Proc. Nat. Amd. Sci. USA Vol. 71, No. 9, pp. 3474-3477, September 1974 r-- A Neuroblastoma x Glioma Hybrid Cell Line with Morphine Receptors ! (cell membranes/narcotics/neurobiology/tissue culture) WERNER A. KLEE AND MARSHALL NIRENBERG Laboratory of General and Comparative Biochemistry, National Institute of Mental Health and the Laboratory of Biochemical Genetics, National Heart and Lung Institute, National Institutes of Health, Beth&a, Maryland 20014 Contributed by Marsh&l Nirenberg, June l&197.4 ABSTRACT A neuroblastoma X glioma hybrid ceR line with welldeveloped neural properties was found that has high-affinity morphine receptors. The average cell contains approximately 3 X 106 receptors. In contrast, parent ceRs and other neuroblastoma or hybrid cell lines tested had few or no morphine receptors. Membrane preparations that bind narcotic analgesics such as morphine and its congeners have been described by a number of laboratories (l-9). Such binding satisfies the criteria first enunciated by Goldstein et al. (10) for a narcotic receptor in that it is specific for the pharmacologically active D (-) en- antiomers of the narcotics. Furthermore, there is a generally satisfactory correlation between binding affinities measured for various opiates and their potencies as analgesics in ti (5) as well as a somewhat more satisfactory correlation be- tween binding strength and narcotic agonist potency mea- sured with isolated intestinal preparations (11). The mem- brane preparations studied to date were isolated from brain or intestine and, therefore, are derived from mixtures of cell types. We hoped to obtain a relatively homogenous cell population for our studies of opiate action by examining clonal lines of neuroblastoma cells (12) for morphine receptors. Somatic cell hybrids also were studied, since previous results showed that genes for neural properties can be activated (13) or deactivated (14) in hybrids formed by the fusion of neuro- blastoma cells with other cell types. New phenotypes thus generated can be inherited and thereby perpetuated in a fairly stable fashion. In this report, we wish to describe cell lines with and with- out morphine receptors. MATERIALS AND METHODS Chemicals were obtained from the following sources: [fH]di- hydromorphine (51.6 Ci/mmole) from New England Nuclear Corp.; morphine sulfate from Merck and Co.; naloxone hydrochloride was donated by Endo Laboratories, Inc.; the other narcotics used were kindly supplied by Dr. Everette May; fetal bovine serum was from Colorado Serum Co.; DMEM (Dulbecco-Vogt modification of Eagle's minimal es- sential medium) was from GIBCO (Catalogue no. H21). Other chemicals were of reagent grade purity. Cell Lines. The neuroblastoma cell lines used were derived from the mouse C-1300 tumor; their properties have been described previously (12). Neuroblastoma X glioma hybrids, NGlO&5 and NGlOS-15, were obtained* by Sendai virusin- o B. Hamprecht, T. Amano, and M. Nirenherg, in preparation. duced fusion of neuroblastoma clone N18TG-2, resistant to 6 thioguanine (14), and CGBu-1, a 5bromodeoxyuridim+re- sistant clone of the rat glioma C6 (15). Clone NLlF is a hy- brid formed by fusion of neuroblastoma clone N4TGl and mouse L cell clone B32 (14). Neuroblastoma and CGBu-1 cells were grown in petri dishes or flasksin 90% DMEM, 10% fetal bovine serum, in a humidi- fied atmosphere of 90% air, 10% COZ, and 37' as described (12). Hybrid cells were grown in the same medium supple- mented with 0.1 mM hypoxanthine, 1 pM aminopterin, and 12 FM thymidine. For drug binding experiments, the cells us- ually were grown to confluency and harvested by scraping (from petri dishes) or shaking (from flasks) in their growth medium. The cells were centrifuged at approximately 500 X g for 5 min at room temperature and the pellets (generally less than 1 ml) were washed