Mu and Delta Opiate Receptors Coupled Negatively to Adenylate Cyclase on Embryonic Neurons from the Mouse Striatum in Primary Cultures

The Journal of Neuroscience, September 1988, 8(9): 3376-3382

Hervi! Chneiweiss, Jacques Glowinski, and JoQl Premont Chaire de Neuropharmacologie, INSERM U 114, Collkge de France, 75231 Paris Cedex 05, France

Primary cultures of pure populations of neuronal or glial cells cyclase. Using this approach, adenylate cyclases sensitive to from the striatum, the cerebral cortex, and the mesenceph- biogenic amines and peptide receptors coupled positively or alon of the mouse embryo were used to look for the presence negatively to the enzyme [vasoactive intestinal peptide (VIP) of opiate receptors coupled to adenylate cyclase. Leu-en- and somatostatin receptors, respectively] were characterized on kephalin inhibited CAMP production in membranes of em- pure neuronal populations (Premont et al., 1983; Chneiweisset bryonic striatal neurons but not in those of other cell types al., 1984, 1985a,b, 1986, 1987). In addition, besidesP-adrener- examined. Mu and delta opiate receptors seemed to be cou- gic receptors, VIP, and somatostatin receptors coupled to ade- pled negatively to adenylate cyclase in embryonic striatal nylate cyclasecould be demonstratedon pure astrocytes (Chnei- neurons. It was found that DTLET (a selective delta agonist), Weisset al., 1984, 1985a,b, 1986). Differencesin the localization as well as DAGO (a selective mu agonist), inhibited CAMP of the receptorswere seenfrom one structure to another either production on these cells. DTLET but not, however, DAGO on neuronsor on glial cells. Finally, information concerningthe produced a similar effect on homogenates from the adult rat eventual colocalization of distinct types of receptorson the same striatum and on membranes from the neuroblastoma x glioma cellular population was obtained by determining whether the hybrid cell line NG 108-l 5, two preparations known to pos- effects of the corresponding agonists on CAMP production in sess only delta receptors negatively coupled to adenylate membrane preparations were additive. For example, all striatal cyclase. The selective kappa agonist U 50.488 was ineffec- neurons possessingdopamine Dl receptorswere also found to tive on all types of membrane preparations used. possessVIP receptorscoupled to adenylate cyclase(Chneiweiss The inhibitory effects of both DTLET and DAGO on basal et al., 1985a). The present investigation was undertaken to de- adenylate cyclase activity in striatal neurons were reversed termine if opioid receptors coupled to adenylate cyclasewere by naloxone with a similar efficacy. Two other selective mu present on either neuronsor astrocytes from the cerebral cortex, agonists, trimu 5 and morphiceptin, inhibited CAMP produc- the striatum, and the mesencephalonof mouse embryos in pri- tion in membranes of striatal neurons as well. The nonad- mary culture. ditivity of the inhibitory effects of DTLET and DAGO on basal Binding studieson tissuesof the adult brain of various species or forskolin-induced activation of adenylate cyclase sug- have distinguished 3 main classesof opioids binding sites: mu, gested that mu and delta receptors were colocalized on a delta, and kappa binding sites, which can be recognizedby spe- similar subpopulation of striatal cells in primary culture. These cific ligands or endogenousopioid peptides (Lord et al., 1977; cells possess dopaminergic receptors of the Dl subtype as Patersonet al., 1983; Simon, 1987). Morphine and the synthetic well since the amplitude of the inhibitory effects of DTLET peptide [D-Ala’, (Me)Phe4, Gly (OH)5] enkephalin (DAGO) are and DAGO on CAMP production was increased in the pres- selective ligands for mu binding sites (Handa et al., 1981). (D- ence of dopamine. The inhibitory effects of DAGO or DTLET Thrz, Leu5,Thr6) enkephalin (DTLET) isone ofthe most specific and of somatostatin on adenylate cyclase activity were ad- ligands for the delta binding sites (Zajac and Roques, 1985). ditive, suggesting that somatostatin receptors coupled neg- Kappa binding sitesare recognized in particular by the alkaloid atively to the enzyme may not be localized on striatal neu- U 50.488 (Piercey et al., 1982). Among pro-enkephalin-derived rons possessing mu and delta receptors. peptides (Met5)-enkephalin is delta selective, the pro-opiome- lanocortin derived beta-endorphin is equipotent at mu and delta In previous studies,we have usedprimary cultures of cells from sites,while the opioids derived from the prodynorphin molecule the cerebral cortex, the striatum, and the mesencephalonof are largely kappa-opiate receptor selective (Martin, 1984; Leslie, mouse embryos in order to characterize and define the cellular 1987). It is now well establishedthat the delta receptor iscoupled localization or colocalization of receptors coupled to adenylate negatively to adenylate cyclase,as has been shown particularly on the neuroblastoma x glioma hybrid cell line NG 108-15 Received Oct. 21, 1987; revised Dec. 28, 1987; accepted Dec. 30, 1987. (Sharma et al., 1975; Law et al., 1981, 1985). Mu and delta We would like to thank particularly Dr. B. Roques and J. M. Zajac for their receptors have also been shown in both central and peripheral generous gifts of mu, delta, and kappa receptor agonists and for stimulating and fruitful discussions. This research was supported by INSERM, DRET (Contract neurons to increase potassium conductance and therefore to 85.078) and Rhbne Poulenc Sante. reduce calcium entry indirectly, whereaskappa receptors seem Correspondence should be addressed to Dr. H. Chneiweiss, Chaire de Neuro- pharmacologic, INSERM U 114, College de France, 11, place Marcelin Berthelot, to reduce calcium currents directly (North, 1986). Using selec- 7523 1 Paris Cedex 05, France. tive ligands we show that, among all cell types examined, the Copyright 0 1988 Society for Neuroscience 0270-6474/88/093376-07$02.00/O embryonic striatal neurons in primary culture are the only ones The Journal of Neuroscience, September 1988, 8(9) 3377

