Introduction

Top Abstract Introduction Methods Results Discussion References

Differences between in vivo and in vitro preparations in studies of receptor down-regulation may help explain some mechanisms of opiate tolerance. Down-regulation has been easier to demonstrate in vitro than in vivo (Zadina et al., 1995). For example, delta agonists down-regulate receptors in cell lines containing only delta receptors (Chang et al., 1982; Law et al., 1982, 1983), and morphine and mu agonists down-regulate receptors in cell lines containing mu receptors (Werling et al., 1989). Down-regulation of both mu and delta receptors has been demonstrated in the human neuroblastoma cell lines SH-SY5Y and SK-N-SH in response to chronic morphine and selective agonists (Carter and Medzhiradsky, 1993; Baumhaker et al., 1993; Zadina et al., 1990, 1993a, 1994a).

Early in vivo studies showed no change in receptor number after exposure to chronic agonists (Klee and Streaty, 1974; Hollt et al., 1975). More recent studies have shown down-regulation in response to chronic etorphine (Tao et al., 1987), PL017 (Tao et al., 1990) and DPDPE (Tao et al., 1991). A major difference exists, however, between in vitro and in vivo studies, because down-regulation in response to morphine has been observed only rarely in vivo (Bhargava and Gulati, 1990). Indeed, under the same conditions in which etorphine down-regulates receptors, morphine increases receptor number (Tao et al., 1987). The only preparation in which down-regulation in response to morphine is observed consistently in vivo is the prenatal or early neonatal rat (Tempel et al., 1988).

These studies highlight two of the important features of opiate receptor down-regulation: a) in vitro preparations, which exhibit down-regulation to many different agonists, may lack one or more components of the adaptation response to chronic drug exposure that is present in mature animals; b) the manner in which cells adapt to chronic morphine seems to depend on the characteristics of the specific cells studied. Different brain regions and cell lines, for example, may have different complements of G-proteins or efficiencies of coupling to intracellular signaling pathways.

One possibility contributing to the tolerance response and to the lack of receptor down-regulation in mature animals is the presence of an endogenous "opiate-modulating system," such as the Tyr-MIF-1 family of peptides. Tyr-MIF-1 (Tyr-Pro-Leu-Gly-NH₂) and Tyr-W-MIF-1 (Tyr-Pro-Trp-Gly-NH₂) have been isolated from human and bovine brain (Hackler et al., 1993; Erchegyi et al., 1992; Horvath and Kastin, 1989, 1990), and they both bind selectively to the mu receptor (K_i = 1 µM for Tyr-MIF-1 and K_i = 70 nM for Tyr-W-MIF-1) (Zadina et al., 1994c). They exhibit opiate agonist and antagonist properties in the guinea pig ileum, with antagonism most easily observed in tolerant tissue (Erchegyi et al., 1992, 1993; Zadina et al., 1992). Tyr-MIF-1 has a significantly longer half-life in neonatal versus adult plasma, a finding that suggests the possible postnatal appearance of its degradative enzymes (Kastin et al., 1994).

In this study, we used an in vitro preparation, the human neuroblastoma cell line SH-SY5Y, in which opiate receptor down-regulation by morphine and selective agonists has already been demonstrated (Zadina et al., 1990, 1993a, 1994a), to test whether Tyr-W-MIF-1 can block this agonist-induced decrease in receptor number. We chose Tyr-W-MIF-1 instead of Tyr-MIF-1 because of its higher affinity for the mu receptor. Serum-free medium was used to avoid degradation of the peptide, and therefore we characterized the suitability of this culture condition for the study of receptor regulation in SH-SY5Y cells.

	Methods

Top Abstract Introduction Methods Results Discussion References

Materials. All trans-retinoic acid was obtained from Sigma Chemical Co. (St. Louis, MO), [³H]DAMGO (60 Ci/mmol) from Amersham (Arlington Heights, IL), [³H]pCl-DPDPE (40 Ci/mmol) from Dupont New England Nuclear (Wilmington, DE) and fetal bovine serum from Intergen (Purchase, NY). Tyr-W-MIF-1 was synthesized in our laboratory by solution-phase methods. PL017 and DPDPE were obtained from Multiple Peptide Systems through the National Institute on Drug Abuse.

