Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Zinc is an essential trace element in most living systems (Aggett and Harries, 1979; Berg and Shi, 1996). Like other d-transition metals, zinc binds to a variety of enzymes and helps to confer enzymatic activity to otherwise inactive metabolic proteins. In contrast, zinc has no ligand field stabilization energy and no redox activity, and consequently, zinc is well suited as a modulator of both protein-protein and protein-DNA interactions (Berg and Shi, 1996). As such, zinc plays a regulatory role in enzyme modulation and gene transcription, which are best exemplified by enzymes such as aspartate transcarbamoylase (Nelbach et al., 1972) and "zinc finger" proteins like transcription factor IIIA (Hanas et al., 1983). Given the apparent ubiquitous nature of both metabolic and regulatory proteins that require zinc for proper functioning, zinc would seem an unlikely candidate for a signaling molecule. However, zinc is highly compartmentalized, and in parts of the central nervous system, zinc is stored in neuronal synaptic vesicles (Perez-Clausell and Danscher, 1985). More importantly, depolarization of the neurons possessing zinc-containing synaptic vesicles results in the release of high concentrations of zinc (ca. 300 µM) into the synaptic cleft, and the released zinc is also taken back up (Assaf and Chung, 1984; Howell et al., 1984) and reloaded into synaptic vesicles (Palmiter et al., 1996). Furthermore, several receptor systems involved in neuronal signaling in the central nervous system are already known to be modulated by micromolar concentrations of zinc, including N-methyl-D-aspartate, -amino butyric acid (GABA; Westbrook and Mayer, 1987; Draguhn et al., 1990),  (Connor and Chavkin, 1992) and glycine receptors (Laube et al. 1995), L-type calcium channels (Winegar and Lansman, 1990), and the dopamine transporter (Norregaard et al., 1998). In the case of the GABA receptors, the modulation by zinc appears to be both subtype- or subunit-selective for GABA_B and GABA_A receptors, respectively (Draguhn et al., 1990; Xie and Smart, 1991), whereas for  receptors, zinc modulation is subclass-selective for the ₂ receptor (Connor and Chavkin, 1992). These results and our recent finding that zinc can allosterically modulate the binding of [³H]antagonists to D_1A and D_2L dopamine receptors (Schetz and Sibley, 1997) hinted that the effect of zinc might differ among the various subtypes or subclasses of dopamine receptors.

In contrast with the two cloned D₁-like dopamine receptors whose sequences, pharmacology, and second messenger systems are very similar, the D₂-like subfamily of dopamine receptors is comprised of three distinct members. Each member of the D₂-like subfamily has a unique pharmacological profile and/or capacity to couple to various signaling pathways, including indirect coupling to adenylyl cyclase via G_i proteins or direct coupling to the Na⁺/H⁺ exchanger (Jackson and Westlind-Danielsson, 1994). Nevertheless, one feature that all D₂-like dopamine receptors share is their ability to bind the D₂-selective antagonist spiperone (and analogs) with picomolar affinity. These shared binding characteristics of [³H]methylspiperone were exploited as a means of standardizing D₂-like dopamine receptor subtype-specific responses to zinc. Chinese hamster ovary (CHO) cells stably expressing either the rat D_2L, D₃, or D₄ receptor were chosen for comparing homologous populations of D₂-like dopamine receptor subtypes, because no specific radioligand binding and low nonspecific binding was observed in untransfected cells. In addition, D_2L receptors expressed in these cells effectively couple to G proteins to inhibit the formation of cyclic AMP (cAMP). Such standardization of experimental variables allowed for the systematic evaluation of the subtype-selective effects of zinc on D₂-like dopamine receptors.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Chemical Reagents. Zinc chloride was purchased from Aldrich Chemical Co. (catalog no. 22,999-7). N-methylspiperone, (+)-butaclamol, ()N-propylnorapomorphine, and dopamine were obtained from Research Biochemicals, Incorporated (Natick, MA). Binding and wash buffer reagents were purchased from Sigma Chemical Company and Fluka (St. Louis, MO). [³H]methylspiperone (NET-856, 82.0-84.4 Ci/mmol) and [³H]raclopride (NET-975, 79.0 Ci/mmol) were purchased from Dupont-NEN (Boston, MA). The cAMP assay reagents, including [³H]cAMP at 23.0 Ci/mmol, were obtained from Diagnostic Products Corporation (catalog no. KAPH2; Los Angeles, CA)

Expression and Screening of Dopamine Receptors. The cDNA encoding the rat D_2L dopamine receptor was subcloned into a modified expression vector, pCD-SR, and clonal CHO cell lines stably expressing high levels of the rat D_2L dopamine receptor were selected as described previously (Zhang et al., 1994). Clonal CHO cells lines expressing the rat D₃ and D₄ dopamine receptors were prepared as described above for the D_2L receptor, except that the D₃ and D₄ cDNAs were subcloned into pcDNA1 and pcDNA3 (Invitrogen, San Diego, CA).

