Proc. Nat. Acud. Set. USA Vol. 72, No. 8, pp. 30923096, August 1975 Biochemistry Dual regulation of adenylate cyclase accounts for narcotic dependence and tolerance (neuroblastomn X glioma hybrid cells/all culture/memory/synaptic transmission/opiate receptors) o Laboratory of Biochemical Genetics, Natiocal Heart and Lung Icstitute. Bethesda, Maryland 20014; and + Laboratory of General acd Comparative Biochemistry, National Institute of Mental Health. Bethesda, Maryland 20014 Contributed by Marshail Nirenberg, June 10.1975 ABSTRACT Narcotics affect adenylate cyclase [ATP pp rophosphate-lyase (cyclizing), EC 4.6.1.1] in two opposing ways, both mediated by the opiate rece is the readily reversible inhibition of tR tor. The first process e enzyme by narcot- its; the second is a compensatory increase in enzyme activity which is delayed in onset and relatively stable. Late positive regulation of the enzyme counteracts the inhibitory influ- ence of morphine and is responsible for narcotic dependence and tolerance. The coupled inhibitory and positive regulato- ry mechanisms for adenylate c clase provide a means of acti- vating and deactivating neura P cucuits hours after the initial event and thus may play a role in a memory process. Recent observations with neuroblastoma X glioma hybrid cells indicate that the binding of morphine and other opiates to narcotic receptors results in an inhibition of adenylate cy- clase [ATP pyrophosphate-lyase (cyclizing), EC 4.6.1.11 ac- tivity (l-3) and a decrease in CAMP levels in intact cells (1, 2, 4, 5). Similar observations have been made with brain (6, 9) but the heterogeneity of cell types present and apparent lability of the brain enzyme may be responsible for conflict- ing observations (8-10). Both dependence upon opiates and tolerance to these compounds were hypothesized to result from either an increase in the number of molecules of ade- nylate cyclase or a long-lived factor which affects the rates of adenylate cyclase activity or turnover (1). In this report, we describe the results of experiments which were designed to test the hypothesis illustrated diagramatically in Fig. 1. The addition of morphine results in the rapid inhibition of adenylate cyclase activity and a resultant decrease in intra- cellular CAMP levels. Further incubation reveals a second regulatory process which involves a compensatory increase in adenylate cyclase activity termed late positive regulation. The increase in adenylate cyclase activity counteracts the in- hibition of enzyme activity by morphine and CAMP levels are restored to the normal value. Cells now are tolerant to morphine and are also dependent upon the narcotic, since withdrawal of the drug or the addition of a specific narcotic antagonist will raise CAMP levels to abnormally high values and secondarily produce a gradual return to the normal level of adenylate cyclase activity. In this communication we report data which demonstrate a rapid inhibition and a late positive regulation of adenylate cyclase which are dependent upon narcotics and account for the phenomena of narcotic dependence and tolerance. METHODS AND MATERIALS The source of each chemical and the medium and growth conditions for neuroblastoma X glioma hybrid NG108-15 were described previously (1). Abbreviations: Ro 20-1724, 4-(3-butoxy-Pmethoxybenzyl)-2imida- zolidinone; PGEr, prostaglandin El t On leave from the Department of Biochemistry, All India Insti- tute of Medical Sciences, New Delhi, India. Assay of CAMP in Intact Cells and Medium. Growth medium was changed 12 hr before assay. Narcotic was added to the growth medium [Dulbecco's modification of Eagle's medium (DMEM), hypoxanthine-aminopterin- thymidine (HAT), and 10% fetal bovine serum] and plates were incubated in a humidified atmosphere of 90% air-lo% COs. At appropriate times the phosphodiesterase inhibitor 4-(3butoxy-4-methoxybenzyl)-2-imidazolidinone (Ro 20-1724), was added (0.5 mM, final concentration) and incubation was continued for 15 min. Reactions were initiat- ed by the addition of narcotic and/or adenosine in Ha0 or prostaglandin Er (PGEr) in ethanol. Ethanol (0.5% with Ro 20-1724 present, or 1.0% when both Ro 2&1724 and PGEl were present) had no effect upon CAMP formation. Reac- tions were terminated by the addition of 1.0 ml of trichloro- acetic acid at 0" (final concentration, 5%). A solution con- taming 5 pmol of ['*C]cAMP (3000 cpm) was added, and the suspension and two washes (each 1 ml) of 5% trichloro- acetic acid were combined and centrifuged. Each superna- tant fraction was applied to an 0.8 X 8 cm column of AC 5OW-X4 resin, 200-400 mesh, H+ form (Bio-Rad) washed with HsO. Each column was washed with 6 ml of Ha0 and CAMP was eluted with an additional 3 ml of HsO, and ap- plied to a 0.8 X 2.5 cm column of AG l-X8 resin, 200400 ADD MORPHINE WITHDRAW ADD MORPHINE WITHDRAW FIG. 1. A model of the role of adenylate cyclase regulation in the development of morphine tolerance and dependence. Part A shows the effects of morphine upon CAMP levels, and part B, the effects of the opiate upon adenylate cyclaae activity as a function of time. 3092 Biochemistry: Sharma et al. Proc. Nat. Acad. Sd. USA 72 (19%) 3093 I 3oo _ INTACT CELLS 01 HOURS FIG. 2. Levels of CAMP in intact NGlO&15 hybrid cells and the medium measured as a function of time in the presence of 10 rM PG& or 10 aM PGEl and 10 aM morphine sulfate. No phos- phodiesterase inhibitor was present. The average 60 mm petri dish contained 3.5 mg of cell protein. mesh, formate form (Bio-Rad) equilibrated with water. The eluate and a subsequent 10 ml He0 wash were discarded; CAMP was eluted with 4 N HCOOH and lyophilized. Each dried sample was taken up in 0.5 ml of He0 and assayed for CAMP by the method of Gilman (11). Values reported are average values obtained with two to four replicate plates and.are corrected to 100% recovery of CAMP. At each time of incubation 6 to 8 dishes were washed free of serum (1) and the cells were transferred quantitatively to a tube. After centrifugation the pellets were dissolved in 1 N NaOH and protein was estimated by a modification of the method of Lowry et ol. (12). The average value of protein at each time point was used to calculate the specific activity of CAMP. The adenylate cyclase assay and preparation of homoge- nates have been described previously (l), except that Tris. HCl was 30 mM and pH 7.5 in the previous as well as this study. Opiate Receptor Assays. Washed cells from two 100 mm culture dishes were hotiogenized in 10 ml of medium Dl (13). centrifuged for 10 min at 20,000 rpm (48,000 X g), and the pellet was resuspended in the original volume of Dl whose pH had been adjusted to 8.0 with 2 M Tris (free base). Specific binding of [3H]naloxone (5 X 10eQ M, 63,500 cpm/ml) was measured as described (13). Results are ex- pressed in fmol of specifically bound naloxone per mg of protein in the whole homogenate. RESULTS The effect of morphine on PGEI-stimulated CAMP accumu- lation in neuroblastoma X glioma hybrid cells, (NG108-15), and CAMP excreted into the medium are shown in Fig. 2 as a function of time of incubation. The results show that mor- phine inhibited PGEl-dependent accumulation of CAMP by more that 90% throughout the 4 hr period of incubation. Since a phosphodiesterase inhibitor was not present in this experiment, at least part of tbe PGEI-dependent increase in CAMP specific activity may represent extracelhdar CAMP. No evidence for cellular tolerance to morphine was observed during the 4 hr incubation period. The effect of morphine upon adenosine-dependent eleva- tion of CAMP levels and basal CAMP levels is shown in Fig. MINUTES FIG. 3. Basal and adenosine-stimulated levels of CAMP in in- tact NGlOE-15 cells measured as a function of time in the presence or absence of morphine sulfate. The concentrations used were 100 PM adenosine and 10 PM morphine sulfate. Bach 100 mm petri dish also contained 0.1 mM Ro 20-1724. The average dish con- tained 16.9 mg of protein. 