed as the accumulation of inositol monophosphate (IPI), inositol 1,4bisphosphate (IP2), or IP3 in the presence of LiCl (Benidge et al., 1982; Xu and Chuang, 1987~).Briefly, the growth medium (2 ml in a six-well cluster) was aspirated from each well of confluent MCEC and replaced with 1 ml of growth medium containing 5.0 pCi/ml my~-[~H]inositol. After labeling for 40-44 h, the cells were washed three times with 1 ml of physiological salt solution comprised of 1 18 mMNaC1, 4.7 mM KCl, 3.0 mM CaC12, 1.2 mM MgS04, 1.2 mM KH2P04, 0.5 mM EDTA, 10 mM glucose, and 20 mM HEPES (pH 7.4). One milliliter of 20 mM LiCl in physiological salt solution was added to each well and the mixture incubated for 30 min at 37°C before appropriate agents were added and incubated at 37°C for the indicated time. The reaction was terminated by adding cold methanol. Watersoluble [3H]IPI, ['H]IP2, and [3H]IP3were extracted with water and chloroform. The accumulated IP, , IP2, and IP3 were separated by a Bio-Rad AG 1 X 8 (100-200 mesh) column. IPI, IP2, and IP3 were eluted with 0.2, 0.4, and 1.0 M ammonium formate in 0.1 M formic acid, respectively. In some experiments, only IPI was eluted (to represent the hydrolysis of phosphatidylinositol 4,5-bisphosphate).

Measurement of PG12 production MCEC grown to confluence in 12-well clusters were washed twice with 1 ml of Hank's balanced buffer containing the following: 5.4 mMKCI, 0.44 mMKH2PO4, 137 mMNaCl, 0.34 mMNa2HP04,5.6 mhfD-glucose, 20 mMHEPES, and 0.5 mMCaC12. After incubation with 1 ml of Hank's buffer at 37OC for 5 min, agents were added and the incubation extended for another 30 min. The incubation supernatant was stored at -20°C for radioimmunoassay of 6[3H]ketoprostaglandin Flu (['H]6-keto-PGF1, ; Amersham, Arlington Heights, IL, U.S.A.), which is a stable metabolite of PG12 (Hsu et al., 1985). The antibody was a gift from Dr. P. Halushka (Department of Pharmacology, Medical University of South Carolina). Hank's balanced buffer with low calcium concentration (0.5 mM) is used routinely in eicosanoid experiments (Xu et al., 1990). Increasing CaC12concentration in Hank's balanced buffer from 0.5 to 3.0 mM does not affect agonist-mediated PG12 formation (data not shown).

RESULTS Effect of carbachol on the accumulation of 3H-IPs in MCEC Carbachol induced a concentration-dependent increase in [3H]IPI,[3H]IP2,and [3H]IP3accumulation in the presence of 20 mM LiCl during a 30-min incubation (Fig. 1). The maximal stimulationsof 13H]IPI, [3H]IP2,and [3H]IP3were 11, seven, and four times the basal values, respectively. The EC50values for carbachol were found to be 34 pM, 100 pM, and 30 pM for IPI , IP2, and IP3, respectively. In the presence of J. Neurochem.. Vol. 58, No. 5, 1992

1932

J. XU ET AL. l200~

IP, 0

S

IP,

800

$ 0

-5 -4 -3 Carbachol ( log H )

-6

601

‘‘I

-2 -10

FIG. 1. Carbachol-induced accumulation of

IPS.The data shown

are means ? SEM of three experiments performed in triplicate. The basal activity measured in the absence of carbachol was designatedas 100% and was 3,360t 330,850f 98,and 1,292k 65 dprn for IP,, IP2, and lP3, respectively.

