- Email: [email protected]
Blockade of morphine analgesia by an antisense oligodeoxynucleotide against the mu receptor
Life Sciences, Vol. 54, No. 21, pp. PL 375-379, 1994 Copyright © 1994 ELsevier Science Ltd Printed in the USA. All rights regclx,ed OO24-3205/94 $6.OO + .00
Pergamon
PHARMACOLOGY LETTERS Accelerated Communication
B L O C K A D E O F M O R P H I N E A N A L G E S I A BY A N A N T I S E N S E OLIGODEOXYNUCLEOTIDE AGAINST THE MU RECEPTOR Grace Rossi, Ying-Xian Pan, Jie Cheng and Gavril W. Pasternak The Cotzias Laboratory of Neuro-Oncoiogy Memorial Sloan-Kettering Cancer Center and Departments of Neurology & Neuroscience and Pharmacology Cornell University Medical College New York, NY 10021 (Submitted January 20, 1993; accepted February 2, 1994; received in finalform March 9, 1994)
Abstract: The recent cloning of mu, delta and kappa~ opioid receptors has provided opportunities in the study of their pharmacology. Using an antisense strategy developed against delta and kappa~ opioid receptors, we designed an antisense oligodeoxynucleotide directed against the 5'-untranslated region ofMOR-1 clone, 51-70 bp upstream from the initiating ATG. Microinjection of this antisense oligodeoxynucleotide directly into the periaqueductal gray on Days 1, 3 and 5 completely blocked the analgesic actions of morphine administered into the periaqueductal gray on Day 6 (p < 0.001), 24 hr after the last antisense treatment. Rats treated with vehicle or with a mismatch oligodeoxynucleotide in which two pairs of bases from the antisense sequence had been switched were not significantly affected. These findings confirm the pharmacological relevance of the MOR-1 clone and its involvement in morphine's actions. Key Words: morphine, antisense treatment, MOR-1 clone, mu opioid receptor
Introduction
Morphine, the classic opioid analgesic, acts through mu opioid receptors (1). The recent cloning of the delta opioid receptor (2,3) quickly led to the identification of clones encoding mu and kappat receptors (4-9). The binding profiles of these expressed receptors is similar to that expected from traditional binding studies in brain membranes and all the receptors are functionally active when transfected into cells lacking opioid receptors. However, correlating these cloned receptors with pharmacological activity requires additional approaches. Antisense strategies provide a method for selectively downregulating specific mRNA's and their proteins (10,11). This approach has been effectively utilized against G-protein receptors and Corresponding author: Gavril W. Pastemak, M.D., Ph.D., Departmentof Neurology,Memorial SloanKettering Cancer Center, 1275 York Avenue,New York, NY 10021
PL-376
Antisense Oligodeoxynucleotideon Analgesia
Vol. 54, No. 21, 1994
has proven valuable in in vivo studies ofneuropeptide Y1 and NMDA receptor activity (12-14). Prior work from our group has demonstrated that antisense oligodeoxynucleotides against a delta opioid receptor selectively blocks 3H-DPDPE binding in a neuroblastoma cell line and both DPDPE and deitorphin II analgesia in mice (15,16). In binding studies antisense oligodeoxynucleotides directed against five different regions of the open reading frame of the delta receptor show similar decreases in binding, indicating that the location of the antisense is not crucial for activity. Detailed studies with mismatch and sense oligodeoxynucleotides confirmed the selectivity of the decrease in binding. In vivo treatments demonstrated a potent blockade of DPDPE and deltorphin II analgesia without interfering with the analgesic actions of either the mu drug morphine or the kappa~ agent U50,488H. Mismatch oligodeoxynucleotides were inactive. Equally important, analgesic sensitivity returned to normal over a period of 5 days following the discontinuation of antisense treatment, ruling out a nonspecific or toxic effect of the treatments. Similar studies demonstrated that an antisense oligodeoxynucleotide directed against the kappa~ receptor given intrathecally blocked U50,488H analgesia, but not DPDPE or morphine analgesia (17). Again, the selectivity of the action was confirmed by the inactivity of a mismatch oligodeoxynucleotide. The role of the cloned mu receptor in morphine pharmacology has not yet been established. We now report that an antisense oligodeoxynucleotide directed towards the 5'-untranslated region of the mu receptor clone microinjected into the periaqueductal gray blocks morphine analgesia. Materials and Methods
