Brain Research, 575 (1992) 93-102 © 1992 Elsevier Science Publishers B.V. All rights reserved. 0006-8993/92/$05.00

93

BRES 17530

Distribution of cannabinoid receptors in rat brain determined with aminoalkylindoles Elizabeth M. Jansen 1, Dean A. Haycock 2, Susan J. Ward 2 and Virginia S. Seybold 1 1Department of Cell Biology and Neuroanatomy and Graduate Program in Neuroscience, University of Minnesota, Minneapolis, MN (U.S.A.) and 2Department of Neurosciences and Inflammation, Sterling Research Group, Rensselear, NY (U.S.A.) (Accepted 29 October 1991)

Key words: Cannabinoid receptor; Aminoalkylindole; Receptor autoradiography; Rat brain

Extensive mapping of the cannabinoid receptor in rat brain has been reported recently using synthetic cannabinoids. Another class of compounds, the aminoalkylindoles (AAIs), does not resemble the cannabinoids structurally. Ligand binding data on isolated membranes, however, indicate that AAIs bind to the cannabinoid receptor. The present experiments compared the binding of AAIs and synthetic cannabinoids in vitro and by receptor autoradiography. The AAIs bound to a receptor in rat cerebellum with high affinity (Kd = 15 nM), and synthetic cannabinoids were potent competitors, for AAI binding sites. In the autoradiographic studies in rat brain, an AAI and a synthetic cannabinoid were used to compete for the binding of a radiolabeled AAI to compare regionally and quantitatively the inhibition of AAI binding by the two classes of compounds. The distribution of the AAI binding was very similar to that reported for synthetic cannabinoid binding. These data add further evidence that the aminoalkylindoles bind to the cannabinoid receptor. Furthermore, the autoradiographic data for AAI binding, in addition to the autoradiographic data for the synthetic cannabinoid, provide a high degree of confidence in the localization of the cannabinoid receptor in the rat brain. INTRODUCTION In addition to their popular psychoactive effects, marijuana (Cannabis sativa) as well as A9-tetrahydrocannab inol (A9-THC), the major psychoactive component of marijuana, have been used as therapeutic agents. Cannabinoids have been used as analgesics, antiemetics for chemotherapy patients, and to relieve intraocular pressure in glaucoma patients (see refs. 7 and 15 for review). Although some of these uses are long standing, elucidation of the cellular mechanisms underlying the physiological effects of cannabinoids has been slow. The lipophilicity of cannabinoids has thwarted attempts to study these drugs in vitro and in vivo and to localize the neurons that may mediate the effects of cannabinoids on sensory, motor and cognitive processing in the nervous system. Progress in understanding a mechanism of cannabinoid action has been advanced by the characterization of a cellular cannabinoid receptor. The biochemical properties of this cannabinoid receptor have been described using A9-THC, other natural cannabinoids as well as synthetic cannabinoids developed by Pfizer 6'16. The synthetic cannabinoid analogues were created on the basis of a three-point agonist-receptor interaction

model postulated for cannabinoid association with receptors that mediate analgesia in the central nervous system 18. CP 55,940 and CP 54,939 (desacetyllevonantradol) are examples of synthetic cannabinoids. These compounds compete with A9-THC and other natural cannabinoids for ligand binding sites in brain and neuronal cell lines 6, and their order of potency in radioreceptor assays is the same as that determined for psychoactive effects, analgesia and inhibition of adenylate cyclase activity 13A6'17. Using autoradiography to visualize [3H]CP 55,940 binding to tissue sections, H e r k e n h a m and co-workers determined the distribution of specific cannabinoid receptor binding in regions of rat, monkey and human brain 13A4. Collectively, these data support the postulate of a cannabinoid receptor, the existence of which was recently confirmed by the cloning and expression of a c D N A whose translated product encodes a cannabinoid sensitive protein 2a. The amino acid sequence and postulated tertiary structure of this protein suggests that it is in the family of G-protein-coupled receptors. A cell line transfected with this c D N A clone exhibited all of the biochemical properties of the cannabinoid receptor described in neuronal cells 23. A second class of compounds, the aminoalkylindoles, is also thought to bind to the cannabinoid receptor.

Correspondence: V. Seybold, Department of Cell Biology and Neuroanatomy, University of Minnesota, 4-135 Jackson Hall, 321 Church St. S.E.,Minneapolis, MN 55455, U.S.A

94 T h e s e synthetic c o m p o u n d s are structurally dissimilar to A 9 - T H C , consisting of an indole nucleus with a substitution on the n i t r o g e n by an a m i n o a l k y l g r o u p and an aroyl g r o u p at position 32. W i n 55212-2 is a p o t e n t agonist of this g r o u p s. L i k e the c a n n a b i n o i d s , these c o m p o u n d s are analgesic 1° and h a v e m o t o r effects of sedation and ataxia 22'29, and several lines of e v i d e n c e indicate that a m i n o a l k y l i n d o l e s bind to c a n n a b i n o i d r e c e p t o r s . Firstly, the o r d e r of p o t e n c y of a m i n o a l k y l i n d o l e s and b o t h natural and synthetic c a n n a b i n o i d s in c o m p e t i n g for [3H]Win 55212-2 binding to n e u r o n a l m e m b r a n e s is c o m p a r a b l e to their o r d e r of p o t e n c y in c o m p e t i n g for [3H]CP 55,940 binding. A high c o r r e l a t i o n also exists b e t w e e n

