Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Steroids are considered to be incapable of interacting with high affinity with the three opioid receptors. Steroids from the androgen, glucocorticoid, mineralocorticoid, and gestagen families are ineffective at displacing bound radioligand at concentrations up to 100 µM (LaBella et al., 1978; Schwarz and Pohl, 1994). At best, certain estrogens (for example diethylstilbestrol, 17-estradiol, and 17-dihydroequilenin) are capable of binding to the µ-opioid receptor with very low affinity (LaBella et al., 1978; LaBella, 1985; Schwarz and Pohl, 1994).

That steroids should exhibit such uniformly low affinity for the opioid receptors is only to be expected from the known structure-activity relationships of opioid ligands (Casy and Parfitt, 1986). Crucially, most of the steroids that have been investigated to date with regard to opioid receptor binding have lacked an amine moiety, thus explaining their low affinity. In contrast, the steroid SC17599 (17-acetoxy-6-dim ethylaminomethyl-21-fluoro-3-ethoxy-pregna-3,5-dien-20- one), which contains a tertiary nitrogen (Fig. 1a), possesses marked antinociceptive potency in vivo as measured by the mouse abdominal constriction, the mouse hot-plate, and the rat tail-flick assay (Craig, 1968; Houshyar et al., 2000). Potency in all cases was less than that of morphine. In addition, SC17599 markedly depressed the respiratory rate and increased pCO₂ in rabbits, caused a reduction in gastrointestinal motility and afforded the Straub tail response in mice. The antinociceptive actions of SC17599 are reversed by the µ-selective antagonist methocinnamox (Houshyar et al., 2000).

View larger version (13K): [in this window] [in a new window]   Fig. 1.   Structure and atom/ring numbering of SC17599 (a), morphine (b), and etorphine (c).

SC17599 could be producing these in vivo effects by acting directly at µ-opioid receptors or indirectly through stimulation of the release of endogenous peptides. However, a closely related analog of SC17599 lacking a methyl substituent in the 10-position, SC22000, has been reported to bind to opioid receptors defined with [³H]naloxone in mouse brain, albeit with affinity 30-fold less than morphine (LaBella et al., 1978). If SC17599 is indeed exerting its antinociceptive and other opioid actions through a direct interaction with the µ-opioid receptor, it represents a highly novel structure for a µ-opioid ligand. Although SC17599 possesses a tertiary nitrogen, it lacks both the critical aromatic feature and a phenolic hydroxyl substituent corresponding to the one typically found in opioid peptides and morphine-like opioids (Fig. 1, b and c; Morgan et al., 1976; Casy and Parfitt, 1986; Lomize et al., 1996).

In the current study, we show the selective interaction of SC17599 with the µ-opioid receptor using radioligand binding and confirm the agonist activity of the compound using [³⁵S]guanosine-5'-O-(3-thio)triphosphate (GTPS) binding assays in a variety of cell lines and in rat brain membranes. Molecular modeling techniques explain the relationship between SC17599 and more traditional opioid ligands and a possible binding mode of the steroid within the µ-opioid receptor binding pocket.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals and Drugs. [³H][D-Ala²,N-Me-Phe⁴,Gly⁵-ol]-Enkephalin (DAMGO; 54.5 Ci/mmol; 2.02 TBq/mmol) and [³H]diprenorphine (45 Ci/mmol; 1.66 TBq/mmol or 58 Ci/mmol; 2.14 TBq/mmol) were from Amersham International (Aylesbury, UK or Piscataway, NJ). [³H]CI 977 [5R-(5,7,8)-N-methyl-N-[7-(1-pyrrolidinyl)-1-oxaspiro[4.5]dec-8-yl]-4-benzofuranacetamide] (21.1 Ci/mmol; 0.78 Tbq/mmol) was a kind gift from Dr. J. C. Hunter (Parke-Davis Neuroscience Research Center, Cambridge, UK). [³⁵S]GTPS (1250 Ci/mmol; 46.25 TBq/mmol) and [³H]triamcinolone (TA; 38 Ci/mmol; 1.41 TBq/mmol) were purchased from DuPont NEN (Hounslow, UK or Boston, MA). The following drugs were generous gifts from the National Institute on Drug Abuse (Rockville, MD): fentanyl HCl, naloxone HCl, and naltrexone HCl. BW 373,U86 [(±)-[1(S*),2,5]-4-{[2,5-dimethyl-4-(2-propenyl)-1-piperazinyl]-(3-hydroxyphenyl)methyl}-N,N-diethyl-benzamide hydrochloride] was from Burroughs Wellcome (Research Triangle Park, NC). Morphine sulfate was purchased from Mallinckrodt (St. Louis, MO). SC17599 free base (17-acetoxy-6-dimethylaminomethyl-21-fluoro-3-ethoxy-pregna-3,5-dien-20-one) was a kind gift from G. D. Searle and Co. (Chicago, IL). 5-Pregnan-3-ol-20-one, 5-pregnan-3-ol-20-one, 17-estradiol, 17-estradiol, estrone, hydrocortisone, and dexamethasone were purchased from Sigma Chemical Co. (St. Louis, MO). U69,593 [5,7,8-()-N-[7-(1-pyrrolidinyl)-1-oxaspiro(4,5)dec-8-yl]benzeneacetamide] was from RBI (Natick, MA). Dulbecco's modified Eagle's medium (without sodium pyruvate; with 4500 mg/l glucose), minimum essential medium (with Earle's salts), fetal calf serum, penicillin/streptomycin, Fungizone, trypsin, EDTA, and Geneticin were all from GIBCO Life Sciences (Grand Island, NY). All other chemicals were of analytical grade and were purchased from Sigma Chemical Co.