Table 1. Inhibition of adenylate cyclase activity by leu-enkephalin STRIATUM on adult rat striatum, mouse striatal neurons and astrocytes in primary culture A 1 Adenylate cyclase activity . (pmol CAMP/ 10 min/mg protein) +Leu-ENK Tissue Basal (lo-‘M) Adult striatum 975 k 56 137 + 62* Neurons Cerebral cortex 19 -t 3 20 + 3 Striatum 69 + 12 48 k 7* Mesencephalon 24 + 4 22 2 5 q Astrocytes Astrocytes o Neurones Cerebral cortex 49 f I 53 k 5 A Adult tissu Striatum 9t1 9&l 70 Mesencephalon 101 + 9 97 2 I ot, ’ 1 1 I J Effect of leu-enkephalin (leu-ENK) on adenylate cyclase activity was determined 0 -9 -8 -7 -6 on adult rat striatal membranes and homogenates from neurons or glial cells grown in primary culture originating from 3 brain areas of the mouse embryo. Inhibition of CAMP production due to leu-enkephalin was found only in the adult striatum Leu-Enkephalin (log[M]) and in striatal neurons (* p < 0.005 as estimated by Student’s t test; n = 4). Each value is the mean f SEM of data obtained in 4 experiments, each individual Figure 1. Inhibitory effects of leu-enkephalin on basal adenylate cy- experiment being performed with triplicate determinations. clase activity in embryonic striatal neurons from the mouse and on adult striatal tissues of the rat. Homogenates of striatal neurons and of astrocytes in primary culture and crude membranes from adult striatal tissue were prepared as described in Materials and Methods. Leu-enkephalin inhibited basal adenylate cyclase activity in striatal neurons possessing delta receptors coupled negatively to adenylate cy- (0) and adult striatal tissue (A) but not in striatal astrocytes (0). The clase.Surprisingly, mu agonistswere also found to inhibit mark- inhibitory effects were reversed by naloxone 10m4 M (0). Naloxone edly adenylate cyclase activity on striatal neurons. Additivity alone induced a 12% decrease in adenylate cyclase activity when used at 10m4M. Since SEMs never exceeded 5% of the mean value, they are experiments suggestedthat the delta and mu receptors were not indicated. Basal adenylate cyclase activity (in pmol/lO min/mg colocalized on striatal neurons. They also revealed that these protein): striatal neurons, 66 ? 2; striatal astrocytes, 9.1 f 0.8; adult opiate receptorscould be colocalized alsowith D 1 dopaminergic striatum, 747 * 53. (DA) but not somatostatin receptorson a population of striatal neurons. Neuroblastoma-glioma hybrid, NG 108-l 5. The hybrid cells were Materials and Methods cultured in Dulbecco’s MEM containing HAT (0.1 mM hypoxanthine, 0.4 PM aminopterin, 16 PM thymidine), sodium bicarbonate (3 mM), Neuronal primary cultures. Cortical and striatal cells (20 x lo6 each) and 10% fetal calf serum. The cells were grown for 1 week after seeding from 16-d-old Swiss mouse embryos (Iffa Credo) and mesencephalic before use in the experiments. cells from 14-d-old mouse embryos were dissociated and plated in the Homogenate preparation. Following 5-6 d for neuronal cells, 1 week absence of fetal calf serum in culture dishes (100 mm diameter, Falcon) for NG 108- 15 cells, and 4 weeks for glial cells, the culture medium previously coated with polyomithine (1.5 pg/ml; MW 40,000; Sigma) was removed and the cells were washed 4 times at room temperature as already described (Di Porzio et al., 1980). The culture medium con- with an isotonic solution (50 mM Tris maleate, 10% sucrose, pH 7.2; sisted of a mixture (1: 1) of modified Eagle’s medium (MEM) and F-12 wt/vol). A further 1.5 ml of the same buffer was then added, and the nutrient (Gibco), supplemented with glucose (33 mM), glutamine (2 mM), cells (20 x 106) were detached using a rubber policeman. The resulting sodium bicarbonate (3 mM), and Hepes (5 mM; Sigma). Fetal calf serum suspension was then homogenized at 4°C by 5 strokes of a Teflon pestle was replaced by a mixture of hormones, proteins, and salts composed in a Potter Elvehjem glass homogenizer. Homogenates were directly of insulin (25 pg/ml; Sigma), transferrin (100 &ml; Sigma), progester- used for adenylate cyclase assay since results were similar to those one (20 nM; Merck), putrescine (60 PM; Sigma), and selenium salt NaSeO, obtained with crude washed membranes (centrifugation at 100,000 x (30 nM; Merck). Under these conditions, according to different mor- g, 15 min). phological and histochemical criteria, the number of non-neuronal cells Membranes from the striata of Sprague-Dawley adult male rats and is reduced to less than 5% of the cell population after 7 d in vitro Swiss adult female mouse were also used. For this purpose, striata were (Prochiantz et al., 1982). dissected at 4°C according to the technique of Glowinski and Iversen Glialprimary cultures. The