two times with 50 ml of Dl [137 mM NaCI, 5.4 mM KCI, 0.17 mM Na*HPG,, 0.22 mM KH2POI, 5.5 mM glucose (IS)] and suspended in a solution of 0.32 M sucrose, 0.01 M Tris. HCl (pH 8) at 3" so that the concentra- tion of protein was close to 0.5 mg/ml. Uniform cell suspen- sions were achieved by aspirating the suspension in a 10-ml plastic pipet three times. Cell suspensions were kept at 3' and usually used immediately for binding studies. Cell sus- pensions that were frozen and stored at -6O* retained the specific binding of [SH]dihydromorphine. Routine Assay Method. A cell suspension prepared as just described (900 ~1) was added to a plastic test tube containing either 10 ~1 of water or 10 ~1 of 0.1 mM morphine sulfate. A solution, 100 ~1, containing [3H]dihydromorphine (10 nM, 51.2 Ci/mmole) dissolved in 0.32 M sucrose was added to each tube in a darkened room and the samples were mixed with a vortex mixer. After incubation at 37' for 10 min, tubes were centrifuged at 19,000 rpm (45,000 X g) for 10 min in a Sorvall RC2B centrifuge at 0' and supernatant fluids were removed with a 5ml syringe equipped with a blunt 19gauge needle and then discarded. Each tube and surface of the pellet was washed with 1 ml of 0.32 M sucrose and the washed pellet was suspended in 1 ml of 1% Triton X-100 and assayed for radioactivity after mixing with 7.5 ml of Triton-toluene scintillation fluid (17). All operations involving dihydromor- phine were performed in very dim, indirect light. Each sample was assayed in triplicate or quadruplicate. Neuroblastoma cell lines used were subcultured 5 to 30 times; hybrid cells were subcultured 10 to 20 times. In most cases a number of different subcultures were tested. Specific binding is defined as fmoles of [JHldihydromorphine bound per mg of protein (upon incubation of cells with nM IaHldi- 3474 Proc. Nat. Acad. Sci. USA 71 (1974) hydromorphine) that are displaced by a lOOO-fold, or higher, excess of unlabeled morphine. Binding affinities of morphine and of other narcotics were estimated from experiments in which a range of concentrations of unlabeled opiates was tested for its ability to displace bound [aH]dihydromorphine as described previously (9). RESULTS The specific binding of [*H]dihydromorphine was tested with various cell lies derived from the nervous system as shown in Table 1. Little or no.specific binding of dihydromorphine was detected with neuroblastoma clone N18TG-2 or rat glioma clone CGBU-1. However, specific binding sites for [*H]dihy- dromorphine were found with a hybrid cell line, NGlO8-15, originating from the fusion of the neuroblastoma N18TG2 and glioma CBBu-I+. Relatively few specific narcotic binding sites were detected, under the conditions used, with a sister neuroblaetoma X glioma hybrid cell line, NGlO&5, neuro- blastoma clones S20 and S29Y, and a neuroblastoma X L cell hybrid, NLlF. The low values may indicate the presence of a reduced number of receptors or receptors $th a reduced affinity for morphine. Narcotic receptors were not detected with neuroblastoma lines NlE, NIE-115, N4, N18, or Nl& TG-2, or with CSBU-1: Thus, two classes of cells were found: cells with morphine receptors, and cells that do not have these receptors in quantities sufficient to detect with the assay used. Characterization of the Narcotic Receptors of NGlOB-16 Hybr$ Cells: A portion, only, of the binding of [*H]dihydro- morphine is saturable in that it is displaced by nonradioactive morphine (Fig: 1)) and indeed, this phenomenon is the basis of our assay procedure. Nonsaturable (that is nondisplaceable) binding of (SH]dihydromorphine has been found in all tissues tested and is thought to be nonspecific binding. Only the saturable component of the binding shown in Fig. 1 is con- sidered to be specific binding. Morphine Receptors in Neural Cell Lines I . . 