Abstract
Presynaptic nicotinic acetylcholine receptors (nAChRs) on striatal synaptosomes stimulate dopamine release. Partial inhibition by the α3β2-selective α-conotoxin-MII indicates heterogeneity of presynaptic nAChRs on dopamine terminals. We have used this α-conotoxin and UB-165, a novel hybrid of epibatidine and anatoxin-a, to address the hypothesis that the α-conotoxin-MII-insensitive subtype is composed of α4 and β2 subunits. UB-165 shows intermediate potency, compared with the parent molecules, at α4β2* and α3-containing binding sites, and resembles epibatidine in its high discrimination of these sites over α7-type and muscle binding sites. (±)-Epibatidine, (±)-anatoxin-a, and (±)-UB-165 stimulated [3H]-dopamine release from striatal synaptosomes with EC50 values of 2.4, 134, and 88 nm, and relative efficacies of 1:0.4:0.2, respectively. α-Conotoxin-MII inhibited release evoked by these agonists by 48, 56, and 88%, respectively, suggesting that (±)-UB-165 is a very poor agonist at the α-conotoxin-MII-insensitive nAChR subtype. In assays of 86Rb+ efflux from thalamic synaptosomes, a model of an α4β2* nAChR response, (±)-UB-165 was a very weak partial agonist; the low efficacy of (±)-UB-165 at α4β2 nAChR was confirmed in Xenopus oocytes expressing various combinations of human nAChR subunits. In contrast, (±)-UB-165 and (±)-anatoxin-a were similarly efficacious and similarly sensitive to α-conotoxin-MII in increasing intracellular Ca2+ in SH-SY5Y cells, a functional assay for native α3-containing nAChR. These data support the involvement of α4β2* nAChR in the presynaptic modulation of striatal dopamine release and illustrate the utility of exploiting a novel partial agonist, together with a selective antagonist, to dissect the functional roles of nAChR subtypes in the brain.
- neuronal nicotinic acetylcholine receptor
- presynaptic nicotinic modulation
- dopamine release
- rat striatal synaptosomes
- Xenopus oocytes
- SH-SY5Y cells
- α-conotoxin
Nicotinic acetylcholine receptors (nAChRs) are widely distributed in the vertebrate CNS. With a few recently reported exceptions (Alkondon et al., 1998; Frazier et al., 1998), most nAChRs in the brain do not appear to mediate synaptic transmission. Instead, their primary function may be modulatory (Role and Berg, 1996). One locus for modulation is the nerve terminal, where presynaptic nAChRs can promote transmitter release and hence influence resting tone or synaptic efficacy.
Presynaptic nAChRs facilitate the release of many neurotransmitters, in numerous brain regions, via various nAChR subtypes (Wonnacott, 1997). nAChR heterogeneity arises from the pentameric assembly of receptors from numerous α and β subunits (α2–α7, β2–β4) expressed in the mammalian brain (Role and Berg, 1996; Lukas et al., 1999). Defining the subunit composition of native nAChR is a major challenge; this quest is hampered by a lack of subtype-specific tools. A recent advance has been the identification of Conus toxins that target particular neuronal nAChR subtypes (McIntosh et al., 1999). α-Conotoxin-MII, with specificity for α3β2-containing nAChRs (Cartier et al., 1996), partially inhibits the nicotinic stimulation of [3H]-dopamine release from striatal preparations (Kulak et al., 1997; Kaiser et al., 1998), indicating heterogeneity of the nAChR mediating this response.
The nicotinic modulation of [3H]-dopamine release from striatal preparations has been exploited as a model system for examining native nAChR responses (Soliakov and Wonnacott, 1996; Grady et al., 1997) and evaluating novel ligands (Holladay et al., 1997) and is pertinent to physiological and pathological processes (Dani and Heinemann, 1996; Decker and Arneric, 1998). Nicotinic agonists elicit dopamine release from rodent striatal synaptosomes and slices in a concentration-dependant manner, and this response is blocked by nicotinic antagonists, including mecamylamine and dihydroβerythroidine (Grady et al., 1992; El-Bizri and Clarke, 1994; Sacaan et al., 1995; Soliakov et al., 1995). Insensitivity to the α7-selective antagonists α-bungarotoxin (αBgt) (Rapier et al., 1990; Grady et al., 1992) and α-conotoxin-ImI (Kulak et al., 1997) argues against the direct involvement of α7 nAChR. The loss of [3H]-nicotine binding sites from the striatum after 6-hydroxydopamine lesion of the nigrostriatal pathway is consistent with α4β2 nAChRs on striatal terminals (Clarke and Pert, 1985). Alternatively, sensitivity to neuronal bungarotoxin was interpreted in favor of α3-containing nAChRs (Grady et al., 1992). These disparate views are reconciled by the nAChR heterogeneity implicit in the partial inhibition by α-conotoxin-MII (Kulak et al., 1997; Kaiser et al., 1998). The selectivity of this toxin is consistent with an nAChR containing α3 and β2 subunits (Cartier et al., 1996; Kaiser et al., 1998); the α4β2* nAChR is a candidate for the α-conotoxin-MII-insensitive component of dopamine release evoked by nicotinic agonists.
Here we report studies with the novel nicotinic ligand UB-165 (Wright et al., 1997) (see Fig. 1) that support this interpretation. UB-165 is a hybrid of anatoxin-a and epibatidine, with potency at rat brain [3H]-nicotine binding sites that is intermediate between the values of the parent molecules. We have extended the characterization of UB-165 and have exploited its limited agonism at α4β2 nAChRs to examine the putative contribution of this subtype to the nicotinic stimulation of striatal dopamine release.
MATERIALS AND METHODS
Materials
Cell culture. SH-SY5Y cells were from European Collection of Animal Cell Cultures (Porton Down, Salisbury, Wiltshire, UK) and were cultured as described byMurphy et al. (1991). M10 cells were provided by Dr. P. Whiting (Merck, Sharp and Dohme Research Center, Harlow, Essex, UK) and were cultured as described previously (Whiteaker et al., 1998). Cell culture media were provided by Life Technologies (Paisley, Renfrewshire, Scotland), and tissue culture plastic ware was obtained from Becton Dickinson UK Ltd. (Oxford, UK) and Sterilin (Stone, Staffordshire, UK).