Preparation of Membranes for Radioligand Binding Assays. Fresh membranes from CHO cells were harvested for each experiment. Cells from confluent 150 mm² flasks were washed once with Earle's balanced saline solution (EBSS), then lifted from the culture substrate by incubation for 15 min in 10 ml EBBS lacking calcium and magnesium ions and supplemented with 5 mM EDTA. Cells were diluted 5-fold in EBSS and centrifuged at 200g for 15 min The resulting cell pellet was resuspended by vortexing in 10 ml of lysis buffer (5 mM Tris-HCl and 5 mM MgCl₂, pH 7.4 at 4°C). After 10 min in lysis buffer the cell lysate was glass-glass homogenized (10-12 strokes in a 15 ml Dounce homogenizer). The homogenate was centrifuged at 35,000g for 30 min. The membrane pellet was resuspended by trituration in 10 to 12 ml of binding buffer (50 mM Tris, pH 7.4 at 4°C) and recentrifuged as above. The final pellet was brought up in the appropriate volume of binding buffer and homogenized with two to three strokes of the homogenizer.

[³H]methylspiperone Equilibrium Binding. [³H]methylspiperone binding affinity (K_D) and the corresponding maximum number of binding sites (B_max) were measured by radioligand saturation isotherm binding, which was assayed by rapid filtration. One and ten micromolar (+)-butaclamol were used to define nonspecific binding for D_2L and D₄ receptors, respectively. Either 10 µM (+)-butaclamol or methylspiperone was used to define nonspecific binding for D₃, with identical results. Saturation isotherm binding of [³H]methylspiperone to D_2L receptors was performed over a concentration range of 7.5 to 960 pM radioligand, and binding to D₃ and D₄ receptors was assayed from 12.5 to 1600 pM. All binding assays were performed using fresh membranes prepared daily in a total volume of 1 ml/sample at protein concentrations of approximately 40 µg/ml (range 25-100 µg/ml). The binding buffer consisted of 50 mM Tris-HCl, pH 7.4 at 25°C and the wash buffer was 50 mM Tris, pH 7.4 at 0°C. After addition of all assay components, each tube was vortexed for 2 s and samples were incubated at 25°C for 90 to 120 min. Membranes were harvested by washing three times with 3.0 volumes of wash buffer onto 0.5% polyethylenimine pretreated GF/C glass filter fiber. Individual filters were placed in scintillation vials containing 4 ml Cytoscint (ICN Pharmaceuticals, Inc., Costa Mesa, CA), mixed by vortexing, and counted on a Beckman TS500 scintillation counter with a counting efficiency of 47%.

[³H]raclopride Equilibrium Binding to D_2L Receptors and the Influence of Sodium Ions. Zinc competition curves of specific [³H]raclopride binding to D_2L receptors were performed as described above for [³H]methylspiperone, except that one group contained no NaCl in either the binding or wash buffer and the second group contained 120 mM NaCl in both the binding and wash buffer. In some experiments, 120 mM N-methyl-D-glucamine was used in place of 120 mM NaCl. The IC₅₀ values measured for zinc under these conditions were subsequently used to determine the effect of zinc on [³H]raclopride saturation isotherms performed in the presence or absence of 120 mM NaCl. Saturation isotherm binding of [³H]raclopride to D_2L was performed over a concentration range of 15 to 1600 pM, and 1 µM (+)-butaclamol was used to define nonspecific binding.

Competition Curve Style Schild-Type Plot Analysis of Zinc Inhibition of Antagonist Binding. The mechanism by which zinc inhibits the binding of the D₂-like selective antagonist, [³H]methylspiperone, was evaluated by a Schild-type competition analysis. Specifically, the shift in IC₅₀ values for zinc inhibition was measured as a function of increasing concentrations of [³H]methylspiperone and then compared to a model of a purely competitive binding interaction. The competitive model was generated by first seeding the Cheng-Prusoff equation for a perfectly competitive interaction with the experimentally determined IC₅₀ values and radioligand concentrations for the lowest concentration of radioligand tested. Next, the model was propagated by formation of an equality, which is simply a rearrangement of the Cheng-Prusoff (1973) equation K_i = IC₅₀ (1 + ([radioligand]/K_D)), at two different radioligand concentrations:

Antagonist Dissociation Rate Kinetics as a Function of Zinc. Dissociation kinetic assays were performed at 37°C using the same buffer conditions as for equilibrium binding. Assays were initiated by equilibrating either D_2L, D₃, or D₄ dopamine receptors with 500 to 1000 pM. [³H]methylspiperone for 1 h. Next, a saturating concentration of excess unlabeled ligand (2 µM nonisotopic methylspiperone) with or without 5 to 10 mM zinc was added at various times from 1 to 60 min and the remaining [³H]methylspiperone specifically bound to dopamine receptors was measured by rapid filtration.

Thermodynamic Analysis of Zinc Binding. [³H]methylspiperone was used as the D₂-selective antagonist to measure the thermodynamics of zinc binding to the D₂-like dopamine receptors. The equilibrium binding conditions and radioligand concentrations were the same as those described above, except that the equilibration time was 2 h and binding was carried out at 0, 10, 23, and 37°C. The temperature of the wash buffer was held constant at 0°C. Both zinc inhibition curves and methylspiperone saturation isotherms were carried out at each of the four temperatures for each of the three receptor subtypes, and the resulting K_D values were used in the calculation of zinc K_i values from zinc IC₅₀ values at the various temperatures. K_i values at each temperature were subsequently used to calculate free energy values.

