Interaction of Ligands with the Opiate Receptors of Brain Membranes: Regulation by Ions and Nucleotides [GMP-P(NH)P/Agonists/Antagonistsi ARTHUR J

Proc. Nati. Acad. Sci. USA Vol. 75, No. 4, pp. 1713-1717, April 1978 Biochemistry

Interaction of ligands with the opiate receptors of brain membranes: Regulation by ions and nucleotides [GMP-P(NH)P/agonists/antagonistsI ARTHUR J. BLUME Department of Physiological Chemistry and Pharmacology, Roche Institute of Molecular Biology, Nutley, New Jersey 07110 Communicated by B. L. Horecker, January 27,1978

ABSTRACT This study shows that nucleotides, as well as this desensitization phenomenon is a result of an alteration in ions, regulate the opiate receptors of brain. GMP-P(NH)P and the agonist-receptor-adenylate cyclase complex which can be Na+ reduce the amount of steady-state specific [3H]dihydromorphine binding and increase the rate of dissociation of the reversed by guanine nucleotides. ligand from the opiate receptor. In contrast, Mn2+ decreases the The data presented in this paper show that guanine nucleo- rate of ligand dissociation and antagonizes the ability of Na+ tides, as well as the ions, Mn2+ and Na+, do regulate brain opiate to increase dissociation. The effects of GMP-P(NH)P on receptors. A similar regulation of opiate receptors by nucleotides steady-state binding and dissociation are not reversed by has recently been observed in the neuroblastoma X glioma washing. Only GTP, GDP, ITP, and IMP-P(NH)P, in addition hybrid cell line NG108-15 (unpublished observations). The to GMP-P(NH)P, increase the rate of dihydromorphine dissociation. The site of nucleotide action appears to have high af- similarities and differences, between nucleotide regulation of finity: <1 &M GMP-P(NH)P produces half-maximal increases opiate receptors compared to the other plasma membrane re- in ligand dissociation. GMP-P(NH)P- and Na+-directed increases ceptors known to be coupled to adenylate cyclase are dis- in dissociation have also been found for the opiate agonists cussed. [3Hketorphine, [3HJLeu-enkephalin, and [3HJMet-enkephalin and the opiate antagonist [3H]naltrexone. Mn2+-directed de- MATERIALS AND METHODS creases in dissociation have been found for the agonist [3H- etorphine and the antagonist [3H]naltrexone. Although the [3H]Dihydromorphine (70.6 Ci/mmol), [3H]etorphine (31 plasma membrane receptors for a number of other neuro- Ci/mmol), and [3H]Leu-enkephalin (45.6 Ci/mmol) were transmitters and hormones are also regulated by guanine nu- obtained from Amersham/Searle, Des Plaines, IL; [3H]Met- cleotides, the opiate receptors appear unique because only they was New show nucleotide regulation of both agonist and antagonist enkephalin (16 Ci/mmol) from England Nuclear, binding. Boston, MA. Naloxone-HCI was obtained from Endo Labs., Garden City, NY; levallorphan was a gift from Hoffman-La Opiate agonists regulate the adenylate cyclase [ATP pyro- Roche Inc., Nutley, NJ. The nucleotides, GMP-P(NH)P, phosphate-lyase (cyclizing); EC 4.6.1.1] in neuroblastoma X AMP-P(NH)P, CMP-P(NH)P, UMP-P(NH)P, and IMP- glioma hybrid cell line NG108-15 (1-3). No clear evidence of P(NH)P, were purchased from ICN Pharmaceuticals, Irvine, opiate regulation of the adenylate cyclase of brain tissue is CA. All other nucleotides were obtained from Sigma Chemical available. However, Klee and Nirenberg (2) have in fact sug- Co., St. Louis, MO. gested that tolerance to opiates in human beings and animals Male Wistar rats (Marland Farms, Hewitt, NJ) weighing is a result of regulation of adenylate cyclase by opiates. As part 250-50 g were decapitated and their brains were rapidly re- of the search for some evidence for a "coupling" between the moved. The cerebellums were then