Cell culture conditions. SH-SY5Y human neuroblastoma cells were cultured, harvested and prepared for binding assays as described previously (Zadina et al., 1993a, 1994a). Cells of passage 24-31 were cultured in a medium (EFF) containing a 1:1 ratio of Eagle's Minimum Essential Medium and F12 (Gibco no. 41500-034 and no. 21700-075; Grand Island, NY) with 10% fetal bovine serum, 100 µg/ml streptomycin and 100 IU/ml penicillin (Gibco). They were grown at 37°C in a humidified atmosphere containing 5% CO₂. At about 90% confluence, cells were differentiated into a neuronal phenotype with RA (10 µM). RA was administered, with the media change, every 2 days for the last 6 days of culture, for a total of three treatments. To avoid metabolism of administered peptides, cells were switched to a serum-free medium (EFN), with the medium supplement N2 [containing insulin (bovine, 5 µg/ml), transferrin (human, 100 µg/ml), progesterone (20 nM), putrescine (100 µM), and Na selenite (30 nM)], with the last RA treatment. Drugs were administered for the last 24 h of culture.

Plasma membrane preparation. Cells were rinsed three times with serum-free medium (EF) and then dissociated with Ca⁺⁺/Mg⁺⁺-free phosphate-buffered saline containing 0.04% EDTA. After centrifugation (183 × g for 7 min at 4°C) for removal of the phosphate-buffered saline and EDTA, cells were reconstituted in 50 mM TRIS, vortexed and homogenized with a polytron at setting 6 (23,000 rpm) for 20 sec. The homogenate was centrifuged at 30,000 × g for 20 min, and the pellet was incubated for 1 h at room temperature in 50 mM TRIS/100 mM NaCl to remove endogenous ligands. After another centrifugation at 30,000 × g for 20 min, the pellet was reconstituted in STEM (0.25 M sucrose, 5 mM TRIS, 0.5 mM EDTA and 1 mM MgSO₄), and protein samples were taken. The sample was divided into equal aliquots and, after a final centrifugation, resuspended in 2 ml STEM for freezing. Protein was measured by the method of Lowry et al. (1951) with bovine serum albumin as standard.

Opiate receptor assays. Saturation binding assays were performed in 50 mM TRIS with 0.1% bovine serum albumin at a final volume of 600 µl. For mu receptor assays, aliquots (200 µl, containing 100 µg protein) of membranes were incubated for 90 min at 23°C with varying concentrations (0.08-5 nM) of [³H]DAMGO. Nonspecific binding was measured in the presence of 1 µM DAMGO. For delta receptor assays, membranes were incubated for 4 h at 23°C in the presence of varying concentrations (0.08-5 nM) of [³H]pClDPDPE, with nonspecific binding measured in the presence of 10 µM naltrexone. Bound and free ligands were separated by filtration. Specific binding at low concentrations of isotope were 70 to 85% for [³H]DAMGO and 50 to 65% for [³H]pCl-DPDPE.

HPLC. Tyr-W-MIF-1 was iodinated by the chloramine T method, and the monoiodinated fraction was isolated by HPLC. The specific activity (2100 Ci/mmol) was diluted with nonradioactive Tyr-W-MIF-1 to a final concentration of 200,000 cpm/30 µM Tyr-W-MIF-1 in 7.5 ml EFN/10 µM RA. SH-SY5Y cells were incubated at 37°C in this medium in T-25 flasks for 1 min or 2, 5 or 24 h. At the end of the incubation, 3.5 ml of medium was transferred to 3.5 ml ice-cold ethanol for 30 min, centrifuged at 3500 rpm for 30 min and the supernatant dried down (Speed-Vac, Savant Instruments, Hicksville, NY). The extract was reconstituted in 400 µl of 0.1% trifluoroacetic acid/water (HPLC buffer A) and fractionated by HPLC on a Vydac C18 no. 201HS54 (46 × 250 mm) column. The gradient for buffer B (methanol in 0.1% trifluoroacetic acid) increased from 20 to 45% during 20 min, then remained at 45% for 30 min, increased to 65% in 0.5 min, remained at 65% for 20 min, then increased to 80%. Retention times of synthetic standards for Tyr-W-MIF-1 and its fragments were determined with the same column and gradient.