D_2L Dopamine Receptor-Mediated Agonist-Induced Inhibition of Forskolin-Stimulated cAMP Production, Reversal by Methylspiperone, and the Effect of Zinc. The same CHO cell lines that were used in binding experiments were also used to test for the effect of zinc on dopamine receptor-regulated cAMP production. For all cAMP assays, approximately 50,000 CHO cells were seeded in 200 µl of F-12 complete culture media per well of a 96-well flat bottomed microtiter plate, and cultured in an incubator at 37°C, 95% humidity, and 5% CO₂ overnight or until about 90% confluent. Ultimately, variations in actual cell density for separate experiments were corrected by expressing cAMP production values in terms of whole cell protein. Whole cell protein was measured with the bicinchoninic acid assay (Pierce Chemical Co., Rockford, IL) by solubilizing EBSS-rinsed cells in the unused wells with reagent B first. Immediately before each assay, each well was rinsed briefly with 300 µl of EBSS at 37°C. The media for all cAMP assays was Dulbecco's modified Eagle's medium supplemented with 20 mM HEPES, pH 7.4 at 37°C; all drugs and zinc solutions were ultimately diluted in this same media supplemented with 25 µM Ro-20-1724 (a selective inhibitor of cAMP phosphodiesterase), 5 µM ()-propranolol, and 250 µM ascorbate for the whole cell cAMP assay. The cAMP levels were measured in a 96-well format using a cAMP detection kit (Diagnostic Products Corporation, Los Angeles, CA), which determines cAMP levels via a cAMP binding protein competition assay in conjunction with a standard curve. To produce an increase in intracellular cAMP in all groups except the one representing basal levels of cAMP, 10 µM Forskolin was added from a dimethyl sulfoxide stock to yield a final dimethyl sulfoxide concentration no greater than 0.05% v/v. Except for the groups marked as basal and Forskolin alone, all groups also contained either the agonist 5 µM ()-N-propylnorapomorphine or 5 µM dopamine, which stimulates D_2L receptors, resulting in a decrease of cAMP accumulation. These concentrations of the two agonists were chosen because they produce about 90% of the maximum attainable response within the defined assay conditions (data not shown). Finally, some groups additionally contained various concentrations of the antagonist methylspiperone (1-10,000 nM). Thus, all reagents and drugs for the cAMP assay were added to cells simultaneously. Using this assay format, the effect of zinc on dopamine receptor function was measured by running two groups in parallel for each experiment: one group in the presence of a saturating concentration of zinc chloride (1 mM) and one in its absence.

Data Analysis and Interpretation. All radioligand binding points were sampled in triplicate or quadruplicate for each experiment, and all cAMP data points were measured in duplicate for each experiment. For all experiments, the averaged values are reported as the arithmetic mean with a standard deviation. When present, the error bars for the data plotted in the graphs represent S.E.M.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Initially, a variety of mono-, di-, and trivalent cations were screened for their effects on [³H]methylspiperone binding to all the cloned D₂-like dopamine receptors. The overall pattern of cations that effectively inhibited [³H]methylspiperone binding (data not shown), was essentially the same as that seen previously for antagonist binding to cloned D_1A and D_2L receptors (Schetz and Sibley, 1997). Table 1 lists dose-response data and the calculated rank order of inhibition of [³H]methylspiperone binding to all D₂-like dopamine receptors by pseudo-noble-gas-configuration (PNGC) cations (Cu²⁺, Zn²⁺, Ag⁺, Cd²⁺, Au³⁺, and Hg²⁺). In every case, the PNGC cations completely inhibited the specific binding of [³H]methylspiperone and competition curves were best fit as having one site, but the pseudo Hill slope coefficients (n_H) varied drastically. For example, for all D₂-like subtypes, n_H  1.0 for copper and n_H  3.0 for silver, but in the case of cadmium, n_H  1.0 at D_2L receptors and n_H  2-3 at D₃ and D₄ receptors. Significant differences in K_i values for silver and copper were observed at the three dopamine receptor subtypes, and for both cations, D₄ receptors were the most resistant and D_2L receptors were the least resistant to inhibition. Remarkably, the rank order binding affinities for cadmium and copper alone reveals a pharmacological distinction between D_2L receptors, and D₃ and D₄ receptors (Table 1). The pharmacological mechanisms of zinc's actions on dopamine receptor subtypes were examined more closely as the n_H values are slightly shallow for zinc at D_2L and D₃, but not D₄, receptors.