mlA00 g body wt.), and perfused transcardially with approximately 200 ml of ice-cold 0.1 M phosphate-buffered saline with 0.16 M sucrose (pH 7.2). Biochemical studies were carried out using the cerebellum. In order to obtain homogeneous tissue sections of the cerebellum with high surface area, the cerebellum was removed, minced and frozen in a plastic tissue-embedding mold. These tissue blocks were sectioned on a cryostat (10 pm), and the sections were thaw-mounted on to acid-cleaned, gelatin-coated slides. Hereafter, these sections will be referred to as cerebellar sections. The cerebellar sections were stored desiccated at -20°C for up to 2 weeks. Autoradiographic studies were performed on i0/~m sections of rat brain and peripheral tissues. After perfusion of the animal the brain, spinal cord, eyes, lungs, kidneys, and heart were removed, trimmed to expose regions of interest and frozen on crushed dry ice. Tissue was cut on a cryostat as described above. The brain was cut in the coronal plane.

the p o t e n c y of this g r o u p in c o m p e t i n g for [3H]CP 55, 940 binding and t h e i r p o t e n c y in inhibiting c o n t r a c t i o n of the m o u s e vas d e f e r e n s 3°. Similar to [3H]CP 55,940 binding to the c a n n a b i n o i d r e c e p t o r , [3H]Win 55212-2 binding is sensitive to G T P 11 indicating that its r e c e p t o r is c o u p l e d to a G - p r o t e i n . F u r t h e r m o r e , W i n 55212-2 inhibits a d e n y l a t e cyclase activity in a G T P - d e p e n d e n t and pertussis toxin-sensitive m a n n e r 25. Finally, and p e r h a p s m o s t i m p o r t a n t in t e r m s of elucidating the physiological significance of c a n n a b i n o i d r e c e p t o r s , the first antagonist of the c a n n a b i n o i d r e c e p t o r is an a m i n o a l k y l i n d o l e analogue 5 . G i v e n the b i o c h e m i c a l d a t a indicating that [3H]Win 55212-2 binds to the c a n n a b i n o i d r e c e p t o r , the p u r p o s e of the p r e s e n t study was to d e t e r m i n e the r e g i o n a l distribution of t h e s e binding sites within the rat brain. Similarities b e t w e e n the regional distribution of [3H]Win 55212-2 binding and the distribution of [3H]CP 55,940 binding 13'14 will p r o v i d e c o m p e l l i n g d o c u m e n t a t i o n of the distribution c a n n a b i n o i d r e c e p t o r s in brain b e c a u s e of the differences in c h e m i c a l structure of the 2 c o m pounds. A s a p r e l u d e to r e c e p t o r a u t o r a d i o g r a p h i c studies, the b i o c h e m i c a l characteristics of [3H]Win 55212-2 binding w e r e d e t e r m i n e d o n s l i d e - m o u n t e d tissue sections. T h e r e g i o n a l distribution of [3H]Win 55212-2 binding sites in rat brain was t h e n d e t e r m i n e d a u t o r a d i o graphically. In o u r a u t o r a d i o g r a p h i c studies we c o m p e t e d the r a d i o l a b e l e d a m i n o a l k y l i n d o l e with a synthetic cann a b i n o i d in a d d i t i o n to a n o t h e r a m i n o a l k y l i n d o l e to exa m i n e the possibility of r e g i o n a l differences in the inhibition of [3H]Win 55212-2 binding and thus the possibility of subtypes o f c a n n a b i n o i d r e c e p t o r s . MATERIALS AND METHODS

Tissue preparation Male Sprague-Dawley rats (Harlan, Madison, WI) weighing 160260 g were used throughout these experiments. Rats were housed in cages in a light-controlled room (lights on 6.00-18.00 h) at 22°C, and food and water were available ad libitum. At the time of tissue collection, rats were anesthetized with a solution of 19.5% sodium pentobarbital and 4% chloral hydrate (0.3

3H-Win 55212-2 binding assay Win 55212-2 was custom-labeled with tritium by DuPonVNEN Research Products (specific activity 59 Ci/rnmol). The purity of the product was monitored by NEN using thin-layer chromatography on silica gel in a solvent system of hexane:ethyl acetate (1:1) and by high performance liquid chromatography on a 25 cm Zorbax ODS column with a mobile phase of acetonitrile:water (65:35). Purity of the radiolabeled ligand was 99%. The buffer for ligand binding was 20 mM HEPES with 0.5% bovine serum albumin (fatty acid-free, #A7030, Sigma Co., St. Louis), pH 7.0. The protocol used for all of the experiments, unless otherwise described, was as follows: sections were initially equilibrated in buffer for 20 min at 30°C. Sections were then incubated with radiolabeled ligand for 80 min at 30°C; nonspecific binding was defined as that which occurred when 1 pM unlabeled ligand was also present in the incubation baths. Following incubation with ligand, four 10 rain washes in buffer at room temperature (RT) were used to remove excess, unbound ligand. Tissue sections were then wiped from the slides with glass microfiber filters (Whatman, GF/A). Filters were placed into scintillation vials with 0.5 ml absolute ethanol and 5 ml Ecoscint Scintillation solution (National Diagnostics, Sommerville). Radioactivity was measured by liquid scintillation spectrometry (Beckman, counting efficiency 49%). Using this protocol, we consistently observed 85-95% specific binding with cerebellar sections. The parameters for this protocol were empirically determined as follows using cerebellar sections. For these experiments, the concentration of the radiolabeled ligand was 1 nM, and nonspecific binding was determined with 1 mM unlabeled ligand. First, pre-incubation time was varied from 5-50 min at both RT and at 30°C. Specific binding was greater when pre-incubations were carried out at 30°C rather than at RT for each time point, but there were no significant differences among pre-incubation times greater than 20 min. The incubation time with radiolabeled ligand was varied from 10-120 min at RT and at 30°C. Binding equilibrated more rapidly at 30°C than at RT; specific binding at 30°C did not change significantly after 80 min. Finally, wash times were varied from 10-60 min, at both RT and at 4°C. Specific binding was not greater at 4°C than at RT for each time point. Specific binding peaked after 40 min of wash time and was stable thereafter. The conditions for optimal specific binding to cerebellar sections closely parallel those determined by Haycock and coworkers 11 for [3H]Win 55212-2 binding to homogenates of rat brain. Saturation studies were run with the standard protocol to estimate the equilibrium dissociation constant (Ka) of [3H]Win 55212-2. Radiolabeled ligand was used at concentrations ranging from 0.25-20 nM (up to i1 different concentrations; 4 experiments). For determination of nonspecific binding, the concentration of the unlabeled ligand (Win 55212-2) was 1/~M for concentrations of radiolabeled ligand less than 2 nM, and 10/~M for concentrations of 2 nM and greater. Equilibrium dissociation constants were also calculated from data for competitive inhibition of [3H]Win 55212-2 by unlabeled Win 55212-2 as described below (4 experiments).