protocol chosen for neurons was also used (1966), and striatal homogenates were prepared as just described for to start the glial cultures. However, the medium included 10% Nu- cells in culture. Particulate fractions (100,000 X g, 15 min) were re- serum (Collaborative Research) and was changed every 3 d for 4 weeks suspended and washed twice before the adenylate cyclase assay. until glial elements had formed a confluent monolayer, devoid of neu- Adenylate cyclase assay. The reaction was initiated by adding 20 ~1 ronal cells. As previously described, the absence of surviving neurons of the homogenates (5-l 5 pg protein) to 40 pl of the incubation medium was checked by indirect immunofluorescence using an antibody against [Tris maleate, 50 mM; MgSO,, 0.5 mM; ATP, 0.05 mM; phosphocreatine, the 70K neurofilament protein (Chneiweiss et al., 1985b). More than 20 mM; creatine kinase, 0.2 mg/ml; BSA fraction V, fatty acid free, 2.5 95% of the cells could be stained significantly by the immunofluores- mg/ml; 3-isobutyl-1-methylxanthine (IBMX) 0.1 mM; GTP, 10 PM; cence technique using a rabbit antibody directed against glial fibrillary papaverine, 0.1 mM; (&P)-ATP, 1 &i; )H-CAMP, 0.002 &i, pH 7.21. acidic protein (GFAP), indicating that cells were mainly mature astro- The membranes were incubated at 30°C for 10 min, and the reaction cytes. A monoclonal antibody from the rat against galactocerebroside ended by the addition of 100 ~1 of a medium containing sodium dodecyl and a rabbit antiserum against fibronectin were used to identify the sulfate, 1% (wt/vol); ATP, 5 mM; CAMP, 2 mM; Tris HCl, 50 mM, pH presence of oligodendrocytes and fibroblasts, respectively. No staining 7.2. The 3ZP-cAMP formed was isolated according to Salomon et al. could be observed, indicating a lack of these cell types in the cultures. (1974). All determinations were performed in triplicate, whose internal The cultures also were checked to confirm the absence of macrophages variability was always less than 5% of the obtained values (relative using a nonspecific fluorescent immunoglobulin. error). 3378 Chneiweiss et al. * Opioid Receptors on Mouse Striatal Neurons

STRIATAL NEURONES

-s .Z Figure 2. Inhibitory effects of mu .E: ti (DAGO) and delta (DTLET) opiate ag- a onists on basal adenylate cyclase activity in embryonic striatal neurons. Left, Dose-response curves obtained with DAGO, DTLET, and U 50.488 which are selective agonists of mu b), delta (6) and kappa (K) opiate receptors, respectively. Right, Hofstee plots of the data obtained with DAGO and DTLET from which their apparent affinities (K,,,,,) of 4.8 x 10m8and 8.7 x 10 m8M . U 50.488 ( k) for DAGO and DTLET were evaluat- ed. Data presented illustrate^_ . results. . of one experiment out ot 5 providing sim- 00 -9 -8 -7 -6 -5 ilar results. Each point is the mean of triplicate determinations. AGONIST (log [M]l

Chemicals. Mu and delta receptor agonists were generous gifts from (Fig. 1). Naloxone ( 1O-4M) antagonized the leu-enkephalin in- Dr. B. Roques. Other chemicals were obtained from the following com- hibitory effect on both preparations (Fig. 1). mercial sources: creatine kinase and creatine phosphate (Boehringer Mannheim); leu-enkephalin and somatostatin (Peninsula). All other Pharmacological characterization of the opioid receptor compounds were purchased from Sigma Chemical Company and were of the highest purity available. negatively coupled to adenylate cyclaseon embryonic striatal Radiochemicals. (a32P)-ATP sodium salt (30-40 Ci/mmol) and 2-8- neurons ‘H-CAMP (25 Ci/mmol) were obtained from New England Nuclear Binding studies have indicated that DAGO, DTLET, and U Corp. 50.488 are selective agonistsfor the mu, delta, and kappa binding sites, respectively, in the adult brain. The kappa agonistU Results 50.488 waswithout effect on adenylate cyclaseactivity in striatal Leu-enkephalin-induced inhibition of adenylate cyclaseactivity embryonic neurons. On the other hand, DAGO and DTLET in mouseembryonic striatal neuronsin primary culture inhibited CAMP production with an almost identical efficacy Primary cultures of pure neuronal or glial cells from the cerebral (Fig. 2). Hofstee plots of the data revealed an interaction of each cortex, mesencephalon,and striatum were usedto look for an agonistwith a singleclass of noninteracting receptors, their ap- effect of leu-enkephalin on adenylate cyclaseactivity in mem- parent affinities (K& being very similar: DTLET, 10.8 * 3.3 branes. As shown in Table 1, using leu-enkephalin concentra- x 1O-8M (n = 5); DAGO, = 3.6 f 0.7 x 1O-8M (n = 5). tions up to 3 x 10m6M, striatal neuronswere the only cells that The inhibitory effectsof DTLET (3 x lo-’ M) and DAGO (3 respondedto the opioid agonist, leu-enkephalin (1O-6M), pro- x 10m7M) were both antagonized by naloxone up to 3 x 1O-6 ducing a 30 + 4% (n = 4) inhibition of the basal adenylate M (Fig. 3). The linear Hofstee plots of the data confirmed that cyclase activity on striatal neuronal membranes.The maximal each agonist interacted with a single class of receptors. The inhibitory effect of the peptide and its apparent affinity for its calculated apparent affinities of naloxone for the DTLET- and receptoron mouseembryonic striatal neuronswere slightly higher DAGO-sensitive receptors were 8.5 f 5.3 x 1O-9and 2.3 f than thoseobserved on striatal homogenatesfrom the adult rat 0.9 x 1O-9M, respectively.

