FR139317, a specific ETA-receptor antagonist, inhibits cerebral activation by intraventricular endothelin-1 in conscious rats

Nwophurmurology

Pergamon

0028-3908(94)00065-4

Vol. 33. No. IO. pp. 1155-l 166, 1994 Copyright Q 1994 Elsevier Science Ltd Printed in Greal Britain. All rights reserved 002%3908/94

$7.00 + 0.00

FR1393 17, A Specific ETA-receptor Antagonist, Inhibits Cerebral Activation by Intraventricular Endothelin- 1 in Conscious Rats P. M. GROSS,‘.‘*

D. F. WEAVER, 3.4 L. T. HO,’

J. J. PANG’

and

L. EDVINSSON’

Neurosurgical Research Unit, Departmenrs of ‘Surgery (Neurosurgery) and ‘Physiology, ‘Medicine (Neurology) and 4Chemistry, Queen’s University and Kingston General Hospital, Kingston, Ontario, Canada K7L 3N6and ‘Department of Internal Medicine, University of Lund Hospital, Lund, Sweden (Accepted 10 June 1994) comprehensive series of time-related behavioral, physiological and cerebral metabolic studies was conducted using conscious Sprague-Dawley rats to discern the anti-endothelin (ET) properties of the specific ET, receptor antagonist, FR139317. Endothelin-1 (9 pmol given by injection into one lateral ventricle, i.c.v.) produced convulsions, acute arterial hypertension, arterial hyperglycemia, and hyperventilation. Brain structures close to the i.c.v. site of injection, such as the caudate nucleus, lateral septal nucleus, corpus callosum and hippocampal CA3 medial lamellae, as well as 14 other individual structures, displayed moderate-to-intense levels of metabolic activation after endothelin. Data were assessed quantitatively by means of the autoradiographic [14C]deoxyglucose technique combined with image analysis. Neural circuits in the efferent projection paths of the stimulated forebrain structures, such as the midbrain oculomotor complex, amygdaloid nuclei, substantia nigra pars reticulata and caudal subicular subregions of the hippocampal formation, were stimulated focally by endothelin. Specific medullary nuclei and cerebellar cortical subregions displayed high rates of glucose metabolism following endothelin injection at the time of maximum behavioral and physiological stimulation. I.c.v. treatment with 2 I4 nmol FRI 39317 before endothelin significantly inhibited the effects produced by the peptide. At the highest dose of FR139317 (28 nmol), there was only mild behavioral stimulation following endothelin injection, and hypermetabolic responses in the brain were abolished except in two specific areas of the cerebellar cortex (approx 40% increases in metabolic activity in the copula pyramis and paramedian lobule). The results indicate that the cerebral stimulatory effects of i.c.v. endothelin are mediated by the A type of endothelin receptor. By itself, i.c.v. FR139317 had no effects on the parameters assessed. Further evaluation of FR139317 is warranted as a possible therapeutic agent for neuropathologies suspected of deriving from central neural or vascular stimulation by endothelin, such as aneurysmal vasospasm, ischemia, excitotoxicity, and peptide-mediated epilepsies.

Summary-A

Keywords-Behavior, central blood pressure regulation, cerebellum, seizure model, hippocampus, neuropeptide, signal transduction,

endothelin-

I, FRI 393 17, endothelin,

The neuropeptide endothelin-1 (ET; Yanagisawa et al., 1988; Kohzuki et al., 1991; Takahashi et al., 1991; Bolger et al., 1992) has stimulatory properties on cerebral function when injected in picomolar doses into a lateral ventricle (i.c.v.) of conscious rats (Gross et al., 1992a, 1993). These effects include a prolonged, repetitive, convulsive syndrome involving barrel-rotations, oculoclonus, nystagmus, forepaw dystonia, and tonic extension of the hindlimbs and tail. The physiological manifestations include hypothermia and arterial hypertension, hypocapnia, and hyperglycemia (Gross et al.,

receptor.

1992a). By applying the quantitative [14C]deoxyglucose method to assesscerebral metabolic activity, we proved that the central neural stimulation by i.c.v. endothelin is expressed widely across the neuraxis as high rates of glucose metabolism among anatomically-connected structures (Gross et al., 1992a, 1993). We speculated that the functions of intrinsic neural circuits were triggered by i.c.v. endothelin as a signal transducer, linking the polysynaptic connections between the likely origin. of stimulation in lateral periventricular forebrain structures with those stimulated nuclei and subreiions of the medulla oblongata and cerebellar cortex (Gross et al., 1993; Chew et al., 1994a). These caudal centers, which *TO whom correspondence should be addressed at: La Salle Building, 146 Stuart Street, Kingston, Ontario, Canada displayed focal intense rates of glucose metabolism, are . K7L 3N6. candidate neural substrates for the behavioral and 1155

1156

P. M. GROSS Ed 01.

physiological effects of endotheiin as a neuropeptide source for the motor convulsions in a new model of complex-partial epilepsy (Gross and Weaver, 1993). The central influence of endothelin (at lateral ventricular doses of ~9 pmol) is primarily neural rather than a vascular response to the potent vasoconstrictor properties of endothelin. Only moderate decreases in blood flow occurred in periventricular tissues which were simultaneously under strong metabolic stimulation by endothelin (Gross et al., 1992b). The resultant uncoupling of the metabolism-blood flow relationship in periventricular structures affected by endothelin presents a condition of metabolic deregulation during which the tissue and microvasculature must be exposed to abnormal levels of ions and transmitter substances probably contributing further to metabolic stimulation. As extreme as these conditions may be in uivo, they nevertheless do not produce acute cytotoxicity at experimental doses of ~9 pmol endothelin (Gross ef al., 1992a). Although insights about the corresponding cellular mechanisms of endothelin are available from literature on peripheral organs (Miller et al., 1993; Thomas et al., 1992), the profound central effects of endothelin have not been clarified. Potential actions include increased conductance of neuronal Ca*+ L-channels sensitive to the dihydropyridine, nimodipine, and interaction with glutamatergic N-methyl-D-aspartate (NMDA) receptors (probably via kinase-induced facilitation of glutamate or aspartate release) (Chuang et al., 1991; Gross et al., 1992a, 1993; Chew et al., 1994a). Although endothelin is a potent activator of voltage-gated Ca2+ channels in several organs and other brain preparations (Miller et al., 1993), it may also affect neural function as an intercellular signaling molecule that couples with effector GTP-binding proteins (G-proteins) to modify second messenger systems and intraneuronal proteins (such as enzymes and ion channels) (Rodland et al., 1991; Sakurai et al., 1992). To extend information about the possible neural mechanisms of endothelin in oiuo, we tested the hypothesis that a specific antagonist of the A class of endothelin receptors, FR139317, would inhibit the central stimulatory effects of endothelin in conscious rats. To date, there have been no reports of anti-ET, activity by FR1393 17 specifically on brain function, whereas blocking effects in constricted cerebral arteries have been attributed to anti-ET,, actions (Nirei et al., 1993; Adner et al., 1993). Antagonism of endothelin effects by FR139317 in several peripheral organ preparations and intact animals has been found (e.g. ovary cells-Aramori et al., 1993; pulmonary arteries in uitro-Cardell et al., 1993; aortic vascular smooth muscle in vitro-Sogabe et al., 1993; systemic pressor responses to endothelinSogabe ef al., 1993; McMurdo et al., 1993; and the glomeruli of failing kidneys-Benigni et al., 1993; further brief reviews in Sakurai et al., 1992; Miller et al., 1993; Thomas et al., 1992). A complete pharmacological study of FRl39317 in cultured aortic membranes,

kidney membranes, aortic strips, and intact conscious rats was repeated by Sogabe et al. (1993). In a manner similar to our previous work with the calcium L-channel blocking agent, nimodipine (Gross er al., 1992a, 1993) and the glutamate NMDA antagonist, MK-801 (Chew et al., 1994a), we examined the behavioral, physiological, and cerebral metabolic effects of i.c.v. endothelin alone and after ventricular pretreatment with FR139317. Showing that FRl39317 interfered in a dose-dependent manner with the cerebral stimulatory effects of i.c.v. endothelin, we speculate that ET, receptors mediate cerebral excitatory functions caused by this peptide. METHODS