Statistical analyses. Values (mean ± S.E.M.) are represented as percent of control to normalize differences in B_max from one assay to the next. Percent attenuation was calculated as follows: [(% decrease in B_max induced by 3 µM MS)  (% decrease in B_max induced by 3 µM MS + Tyr-W-MIF-1)]/% decrease in B_max induced by 3 µM MS. Multiple experiments were analyzed by analysis of variance followed by Duncan's range test.

	Results

Top Abstract Introduction Methods Results Discussion References

Suitability of serum-free culture conditions for study of receptor regulation. To study the role of a peptide in receptor regulation, it was necessary to culture the cells in a serum-free medium to avoid degradation of the peptide. However, because the presence of serum was required for initial seeding and growth of the cells, they were cultured in a serum-containing medium (EFF) until 48 h before harvest. At this point, they were transferred to a medium (EFN) containing the supplement N2 and were allowed to adjust to these new conditions for 24 h before drug treatment. The switch to serum-free conditions caused a significant change in the mu:delta ratio from 1.9:1 in EFF to 1.3:1 in EFN [F(1,12) = 7.90, P < .05]. The serum may contain opiate-like substances, as demonstrated in both binding and guinea pig ileum assays (Zadina et al., 1994a). Therefore, the removal of the serum, rather than the addition of N2, would appear to be the more likely explanation for the change in receptors.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   HPLC chromatogram of 125I-Tyr-W-MIF-1 after 1 min (dashed line) and 24 h (solid line). At 24 h, 79% of the cpm eluted at the position of the intact peptide.

View larger version (32K): [in this window] [in a new window]   Fig. 2.   Mu receptor and delta receptor number expressed as percent of control Bmax after 24 h incubation with Tyr-W-MIF-1 or Met-enkephalin in serum-free medium. Neither mu nor delta receptor number was decreased by Tyr-W-MIF-1 under the same conditions in which Met-enkephalin (Met-enk) reduced mu receptors to 67 ± 1% of control and delta receptors to 31 ± 5% of control (n = 2).

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Up-regulation of mu and delta receptors in serum-containing medium (EFF) and serum-free medium (EFN), expressed as percent increase in Bmax, after 24 h incubation with 10 µM naloxone (Nlx). Values are means ± S.E.M. for three to five experiments.

Attenuation of morphine-induced down-regulation of both mu and delta receptors by Tyr-W-MIF-1. Of the doses of morphine tested for down-regulation of opiate receptors in EFN (1, 3 and 10 µM), the intermediate dose was chosen for the interaction studies because it afforded enough down-regulation to detect enhanced or attenuated down-regulation but avoided the possibility that a supramaximal dose of morphine could prevent detection of attenuation. The dose of Tyr-W-MIF-1 was varied (10, 30, 100 µM) so that it represented 3 times, 10 times or 33 times the dose of morphine, and the two drugs were administered to the cells simultaneously for 24 h. Down-regulation of mu receptors by morphine was attenuated by Tyr-W-MIF-1 in a dose-dependent manner. Figure 4 shows the percent attenuation of morphine-induced down-regulation of mu receptors by increasing doses of Tyr-W-MIF-1. The 10 µM dose of Tyr-W-MIF-1 had no effect, but down-regulation was significantly attenuated by 30 µM (28 ± 9%, P < .05) and 100 µM (49 ± 8%, P < .01) doses of Tyr-W-MIF-1.

View larger version (32K): [in this window] [in a new window]   Fig. 4.   Attenuation of mu receptor down-regulation by Tyr-W-MIF-1. The main figure illustrates increasing percent attenuation with increasing doses of Tyr-W-MIF-1 (*P < .05; **P < .01 compared with control cells given morphine but no Tyr-W-MIF-1). Inset shows the results expressed as percent of control Bmax. Coincubation with increasing doses of Tyr-W-MIF-1 (+ WI0, + W30 and + W100) resulted in smaller decreases from control Bmax. Values are mean ± S.E.M. for two to three experiments.