Drugs and reagents.(−)-[3H]-Nicotine (3.0 TBq/mmol in ethanol) and (±)-[3H]-epibatidine (2.1 TBq/mmol in ethanol) were provided by Dupont NEN (Stevenage, Herts, UK) and stored at −20°C. 86RbCl was obtained from Dupont NEN (Herts, UK or Boston, MA); it was stored at 20°C and used within 1 month. [7,8-3H]-dopamine (specific activity, 1.78 TBq/mmol) was purchased from Amersham International (Buckinghamshire, UK) and stored at −20°C. Na125I from Amersham International was used to iodinate αBgt to a specific activity of 26 TBq/mmol. (±)-Epibatidine was purchased from RBI (Natick, MA), and (±)-anatoxin-a was obtained from Tocris Cookson (Bristol, UK). Racemic UB-165 (Wright et al., 1997) and α-conotoxin-MII (Cartier et al., 1996; Kaiser et al., 1998) were synthesized as previously described. All other drugs and reagents were provided by Sigma (Poole, Dorset, UK).
Tissue preparations
Rat brain membranes. P2 membranes from whole rat brain (minus cerebellum) were prepared as previously described (Davies et al., 1999). Briefly, brains were homogenized (10% w/v) in ice-cold 0.32 m sucrose, pH 7.4, containing 1 mm EDTA, 0.1 mm PMSF, and 0.01% NaN3, and centrifuged at 1000 × g for 10 min. The supernatant fraction (S1) was decanted and retained on ice. The pellet (P1) was resuspended in ice-cold 0.32 m sucrose (5 ml/g original weight) and recentrifuged at 1000 × g for 10 min. The supernatant was combined with S1 and centrifuged at 12,000 × g for 30 min. The pellet (P2) was resuspended (2.5 ml/g original weight) in phosphate buffer (50 mm potassium phosphate, pH 7.4, containing 1 mm EDTA, 0.1 mm PMSF, and 0.01% NaN3), and washed twice by centrifugation at 12,000 × g for 30 min. The washed pellet was resuspended in phosphate buffer (2.5 ml/g original weight) and stored in 5 ml aliquots at −20°C.
Rat muscle extract preparation. A Triton X-100 extract of muscle from the hindlimbs of Wistar rats was prepared as previously described (Garcha et al., 1993).
SH-SY5Y cell membrane preparation. SH-SY5Y cells, grown to confluency in 175 cm2 flasks, were washed briefly with warm PBS containing (in mm): (150 NaCl, 8 K2HPO4, 2 KH2PO4, pH 7.4, 37°C) and scraped into cold phosphate buffer. Cells were washed by centrifugation for 3 min at 500 × g and resuspended in 10 ml of ice-cold phosphate buffer. The suspension was homogenized for 10 sec using an Ultraturax and centrifuged for 30 min at 45,000 ×g. The pellet was resuspended in phosphate buffer (0.5 ml per original flask).
Radioligand binding assays
(−)-[3H]-nicotine competition binding assays: rat brain membranes. P2 membranes (250 μg protein) were incubated in a total volume of 250 μl in HEPES buffer (20 mm HEPES, pH 7.4, containing 118 mm NaCl, 4.8 mm KCl, 2.5 mm CaCl2, 200 mm Tris, 0.1 mm PMSF, 0.01% (w/v) NaN3) (Romm et al., 1990) with 10 nm(−)-[3H]-nicotine and serial dilutions of test drugs. Nonspecific binding was determined in the presence of 100 μm (−)-nicotine. Samples were incubated for 30 min at room temperature followed by 1 hr at 4°C. The reaction was terminated by filtration through Whatman GFA/E filter paper (presoaked overnight in 0.3% polyethyleneimine in PBS), using a Brandel Cell Harvester. Filters were counted for radioactivity in 5 ml Optiphase “Safe” scintillant in a Packard Tricarb 1600 scintillation counter (counting efficiency 45%).
(±)-[3H]-epibatidine competition binding assays. SH-SY5Y cells. SH-SY5Y membranes (30 μg protein) were incubated in a total volume of 2 ml in 50 mmphosphate buffer with 150 pm(±)-[3H]-epibatidine and serial dilutions of test drugs. Nonspecific binding was determined in the presence of 100 μm (−)-nicotine. Samples were incubated for 2 hr at 37°C. They were filtered and counted for radioactivity as described above.
[125I]-αBgt competition binding assays: rat brain membranes. P2 membranes (250 μg protein) were incubated in a total volume of 250 μl in 50 mm phosphate buffer with 1 nm[125I]-αBgt and serial dilutions of test drugs (Davies et al., 1999). Nonspecific binding was determined in the presence of 10 μm αBgt. Samples were incubated for 3 hr at 37°C. Ice-cold PBS (0.5 ml) was then added, and the samples were centrifuged for 3 min at 10,000 × g. Pellets were washed by resuspension in 1.25 ml PBS and centrifugation as before. The resultant pellets were counted for radioactivity in a Packard Cobra II auto-gamma counter.
[125I]-αBgt competition binding assays: rat muscle extract. Rat muscle extract (1.5 mg protein) was incubated in a total volume of 500 μl in 2.5 mm sodium phosphate buffer, pH 7.4, with 1 nm[125I]-αBgt and serial dilutions of test drugs (Garcha et al., 1993). Nonspecific binding was determined in the presence of 10 μm αBgt. Samples were incubated for 2 hr at 37°C. Bound radioligand was separated by filtration through Whatman GF/C filters (presoaked in 0.3% polyethyleneimine in PBS overnight) using a Millipore vacuum manifold. Filters were washed three times with 3 ml cold PBS supplemented with 0.1% BSA and counted for radioactivity in a Packard Cobra II auto-gamma counter.
[3H]-epibatidine binding to M10 cells.Total numbers of nicotinic binding sites in M10 cells were measuredin situ, in cultures in 24-well plates incubated with 500 pm(±)-[3H]-epibatidine for 2 hr, as previously described (Whiteaker et al., 1998). Competition binding assays were performed similarly, using 200 pm[3H]-epibatidine and serial dilutions of test drugs.
Upregulation of [3H]-epibatidine binding sites in M10 cells
M10 cells grown were grown in 24-well plates to ∼70% confluency and were then incubated for 48 hr at 37°C with dexamethasone (to induce nAChR expression) in medium containing serial dilutions of test drugs (Whiteaker et al., 1998). Control cells, treated with dexamethasone in medium without the addition of nicotinic agents, were incubated in parallel. A rigorous washing procedure, in which medium was replaced three times at hourly intervals, was used to ensure removal of nicotinic drugs (Whiteaker et al., 1998) before assaying for [3H]-epibatidine binding sites as outlined above.