Male Sprague-Dawley rats (250 g) were purchased from Charles River (Wilmington, MA). Oligodeoxynucleotides were synthesized by Midland Certified Reagent Co. (Midland, TX). The antisense oligodeoxynucleotide (Table 1) is directed against the 5'-untranslated region (bases -87 to -69 upstream from the initiating ATG). The mismatch oligodeoxynucleotide corresponds to the antisense sequence except for two pairs of bases which have been switched. Rats were cannulated in the periaqueductal gray, as previously described (18-20). Histological studies at the conclusion of the study confirmed the placement. The animals were divided into three groups and all groups received a test dose of morphine to ensure proper cannulae placement. The groups received vehicle or either the antisense or mismatch oligodeoxynucleotides (10 lag in 1 lal) on Days 1, 3 and 5. Twenty-four hours after the last treatment, the animals were tested with morphine. Analgesia was assessed in a graded manner using the tailflick assay (18-20). Results and Discussion
Morphine is a potent analgesic when administered into the periaqueductal gray of rats (18-20; Fig. 1). in the current study, morphine administered prior to any treatment (Control) is analgesic in all three groups. Following the three treatment injections, there is a small decrease in morphine's analgesic potency in the vehicle group (Fig. la), possibly reflecting the effects of multiple injections. The administration of the mismatch oligodeoxynucleotide has little effect on morphine anlagesia (Fig. lb). The small decrease in analgesic effect is quite similar to that observed with vehicle injections. In contrast, treatment with the antisense oligodeoxynucleotide eliminates the analgesic actions of morphine, lowering the tailflick latencies to baseline values (Fig. Ic; p < 0.001).
Vol. 54, No. 21, 1994
Antisense Oligodeoxynudeotideon Analgesia
PL-377
A S
~ntrol
~
,
C~ntrol, 6
fi 0
30
t
t
60 T i m e (rain)
~
I
i 90
i
30
120
~ntroi
8
I
i
60 T i m e (rain)
i
: 90
120
.
6
44 Ba~line 30
L 60 T i m e (mill)
L 90
120
Fig. 1 Effects of vehicle, antisense and mismatch treatments on morphine analgesia: Rats received periaqueductal gray cannulae and were allowed to recover from surgery for at least one week. Animals were divided into three groups and we assessed morphine analgesia (2.5 ~ag in 1 lal) in each animal to confirm the cannulae placement pharmacologically and to establish the morphine sentivity in each group (Control). After waiting another week, the groups then were treated on Days 1, 3 and 5 with either A) vehicle (n=ll), B) mismatch (10 lag, n=5) or C) antisense (10 lag, n=10) oligodeoxynucleotides. Morphine then was injected on Day 6 and analgesia compared to the control curve from the same group. Control and treatment testing do not differ significantly for either the vehicle or the mismatch group, comparing either peak effects (30 min) or areas under the curves. Antisense treatment significantly reduces morphine analgesia, looking at either peak effects (p < 0.00l) or areas under the curves (p < 0.001). Our current study dearly demonstrates the ability of an antisense oligodeoxynucleotide to the mu receptor to interfere with morphine analgesia. The inability of the mismatch oligodeoxynucleotide to influence morphine anlagesia confirms the specificity of the response. We chose to examine these actions in the periaqueductal gray for its sensitivity to morphine and the well-localized site of action. We previously utilized a similar approach for the study of pertussis and cholera toxins and morphine analgesia (20). Intracerebral injections also avoid the dilution and rapid clearance associated with intracerebroventricular administration.