1 I I I I 10-D 10-0 Id' IO-@ Id' MoRPHIpE 04) Fm. 1. The displacement of bound [*H]dihydromorphine by dZerent concentrations of unlabeled morphine from a suspension of NGlOS-15 cells (subculture 10, 468 rg of protem per tube). Each tube contained 1 n&l [*H]dihydromorphine and the indi- cated concentration of morphine. The specific binding sites discriminate between (+) and (-) enantiomers of the narcotics as shown for the pair dextror- phan and levorphanol in Fig. 2. The latter compound is seen to displace specifically bound dihydromorphine at concentra- tions more than two orders of magnitude lower than does its analgesically inactive enantiomer, dextrorphan (18). A similar result has been obtained wjth the (-) and (+) enantiomers of metasocine, the latter of which is inactive as an analgesic (19) and as a displacer of dihydromorphine, at least at con- centrations of PM and below. The number of binding sites present in suspensions of NG198-15 can be estimated by measuring the amount of specific binding as a function of [SHldihydromorphine con- centration. We find that, at saturation, approximately 0.6 pmoles of dihydromorphine are bound specifically per mg of cell protein (Fig. 3). The insert to the Figure shows the data TABLE 1. Specify bindins of [lH]dihydromorphine to ncuroblastomcl and hybrid cell lines fmolea of (`HI- No. of dihydromorphine experi- Neurotranamitter Cell lines bound/mg of proteina menta synthesis Ref. A. Parents Mouse neuroblaatoma N18TG-2 Of1 2 None Rat glioma CGBU-1 If1 5 None & B. Nl8TG.2 x C6BU-1 hybrids NGlO815 17 f6 13 Acetylcholine b NGlO8-5 4f2 I1 Acetylcholine b C. Neuroblastama NS20 3f2 3 Acetylcholine (12) NS20Y 4f2 a, 7 Acetylcholine (12) NlE 0 1 Catechols (1% NlE115 2fl 9 Catechols (12) N4 0 1 None (12) N18 2fl 6 None (12) D. Neuroblaatonaa X L cell hybrid (N4TG-1) x (B82) NLlF 6f3 3 None (14) 0 Specific binding of ["Hldiiydromorphine is that amount of the bound radioactivity which is displaced by 1 PM unlabeled morphine (see the Mater& and Methods section for details). b B. Hamprecht, T. Amano, and M. Nirenberg, in preparation. 3476 Biochemistry: Klee and Nirenberg Proc. Nat. Acad. Sci. ~S;l il (,19;.$) A 700 , , , I 0 lo-`O lo9 10-o lo-' lo* 1o-6 MOIAMTY OF ADDED DFiUG FIQ. 2. The displacement of bound ["Hldihydromorphine by levorphanol (0), [( - )%hydroxy-N-methyhnorphinan] and dex- trorphan( A) [ (+ )3-hydroxy-N-methyhnorphinan] at the concen- trations shown from a suspension of NGlOS-15 cells (subculture 15). plotted according to Scatchard (20) and is consistent with there being only one class of dihydromorphine binding site with a dissociation constant of 20-30 nM. We have examined the ability of a number of other narcotic analgesics as well as of the specific antagonist, naloxone, to bind to the receptors found in NGlO&15 suspensions, as measured by their ability to displace specifically bound [SH]dihydromorphine. In Table 2 the binding affinities mea- sured in this way with NGlO&15 suspensions and also with a crude mitochondrial fraction of rat brain as the source of receptors are shown. Note that the binding affinities of the NGlO&15 preparations are generally somewhat lower than those of the brain preparation, but that the relative binding strengths are similar in the two preparations. The order of binding affinities is seen to be well correlated with the afhnities found by Kosterlitz and Watt (11) for the pharmacological action of these agents on segments of guinea pig ileum. An important question concerns the location of the narcotic receptor sites. Trypsin is often used as an aid in detaching TABLE 2. Binding aJSties of a series of narcotics to suspensions of NG108-16 and to brain homogenates Compound Apparent dissociation constant0 (+I) Activity in NGIO8-15 Rat brainb guinea pig ileumc Levorphanol 0.3 0.8 7 Morphine 20 3 I38 Pentszocine 50 10 150 Methadone EiO 20 - Codeine 