3A, and B, respectively. Cells were preincubated with a phosphodiaterase inhibitor, Ro 20-1724, prior to the addi- tion of adenosine and/or morphine. The addition of adeno- sine evoked a large increase in CAMP and morphine inhibit- ed adenosine-dependent CAMP formation. Morphine also re- duced the basal level of CAMP (Fig. 3B). The effect of culturing cells in the presence of morphine for O-4 days upon the specific activity of basal or PGEl- stimulated adenylate cyclase activity is shown in Fig. 4A and B, respectively. Both basal and PGEl-stimulated adenylate cyclase specific activities increased gradually during the pe- riod that cells were exposed to morphine. After 23 days of incubation, the specific activity of adenylate cyclase, as- sayed in the presence of morphine, equaled values found at zero time in the absence of morphine. Thus, culturing cells with morphine results in tolerance to the narcotic. However, a marked increase in adenylate cyclase specific activity was observed when enzyme activity was determined in the ab- sence of morphine. Thus, morphine still inhibits adenylate cyclase activity but the inhibition is masked by a gradual, compensatory increase in the specific activity of the en- zyme. Similar results were obtained in other experiments (not shown) with cells in the logarithmic phase of growth and with confluent cultures of cells, which multiply at a greatly reduced rate. One experiment of this type is shown in Fig. 5. Homogenates prepared from cells cultured with or L.TyST' z P BASAL PGEI STIMULATED 4D ADENYLATE CYCLASE ACTIVITY NYLATE CYCLASE ACTIVITY Zoo% FIG. 4. Basal and PGE~-stimulated adenylate cyclasa activity of homogenates of NG108-15 cells cultured for the times shown in the presence of morphine. Confluent cultures (30-S%) in 100 mm petri dishes were divided into four groups; 10 rM morphine sulfate was first added to one group of cultures on day 0, to a second group on day 1, and to a third group on day 2. Media were changed and fresh morphine added daily. Cells were harvested, washed, homog- enized, and assayed on day 4. 3094 Biochemistry: Sharma et al. Proc. Nat. Aced. Sd. USA 72 (1975) z -t5or & -125; I - 100% - 75 I 2 - 5oz 0 d - z5zi P 01 I I "I " " " " " " " " " "1 12 24 36 480 12 24 36 48 ' HOURS FIG. 5. Basal and PGEr-stimulated adenylats cyclase activity of homogenates of NG108-15 cells cultured and assayed in the presence or absence of 10 pM morphine sulfate and/or PGEl as indicated. Cells 90% confluent in 100 mm petri dishes were cultured for the times indi- cated and were harvested, washed, and assayed immediately. Protein values ranged from 8.2 to 9.2 mg per dish. Filled symbols, addicted cells; empty, normal; triangles, plus morphine; circles, minus morphine. without morphine for 048 hr were assayed in the presence or absence of morphine and/or PGEr as indicated in the fig- ure. Some variation in the specific activity of adenylate cy- clase was observed both with normal and addicted cells. The specific activities of adenylate cyclase of normal and addict- ed cells did not differ appreciably during the first I2 hr of incubation, whereas exposure of cells to morphine for 2448 hr resulted in an increase in adenylate cyclase activity. Thus the delayed increase in adenylate cyclase specific activity requires more than 12 hr of exposure to the drug. The effects of withdrawal, a narcotic antagonist, and stereoisomeric narcotics on positive regulation of adenylate cyclase are shown in Table I. Cells were cultured with mor- phine for S days and then for an additional day with or without morphine to test the effects of withdrawal. The re- sults show that the morphine-dependent increase in adenyl- ate cyclase activity was reversed, ahnost completely, 24 hr after withdrawal of the drug. In another experiment, not shown here, withdrawal for SO min did not restore adenylate cyclase activity to normal levels. Thus the increase in ade- nylate cyclase activity due to narcotic-dependent positive regulation is stable > SO min. Table 1. Effects of withdrawal, a narcotic antagonist, and stereoisomeric narcotics on positive regulation of adenylate cyclase pmol CAMP formed/ Additions during cell culture Cells ctiltured 4 days HZ0 Morphine 3 Days Morphine 1 Day Withdrawal Morphine + Naloxone Cells cultured 3 days I-40 Naloxone Levorphanol Dextrorphan min per'mg protein I-40 .PGE, 18 95 27 115 21 98 18 79 18 100 