PL-378
Antisense Oligodeoxynudeotideon Analgesia
Vol. 54, No. 21, 1994
TABLE 1 Oligodeoxynucleotide sequences
Oligodeoxynucleotide (5' to 3') Sense
AGA GGA AGA C_rGCTGG GGC G
Antisense
CGC CCC AGC CTC TTC CTC T
Mismatch
CGC CCC GAC CTC TTC CCT T
Prior work in mice has shown that antisense oligodeoxynucleotides directed against the open reading frames of delta or kappal receptors selectively block the analgesic actions of their respective ligands, but not morphine (15-18). Our binding studies indicate that the region of the open reading frame targeted is not important. Antisense oligodeoxynucleotides directed at five different regions of the open reading frame ofDOR-1 downregulate delta binding in the NGI08-15 cells to the same extent. In our current study we targeted the antisense oligodeoxynucleotide to the 5'-untranslated region, approximately 50 bp upstream from the initiating ATG. Untranslated regions are specific for the gene in question and do not show the homology typically seen in the open reading frames among related gene products. Furthermore, Southern analysis appears to indicate the presence of only a single MOR-1 gene (17). Thus, the downregulation of morphine analgesia by our antisense oligodeoxynucleotideimplies that the gene encoding the MOR-1 clone is involved with mu analgesia. However, some questions still remain. Mu~ receptors mediate morphine analgesia in the periaqueductal gray (18-20), suggesting that the MOR-1 clone might correspond to the mu I receptor. Mounting evidence suggests the presence of alternative splicing in the processing of opioid receptor clones and it is possible that subtypes might reflect alternative splicing of a single gene. These alternatively spliced mRNA species might contain the same 5'-untranslated sequence as MOR-1 and also would be downregulated. Although the results using the antisense oligodeoxynucleotide directed toward the region indicate that the MOR-1 clone derives from the gene encoding mu~ receptors, we do not yet know whether the MOR-1 clone represents the mu] receptor or an alternatively spliced variant of the same gene.
Acknowled2ments We thank Dr. J. Posner for his support of this research and Drs. C. Wahlestedt and H. Furneaux for helpful discussions. This work was supported, in part, by grants from NIDA to GWP (DA02615 and DA07242) and a core grant from the NCI to MSKCC (CA08748). GWP is supported by an RSDA from NIDA (DA000138), GR by a Training Grant from NIDA (DA07274) and YXP by a Fellowship from the Aaron Diamond Foundation.
References 1. G.W. PASTERNAK, Clin. Neuropharmacol. 16 1-18, (1993). 2. C.J. EVANS, D.E. KEITH, JR., H. MORRISON, K. MAGENDZO and R.H. EDWARDS, Science 258 1952-1955 (1992). 3. B.L. KIEFFER, K. BEFORT, C. GAVERIAUX-RUFF and C.G. HIRTH, Proc. Nat. Acad. Sci. USA 8__9912048-12052 (1992)
Vol. 54, No. 21, 1994
Antlsense Oligodeoxynucleotideon Analgesia
PL-379
4. Y. CHEN, A. MESTEK, J. LIU, J.A. HURLEY and L. YU, Mol. Pharmacol. 44 8-12 (1993). 5. J.-B.WANG, Y. MEI, C.M. EPPLER, P. GREGOR, C.E. SPIVAK and G.R. UHL, Proc. Nat. Acad. Sci. USA 90 10230-10234 (1993). 6. R.C. THOMPSON, A. MANSOUR, H. AKIL and S.J. WATSON, Neuron 11 1-20 (1993). 7. F. MENG, G.-X. XIE, R.C. THOMPSON, A. MANSOUR, A. GOLDSTEIN, S.J. WATSON and H. AKIL, Proc. Nat. Acad. Sci. USA, in press. 8. M. MINAMI, T. TOYA, Y. KATAO, K. MAEKAWA, S. NAKAMURA, T. ONOGI, O. KANEK and M. SATOH, FEBS Lett. 329 291-295 (1993). 9. K. YASUDA, K. RAYNOR, H. KONG, C.D. BREDER, J. TAKEDA, T. REISINE and G.I. BELL, Proc. Nat. Acad. Sci. USA 90 6736-6740 (1993) 10. C.A. STEIN,. Leukemia 6 967-974 (1992)~ 11. H.H. WEINTRAUB, Scientific American January_40-46 (1990). 