95 O

20

°L

t,_

15

,,,.,

Kd= 24 nM

CP55940

•

Win 55212-2

•

Win 55212-3

n

•£

80

¢.j

10

O

~

Desaeetyllevonantrodol

A

100-

O t_

~,

Win 55225

x

5

= o 0

.

0

i

.

.

100 Bound

i

200

eo

60-

N

40

"3 ~

20

•

.

!

!

300

(fmol/mg

40(1

500

0

// 0

-lb

-~

.~

AX...A _~

.~

log [drug] (M)

protein)

Fig. 1. Representative Scatchard plot of [3H]Win 55212-2 binding to tissue sections of cerebellum. Radiolabeled ligand was used at a concentration of 1 nM and competed with concentrations of unlabeled Win 55212-2 that ranged from 1 pM to 1 nM. See Materials and Methods for additional experimental details.

Fig. 2. Competition of aminoalkylindoles and synthetic cannabinoids for [3H]Win 55212-2 binding. Competition studies were performed on sections of minced cerebellum using the protocol described in Materials and Methods. [3H]Win 55212-2 was used at a concentration of 1 nM. Each point is the mean of 3 experiments.

Competition studies were performed to determine the selectivity of radioligand binding. In these studies, [3H]Win 55212-2 was used at a concentration of 1 nM, and competitors were used at concentrations ranging from 1 pM to 1 #M (9 concentrations). The competing compounds were Win 55212-2, Win 55212-3, Win 55225, CP 54,939 (desacetyllevonantradol) and CP 55,940. The K d and the inhibition constant for each competitor (Ki) of [3H]Win 55212-2 binding were calculated using nonlinear regression analysis of the data with the McPherson version of LIGAND 24. Protein content of cerebellar sections was determined by the Lowry protein assay 21, and bovine serum albumin served as the protein standard. The average protein content of cerebellar sections was 107/~g. All experiments were run in triplicate, and repeated at least 3 times using tissue from different animals, to ensure reliability,

toradiogram that corresponded to specific regions of the rat brain 28. Autoradiograms from 2 or 3 groups of near-adjacent sections prepared for total and nonspecific binding were analyzed at each level of the brain sampled from 3 different animals. Anatomical regions of the rat brain were identified based on consultation of The Rat Brain in Stereotaxic Coordinates26. Autoradiograms of the Microscale standards were used to confirm that quantitative data were collected from the linear response range of the emulsion, and to convert densities of silver grains to fmol ligand bounding tissue equivalents. Statistical comparisons were made between the inhibition of [3H]Win 55212-2 receptor binding by desacetyllevonantradol and by Win 55225 using a paired t-test across the regions quantified. RESULTS

Autoradiography Tissue sections used for the determination of total binding were incubated in 1 nM [3H]Win 55212-2; near-adjacent sections were incubated in the same concentration of radiolabeled ligand in the presence of 1/~M unlabeled ligand to estimate nonspecific binding. The two structurally different ligands that were used to compete with [3H]Win 55212-2 for receptor binding sites were an aminoalkylindole (Win 55225) and a synthetic cannabinoid (desacetyllevonantradol). Following the wash described in the above binding protocol, slides were dipped in ice-cold distilled water (to remove excess salts) and then dried under a stream of cool, dried air. Autoradiograms were prepared on emulsion-coated coverslips (NTB-3, Eastman Kodak, Rochester) according to the method of Kuhar and Unnerstal119. The slides were stored with desiccant in a light-tight box at 4°C for 35 days. Tritiated Microscale standards (Amersham, Arlington Heights) were also apposed to emulsion-coated coverslips for purposes of quantification. The emulsion was developed in Kodak D-19, fixed in Kodak Rapid Fix (Eastman Kodak), and the tissue was counterstained with thionin. Finally, the coverslip was sealed to the slide with Entellan Mounting Media (E.M. Science, Cherry Hill). Quantitative data were obtained from the autoradiograms by computer-assisted analysis of grain densities in regions of the au-