excised, and the remainder opiate receptors of brain and its adenylate cyclase, I have in- was placed in 10 volumes of iced 50 mM Tris-HCl, pH 7.4, and vestigated the effects of nucleotides on opiate binding to rat homogenized (twice for 10 sec) in a Brinkman Polytron at 40. brain membranes. These studies were initiated since it is known The homogenates were then centrifuged at 27,000 X g for 30 that glucagon (4-6), prostaglandin E1 (7), and ,B-adrenergic (8, min at 40; the resulting pellet was washed twice with buffer and 9) receptors, which regulate adenylate cyclases, are all regulated finally suspended in 50 mM TrissHCI, pH 7.4, (2-5 mg of themselves by nucleotides. For these three classes of agonists, protein per ml). This preparation (referred to as the "Tris- agonist occupation of its receptor in the presence of guanine washed" membranes) was made 25 mM NaCl, incubated at 370 nucleotide (10-12) leads to increased adenylate cyclase activity, for 30 min, and centrifuged at 27,000 X g; the pellet was yet the presence of guanine nucleotides selectively decreases washed twice at 40 with 50 mM Tris-HCl, pH 7.4. These pellets agonist but not antagonist affinity and increases agonist disso- were then suspended in 50 mM Tris-HCI, pH 7.4, (2-5 mg ciation from the receptor. In addition, the occupation of the protein per ml) and are referred to as "Na+-washed" mem- prostaglandin E1 and f,-adrenergic receptors by only agonists branes. Membrane preparations were frozen and stored under for extended periods of time leads to a condition (termed "de- N2 until they were used. sensitization") in which agonist no longer can activate the Binding assay reactions contained 200-400 l of membranes, adenylate cyclase complex* to the same extent as they did be- 50 mM Tris-HCI (pH 7.4), [3H]ligand with or without 10 gM fore the long-term exposure to the agonist (7, 9, 13, 14). naloxone or levallorphan, and other additions (as noted in the Lefkowitz and coworkers (7, 9) have recently suggested that text) in a final volume of 0.5 ml. Reactions were terminated by addition of 2 ml of standard buffer (room temperature) and The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked * The components responsible for the catalytic activity and the binding "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate of guanine nucleotides are designated as the adenylate cyclase this fact. complex only for reasons of simplicity. 1713 Downloaded by guest on October 4, 2021 1714 Biochemistry: Blume Proc. Natl. Acad. Sci. USA 75 (1978)

c 100 ._ c 4 a, 80 80 10 A BI o 0 I 60 60 a, 0 %I- ~0 0E ; 40 C :s la 0 10- 0 m .0 .0)C 0 2 I 20 0-0E 10 10 .~ X a, W0 1 2 3 4 0 20 40 60 80 O -' A,, a0. D) [3H] Dihydromorphine, [3H] Dihydromorphine nM bound FIG. 1. Tris-washed membranes were incubated with various concentrations of [3H]dihydromorphine with the following additions:

0, none; 0, 25 mM NaCl; *, 10 MM GMP-P(NH)P; *, NaCl and GMP-P(NH)P. (A) Plot of the specific [3H]dihydromorphine bound 2 4 6 8 10 0 2 4 6 8 K0 (i.e., difference ± 10MgM levallorphan). (B) Scatchard plot for the data Time, min Time, min in panel A. Ordinate, fmol of dihydromorphine bound per mg of protein/[3H]dihydromorphine in nM; abscissa, fmol of dihydromor- FIG. 2. Time course of dissociation of [3Hjdihydromorphine (A) or (B). Na+-washed membranes were incubated with phine bound per mg of protein. The lines drawn were obtained by [3H]naltrexone either 2.8 nM or 4.8 nM for 20 unweighted linear regression analysis and have a correlation coeffi- [3H]dihydromorphine [3H]naltrexone cient of 20.9. min at 32°. At this point (t = 0) 10MgM levallorphan and 25 mM NaCl, or 10 WM GMP-P(NH)P, or 1 mM MnCl2, or NaCl and GMP- P(NH)P, or NaCl and MnCl2 were added and [3H]ligand binding was followed with time. Binding plotted on the ordinate refers only to filtration over Whatman GF/B filters. Filters were washed specific binding, with that seen at t = 0 taken as 100%. Specific three times with 3 ml of standard buffer (<10 sec), dried, and binding at t = 0 of [3H]dihydromorphine and of [3H]naltrexone is 74 suspended in 10 ml of Aquasol; radioactivity was determined and 163 fmol/mg of protein, respectively. 