12. I. HOLOPAINEN and W.J. WOJCIK, J. Pharmacol. Exp. Ther. 264 423-430 (1993). 13. C. WAHLESTEDT, E.M. PICH, G.F. KOOB, F. YEE AND M. HELLIG, Science 259 528531 (1993). 14. C. WAHLESTEDT, E. GOLANOV, S. YAMAMOTO, F. YEE, H. ERICSON, H. YOO, C.E. INTURRISI AND D.J. REIS, Nature 363 260-263 (1993). 15. K.M. STANDIFER, C.-C. CHIEN, C. WAHLESTEDT and G.W. PASTERNAK, Soc. Neurosci. 19 37.13 16. K.M. STANDIFER, C.-C. CHIEN, C. WAHLESTEDT, G.P. BROWN and G.W. PASTERNAK, Neuron, in press. 17. G. UHL, S.R. CHILDERS and G.W. PASTERNAK, Trends in Neurosci., in press 18. C.-C. CHIEN, G. BROWN, Y.-X. PAN and G. W. PASTERNAK, Eur. J. Pharmacol., in press. 19. R.J. BODNA1L C.W. WILLIAMS, S.J. LEE and G.W. PASTERNAK, Brain Res. 447 25-37 (1988). 20. G.C. ROSSI, G.W. PASTERNAK and R.J. BODNAR, Brain Res. 624 171-180 (1993). 21. R.J. BODNAR, D. PAUL, M. ROSENBLUM, L., LIU and G.W. PASTERNAK, Brain Res. 529 324-328 (1990).
Pergamon
PHARMACOLOGY LETTERS Accelerated Communication
B L O C K A D E O F M O R P H I N E A N A L G E S I A BY A N A N T I S E N S E OLIGODEOXYNUCLEOTIDE AGAINST THE MU RECEPTOR Grace Rossi, Ying-Xian Pan, Jie Cheng and Gavril W. Pasternak The Cotzias Laboratory of Neuro-Oncoiogy Memorial Sloan-Kettering Cancer Center and Departments of Neurology & Neuroscience and Pharmacology Cornell University Medical College New York, NY 10021 (Submitted January 20, 1993; accepted February 2, 1994; received in finalform March 9, 1994)
Abstract: The recent cloning of mu, delta and kappa~ opioid receptors has provided opportunities in the study of their pharmacology. Using an antisense strategy developed against delta and kappa~ opioid receptors, we designed an antisense oligodeoxynucleotide directed against the 5'-untranslated region ofMOR-1 clone, 51-70 bp upstream from the initiating ATG. Microinjection of this antisense oligodeoxynucleotide directly into the periaqueductal gray on Days 1, 3 and 5 completely blocked the analgesic actions of morphine administered into the periaqueductal gray on Day 6 (p < 0.001), 24 hr after the last antisense treatment. Rats treated with vehicle or with a mismatch oligodeoxynucleotide in which two pairs of bases from the antisense sequence had been switched were not significantly affected. These findings confirm the pharmacological relevance of the MOR-1 clone and its involvement in morphine's actions. Key Words: morphine, antisense treatment, MOR-1 clone, mu opioid receptor
Introduction
Morphine, the classic opioid analgesic, acts through mu opioid receptors (1). The recent cloning of the delta opioid receptor (2,3) quickly led to the identification of clones encoding mu and kappat receptors (4-9). The binding profiles of these expressed receptors is similar to that expected from traditional binding studies in brain membranes and all the receptors are functionally active when transfected into cells lacking opioid receptors. However, correlating these cloned receptors with pharmacological activity requires additional approaches. Antisense strategies provide a method for selectively downregulating specific mRNA's and their proteins (10,11). This approach has been effectively utilized against G-protein receptors and Corresponding author: Gavril W. Pastemak, M.D., Ph.D., Departmentof Neurology,Memorial SloanKettering Cancer Center, 1275 York Avenue,New York, NY 10021
PL-376
Antisense Oligodeoxynucleotideon Analgesia
Vol. 54, No. 21, 1994