Preparation of rat3 and test agents

Fifty-five adult male Sprague-Dawley rats (2 15-392 g) were anesthetized with 65 mg/kg i.p. sodium pentobarbital, positioned in a stereotaxic frame with the skull leveled, and implanted in the left lateral ventricle with stainless steel cannulae. Using bregma as the reference, we trephined a hole in the parietal bone to expose the dural surface at coordinates 1.8 mm L and - 1.O mm P. A 22-gauge, guide cannula was lowered 3.7 mm ventral from the dural surface into the lateral ventricle at the level of the septofimbrial and caudate nuclei (approximate to Plate 22 of Paxinos and Watson, 1986). Three jeweler’s screws were positioned in the skull to anchor a cranioplastic cap to hold the cannula through which we inserted a 28 gauge obturator with a nylon screw fitting to maintain patency (Plastic Products Co., Roanoke, Virginia). The rats were placed in individual cages with food and water ad libitum and allowed at least 24 h for recovery. I.c.v. injections were made at the rate of 1 pI/min in volumes < 3 ~1 that produced minimal ventricular dilatation and no histological artifact (Gross et al., 1992a, b). Control studies involving injection of saline alone were included. The solution containing FR1393 17 (Fujisawa Pharmaceuticals, Tsukuba, Japan, obtained by LE) was made in 10% ethanol in saline as the vehicle. Endothelin (human ET-l, Peptides International, Lexington, KY) was dissolved in saline. All injected solutions were marked with Evans blue dye to trace the distribution of the injected material within the ventricular system at post-mortem. Behavioral, cardiovascular and blood chemical studies

In preparation for the behavioral and physiological analyses, 30 rats with ventricular cannulae were fasted overnight before being anesthetized briefly with a mixture of 1.5% halothane and 2: 1 nitrous oxide and oxygen. Femoral venous and arterial catheters were then inserted (PE SO). The rats were given 500 U of heparin i.v., a loose-fitting plaster cast was placed around the hindquarters for restraint, and the anesthetic was removed. The rat was mounted on a lead block to

Functional ET, receptors in the rat brain facilitate manipulation of the catheters according to the original procedure of the [‘4C]deoxyglucose technique for measuring rates of brain glucose metabolism in conscious rats (technique described below for other animals; SokololT et al., 1977). Preliminary work comparing behavioral responsesin unrestrained rats had been conducted to establish the full behavioral response to i.c.v. endothelin (Gross et al., 1992a). Body temperature was maintained at 37°C by using a heat lanip connected to a rectal thermistor. One-two hr were allowed for recovery from the anesthetic at which time the appropriate i.c.v. injections were given according to the organization of groups as noted below. The technologist making the i.c.v. injections was unaware of the contents of the syringe. Except for those rats in the Saline group (which received one continuous injection of 3 ~1 at the rate of 1pl/min), each rat was given two successivei.c.v. injections of the antagonist + test substance (total i.c.v. volume c 3 ~1) (see Groups below). For this procedure, the catheter containing the first solution was gently removed from the protruding injection cannula and. replaced with a catheter containing the second agent. A 20 min period was interposed between injections following which was another period of 20min when arterial blood pressureand the behavioral patterns were monitored for stability. Measurements were then obtained for the following 45 min to coincide with the duration of the [“‘Cldeoxyglucose procedure in other animals. Before and after an i.c.v. injection, physiological measurements,including arterial blood pressure, blood gasesand plasma glucose concentration were measured periodically. Behavioral assessmentswere recorded on a grading scale of O-3: 0, no response; 1, mild; 2, moderate; and 3, severe (results in Table 1). The groups (n = 5 each) for the behavioral and physiological studies were: (i) saline; (ii) the vehicle solution for FR1393 17 (10% ethanol + saline) + 9 pmol endothelin; (iii) 28 nmol FRl39317 + saline; (iv) 6 nmol 14 nmol FRl39317 + 9 pmol endothelin; (v) FR139317 + 9 pmol endothelin; 28 nmol (4 FR1393 17 + 9 pmol endothelin.

1157

These dosesof FR 139317were applied first in preliminary i.c.v. studies after approximating effective blocking concentrations against endothelin-1 in the intact rat (20.1 mg/kg or about 2 50 nmol i.v.) as reported by Sogabe et al. (1993). Cerebral tnetabolic studies Twenty-five rats instrumented with ventricular cannulae were used to separately assessthe regional rates of tissue glucose metabolism in uivo as described below. [‘4C]Deoxyglucose (2-deoxy-D-[l-‘4C]glucose, sp. act. 50-60 mCi/mmol; American Radiolabeled Chemicals,St Louis, MO) was injected as a bolus i.v. (125 pCi/kg in 0.5 ml saline)after which I6 timed arterial blood samples were drawn at predetermined intervals to determine plasma concentrations of 14Cand glucose. After a 45min circulation time, the rat was killed by i.v. overdose with sodium pentobarbital and the brain rapidly extracted and frozen in isopentane at -40°C. The brain was sectioned at - 18°C in a cryostat, with sectionscut in triplicate 20 pm-thick in the region of the cannula placement (f400 pm in the rostrocaudal plane). Individual sections were collected on glass coverslips and dried on a hotplate. In sequence, the coverslips were glued to cardboard and placed with [‘4C]methylmethacrylate standards (American Radiolabeled Chemicals) in cassettesfor 14 days with Kodak OM-1 film (Rochester, New York). A fourth section was taken for staining with thionin at 60°C to aid in the histological identification of structures of interest using light microscopy, enhancement features of an image analysis system (Imaging Research Inc., St Catharines, Ontario, Canada), and rat brain atlases (Paxinos and Watson, 1986; Swanson, 1992).The imaging systemwas used to magnify the autoradiographic images, compare the image with an adjacent histological section in an adjoining channel, sample optical densitieswithin individual regions, and compute rates of tissue glucose metabolism according to the operational equation of the [‘4C]deoxyglucose method (Sokoloff et al., 1977). As.3 only the experiment number was recorded on each film, the analyst had no knowledge of the treatment being studied.

TableI. Behavioralvariablesfor differentdoses of FR139317 combined with Saline or 9 pmol endothelin FR 28 nmol+ Vehicle+ FR 6nmol+ FR 14nmol+ FR 28nmol+ ET ET ET ET Variables Saline Hindlimb/tail

extension

Piloerection Forepawdystonia Nys&nus-

Facialclonus Mastication

Agitation “Wet-dog”

shaking

0.2 * 0.2 0.2 f 0.2 0 0 0 0.8 f 0.4 0.8 If: 0.2 0 0

2.8 f 0.2 1.8 +0.2 2.8 + 0.2 2.2 + 0.2 3.0 + 0.0 2.2 + 0.6 2.6 k 0.3 2.3 + 0.4 515

2.2 * 0.3 1.8 f 0.3 2.0 f 0.6 1.5+0.5 1.8kO.6 2.0 + 0.6 3.0 * 0.0 0.6 & 0.6* 315

0.6 + 0.4* 0.6 f 0.6’ 0* 0* 0.2 f 0.2* 0.8 k 0.4* 0.4 * 0.4* 0.2 f 0.2* 0*

0.6 f 0.3’ 0* 0* 0.2 f 0.2’ 0* 0.4 * 0.4* 0.6 f 0.4* 0’ 0*

Barrel-rolling (No. responding) Data represent scores assigned on the followingscale:0, no response; 1, mild; 2, moderate; 3, severe. n = 5 pergroup.Data wereobtained20 min followinglateralventricularinjection(volumes93 /II). *Differentfrom response to “Vehicle/ET”injection,P < 0.05,i.e. blockingeffectof FR139317.