Figure 3. Antagonism by naloxone of STRIATAL NEURONES DAGO- and DTLET-induced inhibition of adenylate cyclase activity on embryonic striatal neurons in primary cultures. Left, Increasing concentra- tions of naloxone reversed the inhibitory effects of DAGO (3 x lo-’ M) and DTLET (3 x lo-’ M) on CAMP production. Right, Hofstee plots ofthe data from which IC,, values of 3 x lo-* and 4 x 10-s M were obtained in the presence of DAGO and DTLET, respectively. K, values were determined according to the formula K, = IC,,/l + (AgIK,) in which Ag is the concentration of DAGO or DTLET and Kd their dissociation constants. Data presented correspond to results from one experiment out of three giving similar results. Jy-+y+-y eO Each point is the mean of triplicate determinations. NALOXONE (log [M]) I/S The Journal of Neuroscience, September 1988, 8(9) 3379

I I NG 108-15 Adult Striatum Figure 4. Comparisonof the effectsof DAGO and DTLET on basal adenylate cyclase activity on membranes from the neuroblastoma x glioma hybrid cell line NG 108-15 and from the rat striatum. DTLET significantly inhibited adenylate cyclase activity in both types of preparations (54 f 4% and 21 + 2%, respectively, for NG 108-15 cells and adult striatal tissue at a concentration of 10e6 M). The maximal inhibition due to DAGO on striatal membranes was always less than 12% of basal activity. Data illustrateresults obtained in one experiment out of three yielding similar results. The apparent affinities of DTLET (K,,,,,): 6.3 f 1.2 x lo-* and 2.5 + 0.5 x 10m’~forNG 108-15cells and striatal adult tissue. resoectivelv. were obtained from Hofitee plots(n = AGONIST (log[M]) 3).

For comparison, the effects of DTLET and of DAGO were only a weak effect, closeto that of DAGO, was found on striatal also examined on the neuroblastoma x glioma hybrid cell line adult tissue(11 f 2% and 8 -t 2% for trimu 5 and morphiceptin, NG 108-l 5, which is known to possessonly delta receptors. respectively; 12= 3). Experiments were also made on striatal homogenatesof the adult rat, which possessboth mu and delta binding sitesand in Cellular localization of mu and delta receptorson embryonic which only delta receptorshave been reported to be negatively striatal neuronsin primary culture coupled to adenylate cyclase.As illustrated in Figure 4, DTLET, As previously reported (Premont et al., 1983), indications about but not DAGO, markedly inhibited basaladenylate cyclaseac- the cellular localization of different types of receptorscoupled tivity on NG 108-15 membranes, its apparent affinity being to adenylate cyclasecan be obtained by experiments which de- similar to that observed on embryonic striatal neurons.DTLET termine whether the additive effects of the correspondingago- was much more potent than DAGO in inhibiting CAMP pro- nistsare present. Complete additivity of the effectsof 2 distinct duction on striatal membranesof the adult rat. Naloxone an- agonistsfavors a localization of the correspondingreceptors on tagonized the inhibitory effect of DTLET monophasically (data distinct cellular populations or on distinct cellular domains. not shown) in both types of preparations, allowing the respective IC,, and K, values to be calculated (Table 2). In order to determine whether the inhibitory effect of DAGO STRIATAL NEURONES on adenylate cyclaseactivity in embryonic striatal neuronsfrom the mouse was not speciesdependent, additional experiments were made using striatal membranesof the adult mouse. As with Forskolin(lO% observed on striatal membranesof the adult rat, DTLET, but not DAGO, produced a marked inhibition of CAMP production (data not shown). Finally, 2 other potent mu agonists,trimu 5 and morphiceptin, were usedand their effectsup to 1O-6M were testedon both homogenatesfrom striatal neuronsin culture and adult rat striatal membranes.A significant inhibition of basaladenylate cyclase activity was observed with both trimu 5 (27 f 3%, n = 3) and morphiceptin (26 f 4%, n = 3) on neurons, whereas

Table 2. Comparison of the efficacy of naloxone in reversing inhibition of CAMP production by DTLET on striatal neurons in primary culture, NG 108-15 cells, and adult striatum DAGO (3xlO’hl) + + f + + + + + Tissue G bf) K MM) DTLET( 3xlO’M) Striatal neurons 25 f 11 8.5 + 5.3 Figure 5. Nonadditivity of the inhibitory effects of DTLET and DAGO Adult rat striatum 34 + 12 15.5 f 8.6 on basal and forskolin-stimulated adenylate cyclase activity. Adenylate cyclase activity was estimated on homogenates of striatal neurons grown NG 108-15 130 +- 35 22.6 + 10.4 in primary culture in the absence or presence of forskolin, DAGO (3 x IC,, and K, values for naloxone were determined from Hofstee plots. The lo-’ M), and DTLET (3 x lo-’ M) being added alone or simultaneously. concentration of DTLET used was 3 x lo-’ M. Data are means + SEM of results Data are means + SEM of results obtained in 3 independent experi- obtained in 3 experiments. ments. 3380 Chneiweiss et al. * Opioid Receptors on Mouse Striatal Neurons