0, No additions; ,, MnCl2; 0, NaCl; *, GMP-P(NH)P; *, NaCl and GMP-P(NH)P; v, MnCl2 at 37% counting efficiency. The binding data presented rep- and NaCl. resent only the observed specific binding of the [3H]ligand (as defined by the difference :E10 MiM nonradioactive naloxone or levallorphan). For the experiments monitoring the dissociation The presence of two different [3Hjdihydromorphine binding of [3Hlligand, binding was first allowed to proceed for 20 min sites in rat brain membranes is further supported by the ob- at 320 under the above conditions. This time was then taken as servation of the biphasic nature of the dissociation of bound t = 0. Levallorphan (10 AM) and the compounds under study dihydromorphine (Fig. 2). After a 20-min incubation in Tris (noted in text) were then added at t = 0 without significant buffer, although the major amount of the bound [8H]dihy- dilution of the reaction mixture and the amount of [3H]ligand dromorphine dissociates slowly with a t1/2 of 4.5 min, a sig- amount can seen to remaining bound was followed with time (at 320) by with- nificant (<40%) be dissociate faster with drawing aliquots and filtering them. Similar additions of test a t1/2 of <1 min. GMP-P(NH)P and Na+ (and Mn2+ as well) compounds were made at t = 0 to the tubes that had contained influence the dissociation of dihydromorphine in a complex 10MuM levallorphan during the initial 20-min period. Binding manner. Both GMP-P(NH)P and Na+ added separately cause data refer only to the specific binding (i.e., difference 4 le- the major portion (>60%) of the bound dihydromorphine to vallorphan) seen at each successive time point and are given as dissociate rapidly now with a t1/2 of <0.2 min. The remaining 4-5 mi. a percentage of the specific binding observed at t = 0. dihydromorphine still dissociates with a slower t1/2 of As already seen with steady-state binding, GMP-P(NH)P and Na+ produce a larger effect on dissociation when added together than when added separately. In the presence of both RESULTS GMP-P(NH)P and Na+, almost all (285%) of the bound When the steady-state binding of [3H]dihydromorphine to rat dihydromorphine is converted into a form that dissociates very brain membranes is followed over [3H]dihydromorphine con- rapidly with a t1/2 of <0.2 min. The effects of Mn2+ on dihy- centrations from 0.1 to 4.6 nM, 50 MiM GMP-P(NH)P alone or dromorphine dissociation are opposite to those produced by 25 mM NaCi alone reduces the amount of specific binding by Na+ or GMP-P(NH)P. In the presence of 1 mM Mn2+, although approximately one-half (Fig. 1). Only one-third of control dissociation is still biphasic, most of the bound dihydromorphine binding is observed over this range of [3H]dihydromorphine (i.e., 2 70%) now dissociates with the very slow t1/2 of 30 min. when GMP-P(NH)P and NaCl are both present. Analysis of The remaining amount of dihydromorphine dissociates with these data by Scatchard plots reveals that in the Tris buffer there a t1/2 of 4 min and no dihydromorphine dissociates rapidly with are two discernible binding components for [3H]dihydromor- a t1/2 Of <1 min. phine contributing to the total number of dihydromorphine The potency of GMP-P(NH)P to increase the dissociation of binding sites of 76 fmol/mg of protein. One class