1000 800 8800 Naloxone 10 1 1.2 Dextrorphan 156 160 - 0 Binding wse measured as the concentration required to dis- place l/z of the specifically bound [`Hldihydromorphine in our usual sssay system. Under our conditions of assay this number should be close to a true dissociation constant. b Measured as described in Mat&b and Method.9 and (9). (R. A. Streaty and W. A. Hlee, unpublished observations.) c These are apparent dissociation constants (nM) for the antagonist activity cf these compounds in the isolated ileum (11). e . I g 0.4- a s g 0.3- m 0 pj 0.2- I I I I I J 0 I 2 3 DIHYDROMORPHINE ( M x 10'1 FIQ. 3. The specific binding of [`Hldihydromorphine to s suspension of NGlOgl5, measured as the difference between the amount of binding found in the absence and presence of 10 PM unlabeled morphine, as a function of [`Hldihydromorphine con- centration. The cells used were subculture 12 and were grown in the presence of dibutyryl cyclic AMP. cells from surfaces and is believed not to penetrate the cell membrane. Treatment of a cell suspension of NGIo&15 with 0.05% trypsin for 10 min at 37' resulted in a decrease of the specific binding of [3H]dihydromorphine from 26 to 8 fmoles/ mg of protein under standard conditions. Although most of our experiments were performed with cell suspensions, we found that the specific binding of [3H]dihydromorphine to homogenates of NGlO&15, prepared in 0.32 M sucrose, was unchanged when compared with whole cells. When a homog- enate of NG108-15 was centrifuged to equilibrium in a 5% 2Oyc w/w continuous sucrose gradient, the bulk of the mor- phine receptors were found at a sucrose concentration (mea- sured by refractometry) equivalent to a density of 1.16 g/ml. These results are consistent with the working hypothesis that the receptor is localized on the plasma membrane. DISCUSSION The results show that a neuroblastoma X glioma hybrid cell line (NG108-15) has the ability to bind narcotic analgesics in a stereospecific manner and with high affinity, whereas little or no specific binding was observed with parental cells and other cell lines tested, The relative binding affinities of narcotic analogues to NGlOS-15 cells were approximately the same as those found with rat brain homogenates and match well the pharmacological potency of the narcotics reported for isolated intestinal segments. Thus, there is good reason to believe that the neuroblastoma X glioma hybrid cells have morphine receptors similar to those found in brain and guinea pig ileum. In addition to morphine receptors, NGlO&15 cells also have other neuronal characteristics that are not found with parental cells, such as choline acetyltransferase*, intracellulzl acetylcholine, and clear and dense core vesicles, 500 and 1560 d in diameter (21). In addition, the cells have long neurites, electrically excitable membranes, and nicotinic acetylcholirlt' receptors. t t B. Hamprecht, T. Amano, W. Kemper, and M. Nirenbcrg, unpublished observations. Proc. Nat. Ad. Sci. USA 71 (j97.4) Morphine Receptors in Neural Cell Lines 3477 Many properties of neuronal cell lines in culture have been shown to be regulated (22). Thus, the specific activities of choline acetyl transferase and choline&erase* as well as the extent of morphological differentiation of NGlO&15 cells can be increased many-fold in cultures which are grown in the presence of dibutyryl CAMP or are maintained in stationary phase (21). We did not detect regulation of the number of morphine receptor sites, since logarithmically dividing NG- 108-15 cells had the same number of morphine receptors as sister cultures maintained in the presence of dibutyryl CAMP that were highly differentiated with respect to cell mor- phology. On the other hand, we observed with one popula- tion of NGlO&15 that the cells lost much of their ability to bind dihydromorphine in a specific manner after 19 to 20 subcultures. Thus, it should be possible to obtain defective cell lines with respect to the synthesis, structure, or function of morphine receptors. possibly, phenomena such as morphine dependence and tolerance. We thank Mr. Doyle Mullinex and Mrs. Theresa M. Caryk for growing cells and Dr. Everette May for samples of the narcotics teated. Pert, C. B. & Snyder, S. H. (1973) Science 179,1011-1614. Terenius. L. (1973) Actu Phunnacol. Tozic. 