Zinc inhibition of antagonist binding to rat D_2L, D₃, and D₄ dopamine receptors was assessed by measuring IC₅₀ values for zinc at progressively increasing concentrations of [³H]methylspiperone, i.e., competition curve, Schild-type of analysis (Fig. 1). In all cases, zinc completely and dose-dependently inhibited specifically bound antagonist. For all three receptor subtypes, the measured IC₅₀ values increased only 1.4- to 2-fold for about a 100-fold change in radiolabeled antagonist concentration, and for all curves at all concentrations a one-site competition model with a variable pseudo Hill slope was the best fit to the data. Theoretical (expected) curves arising from a perfectly competitive model of inhibition are plotted along with the data for comparison. Notably, there are large deviations of the measured IC₅₀ values from those derived from a perfectly competitive model of inhibition, especially at higher concentrations of radiolabeled antagonist. When nonisotopic methylspiperone was used as the inhibitor of [³H]methylspiperone binding for comparison, the measured rightward shift in IC₅₀ values at progressively higher concentrations of radioligand closely approximates the theoretical model for a perfectly competitive inhibitor (Fig. 1D).

View larger version (31K): [in this window] [in a new window]   Fig. 1.   Dose-response inhibition of [3H]methylspiperone binding to D2-like dopamine receptors by zinc as a function of increasing concentrations of [3H]methylspiperone. The effect of zinc on [3H]methylspiperone binding was determined by rapid filtration of membranes prepared from CHO cells stably expressing either the rat D2L, D3, or D4 dopamine receptors. For each of the various fixed concentrations of radioligand, the percent specific binding was determined from the total binding in the presence (or absence) of various concentrations of zinc minus the nonspecific binding represented by 1 to 10 µM (+)-butaclamol or 10 µM methylspiperone (n = 3). Zinc inhibition curves of [3H]methylspiperone binding to each D2-like receptor were generated at three different concentrations of [3H]methylspiperone. The open, gray, and solid circles correspond to the experimentally measured binding values from the lowest to highest concentration of radioligand, respectively. The thick solid, fine dashed, and large dashed lines correspond to the expected theoretical binding curves for a purely competitive inhibitor from the lowest to highest concentration of radioligand, respectively. All values listed are the mean ± S.D. A, at D2L receptors, the measured zinc IC50 and nH values at [3H]methylspiperone concentrations of 25 ± 0.3, 263 ± 8, and 2795 ± 35 pM are 55 ± 11 nM and 0.82 ± 0.07, 130 ± 27 nM and 0.88 ± 0.09, and 109 ± 62 nM and 0.54 ± 0.17, respectively. A purely competitive inhibitor has an nH value equal to one and IC50 values equal to 55, 325, and 3200 nM from lowest to highest radioligand concentration. B, at D3 receptors, the measured zinc IC50 and nH values at [3H]methylspiperone concentrations of 63 ± 3, 561 ± 74, and 5925 ± 912 pM are 11 ± 0.2 nM and 1.02 ± 0.27, 16 ± 4 nM and 1.02 ± 0.23, and 16 ± 1 nM and 0.89 ± 0.19, respectively. A purely competitive inhibitor has an nH value equal to one and IC50 values equal to 11, 27, and 201 nM from lowest to highest radioligand concentration. C, at D4 receptors, the measured zinc IC50 and nH values at [3H]methylspiperone concentrations of 52 ± 11, 540 ± 99, and 5610 ± 778 pM are 34 ± 13 nM and 1.19 ± 0.21, 41 ± 13 nM and 1.41 ± 0.23, and 47 ± 6 nM and 1.46 ± 0.22, respectively. A purely competitive inhibitor has an nH value equal to one and IC50 values equal to 34, 61, and 350 nM from lowest to highest radioligand concentration. D, at D4 receptors, the measured methylspiperone IC50 and nH values at [3H]methylspiperone concentrations of 55 ± 5, 542 ± 49, and 5970 ± 325 pM are 645 ± 350 pM and 0.93 ± 0.21, 792 ± 410 pM and 1.03 ± 0.19, and 4470 ± 2390 pM and 1.04 ± 0.18, respectively. A purely competitive inhibitor has an nH value equal to one and IC50 values equal to 645, 1172, and 7046 nM from lowest to highest radioligand concentration. The measured and theoretical IC50 and nH values for methylspiperone inhibition of [3H]methylspiperone binding to D4 were not significantly different, which is consistent with a purely competitive binding interaction.

The effects of zinc on ligand affinity and receptor density values were measured by saturation isotherm analysis in the presence and absence of approximately an IC₅₀ concentration of zinc (Fig. 2). At these concentrations, zinc primarily reduces antagonist affinity for D_2L and D₃ receptors (3- to 4.5-fold) whereas at D₄ receptors, zinc effectively reduces receptor density (Fig. 2). In all cases, saturation isotherms were best fit to a model of one-site saturable binding (Fig. 2). To confirm that the primary effect of zinc on the D₄ receptor is a reduction in B_max, saturation isotherms were also performed in the presence of a 10,000-fold concentration range of zinc (Fig. 3).