View larger version (30K): [in this window] [in a new window]   Fig. 5.   Attenuation of delta receptor down-regulation by Tyr-W-MIF-1. The main figure illustrates percent attenuation with increasing doses of Tyr-W-MIF-1 (*P < .05 compared with control cells given morphine but no Tyr-W-MIF-1). Inset shows the results expressed as percent of control Bmax. Coincubation with 30 µM and 100 µM (+ W30 and + W100) but not 10 µM (+ W10) Tyr-W-MIF-1 resulted in smaller decreases from control Bmax. Values are mean ± S.E.M. for two to three experiments.

Effects of Tyr-W-MIF-1 on selective agonist-induced down-regulation of opiate receptors. In addition to morphine, selective agonists also have been shown to down-regulate opiate receptors in SH-SY5Y cells (Zadina et al., 1994a). We therefore tested the ability of Tyr-W-MIF-1 to attenuate this down-regulation by selective agonists. Figure 6 shows the dose-responsive attenuation by Tyr-W-MIF-1 of mu receptor down-regulation induced by PL017, a mu selective agonist. PL017 alone (1 µM) decreased mu receptor B_max by 52%, in agreement with our earlier work (Zadina et al., 1994a). Tyr-W-MIF-1 attenuated this effect at 30 µM (29 ± 1%) and 100 µM (68 ± 2%) doses. This pattern parallels that seen with morphine-induced down-regulation of mu receptors.

View larger version (23K): [in this window] [in a new window]   Fig. 6.   Attenuation by Tyr-W-MIF-1 of PL017-induced down-regulation of mu receptors. Results are expressed as percent attenuation (mean ± S.E.M. for two separate experiments) with increasing doses of Tyr-W-MIF-1.

	Discussion

Top Abstract Introduction Methods Results Discussion References

In this study, we established the suitability of serum-free culture conditions for the study of opiate receptor regulation and used this preparation to examine the effects of the opiate-modulating peptide Tyr-W-MIF-1 on agonist-induced decreases in receptor number. Tyr-W-MIF-1 remained largely intact during 24 h of culture in serum-free medium, but did not affect receptor number when administered alone. However, Tyr-W-MIF-1 dose-dependently attenuated morphine-induced down-regulation of both mu and delta receptors. The effect of Tyr-W-MIF-1 on selective agonist-induced down-regulation revealed differences between mu and delta receptors; the PL017-induced down-regulation of mu receptors was attenuated by Tyr-W-MIF-1, but the DPDPE-induced down-regulation of delta receptors was not. These studies represent the first demonstration of attenuation by an endogenous ligand of morphine-induced changes in receptor number and suggest a mechanism whereby endogenous antiopiates can counteract the effects of chronic morphine.

Several peptides have been proposed to play a role in opiate tolerance and dependence by acting as "antiopiates." According to the proposed model, administration of exogenous opiates, such as morphine, causes the release of endogenous antiopiate peptides. These peptides counteract the effects of morphine, thereby causing a need for increased doses of the drug to achieve a given effect (tolerance). When the drug is removed, the continued presence of the antiopiate peptide in the absence of opiate drug causes the symptoms of withdrawal (Rothman, 1992; Zadina et al., 1994b). When we introduced the antiopiate concept with MIF-1 (Pro-Leu-Gly-NH₂) in 1979, we predicted that other types of antiopiates would be found (Kastin et al., 1979). Among the proposed antiopiates is neuropeptide FF, whose actions in opiate tolerance and dependence have been relatively well characterized. The level of this peptide in the cerebrospinal fluid is increased after chronic morphine (Malin et al., 1990b), and it can attenuate morphine antinociception (Yang et al., 1985) and precipitate withdrawal (Malin et al., 1990a). Antibodies to this peptide can attenuate naloxone-induced withdrawal (Malin et al., 1990b) and reverse morphine tolerance (Lake et al., 1991). It does not, however, bind to opiate receptors and therefore must exert its antiopiate actions through a homeostatic process (Allard et al., 1989; Raffa et al., 1994).