Superfusion of rat striatal synaptosomes for [3H]-dopamine release
P2 synaptosomes were prepared from rat striata, loaded with [3H]-dopamine (0.1 μm, 0.132 MBq/ml) for 15 min at 37°C, and superfused in open chambers as previously described (Soliakov et al., 1995; Kaiser et al., 1998). Synaptosomes were superfused with Krebs bicarbonate buffer containing (in mm): 118 NaCl, 2.4 KCl, 2.4 CaCl2, 1.2 MgSO4, 1.2 K2HPO4, 25 NaHCO3 and 10 glucose, titrated to pH 7.4 with 95% O2/5% CO2, supplemented with 1 mm ascorbic acid, 8 μmpargyline, and 0.5 μm nomifensine to prevent dopamine degradation and reuptake. Agonists were applied for 40 sec. In antagonist studies, α-conotoxin-MII (112 nm) or mecamylamine (10 μm) was added to the superfusion buffer 10 min before the application of agonist and maintained throughout the remainder of the experiment. Experiments always included chambers stimulated in parallel with 1 μm (±)-anatoxin-a, as a standard for normalization of data between experiments.
86Rb+ efflux from thalamic synaptosomes
86Rb+efflux experiments were performed essentially as described by Marks et al. (1996). P2 synaptosomes were prepared from mouse or rat thalamus by homogenization in 0.32 m sucrose in 5 mm HEPES, pH 7.5, and differential centrifugation (Soliakov et al., 1995). Synaptosomes were loaded with86Rb+(sufficient to give ∼70 MBq per chamber) by incubation for 30 min at 22°C in uptake buffer containing (in mm): 140 NaCl, 1.5 KCl, 2.0 CaCl2, 1.0 MgS04, 25 HEPES, 20 glucose, pH 7.5. Uptake was terminated, and unincorporated86Rb+ was removed by transferring aliquots (mouse, 25 μl; rat, 35 μl) to glass fiber filters for gentle filtration and washing. One thalamus provided sufficient material for up to 8 (mouse) or 12 (rat) filters. Each filter was placed in an open chamber of a superfusion apparatus (Marks et al., 1993; Soliakov et al., 1995). Samples were perfused at a rate of 2.5 ml/min with physiological buffer [135 mmNaCl,, 1.5 mm KCl, 2.0 mmCaCl2, 1.0 mmMgS04, 20 mm glucose, 25 mm HEPES, pH 7.5, containing 0.1% (w/v) BSA; 5 mm CsCl and 50 nM tetrodotoxin]. After perfusion for 6 min, 12 samples were collected at 30 sec intervals. Agonist stimulation (60 sec) was given 90 sec after the start of sample collection. Where used, antagonists were present in the perfusion buffer throughout the experiment.
Xenopus oocyte preparation and recording
Stage V–VI oocytes were isolated from anesthetizedXenopus laevis frogs and enzymatically defolliculated by gentle shaking with collagenase [Worthington (Freehold, NJ), Type II, 1.7 mg/ml for 90 min; then Sigma (St. Louis, MO), Type II, 1.7 mg/ml for 30 min] in a Ca2+-free Barth's solution. After defolliculation, oocytes were incubated at 16–19°C in a solution containing (in mm): 77.5 NaCl, 2 KCl, 1.8 CaCl2, 1 MgCl2, 5 HEPES, pH 7.5, adjusted with NaOH, and supplemented with 100 U/ml penicillin, 100 μg/ml streptomycin, 50 mg/ml gentimycin, and 5% heat-inactivated horse serum. Oocytes were injected the next day with 10–50 ng of mRNA encoding αx and βx nAChR subunits in an injection volume of 50 nl. The human nAChR subunits α2, α3, α4–2, β2, β4, and α7 were cloned from cDNA libraries prepared from human brain and the human IMR32 neuroblastoma cell line (Elliott et al., 1996). For two-way combinations, RNA was injected at a ratio of 1:1 (25 ng of each subunit per egg). RNA for α7 was injected at a concentration of 25 ng per egg.
Oocytes were examined for functional expression 2–5 d after mRNA injection using two-electrode voltage-clamp techniques described previously (Chavez-Noriega et al., 1997). Agonist-induced currents were elicited at a holding potential of −60 mV. The recording solution contained (in mm): 115 NaCl, 2.5 KCl, 1.8 CaCl2, 10 HEPES, 0.001 atropine, pH 7.3. Recording electrodes (0.5–2.5 MΩ resistance) were filled with 3m KCl. Perfusion solutions were gravity fed into the recording chamber (capacity, 100 μl) at a rate of ∼6–10 ml/min. All recordings were performed at room temperature (19–23° C). Signals were amplified, digitized (100–500 Hz), and filtered (at 40–200 Hz).
Calcium fluorimetry
SH-SY5Y cells, grown to confluency in 175 cm2 flasks, were removed by incubation in Ca2+-free PBS for 3 min at 37°C. The cell suspension was centrifuged for 3 min at 500 ×g, and the pellet was resuspended in 3–4 ml Ca2+-free HEPES buffer containing (in mm): 145 NaCl, 5 KCl, 1 MgCl2, 0.5 Na2HPO4, 5.5 glucose, and 10 HEPES, pH 7.4, containing 5 μm fura-2 AM. The suspension was incubated in darkness at room temperature for 45 min, and excess fura-2 AM was removed by centrifugation for 3 min at 500 × g followed by three wash/centrifugation steps. Cell density was adjusted to 1–2 × 106 cells/ml, and a 2 ml aliquot of cell suspension was placed in a cuvette in a PTI dual-excitation spectrophotofluorimeter (Photon Technology International, South Brunswick, NJ). To the cell suspension, 2 mmCaCl2 was added before applying a nicotinic agonist ((±)-UB-165 or (±)-anatoxin-a). Where used, the antagonist α-conotoxin-MII (112 nm) or mecamylamine (10 μm) was added to the cuvette 5 min before the addition of agonist. Excitation at 340 and 380 nm and emission at 510 nm were monitored. Calibration was performed in each experiment by adding sequentially 0.5% (v/v) Triton X-100 and 10 mm EDTA to derive maximum and minimum fluorescence ratios, respectively. Results were normalized within each experiment with respect to a maximally stimulating concentration of agonist.