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Saturation isotherm analysis of zinc inhibition of [3H]methylspiperone binding to D2-like dopamine receptors and reversal with EDTA. Membranes were pretreated for 1 h at 25°C with approximately an IC50 concentration of zinc (i.e., 150 µM, 18 µM, and 50 µM zinc chloride for D2L, D3, and D4, respectively), before addition of EDTA to some groups at a final concentration of 2.5 mM and radioligand. A representative example of the effect of zinc pretreatment and the subsequent recovery of binding by EDTA is shown for each of the D2-like subtypes.  and , binding in the absence and presence of zinc, respectively. , binding in the presence of zinc and EDTA. Averaged ratios of KD and Bmax values in the absence of zinc divided by these values after recovery with EDTA were 88 ± 33 and 89 ± 2% for D2L, 89 ± 30 and 90 ± 25% for D3, and 103 ± 24% and 86 ± 19% for D4. The average change in KD and Bmax values in the presence of approximately the IC50 for zinc at each receptor were +4.5-fold and 1.4-fold change for D2L, +2.7-fold and 1.4-fold change for D3, and +1.5-fold and 1.9-fold change for D4, respectively (n = 2-3).

View larger version (22K): [in this window] [in a new window]   Fig. 3.   Saturation isotherm analysis of zinc inhibition of [3H]methylspiperone binding to the D4 receptor subtype as a function of increasing concentrations of zinc. A representative experiment for the effect of zinc on [3H]methylspiperone saturation isotherm binding to D4 receptor over a 10,000-fold concentration range of zinc. , binding in the absence of zinc with corresponding KD = 370 pM and Bmax = 2.78 pmol/mg protein. The remaining symbols represent binding in the presence of increasing concentrations of zinc: 2.5 (), 25 (), 250 (), 2500 (), and 25,000 µM (). The respective KD and Bmax values are 330 pM and 2.74 pmol/mg, 648 pM and 2.20 pmol/mg, 544 pM and 0.90 pmol/mg, 654 pM and 0.48 pmol/mg, and 763 pM and 0.22 pmol/mg.

Because equilibrium measures indicated a difference in the mechanism of zinc inhibition, the effect of zinc on the dissociation rate of specifically bound [³H]methylspiperone was investigated for D_2L, D₃, and D₄ receptors (Fig. 4). In the absence of zinc, the measured dissociation rate for specifically bound [³H]methylspiperone at the D_2L receptor was monophasic from 0 to 60 min at 37°C, whereas dissociation from D₃ and D₄ receptors was biphasic over this same time frame. However, in the presence of saturating concentrations of zinc (5-10 mM), [³H]methylspiperone dissociation kinetics for the D_2L receptor became biphasic with the addition of a second faster off rate. The normally biphasic dissociation kinetics for D₃ receptors remained biphasic in the presence of zinc, but one of the phases was greatly accelerated. In contrast to D_2L and D₃ receptors, [³H]methylspiperone dissociation from D₄ receptors was not accelerated in the presence of zinc.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Effect of zinc on the rate of [3H]methylspiperone dissociation from D2-like dopamine receptors. Dissociation rate data for [3H]methylspiperone binding to either D2L, D3, or D4 receptors was transformed to fit the function ln(B/Bo) verses time and plotted. The dissociation rates were measured by first equilibrating dopamine receptors with radioligand at 37°C followed by addition of excess nonisotopic antagonist (2 µM methylspiperone) without significant dilution either in the absence () or presence () of 5 to 10 mM zinc chloride. The averaged concentrations of radioligand used for the dissociation experiments were 668 ± 15 pM for D2L receptors, 831 ± 49 pM for D3 receptors, and 1034 ± 94 pM for D4 receptors. The calculated koff rates for D2L receptors were 0.056 ± 0.023 min1 in the absence and k1 = 0.38 ± 0.16 min1 and k2 = .029 ± 0.011 min1 in the presence of zinc; koff rates for D3 receptors were k1 = 0.172 ± 0.012 min1 and k2 = 0.026 ± 0.025 min1 in the absence of zinc, and k1 = 0.210 ± 0.053 min1 and k2 = 1.96 ± 0.311 min1 in the presence of zinc; and the koff rates for D4 receptors are k1 = 0.23 ± 0.017 min1 and k2 = .029 ± 0.004 min1 in the absence of zinc, and k1 = 0.23 ± 0.028 min1 and k2 = 0.027 ± 0.003 min1 in the presence of zinc (n = 3). A small difference in the proportion of the two rates accounts for the minor slowing of radioligand dissociation in the presence of zinc, however, zinc clearly does not accelerate dissociation.

Because zinc's actions on all subtypes are reversible (Fig. 2) and both kinetic and equilibrium measures of antagonist binding indicate that zinc interacts with the D₄ receptor in a manner distinct from its interaction with D_2L and D₃ receptors, the thermodynamic forces underpinning zinc's interactions with the D₂-like dopamine receptors were examined. The thermodynamics of zinc binding were determined for each D₂-like subtype: G values were calculated from K_i values determined at four different temperatures, H values were calculated from the slope of the van't Hoff plots, and S values were calculated via the second law of thermodynamics. Remarkably, all the D₂-like subtypes displayed entropy-driven favorable free energy changes for zinc binding. The only differences were significantly smaller H and S values for zinc binding to D_2L receptors in comparison to D₃ and D₄ receptors (Fig. 5). Notably, the free energy changes for zinc inhibition of antagonist binding to all D₂-like subtypes were relatively small (G = 1.7 to 2.6 kcal/mol), which is commensurate with the relatively low affinity of zinc binding (micromolar range).