LiCl, the carbachol-induced IPI accumulation increased in a linear manner during the first 30 rnin of incubation and continued to increase in a nonlinear fashion (Fig. 2). IP2 and IP3 increased and reached a plateau at 30 rnin and then declined at 60 min. Therefore, an incubation time of 30 rnin was chosen for the experiment. Atropine, a nonselective muscarinic (M) antagonist, and pirenzepine, a putative M I antagonist, were assessed for their ability to block the increased accumulation of IPI evoked by carbachol. Figure 3 shows that atropine at 0.3 pMand pirenzepine at 10 pM block the IPI accumulation by over 95%.Apparent Ki values calculated from their respective ICso values were 0.3 nM and 30 nM for atropine and pirenzepine, respectively. Effect of norepinephrine (NE) on the accumulation of 3H-IPs in MCEC NE increased [’H]IPI, [3H]IP2,and [3H]IP3accumulation in a dose-dependent manner in the presence of 20 mM LiCl during a 30-min incubation (Fig. 4). The EC5*values for NE calculated from dose-response curves were 0.16 pM, 0.1 pM, and 0.21 pM, respec-

1400 1

1200 1000

i

-9 -8 -7 -6 Antagonists ( log M )

-5

FIG. 3. Effect of muscarinic antagonists on carbachol-induced [3H]lP, accumulation. Cells were preincubated for 5 min with the

concentrationof atropine or pirenzepine indicatedprior to addition of 100 pJW carbachol. The data shown are means f SEM of three experiments performed in triplicate. The 100% value was 32,832 f 108 dpm.

tively. The maximal stimulations of IPI, IP2, and IP3 were 1 1, six, and four times the basal level, respectively. The time course of accumulation of IPI , IP2, and IP3 was examined using 1 pM NE in the presence of 20 mM LiCl. IPI increased linearly for 30 rnin and continued to rise in a nonlinear manner at 60 rnin (Fig. 5). IP2 and IP3 reached plateaus at 10 rnin and 5 min, respectively. Prazosin, a selective a I antagonist, and yohimbine, an a2antagonist, were assessed for their ability to block the accumulation of IPI evoked by NE. Figure 6 shows that prazosin (lo-’ M) inhibited NEinduced IPI accumulation completely (Ki, 0.1 nM), whereas yohimbine (lo-’ M) had little effect on the IP, accumulation. Effects of other agonists upon 13H]IPIaccumulation and PGIl production in MCEC Several agonists were examined at multiple concentrations for their effects upon 13H]IPIand PG12 production. Maximal responses are presented in Table 1. Carbachol and NE stimulated IPI to a maximum accumulation of approximately 1 1 times basal levels, but had only small effects upon PG12 production. In contrast, A23 187 increased PG12 secretion by about fourfold over basal levels, but did not stimulate [3H]IPI accumulation. Bradykinin and thrombin, on the other hand, increased both [3H]IP1and PGI2 production, by

13001

I

0 4 . I 0 10

.

,

20

.

I

30

.

,

40

. , . , . 50

60

, 70

Time of Incubation (min.) FIG. 2. Time course of carbachol-induced 3H-IP accumulation in MCEC. Experimental conditions were as described in the text, except that cells were incubated for different intervals in the presence of 100 pJW carbachol. Data presentedare means of triplicate experimentswhich were repeatedthree times with similar results. The basal activity was 3,5002 330 dpm.

J. Neurochem., Vol. 58, No. 5. 1992

IP.

4

1 ’

-9

-8 -7 -b -5 Norepinephrine ( log H )

-4

FIG. 4. NE-induced accumulation of IPS. The data shown are

means f SEM of three experiments performed in triplicate. The basal activity measured in the absence of NE was designated as 100% and was 3.870 f 162,750r 78,and 1.092 f 60 dpm for IP,, IP2, and lP3, respectively.

RECEPTOR-LINKED PI HYDROLYSIS AND PGI, PRODUCTION 1600

-e

-g '

1400

-

-

1000-

0

800

TABLE 1. Agonist-induced f 3H]IPl accumulation and PG12 production

-

1200-

-

A

Compound

IP2

IP3

Control Carbachol

400E==== 600

NE

200 .' "

1

0

~

1

10

-

20

1

~

30

1

40

-

1

50

-

1

60

-

1

~

70

Time of incubation (min.) FIG. 5. Time course of NE-induced 3H-IP accumulation in MCEC. Experimentalconditions were as described in the text, except that cells were incubated for different intervals with 1 pM NE. Data shown are means of triplicate experiments which were repeated three times with similar results. The basal activity was 3,650 k 350 dpm.