Biochemical studies O n c e r e b e U a r sections, [3H]Win 55212-2 b i n d i n g was s a t u r a b l e , and the e q u i l i b r i u m b i n d i n g d a t a w e r e best fit by a o n e - s i t e m o d e l s h o w i n g high affinity (Fig. 1). T h e e s t i m a t e o f t h e steady-state e q u i l i b r i u m dissociation was 15.0 + 4.12 n M ( m e a n + S . E . M . , n = 8); t h e Bma x was 436 + 142 fmoVmg p r o t e i n , and t h e Hill coefficient was a p p r o x i m a t e l y 1. B i n d i n g to a specific r e c e p t o r was ind i c a t e d by t h e i n e f f e c t i v e n e s s o f W i n 55212-3 (a biologically inactive s t e r e o i s o m e r of W i n 55212-225'3°) in inhibiting [3H]Win 55212-2 binding (Table I). B i n d i n g of the r a d i o l a b e l e d a m i n o a l k y l i n d o l e to a c a n n a b i n o i d r e c e p t o r was i n d i c a t e d by t h e high p o t e n c y o f the synthetic cann a b i n o i d s C P 55,940 and d e s a c e t y l l e v o n a n t r a d o l in c o m p e t i n g for b i n d i n g (Fig. 2). In fact, the o r d e r of p o t e n c y s h o w e d that the synthetic c a n n a b i n o i d s w e r e m o s t effective in c o m p e t i n g for [3H]Win 55212-2 binding: C P 55,

96

Fig. 3. [3H]Win 55212-2 binding to rat brain at the level of paraventricular nucleus of the hypothalamus. A: Darkfield photomicrograph of an autoradiogram of total binding. B: Brightfield photomicrograph of the corresponding thionin-stained tissue section. High densities of silver grains were associated with the globus pallidus (GP), entopeduncular nucleus (EP) and hippocampal formation (HIP). Moderate densities of silver grains were associated with caudate/putamen (CP), the superficial and deep layers of the cerebral cortex and fornix (f). cc, corpus callosum; cg, cingulum; f, fornix; THAL, thalamus. Bar in A = i00 ~m.

940 > desacetyllevonantradol, Win 55212-2, Win 55225 > > > Win 55212-3.

Autoradiographic studies The differential distribution of [3H]Win 55212-2 binding within the rat brain was visualized by autoradiography. In the brainstem, radioligand binding was very low, below the level of the midbrain, and very low levels of binding were also observed in the spinal cord. The areas quantified represent regions determined qualitatively under low magnification to have relatively high levels of specific binding (Table II). The highest densities of binding sites were associated with the substantia nigra pars reticulata (Fig. 5C,D), the globus pallidus (Fig. 3A,B) and the entopeduncular nucleus (Fig. 4A,B). The next highest levels were associated with regions within the stratum oriens (polymorphic layer) of the hippocampus (Fig. 4C,D). In addition to the high densities of binding

sites associated with the substantia nigra, globus pallidus and entopeduncular nucleus, moderate levels of binding sites were associated with other regions involved in motor control: caudate/putamen (Fig. 3A,B) and the molecular layer of the cerebellum (Fig. 5A,B). Moderate levels of [3H]Win 55212-2 binding also occurred in the amygdala and its associated pathway, the stria terminalis (Fig. 3A,B), as well as superficial and deep layers of the cerebral cortex (Fig. 3A,B). Moderate levels of [3H]Win 55212-2 binding were found in several myelinated fiber bundles which included the anterior commissure, corpus callosum and the fornix. Finally, the synthetic cannabinoid desacetyllevonantradol and the aminoalkylindole Win 55225 did not show regional differences in their competition for [3H]Win 55212-2 binding among the regions quantified (Table II). For each region quantified, the amount of specific binding determined with desacetyllevonantradol was signifi-

97

Fig. 4. A and B: [all]Win 55212-2 binding within the entopeduncular nucleus: (A) darkfield photomicrograph of an autoradiogram of total binding; (B) brightfield photomicrograph of the corresponding thionin-stained tissue sections. Note the high densities of silver grains associated with the entopeduncular nucleus and globus pallidus. VB, ventrobasal nuclear complex of the thalamus; ZI, zona incerta; AMG, amygdala; ot, optic tract. Bar in A = 100/zm. C and D. [all]Win 55212-2 binding within the hippocampal formation: (C) darkfield photomicrograph of an autoradiogram of total binding; (D) brightfield photomicrograph of the corresponding thionin-stained tissue sections. Note that high densities of silver grains were associated with the polymorphic layer of the hippocampus (CAI~-CA2>>CA3) and the molecular layer, but not the pyramidal cell or granule cell layer of the dentate gyrus (DG). Relatively low densities were associated with stratum lacunosum of the hippocampus. Bar in C = 100/~m.

cantly greater ( P < 0.001) than that d e t e r m i n e d with Win 55225,

DISCUSSION

A m o n g the p e r i p h e r a l tissues examined, no tissue exhibited binding above background.