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Thermodynamic analysis of zinc inhibition of antagonist binding to D2-like dopamine receptors. van't Hoff plot analysis of zinc inhibition of [3H]methylspiperone binding to D2-like receptors was performed and the macroscopic enthalpy changes due to zinc binding were extracted from the slope of the line which equals H/KD. The corresponding macroscopic changes in entropy were calculated from the Gibbs free energy equation, G = H  T S, by solving for the measured changes in free energy according to the equation G = RT × lnKD. These calculated free energy, enthalpy, and entropy values of zinc binding for D2L, D3, and D4 receptors, respectively, were G = 1.66, 2.59, and 2.61 kcal/mol, H = 3.05, 5.17, and 7.09 kcal/mol, and S = 1.73, 3.55, and 3.16 cal/K · mole. The average concentration of [3H]methylspiperone used in the zinc competition assays for the D2L, D3, and D4 subtypes was 0.3, 0.8, and 1.1 nM, respectively (n = 2).

Next, an antagonist of a different structural class was assayed for its sensitivity to zinc. The substituted benzamide antagonist, [³H]raclopride, was chosen as the radioligand because it is used in clinical studies. [³H]Raclopride binds to the cloned rat D_2L dopamine receptor with reasonably high affinity (K_D  1 nM) and like most substituted benzamide antagonists, its binding was enhanced by sodium ions (Fig. 6B). Notably, however, the affinity of [³H]raclopride binding was not changed in the presence of 120 mM NaCl, rather the B_max was significantly increased. In the presence of 120 mM sodium, zinc inhibition curves of [³H]raclopride were shifted to the right, indicating about an 8-fold decrease in zinc binding affinity (Fig. 6A). Although 120 mM NaCl decreased zinc binding affinity, the mechanism of the zinc effect on [³H]raclopride binding was the same (Fig. 6B). Because [³H]raclopride binding is, itself, sensitive to sodium, the effect of sodium on zinc binding was tested with the sodium-insensitive radioligand [³H]methylspiperone. In parallel experiments, N-methyl-D-glucamine was used as a replacement electrolyte for sodium to examine simple charge effects. Surprisingly, the potency of zinc inhibition of [³H]methylspiperone binding was decreased approximately 5-fold in the presence of 120 mM NaCl, whereas 120 mM N-methyl-D-glucamine had essentially no effect (Fig. 6C).

View larger version (22K): [in this window] [in a new window]   Fig. 6.   Zinc inhibition of [3H]raclopride binding to the D2L dopamine receptor and modulation by 120 mM NaCl. A, zinc inhibition of [3H]raclopride binding to D2L receptors in the presence or absence of 120 mM NaCl (n = 4). The corresponding IC50 and nH values are 18 ± 12 µM and 1.7 ± 0.6 in the absence and 166 ± 56 µM and 1.0 ± 0.2 in the presence of sodium. The average concentration of [3H]raclopride used in these competition assays was 724 ± 60 pM. B, saturation isotherm analysis of [3H]raclopride binding to D2L receptors as a function of 120 mM sodium and zinc. The averaged KD and Bmax values in the presence (solid symbols) and absence (open symbols) of sodium were 860 ± 97 pM and 1.86 ± 0.71 pmol/mg protein and 957 ± 76 pM and 3.26 ± 0.70 pmol/mg protein, respectively. When present ( and ), zinc was added at a concentration of 18 and 180 µM in the absence and presence of 120 mM NaCl, respectively. The averaged fold changes in KD and Bmax values in the presence and absence of zinc (+Zn/Zn) and sodium were +3.2 ± 1.2- and 1.7 ± 0.2-fold change due to zinc in the absence of sodium and +2.4 ± 0.3- and 1.4 ± 0.3-fold change due to zinc in the presence of sodium, respectively (n = 3). C, zinc inhibition of [3H]methylspiperone binding in the absence or presence of either 120 mM NaCl or 120 mM N-methyl-D-glucamine (n = 2). The corresponding IC50 values in the absence of sodium () or in the presence of sodium () or N-methyl-D-glucamine () are 128 ± 19 µM, 691 ± 98 µM, and 125 ± 67 µM, respectively. The average concentration of [3H]methylspiperone used in these competition assays was 513 ± 28 pM. The dashed line corresponds to the best fit curve through the N-methyl-D-glucamine binding data.

The functional effects of zinc were assessed by measuring the ability of zinc to alter D_2L receptor-regulated inhibition of Forskolin-stimulated cAMP production in intact cells (Fig. 7). Only D_2L receptors were examined in CHO cells because D₃ receptors are not clearly involved in the regulation of cAMP production and the D₄ receptors responded weakly. Zinc had no significant effect on the ability of agonists to stimulate inhibition of cAMP production via D_2L receptors. In contrast, 1 mM zinc retarded the methylspiperone-induced reversal of both the propylnorapomorphine- and dopamine-stimulated cAMP responses at the D_2L receptor. Although these patterns of antagonist reversal of cAMP production in the presence and absence of zinc remained constant between experiments, the absolute cAMP values varied greatly. A significant inhibitory effect of zinc (P < .05) was observed at higher concentrations of methylspiperone, i.e., when functional antagonism was maximal. The inhibition of D_2L receptor functional antagonism by zinc appears to be due to a reduction in the maximal response as opposed to a shift in potency (EC₅₀).