1158

P. M. GROSS er al. Table

2. Key physiological

variables

Assessments

Saline

Mean arterial pressure (mmW Arterial PCO, (mmW Arterial glucose concentration W@-Jl)

113+5 37*

for different

FR 28 nmol + Saline 119*4 I

146 k 6

36k

I

142 + I3

doses of FRl39317

with

Vehicle +ET

FR 6 nmol +ET

143 * 13**

133 + 7**

23 * I** 174rf:

II**

26 f 2++ I61 +8

9 pmol FR

endothelin

I4 nmol +ET

ll4*6* 32 f 2+ 133*5*

FR 28 nmol +ET 123+4* 32 f 2’ 144&6*

Data

are means + SE for 5 rats per group. Measurements were taken 20 min after i.c.v. injection. <3 ~1 i.c.v. injection volume. *Significantly different from response to Vehicle/ET injection, P < 0.05 (i.e. blocking effect of FRl39317). **Significantly different from response to Saline alone or FR1393 I7 + Saline injection, P < 0.05.

Because FR1393 17 proved to have dose-related inhibitory effects (above 6 nmol) on the i.c.v. injection of 9 pmol endothelin (Results in Table l), we conducted a 5-group series of [‘4C]deoxyglucose studies to compare the cerebral metabolic effects of: (1) saline, (2) FR1393 17 (28 nmol) + saline, (3) endothelin alone, (4) FR1393 17 (14 nmol) + endothelin, and (5) FR1393 17 (28 nmol) + endothelin. Metabolic studies were not completed with the 6 nmol dose of FR139317 + endothelin because the behavioral and physiological results indicated little difference from the endothelin effect alone (Tables 1 and 2). Selection of brain structures for image analysis

The autoradiographic [‘4C]deoxyglucose method provides the powerful advantage of structure-by-structure investigation of brain metabolic activity in uivo (Sokoloff et al., 1977). Aimed at defining the ability of a putative antagonist to interrupt endothelin effects in the brain, we limited our choices for analyses in the present work to 3 sets of structures (total of 18) selected on the following reasoning. “Plate” numbers cited in this section refer to those in the atlas by Paxinos and Watson (1986), whereas mention of previously demonstrated hypermetabolic responses refers to our previous publications on endothelin as listed in the Reference section. All of the structures were studied on the side of the brain ipsilateral to the injection of the test agents. (1) Periuentricular tissues(i.e. structures forming the contiguous walls of the injected lateral ventricle) were included because they are either near the site of injection or border the ipsilateral lateral and third cerebral ventricles and cerebral aqueduct. Because these structures likely make contact with the highest concentrations of endothelin in the CSF, they also may be the neural sources of functional activation by endothelin in the brain (Gross and Weaver, 1993; Gross et al., 1993). Furthermore, their structural differences permit comparisons of(i) gray vs white matter, (ii) segregated neural circuit responsiveness to i.c.v. endothelin and FR1393 17, and (iii) structures harboring different densities of endothelin receptors (Kohzuki et al., 1991; reviewed previously in Gross et al., 1992a). Accordingly, we chose two gray matter nuclei with different efferent neural radiations (the caudate nucleus

and lateral septal nucleus) and two white matter tracts (the corpus callosum and the fimbria hippocampus) bordering the injected lateral ventricle (Plates 20-24). We also analyzed a medial subregion of the hippocampus-the alveus, oriens and pyramidale strata of Ammon’s horn; see Plates 38-39 and Fig. 36 of Swanson, 1992. Central gray matter surrounding the cerebral aqueduct at the level of the midbrain raphe was also analyzed (Plates 44-46). (2) Limbic and midbrain structures, surmised to be within the efferent neural projection paths of the forebrain structures activated by i.c.v. endothelin, included the hypothalamic ventromedial nucleus and the posterior and medial subnuclei of the amygdaloid complex (read at the level of Plates 30-32) pars reticulata of the substantia nigra, the oculomotor Darkschewitsch nucleus, and the medial terminal nucleus of the accessory optic tract (which harbors high-density binding sites for endothelin, Kohzuki et al., 1991) (see Plates 38-40). Two specific hippocampal caudal subregions previously demonstrated to react to i.c.v. endothelin with vigorous metabolic stimulation-the infrapyramidal ventral blade of the dentate gyrus (Plates 36-39) and the parasubiculum (Plates 49-51), were included. (3) Medulla oblongata and cerebellum, within which are subregions and individual nuclei forming neural circuits putatively involved in the behavioral and cardiovascular responses to central endothelin (Gross et al., 1993), were analyzed topographically according to the mossy and climbing fiber innervation of the cerebellar cortex (Ito, 1982; Chew et al., 1994a). The medullary inputs to the cerebellum-the cuneate nucleus (Plates 73-74), the medial vestibular nucleus (Plates 61-65), and the inferior olivary complex (Plates 69-72), were analyzed individually. Two subregions of the cerebellar cortex, the copula pyramis (Plates 69-72) and the cortex of the paramedian lobule (Plates 69-73), were included in the analyses. In summary, by selection of the above 18 individual structures, we tested the ability of FR139317 to inhibit endothelin-induced metabolic stimulation near the site of lateral ventricular injection and also over integrated neural circuits activated by endothelin to engage the function of structures as caudal as the cerebellar paramedian cortex.

Functional ET, receptors in the rat brain Statistical analyses

Metabolic and physiological data across the different rat groups were tested by analysis of variance in a completely randomized design with a modified r-statistic procedure applied to evaluate intergroup differences for significant F ratios at P < 0.05 (Bruning and Kintz, 1977). The behavioral data, which represent our subjective gradings of behavioral phenomena, were tested for significance with the nonparametric Kruskal-Wallis test.

RESULTS

‘Behavior

The data in Table 1 indicate the status of 10 behavioral variables for the rats 20 min following i.c.v. injection of saline or the vehicle for the FRI 39317 (90% saline, 10% ethanol), endothelin, and endothelin in the presence of 3 doses of FRI 39317. This 20-min time delay corresponded to the timing for initiation of the [i4C]deoxyglucose procedure performed in separate animals (begun 20 min after i.c.v. injection of the test solution when behavioral and physiological responses were stable; see results below and Gross et al., 1992a). We do not show behavioral data for the Saline group as these rats displayed no changes from baseline behavior after injection. Those animals injected i.c.v. with the ethanolbased vehicle for FR139317 + saline exhibited only a few signs of behavioral activation, such as increased gradings for sniffing and agitation, although most behavioral variables were unaffected by this procedure (Table 1). Behavioral signs following i.c.v. injection of endothelin (9 pmol) confirmed the convulsive effects of the neuropeptide upon central administration (Ouchi et al., 1989; Lecci et al., 1990; Gross et al., 1993). The gradings produced by endothelin were all in the moderate-tosevere range and included 5/5 rats that attempted to barrel-roll with their hindlimbs and tails extended (hindquarters immobilized, see Methods). Dystonia of the forepaws, nystagmus, facial clonic and masticatory activity, and “wet-dog” shaking were prevalent. In the presence of 6 nmol FR139317, endothelin produced barrel-rolling in 3/5 rats but generally caused the same degree of behavioral stimulation among the other variables assessed (Table 1). Only the amount of “wetdog” shaking was reduced in the endothelin-injected rats treated with 6nmol FR139317. At the higher doses, 14 and 28 nmol, FR 1393 17 ‘prevented endothelin-induced barrel-rolling and effectively inhibited the other types of behavioral stimulation produced by endothelin (Table 1). Physiology