seven- and fourfold for [3H]IPIand four- and twofold for PG12 over control levels, respectively. Other receptor agonists, such as histamine, glutamate, serotonin, substance P, platelet-activatingfactor, and vasopressin, did not have significant effects upon either IP, or PG12 production in MCEC (Table 1). DISCUSSION

We have examined the metabolism of PI and the production of PGIz in agonist-stimulatedMCEC. Carbachol (a muscarinic agonist), NE (an adrenergic agonist), bradykinin (an inflammatory mediator), and thrombin (a serine protease) stimulated major increases in PI turnover in a dose-dependent manner. In contrast, only bradykinin and thrombin induced large increases in PG12production. To our knowledge, the results provide the first evidence for the presence in cultured murine EC of muscarinic and adrenergic receptors that are linked to PI turnover. In addition, they suggest that PI metabolism and PG12production in MCEC are regulated separately. The presence of muscarinic receptors on cerebral endothelium has been suggested previously by binding

-11

-10 -9 -8 -7 Antagonists (lag H)

1933

-6

FIG. 6. Effect of adrenergic antagonists on NE-induced [3H]IP1 accumulation. Cells were preincubated with the indicated concentrations of antagonists prazosin and yohimbine. Data shown are means f SEM of three experiments performed in triplicate. The 100% value was 30,492 f 378 dpm.

1

A23 187 Bradykinin Thrombin' Histamine Glutamate Vasopressin Platelet-activating factor Serotonin Substance P

Concentration

1 mM 10 KM

10 l*M 10 p M 2 units/ml 1 mM 1 mM 100 /.lM 10 p M

100 10 "44

['HIIP, ( W )

1,097 f 25" 1,110 f 80" 152f20 708 f 50" 401 f 61" 135 t 9 114+6 1 5 4 f 10 117f5 154f9' 114+6

6-ketoPGF,, (%) 100 152 i 24 I45 f 15 410f I l n 378 k 57" 208 k 9" 154 f 3'

105f3 100+5 100 f 5 115f 10 140 f 5'

The data presented are means f SEM of three experiments performed in triplicate. The basal activity (control) measured in the absence ofstimuli was designated as 100%.The 100%value for ['HIIP, was 3,360 f 330 dpm, and the value of basal release of PGIz was 1,233 f 2 19 pg/ml/well/30 min. Although multiple concentrations were tested for each agonist, the concentration producing the maximal stimulation was chosen for comparison of the effect of various stimuli upon ['HIIP, and PGIz in MCEC presented in this table. Effects of agonists were determined for statistical significance using analysis of variance, followed by post-hoc t test. " p < 0.01; ' p < 0.05, significantly different from control. 'The concentration of thrombin is expressed in NIH units. One NIH unit is defined as that amount of thrombin which will clot 1.0 ml of standardized fibrinogen solution in 15 f 0.5 s at 28 f I.0"C.

studies in isolated brain microvessels (Estrada and Krause, 1982; Grammas et al., 1983), by autoradiographic studies (Tsukahara et al., 1989) and by physiological studies similar to those of Kontos et al. (1990) in cat pial vessels where acetylcholine stimulates endothelial-dependent dilation. However, receptor coupling to second messenger systems has not been investigated. The present study indicates that muscarinic receptors on cerebral endothelium mediate large rises in intracellular IPS. In fact, the carbachol-induced IP accumulation in MCEC to 11 times basal levels was much greater than that found in other systems, including brain slices (Batty and Nahorski, 1985),astrocytes (Pearce et al., 1989, the 1321N astrocytoma cell line (Masters et al., 1985), AtT20/D16-16 cells (Akiyama et al., 1986), and NCB-20 cells (Chuang, 1986), but was less than that reported for cerebellar granule cells (Xu and Chuang, 1987a). The coupling of MCEC muscarinic receptors to PLC suggests that an MI or M3 pharmacologic subtype is present (Hulme et al., 1990). The carbachol-induced IP, response was inhibited completely by two muscarinic receptor antagonists, atropine and pirenzepine, and the Ki value for pirenzepine suggests that the MCEC receptors may not be M I . M3 receptors are present on EC of the rabbit ear artery and smooth muscle cells of the bovine coronary artery (Duckles, 1990). Additional studies will be necessary to classify further the muscarinic receptor subtype(s) on MCEC. J. Neurochem., Vol. 58, No. 5, 1992