Data analysis
Competition binding. IC50values were calculated by fitting data points to the Hill equation, using the nonlinear least squares curve fitting facility of Sigma Plot V2.0 for windows. Ki values were derived from IC50 values according to the method of Cheng and Prusoff (1973), assumingKd values of 10 and 1 nm for [3H]-nicotine and [125I]-αBgt binding to rat brain membranes and 0.12 nm for [125I]αBgt binding to rat muscle, respectively.
86Rb+ efflux.Agonist-induced86Rb+ efflux was calculated as the fractional release above baseline on agonist stimulation (Marks et al., 1996). Basal efflux was defined as cpm collected in the fractions immediately before and after stimulation, and the basal rate of86Rb+ efflux was determined as an exponentially decaying curve fitted to these data points. Agonist-stimulated efflux was calculated as the cpm above the calculated baseline during the period of agonist application. To correct for interexperimental variation in synaptosomal86Rb+ uptake, agonist-stimulated efflux was divided by the calculated baseline efflux. Values for86Rb+ efflux from mouse thalamic synaptosomes are more than double those from corresponding rat preparations (see Fig. 5), reflecting the higher density of [3H]-nicotine binding sites in mouse thalamus. Curve fitting of86Rb+ efflux data from mouse thalamic synaptosomes was performed using the nonlinear least squares curve fitting facility of Sigma Plot V5.0 for DOS.
Upregulation. Upregulation profiles were fit to a logistic equation (Whiteaker et al., 1998) using the nonlinear least squares curve fitting facility of Sigma Plot V2.0 for Windows, to give the EC50 for the upward phase of the dose–response curve and Umax, the maximum upregulation produced by the drug tested.
Dopamine release. Evoked [3H]-dopamine release was calculated as the area under the peak, after subtraction of the baseline, as described previously (Kaiser et al., 1998). Data points for agonist dose–response relationships (after subtraction of nonspecific release determined in the presence of mecamylamine) were fitted to the Hill equation y = (a −d) /[1 + (k/x)n] + d, wherea is the asymptotic maximum, d is the asymptotic minimum, k is the agonist concentration at the inflection point (EC50), x is the ligand concentration, and n is the slope parameter (Hill number).
Data for inhibition by antagonist are represented as percentages of the corresponding controls, assayed in parallel in the absence of antagonist. One-way ANOVA–post hoc Bonferroni test was used to determine the significance of differences from control.
Two-electrode voltage-clamp recordings. Full dose–response curves were obtained from individual oocytes and normalized relative to the response to an EC50 concentration of ACh recorded in the same oocyte. Sigmoidal concentration–response curves were fit to the Hill equation using Origin 4.0 (Microcal Software).
Calcium fluorimetry. The intracellular Ca2+ concentration ([Ca2+]i) was calculated from the fluorescence ratio of fura-2 (A340/A380, given as R below) according to the Grynkiewicz equation (Grynkiewicz et al., 1985) [Ca2+]i =Kd * [(R −Rmin)/(Rmax−R)]*(Sf2/Sb2), where Kd is the dissociation constant for Ca2+ binding to fura-2,Rmin is the fluorescence ratio under nominally “zero” free Ca2+ conditions,Rmax is the fluorescence ratio under saturating Ca2+ conditions, and Sf2/Sb2 is the ratio of fluorescence values of Ca2+-free and Ca2+-saturated fura-2, measured at the wavelength used to monitor Ca2+-free fura-2.
For dose–response curves, agonist responses were calculated as the percentage of the change in intracellular Ca2+ produced by a maximally stimulating concentration of the same agonist, assayed in parallel. EC50 values were calculated by fitting data points to the Hill equation, using the nonlinear least squares curve fitting facility of Sigma Plot V2.0 for Windows.
RESULTS
Competition binding assays
(±)-UB-165 was compared with (±)-anatoxin-a and (±)-epibatidine (see Fig. 1 for structures) for its ability to displace the binding of a number of nicotinic radioligands (Fig.2, Table1). Radioligand binding assays, at least those using radiolabeled agonists, reflect the desensitized state of the nAChRs. Most high-affinity [3H]-nicotine binding sites in rat brain are considered to represent the α4β2 subtype (Zoli et al., 1998). (±)-Anatoxin-a and (±)-epibatidine inhibited [3H]-nicotine binding withKi values of 1.25 ± 0.20 and 0.02 ± 0.005 nm, respectively. (±)-UB-165 displayed an intermediate potency, with aKi value of 0.27 ± 0.05 nm (Fig. 2A, Table 1). Comparable Ki values were derived for inhibition of [3H]-epibatidine binding to chicken α4β2 nAChRs in M10 cells (Table 1). (±)-Anatoxin-a and (±)-epibatidine competed with [125I]-αBgt for binding to putative α7-type nAChRs in rat brain membranes, withKi values of 1840 ± 260 and 233 ± 69 nm, respectively. (±)-UB-165 exhibited slightly lower potency than (±)-anatoxin-a, having aKi value of 2790 ± 370 nm (Fig. 2B).
Molecular structures of (−)-epibatidine and (+)-anatoxin-a and the enantiomer of UB-165 that corresponds to (+)-anatoxin-a.
Competition binding assays for [3H]-nicotine binding sites (A) and [125I]-αBgt binding sites (B) in rat brain membranes, [3H]-epibatidine binding sites (C) in human SH-SY5Y cell membranes, and [125 I]-αBgt binding sites (D) in rat muscle extract. Preparations were incubated with radioligand and increasing concentrations of (±)-UB-165 (▴), (±)-anatoxin-a (○), or (±)-epibatidine (■) as described in Materials and Methods. Each point is the mean of at least three separate determinations ± SEM. Data points were fitted to the Hill equation.
Affinities for radioligand binding sites corresponding to various nAChR subtypes
[3H]-Epibatidine labels α3-containing nAChRs in human neuroblastoma SH-SY5Y cells (Wang et al., 1996). The concentration of [3H]-epibatidine used (150 pm) was chosen to preferentially label α3β2-containing nAChRs [including α3β2 and α3β2α5 combinations (Wang et al., 1996)], but Kivalues have not been derived because of the inherent nAChR heterogeneity. (±)-UB-165 displaced [3H]-epibatidine from SH-SY5Y cell membranes with an IC50 value of 20 ± 0.7 nm, intermediate between that of (±)-anatoxin-a (IC50 = 155 ± 25 nm) and (±)-epibatidine (IC50 = 0.34 ± 0.06 nm) (Fig. 2C). (±)-UB-165 was also examined for its ability to displace [125I]-αBgt from rat muscle extract (Fig. 2D). Its Kivalue of 990 ± 240 nm was similar to that of (±)-epibatidine (Ki = 610 ± 160 nm), whereas (±)-anatoxin-a was the most potent competing ligand, with a Kivalue of 85 ± 41 nm.