All the rats were in normal blood gas and acid-base balance before experimentation (range of arterial PO1

1159

of 86-98 mmHg and arterial pH of 7.43-7.51); body temperature was 37.5 + 0.3”C (mean If: SE, n = 30). In previous evaluations of the stimulatory state produced by i.c.v. endothelin (e.g. Gross and Weaver, 1993), we focused our analysis of physiological responses to the three indicators proven to be the most reliable reflections of central activation-mean arterial pressure, arterial PCO,, and the arterial plasma glucose concentration. Saline and Vehicle/Saline injections did not alter these three key indicators of physiological status, whereas endothelin, as previously demonstrated by others and us (Ouchi ef al., 1989; Matsumura et al., 1991; Gross et al., 1992a), increased arterial pressure by 27%, caused hyperventilation and reduced arterial PCOz by 38%, and increased the concentration of glucose in arterial plasma by 16% (Table 2). Pretreatment with 6 nmol FR139317 did not significantly block the physiological responses to endothelin (i.e. the effects were similar to those in the endothelininjected group, Table 2), whereas 14 and 28 nmol doses of FRl39317 were effective in inhibiting the physiological activation of all three variables by 9 pmol endothelin (Table 2). Cerebral glucose metabolism

Metabolic image analyses were performed selectively on 6 periventricular, 7 limbic or midbrain, and 5 hindbrain structures in each rat. There were no significant differences in the rates of glucose utilization for any structure between the groups injected i.c.v. with either Saline or the high dose of FR139317 (28 nmol) + Saline (Table 3). Values shown in Table 3 for these two groups are similar to those for control animals in endothelin studies reported previously (Gross et al., 1992a, 1993) and for normal untreated conscious rats (Sokoloff er al., 1977). As demonstrated previously for these and numerous ’ other brain regions (Gross et al., 1992a, 1993), i.c.v. injection of 9 pmol endothelin produced intense, widespread metabolic activation among all 18 structures included in this analysis. The range of increases in metabolic activity (compared to the corresponding values in the Saline group) was from 21% (medial vestibular subnucleus) to 350% (fimbria hippocampus; white matter, near the site of endothelin injection). Three periventircular structures close to the endothelin injection site-the caudate and lateral septal nuclei (both gray structures) and the corpus callosum (white)-were stimulated strongly (+ 103 to 240%) [Fig. l(A), top row]. The hippocampal CA3 region showed highly focal, intense hypermetabolism in its medial subregions [HC in Fig. l(A), bottom row; Table 31. Further caudally in the ventricular system at the level of the cerebral aqueduct, central gray matter was significantly activated (+60%; Table 3). All the remaining 12 structures in the analysis were hypermetabolic following injection of endothelin, yet only one of these--the vestibular nucleus-has contact

1160

P. M. GROSS CI al.

with cerebrospinal fluid which could transport endothelin caudally into the fourth ventricle. All these 12 structures, however, are anatomically connected with striatal, septal or hippocampal neural efferent systems (summarized in Swanson et al. 1987; Gross et al., 1993) that may have accounted for their metabolic stimulation. Such functional activation via neural projections, therefore, was considered as part of the hypothesis testing of whether FR139317 could be antagonistic at these efferent sites of stimulation. Focal hypermetabolic responsesto endothelin occurred in the ventral blade of the dentate gyrus [ +90%, bilateral stimulation, Fig. 2(A), top row], the parasubiculum [ +74%, bilateral stimulation, Fig. 2(A), middle row], and the copula pyramis [+68%, Fig. 2(A), bottom row] and paramedian lobular cortex of the cerebellum [+ 116%, Fig. 2(A), bottom row]. Because of the behavioral and physiological studies demonstrated little antagonistic activity of 6 nmol FR139317 against endothelin (Table I). we evaluated metabolic responsesin the 18 structures to endothelin after the higher antagonist dosesof either 14 or 28 nmol FR139317. Both doses were effective at blocking the endothelin stimulatory effect on glucose metabolism in the 6 lateral “periventricular” structures (Table 3). Notably, the caudate and lateral septal nuclei, which are gray matter structures close to the endothelin injection site and are highly sensitive to endothelin, did not show significant metabolic increases following 14 nmol FR139317, although the caudate nucleus appeared to be slightly stimulated in 3 rats [e.g. Fig. l(B)]. Except for the parasubiculum (35% increase, Fig. 2 middle row), all

regions in the “limbic/midbrain” category were prevented from endothelin-induced stimulation by both dosesof FRl39317 (Table 3). Although not statistically significant as a group, responses to endothelin after I4 nmol FR139317 in the dendate gyrus ventral blade were present in 3/5 rats, but only on the brain side ipsilateral to stimulation [Fig. 2(B, C)] as opposed to the bilateral responsesseen in rats injected with endothelin alone [Fig. 2(A)]. At 14 nmol FR 139317, all the “medullary/cerebellar” structures tended to be stimulated by endothelin, although only the cuneate nucleus (+ 29%) and the two regions of cerebellar cortex achieved statistical significance (copula pyramis, + 60%; paramedian cortex, +57%; Table 3). Three out of five rats treated with 14 nmol FR139317 displayed focal increases in metabolic activity after endothelin in the hippocampal CA3 region [Fig. l(B, bottom)] and the inferior olive [Fig. 2(B, bottom); Table 31, but these values were not significant when grouped with nonstimulated animals. After 28 nmol FRI 39317, all structures were blocked from responding metabolically to endothelin [Figs l(C) and 2(C); Table 31 except the two cerebellar cortical regions, copula pyramis (+ 35%) and paramedian lobule (+43%, Table 3). There was a tendency for elevated metabolic activity in the caudate and lateral septal nuclei and medial CA3 region of the hippocampus, but these values did not achieve significance (Table 3). The trend of these data, however, may hold importance for interpretation of persisting behavioral responses to endothelin after the 28 nmol treatment with FR139317 (discussion below).

Table 3. Rates of glucose metabolism in individual brain structures following lateral ventricular injection of endothelin (ET) alone or after pretreatment with FRl39317 FR (I4 nmol) FR (28 nmol) FR (28 nmol) Brain structures Saline + saline ET (9 pmol) % + A +ET % +A +ET %+A Periventricular

Tissues

Corpus callosum (white) Caudate nucleus, medial margin Lateral septal nucleus CA3 (alveus/oriens/pyramidale)’ Fimbria hippocampus (white) Periaqueductal gray* LimbiclMidbrain

Oblongara

0.3 1 + 0.03 0.77 kO.05 0.42 f 0.06 0.54 f 0.05 0.36 +0.05 0.50 kO.03

0.87 kO.07; 1.42&0.13* 1.46 & O.lO* 1.20 + 0.08* 1.26+0.10* 0.83 + 0.08*

200 103 240 Ill 350 60

0.48 f 0.05 0.58 f 0.06 0.50*0.03 0.80+0.04 0.50+0.05 0.62& 0.02 0.74kO.06

0.41 + 0.04 0.59 f 0.04 0.44 f 0.02 0.67 f 0.03 0.5 I f 0.03 0.56f0.04 0.65 f 0.03