STRIATAL NEURONES Figure 6. Additivity of the inhibitory effects of leu-enkephalin and somatostatin on basal adenylate cyclase activity in embryonic striatal neurons in primary culture. The inhibitory effects of somatostatin alone or in the uresence of leu-enkephalin ( 10m6M) (/efl) and of leu-enkephalin alone or in the presence of somatostatin (3 x 10m6M) (right) on homogenates of striatal neurons. Nei- ther the maximal effect nor the apparent affinity of each peptide was affected If’: 0 SOMATOSTATIN ALONE by the presence of the other, indicating 4 ul the complete additivity of their inhib- zif . +LEU-ENK(lC?#) itory effects. Data are means of results x E a-” 30 obtained in 3 independent experi- i 30 b 0% ’ I I I 0-h ’ I I / ments. SinceSEMs were always less than -6 0 -9 -6 -7 -6 7% of the mean values, they are not 0 -9 -8 -7 represented. SOMATOSTATIN (log[hl]) LEU-ENKEPHALIN ( log [M] 1

Nonadditivity demonstrates a colocalization of the 2 types of (1 Om5M) alone markedly stimulated adenylate cyclase activity. receptors on the same cells. In the presence of dopamine (1 O-5 M), the decreases in CAMP As shown in Figure 5, the inhibitory effects of DAGO (3 x production induced by DTLET and DAGO were about twice 1Om7 M) and DTLET (3 x lo-’ M) on basal adenylate cyclase those observed in the absenceof dopamine (Table 3). These activity were nonadditive. Similar results were obtained in the observations favor the colocalization of opiate receptors with presence of forskolin (10m5 M), an activator of the adenylate dopaminergic D 1 receptors on a subpopulation of striatal neu- cyclase catalytic subunit. This result suggested that mu and delta rons. receptors were located on the same population of striatal embryonic neurons. Discussion Complementary experiments were made with somatostatin, The present study demonstratesthat 16-d-old embryonic neu- which has also been shown to inhibit basal adenylate cyclase rons from the mouse striatum grown for 6 d in primary culture activity in embryonic striatal neurons(Chneiweiss et al., 1985b). possessdelta opiate receptors negatively coupled to adenylate Interestingly, the inhibitory effects of leu-enkephalin and so- cyclase. In addition, experiments with DAGO, a selective mu matostatin were completely additive (Fig. 6). These resultsin- agonist, suggestthat mu receptors coupled negatively to the dicate that additive inhibitory effects can be demonstratedand enzyme are also present on these cells. Both types of opiate suggestthat opiate and somatostatin receptorsare not colocal- receptors (mu and delta) seem to be colocalized on the same ized on the same population of striatal cells or on the same subpopulation of striatal neurons.Interestingly, leu-enkephalin, morphological domain of the cells. which has been shown in electrophysiologic studies to be as Finally, since dopamine has been shown to modify the char- potent on mu and delta receptors(North, 1986), did not modify acteristics of the inhibitory responseto somatostatin (Chnei- CAMP production either in embryonic neurons from the cere- Weisset al., 1985b), similar experiments were made usingeither bral cortex or mesencephalon,or in astrocytes from the 3 brain DTLET or DAGO, or both. As previously shown, dopamine structures examined. As shown in the adult mouse, rat, and other species,the striatum contains both mu and delta binding sites(Quit-ion et Table 3. Potentiation by dopamine of the DTLET- and DAGO- al., 1983; Moskowitz and Goodman, 1984; Eghbali et al., 1987). induced decrease in CAMP production on striatal neurons in primary Delta binding siteshave a diffuse distribution, while the local- culture ization of mu binding sitesis restricted to the striosomes(Good- Adenylatecyclase man et al., 1980; Herkenham and Pert, 1981). An opiate-sen- activity (pmolCAMP/ 10 sitive adenylate cyclase is present on striatal membranes(Law CAMP production min/mg protein) et al., 1981; Cooper et al., 1982) and as confirmed in the present Basalactivity 68 k 5 study, delta receptors are negatively coupled to the enzyme in + DTLET (1Om6 M) 49 f 4(100) both adult rats and adult mice. As observed in adult animals, + DAGO (1O-6M) 47 f 3 (110) DTLET, a selective agonist of delta receptors (Garbay-Jaure- + Dopamine(1 O-5 M) 153 + 9 guiberry et al., 1984) inhibited basalCAMP production in mem- + Dopamine(1 O-5 M) branesof embryonic striatal neurons, and this effect was antag- + DTLET (1O-6 M) 115 ? 7 (200) onized by naloxone. Although the patterns of the inhibitory + Dopamine(1 O-5 M) responseswere comparable,the apparent affinity of DTLET for + DAGO (10m6M) 110 + 8 (205) the embryonic striatal delta receptorsseemed to be slightly higher than that observed in the adult rat. This could be linked to the Data are means + SEM ofresults obtained in 4 (DTLET) or 2 (DAGO) experiments, each independent experiment being performed in triplicate. The absolute decrease presenceof supersensitivedelta opiate receptorson embryonic in CAMP production induced by DTLET and DAGO was 19 f 6 and 21 k 5 neurons in primary culture, as has been observed with other pm01 cAMP/lO min/mg protein, respectively, in the absence ofdopamine and 38 types ofreceptors (Premont et al., 1983; Chneiweisset al., 1984). 