exhibits high [3H]dihydromorphine was investigated; a concentration of affinity (Kd < 0.3 nM) and makes up less than half of the total between 0.7 and 1 MM appears to produce a half-maximal in- sites, while the major class of sites exhibits a much lower affinity crease in the dissociation of dihydromorphine (Fig. 3). A test (Kd 2 3.5 nM). However, in the presence of GMP-P(NH)P or of the specificity of the nucleotide effect indicates that at 50 NaCl, only a single class of low-affinity sites (Kd > 3.6 nM) is ,MM the only ones that increase [3H]dihydromorphine disso- observable and the maximum number of these components is ciation besides GMP-P(NH)P are GTP, GDP, ITP, and IMP- now only 44 fmol/mg of protein. The combined presence of P(NH)P (Table 1). None of the adenine nucleotides, be they both GMP-P(NH)P and NaCl leads to an even greater reduction mono-, di-, or triphosphates or even AMP-P(NH)P, influence in the maximum number of sites to 28 fmol/mg of protein and dihydromorphine dissociation. GMP is also without effect. these too are only of low affinity (Kd 2 3.7 nM). Therefore, although not identical, the specificity and affinity Downloaded by guest on October 4, 2021 Biochemistry: Blume Proc. Natl. Acad. Sci. USA 75 (1978) 1715

E 3000 Table 2. "Irreversible" nature of the effect of GMP-P(NH)P on the interaction of [3H]dihydromorphine and opiate receptor C [3H]Dihydro- Dissociation t1/2 for 0 morphine* conditions dissociation, min *'-o 2000\- Preincubation with Mn2+ 20 109 ± 15 No additions 4.8 (80%) + .0.3 (20%) E ONNIO GMP-P(NH)P 4.5 (50%) + <0.3 (50%) Na+ 3.6 (50%) + <0.3 (50%) *E 1000 Na+ + GMP-P(NH)P 1.4 (44%) + <0.3 (56%) 0 Preincubation with Mn2+ + GMP-P(NH)P 52 ± 8 No additions 4.5 (100%) Na+ 1.4 (10096) Cn ° o -7 10- 6 10- S 0-44 Mn2+ 8.0 (100%) GMP-P(NH)P, M Na+-washed membranes were incubated with 1 mM MnCl2 or 10 FIG. 3. Increased dissociation of [3H]dihydromorphine producecd MM GMP-P(NH)P plus MnCl2 for 20 min at 320, then extensively by GMP-P(NH)P. Na+-washed membranes were incubated with 2. 8 washed at 40 with Tris-HCl and resuspended in Tris-HCl. Specific nM [3H]dihydromorphine + levallorphan (10juM) for 20 min at 32°). [3H]dihydromorphine binding was then determined at 320 with 5 nM Levallorphan (10 gM) and various concentrations of GMP-P(NH)Fp [3H]dihydromorphine : levallorphan (10 MM) and an incubation time were added and the incubation was continued at 32° for an additionaII of20 min. Dissociation ofthe [3H]dihydromorphine bound at this time 1.5 min. At this point the amount of specific [3H]dihydromorphince was then followed. Additions during dissociation: 10 uM GMP- bound (i.e., difference in cpm ± levallorphan) was determined. Or- P(NH)P, 25 mM NaCl, and 1 mM MnCl2. The times for half of the dinate, specific [3H]dihydromorphine still bound at t = 1.5 min irin specifically bound [3H]dihydromorphine to dissociate are given (t1/2) cpm. along with the percentage of ligand dissociating in that manner. The actual dissociation curves are shown in Fig. 4. * fmol bound/mg of protein + SD. of the nucleotide site in rat brain membranes that influencee dihydromorphine binding resembles that which hars been de- decreased binding could be a simple reflection of the previously scribed for the guanine nucleotide site in plasma membranes observed reduction in binding produced under steady-state that regulate the activity of adenylate cyclase (10, 15).