0.85 kO.08’ 1.07 & 0.13* 0.88 +0.15* 0.97 +0.06* 0.74+0.06* I.18 +0.20* 1.29+0.09*

i: 76 21 48 90 74

0.37 f 0.05 0.84 f 0.12. 0.43 * 0.05 0.75 + 0.12. 0.38 f 0.05 0.56 f 0.07

ns ns ns ns ns ns

0.34& 0.78 f 0.53 + 0.67 + 0.37 f 0.59 f

0.02 0.03 0.08 0.02 0.05 0.02

ns ns ns ns ns ns

0.43 + 0.42 f 0.51 f 0.79 f 0.61 + 0.70 f 1.00 +

ns ns ns ns ns

0.51 +0.03 0.55 + 0.03 0.53 + 0.02 0.87 + 0.03 0.62 + 0.02 0.69 & 0.01 0.68 f 0.05

ns ns ns ns ns ns ns

Srrucrures

Hypothalamic ventromedial nucleus Amygdala (posteromedial nucleus) Substantia nigra, pars reticulata Darkschewitsch nucleus Medial terminal nucleus, AOT3 Dentate gyrus, ventral blade Parasubiculum Medulla

0.29 + 0.03 0.70 f 0.07 0.43 * 0.03 0.57 + 0.04 0.28 * 0.04 0.52 + 0.04

0.06 0.05 0.07 0.09 0.12 0.15. 0.16’

;s

and Cerebellum

0.71 f 0.08* 29 Cuneate nucleus 0.55 + 0.03 0.60 kO.03 0.73 +0.06* 33 0.61 f 0.02 ns 0.87 f O.lO* 30 0.79fO.ll* ns 0.79 * 0.02 ns Inferior olivary complex 0.67 + 0.06 0.59 +0.05 1.30 f 0.09* 21 l.l6&0.12 I.13 +0.05 Medial vestibular subnucleus 1.07 L-o.07 0.98 +0.06 Copula pyramis, cerebellum 0.57 rto.05 0.55 kO.03 0.96* 0.09, 68 0.91 f 0.15* :i 0.77fo.l1* 3”s Cerebellar paramedian cortex 0.51 + 0.06 0.48 kO.03 1.10 f 0.05* II6 0.80~0.16* 57 0.73 f 0.10* 43 Values are the mean + SE in units of pmol g-’ min-’ for 5 rats per group. ‘Medial layers of Ammon’s horn at the level of Plates 38-39 in Paxinos and Watson (1986). sMedia1 central gray, Plates 44-46, Paxinos and Watson (1986). )AOT, accessory optic tract; Plates 38-39, Paxinos and Watson (1986). *Significantly different from Saline group, P < 0.05; ns, not significant. l 3/5 rats displayed focal stimulation.

Functional ET,, receptors in the rat brain DISCUSSION

New findings from this study were: (1) the etfeetive inhibition by FRI 39317 (doses > 14 nmol) of the endothelin hypermetabolic response indicates that endothelin-I mediates its neurostimulatory properties via ET,, receptors; (3) the efferent neural circuit activation. thought to originate from periventricular fore- and midbrain sites after cndothelin injection, persists to reflect metabolic stimulation in focal regions of the cerebellar cortex. a distant downstream target, even after diminution of the endothelin-induced convulsions; and (3) given the anti-endothelin efficacy of FRl39317 in the present neural studies, and the absence of physiological, or cerebral metabolic ‘behavioral, effects of 38 nrnol FRl39317 itself following central administration, this drug has promise for treating endothelin-related neuropathologies without deleterious side-effects.

1161

PIIN~t~l~IL.oI~~iL’L~I clu~r~~cteris~ics qf FR 139317 Only two previous studies have evaluated the antiendothelin blocking efficacy of FRl39317 in brain preparations. but both were demonstrations of antagonistic activity in cerebral arteries constricted by endothelin (Adner et [I/., 1993; Nirei et al., 1993). Until the present work describing regional brain metabolic activity, there have been no published characterizations of FR 1393 I7 effects on endothelin-induced neural activation. Because FRl39317 is a hydrophilic tetrapeptide of 605 MW (Sogabe CI crl., 1993; Miller rf al.. 1993). its physicochemical characteristics are not favorable for ready passage from the blood across cerebral capillary walls. Therefore, intraventricular administration, a route bypassing the exclusive endothelial features that restrict solute permeability at the blood-brain interface, was exploited in the present study.

Fig. I. Coronal [‘JC]deoxyglucose autoradiographs at two brain levels that likely are the sources of neural circuit stimulation by i.c.v. endothelin. Sections are from rats treated via the lateral ventricle on the left side with 9 pmol endothelin alone (A), 14 nmol FRl39317 + endothelin (B), and 28 nmol FRl39317 + endothelin (C) (see Table 3 for quantitative metabolic data). The regional image analysis was confirmed by reference to the corresponding histological sections. Top roar is at the level of the caudate nucleus (CN), lateral septal nucleus (LS). and corpus callosum (CC, white matter). See Plate I7 of Paxinos and Watson (1986). Cannula tracts are above CC in each of the top row figures. Bar beneath A = I mm. In the CN, LS and CC, endothelin alone caused increased rates of glucose metabolism (A) (compare to contralateral side of the brain; O.D. in the image is proportional to the rate of glucose metabolism). Within groups of rats, this hypermetabolic effect in the three structures was significantly inhibited by I4 or 28 nmol FRl39317 (B, C) (Table 3), although subjectively there was a mild degree of stimulation by endothelin in the CN and LS in 3/5 rats treated with 14nmol FRl39317 (e.g. CN and LS in B: Table 3). Borronr I’OIVdisplays responses more caudal in the rostrocaudal plane approximately to Plate 40 of Paxinos and Watson (1986). Focal stimulation by endothelin occurred in the periaqueductal central gray (PC). Darkschewitsch nucleus (DK, oculomotor complex), hippocampal lamellae alveus. oriens, and pyramidale (HC), and substantia nigra (SN) pars reticulata (A). Bilateral responses to endothelin were evident in the HC, DK and SN (A; Gross rt ol., l992a, 1993). Both doses inhibited the endothelin effect in the PG. DK and SN (B and C), but hippocampal stimulation was still present after I4 nmol FRl39317 (B), although the area and magnitude of HC stimulation were reduced (B vs A). Only a very focal amount of hippocampal stimulation by endothelin resulted after pretreatment with 28 nmol FRl39317 (C). We speculate that neural efferent projections from structures contacted and stimulated by i.c.v. endothelin via its A-type receptors, such as the CN. LS, PG or HC, formulate the activated network of regions involved in endothelin-induced convulsions (Gross er al., l992a, 1993; Gross and Weaver, 1993). FRl39317 effectively inhibited these putative sources of stimulation (B,C). NP31IO-C

1162

P. M.

GROSS

Before discussing our results further. however, we shall consider briefly some background from other investigators using peripheral organ preparations in which a high degree of anti-endothelin efficacy by FR1393 I7 was proven. We shall also comment on the possible anti-endothelin mechanism of FR139317 in these tissues. As confirmed by our studies showing no behavioral, physiological, or cerebral metabolic effects of FR I393 I7 by itself, this blocking agent is without endothelin agonist activity (Sogabe et al., 1993), but displays com-

1’1 td.

petitive blockade specifically of ET, receptors in such preparations as porcine and rabbit aortic membranes, transfected ovary cells, porcine and guinea pig pulmonary arteries. and guinea pig cerebral arteries (Sogabe et nl., 1993; Cardell et ul., 1993; Sudjarwo ~JI ul., 1993; Adner r~ (I/., 1993). Such binding work has not yet been reported for brain tissue treated with FR 1393 17. Among the vascular preparations used, contractile responses to endothelin were examined in the presence of a calcium channel antagonist or a kinase C inhibitor, neither of