Thus the rank order of potencies at the rat and chicken α4β2 and human α3-containing nAChR subtypes is (±)-epibatidine > (±)-UB-165 > (±)-anatoxin-a. In contrast, (±)-UB-165 was the least potent at α7-type nAChRs and muscle nAChRs: rank orders of potencies are (±)-epibatidine > (±)-anatoxin-a ≥ (±)-UB-165 and (±)-anatoxin-a > (±)-epibatidine ≥ (±)-UB-165, respectively. The ability of each of these ligands to discriminate between nAChR subtypes was expressed as an affinity ratio, relative to the value at the rat brain [3H]-nicotine binding site, which was defined as 1 (Table 1). From comparison of these affinity ratios, (±)-UB-165 resembles (±)-epibatidine in its marked preference for α4β2* nAChRs compared with α7-type and muscle nAChRs.
Upregulation of α4β2 nAChRs
(±)-Epibatidine and (±)-anatoxin-a differ in their abilities to upregulate α4β2 nAChRs in the M10 cell line: (±)-epibatidine is a partial upregulator whereas (±)-anatoxin-a is fully efficacious (Whiteaker et al., 1998). (±)-UB-165 was compared with (±)-anatoxin-a and (±)-epibatidine in this assay, to determine which of the parent compounds it most resembles with respect to upregulation (Fig.3). Dose–response profiles for upregulation produced by (±)-UB-165 and (±)-anatoxin show that they are similarly efficacious, with maximum upregulation of 237 ± 47 and 221 ± 27% above control levels, respectively. (±)-UB-165 was a more potent upregulator than (±)-anatoxin-a, with EC50 values of 25.3 ± 17.7 and 985 ± 515 nm, respectively. In contrast, (±)-epibatidine achieved a maximum upregulation of only 76 ± 6% above control, although it was the most potent of the three ligands, with an EC50 value of 1.2 ± 0.2 nm. The ability of (±)-UB-165 to upregulate nicotinic binding sites is consistent with it behaving as an agonist at α4β2 nAChRs, because all agonists that we have examined have produced some degree of upregulation in this system, whereas antagonists produce little if any upregulation (Whiteaker et al., 1998).
Upregulation of [3H]-epibatidine binding sites expressed in M10 cells. M10 cells were cultured in the presence of dexamethasone and (±)-UB-165 (▴), (±)-anatoxin-a (○), or (±)-epibatidine (■) for 48 hr. After extensive washing, [3H]-epibatidine binding was performed on M10 cells in situ. Upregulation was expressed as specific [3H]-epibatidine binding above that of control cells treated in parallel with dexamethasone alone. (±)-Epibatidine was a partial agonist for upregulation compared with (±)-UB-165 and (±)-anatoxin-a. Each point is the result of at least three separate determinations ± SEM. Data points were fitted to a logistic equation (Whiteaker et al., 1998), giving EC50 values of 1.2 ± 0.2, 25.3 ± 17.7, and 985 ± 515 nm, and percentage maximum upregulation above control of 76.4 ± 5.7, 237 ± 47, and 221 ± 28%, for (±)-epibatidine, UB-165, and (±)-anatoxin-a, respectively.
Presynaptic nicotinic stimulation of [3H]-dopamine release from rat striatal synaptosomes
(±)-UB-165 was compared with (±)-anatoxin-a, (±)-epibatidine, and (−)-nicotine for their abilities to promote [3H]-dopamine release from rat striatal synaptosomes (Fig. 4A). Of these four compounds, (±)-epibatidine was the most efficacious and most potent (EC50 value = 2.42 ± 0.44 nm), whereas (−)-nicotine (EC50 value = 1.59 ± 0.38 μm) was the least potent but had higher efficacy than (±)-anatoxin-a. Although (±)-UB-165 was slightly more potent than (±)-anatoxin-a (EC50 values = 88 ± 18 and 134 ± 26 nm, respectively), it was much less efficacious than the other agonists, achieving a maximum of specific, nAChR-mediated [3H]-dopamine release that was only 42% of that of (±)-anatoxin-a and 21% of that of (±)-epibatidine.
Nicotinic stimulation of [3H]-dopamine release from rat striatal synaptosomes. A, Dose–response curves for the release of [3H]-dopamine evoked by increasing concentrations of (±)-epibatidine (■), (−)-nicotine (●), (±)-anatoxin-a (○), and (±)-UB-165 (▴). Each agonist concentration was tested in the presence and absence of mecamylamine to determine specific nAChR-mediated release, as described in Materials and Methods. Responses were normalized to the response to 1 μm (±)-anatoxin-a, determined in parallel. Values are the mean ± SEM of at least three independent assays. The rank order of potencies (EC50 values) is (±)-epibatidine (2.4 ± 0.4 nm) > (±)-UB-165 (88 ± 18 nm) ≥ (±)-AnTx-a (134 ± 26 nm) > (−)-nicotine (1595 ± 377 nm). B, The inhibition by α-conotoxin-MII (112 nm; hatched bars) of [3H]-dopamine release evoked by maximally effective concentrations of agonists is compared with inhibition by mecamylamine (10 μm; black bars). Synaptosomes loaded with [3H]-dopamine were perfused with antagonist for 10 min before stimulation with (±)-epibatidine (0.1 μm), (−)-nicotine (10 μm), (±)-anatoxin-a (1 μm), or (±)-UB-165 (1 μm). Responses were normalized to the response to 1 μm (±)-anatoxin-a, determined in parallel. Values are the mean ± SEM of at least three independent assays.