has proven valuable in in vivo studies ofneuropeptide Y1 and NMDA receptor activity (12-14). Prior work from our group has demonstrated that antisense oligodeoxynucleotides against a delta opioid receptor selectively blocks 3H-DPDPE binding in a neuroblastoma cell line and both DPDPE and deitorphin II analgesia in mice (15,16). In binding studies antisense oligodeoxynucleotides directed against five different regions of the open reading frame of the delta receptor show similar decreases in binding, indicating that the location of the antisense is not crucial for activity. Detailed studies with mismatch and sense oligodeoxynucleotides confirmed the selectivity of the decrease in binding. In vivo treatments demonstrated a potent blockade of DPDPE and deltorphin II analgesia without interfering with the analgesic actions of either the mu drug morphine or the kappa~ agent U50,488H. Mismatch oligodeoxynucleotides were inactive. Equally important, analgesic sensitivity returned to normal over a period of 5 days following the discontinuation of antisense treatment, ruling out a nonspecific or toxic effect of the treatments. Similar studies demonstrated that an antisense oligodeoxynucleotide directed against the kappa~ receptor given intrathecally blocked U50,488H analgesia, but not DPDPE or morphine analgesia (17). Again, the selectivity of the action was confirmed by the inactivity of a mismatch oligodeoxynucleotide. The role of the cloned mu receptor in morphine pharmacology has not yet been established. We now report that an antisense oligodeoxynucleotide directed towards the 5'-untranslated region of the mu receptor clone microinjected into the periaqueductal gray blocks morphine analgesia. Materials and Methods
Male Sprague-Dawley rats (250 g) were purchased from Charles River (Wilmington, MA). Oligodeoxynucleotides were synthesized by Midland Certified Reagent Co. (Midland, TX). The antisense oligodeoxynucleotide (Table 1) is directed against the 5'-untranslated region (bases -87 to -69 upstream from the initiating ATG). The mismatch oligodeoxynucleotide corresponds to the antisense sequence except for two pairs of bases which have been switched. Rats were cannulated in the periaqueductal gray, as previously described (18-20). Histological studies at the conclusion of the study confirmed the placement. The animals were divided into three groups and all groups received a test dose of morphine to ensure proper cannulae placement. The groups received vehicle or either the antisense or mismatch oligodeoxynucleotides (10 lag in 1 lal) on Days 1, 3 and 5. Twenty-four hours after the last treatment, the animals were tested with morphine. Analgesia was assessed in a graded manner using the tailflick assay (18-20). Results and Discussion
Morphine is a potent analgesic when administered into the periaqueductal gray of rats (18-20; Fig. 1). in the current study, morphine administered prior to any treatment (Control) is analgesic in all three groups. Following the three treatment injections, there is a small decrease in morphine's analgesic potency in the vehicle group (Fig. la), possibly reflecting the effects of multiple injections. The administration of the mismatch oligodeoxynucleotide has little effect on morphine anlagesia (Fig. lb). The small decrease in analgesic effect is quite similar to that observed with vehicle injections. In contrast, treatment with the antisense oligodeoxynucleotide eliminates the analgesic actions of morphine, lowering the tailflick latencies to baseline values (Fig. Ic; p < 0.001).