Fig. 2. Coronal [‘4C]deoxyglucose autoradiographs at three brain levels that likely are in the neural projection paths of forebrain structures stimulated by i.c.v. endothelin. Sections are from rats treated via the lateral ventricle on the left side with 9 pmol endothelin alone (A), I4 nmol FRl39317 + endothelin (B), and 28 nmol FRl39317 + endothelin (C) (see Table 3 for the quantitative metabolic data). The regional image analysis was confirmed by reference to the corresponding histological sections. Top ~OIVis at the level of the medial terminal (MT) nucleus of the accessory optic tract, a region dense with endothelin binding sites (Kohzuki PI al., 1991) (see Paxinos and Watson, 1986, Plate 38). Other abbreviations: DC, dentate gyrus infrapyramidal ventral blade and PC, periaqueductal central gray (a structure probably stimulated by endothelin in the cerebral aqueduct). Bilateral responses that occurred in the MT and DC after endothelin alone (A) were reduced in a dose-dependent way to a unilateral (B) or no effect (C) after FRl39317. Bar under A = I mm. Middle IYIN’,corresponding to Plate 49 of Paxinos and Watson (1986), shows the hypermetabolic effect of endothelin bilaterally in the parasubiculum (PS) and PC at this level (A). At both doses, FRl39317 inhibited the endothelin-induced metabolic stimulation (B, C; Table 3). Bottom ~011’(Plate 69 of Paxinos and Watson, 1986) demonstrates several subregions of focal hypermetabolism in the inferior olivary complex (IO) and the cerebellar cortex bilaterally, such as the paramedian lobule (PL) and copula pyramis (CP). The low dose of FRl39317 (I4 nmol) limited the extent and magnitude of olivary and cerebellar cortical stimulation (B), whereas the higher dose (28 nmol) abolished the metabolic stimulation in the inferior olive and reduced it further in the cerebellar cortex (C) (Table 3). We speculate that intrinsic endothelinergic or glutamatergic neural circuits are involved in the metabolic stimulation of these focal sites by i.c.v. endothelin (Gross et al., 1992a, 1993; Gross and Weaver, 1993; Chew et al., 1994a). By specific blockade of ET,, receptors, FR139317 effectively inhibited these responses.

Functional ET,, receptors in the rat brain which fully curbed the endothelin-induced vasoconstriction, whereas FR139317 was effective as an inhibitor of endothelin vasoconstriction (reviewed by Sogabe et al., 1993). These findings indicate that vascular ET, receptors mediate responses that are not entirely dependent on transmembrane calcium conductance or intracellular messenger systems involving kinases. FR 1393 17 may act mainly on ET, receptors coupled to a G-protein that, in cells not treated with FR 1393 17, activates phospholipase C, inositol triphosphate, and mobilization of calcium from sequestered stores to elevate the cytosolic calcium concentration (Sakurai et al., 1992; Berridge, 1993; Thomas et al., 1992). Although the above mechanism applies to vascular smooth muscle, a different sequence of events may occur following endothelin binding to its A receptor in neuronal membranes. For example, we speculated previously that extracellular calcium was important in mediating the hypermetabolic effects of i.c.v. endothelin which were blocked by nimodipine, a dihydropyridine L-channel inhibitor (Gross ef al., 1992a, 1993). Hypermetabolic responses to i.c.v. endothelin in periventricular brain tissues were reduced in a manner similar to the effect elicited by FR139317 in the present study. Whether produced from inositol triphosphate-sensitive stores intracellularly or from extracellular sources across membrane channels, rising cytosolic calcium levels in brain cells likely evoke kinase C activity (Alberts ef al., 1989), nitric oxide formation (Reiser, 1990), or synaptic release of neurotransmitters, such as acetylcholine or glutamate (Alberts et al., 1989; Chuang et al., 1991). Transmitter molecules like glutamate and nitric oxide could be involved in further signaling of nearby cells (Garthwaite, 1991) and in stimulation of energy metabolism in a sphere of tissue surrounding the original site of stimulation. Neural efferents could then engage the metabolism of downstream targets. We demonstrated that (1) nitric oxide (using the donors, sodium nitroprusside and S-nitroso-N-acetylpenicillamine) has stimulatory effects on white and gray matter glucose metabolism (Gross et al., 1994) and (2) the glutamate NMDA receptor antagonist. MK-801, inhibits endothelin-induced convulsions and cerebral hypermetabolism in a way similar to that of FR139317 (Chew et al.‘, 1994a). The results of this latter study suggest that enddthelin causes release of glutamate which may then become the neuroactive species evoking the cerebral stimulatory effects described here and previously (Gross et al., 1992a, 1993). Endothelin, nitric oxide and glutamate are regarded as colocalized, comodulators of cerebellar cortical functions (Ross et al., 1990). It is interesting that lateral ventricular injection of glutamate analogs produces convulsions (Chiamulera et al., 1992) similar to those observed with endothelin-1 (although endothelin-1 is considerably more potent for ‘evoking behavioral responses; Chew er al., 1994b). A valuable study extending from the present experiments

1163

would be to test the specificity of endothelin’s actions at ET, receptors by injecting a glutamate analog following treatment with FR 1393 17. Expanding our understanding of the molecular characterization of endothelin’s neuronal effects, and identifying the specific mechanism of FR139317 inhibition of neuronal ET, receptors, are important areas of future research. ET, receptors mediateneurostimula!ory processesevoked in the brain by endothelin-1

At the two FR139317 doses 2 14 nmol, the central stimulatory effects of endothelin were clearly inhibited. Three areas of evidence from FRl39317-treated animals injected i.c.v. with endothelin produce this conclusion: (1) convulsive responses spawned by i.c.v. endothelin were mitigated or blocked completely (Table 1); (2) the pressor, hyperventilatory, and hyperglycemic responses to i.c.v. endothelin were effectively inhibited (Table 2); and (3) the extent and magnitude of metabolic stimulation in the brain regions examined were abolished in most structures and diminished in others (Table 3; Figs 1 and 2). These findings permit the novel conclusion that ET, receptors sensitive to FR139317 are responsible for most of the stimulatory effect produced in the brain by endothelin-1 after i.c.v. injection. We note, however, that ET, receptors have been localized in the cerebellum (Hagiwara et al., 1993) and may have roles in the neurostimulatory actions of endothelin that are not yet identified. Future experiments assessing brain stimulation following i.c.v. injection of an ET, agonist, alone and after FR1393 17, would be helpful. Such studies might identify both the neurostimulatory properties of the ET, receptor and, in the presence of a blocking dose of FR139317 at ET, receptors, the proportion of the cerebral activation by endothelin-1 that occurs at ET, receptors. It is appropriate to consider mechanisms through which binding of ET, receptors may lead to stimulation of the brain. First, the pattern and magnitude of hyper-. metabolism in periventricular structures after i.c.v. injection of endothelin indicate that such stimulation involves more than just binding to ET, receptors along the ventricular wall of different structures. If only binding were needed to activate brain function, metabolic stimulation would be reasonably uniform and strong among regions with a high density of endothelin binding sites (septum and hippocampus) and weaker among those with a low of moderate density of receptors (corpus callosum, caudate nucleus, periaqueductal gray; see Kohzuki et al., 1991 for reference). Such was not the case because all periventricular structures, especially those closest to the site of injection, had robust hypermetabolic responses to endothelin (Table 3). It is likely; then, that after endothelin binds to its A receptors in brain structures, it initiates a cascade of excitatory events possibly involving other stimulatory transmitter species. One potentially important neural mechanism of endothelin-1