21 90 30 - 21 - NGlO&15 hybrid cells were cultured in 100 mm petri dishes for 3 or 4 days in the presence or absence of 10 pM narcotic as indicated. Cultures were 50% confluent when narcotic was first added. The medium. was replaced each day. For the withdrawal experiment cells were cultured 3 days with 10 pM morphine sulfate, then plates were washed three times with growth medium and then cultured for an additional 24 hr in growth medium devoid of morphine. The average protein per dish was 7.2 mg after 3 days and 8.2 mg after 4 days. Adenylate cyclase activity was assayed in the presence of 10 pM naloxone. The final concentration of PGEI was 10 PM. The activities found with morphine and levorphanol in the presence of naloxone were higher than those observed in the absence of nalox- one; however, in all other cases little or no difference was observed in the absence of naloxone (data not shown). Table 2. CAMP levels of hybrid cells grown in the presence or absence of morphine pm01 cAMP/mg protein Hours Control Addicted cell cells, cells, Exp. cul- Additions to test cell no morphine no. ture responses morphine present 1. Tested withoutphosphodiesterase inhibitor 48 H,O 21 23 48 Naloxone 21 37 48 PGE, 264 81 48 PGE, + Naloxone 241 1183 48 Adenosine 103 65 48 Adenosine + Naloxone 72 217 2. Tested with phosphodiesterase inhibitor 0 H," 89 - 12 Hz0 54 40 12 Naloxone 61 125 24 H,O 61 40 24 Naloxone 60 285 48 Hz0 63 38 48 Adenosine 2000 ,859 48 Adenosine + Naloxone 1430 3020 NG108-15 celis were cultured in 60 mm petri dishes until 50% con- fluent. Then 10 .uM morphine sulfate or HsO was added and incu- bation was continued as indicated in the table. In Exps. 1 and 2 at appropriate times cells were incubated for 15 min in the presence or absence of 0.5 mM Bo 20-1'724 (final concentration). To test cell responses, 10 MM naloxone and/or PGE,; 100 pM adenosine, or HsO were added and incubation was continued for an additional 10 min. Reactions were terminated as described under Methods and Materials. Average mg of protein per dish at 0, 12, 24, and 48 hr were2.6, 2.8,3.5, and4.4, respectively. Biochemistry: Sharma et al. Proc. Nat. Ad. Sd. USA 72 (1975) 3095 Table 3. Opiate receptors of hybrid cells cultured with or without morphine Cell culture conditions fmol of [ sH]naloxone specifically bound/mg of protein Control cells Cells cultured with morphine 78 (67-92) 72 (63-80) Cells were cultured for 26 hr in the uresence or absence of 10-s M morphine. Homogenates of washed cells were assayed for opiate re- ceptors as described in the Methods and Materials section. The values are based on duplicate determinations and are the average of four sets of duplicate dishes. Numbers in parentheses represent the range found. Naloxone, a specific opiate antagonist with high affinity for the narcotic receptor, when present together with mor- phine reduced adenylate cyclase activity to levels below those of the control cells. Naloxone also reduced the PGEr- dependent adenylate cyclase activity when cells were cul- tured with this compound. These results possibly suggest ei- ther that the cells synthesize or serum contains a morphine- like factor. Levorphanol, a potent narcotic, increased adenylate cy- &se activity of cells cultured in its presence for 3 days, whereas under the same conditions dextrorphan, the phar- macologically inactive stereoisomer, had little or no effect on basal adenylate cyclase activity. In Table 2, CAMP levels of cells grown in the presence and absence of morphine.and tested for 10 min with nalox- one and/or activators of adenylate cyclase are shown. Cell response was tested in the absence of a phosphodiesterase in- hibitor in Exp. 1. The data show that cells cultured in the presence of morphine have normal basal CAMP levels. In the presence of added naloxone, however, basal, PGEr-, and adenosine-stimulated CAMP levels were considerably higher than those found with control cells. These data demonstrate morphine-dependent positive regulation of CAMP levels in intact cells, which agrees well with the increased adenylate cyclase activity found with homogenates. The data also