Vol. 54, No. 21, 1994
Antisense Oligodeoxynudeotideon Analgesia
PL-377
A S
~ntrol
~
,
C~ntrol, 6
fi 0
30
t
t
60 T i m e (rain)
~
I
i 90
i
30
120
~ntroi
8
I
i
60 T i m e (rain)
i
: 90
120
.
6
44 Ba~line 30
L 60 T i m e (mill)
L 90
120
Fig. 1 Effects of vehicle, antisense and mismatch treatments on morphine analgesia: Rats received periaqueductal gray cannulae and were allowed to recover from surgery for at least one week. Animals were divided into three groups and we assessed morphine analgesia (2.5 ~ag in 1 lal) in each animal to confirm the cannulae placement pharmacologically and to establish the morphine sentivity in each group (Control). After waiting another week, the groups then were treated on Days 1, 3 and 5 with either A) vehicle (n=ll), B) mismatch (10 lag, n=5) or C) antisense (10 lag, n=10) oligodeoxynucleotides. Morphine then was injected on Day 6 and analgesia compared to the control curve from the same group. Control and treatment testing do not differ significantly for either the vehicle or the mismatch group, comparing either peak effects (30 min) or areas under the curves. Antisense treatment significantly reduces morphine analgesia, looking at either peak effects (p < 0.00l) or areas under the curves (p < 0.001). Our current study dearly demonstrates the ability of an antisense oligodeoxynucleotide to the mu receptor to interfere with morphine analgesia. The inability of the mismatch oligodeoxynucleotide to influence morphine anlagesia confirms the specificity of the response. We chose to examine these actions in the periaqueductal gray for its sensitivity to morphine and the well-localized site of action. We previously utilized a similar approach for the study of pertussis and cholera toxins and morphine analgesia (20). Intracerebral injections also avoid the dilution and rapid clearance associated with intracerebroventricular administration.
PL-378
Antisense Oligodeoxynudeotideon Analgesia
Vol. 54, No. 21, 1994
TABLE 1 Oligodeoxynucleotide sequences
Oligodeoxynucleotide (5' to 3') Sense
AGA GGA AGA C_rGCTGG GGC G
Antisense
CGC CCC AGC CTC TTC CTC T
Mismatch
CGC CCC GAC CTC TTC CCT T
Prior work in mice has shown that antisense oligodeoxynucleotides directed against the open reading frames of delta or kappal receptors selectively block the analgesic actions of their respective ligands, but not morphine (15-18). Our binding studies indicate that the region of the open reading frame targeted is not important. Antisense oligodeoxynucleotides directed at five different regions of the open reading frame ofDOR-1 downregulate delta binding in the NGI08-15 cells to the same extent. In our current study we targeted the antisense oligodeoxynucleotide to the 5'-untranslated region, approximately 50 bp upstream from the initiating ATG. Untranslated regions are specific for the gene in question and do not show the homology typically seen in the open reading frames among related gene products. Furthermore, Southern analysis appears to indicate the presence of only a single MOR-1 gene (17). Thus, the downregulation of morphine analgesia by our antisense oligodeoxynucleotideimplies that the gene encoding the MOR-1 clone is involved with mu analgesia. However, some questions still remain. Mu~ receptors mediate morphine analgesia in the periaqueductal gray (18-20), suggesting that the MOR-1 clone might correspond to the mu I receptor. Mounting evidence suggests the presence of alternative splicing in the processing of opioid receptor clones and it is possible that subtypes might reflect alternative splicing of a single gene. These alternatively spliced mRNA species might contain the same 5'-untranslated sequence as MOR-1 and also would be downregulated. Although the results using the antisense oligodeoxynucleotide directed toward the region indicate that the MOR-1 clone derives from the gene encoding mu~ receptors, we do not yet know whether the MOR-1 clone represents the mu] receptor or an alternatively spliced variant of the same gene.