1934

J. X U ET AL.

Numerous previous studies of cerebral blood vessels and cultured cerebral endothelium have documented the presence of adrenergic receptors linked to adenylate cyclase (see Jo6, 1985; Jo6 and Klatzo, 1989), but potential linkage to PLC has not been reported. Substantial increases in IPSwere elicited by NE in the present study. Like muscarhic stimulation, NE-induced IPI accumulation (1 1 times the control value) was also much greater than that found in other tissues and cells, including cortex slices (Ninomiya et al., 1989), astrocytes (Pearce et al., 1985), and cerebellar granule cells (Xu and Chuang, 1987b). NE-stimulated IP1 accumulation was inhibited completely by prazosin, an a I antagonist, suggesting that a1receptors are linked to PLC in MCEC. Unlike their substantial stimulus of PLC, carbachol and NE did not stimulate large increases in PG12 synthesis, suggesting that receptor-mediated activation of PLC by carbachol and NE is not coupled to activation of PLA2 or PLC-DG-lipase pathways. In contrast to the effects with carbachol and NE, bradykinin and thrombin stimulated both IPI accumulation and PGI2 synthesis in MCEC. This finding suggests that these agonists both stimulate PLC and lead to the release of arachidonic acid. There are several possible mechanisms. Because IP3 mobilizes calcium from the endoplasmic reticulum (Abdel-Latif, 1986; Chuang, 1989), the increase produced in cytoplasmic calcium has been proposed to stimulate PLA2 or DGlipase activity (Nishizuka, 1984; Martin and Wysomershi, 1987).The report of Lambert et al. ( 1986)supports such an arachidonic acid release in porcine aortic endothelium. Alternatively, DG activation of protein kinase C may stimulate PLA2(Hong and Deyki, 1982). Direct coupling of receptors to PLA2has also been described (Axelrod, 1990), and the results of our studies would support this mechanism in MCEC. Thus, bradykinin and thrombin probably stimulate PLC and PLA2 via separate mechanisms. Calcium ionophore A23 187 also increased PG12 synthesis in MCEC, as has been reported previously (Moore et al., 1988),but A23 187 did not stimulate IPI accumulation. Similar results have been reported in bovine aortic and porcine aortic EC and in bovine cerebral artery segments (Derian and Moskowitz, 1986; Lambert et al., 1986). Moscat et al. (1988), on the other hand, reported that A23 187 stimulated release of IPS in porcine aortic EC and suggested that calcium influx activates PLC. Our results, however, indicate that activation of PLC in MCEC is dependent upon more than merely increasing intracellular calcium. In addition to providing further details concerning agonist-mediated PI metabolism and PG12 production, our study may offer insight into receptor-mediated events underlying the differences in endothelial-dependent cerebrovascular dilation stimulated by bradykinin and acetylcholine. Kontos et al. (1990) concluded that bradykinin induces cerebral vasodilation via a cyclooxygenase-dependent pathway that yields J. Neurochem., Vol. 58, No. 5, 1992

vasoactive oxygen radicals in addition to prostaglandins. Acetylcholine, on the other hand, acts independently of cyclooxygenase probably by releasing a nitric oxide-like endothelium-derived relaxing factor similar to that generated in other vascular beds (Ignarro, 1990). In our experiments, bradykinin stimulated arachidonic acid metabolism with the potential for concurrent generation of oxygen radicals as proposed by Kontos et al. (1990). By contrast, the carbachol action is limited primarily to the PLC pathway and only minimally affects arachidonic acid metabolism in MCEC. In summary, the results show that carbachol and NE primarily stimulate PI turnover, whereas A23 187 increases solely PG12 synthesis. Bradykinin and thrombin, on the other hand, induce both PI metabolism and PG12 production. The difference in the effectiveness of these agonists to stimulate formation of IPS and PG12 suggests that the activation of PLC and PLA2in MCEC is regulated by separate and sometimes independent mechanisms. Acknowledgment: This work was supported by NIH grants NS11066, NSO1096, NS24621, NS25545, andNS28995. Dr. Steven A. Moore is the recipient of an American Heart Association Grant-in-Aid. We thank Joan Kingsley and Gay Fullerton for preparation of the manuscript.