nAChRs mediating the presynaptic nicotinic stimulation of [3H]-dopamine release from striatal terminals appear to be heterogeneous, based on the partial inhibition by the α3β2-selective toxin α-conotoxin-MII (Kulak et al., 1997;Kaiser et al., 1998). A maximally effective concentration of this toxin (112 nm) was compared with mecamylamine (10 μm) for their abilities to inhibit [3H]-dopamine release evoked by maximally effective concentrations of the four agonists mentioned above (Fig. 4B). The results confirm the differences in efficacy noted above. Although mecamylamine inhibited agonist-evoked [3H]-dopamine release to the same extent in each case (the residual release being the nonspecific component that is elicited by a buffer pulse without agonist), α-conotoxin-MII produced varying degrees of inhibition. The mecamylamine-sensitive response to (±)-anatoxin-a was reduced by 56% by the toxin, in agreement with our previous findings (Kaiser et al., 1998), whereas the mecamylamine-sensitive [3H]-dopamine release evoked by (±)-epibatidine and (−)-nicotine was inhibited by 47.9 and 32.4%, respectively. However, the mecamylamine-sensitive response to (±)-UB-165 was almost completely blocked (87.8%) by α-conotoxin-MII. This suggests that (±)-UB-165 has very little efficacy at the other subtype(s) of nAChRs responsible for [3H]-dopamine release. Because the α4β2 subtype is a candidate for this role, we examined the behavior of (±)-UB-165 in a neurochemical assay for this putative subtype.
86Rb+ efflux from thalamic synaptosomes
The efflux of86Rb+ from preloaded mouse thalamic synaptosomes in response to nicotinic agonists is proposed to reflect the activation of α4β2 nAChRs (Marks et al., 1993, 1996, 1999). (±)-UB-165 was compared with (±)-epibatidine for its ability to elicit86Rb+ efflux from this preparation (Fig.5A,B). Although (±)-epibatidine was a potent and efficacious agonist in this assay (Fig. 5A), as reported previously (Marks et al., 1996), (±)-UB-165 produced very little response when tested over the concentration range examined in the [3H]-dopamine release assay (Fig.5B). Indeed, 1 μm (±)-UB-165 elicited <15% of the maximum86Rb+ efflux provoked by (±)-epibatidine. The interpretation that (±)-UB-165 is a partial agonist with respect to86Rb+ efflux is supported by the ability of (±)-UB-165 (1 μm) to shift the dose–response curve for (±)-epibatidine to the right (Fig. 5A). Moreover, increasing concentrations of (±)-UB-165 progressively inhibited the response to a maximally effective concentration of (±)-epibatidine (100 nm) (Fig.5B). The low efficacy of (±)-UB-165 in eliciting86Rb+ efflux was verified in rat thalamic synaptosomes (Fig. 5C). Maximally effective concentrations of (±)-UB-165, (±)-anatoxin-a, and (±)-epibatidine with respect to [3H]-dopamine release were compared. Their relative efficacies in stimulating86Rb+ efflux resemble their efficacies in evoking α-conotoxin-MII-insensitive [3H]-dopamine release (Fig.4B). Indeed, the slight increase in86Rb+ efflux in response to (±)-UB-165 was blocked by mecamylamine but was insensitive to α-conotoxin-MII (Fig. 5C).
Nicotinic stimulation of86Rb+ efflux from thalamic synaptosomes. P2 synaptosomes from mouse (A, B) or rat (C) thalamus were loaded with86Rb+ and superfused;86Rb+ efflux is a measure of nAChR activation, primarily of the α4β2 subtype (Marks et al., 1993).A, Dose–response curve of86Rb+ efflux stimulated by (±)-epibatidine, in the absence (▪) or presence (■) of 1 μm (±)-UB-165. B, Dose–response curve of86Rb+ efflux stimulated by (±)-UB-165, in the absence (▴) or presence (▵) of 100 nm(±)-epibatidine. C, Comparison of86Rb+ efflux from rat thalamic synaptosomes stimulated by maximally effective concentrations of (±)-UB-165 (1 μm), (±)-anatoxin-a (10 μm), and (±)-epibatidine (100 nm) (and buffer control). (±)-UB-165 and (±)-anatoxin-a were also tested in the presence of α-conotoxin-MII (112 nm; hatched bars) and mecamylamine (10 μm; black bars). Values are the mean ± SEM of at least three independent assays.
Activation of inward currents in Xenopus oocytes expressing nAChRs of defined subunit combination
To confirm that UB-165 has low efficacy at α4β2 nAChRs, and to assess its efficacy at other nAChR subtypes, it was examined for its ability to elicit inward currents in Xenopus oocytes expressing pairwise combinations of human α and β subunits, or homo-oligomeric nAChRs of α7 subunits (Fig.6). (±)-UB-165 was most potent in activating α4β4 and α2β4 nAChRs (EC50 = 0.05 μm), followed by α3β4 (0.27 μm), α3β2 (3.9 μm), and α7 (6.9 μm), and was similarly efficacious at these subtypes. In contrast, (±)-UB-165 failed to elicit significant currents from oocytes expressing α4β2 or α2β2 nAChRs.
Concentration–response curves for (±)-UB-165 at recombinant human nAChRs. Pairwise combinations of human α and β subunits, or α7 alone, were expressed in Xenopusoocytes. Current responses in each oocyte were normalized to an EC50 concentration of ACh. EC50 values yielded by logistical curve fits were 3.9 μm for α3β2 (■), 0.05 μm for α2β4 (●), 0.27 μm for α3β4 (▪), 0.05 μm for α4β4 (▴), and 6.9 μm for α7 (▿). There was little activation of α2β2 (○) and α4β2 (▵) subtypes. Data points represent mean ± SD of three to six experiments.
Activation of Ca2+ fluxes in the SH-SY5Y cell line
The functional potency of (±)-UB-165 at native α3-containing nAChRs was investigated in the human neuroblastoma cell line SH-SY5Y. Because this cell line does not express α4 subunits, it was used to examine the efficacy of UB-165 at native nAChRs other than the α4β2* subtype. Receptor activation was assayed as an increase in intracellular Ca2+, measured quantitatively in suspensions of SH-SY5Y cells using fura-2 (Fig.7). (±)-UB-165 and (±)-anatoxin-a were compared: both agonists increased intracellular Ca2+, and this effect was totally blocked by 10 μm mecamylamine (Fig.7A,B). Determination of dose–response relationships showed that (±)-UB-165 was more potent than (±)-anatoxin-a, with EC50 values of 154 ± 20 and 530 ± 92 nm, respectively (Fig. 7C). Comparison of maximally effective concentrations of each drug (1 μm (±)-UB-165 and 10 μm (±)-anatoxin-a) showed them to be comparably efficacious in this assay (Fig. 7D). Increases in intracellular Ca2+ evoked by these maximally stimulating drug concentrations were partially blocked to the same extent by 112 nm α-conotoxin-MII (43.9 ± 8.3 and 49.8 ± 10.0% inhibition of (±)-UB-165- and (±)-anatoxin-a-evoked Ca2+ responses, respectively) (Fig. 7D).