- If thesee conditions by the presence of GMP-P(NH)P (see Fig. 1). two nucleotide sites are the same or closely related, GMP- Dissociation of dihydromorphine from these Mn2+-treated P(NH)P would be expected to produce an "irreversible" effe~ctt membranes (Table 2 and Fig. 4) resembles that previously on the specific binding of dihydromorphine, since GMP- shown (Fig. 2A) for membranes that had not been treated with P(NH)P produces an "irreversible" activation of adenylate Mn2+. Both are biphasic: the major component dissociates with cyclase (15). When the opiate receptors in rat brain membraness a t1/2 of 4.5 min and the minor component dissociates with a incubated at 32° for 30 min with Mn2+ are compared with t1/2 of <0.3 min. GMP-P(NH)P addition increases to about 50% those treated with GMP-P(NH)P and Mn +, the GMP that amount dissociating rapidly (i.e., t1/2 < 0.3 min) and the P(NH)P-treated membranes exhibit only half as much specific copresence of Na+ and GMP-P(NH)P also induces the re- binding of [3H]dihydromorphine as seen with the membranes maining 50% to dissociate faster. In contrast, there is only one that have not been treated with GMP-P(NH)P when both are dissociation component observable from the membranes treated assayed in Tris buffer after extensive washing (Table 2). This with Mn2+ plus GMP-P(NH)P. In Tris buffer, dissociation occurs slowly with a t1/2 of 4.5 min and when Na+ is added, dis- Table 1. Specificity of the nucleotide effect on dissociation of sociation occurs with the faster t1/2 of 1.4 min. From these ex- [3H]dihydromorphine periments, it appears that after GMP-P(NH)P treatment % control [3H]dihydromorphine binding to approximately half of the Nucleotide added [3H]dihydromorphine bound I SD* membrane receptor sites is "irreversibly" prevented. Addi- tionally, these "lost" sites seem to correspond to those in control None 100 ± 6 membranes from which dihydromorphine dissociates rapidly ATP 99:1:5 (i.e., with a t1/2 of <0.3 min) when either Na+ or GMP-P(NH)P ADP 105+8 or both are added. Those sites remaining after GMP-P(NH)P AMP 99: 2 treatment are still affected by Na+ (an increase in the rate of AMP-P(NH)P 100:1 5 and Mn2+ UMP-P(NH)P 101 i 2 dissociations (a slowing of the dissociation rate). CMP-P(NH)P 99:+ 4 Therefore, the above data do indicate that the action of GMP 9713 GMP-P(NH)P with its nucleotide site leads to an "irreversible" IMP-P(NH)P 62 + 4 alteration in dihydromorphine binding sites. GMP-P(NH)P 37 1:3 In order to investigate the general nature of these nucleotide GTP 34±4 and ion effects on the opiate receptor, dissociation of three other GDP 29:1 5 opiate agonists (ie., etorphine, Leu-enkephalin, and Met- enkephalin) and the opiate antagonist, naltrexone, have also After the incubation of Na+-washed membranes for 20 min with been examined (Table 3). Although [3H]etorphine dissociates 2.8 nM [3H]dihydromorphine (at 320), various nucleotides at 50 gM were added along with 10 ,gM levallorphan and dissociation was al- much more slowly than [3H]dihydromorphine, etorphine dis- lowed to proceed at 320 for 1.5 min (t = 1.5). sociation is slowed by Mn2 , sped up by Na+ or GMP-P(NH)P, * The amount of [3H]dihydromorphine specifically bound at t = 1.5 and greater when both Na+ and GMP-P(NH)P are present than in the absence of added nucleotides (control) was 56.9 fmol/mg of when either is present alone. The two peptide opiate agonists, protein. Leu- and Met-enkephalin, are also induced to dissociate faster Downloaded by guest on October 4, 2021 1716 Biochemistry: Blume Proc. Natl. Acad. Sci. USA 75 (1978)

ways monophasic with either Na+- or Tris-washed membranes. This is not the case for dihydromorphine (see Figs. 2 and 3) nor always for etorphine or Leu- and Met-enkephalin. Whether or not these last three agonists dissociate in a mono- or biphasic manner in Tris buffer depends upon whether the membranes are Na+- or Tris-washed (unpublished data). However, all of .0 the [3H]agonist ligands dissociate in a biphasic manner when C 20 Na+ or GMP-P(NH)P or both are present.