5 6 and 43 + 8 pmol cAMP/lO min/mg protein in the presence of dopamine. Values in parentheses indicate percentage of the decrease in CAMP production However, the apparent affinity of DTLET for inhibiting ade- induced by DTLET on basal activity. nylate cyclase activity in both embryonic neurons and adult The Journal of Neuroscience, September 1988, 8(9) 3381 striatum was lower than that reported in binding studies. Most striatal neuronspossessing mu and delta receptorsalso possess likely, this is due to the presence of the high concentration of DA D 1 receptors. Indeed, the amplitude of the inhibitory effects GTP required in the adenylate cyclase assay for coupling be- of DAGO or DTLET was increasedin the presenceof DA. These tween the agonist receptor complex, the G protein, and the results can also be discussedin light of those of Weiss et al. catalytic subunit of adenylate cyclase (Chneiweiss et al., 1987). (1985) who reported that the inhibitory effects of opiate and Such a sensitivity of opiate agonist binding to GTP has been dopaminergicD2 agonistson CAMP production in intact striatal observed previously on membranes from adult brain tissues and neurons in primary culture were nonadditive. Such a nonad- the neuroblastoma x glioma hybrid cell line NG 108- 15 (Law ditivity favors colocalization of dopaminergic D2 and opiate et al., 1985; Zajac and Roques, 1985). receptorson neuronal subpopulations.This allowsus to suggest, Surprisingly, DAGO, a potent selective agonist of mu recep- on the basis of our data, that this subpopulation of sttiatal tors, which had a weak effect on striatal membranes of adult neurons possessesDl and D2 dopaminergic receptors, as well animals (rat or mouse) and no effect on the hybrid cell line NG asmu and delta opioid receptorscoupled to an adenylate cyclase. 108-l 5, markedly inhibited CAMP production in membranes This conclusion may seemto contradict autoradiographic find- from embryonic striatal neurons in primary culture with an ings indicating a distinct distribution of Dl (striosomes)and efficacy similar to that observed with DTLET. In contrast, the D2 (matrix) receptorsin the striatum of adult mammals(Joyce kappa-selective agonist U 50.488 was ineffective. The kinetic et al., 1986; Bessonet al., 1988). However, as observed for the characteristics of the DAGO-induced inhibition of adenylate inhibitory effects of DAGO or DTLET, we have previously cyclase activity were similar to those observed with DTLET, shown that the inhibitory effect of somatostatin on adenylate DAGO even being slightly more potent. Naloxone, an antago- cyclaseactivity increasesin the presenceof DA (Chneiweisset nist known to be more effective on mu than delta opiate recep- al., 1985b), indicating colocalization of Dl and somatostatin tors, reversed the inhibitory effect of DAGO on CAMP pro- receptors. Thus, the present resultssuggest that the Dl receptors duction with a higher efficacy than that observed with DTLET. are either colocalized with mu and delta opioid receptors or Other selective mu agonists,such as trimu 5 and morphiceptin, with somatostatin receptorson 2 distinct subpopulationsof em- inhibited basaladenylate cyclase activity in striatal embryonic bryonic striatal neurons. This conclusion is supported by the neurons with a potency similar to that of DAGO. Our data completeadditivity of the inhibitory effectsof somatostatinand clearly indicate a coupling of mu receptors with adenylate cy- DAGO or DTLET on basaladenylate cyclaseactivity. clasein central neurons.It should be recalledthat suchcoupling Besidesthe striatum, several structures (including the cerebral hasbeen observed in tumor cell line 73 15~(Frey and Kebabian, cortex and mesencephalon)possess delta and mu opiate binding 1984) and that mu agonists have been shown to reduce the sites in the adult brain (Quirion et al., 1983; Moskowitz and CAMP efflux from striatal slicesof the adult rat (Schoffelmeer Goodman, 1984). In fact, the cerebral cortex is one of the struc- et al., 1985, 1987). In addition, it has recently been reported tures with the greatestdensity ofdelta binding sites.Surprisingly, that DAGO inhibits CAMP production in homogenatesfrom in contrast to resultsobtained in the striatum, opiate receptors the neonatal rat forebrain (Milligan et al., 1987) and the adult coupled negatively to adenylate cyclasewere not found on em- guinea pig cochlea (Eybalin et al., 1987). bryonic neurons from the cerebral cortex and the mesenceph- Electrophysiological studies made in the periphery or CNS alon in primary culture. This is not due to a delayed appearance indicate that mu and delta opiate receptors can be colocalized of functional adenylate cyclase in these cells, since VIP and on the samecells (Egan and North, 1981; Zieglgansbergeret al., somatostatin receptors, respectively, stimulate and inhibit the 1982). Such colocalization of both types of receptorsnegatively enzyme in these cells (Chneiweisset al., 1985a, b). It remains coupled to adenylate cyclase also seemsto occur on the same to be determined whether opiate receptorsare localized on in- neuronal subpopulation in the striatum of the mouse embryo, trinsic neuronsin thesestructures or whether specificfactors are since the inhibitory effects of DAGO and DTLET on basalor required for the appearancein vitro of functional opiate recep- forskolin stimulated activity of adenylate cyclasewere nonad- tors on neurons from the