Nicotinic stimulation of increases in intracellular Ca2+ in SH-SY5Y cells. SH-SY5Y cells were loaded with 5 μm fura-2 AM in Ca2+-free buffer for 45 min before addition of 2 mm CaCl2 followed by (±)-anatoxin-a or (±)-UB-165. Antagonists were added 5 min before the addition of Ca2+ and agonist. Representative fluorimetry traces in response to 10 μm (±)-anatoxin-a (A) and 1 μm (±)-UB-165 (B). The bottom trace inA and B shows the response in the presence of 10 μm mecamylamine. C, Dose–response relationships of increases in intracellular Ca2+ induced by (±)-anatoxin-a (○) and (±)-UB-165 (▴). For each curve, results were normalized to the responses produced by a maximally effective concentration of agonist, determined in parallel in the same experiment. Data points are the mean ± SEM of three independent assays and were fitted to the Hill equation, giving EC50 values of 154 and 530 nm for (±)-UB-165 and (±)-anatoxin-a, respectively.D, Comparison of efficacies of maximally stimulating concentrations of (±)-UB-165 (1 μm) and (±)-anatoxin-a (10 μm), and inhibition by α-conotoxin-MII (112 nm, black bars). Values are the mean of three independent experiments ± SEM and were normalized to the response to (±)-UB-165 within each separate experiment.
DISCUSSION
The potent nicotinic agonists epibatidine and anatoxin-a are structurally related in that they both incorporate an azobicyclic core (Fig. 1); UB-165 is a hybrid molecule comprising the bulkier azobicyclo[4.2.1]nonane moiety of anatoxin-a attached to the chloropyridyl substituent of epibatidine (Wright et al., 1997). At α4β2* and α3-containing nAChR binding sites (defined by [3H]-nicotine binding to brain membranes and [3H]-epibatidine binding to SH-SY5Y cell membranes, respectively), UB-165 has intermediate potency, compared with the parent compounds (Fig. 2). It resembles epibatidine in its high degree of discrimination between α4β2* binding sites on the one hand versus α7 and muscle sites on the other (Table 1). In contrast, UB-165 is more like anatoxin-a with regard to its enantiospecificity (Wright et al., 1997) and its efficacy in upregulating α4β2 nAChRs in M10 cells (Fig. 3). From these properties and its structural features it was not possible to predict the very low efficacy of UB-165 with regard to striatal [3H]-dopamine release, measured in perfused synaptosome preparations in vitro. Use of the α3β2-selective α-conotoxin-MII (Cartier et al., 1996) has enabled us to show that this low efficacy of UB-165 arises from its almost complete inability to activate the α-conotoxin-MII-insensitive component of nAChR-stimulated [3H]-dopamine release. The failure of UB-165 to activate α4β2*-mediated86Rb+ efflux from thalamic synaptosomes and to elicit responses in oocytes expressing α4β2 nAChRs provides support for the proposition that presynaptic α4β2* nAChRs on striatal dopaminergic terminals contribute to the nicotinic stimulation of [3H]-dopamine release, in addition to α3β2-containing nAChRs (Kulak et al., 1997; Kaiser et al., 1998).
Neurons in the substantia nigra pars compacta of the rat express mRNA for α3, α4, α5, α6, β2, and β3 (Wonnacott, 1997) and possibly β4 and α7 subunits (Charpantier et al., 1998), and some or all of these may contribute to nAChRs on dopaminergic terminals in the striatum. The challenge of defining the subunit composition of native nAChRs responsible for particular physiological responses is important for understanding the significance of subunit heterogeneity, rules of assembly, and functional implications arising from receptor subtypes with differing properties. Nicotine-evoked [3H]-dopamine release from striatal synaptosomes is absent from β2 null mutant mice (Grady et al., 1998), implicating the β2 subunit in all nAChRs in dopaminergic terminals governing this response in the mouse. Sensitivity to α-conotoxin-MII (Kulak et al., 1997; Kaiser et al., 1998), a toxin with specificity for the α3β2 subunit combination expressed in Xenopusoocytes (Cartier et al., 1996; Kaiser et al., 1998) is compatible with the requirement for β2 to be present. This does not exclude the presence of additional types of subunit in an nAChR containing α3 and β2 subunits. The restricted distributions of α6 and β3 mRNA in the CNS, with high expression in the substantia nigra pars compacta (Deneris et al., 1989; Le Novère et al., 1996; Charpantier et al., 1998), together with the erstwhile “orphan” status of the α6 and β3 subunits, makes them prime candidates for complex combinations with other subunits such as α3 and β2 (Fucile et al., 1998;Groot-Kormelink et al., 1998).
A striking observation from the studies with α-conotoxin-MII is its incomplete inhibition of nicotinic agonist-evoked [3H]-dopamine release. Kulak et al. (1997) found a block of 34–49% of (−)-nicotine-stimulated [3H]-dopamine release, whereas Kaiser et al. (1998) reported 56% inhibition of mecamylamine-sensitive release evoked by (±)-anatoxin-a. The present study reproduces these findings and extends the analysis to other agonists. A partial inhibition (48%) of [3H]-dopamine release evoked by (±)-epibatidine was observed, whereas the response evoked by (±)-UB-165 was almost totally inhibited by α-conotoxin-MII (Fig. 4). This suggests that (±)-UB-165 primarily activates only the α-conotoxin-MII-sensitive nAChRs associated with dopaminergic terminals, which accords with an α3β2-containing nAChR. The magnitude of the α-conotoxin-MII-sensitive portion of evoked [3H]-dopamine release is rather similar between agonists (with the exception of (±)-epibatidine, for which considerable variability was encountered), whereas the α-conotoxin-MII-insensitive portion varies in the ratio 1:0.88:0.40:0.04 for (±)-epibatidine, (−)-nicotine, (±)-anatoxin-a, and (±)-UB-165, respectively. This variation in the α-conotoxin-MII-insensitive responses to different agonists is largely responsible for their different efficacies with respect to total release (Fig. 4B).