Cell Membrane Preparation. Undifferentiated human neuroblastoma SH-SY5Y cells were a kind donation from Dr. D. Lambert (Department of Anesthesia, Leicester University, UK). Cells (passages 75-90) were cultured in minimum essential medium supplemented with 10% fetal calf serum, 2.5 µg/ml amphotericin B (Fungizone), 50 µg/ml penicillin/streptomycin, and 250 µg/ml L-glutamine at 37°C in a humidified 5% CO₂ atmosphere. C6 glioma cells transfected with the cloned rat µ-opioid receptor (C₆µ cells) were a kind donation from Dr. Huda Akil (Mental Health Research Institute, University of Michigan, MI). Cells (passages 15-25) were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum at 37°C under a 5% CO₂ atmosphere. For subculture, one flask from each passage was grown in the presence of 1 mg/ml Geneticin. Cells used for experiments were grown in the absence of Geneticin with no significant reduction in receptor number. Chinese hamster ovary (CHO) cells transfected with the mouse -opioid receptor were a kind gift from Dr. C. J. Evans (Department of Psychiatry, UCLA, Los Angeles, CA). Cells were cultured in Dulbecco's modified Eagle's medium supplemented with 5% fetal calf serum, 2.5 µg/ml Fungizone, 50 U/ml penicillin, 50 µg/ml streptomycin, and 258 µg/ml L-glutamine at 37°C in a humidified 5% CO₂ atmosphere. CHO cells expressing the rat µ-opioid receptor and its D147N mutant (Li et al., 1999) were established as previously described. Cells were grown in Dulbecco's modified Eagle's medium/Ham's F12 medium supplemented with 10% fetal calf serum and 0.2 mg/ml Geneticin at 37°C in a humidified 5% CO₂ atmosphere. In all cases, cells were grown to confluence and then harvested in HEPES (20 mM, pH 7.4)-buffered saline containing 1 mM EDTA, dispersed by agitation and collected by centrifugation at 1600 rpm. The cell pellet was suspended in 50 mM Tris-HCl buffer, pH 7.4, and homogenized using a Tissue Tearor. The resultant homogenate was centrifuged for 15 min at 18,000 rpm at 4°C and the pellet collected, washed, resuspended, and recentrifuged. The pellet was resuspended in 50 mM Tris-HCl buffer, pH 7.4, at a final protein concentration of 100 to 200 µg/ml (Lowry et al., 1951). Cytosol fractions from Sf9 cells stably transfected with the glucocorticoid receptor were kindly supplied by Dr. W. Pratt (Department of Pharmacology, University of Michigan, MI).

Brain Membrane Preparation. Male Sprague-Dawley rats and Duncan-Hartley guinea pigs, (300 g; Harlan, Indianapolis, IN) were housed in standard laboratory cages (three per cage) in a temperature-controlled colony room maintained on a 12-h light/dark cycle. Food (Purina Rodent Chow; Purina Mills, St. Louis, MO) and water were available ad libitum until testing. Studies were carried out in accordance with the Declaration of Helsinki and with the Guide for the Care and Use of Laboratory Animals as adopted and promulgated by the National Institutes of Health.