mesencephalonor cerebral cortex in ditive. Interestingly, it has been suggestedthat the different to- particular. pographic distribution of the mu (striosomes)and delta (diffuse) During the last few years, an increasingnumber of functional opiate binding sites in the adult striatum could be modulated receptorshave beenidentified on astrocytes(Murphy and Pearce, by ions and GTP (Bowen et al., 1981) and that these binding 1987) and it has been proposed that they could play an im- sitesmay be allosterically coupled and may eventually share a portant role in the interactions betweenastrocytes and neurons. common component (Rothman et al., 1985; Bidlack and O’Mal- This is the case particularly for the VIP receptors coupled to ley, 1986). According to this model, differencesin the ontogeny adenylate cyclase that have been characterized on astrocytes or regional distribution in the adult of these2 types of binding from several structures of the mouse brain (Chneiweisset al., sites could be related to some environmental changesin the 1985a, 1986). The absenceof opiate receptors coupled to ade- membrane composition or cellular activity. It should also be nylate cyclase on astrocytes of all structures examined in the pointed out that the density of mu receptors does not change present study suggestthat not all neuropeptidesare involved in after birth, whereas that of delta receptors increases40-fold neuron-glia interactions or that opioids may act on glial cells (Milligan et al., 1987). This difference in the density of mu and through a second messengerother than CAMP. delta receptorscould explain the much weaker inhibitory effects of DAGO, trimu 5, and morphiceptin on CAMP production in References adult striatal membranescompared with that of DTLET. Besson, M. J., A. M. Graybiel, and M. A. Nastuk (1988) (3H) SCH A recent investigation in the adult rat striatum has suggested 32 390 binding to Dl dopaminergic receptors in basal ganglia of cat and primate: Delineation of striosomal compartment and pallidal and the occurrence of heteroregulation between DA Dl receptors nigral subdivision. J. Neurosci. (in press). and opiate receptors (Schoffelmeer et al., 1986). Our results Bidlack, J. M., and W. E. O’Malley (1986) Inhibition of mu and delta favor the existence of such heteroregulation and indicate that but not kappa-opioid binding to membranes by Fab fragments from 3382 Chneiweiss et al. - Opioid Receptors on Mouse Striatal Neurons

a monoclonal antibody directed against the opioid receptor. J. Biol. Leslie, F. M. (1987) Methods used in the study of opioids. Pharmacol. Chem. 261: 15844-15849. Rev. 39: 197-249. Bowen, W. D., S. Gentleman, M. Herkenham, and C. B. Pert (1981) Lord, J. A. H., A. A. Waterfield, J. Hughes, and H. W. Kosterlitz (1977) Interconverting mu and delta forms of the opiate receptor in rat Endogenous opioid peptides: Multiple agonists and receptors. Nature striatal patches. Proc. Natl. Acad. Sci. USA 78: 48184822. 267: 495499. Chneiweiss, H., A. Prochiantz, J. Glowinski, and J. Premont (1984) Martin. W. R. (1984) The uharmacologv-_ of opioids._ Pharmacol. Rev. Biogenic amines sensitive adenylate cyclases in primary culture of 35: 283-323: - neuronal or glial cells from mesencephalon. Brain Res. 302: 363-370. Milligan, G., R. A. Streaty, P. Gierschik, A. M. Spiegel, and W. A. Klee Chneiweiss. H.. J. Glowinski. and J. Premont (1985a) VIP recenters (1987) Develonment of opiate receptors and GTP-binding regulatory linked to an adenylate cyclase and their relationship to’biogenic amines proteins in neonatal rat brain. J. Biol. Chem. 262: 8626-8650. - and somatostatin-sensitive adenylate cyclases on central neuronal and Moskowitz, A. S., and R. R. Goodman (1984) Light microscopic glial cells in primary cultures. J. Neurochem. 44: 779-786. autoradiographic localization of mu- and delta-opioid binding sites Chneiweiss, H., J. Glowinski, and J. Premont (1985b) Modulation by in the mouse central nervous system. J. Neurosci. 4: 133 l-l 342. monoamines of somatostatin-sensitive adenylate cyclase on neuronal Murphy, S., and B. Pearce (1987) Functional receptors for neurotrans- and glial cells from mouse brain in primary cultures. J. Neurochem. mitters on astroglial cells. Neuroscience 22: 38 l-394. 44: 1825-1831. North, R. A. (1986) Opioid receptor types and membrane ion chan- Chneiweiss, H., J. Glowinski, and J. Premont (1986) Do secretin and nels: Trends ‘Nemo& 9: 114-117. - VIP have independent receptors on striatal neurones and glial cells Paterson. S. J.. L. E. Robson. and H. W. Kosterlitz (1983) Classifi- in primary cultures? J. Neurochem. 47: 608-6 13. cation’of opioid receptors. Br. Med. Bull. 39: 3 l-36: ’ Chneiweiss, H., P. Bertrand, J. Epelbaum, C. Kordon, J. Glowinski, J. Piercey, P. F., R. A. Lahti, L. A. Schroeder, F. J. Einspahr, and C. Premont, and A. Enjalbert (1987) Somatostatin receptors on cortical Barsuhn (1982) LJ 50488 H, a pure kappa receptor agonist with neurones and adenohypophysis: Comparison between specific binding spinal analgesic loci in the mouse. Life Sci. 31: 1197-1200. and adenylate cyclase inhibition. Eur. J. Pharmacol. 