show that cells grown in the presence of morphine are tolerant to morphine since their basal CAMP level is similar to that of control cells. Note that the CAMP assays were performed with cells still in the presence of growth medium and in the case of the addicted cells with morphine. The increase in CAMP levels found after short exposure of the cells to nalox- one indicates that the cells have become dependent upon the presence of morphine for the preservation of normal CAMP levels. This phenomenon is, we believe, the biochemical counterpart of the abstinence syndrome seen with animals and man upon administration of antagonists to dependent individuals. In Exp. 2, the phosphodiesterase inhibitor Ro 26-1724 was present when cell responses were tested. Cells grown in the presence of morphine for 42 hr had higher CAMP levels than control cells in the presence of naloxone. The difference was greater after 24 hr of growth with morphine. However, complete tolerance to morphine was not achieved under the conditions used even after 48 hr of culture with morphine. In the presence of naloxone, adenosine-stimulated CAMP levels of addicted cells were higher than those of control cells, similar to the observations made in the absence of Bo 20-1724. In Fig. 6, morphine is shown to have no effect upon the rate of protein synthesis during 9 days of growth. The 1 1 I I , , , , ( 30 MG PROTEIN /DISH 1 L II IIIII, a30 2 4 6 6 10 DAYS FIG. 6. Lack of effect of morphine on cell growth. Replicate cultures, each with 1.5 X 106 NG108-15 cells per 150 mm petri dish, were maintained in the presence or absence of 10 aM mor- phine sulfate. Cells were washed three times and the amount of protein per dish was determined. amount of protein synthesized per dish is used as a measure of cell growth and replication. Thus, the effects of morphine upon adenylate cyclase activity of cells made tolerant to the narcotic are specific ones and are not due. to an increase in total protein synthesis. Another way in which the cells might have adapted to the presence of morphine is by changingthe number-or proper- ties of opiate receptors. The data shown in Table 3 indicate that such changes have not taken place. The number of mor- phine receptors measured in homogenates prepared from well-washed hybrid cells grown in the presence of morphine is, within experimental error, unchanged from those of the control cells. The number of opiate receptors found in mor- phine-dependent rats also is similar to that of control ani- mals (14). DISCUSSION Narcotics affect adenylate cyclase in two ways, both mediat- ed by the opiate receptor. The first process is a readily re- versible inhibition of the enzyme by narcotics observed with homogenates and, indirectly, with intact cells The second phenomenon, late positive regulation, results in an increase in adenylate cyclase specific activity which is dependent upon incubation of cells with narcotics for 12 or more hours. The effects of narcotics on both inhibition and positive regu- lation of adenylate cyclase are stereospecific and are re- versed by naloxone. After culture with morphine for l-2 days cells are tolerant to the narcotic, since the increase in adenylate cyclase activity due to positive regulation com- pensates for the inhibition of enzyme activity observed in the presence of the narcotic. Cells are also dependent upon morphine, since withdrawal, or displacement of morphine from the opiate receptor by the antagonist, naloxone, in- creases basal, PGEi-. and adenosine-stimulated adenylate cyclase activities, which results in an overproduction of CAMP. This phenomenon can be likened to the abstinence syndrome in animals. Thus, NG103-15 cells become depen- dent upon morphine in 12-24 hr. The .mechanism of positive regulation of adenylate cy- clase is not known. The long incubation in the presence of 3096 Biochemistry: Sharma et al. Proc. Nut. Ad. Sd. USA 72 (1975) narcotics required to increase adenylate cyclase activity, the fact that increases in basal, adenosine-. and PGEr-stimulated adenylate cyclase activities are observed, and the relative stability of the elevated