The hypothesis, elaborated in the introductory remarks, states that the α-conotoxin-MII-insensitive component of nAChR-evoked striatal dopamine release is mediated by α4β2* nAChRs. The low efficacy, and complete block by α-conotoxin-MII, of UB-165-evoked [3H]dopamine release (Fig. 4) leads to the prediction that (±)-UB-165 should be a very poor agonist at this receptor subtype. nAChR-evoked86Rb+ efflux from mouse thalamic synaptosomes is attributed to an α4β2* nAChR (Marks et al., 1993, 1996), and, in agreement with this prediction, UB-165 is a very weak, partial agonist in this assay (Fig. 5). In rat thalamic synaptosomes, (±)-epibatidine, (±)-anatoxin-a, and (±)-UB-165 stimulated mecamylamine-sensitive86Rb+ efflux with relative efficacies of 1:0.39:0.08, very comparable to the ratio of efficacies for α-conotoxin-MII-insensitive [3H]-dopamine release (see above). Moreover, Type II responses in rat hippocampal neurons, attributed to an α4β2 nAChR, show different efficacies with different agonists, with the ratio 1:0.6:0.3 for (+)-epibatidine, (−)-nicotine, and (+)-anatoxin-a, respectively (Alkondon and Albuquerque, 1995), similar to those found here for these agonists, with respect to α-conotoxin-MII-insensitive [3H]-dopamine release and86Rb+efflux.
Results from Xenopus oocytes expressing various recombinant nAChR subtypes verify that (±)-UB-165 has low efficacy at α4β2 nAChRs. The only other subunit combination examined at which (±)-UB-165 was inefficacious was α2β2 (Fig. 6), but the lack of expression of α2 in substantia nigra (Wada et al., 1989) excludes this as a contributor to the presynaptic modulation of striatal dopamine release. The electrophysiological analysis confirms that UB-165 is a full agonist at α3β2 nAChRs. This is consistent with the data from a functional assay for native α3-containing nAChRs [increases in intracellular Ca2+ in SH-SY5Y cells (Fig. 7)], in which (±)-anatoxin-a and (±)-UB-165 were equally efficacious and equally sensitive to α-conotoxin-MII. SH-SY5Y cells expresses mRNA for α3, α5, α7, β2, and β4 nAChR subunits (Lukas et al., 1993; Peng et al., 1994). Therefore the α4β2 nAChR is not present in this cell line. Nevertheless, inhibition by α-conotoxin-MII was incomplete (∼50%) (Fig. 7), and this presumably reflects the presence of additional nAChR subtypes that lack an α3β2 interface, in particular β4-containing nAChRs. Heterogeneous combinations of α3, α5, β2, and β4 subunits in SH-SY5Y cells have been proposed, on the basis of immunoisolation of nAChRs using subunit-specific antibodies (Wang et al., 1996). Notably, 56% of α3-containing nAChRs were found to contain the β2 subunit. The α-conotoxin-MII-insensitive response is likely to arise from activation of α3β4-containing nAChRs that lack the β2 subunit (although a contribution from α7-type nAChRs cannot be ruled out). α3β4 nAChRs expressed in Xenopusoocytes are insensitive to the concentration of α-conotoxin-MII used in this study (Cartier et al., 1996; Kaiser et al., 1998) but are fully activated by UB-165 (Fig. 6). Moreover, the comparable efficacies of (±)-UB-165- and (±)-anatoxin-a demonstrate that the α-conotoxin-MII-insensitive component of nAChR-mediated Ca2+ increases in SH-SY5Y cells is clearly different from the α-conotoxin-MII-insensitive component of nAChR-mediated [3H]-dopamine release from striatal synaptosomes.
The weak partial agonism of UB-165 at α4β2* nAChRs contrasts with its full efficacy in upregulating α4β2 nAChRs in M10 cells (Fig.3). Lower concentrations of agonist are needed to induce upregulation, compared with those required to elicit functional responses. Together with the general inability of antagonists to prevent agonist-induced upregulation, a higher affinity state of the nAChRs than the active conformation is implicated as the trigger for the upregulation of binding sites (Whiteaker et al., 1998; Fenster et al., 1999). Whether this state is the high-affinity desensitized state measured in ligand binding assays remains controversial. This result demonstrates that although UB-165 lacks functional efficacy at α4β2* nAChRs, long-term exposure to this drug would upregulate this receptor subtype with potential functional consequences.
UB-165 is not unique as a partial agonist in stimulating [3H]-dopamine release (Holladay et al., 1997) and86Rb+ efflux (Marks et al., 1996). For example RJR 2429 has a structure comparable to UB-165, comprising an azobicyclo[2.2.2]octane core coupled to a chloropyridyl moiety (Bencherif et al., 1998), and has a profile similar to UB-165 in functional and binding assays, except for its much higher potency at muscle nAChRs. However, application of novel ligands in conjunction with selective antagonists is necessary to maximize their potential as tools to discriminate receptor subtypes. This study has exploited the exquisite subtype selectivity of α-conotoxin-MII together with the partial agonism of UB-165 to arrive at a better definition of the nAChR subtypes present on dopamine terminals in striatal preparations. The conclusion that both α4β2*- and α3β2-containing nAChRs contribute to [3H]-dopamine release, either on the same or separate populations of terminals (varicosities), enhances our understanding of this model system and raises further questions about the physiological purpose of nAChRs in the mammalian brain.
Footnotes
This work was supported by Biotechnology and Biological Sciences Research Council (BBSRC) Grants 7/MOLO4724 (T.C.) and 86/B11785 (S.W. and T.C.); BAT Co. Ltd. (S.W.); National Institute on Drug Abuse Grants DA-03194 and DA-00197 (A.C.C.); postgraduate studentships from BBSRC and Engineering and Physical Sciences Research Council to C.G.V.S. and J.A.S., respectively; and NATO Collaborative Research Grant CRG 971632 (S.W. and A.C.C.). We are grateful to Adrian Mogg for the assays on rat muscle.
Dr. Kaiser's present address: Department of Biology, University of California, San Diego, 9500 Gilman Drive, La Jolla, CA 92093-0357.
Correspondence should be addressed to Dr. S. Wonnacott, Department of Biology and Biochemistry, University of Bath, Bath, BA2 7AY, UK. E-mail: s.wonnacott{at}bath.ac.uk.