138: 249-255. Premont, J., M. C. Daguet de Montety, A. Herbet, J. Glowinski, J. Cooper, D. M. F., C. Londos, D. L. Gill and M. Rodbell (1982) Opiate Bockaert, and A. Prochiantz (1983) Biogenic amines and adenosine receptor-mediated inhibition of adenylate cyclase in rat striatal plas- sensitive adenylate cyclases on striatal neurones. Dev. Brain Res. 9: ma membranes. J. Neurochem. 38: 1164-l 167. 53-61. Di Porzio, U., M. C. Daguet, J. Glowinski, and A. Prochiantz (1980) Prochiantz, A., A. Delacourte, M. C. Daguet, and D. Paulin (1982) Effect of striatal cells on in vitro maturation of mesencephalic do- Intermediate filament proteins in mouse brain cells cultured in the paminergic neurones grown in serum free conditions. Nature 288: presence or absence of fetal calf serum. Exp. Cell Res. 139: 4Ou 10. 370-373. Quirion, R., J. M. Zajac, J. L. Morgat, and B. P. Roques (1983) Au- Egan, T. M., and R. A. North (198 1) Both mu and delta-opiate re- toradiographic distribution of mu and delta opiate receptors in rat ceptors exist on the same neuron. Science 214: 923-924. brain using highly selective ligands. Life Sci. Suppl. I 227-230. Eghbali, M., C. Santoro, W. Paredes, E. L. Gardner, and R. S. Zukin Rothman, Ry B.,W. D. Bowen,M. Herkenham, A. E. Jacobson, K. C. (1987) Visualization of multiple opioid receptor types in rat striatum Rice. and C. B. Pert (1985) A auantitative studv of ‘H(D-AlaZ-D- after suecific mesenceuhalic lesions. Proc. Natl. Acad. Sci. USA 84: Leu$ enkephalin binding to rat brain membranes, evidence that ex- 6482-6586. - omorphine is a non-competitive inhibitor of the lower affinity delta- Eybalin, M., R. Pujol, and J. Bockaert (1987) Opioid receptors inhibit binding site. Mol. Pharmacol. 27: 399-408. the adenylate cyclase in guinea pig cochleas. Brain Res. 421: 336- Salomon, Y., C. Londos, and M. Rodbell (1974) A highly sensitive 342. adenylate cyclase assay. Anal. Biochem. 58: 541-548. Frey, E. A., and J. W. Kebabian (1984) A mu-opioid receptor in 7325~ Schoffelmeer, A. N. M., H. A. Hansen, J. C. Stoof, and A. H. Mulder tumor tissue mediates inhibition of immunoreactive nrolactine re- (1985) Inhibition of dopamine-stimulated cyclic AMP efflux from lease and adenylate cyclase activity. Endocrinology llj: 1797-1804. rat nebstriatal slices by activation of mu- and-delta-opioid receptors. Garbay-Jaureguiberry, C., A. Robichon, V. DaugeT-P. Rossignol, and Eur. J. Pharmacol. 118: 363-366. B. Roaues f1984) Hiahlv selective Dhotoaffinitv labeling of mu and Schoffelmeer. A. N. M., H. A. Hansen, J. C. Stoof, and A. H. Mulder delta opioid‘ receptors.-Proc. Natl. Acad. Sci. USA 81: 7’i 18-7722. (1986) Blockade of dopamine D2 receptors strongly enhances the Glowinski, J., and L. L. Iversen (1966) Regional studies of catechol- potency of enkephalins to inhibit dopamine-sensitive adenylate cy- amines in the rat brain. I. Disposition of )H-norepinephrine, 3H- clase in rat neostriatum: Involvement of mu- and delta-opioid re- dopamine and )H-DOPA in various regions of the brain. J. Neuro- ceptors. J. Neurosci. 6: 2235-2239. them. 13: 655-669. Schoffelmeer, A. N. M., F. Hogenboom, and A. H. Mulder (1987) Goodman. R. R.. S. H. Snvder. M. Kuhar. and W. S. Youna (1980) Inhibition of dopamine-sensitive adenylate cyclase by opioids: Pos- Differentiation’of mu and delta opiate receptor localization; by light sible involvement of physically associated mu- and delta-opioid re- microscopic autoradiography. Proc. Natl. Acad. Sci. USA 77: 6239- ceptors. Nanyn Schmiedberg’s- Arch Pharmacol. 335: 278-284. 6243. Sharma. S. K.. M. Nirenbera. and W. A. Klee (1975) Morphine re- Handa, B. K., A. C. Cane, J. A. H. Lord, B. A. Morgan, M. J. Rance, ceptors as regulators of adenylate cyclase activity. Proc. Natl. Acad. and C. F. C. Smith (1981) Analogues of P-LPH,,,, processing se- Sci. USA 72: 590-594. lective against activity of mu-opiate receptors. Eur. J. Pharmacol. 70. Simon, E. J. (1987) Subunit structure and purification of opioid re- 53 l-540. ceptors. J. Receptor Res. 7: 105-132. Herkenham, M., and C. B. Pert (198 1) Mosaic distribution of opiate Weiss, S., M. Sebben, J. A. Garcia-Sainz, and J. Bockaert (1985) D2- receptors, parafascicular projections and acetylcholinesterase in rat dopamine receptor mediated inhibition of cyclic AMP formation in striatum. Nature 291: 4 154 17. striatal neurons in primary culture. Mol. Pharmacol. 27: 595-599. Joyce, J. N., D. W. Sapp, and J. F. Marshall (1986) Human striatal Zajac, S. M., and B. P. Roques (1985) Differences in binding properties dopamine receptors are organized in compartments. Proc. Natl. Acad. of mu- and delta-opioid receptor subtypes from rat brain: Kinetic Sci. USA 83: 8002-8006. analysis and effects of ions and nucleotides. J. Neurochem. 44: 1605- Law, P. Y., J. Wu, J. Koehler, and H. H. Loh (1981) Demonstration 1614. and characterization of opiate inhibition of the striatal adenylate cy- Zieglgansberger, W., E. D. French, N. Merecuri, F. Pelayo, and J. T. clase. J. Neurochem. 36: 1834-1846. Williams ( 1982) Multiple opiate receptors on neurons of the mam- Law, P. Y., D. S. Horn, and H. H. Loh (1985) Multiple affinity states malian central nervous system. In vivd and in vitro studies. Life Sci. of opiate receptor in Neuroblastoma x Glioma NG108-15 hybrid 31: 2343-2346. cells. J. Biol. Chem. 260: 3561-3569.