The Tyr-MIF-1 peptides display antiopiate activity and can be described as "opiate modulating," because the peptides exhibit both agonist and antagonist properties, depending on the preparation. As an example of an opiate agonist property, Tyr-W-MIF-1 induces naloxone-reversible analgesia after either intracerebroventricular or intrathecal routes of administration (Gergen et al., 1996a; Zadina et al., 1993b). Also, in SH-SY5Y cells, Tyr-MIF-1 inhibits cAMP in a naloxone-reversible manner (Zadina et al., 1991). Examples of antagonist properties are the ability of Tyr-MIF-1 to precipitate withdrawal (Malin et al., 1993) and to antagonize stress-induced (Galina and Kastin, 1987) as well as morphine-induced analgesia (Kastin et al., 1984).

The opiate modulating nature of the peptides is best demonstrated within a single preparation in which the peptides can exhibit either agonism or antagonism, depending on the conditions of testing. This was first demonstrated with the guinea pig ileum. In ilea from naive animals, the peptides inhibit electrically induced contractions, but in ilea from animals made tolerant to morphine or with "receptor reserve" reduced by alkylating agents, they antagonize opiate-induced contractions (Erchegyi et al., 1992, 1993; Zadina et al., 1992). Tyr-W-MIF-1 can bind to both mu₁ and mu₂ receptors (Zadina et al., 1996) and can induce analgesia in the mouse through a mu₂ mechanism but antagonize mu₁ analgesia induced by morphine and DAMGO (Gergen et al., 1996b).

A key characteristic of Tyr-MIF-1 peptides is their ability to bind to opiate receptors. For an antiopiate to account for the high degree of opiate tolerance that can be attained, it must be able to bind to the opiate receptor (Smith et al., 1988). Tyr-MIF-1 peptides are the only proposed antiopiates (excluding forms of opioids themselves) that have the ability to bind to opiate receptors. Both Tyr-MIF-1 and Tyr-W-MIF-1 bind selectively to mu over delta and kappa receptors (Zadina et al., 1994c). According to the antiopiate peptide model of tolerance and dependence, such peptides counteract the actions of administered opiates. However, the mechanism of this counteraction is unknown and has not been studied at the cellular level. The use of a preparation in which morphine-induced down-regulation has been demonstrated (SH-SY5Y cells) and of a peptide (Tyr-W-MIF-1) which binds to the mu opiate receptor affords a unique opportunity to study this endogenous peptide's ability to counteract the actions of morphine at the cellular and receptor level during chronic administration.

Tyr-W-MIF-1 attenuated morphine-induced down-regulation of the mu receptor in a dose-dependent manner, but it did not affect receptor number when administered alone. These results suggest that it may be working as a partial agonist. Although it binds to the mu receptor, the affinity of Tyr-W-MIF-1 (70 nM) is far less than the affinity of morphine for this receptor (Zadina et al., 1994c). However, at the two higher doses used here, when it is present in a 10-fold or 33-fold excess over morphine, it may be able to compete with morphine for binding to the receptor. If Tyr-W-MIF-1 has lower efficacy at this receptor, it could prevent morphine's agonist activity without having appreciable activity of its own. Thus, by acting as a partial agonist, it could attenuate morphine's actions at the mu receptor and the subsequent down-regulation of this site. This would be in agreement with studies involving the guinea pig ileum, in which Tyr-MIF-1 and Tyr-W-MIF-1 increase their antagonist activity when the receptor reserve is low, such as in tolerant tissue (Erchegyi et al., 1992,1993; Zadina et al., 1992). Met-enkephalin has been shown to negatively modulate morphine analgesia, but this action is blocked by the delta antagonist ICI 174864 and thus does not result from partial agonism at the mu receptor (Jiang et al., 1990). The partial agonism by an endogenous ligand suggested by the present studies therefore is an unusual phenomenon at the opiate receptor.