E DISCUSSION 10 The data presented in this paper clearly demonstrate that nucleotides are essential regulators of opiate receptors in brain. Previous work on the possible existence of effectors of opiate

Z.4 receptors had been limited to studies dealing primarily with ions (16-20). These data had led to the hypothesis that Na+ ions induce an opiate receptor conformation favorable to antagonist 2 binding and Mn2+ ions induce one favorable to an agonist binding. This paper shows that as little as 50,uM GMP-P(NH)P decreases by half the specific binding of [3H]dihydromorphine. a at 25 mM. 0 2 4 6 8 10 Sodium has similar affect when tested The effects Time, min of GMP-P(NH)P are like those of Na+, and appear to be on both the affinity and number of dihydromorphine binding sites. As FIG. 4. Dissociation of [3H]dihydromorphine from membranes reported by others (21, 22) and confirmed here' steady-state that had been treated with Mn2+ (open symbols) or GMP-P(NH)P binding in Tris buffer indicates the existence of two dihydro- and Mn2+ (closed symbols) as described in the legend to Table 2. morphine binding components in rat brain membranes: one of One-hundred percent specific [3H]dihydromorphine bound refers to low affinity (Kd 3.5 nM) and one of high affinity (kd 0.3 that seen at t = 0 with the Mn2+-treated membranes; all binding is given as a percentage of this amount. For Mn2+-treated membranes: nM). GMP-P(NH)P, like Na+, eliminates the class of high- dissociation with no additions (0), with 10MuM GMP-P(NH)P (a), affinity sites and reduces the number of total binding sites. The and with 25 mM NaCl and GMP-P(NH)P (v). For membranes reduction in the number of sites is greatest when both GMP- treated with Mih2+ plus GMP-P(NH)P: dissociation with no additions P(NH)P and Na+ are added together. (A), with 25 mM NaCI (v), and with 1 mM MnCl2 (-). This nucleotide effect on dihydromorphine binding was further characterized by investigating dissociation of [3H]- by Na+ or GMP-P(NH)P. In contrast to the reaction see with dihydromorphine with and without nucleotides. Dissociation etorphine and dihydromorphine, Na+ appears more potent of [3Hjdihydromorphine from rat brain membranes is biphasic, than GMP-P(NH)P with these peptide opiates and no extra in support of steady-state binding data indicating the existence effect over that seen with only Na+ is noticed when both of two classes of dihydromorphine sites. The effects of GMP- GMP-P(NH)P and Na+ are present. What is also clear from the P(NH)P and Na+ on dissociation were similar. Although data in Table 3 is that dissociation of the opiate antagonist, complex in nature, in general both GMP-P(NH)P and Na+ naltrexone, is regulated much like that of the agonists dihy- increase, in a population of previously formed receptor-dihydromorphine and etorphine. A detailed illustration of the dis- dromorphine complexes, the percentage that dissociates rapidly. sociation of [3H]naltrexone is given in Fig. 2B. GMP-P(NH)P However, both increases in the amount of dihydromorphine speeds up naltrexone dissociation, as does Na+. GMP-P(NH)P dissociating rapidly and the actual rate of dihydromorphine appears to be the more potent of the two and together with Na+ dissociation have been observed. Although the increases in produces the largest increase in naltrexone dissociation (which dihydromorphine dissociation and decreases in steady-state is 34 times faster than in controls). In addition, Mn2+ slows the binding produced by both Na+ and GMP-P(NH)P are in gross dissociation of naltrexone and antagonizes the effects of Na+. agreement, additional information is needed on association rates Under all these conditions the dissociation of naltrexone is al- before a complete understanding of the effects of GMP-P(NH)P