View larger version (30K): [in this window] [in a new window]   Fig. 7.   Effect of zinc on agonist-induced inhibition of cAMP production in attached, intact CHO cells expressing the D2L dopamine receptor. The effect of zinc on D2L dopamine receptor function was assessed by measuring dopamine receptor-mediated changes in cAMP production. Attached CHO cells were incubated in 20 mM HEPES/Dulbecco's modified Eagle's medium, pH 7.4 at 37°C in a CO2 incubator for 15 min with various agonist and antagonist either in the presence or absence of 1 mM zinc (n = 3). The brackets below each graph indicate the other reaction components or drugs added to each group. MSP, methylspiperone. *group containing zinc (solid columns) is significantly different at 95% CL (Neuman-Keuls) from the group with no zinc added (open columns). The approximate EC50 values for functional antagonism of the D2L receptor by MSP were 1 and 2 µM in the presence and absence of zinc, respectively, when propylnorapomorphine was used as agonist, and 12 nM and 10 nM in the presence and absence of zinc, respectively, when dopamine was the agonist.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Previous studies of the effect of heavy metal cations on antagonist binding to D₂-like dopamine receptors used membranes prepared from brain tissues that contained various proportions of dopamine receptor subtypes as well as other receptors (e.g., serotonin receptors) that bind the radioligands used (Oliveria et al., 1983, Scheuhammer and Cherian, 1985). Here, CHO cell lines expressing a single subtype of cloned D₂-like receptor were used so that any subtype-selective effects might be clearly distinguished. All PNGC cations inhibited [³H]methylspiperone binding to D₂-like dopamine receptors, and on the basis of copper and cadmium rank order binding affinities alone, the D_2L receptor subtype was significantly unique from D₃/D₄ receptors. These distinctions in binding affinity are due to the significant gradient of copper sensitivity for the D₂-like subtypes. Significant deviations of pseudo Hill slope values from unity were observed for some PNGC cations, and in the case of cadmium, n_H values could distinguish D_2L receptors from D₃/D₄ receptors as well.

In a recent study, we demonstrated that zinc was allosterically modulating [³H]antagonist binding to D_1A and D_2L receptors (Schetz and Sibley, 1997), but did not rule out the possibility that zinc might somehow first complex with radioligand before binding dopamine receptors. To resolve this issue, the effects of zinc on the D₂-like receptor subtypes were analyzed by competition curve style Schild-type plots. The basic premise of this type of analysis is to perform competition assays at increasingly higher concentrations of radioligand, then to compare the degree and direction of the shift in IC₅₀ values. For an uncompetitive binding interaction, a leftward shift in the competition curve would be expected because more radioligand would be present to form a complex with inhibitor before acting on the receptor (Cheng and Prusoff, 1973). However, no such increase in zinc binding affinity was observed at higher concentrations of radioligand. Instead, all receptor subtypes showed apparent decreases in binding affinity, albeit the magnitude of the IC₅₀ shifts were far less than that expected for a purely competitive binding interaction (Cheng and Prusoff, 1973; Ehlert, 1988). A purely noncompetitive binding interaction would result in absolutely no shift in IC₅₀ values as the affinity for a noncompetitive site is independent of the concentration of primary ligand. Alternatively, a negative heterotropic cooperative binding interaction with a low cooperativity value () would result in a small rightward shift that may be difficult to differentiate from noncompetitive binding. Because the magnitude of the measured rightward shifts in zinc competition curves were relatively small, the n_H values for D_2L and D₃ were shallow, and high concentrations of zinc can also reduce apparent D_2L receptor density (Schetz and Sibley, 1997), the mechanism of zinc inhibition was additionally examined by saturation isotherm analysis.

Saturation isotherm binding of [³H]methylspiperone in the presence of an IC₅₀ concentration of zinc results in a decrease in K_D values for all D₂-like receptor subtypes. Notably in the case of D₄ receptors, the change in the K_D was smaller than the change in B_max. Thus, at lower zinc concentrations (IC₅₀), a minor component of the effect of zinc on D₄ receptors appears to be similar to the primary mechanism for zinc inhibition of antagonist binding to D_2L and D₃ receptor subtypes, i.e., a reduction in K_D. At saturating concentrations of zinc, the maximal number of [³H]methylspiperone binding sites to D_2L receptors is substantially reduced (Schetz and Sibley, 1997), which is similar to the primary effect of zinc on D₄ receptors.

In general agreement with these equilibrium studies, the rates of [³H]methylspiperone dissociation in the presence of zinc were accelerated for D_2L and D₃ receptors, but not for D₄ receptors. At D₃ receptors, one of the two dissociation phases in the absence of zinc is accelerated in the presence of zinc, whereas at the D_2L receptor, zinc actually appears to induce the additional, accelerated phase. One interpretation is that the transition state induced by zinc at D_2L receptors is normally present in D₃ receptors and that zinc exaggerates this transition further. Even though zinc did not affect either of the two [³H]methylspiperone dissociation rates at the D₄ receptor, the small but noticeable differences observed for the kinetic rate curves in the presence and absence of zinc are due to an alteration in the proportion of the two dissociation rates (K_off).