Radioligand Binding Assays. Displacement of bound radioligand from cell membranes or guinea pig brain homogenates was performed as follows. C₆µ cell membranes (30-60 µg of protein), SH-SY5Y cell membranes (100-150 µg of protein), CHO cell membranes (50-80 µg of protein), or guinea pig brain homogenates (400 µg of protein) were incubated at 25°C in 50 mM Tris-HCl buffer, pH 7.4, for 1 h with radiolabeled ligand and varying concentrations of competing ligand to give a final volume of 1 ml. Radiolabeled ligand concentrations used were: [³H]DAMGO, 1.0 nM; [³H]diprenorphine, 0.2 nM; [³H]CI 977, 0.5 nM. For CHO cells expressing the wild-type µ-opioid receptor or the D147N mutant, binding was determined as above using 35 to 55 µg of protein and either 0.2 nM [³H]diprenorphine (wild-type) or 2 nM [³H]diprenorphine (D147N mutant) approximating to the affinity of [³H]diprenorphine for the two receptors (K_d wild-type = 0.24 nM, K_d mutant = 2.4 nM). Wild-type and mutant receptors were expressed at a similar level (approximately 1.5 pmol/mg of protein; Li et al., 1999). In all cases, total binding was determined in the absence of unlabeled ligand, and nonspecific binding was defined by naloxone (10 µM). Bound and free radioligand were separated by vacuum filtration through glass fiber filters and quantified by liquid scintillation counting.

[³⁵S]GTPS Binding Assay. Agonist stimulation of [³⁵S]GTPS binding in cell lines containing cloned receptors was measured as previously described (Traynor and Nahorski, 1995). Briefly, C₆µ cell membranes (30-60 µg of protein) were incubated at 30°C in buffer A, pH 7.4, for 1 h with [³⁵S]GTPS (100 pM), GDP (10 µM), and varying concentrations of ligand to give a final volume of 1 ml. Basal binding of [³⁵S]GTPS was determined in the absence of unlabeled ligand, and maximal stimulation was defined by fentanyl (10 µM). Bound and free [³⁵S]GTPS were separated by vacuum filtration through glass fiber filters and quantified by liquid scintillation counting. Buffer A comprised HEPES (20 mM), MgCl₂·6H₂O (10 mM), and NaCl (100 mM).

Molecular Modeling. All computations were carried out on either an SGI Oxygen R10,000 workstation or an SGI Octane workstation (Silicon Graphics, Mountain View, CA) using SYBYL 6.4.3 from Tripos Inc. (St. Louis, MO). In all calculations, parameters were as default except where noted.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Radioligand Binding Studies. SC17599 displaced [³H]diprenorphine from membranes of C₆µ cells with a K_i value of 62.3 ± 5.4 nM (Fig. 2a), approximately 5.5-fold lower than that shown by morphine (K_i = 11.7 ± 2.6 nM). In membranes from SH-SY5Y cells, SC17599 displaced both the selective µ-agonist [³H]DAMGO and the nonselective antagonist [³H]diprenorphine in a concentration-dependent manner, giving higher affinity values than using C₆µ cells (Table 1). When the binding buffer was changed from Tris buffer to buffer A, which contains 100 mM Na⁺ ions, the concentration-response curve for the displacement of [³H]diprenorphine from C₆µ cell membranes was shifted to the right in parallel fashion by approximately 7.7-fold (Fig. 2b). Similarly, a 7.0-fold rightward parallel shift was seen in the displacement of [³H]diprenorphine in SH-SY5Y cell membranes (Table 1). SC17599 produced a concentration-dependent displacement of both [³H]diprenorphine from CHO cell membranes and of [³H]CI 977 from guinea pig brain homogenate but with very low affinity, affording K_i values of 2348 ± 509 nM and 1950 ± 928 nM at - and -receptors, respectively (Fig. 2c). This gives greater than 100-fold selectivity for the µ-opioid receptor over the - and -opioid receptors (Table 1).