The biochemical p r o p e r t i e s of [3H]Win 55212-2 binding to tissue sections of rat cerebellum are consistent with its binding to a cannabinoid receptor. Binding of the radioligand exhibited high affinity, and competition for binding was stereospecific. The equilibrium dissociation constant of 15.0 nM d e t e r m i n e d for [3H]Win 55212-2 binding on cerebellar sections is higher than the value of 1.9 n M d e t e r m i n e d for the same ligand on cerebellar h o m o g e n a t e s 11. A similar discrepancy b e t w e e n binding on tissue sections and m e m b r a n e s isolated from rat brain was n o t e d for the cannabinoid [3H]CP 55,9406'13. The reason for the discrepancy in [all]Win 55212-2 binding b e t w e e n assays is not apparent. Ligand binding conditions are not likely to contribute to the difference since buffers and incubation times were the same. The differences m a y stem from t r e a t m e n t of m e m b r a n e s prior to

TABLE I

Competition of aminoalkylindoles and cannabinoids for [sH]Win 55212-2 binding Compound

Ki~M)*

n

CP 55,940 CP 54,939 Win 55212-2 Win 55225 Win 55212-3

1.0 ___0,1 8.6 + 4.0 8.8 + 0.4 14 + 3.4 >>1000

3 3 3 3 3

* Values represent the mean + S.E.M.

98

Fig. 5. A and B: [3H]Win 55212-2 binding within the cerebellar cortex: (A) darkfield photomicrograph of an autoradiogram of total binding; (B) brightfield photomicrograph of the corresponding thionin-stained tissue sections. High densities of silver grains were associated with the molecular layer of the cerebellum (mol). gc, granule cell layer; win, white matter. Bar in A = 100/~m. C and D: [3H]Win 55212-2 binding within the midbrain: (C) darkfield photomicrograph of an autoradiogram of total binding; (D) brightfield photomicrograph of the corresponding thionin-stained tissue sections. In the midbrain, the highest densities of silver grains were associated with the substantia nigra pars reticulata (SNr) but not the pars compacta (SNc). Small arrows point to the cerebral peduncle, and the arrowheads point to the alveus of the hippocampus. DpMe, deep mesencephalic nuclei; PAG, periaqueductal gray. Not pictured: there were generally low levels of grain densities in the brain stem and spinal cord. Bar in C = 100/~m.

ligand binding (our data were derived from frozen tissue) or differences in rinses following ligand incubation. The synthetic cannabinoids were potent competitors for [3H]Win 55212-2 binding, with CP 55,940 exhibiting a K i lower than that of Win 55212-2. The greater potency of CP 55,940 in competing for [3H]Win 55212-2 binding is consistent with the lower equilibrium dissociation constant of [3H]CP 55,940 calculated from binding assays using membranes isolated from rat brain (50-200 pM) 3'6 or tissue sections of whole rat brain (2.6 nM in a buffer containing 1% BSA, approximating the assay conditions used in the present study) 14. Furthermore, the Bmax of [3H]Win 55212-2 binding on the cerebellar sections (436 fmol4"ng protein) is in agreement with the Bmax of [3H]CP 55,940 binding to homogenates of rat cerebellum (200 fmol/mg protein) 3. In total, these data provide

further support for the hypothesis that [3H]Win 55212-2 binds to a cannahinoid receptor. The overall distribution of binding sites determined autoradiographically in rat brain for [3H]Win 55212-2, an aminoalkylindole, was very similar to the distribution of binding sites for [3H]CP 55,940, a synthetic cannabinoid, described by Herkenham and co-workers 13'14. A m o n g the regions quantified in the present study, CP 54,939 (desacetyllevonantradol) competitively inhibited [3H]Win 55212-2 binding to a greater extent than did Win 55225 (Table II). These data are consistent with the higher affinity of CP 55,940 for the cannabinoid site and underscore the parallel distribution of binding among the regions quantified. The highest densities of cannabinoid receptors occurred in the globus pallidus, the entopeduncular nucleus and pars reticulata of the substantia nigra.

99 TABLE II

Density of [3H]Win 55212-2 binding sites in regions of rat brain Specific [3H]Win 55212-2 binding was determined by competition with (A) 1 #M Win 55225 or (B) 1 /zM CP 54,939 in order to assess whether differences occurred in the sites recognized by the 2 classes of compounds. [3H]Win 55212-2 was used at 1 nM. Assuming a K,, of 15 nM, approximately 7.5% of the receptors were occupied at this ligand concentration. Amounts of specific binding represent the mean + S.E.M. of measurements for each region from 3 animals.

Brain region

[3H]Win 55212-2 specific binding A fmol/mg tissue

Telencephalon Amygdala Anterior commissure Cingulum Corpus callosum Hippocampal formation hippocampus pyramidal cells (CA3) stratum lacunosum polymorphic layer/stratum oriens CA1 CA2 CA3 dentate gyrus molecular layer granule ceils Stria terminalis Fornix Basal ganglia and related areas Caudate/putamen Entopedunclar nucleus Globus pallidus Susbstantia nigra pars compacta pars reticulata Cerebral peduncle Cerebellum molecular layer granule cell layer white matter

High densities were also associated with the hippocampus and the molecular layer of the cerebellum, whereas somewhat lower densities occurred in the cerebral cortex, amygdala and m o t o r areas such as the caudate/putam e n nuclei. For each area from which quantitative data were obtained, desacetyllevonantradol competed more effectively than Win 55225 for [3H]Win 55212-2 binding, making it likely that the 2 classes of compounds bind to the same, single site. The fact that 2 structurally dissimilar compounds showed similar patterns of distribution provides a high degree of confidence in these data for the distribution of c a n n a b i n o i d receptors. F u r t h e r m o r e , the pattern of distribution of c a n n a b i n o i d binding is consistent with data from in situ hybridization studies using a nick-translated fragment of the nucleotide sequence encoding the c a n n a b i n o i d receptor 23. Within the hippo-