REFERENCES Abbott N. J. and Revest P. A. (1991) Control of brain endothelial permeability. Cerebrovnsc. Brain Metab. Rev. 3, 39-72. Abdel-Latif A. ( 1986) Calcium-mobilizing receptors, polyphosphoinositides, and the generation of second messengers. J. Pharmacol. Rev. 38,227-272. Akiyama K., Vickroy T. W., Watson M., Roeske W. R., Reisine T. D., Smith T. L., and Yamamura H. I. (1986) Muscarinic cholinergic ligand binding to intact mouse pituitary tumor cells (AtT20/D 16-16) coupling with two biochemical effectors: adenylate cyclase and phosphatidylinositol turnover. J. Pharmacol. Exp. Ther. 236,653-66 1. Axelrod J. (1990) Receptor-mediated activation of phospholipase A2 and arachidonic acid release in signal transduction. Biochem. SOC.Trans. 18, 503-507. Bartha K., Muller-Peddinghaus R., and van Rooijen L. A. A. (1989) Bradykinin and thrombin effects on polyphosphoinositide hydrolysis and prostacyclin production in endothelial cells. Biochem. J. 263, 149-155. Batty 1. and Nahorski S. R. (1985) Differential effects of lithium on muscarinic receptor stimulation of inositol phosphates in rat cerebral cortex. J. Neurochem. 45, 15 14-1 52 I . Bemdge M. J. and Irvine R. F. (1984) Inositol triphosphate, a novel second messenger in cellular signal transduction. Nature 312, 3 15-32 1. Bemdge M. J., Downes C. P., and Hanley M. R. (1982) Lithium amplifies agonist dependent phosphatidylinositol responses in brain and salivary glands. Biochem. J. 206,587-595. Chuang D. M. (1986) Carbachol-induced accumulation of inositol1-phosphatein neurohybridoma NCB-20 cells. Effects of lithium and phorbol esters. Biochem. Biophys. Res. Commun. 136,622629. Chuang D. M. (1989) Neurotransmitter receptors and phosphoinositide turnover. Annu. Rev. Pharmacol. Toxicol. 29, 71-1 10. Clark M. A., Conway T. M., Bennett C. F., Gooke S. T., and Stadel J. M. (1986) Islet-activating protein inhibits leukotriene D4 and leukotriene C, but not bradykinin or calcium ionophore-induced prostacyclin synthesis in bovine endothelial cells. Proc. Natl. &ad. Sci. USA 83,7320-7324. DeBault L. E., Kahn L. E., Fromomes S. P., and Cancilla P. A.

RECEPTOR-LINKED PI HYDROLYSIS AND PGI2 PRODUCTION (1979) Cerebral microvessels and derived cells in tissue culture: isolation and preliminary characterization. In Vitro 15, 473487. DeBault L. E., Henriquez E., Hart M. N., and Cancilla P. A. (1981) Cerebral microvessels and derived cells in tissue culture. 11. Es-

tablishment, identification and preliminary characterization of an endothelial cell line. In Vitro 17, 480-494. Demolle D. and Boeynaems J. M. (1988) Role of protein kinase C in the control of vascular prostacyclin: study of phorbol esters' effect in bovine aortic endothelium and smooth muscle. Prostaglandins 35, 243-257. Derian C. K. and Moskowitz M. A. (1986) Polyphosphoinositide hydrolysis in endothelial cells and carotid artery segments. J. Biol. Chem. 261,3831-3837. DOrlkans-Juste P., de Nucci G., and Vane J. R. (1989) Kinins act on B, or B2 receptors to release conjointly endothelium-derived relaxing factor and prostacyclin from bovine aortic endothelial cells. Br. J. Pharmacol. 96, 920-926. Duckles S. P. (1990) P-Fluoro-hexahydro-sila-difenidol: affinity for vascular muscarinic receptors. Eur. J. Pharmacol. 185, 227230.