The overall effect of zinc on all subtypes was entropy-driven, with the largest entropy changes observed at D₄ receptors. Thus, macroscopic quantities (thermodynamics and reversibility) did not reveal the subtype-selective differences that were observed for the molecular mechanisms of [³H]antagonist binding. Favorable entropy changes associated with zinc binding are remarkable because it is the bulky, lipophilic dopamine receptor ligands that are known to increase entropy upon binding (Kilpatrich et al., 1986) and zinc is small and charged.

Like [³H]methylspiperone, [³H]raclopride binding affinity was reduced in the presence of zinc. Addition of 120 mM NaCl to the binding and wash buffers enhanced specific raclopride binding, but the increased binding was due to a 2-fold increase in receptor density rather than the 2-fold increase in affinity that has been reported previously (Malmberg et al., 1993). Although the apparent K_i for zinc is decreased 8-fold in the presence of sodium, [³H]raclopride saturation isotherm analysis in the presence of dose-equivalent concentrations of zinc and in the presence and absence of sodium demonstrates that the basic mechanism of zinc inhibition of raclopride binding (via a reduction in affinity) is indifferent to sodium ions. More important, when the sodium-insensitive antagonist [³H]methylspiperone was used as the radioligand for generating zinc competition curves, the affinity of zinc inhibition was decreased approximately 5-fold in the presence of 120 mM NaCl but not 120 mM N-methyl-D-glucamine. Thus, sodium regulation of zinc binding to D_2L receptors is not a simple charge effect. Moreover, two lines of evidence suggest that the binding interactions between zinc and sodium are likely to be allosteric (indirect), rather than zinc directly competing with sodium for the sodium binding site (i.e., aspartate at amino acid 80 of the rat D_2S receptor; Neve et al., 1991). First, sodium increases [³H]raclopride B_max values, whereas zinc decreases [³H]raclopride binding affinity. Second, zinc but not sodium affects [³H]methylspiperone binding.

In addition to its effects on drug binding, zinc inhibited functional antagonism by methylspiperone, even though zinc alone did not act as an agonist (or antagonist) of D_2L receptors, i.e., zinc has no efficacy. In intact cell functional assays, zinc suppressed methylspiperone reversal of the agonist-induced D_2L receptor response by lowering the apparent maximal effect of methylspiperone reversal. This was surprising because an IC₅₀ concentration of zinc primarily reduced antagonist binding affinity to membranes. This data, however, is consistent with our previous studies (Schetz and Sibley, 1997) in which a notable decrease in B_max was observed in antagonist Schild-type plots of D_2L at saturating concentrations of zinc. These discrepancies between the whole cell functional assays and the membrane binding assays could ultimately reflect differences in sidedness (intra- versus extracellular) or accessibility to the zinc site(s). Future studies with intact cells and zinc ionophores might help resolve this issue.

In summary, zinc modulates antagonist binding to the entire subfamily of D₂-like dopamine receptors. These subtype-selective differences to zinc are revealed only in terms of the molecular mechanisms of inhibition of antagonist binding rather than general differences in the macroscopic binding properties of zinc. At D_2L receptors, zinc inhibits both functional antagonism and the binding of radiolabeled antagonists, and this inhibition by zinc is sodium-sensitive. Interestingly, the calculated affinity values of zinc for the allosteric site on dopamine receptors is in the low micromolar range. This is approximately in the same concentration range in which zinc reportedly allosterically modulates other receptor systems that are also implicated in basal ganglia circuitry (e.g., N-methyl-D-aspartate and GABA receptors), and is still well below the estimated concentrations of zinc that one might expect to find in the synapse after depolarization-induced synaptic release of zinc (ca. 300 µM). Furthermore, these concentrations of zinc would still be relevant, even though zinc binding is sodium-sensitive, because in 120 mM NaCl the K_i for zinc binding to the D_2L receptor would be approximately 50 to 60 µM.

A neuromodulatory role for zinc has been proposed previously (Hesse, 1979; Assaf and Chung, 1984; Howell et al., 1984) and subsequent histochemical studies have indicated that zinc is present in synaptic vesicles in several brain regions including the striatum and amygdala (Perez-Clausell and Danscher, 1985). Recently, Norregaard et al. (1998) demonstrated that micromolar concentrations of zinc negatively modulate the dopamine transporter in vitro and potentiate the binding of an antagonist of the dopamine transporter, WIN 356,428. In the present report, we provide the first evidence to suggest that direct neuromodulation of dopamine receptors by zinc could possibly impact on the effectiveness of drug therapies that rely upon dopamine receptor antagonists. Future in vivo studies with a specific zinc chelator may help to clarify whether zinc plays a physiologically relevant role in dopaminergic drug therapies in addition to it already being useful as a probe of the molecular structure and molecular pharmacology of dopamine receptor proteins.