12.09 + 39.73 + 3.40 + 14.96 +

3.69 11.27 2.55 0.62

% SB

B fmol/mg tissue

57 76 22 59

13.69 + 45.81 + 8.50 + 18.38 +

18.34 + 3.43 22.27 _+ 5.22

61 59

22.15 _+ 2.54 30.57 + 2.76

73 81

66.41 + 2.64 62.31 + 8.89 39.02 + 3.29

85 82 78

70.81 + 2.14 68.51 + 7.81 43.22 + 3.23

90 90 87

38.64 3.46 9.92 38.83

74 20 54 82

43.90 + 9.66 + 11.12 + 40.03 +

84 57 60 84

+ 5.33 _ 0.36 + 1.12 _+ 4.39

2.14 11.03 1.29 2.05

% SB

6.09 0.36 0.86 4.98

65 88 54 73

24.95 + 2.32 160.39 + 13.34 198.55 + 20.46

72 91 95

27.65 + 2.58 168.85 + 17.48 201.25 + 20.47

80 96 97

1.17 + 0.20 165.63 + 31.64

12 81

2.86 + 0.67 196.94 + 40.80

29 97

1.76 + 1.02

21

2.24 + 1.29

26

27.06 + 1.42 5.14 + 1.27 2.14 + 0.81

72 39 24

30.83 + 2.52 5.72 + 1.03 2.17 _ 0.81

82 43 24

campal formation, high levels of m R N A for the cannabinoid receptor were found in granule cells of the dentate gyrus in addition to cells in the pyramidal and molecular layers of the hippocampus. Message for the cannabinoid receptor was also concentrated in cells within the superficial and deep layers of the cerebral cortex, as well as in the amygdala. A n additional similarity between the pattern of [3H]Win 55212-2 binding and that reported for [3H]CP 55,94014 is the occurrence of high amounts of [3H]Win 55212-2 binding associated with the anterior commissure. In contrast to data reported by H e r k e n h a m and co-workers 13'14, moderate levels of [3H]Win 55212-2 binding also occurred in conjunction with the corpus callosum. This difference may reflect differences in the regions of the corpus callosum that were sampled in the studies. We

100 believe that the localization of [3H]Win 55212-2 binding to these structures represents specific binding to receptors, and not artifactual, lipophilic association with myelin, because very low amounts of binding (not different from background) were found over the cerebral peduncle in the midbrain (Fig. 5C,D), the white matter of the cerebellum (Fig. 5A,B) and over myelinated areas within the caudal brainstem and spinal cord. In addition, quantitative data showed that the synthetic cannabinoid desacetyllevonantradol competed for [3H]Win 55212-2 binding within the anterior commissure and corpus callosum. It is possible that the binding sites associated with these fiber bundles represent cannabinoid receptors associated with axons since these tracts are comprised of efferent axons from nuclei that exhibited moderate to high levels of binding. Additional support for cannabinoid receptors on axons rather than nonspecific lipophilic associations with myelin is the specific [3H]Win 55212-2 binding that was associated with the stria terminalis, which is an unmyelinated pathway. Furthermore, data from lesion experiments indicate that [3H]CP 55,940 binding in the substantia nigra occurs on axons of striatal neurons 12. The association of [3H]Win 55212-2 binding with axons suggests that some of the effects of cannabinoids on neurons may be presynaptic. We cannot exclude the possibility, however, that [3H]Win 55212-2 binds differentially to subpopulations of glial cells that are preferentially associated with these bundles of axons. Although the overall pattern of distribution is very similar for specific [3H]Win 55212-2 and [3H]CP 55,940 binding 14, notable differences in amounts of binding are apparent between the studies. For example, if one corrects for receptor occupancy and converts our values to pmol/mg protein (assuming tissue is 10% protein), maximum specific [3H]Win 55212-2 binding was approximately 3.6 pmol/tng protein in the molecular layer of the cerebellum, and 26.4 pmol/mg protein in the globus pallidus. These values are lower and higher, respectively, than values reported by Herkenham et a1.14. We do not think that the higher values reflect visualization of more cannabinoid binding sites with [3H]Win 55212-2 since CP 54,939 competed more effectively than Win 55225 for [3H]Win 55212-2 binding among all of the regions quantified (Table II). Variations in sampling or technical aspects of autoradiography (e.g. use of liquid emulsion vs. tritium-sensitive film; grain counting vs. densitometry) could contribute to the different values reported. However, we cannot exclude the possibility that lower values in our study reflect that [3H]Win 55212-2 did not bind to all of the sites visualized with [3H]CP 55,940. Differences in the relative amounts of [3H]Win 55212-2 and [3H]CP 55,940 binding among CA1, CA2 and CA3 of the hippocampus suggest this possibility.