Table 3. Ion and nucleotide effects on dissociation of opiate agonists and antagonists Additions made at % initial [3H~ligand bound* t = 0 Dihydromorphine Etorphine Met-enkephalin Leu-enkephalin Naltrexone Mn2+ 76+ 4 95 ± 5 ND ND 70± 2 Mn2+ + Na+ ND 87 ± 3 ND ND 62± 2 None 53 7 77 : 8 81 90 54 ±6 Na+ 29 1 59 + 6 50 54 45 + 5 GMP-P(NH)P 11 3 52 + 2 63 80 26±1 Na++GMP-P(NH)P 4±3 30I7 50 54 17±9 Rat brain membranes were incubated at 320 in 50 mM Tris.HCl (pH 7.4) for 20 min with 2.3-3 nM [3H]dihydromorphine, 1.6-6.4 nM [3H]- etorphine, 7.4 nM [3H]Met-enkephalin, 10.3 nM [3H]Leu-enkephalin, and 4.8-17.2 nM [3H]naltrexone, all ± 10MAM levallorphan. At this time (t = 0) 10 WM levallorphan was added to the remaining tubes and dissociation was followed with time at 32°. With the exception of the data for Met- and Leu-enkephalin, the data are averages from experiments after dissociation from both Na+- and Tris-washed membranes. For [3H]opiate peptides, only Tris-washed membranes were used. Specific binding at t = 0, in fmol/mg of protein, was 60-70 (dihydromorphine), 170-380 (etorphine), 85 (Met-enkephalin), 121 (Leu-enkephalin), and 115-270 (naltrexone). ND, not determined. * All data are taken at dissociation time of 2 min with the exception of [3H]etorphine, for which dissociation was for 10 min. Downloaded by guest on October 4, 2021 Biochemistry: Blume Proc. Natl. Acad. Sci. USA 75 (1978) 1717 on the-opiate receptor is possible. As with the Nat and Mn2+ covered that nucleotides also regulate the opiate receptors of ions, these effects of GMP-P(NH)P are not limited to only the neuroblastoma X glioma hybrid NG108-15 cell (unpub- dihydromorphine binding. GMP-P(NH)P- and Na+-directed lished observations), and in these cells opiates inhibit adenylate increases in dissociation have also been found with the opiate cyclase (1-3). A comparison of the opiate receptors in these cells agonists [3H]etorphine, [3H]Leu-enkephalin, [3H]Met-en- with those in brain should help us understand the nature of the kephalin and the opiate antagonist [3H]naltrexone. Mn2+-di- guanine nucleotide component regulating opiate receptors. rected decreases in dissociation have been observed for [3H]- dihydromorphine, [3H]etorphine, and [3H]naltrexone. It I thank Gloria Boone for her excellent technical assistance. therefore appears that nucleotides are as essential regulators of the opiate receptor as are ions. 1. Sharma, S. K., Nirenberg, M. & Klee, W. (1972) Proc. Nati. Acad. The activity of plasma membrane adenylate cyclase is also Sci. USA 72,590-594. regulated by a membrane component that binds guanine nu- 2. Klee, W. A. & Nirenberg, M. (1976) Nature 263,609-612. cleotides with high affinity and selectivity (10). The following 3. Goldstein, A., Cox, B. M., Klee, W. A. & Nirenberg, M. (1977) observations have been made for this site: (a) GTP, GMP- Nature 265,362-363. P(NH)P, and GDP are bound with the best affinity; GMP and 4. Sutherland, E. W. & Rall, T. W. (1960) Pharmacol. Rev. 23, adenine nucleotides are bound with little affinity or not at all 265-299. 5. Birnbaumer, L., Pohl, S. L. & Rodbell, M. (1969) J. Biol. Chem. (10). (b) Whereas GTP and GMP-P(NH)P activate the enzyme, 244,3468-3476. GDP does not but instead blocks activation of the enzyme by 6. Lad, P. M., Welton, A. F. & Rodbell, M. (1977) J. Biol. Chem. GMP-P(NH)P (23); (c) ITP and IMP-P(NH)P are in some in- 252,5942-5946. stances able to activate the enzyme and in others not able to do 7. Lefkowitz, R. J., Mullikin, D., Wood, C. L., Gove, T. B. & Mu- so (24-26). (d) Activation of adenylate cyclase by GMP-P(NH)P kherjee, C. (1977) J. Biol. Chem. 252,5295-5303. is not readily reversed by washing (15). The following obser- 8. Maguire, M. E., Van Arsdale, P. M. & Gilman, A. G. (1976) Mol. vations have-been made for the nucleotide site regulating opiate Pharmacol. 