1164

P. M. GROSS et al.

may be to inhibit membrane K+ conductance, as it does in coronary smooth muscle (Miyoshi et al., 1992). Chemicals that block neuronal K+ channels, such as the peptide from mamba snake venom, a-dendrotoxin, are powerful seizure-producing agents (Bagetta et al., 1992). Secondly, endothelin and FR139317 could diffuse from the cerebrospinal fluid to bind at sites both in periventricular and deeper cerebral nuclei connected by innervation with neural circuits responsible for the behavioral and physiological responses to i.c.v. endothelin. Based on the specificity of the metabolic responses (Figs 1 and 2), we discussed this premise previously as a network response of numerous anatomically-linked structures engaged in the regulation of visuovestibular, oculomotor and proprioceptive functions (Gross et al., 1993; Gross and Weaver, 1993). It is even possible that the activity of intrinsic “endothelinergic” nerves is radiated among a circuit of structures using the peptide as a common transmitter (Gross et al., 1993). All the limbic/midbrain, medullary and cerebellar structures analyzed in the present work (Table 3) contain variable amounts of endothelin receptors (Kohzuki ef al., 1991) and are within neural systems emanating from forebrain regions metabolically stimulated by i.c.v. endothelin, viz. the caudate and lateral septal nuclei and medial subregions of the hippocampus. Each of these structures is implicated in the initiation and downstream regulation of endothelin-induced convulsions and physiological activation over intact neural networks (Gross er al., 1993; Gross and Weaver, 1993 and references therein). Key components of this putative circuit activation may be the substructures of the hippocampal formation, such as the medial CA3 lamellae, dentate gyrus ventral blade, and parasubiculum (all of which were selectively stimulated, Figs 1 and 2). Through the complexity of the intrahippocampal association pathways (Swanson et al., 1987), these structures are parts of the circuitry likely involved in the modification of various sensory and motor effects that could be linked to central endothelin. Clinical

blood flow-metabolism relationship in the brain, i.e. the tissue metabolic stimulation was well out of proportion to the level of perfusion which was reduced moderately (Gross et al., 1992b). Such conditions locally deregulate the balance of vaso- and neuroactive chemical species and may resemble those occurring in the core and penumbra of cerebral ischemia. Endogenous endothelin may create these opposing effects that would promote a pathological outcome. Under such pathophysiological conditions in which endothelin is active, centrally-administered FR 1393 17 might be therapeutically beneficial by antagonizing ET, receptors responsible both for the neural stimulation and for the vasoconstriction. The present work has emphasized the neural, rather than the vascular, responses to endothelin and the effective blockade by FR139317 of these neurostimulatory effects. In addition to the hypothetical role for endothelin in a model of complex-partial epilepsy that we are testing in the rat (Gross and Weaver, 1993), endothelin could influence other transmitter systems affecting mood and may be involved in the origin of cerebral neoplasms (Miller et al., 1993). The present results have demonstrated that centrally administered FR139317 (doses of 6-28 nmol i.c.v.) does not have important systemic physiological, behavioral or cerebral metabolic effects of its own, yet its higher doses were effective in reducing different forms of cerebral activation produced by i.c.v. endothelin. These findings establish the utility of intraventricular administration in the rat as a model test system and constitute an experimental basis for further development of receptorspecific endothelin antagonists having potential for the treatment of endothelin-related neuropathologies. thank Dan Wainman, Ben Chew, Carol MacNeil and Jennifer Hamilton for assistanceand Ethicon Ltd. of Peterborough,Ontario for donating supplies. The studiesweresupportedby the Medical ResearchCouncil of Canada. DFW is a Career Scientist with the Ontario Ministry of Health. Acknowledgements-We

significance

Although mainly studied as a peripheral vascular autacoid (Yanagisawa et al., 1988; Sakurai et al., 1992), endothelin is implicated as a neurovascular factor in several neuropathologies, such as aneurysmal vasospasm, stroke, head injury, post-surgical trauma, brain tumors, depression and epilepsy (Kraus et al., 1991; Miller et al., 1993; Gross and Weaver, 1993; Barone et al., 1994). Specifically, potent binding by endothelin at its A type of receptors in cerebral arterial smooth muscle (Adner et al., 1993) confirms that endothelin likely has critical roles in the regulation of vascular resistance in health and in pathological spasmogenic states such as after subarachnoid hemorrhage (Nirei et al., 1993). We suggested previously that the neural hypermetabolic effect of endothelin, combined simultaneously with localized vasoconstriction produced by endothelin in the same structure, would uncouple the

REFERENCES

Adner M., You J. and EdvinssonL. (1993)Characteristicsof endothelin-A receptorsin the cerebral circulation. Neuroreport 4: 441-443. Alberts B., Bray D., LewisJ., Raff M., RobertsK. and Watson J. D. (1989)The nervoussystem.In: Molecular Biology ofthe Cell, 2nd ed., Chap. 19,pp. 1059-l 136.Garland, New York. Aramori I., Nirei H., Shoubo M., SogabeK., Nakamura K., Kojo H.;Notsu Y., Ono T. andNakanishiS. (1993)Subtype selectivity of a novel endothelinantagonist,FRl39317, for the two endothelinreceptorsin transfectedChinesehamster ovary cells.Molec. Pharmacol. 43: 127-131. Bagetta G., Nistico G. and Dolly J.’ 0. (1992)Production of seizuresand brain damagein rats by a-dendrotoxin, a selectiveK+ channel blocker. Neurosci. Letf. 139: 34-40. Barone F. C., Globus M. Y.-T., Price W. J., White R. F., Storer B. L., FeuersteinG. Z., Busto R. and Ohlstein E. H.

Functional ET, receptors in the rat brain (1994) Endothelin levels increasein rat focal and global ischemia.J. Cereb. Blood FIOW Me&b. 14: 337-342. Benigni A., Zola C., Corna D., Drisio S., Longaretti L., Bertani T. and Remuzzi G. (1993) A specificendothelin subtype A receptor antagonist protects against injury in renal diseaseprogression.Kidney Int. 44: 440-444. Berridge, M. J. (1993) Inositol triphosphate and calcium signaling.Nature 361: 3 15-325. BolgerG. T., Berry R. and JaramilloJ. (1992)Regionaland subcellulardistribution of [‘*‘I]endothelinbindingsitesin rat brain. Brain Res. Bull. 28: X9-797. BruningJ. L. and Kintz B. L. (I 977) Computational Handbook q/ Statistics, pp. I 13-l 16. Scott, Foresman,Glenview, IL. Cardell L. O., Uddman R. and EdvinssonL. (1993)A novel ‘ETA-receptor antagonist, FRl39317, inhibits endothelininducedcontractionsof guinea-pigpulmonary arteries,but not trachea. Br. J. Pharmacol. 108: 448-452. Chew B. H., Weaver D. F., BalabanC. D. and GrossP. M. (I 994a) NMDA-mediated metabolicactivation of the cerebellarcortex in behavingrats by the neuropeptideendothelin-I, Brain Res. 647: 345-352. ChewB. H., Weaver D. F. and Gross P. M. (1994b)Doserelated potent cerebral stimulation by the neuropeptide endo-thelin-I after intraventricular administration in consciousrats, Pharmac. Biochem. Behav. In press. Chiamulera C., Costa S., Valerio E. and Reggiani A. (1992) Domoic acid toxicity in rats and mice after intracerebroventricularadministration: comparisonwith excitatory amino acid agonists. Pharmacol. Tox. 70: 115-120. ChuangD.-M., Lin W.-W. and Lee C. Y. (1991)Endothelininduced activation of phosphoinositideturnover, calcium mobilization, and transmitter releasein cultured neurons and neurally relatedcell types.J. Cardiouas. Pharmacol. 17 (Suppl. 7) S85-S88. Garthwaite J. (1991) Glutamate, nitric oxide and cell-cell signalling in the nervous system. Trends Neurosci. 14: 60-67.