Of the tissues examined, [3H]Win 55212-2 binding only occurred in the brain. No binding was detected autoradiographically in eye, kidney, lung or heart. These observations are consistent with the undetectable levels of mRNA for the cannabinoid receptor reported as determined by Northern analysis of tissue mRNA for heart and kidney 23. In addition, mRNA for the cannabinoid receptor was undetectable in liver, spleen and intestine. This tissue distribution of cannabinoid receptors is consistent with observations that the multiplicity of effects of cannabinoids are largely limited to activity of the drugs within the central nervous system7. The distribution of [3H]Win 55212-2 binding sites visualized autoradiographically parallels the pharmacology of cannabinoids. In rodents, cannabinoid administration results in a characteristic decrease in spontaneous activity accompanied by hyperreflexia at low doses and catalepsy at higher doses 7. Similarly, some aminoalkylindoles cause hyperreflexia and jumping in rodents at lower doses than those required to produce ataxia zz. The synthetic cannabinoids CP 55,940 and desacetyllevonantradol decreased spontaneous activity in mice 2°. These effects on motor function may be mediated by the cannabinoid receptors localized to the basal ganglia and cerebellar cortex. Cannabinoid receptors within the basal ganglia may be linked specifically with catalepsy in that intrastriatal or intrapallidal injection of A9-THC in rats causes catalepsy in a dose-dependent manner 9'27. Cannabinoid receptors in the cerebellum, however, have been linked to the ataxia associated with administration of cannabinoids ~3. Furthermore, the very low levels of [3H]Win 55212-2 binding that were observed in the brainstem are consistent with the distribution in cannabinoid receptors determined by Herkenham and coworkers 13A4. These authors suggested that the apparent low density of cannabinoid receptors in regions of the brainstem involved in the regulation of respiration and cardiovascular function could account for the low toxicity of cannabinoids. The psychoactive effects of cannabinoids have been discerned largely from studies in humans. These effects include perceptual changes in vision and hearing as well as disruption in concentration, reasoning and short-term memory 15. Cannabinoid disruption of tasks requiring short-term memory in monkeys is similar to that seen in monkeys bearing lesions in the limbic system, suggesting that the disruption of memory by cannabinoids is mediated by the cannabinoid binding sites localized to the cerebral cortex and the hippocampus ~. In addition, several animal studies have attributed anticonvulsant activity to cannabinoids aS. These effects are consistent with cannabinoid depression of electrical activity in the hippocampus 4 which may occur via the cannabinoid recep-

101 tors localized by autoradiography (present data and those of Herkenham et a1.13'14). In contrast to the good correlation between the distribution of binding sites for cannabinoids and behavioral effects of the drugs addressed above, there is not a clear morphological correlate for the analgesic effects of cannabinoids TM. Whereas potent analgesic effects of the synthetic cannabinoid CP 55,94017'2° and the aminoalkylindoles ~° have been described, binding sites for these compounds occur in very low densities in regions of the brainstem and spinal cord associated with transmission of nociceptive information and analgesia. The efficacy of intrathecal administration of desacetyllevonantradol in analgesic assays in rats 31 suggests that the low densities of receptors in the spinal cord and brainstem are sufficient to effect analgesia. These data, however, do not exclude the possibility that cannabinoids act peripherally

on sensory neurons or at higher levels of the brain to produce analgesia. In conclusion, [3H]Win 55212-2 binding to rat brain was specific, and competitive binding studies indicated that the radioligand bound to sites recognized by synthetic cannabinoids. The pattern of distribution of [3H]Win 55212-2 binding in rat brain was similar to that determined for [3H]CP 55,940, whereas no binding above background was detected in the peripheral tissues sampied. Some of the effects of cannabinoids on the central nervous system may be mediated by the receptors visualized with [3H]Win 55212-2.

REFERENCES

305. 12 Herkenham, M., Lynn, A.B., de Costa, B. and Richfield, E.K., Neuronal localization of cannabinoid receptors in the basal ganglia, Soc. Neurosci. Abst., 15 (1989) 905. 13 Herkenham, M., Lynn, A.B., Little, M.D., Johnson, M.R., Melvin, L.S., de Costa, B.R., and Rice, K.C., Cannabinoid receptor localization in brain, Proc. Natl. Acad. Sci. U.S.A., 87 (1990) 1932-1936. 14 Herkenham, M., Lynn, A.B., Johnson, M.R., Melvin, L.S., de Costa, B.R., and Rice, K.C., Characterization and localization of cannabinoid receptors in rat brain: A quantitative in vitro autoradiographic study, J. Neurosci., 11 (1991) 563-583. 15 Hollister, L.E., Health aspects of cannabis, Pharmacol. Rev., 38 (1986) 1-20. 16 Howlett, A.C., Johnson, M.R., Melvin, L.S. and Milne, G.M., Nonclassical cannabinoid analgetics inhibit adenylate cyclase: development of a cannabinoid receptor model, Mol. Pharmacol., 33 (1988) 297-302. 17 Howlett, A.C., Bidaut-Russell, M., Devane, W.A., Melvin, L.S., Johnson, M.R. and Herkenham, M., The cannabinoid receptor: biochemical, anatomical and behavioral characterization, Trends Neurosci., 13 (1990) 420-423. 18 Johnson, M.R. and Melvin, L.S., The discovery of nonclassical cannabinoid analgetics. In R. Mechoulam (Ed.), Cannabinoids as Therapeutic Agents, CRC Press, Boca Raton, 1986, pp. 121145. 19 Kuhar, M.J. and Unnerstall, J.R., Receptor autoradiography. In H.I. Yamamura, S.J. Enna and M.J. Kuhar (eds), Methods in Neurotransminer Receptor Analysis, Raven Press, New York, 1990, pp. 177-218. 20 Little, P.J., Compton, D.R., Johnson, M.R., Melvin, L.S. and Martin, B.R., Pharmacology and stereoselectivity of structurally novel cannabinoids in mice, J. Pharmacol. Exp. Ther., 247 (1988) 1046-1051 21 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J., Protein measurement with the Folin phenol reagent, J. Biol. Chem., 193 (1951) 265-275. 22 Luttinger, D., Baizman, E.R., Miller, M.S. and Ward, S.J., The in vivo effects of aminoalkylindoles, a new class of compounds which interact with a cannabinoid binding site, Proceedings of the 1990 Committee on Problems of Drug Dependence, 105 (1991). 23 Matsuda, L.A., Lolait, S.J., Brownstein, M.J., Young, A.C. and Bonner, T.I., Structure of a cannabinoid receptor and functional expression of the cloned eDNA, Nature, 346 (1990) 561-