12, 335-339. receptors: (a) The amount of GMP-P(NH)P required to half- 9. Mukherjee, C. & Lefkowitz, R. J. (1976) Proc. Natl. Acad. Sci. maximally increase dihydromorphine dissociation is < 1 uM. USA 73, 1494-1498. (b) GDP as well as GTP, ITP, and at 10. Londos, C., Salomon, Y., Lin, M. C., Harwood, J. P., Schramm, IMP-P(NH)P, all 50,gM, M., Wolff, J. & Rodbell, M. (1974) Proc. Natl. Acad. Sci. 71, are equally able to replace GMP-P(NH)P. (c) The effects of 3087-3090. GMP-P(NH)P on steady-state binding and dissociation are not 11. Maguire, M. E., Ross, E. M. & Gilman, A. G. (1977) in Advances reversed by extensive washing. Therefore, although similar, in Cyclic Nucleotide Research, eds. Greengard, P. & Robison, there are differences between the nucleotide site associated with G. A. (Raven, New York), Vol. 8, pp. 1-84. adenylate cyclases and that associated with opiate receptors. 12. Kimura, N. & Nagata, N. (1977) J. Biol. Chem. 252, 3829- Further data are needed to determine the position in the plasma 3835. membrane (i.e., inner or outer surface) of these two sites, 13. Makman, M. H. (1971) Proc. Natl. Acad. Sci. USA 68, 805- whether these two sites are in fact related in some manner, and 809. whether or not opiates can influence adenylate cyclase through 14. Kebabian, J. W., Zatz, M., Romero, J. A. & Axelrod, J. (1975) Proc. these nucleotide sites. Natl. Acad. Sci. USA 72,3735-3739. 15. Birnbaumer, L., Pohl, S. L. & Rodbell, M. (1971) J. Biol. Chem. Regardless of the final answers to the above questions, it is 246, 1857-1860. important to note that there is also a guanine nucleotide site on 16. Pert, C. B., Pasternak, G. & Snyder, S. H. (1973) Science 182, the plasma membrane that regulates the plasma membrane 1359-1361. receptors for at least three known activators of adenylate cy- 17. Pasternak, G. W., Snowman, A. M. & Snyder, S. H. (1975) Mol. clase. Glucagon, prostaglandin E1, and fl-adrenergic agonists Pharmacol. 11, 735-744. all dissociate faster and have higher Kd values when GMP- 18. Simatov, R., Snowman, A. M. & Snyder, S. H. (1976) Mol. P(NH)P is present (4-9). Nucleotide control of the opiate re- Pharmacol. 12, 977-986. ceptor is, however, different in at least two ways from its control 19. Simatov, R. & Snyder, S. H. (1976) Mol. Pharmacol. 12, 987- over these other receptors: (a) with only the opiate receptor, it 998. is possible to 20. Wong, D. T. & Horng, J. S. (1976) in Tissue Responses to Ad- observe nucleotide regulation of receptor-antag- dictive Drugs, eds. Ford, D. H. & Clouqet, D. H. (Spectrum onist interaction; and (b) GMP-P(NH)P alteration of the glu- Publications, Inc., New York), pp. 391-407. cagon receptor is reversible (27) but that of the opiate receptor 21. Birnbaumer, L., Pohl, S. L. & Rodbell, M. (1971) J. Biol. Chem. appears to be irreversible. 246, 1857-1860. As proposed for the fl-adrenergic and prostaglandin E1 re- 22. Pasternak, G. W. & Snyder, S. H. (1975) Nature 253, 563- ceptors (7-9, 11), only agonist interaction is sensitive to nucle- 565. otide regulation because only agonists can induce a "coupling" 23. Rodbell, M., Lin, M. C. & Salomon, Y. (1974) J. Biol. Chem. 246, between receptor and the guanine nucleotide component as- 59-65. sociated with the adenylate cyclases. The unique sensitivity of 24. Lefkowitz, R. J. (1974) J. Biol. Chem. 249, 6119-6124. the opiate receptor to nucleotide control over antagonist in- 25. Pfenffer, T. & Helmreich, E. J. M. (1975) J. Biol. Chem. 250, teraction could 867-876. indicate that the brain's opiate receptors have 26. Londos, C. & Rodbell, M. (1975) J. Biol. Chem. 250, 3459- already been "coupled" due to the presence of endogenous 3465. opiate agonists or, instead, due to the basic difference between 27. Lad, P. M., Welton, A. F. & Rodbell, M. (1977) J. Biol. Chem. the two nucleotide binding sites involved. I have recently dis- 252,5942-5946. Downloaded by guest on October 4, 2021