activity suggests that positive regula- tion represents an increase in the number of molecules of adenylate cyclase. The PGEr- and adenosine-dependent ele- vations of CAMP levels in NGlOgl5 cells are not additive% thus adenylate cyclase molecules rather than PGEr or aden- osine receptors probably are limiting in these cells. We do not rule out the possibility that the enzyme activity is in- creased by the formation of a relatively long-lived modula- tor; however, no evidence for the formation of a diffusable activator or inhibitor of adenylate cyclase was detected when homogenates prepared from normal and narcotic-de- pendent cells were combined and assayed for adenylate cy- clase activity. Goldstein and Goldstein (15) as well as Shuster (16) in 1961 proposed that enzyme induction might be a mechanism for drug tolerance and dependence. Positive regulation of adenylate cyclase represents a new type of receptor-mediated control of adenylate cyclase ac- tivity. Inhibition of adenylate cyclase activity and late posi- tive regulation are coupled, since both are initiated by inter- actions of narcotic with the opiate receptor. The inhibition of adenylate cyclase activity and concomitant reduction in CAMP levels may be required for positive regulation; that is, the concentration of CAMP or the activity of adenylate cy- clase may regulate the number of adenylate cyclase mole- cules in a sequential feedback fashion. If so, each transmit- ter, hormone, or effector of adenylate cyclase is, in fact, a dual regulator, for transient stimulation or inhibition of ade- nylate cyclase may evoke a regulatory process in the oppo- site direction hours after the initial event, The slowly expressed, relatively stable, positive regulatory process is a form of memory. Shifts in the number of mole- cules of adenylate cyclase or a long-lived modulator of en- zyme activity would alter the sensitivity of cell responses to other activators or inhibitors of adenylate cyclase, and thus would profoundly affect the efficiency of tram-synaptic communication, 4 H. Matsuzawa and M. Nirenberg. in prepuation. We think it likely that the endogenous morphine-like fac- tor termed enkaphahne, a newly discovered peptide or fam- ily of peptides found in the nervous system which mimic the effects of opiates (17), will evoke both transient inhibition and late positive regulation of adenylate cyclase and, thus, may be involved in a memory process. Certainly any reac- tion affecting either the number of molecules needed for sending or receiving neural information or molecules that regulate their activities provides a potential for activating or deactivating neural circuits and thus may play a role in a memory process. We thank Doyle Mullinax, Deborah Carper, and Linda Lee for their excellent assistance, and Mary Ellen Miller for typing the manuscript. S.K.S. is a Fogarty International Fellow at the National Institutes of Health. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. Sharma, S. K. Nirenberg, M. & Klee, W. A. (1975) Pmt. Nut. Acid. Sci. USA 72, SQO-594. Klee, W. A., Sharma, S. K. & Nirenberg, M. (1975) Ufe Sci., in press. Blosser, J. C., Abbott, J. R. & Sham, W. (1975) Fed. Proc. 34, 713. Traber, J., Fischer, K., Latzin, S. & Hamprecht, B. (1975) Nu- ture 253,120-122. Traber, J., Reiser, G., Fischer, K. EL Hamgrecht, B. (1975) FEBS Lett. 52,327-332. Collier, H. 0. J. & Roy, A. C. (1974) Nature 248,24-27. Collier. H. 0. J. & Roy, A. C. (1974) Prostaghdtns 7,361- 376. Chou, W. S., Ho. A. K. S. & Leh, H. H. (1971) Proc. West. Phawnacol. Sot. 14,42-46. &xi, S. K., Cechin, J. & Volicer, L. (1975) Ltfe Set. 16,75Q- c?i C P Pastemak. G. W. h Cuatrecasas, P. (1975) FEBS Let;. 5i, &2-245. Gilman,.A. G. (1970) Proc. Nat. Acud. Set. USA 67,305-312. Lowry, 0. H., Rasebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J, Btol. Chem. 193,265-275. Klee, W. A. & Nirenberg. M. (1974) Proc. Nat. Acad. Set. USA 71,3474-3477. - Klee, W. A. L Streaty, R. A. (1974) Nature 248,61-66. Goldstein, D. B. 81 Goldstein, A. (1961) B&&em. Pharmawl. 8,48. Shuster, L. (1961) Nuture 189,314-315. Hughes. J. (1975) Brain Res. 88,295-308.