1 Aigner, T.G., Delta-9-tetrahydrocannabinol impairs visual recognition memory but not discrimination learning in rhesus monkeys, Psychopharmacology, 95 (1988) 507-511 2 Bell, M.R., D'Ambra, T.E., Kumar, V., Eissenstat, M.A., Herrmann Jr., J.L., Wetzel, J.R., Rosi, D., Philion, R.E., Daum, S.J., Hlasta, D.J., Kullnig, R.K., Ackerman, J.H., Haubrich, D.R., Luttinger, D.A., Baizman, E.R., Miller, M.S. and Ward, S.J., Antinociceptive aminoalkylindoles, J. Med. Chem., 34 (1991) 1099-1110. 3 Bidaut-Russell, M., Devane, W.A. and Howlett, A.C., Cannabinoid receptors and modulation of cyclic AMP accumulation in the rat brain, J. Neurochem., 55 (1990) 21-26. 4 Campbell, K.A., Foster, T.C., Hampson, R.E. and Deadwyler, S.A., Effects of d9-tetrahydrocannabinol on sensory-evoked discharges of granule cells in the dentate gyrus of behaving rats, J. Pharmacol, Exp. Ther., 239 (1986) 941-945. 5 Casiano, E, Arnold, R., Haycock, D.A., Kuster, J. and Ward, S.J., Putative aminoalkylindole (AAI) antagonists, Proceedings of the 1990 Committee on Problems of Drug Dependence, 105 (1991) 295-296. 6 Devane, W.A., Dysarz III, F.A., Johnson, M.R., Melvin, L.S. and Howlett, A.C., Determination and characterization of a cannabinoid receptor in rat brain, Mol. Pharmacol., 34 (1988) 605-613. 7 Dewey, W.L., Cannabinoidpharmacology, Pharmacol. Rev., 38 (1986) 151-178. 8 Eissenstat, M.A., Bell, M.R., D'Ambra, T.E., Estep, K.G., Haycock, D.A., Olefirowicz, E.M., and Ward, S.J., Aminoalkylindoles (AAIs): structurally novel cannabinoid-mimetics. Proceedings of the 1990 Committee on Problems of Drug Dependence, 105 (1991) 427-428. 9 Gaugh, A.L. and Olley, J.E., Catalepsy induced by intrastriatal injections of A9-THC and ll-OH-Ag-THC in the rat, Neuropharmacology, 17 (1978) 137-144. 10 Haubrich, D.R., Ward, S.J., Baizman, E., Bell, M.R., Bradford, J., Ferrari, R., Miller, M., Perrone, M., Pierson, A.K., Saelens, J.K. and Luttinger, D., Pharmacology of pravadoline: a new analgesic agent, J. Pharmacol Exp. Ther., 255 (1990) 511-522. 11 Haycock, D.A, Kuster, J.E., Stevenson, J.I, Ward, S.J. and D'Ambra, T., Characterization of aminoalkylindole binding: selective displacement by cannabinoids. Proceedings of the 1990 Committee on Problems of Drug Dependence, 105 (1991) 304-

Acknowledgements The synthetic cannabinoids CP 54,939 (CP 54,939-1; desacetyllevonantradol) and CP 55,940 were gifts of Pfizer Inc. (Groton, CT)

102 564. 24 McPherson, G.A., A practical computer-based approach to the analysis of radiolabeled ligand binding experiments, Comput. Programs Biomed.,17 (1983) 107-114. 25 Pacheco, M., Childers S.R., Arnold, R., Casiano, E and Ward, S.J., Aminoalkylindoles: actions on specific G-protein-coupled receptors, J. Pharmacol. Exp. Ther., 257 (1991) 170-183. 26 Paxinos, G. and Watson, C., The Rat Brain in Stereotaxic Coordinates, 2nd edn., Academic Press, San Diego, 1986. 27 Pertwee, R.G. and Wickens, A.P., Enhancement by chlordiazepoxide of catalepsy induced in rats by intravenous or intrapallidal injections of enantiomeric cannabinoids, Neuropharmacology, 30 (1991) 237-244. 28 Stevens, C.W., Kajander, K.C., Bennett, G.J. and Seybold, V.S., Bilateral and differential changes in spinal mu, delta and

kappa opioid binding in rats with a painful, unilateral neuropathy, Pain, 46 (1991) 315-326. 29 Ward, S.J., Miller, M., Luttinger, D., Eissenstat, M.A. and Bell, M., Inhibitory activity of analogs of Win 48098 in isolated tissue preparations in vitro is reflective of a mechanism of antinociception. Soc. Neurosci. Abst., 14 (1988) 324. 30 Ward, S.J., Baizman, E., Bell, M., Childers, S., D'Ambra, T., Eissenstat, M., Estep, K., Haycock, D., Howlett, A., Lutfinger, D., Miller, M. and Pacheco, M., Aminoalkylindoles (AAIs): a new route to the cannabinoid receptor? Proceedings of the 1990 Committee of Problems of Drug Dependence, 105 (1991) 425-426. 31 Yaksh, T.L., The antinociceptive effects of intrathecally administered levonantradol and desacetyllevonantradol in the rat. J. Clin. Pharmacol., 21 (1981) 334S-340S.