32. 317320. Simon, 6. J., Hiller, J. M. & Edelman, I. (1973)Proe. Nat. Acud. Sci. USA 70, 1947-1949. Pert, C. B. Br: Snyder, S. H. (1973) Proc. Nat. Awd. Sci. USA 70,2243-2247. Wong, D. T. 8c Horng, J. S. (1973) Life Sci. 13, 1543-1556. Kuhsr, M. J., Pert, C. B. & Snyder, S. H. (1973) Nature 245, 447-456. At saturation of dihydromorphine we estimate that there are 0.6 pmoles of bound morphine per mg of cell protein. This is approximately twice the specific activity of rat brain homogenates measured under identical conditions. Since 1 mg of protein corresponds to about 1.2 X 1V cells, there are approximately 300,000 receptor sites per cell. This is a large number and is similar to the number of nicotinic acetylcholine receptors of cultured sympathetic ganglion neurons of the chick (23). One mechanism that has been proposed for the action of morphine is to inhibit the release of acetylcholine (24). Since NGlO8-15 cells both synthesize acetylcholine and release this compound into the medium, it should be possible to test this hypothesis. In addition, the recent demonstration by Collier and Roy of the specific reversal by narcotic analgesics of prostaglandin Estimulated CAMP formation in rat brain homogenates (25) provides a possible mechanism for the action of these drugs. In collaboration with Shail K. Sharma, we have confirmed these findings with homogenates prepared from NGlO&15, in that morphine was found to inhibit adenyl- ate cyclase activity of NGlO3-15. X: 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. j6. Hiller, J. M., Pearson, J. & Simon, E. J. (1973) Res. Com- mun. Chem. Paulol. Pharmawl. 6, 1052-1062. Pert, C. B., Pasternak, G. 6c Snyder, S. H. (1973) Science 182, 1359-1361. Klee, W. A. & Streaty, R. (1974) Nature 248, 61-63. Goldstein, A., Lowney, L. I. 6 Pal, B. K. (1971) Proc. Nat. Awd. Sci. USA 68, 1742-1747. Kosterlitz, H. W. & Watt, A. J. (1968) &it. J. Pharmacol. Chemother. 33,266-276. Amano, T., Richelson, E. & Nirenberg, M. (1972) Proc. Nut. Acud. Sci. USA 69,258-263. Minns, J., Nelson, P., Peacock, J., Glazer, D. & Nirenberg, M. (1971) Proc. Nat. Acad. Sci. USA 68,234-239. Mmns, J., Glazer, D. & Nirenberg, M. (1972) Nature New Btil. 235,225-231. Amano, T., Hamprecht, B. t Kemper, W. (1974) Exp. Cell Res., in press, Ham, R. G. 8 Puck, T. T. (1962) in Methods in Enzymology, eds. Colowick, S. P. &Kaplan, N. 0. (Academic Press, New York), Vol. 5, pp. 96-119. Patterson, M. S. & Greene, R. C. (1965) Anal. Chem. 37, 854-857. The availability of relatively homogeneous populations of cells with and without morphine receptors and with other neuronal properties provides opportunities to explore ques- tions relating to the mechanism of action of morphine and, 17. 18. 19. 20. 21. 22. 23. 24. Eddy, N. B., Halbach, H. & Braenden, 0. J. (1957) Bull. W.H.O. 17,569-863. May, E. L. & Eddy, N. B. (1959) J. Org. Chem. 24, 1435 1437. Scatchard, G. (1949) Ann. N.Y. Acud. Sci. 51, 660-672. Daniels, M. & Hamprecht, B. (1974) J. Cell Biol., in press. McMorris, F. A., Nelson, P. G. & Ruddle, F. H. (1973) Neurosci. Rea. Progr. Bull. 412-536. Greene, L. A., Sytkowski, A. J., Vogel, Z. & Nirenberg, M. W. (1973) Nature 243, 163-166. Smith, C. B. (1972) in Chemical and Biological Aspects of Drug Dependence, eds. Mule, S. J. & Brill, H. (CRC Press, Cleveland, Ohio), pp. 495-564. 25. Collier, H. 0. J. & Roy, A. C. (1974) Nature 248, 24-27.