Estrada C. and Krause D. N. (1982) Muscarinic cholinergic receptor sites in cerebral blood vessels. J. Pharmacol. Exp. Ther. 221, 85-90.

Fajardo L. F. (1989) The complexity of endothelial cells. Am. J. Clin. Pathol. 92, 241-250. Forsberg E. J., Feuerstein G., Shohami E., and Pollard H. B. (1987) Adenosine triphosphate stimulates inositol phospholipid metabolism and prostacyclin formation in adrenal medullary endothelial cells by means of P2-purinergic receptors. Proc. Natl. Acad. Sci. USA 84,5630-5634. Garcia J. G., Painter R. G., Fenton J. W., English D., and Callahan K. S. (1990) Thrombin-induced prostacyclin biosynthesis in human endothelium: role of guanine nucleotide regulatory proteins in stimulus/coupling responses. J. Cell. Physiol. 142, 186-193. Gerritson M. E. and Printz M. P. (1981) Sites of prostaglandin synthesis in the bovine heart and isolated bovine coronary microvessels. Circ. Res. 49, 1152-1 163. Grammas P., Diglio C. A., Marks B. H., Giacomelli F., and Wiener J. (1983) Identification of muscarinic receptors in rat cerebral cortical microvessels. J. Neurochem. 40, 645-65 1. Hong S. L. and Deyki D. (1982) Activation of phospholipases A2 and C in pig aortic endothelial cells synthesizing prostacyclin. J. Biol. Chem. 257,7151-7154. Hsu C. Y., Halushka P. V., Hogan E. L., Lee W. A., and Perot P. L. Jr. (1985) Altered thromboxane and prostacyclin levels in experimental spinal cord injury. Neurology 35, 1003- 1009. Hulme E. C., Birdsall N. J., and Buckley N. J. (1990) Muscarinic receptor subtypes. Annu. Rev. Pharmacol. Toxicol. 30, 633-673. Ignarro L. J. (1990) Biosynthesis and metabolism of endotheliumderived nitric oxide. Annu. Rev. Pharmacol. Toxicol. 30,535560.

Jaffe E. A., Grulich J., Weksler B. B., Hampel G., and Watanabe K. ( 1987) Correlation between thrombin-induced prostacyclin production and inositol triphosphate and cytosolic free calcium I Biol. Chem. 262, levels in cultured human endothelial cells. .

1935

Martin T. W. and Wysomershi B. (1987) Ca2+-dependentand Ca2+independent pathways for release of arachidonic acid from phosphatidylinositol in endothelial cells. J. Biol. Chem. 262, 13086-1 3092.

Masters S. B., Martin M. W., Harden T. K., and Brown J. H. (1985) Pertussis toxin does not inhibit muscarinic receptor mediated phosphoinositide hydrolysis or calcium mobilization. Biochem. J. 227,933-937. Mattila P. and Renkonen R. (1989) Gamma-interferon induces prostacyclin and prostaglandin but not lipoxygenase product synthesis in rat endothelial cells. Immunol. Lett. 22, 59-64. Moncada S., Herman A. G., Higgs E. A., and Vane J. R. (1977) Differential formation of prostacyclin (PGx or PGI2) by layer of the arterial wall: an explanation for the antithrombic properties. Thromb. Res. 11, 323-344. Moore S. A,, Spector A. A., and Hart M. N. (1988) Eicosanoid metabolism in cerebromicrovascular endothelium. Am. J. Physiol. 254, C37-C44.

Moore S. A., Yoder E., and Spector A. A. (1 990) Role of the bloodbrain bamer in the formation of long-chain w-3 and w-6 fatty acids from essential fatty acid precursors. J. Neurochem. 55, 39 1-402.