GrossP. M., Wainman D. S., EspinosaF. J., Nag S. and WeaverD. F. (1992a)Cerebralhypermetabolismproduced by intraventricular endothelin-I in rats: inhibition by nimodipine,Neuropeptides 21: 21 l-223. GrossP. M., Zochodne D. W., Wainman D. S., Ho L. T., EspinosaF. J. and Weaver D. F. (1992b)Intraventricular endothelin-I uncouplesthe blood flow: metabolismrelationship in periventricular structuresof the rat brain: involvement of L-type calcium channels. Neuropeptides 22: 155-165. Gross P. M. and Weaver D. F. (1993)A new experimental mod$lof epilepsybasedon the intraventricular injectionof endothelin. J. Cardiovasc. Pharmacol. 22 (SuppI. 8) S282-S287.

1165

Hagiwara H., NagasawaT., Yarn&mot0 T., Lodhi K. M., Ito T., TakemuraN. and HiroseS. (1993)Immunochemical characterizationand localization of endothelinETa receptor. Am. J. Physiol. 264: R777-R783. Ito M. (I 982) The Cerebellum and Neural Control. New York, Raven. Kohzuki M., Chai S.Y., PaxinosG., KaravasA., CasleyD. J., JohnstonC. I. and MendelsohnF. A. 0. (1991)Localization and characterization of endothelin receptor binding sites in the rat brain visualized by in vitro autoradiography. Neuroscience 42: 245-260.

Kraus G. E., Bucholz R. D., Yoon K.-W., Knuepfer M. M. and SmithJr., K. R. (1991)Cerebrospinalfluid endothelin-l and endothelin-3 levels in normal and neurosurgical patients:a clinical study and literature review. Surg. Neurol. 35: 20-29.

Lecci A., Maggi C.-A., Rover0 P., Giachetti A. and Meli A. (1990)Effect of endothelin-I, endothelin-3,and C-terminal hexapeptide,endothelin(16-21) on motor activity in rats. Neuropeptides

16: 21-24.

MatsumuraK., Abe I. and TsuchihashiT. (1991)Centraleffect of endothelin on neurohormonalresponsesin conscious rabbits. Hypertension 17: 1192-1196. McMurdo L., Lidbury P. S., ThiemermannC. and Vane J. R. (1993) Mediation of endothelin-inducedplatelet aggregationvia the ET, receptor. Br. J. Pharmacol. 109: 530-534.

Miller R. C., Pelton J. T. and HugginsJ. P. (1993)Endothelins-from receptorsto medicine.Trends Pharmacol. Sci. 14: 54-60.

Miyoshi Y., Nakaya Y., WakatsukiT., Nakaya S., Fujino K., Saito K. and Inoue I. (1992) Endothelin blocks ATP-sensitiveK+ channelsand depolarizessmoothmuscle cellsof porcine coronary artery. Circ. Res. 70: 612-616. Nirei H., HamadaK., ShouboM., SogabeK., Notsu Y. and Ono T. (1993) An endothelin ET,, receptor antagonist, FR139317,ameliorates cerebralvasospasm in dogs.Life Sci. 52: 1869-l874. Ouchi Y., Kim S., SouzaA. C., Iijima S., Hattori A., Orimo H., Yoshizumi M., Kurihara H. and Yazaki Y. (1989) Central effect of endothelinon blood pressurein conscious rats. Am. J. Physiol. 256: Hl747-Hl751. PaxinosG. and WatsonC. (1986) The Rat Brain in Stereotaxic Coordinates, 2nd edn. AcademicPress,Sydney. ReiserG. (1990)Endothelinand a Ca*+ionophoreraisecyclic GMP levelsin a neuronalcell line via formation of nitric oxide. Br. J. Pharmacol. 101: 722-726. Rodland K: D., Muldoon L. L. and Magun B. E. (1991) Regulation of intracellular Ca*+ and gene expressionb: endothelin-I. J. Cardiouasc. Pharmacol. 17 (Suppl. 7 S89-s95. RossC. A., Bredt D. and Snyder S. H. (1990) Messenger molecules in the cerebellum. Trends Neurosci. 13: 2 16-222. SakuraiT., YanagisawaM. and Masaki T. (1992)Molecular characterizationof endothelinreceptors.Trends Pharmacol.

GrossP. M., BeningerR. J., Shaver S. W., Wainman D. S., EspinosaF. J. and Weavei D. F. (1993) Metabolic and neuroanatomicalcorrelatesof barrel-rollingand oculoclonic convulsions induced by intraventricular endothelin-1: a Sci. 13: 103-108. novel peptidergic signalingmechanismin visuovestibular and oculomotor regulation? Exp. Brain Res. 95: SogabeK., Nirei H., ShouboM., Nomoto A., Ao S., Notsu Y. and Ono T. ( 1993)Pharmacologicalprofile of FR13931.7, 397-408. a novel, potent endothelin ET, receptor antagonist. GrossP. M., WeaverD. F., Nag S., Ho L. T., PangJ. J. and J. Pharmacol. Exp. Ther. 264: 1040-1046. ‘EspinosaF. J. (1994)Neurotoxicity in consciousratsfollowSokoloff L., Reivich M., Kennedy C., Des RosiersM. H., ing intraventricular SNAP, a nitric oxide donor. NeuroPatlak C. S., Pettigrew K. D., Sakurada 0. and pharmacology 33: 9 15-927.

P. M.

1166

GROSS

Shinohara M. (1977) The [‘4C]deoxyglucose method for the measurement of local cerebral glucose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat. J. Neurochem. 28: 897-916. Sudjarwo S. A., Hori M., Takai M.. Urade Y., Okada T. and Karaki H. (1993) A novel subtype of endothelin B receptor mediating contraction in swine pulmonary vein. Life Sci. 53: 431-437. Swanson L. W., Kiihler C. and Bjiirklund A. (1987) The limbic region I: The septohippocampal system. In: Handbook qf Chemical Neuroanatomy, Vol. CNS, Part I: Hypothalamus, Retina (Bjiirklund A., Hiikfelt

Amsterdam,

5, Integrated Hippocatnpus,

Systems qf the Amygdala,

T. and Swanson L. W., Eds), Elsevier, pp. 125-277.

et nl. Swanson L. W. (I 992) Brain Maps: Structure of the Rot Brain. Amsterdam, Elsevier. Takahashi K., Ghatei M. A., Jones P. M., Murphy J. K., Lam H.-C., O’Haloran D. J. and Bloom S. R. (1991) Endothelin in human brain and pituitary gland: presence of immunoreactive endothelin, endothelin messenger ribonucleic acid, and endothelin receptors. J. Clin. Endocrinol. Metab. 72: 693-699.

Thomas C. P., Simonson M. S. and Dunn M. J. (1992) Endothelin: receptors and transmembrane signals. Neltvs Physiol. Sri. 7: 207-21 I . Yanagisawa M., Kurihara H.. Kimura S., Tomobe Y., Kobayashi M., Mitsui Y., Yazaki Y., Goto K. and Masaki T. (1988) A novel potent vasoconstrictor peptide produced by vascular endothelial cells. Name 332: 41 I-415.