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Radioligand binding of SC17599. a, displacement of the binding of [3H]diprenorphine (0.2 nM) from membranes of C6µ cells by morphine () and SC17599 () in Tris buffer. Data shown for morphine were taken from previous work in this laboratory (Lee et al., 1999). b, displacement of the binding of [3H]diprenorphine (0.2 nM) from membranes of C6µ cells by SC17599 in the absence () (Tris-HCl buffer) and presence of 100 mM Na+ () (buffer A). For buffer composition, see Materials and Methods. c, displacement by SC17599 of either bound [3H]diprenorphine from membranes of CHO cells () or bound [3H]CI 977 from guinea pig brain homogenates (). All assays were performed as described under Materials and Methods. Data represents the mean ± S.E.M. from three or more separate experiments performed in duplicate.

View this table: [in this window] [in a new window]   TABLE 1 Displacement of bound radioligands by SC17599 in SH-SY5Y cell membranes Membranes from SH-SY5Y cells were incubated with radioligand ([3H]diprenorphine 0.2 nM or [3H]DAMGO 1.0 nM) and varying concentrations of SC17599 in either 50 mM Tris-HCl buffer or buffer A containing 100 mM Na+ ions (see Materials and Methods). Values are means ± S.E.M. from three or more separate experiments performed in duplicate. Selectivities were calculated using affinities for the - and -receptors, determined as described in Fig. 2c.

[³⁵S]GTPS Binding Studies. In SH-SY5Y cell membranes SC17599 stimulated the binding of [³⁵S]GTPS with an EC₅₀ of 282.3 ± 42.4 nM and maximal response equivalent to that of the µ-agonist fentanyl (10 µM) (Fig. 3a). This effect was antagonized by naloxone (10 nM), which shifted the concentration-response curve to the right by approximately 5.6 fold, affording an apparent K_e value for naloxone of 2.2 nM (Fig. 3a). In membranes from C₆µ cells, SC17599 stimulated [³⁵S]GTPS binding with an EC₅₀ value of 110 ± 6.2 nM, compared with DAMGO (EC₅₀ = 38.3 ± 3.4 nM) and morphine (EC₅₀ = 23.9 ± 2.2 nM). The maximal degree of stimulation afforded by SC17599 was 95.9% (95% confidence limits, 91.9-99.9). This was not significantly different from the 95.5% (95% confidence limits, 91.9-99.9) stimulation afforded by morphine or the 104.8% (95% confidence limits, 92.8-117.0) stimulation evoked by the highly efficacious µ-peptide DAMGO (Fig. 3b).

View larger version (17K): [in this window] [in a new window]   Fig. 3.   a, stimulation of [35S]GTPS (100 pM) binding to membranes of SH-SY5Y cells by SC17599 in the absence () and presence () of naloxone (10 nM). b, stimulation of [35S]GTPS (100 pM) binding to membranes of C6µ cells by SC17599 (), morphine (), and DAMGO (). Assays were performed as described under Materials and Methods. Data represent the mean ± S.E.M. from three or more separate experiments each performed in duplicate.

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Stimulation of [35S]GTPS binding in rat brain membranes by DAMGO (1 µM) and SC17599 (1 µM) and reversal by naltrexone (30 µM). Assays were performed as described under Materials and Methods. Values are means ± S.E.M. of three experiments each performed in triplicate.

Molecular Modeling. Various combinations of morphine, the potent opioid etorphine, and SC17599 (structures shown in Fig. 1) were analyzed using the GASP (Genetic Algorithm Superposition Program) module of SYBYL. Automated pharmacophore generation using morphine and SC17599 alone gave very poor steric overlap (data not shown), whereas overlap between the more complex etorphine and SC17599 was excellent (Fig. 5). The A, B, and C rings of the steroid coincide with, and are broadly coplanar with, the A, B, and E rings of etorphine. The quaternary nitrogens of each molecule are in close proximity and interact with the same hypothetical site point. The steroid D ring and its substituents project beyond the volume occupied by the morphinan skeleton, into the space corresponding to the etorphine 7-substituent. Here there is another theoretical site point that interacts with the oxygen atoms present in the 19-hydroxyl substituent of etorphine and the 17-acetoxyl group of SC17599. Overall steric overlap is excellent; the only moieties that project beyond the shared volume of the two molecules are the 19-methyl substituent of etorphine and the 17-fluoroacetone and 3-ethoxy groups of SC17599. Pharmacophore generation using morphine, etorphine, and SC17599 together gave an overlap very similar to that produced by the analysis of etorphine and SC17599 alone (data not shown).