The attenuation of morphine-induced down-regulation of the delta receptor was unexpected and suggests that the mechanism of action of Tyr-W-MIF-1 at this receptor may not be simple partial agonism. The affinity of Tyr-W-MIF-1 for the delta receptor (15 µM) is 200-fold lower than its affinity for the mu receptor (Zadina et al., 1994c). Even in a 10- and 33-fold excess over morphine, it seems unlikely that the peptide could compete for binding at the delta receptor. It has been suggested that down-regulation of the delta receptor by morphine in SK-N-SH cells occurs through the mu receptor (Baumhaker et al., 1993), which might account for the ability of a mu-acting peptide to attenuate delta down-regulation. However, we have shown previously that in SH-SY5Y cells morphine down-regulates delta receptors even when the mu receptor is blocked by the selective antagonist CTAP (Zadina et al., 1994a). Thus, attenuation of delta receptor down-regulation in this preparation seems unlikely to occur through the mu receptor.

Alternatively, Tyr-W-MIF-1 could be acting through a nonopiate, high-affinity binding site which it shares with Tyr-MIF-1 and which is present in SH-SY5Y cells (Zadina et al., 1990, 1993a). However, Tyr-MIF-1 inhibits cAMP in SH-SY5Y cells in a naloxone-reversible manner, which suggests activity at opiate receptors rather than at this nonopiate site (Zadina et al., 1991). Tests of other signaling pathways and development of antagonists for this site will be necessary to define its role, if any, in the modulation of down-regulation observed in the present studies.

The effects of Tyr-W-MIF-1 on selective agonist-induced down-regulation reveal differences between mu and delta receptors, as well as between morphine and other ligands. The PL017-induced down-regulation of mu receptors was attenuated by Tyr-W-MIF-1, but the DPDPE-induced down-regulation of delta receptors was not. The lack of attenuation at delta receptors could be caused by the high affinity of DPDPE for these receptors (Akiyama et al., 1985) or by its high efficacy in down-regulating delta receptors. We have shown that in SH-SY5Y cells DPDPE down-regulates delta receptors with an IC₅₀ of 0.5 nM and a maximal effect of about 80% reduction of B_max, compared with an IC₅₀ of 179 nM and maximal effect of 62% reduction for PL017 (Zadina et al., 1994a). It could be argued that the inability of Tyr-W-MIF-1 to attenuate DPDPE-induced down-regulation of delta receptors is caused by the high degree of down-regulation achieved by the dose (5 nM) used here. However, this dose is just enough to produce maximal down-regulation, whereas the dose of PL017 chosen (1 µM) is well above that required to achieve maximal down-regulation of mu receptors (Zadina et al., 1994a).

The ability to modulate actions of morphine but not of other ligands, including peptides, has been demonstrated previously for the enkephalin analog DPDPE (Heyman et al., 1989). Such findings, as well as those presented here, suggest differences in the ways cells respond to morphine as compared with other opiate ligands. As mentioned, in vivo studies have demonstrated down-regulation of opiate receptors in response to many ligands, but rarely to morphine. More recently, it has been demonstrated that cloned mu and delta receptors expressed in HEK 293 cells internalize in response to enkephalins and etorphine but not in response to morphine under the same conditions (Keith et al., 1996; Arden et al., 1995). Thus, morphine may induce unique changes in a cell's responsiveness, and the type of cell and presence of modulators can affect the responsiveness.

In summary, we have demonstrated that under appropriate culture conditions an endogenous peptide can attenuate morphine-induced down-regulation. In addition, these findings demonstrate another example of different responses of cells to morphine and other opiate ligands, illustrated by the failure of Tyr-W-MIF-1 to attenuate DPDPE-induced down-regulation even though it can attenuate morphine-induced down-regulation of delta receptors. The in vitro condition, with the addition of this endogenous peptide, may help clarify the in vivo condition, in which significant down-regulation in response to morphine is not easily demonstrated. These findings, therefore, indicate a role for endogenous opiate modulators in the mechanism of tolerance and dependence and suggest a specific cellular mechanism, down-regulation of receptors, that is altered by these peptides.