Moscat J., Moreno F., Herrero C., Lopez C., and Garcia-Barreno P. (1988) Endothelial cell growth factor and ionophore A23187 stimulation of production of inositol phosphates in porcine aorta endothelial cells. Proc. Natl. Acad. Sci. USA 85, 659-663. Ninomiya H., Taniguchi T., Fujiwara M., and Kameyama M. (1989) Effects of oxygen depletion on norepinephrine- and carbacholstimulated phosphoinositide turnover in rat brain slices. J. Neurochem. 53, 183-190. Nishizuka Y. (1984) The role of protein kinase C in cell surface signal transduction and turnover promotion. Nature 308,693-698. Pearce B., Cambray-Deakin M., Marrow C., Grimble J., and Murphy S. (1985) Activation of muscarinic and a,-adrenergic receptors on astrocytes results in the accumulation of inositol phosphates. J. Neurochem. 45, 1534-1 540. Resink T. J., Grigorian G. Y., Moldabaeva A. K., Danilov S . M., and Buhler F. R. (1987) Histamine-induced phosphoinositide metabolism in cultured human umbilical vein endothelial cells. Association with thromboxane and prostacyclin release. Biochem. Biophys. Res. Commun. 144, 438-446. Sahagun G., Moore S. A., Fabry Z., Schelper R. L., and Hart M. (1989) Purification of murine endothelial cell cultures by flow cytometry using fluorescein-labeled Grifonia simplicifolia agglutinin. Am. J. Pathol. 134, 1227-1232. Salzman M. E., Wolomon J. A., and Moncada S. (1980) Prostacyclin and thromboxane A2 synthesis by rabbit pulmonary artery. J. Pharmacol. Exp. Ther. 215,240-247. Tsukada T., Tippens D., Gordon D., Ross R., and Gown A. M. ( 1987) HHF35, a muscle-action specific monoclonal antibody. 1. Immunocytochemical and biochemical characterization. Am. J. Pathol. 126, 51-60. Tsukahara T., Kassell N. F., Hongo K., Vallmer D. G., and Ogawa H. (1989) Muscarinic cholinergic receptor on the endothelium of human cerebral arteries. J. Cereb. Blood Flow Metab. 9, 748753.

Xu J. and Chuang D. M. (1987a) Muscarinic acetylcholine receptormediated phosphoinositide turnover in cultured cerebellar granule cells: desensitization by receptor agonists. . I Pharmacol. . Exp. Ther. 242, 238-244. Xu J. and Chuang D. M. (1987b) Serotonergic, adrenergic and histaminergic receptors coupled to phospholipase C in cultured cerebellar granule cells of rats. Biochem. Pharmacol. 36,2353-

8557-8565. Jo6 F. (1 985) The blood-brain barrier in vitro: ten years of research on microvessels isolated from the brain. Neurochem. Int. 7, 125. Jo6 F. and Klatzo I. (1989) Role of cerebral endothelium in brain oedema. Neural Res. 11,67-75. Kontos H. A., Wei E. P., Kukreja R. C., Ellis E. F., and Hess M. L. ( 1990) Differences in endothelium-dependent cerebral dilation by bradykinin and acetylcholine. Am. J. Physiol. 258, H1261H1266. Lambert T. L., Kent R. S., and Whorton A. R. (1986) Bradykinin

Xu J., Hsu C. Y., Liu T. H., Hogan E. L., Perot P. L. Jr., and Tai H.-H. (1990) Leukotriene B4 release and polymorphonuclear cell infiltration in spinal cord injury. J. Neurochem. 55, 907-

stimulation of inositol polyphosphate production in porcine aortic endothelial cells. J. Biol. Chem. 261, 15288-15293. Leszczynski D. and Ustinov J. (1990) Protein kinase C-regulated production of prostacyclin by rat endothelium is increased in the presence of lipoxin A4. FEBS Lett. 263, 1 17-120.

Yoshimoto S., Ishizaki Y., Kurihara H., Sasaki T., Yoshizumi M., YanagisawaM., Yazaki Y., Masaki T., Takakuya K., and Murota S. (1990) Cerebral microvessel endothelium is producing endothelin. Brain Res. 508,283-285.

2358.

912.

J. Neurochem., Vol. 58, No. 5, 1992