View larger version (27K): [in this window] [in a new window]   Fig. 5.   Overlap of SC17599 (orange) and etorphine (magenta) generated by GASP. Nitrogen is shown in blue, and hypothetical site points in green.

View larger version (40K): [in this window] [in a new window]   Fig. 6.   Docking of morphine (magenta) and SC17599 (orange) to the µ-opioid receptor model. Only selected residues in the binding pocket are shown for clarity (white). Nitrogen is shown in blue, oxygen in red, and hydrogen in cyan. Shown in yellow is possible hydrogen bonding between the N+ of both molecules and Asp-147 in TMD III.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Radioligand Binding Studies. SC17599 bound with good affinity and marked preference for the µ-opioid receptor over the - and -opioid receptors. In membranes from C₆µ cells and CHO cells expressing the µ-opioid receptor, affinity remained good but was somewhat reduced. The reasons underlying this discrepancy are unclear, but may relate to the use of the oripavine [³H]diprenorphine in the radioligand binding assays (Lee et al., 1999). The binding of morphine and SC17599 to the µ-opioid receptor in CHO cells was highly dependent on the presence of Asp-147 in TMD III, because mutation of this residue to Asn considerably reduced binding affinity.

[³⁵S]GTPS Binding Studies. The [³⁵S]GTPS binding assay provides a functional measure of agonist occupation of µ-opioid receptors, allowing for determination of potency and relative efficacy (Traynor and Nahorski, 1995). The µ-agonist properties of SC17599 were confirmed by its ability to stimulate [³⁵S]GTPS binding in membranes from both SH-SY5Y and C₆µ cell lines. The potency of SC17599 was 3- to 4-fold lower than morphine or DAMGO but maximal stimulation was equivalent to that afforded by the full µ-agonist DAMGO.

Molecular Modeling. SC17599 is able to bind to µ-opioid receptors despite lacking any aromatic or hydroxyl feature, making it a highly unusual opioid agonist. An aromatic feature is considered critical in traditional opioid pharmacophores, whereas a phenolic hydroxyl group is thought to be crucial to the activity of morphine-like and peptide opioid ligands (Casy and Parfitt, 1986). On the other hand, SC17599 does have a tertiary nitrogen, several oxygen functionalities, and a relatively planar electron-rich area that may be able to substitute for the pharmacophoric elements of more traditional opioid ligands.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Revised pharmacophore for µ-opioid ligands, based on comparisons by GASP (see Materials and Methods) of the structures of SC17599, morphine, and etorphine.

Ligand Docking. To examine whether this pharmacophore was useful in predicting the binding mode of SC17599, the docking of this unusual ligand to the µ-opioid receptor was examined. Using a genetic algorithm for ligand docking, two cavities were identified within the µ-opioid receptor. Only the larger pocket within the TMD helix bundle is likely to be involved in ligand binding because the other lies in the extracellular loops and is very small. Morphine docks almost entirely within the identified binding pocket in an orientation that gives very close agreement to one binding mode for morphine to the same model (Pogozheva et al., 1998). In this position the quaternary nitrogen of morphine is involved in a hydrogen-bonding interaction with Asp-147 in TMD III. Despite its larger steric bulk, SC17599 is also able to bind within the same pocket in an orientation that allows interaction between the quaternary nitrogen and Asp-147. Comparison of the docked alignments of the two ligands shows that the A ring of SC17599 is shifted approximately 3.25 Å within the plane of both ligands such that it now overlaps the B ring of morphine. Despite this shift in the position of the steroid skeleton, the quaternary nitrogen is still appropriately positioned to form a hydrogen bond with Asp-147 with an H-O distance of 2.85 Å and an N-H-O angle of 162.6°. Alternatively, alteration of the configuration of the side chain of Asp-147 allows for an ionic interaction between the steroid N⁺ and the Asp carboxyl, with a N⁺... . O distance of 5.1 Å. The importance of Asp-147 in the binding of SC17599 to the µ-opioid receptor was confirmed by the large loss in affinity seen at the D147N mutant compared with the wild-type receptor. The finding that a larger loss of affinity was seen with morphine than with SC17599 probably relates to the fact that SC17599 contacts more residues than morphine and so the contribution of the interaction with Asp-147 to the overall binding affinity is reduced.