Acknowled2ments We thank Dr. J. Posner for his support of this research and Drs. C. Wahlestedt and H. Furneaux for helpful discussions. This work was supported, in part, by grants from NIDA to GWP (DA02615 and DA07242) and a core grant from the NCI to MSKCC (CA08748). GWP is supported by an RSDA from NIDA (DA000138), GR by a Training Grant from NIDA (DA07274) and YXP by a Fellowship from the Aaron Diamond Foundation.
References 1. G.W. PASTERNAK, Clin. Neuropharmacol. 16 1-18, (1993). 2. C.J. EVANS, D.E. KEITH, JR., H. MORRISON, K. MAGENDZO and R.H. EDWARDS, Science 258 1952-1955 (1992). 3. B.L. KIEFFER, K. BEFORT, C. GAVERIAUX-RUFF and C.G. HIRTH, Proc. Nat. Acad. Sci. USA 8__9912048-12052 (1992)
Vol. 54, No. 21, 1994
Antlsense Oligodeoxynucleotideon Analgesia
PL-379
4. Y. CHEN, A. MESTEK, J. LIU, J.A. HURLEY and L. YU, Mol. Pharmacol. 44 8-12 (1993). 5. J.-B.WANG, Y. MEI, C.M. EPPLER, P. GREGOR, C.E. SPIVAK and G.R. UHL, Proc. Nat. Acad. Sci. USA 90 10230-10234 (1993). 6. R.C. THOMPSON, A. MANSOUR, H. AKIL and S.J. WATSON, Neuron 11 1-20 (1993). 7. F. MENG, G.-X. XIE, R.C. THOMPSON, A. MANSOUR, A. GOLDSTEIN, S.J. WATSON and H. AKIL, Proc. Nat. Acad. Sci. USA, in press. 8. M. MINAMI, T. TOYA, Y. KATAO, K. MAEKAWA, S. NAKAMURA, T. ONOGI, O. KANEK and M. SATOH, FEBS Lett. 329 291-295 (1993). 9. K. YASUDA, K. RAYNOR, H. KONG, C.D. BREDER, J. TAKEDA, T. REISINE and G.I. BELL, Proc. Nat. Acad. Sci. USA 90 6736-6740 (1993) 10. C.A. STEIN,. Leukemia 6 967-974 (1992)~ 11. H.H. WEINTRAUB, Scientific American January_40-46 (1990). 12. I. HOLOPAINEN and W.J. WOJCIK, J. Pharmacol. Exp. Ther. 264 423-430 (1993). 13. C. WAHLESTEDT, E.M. PICH, G.F. KOOB, F. YEE AND M. HELLIG, Science 259 528531 (1993). 14. C. WAHLESTEDT, E. GOLANOV, S. YAMAMOTO, F. YEE, H. ERICSON, H. YOO, C.E. INTURRISI AND D.J. REIS, Nature 363 260-263 (1993). 15. K.M. STANDIFER, C.-C. CHIEN, C. WAHLESTEDT and G.W. PASTERNAK, Soc. Neurosci. 19 37.13 16. K.M. STANDIFER, C.-C. CHIEN, C. WAHLESTEDT, G.P. BROWN and G.W. PASTERNAK, Neuron, in press. 17. G. UHL, S.R. CHILDERS and G.W. PASTERNAK, Trends in Neurosci., in press 18. C.-C. CHIEN, G. BROWN, Y.-X. PAN and G. W. PASTERNAK, Eur. J. Pharmacol., in press. 19. R.J. BODNA1L C.W. WILLIAMS, S.J. LEE and G.W. PASTERNAK, Brain Res. 447 25-37 (1988). 20. G.C. ROSSI, G.W. PASTERNAK and R.J. BODNAR, Brain Res. 624 171-180 (1993). 21. R.J. BODNAR, D. PAUL, M. ROSENBLUM, L., LIU and G.W. PASTERNAK, Brain Res. 529 324-328 (1990).