Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Although the causes of schizophrenia remain elusive, many neurotransmitter systems have been implicated in the pathophysiology of schizophrenia. The most widely used therapeutics in the treatment of schizophrenia have included those that block transmission at dopamine receptors (for review, see Seeman and Van Tol, 1994). Five pharmacologically distinct dopamine receptors and variants have been identified by use of molecular cloning techniques (for review, see Civelli, 1995). On the basis of structural and functional similarities, all these dopamine receptor subtypes are in one of two categories, designated D₁-like (D₁ and D₅) and D₂-like (D₂, D₃ and D₄). Receptors in the D₁-like family couple positively to the enzyme adenylyl cyclase and increase intracellular cAMP, whereas D₂-like receptors exert an inhibitory effect on this enzyme (Sibley and Monsma, 1992). Unfortunately, other functions of these dopamine receptor subtypes are not yet completely understood (Civelli, 1995).

The dopamine D₄ receptor has been proposed as a therapeutic target for the treatment of schizophrenia for several reasons. The atypical antipsychotic, clozapine, has a 10-fold greater affinity for the D₄ receptor subtype than for the D₂ receptor subtype (Van Tol et al., 1991). Furthermore, the clinically efficacious dose of clozapine correlates well with the affinity of clozapine at the D₄ receptor (Seeman, 1995; Seeman and Van Tol, 1995). Additional evidence reported by several laboratories includes selective increases in D₄ receptor binding density in postmortem brain tissue from schizophrenics (Seeman et al., 1993; Sumiyoshi et al., 1994; Murray et al., 1995). This latter finding remains controversial (Lahti et al., 1996; Reynolds, 1996; Reynolds and Mason, 1995). Because of the lack of selective ligands for the dopamine D₄ receptor, the density of D₄ receptors in brain tissue has been measured only indirectly as the difference in maximal binding density between [³H]YM-09151-2 (with affinity for D₂, D₃ and D₄ receptor subtypes) and [³H]raclopride (with affinity for only D₂ and D₃ receptor subtypes). Although this indirect methodology has been informative, the direct characterization of dopamine D₄ receptors in brain tissue requires the development of subtype-selective ligands for the D₄ receptor.

This paper reports the identification of NGD 94-1 (2-phenyl-4(5)-[4-(2-pyrimidinyl)-piperazin-1-yl)-methyl]-imidazole dimaleate), a novel, highly selective antagonist at the dopamine D₄ receptor. NGD 94-1 was shown to bind with high affinity to human D_4.2 receptors expressed in mammalian cells whereas its affinity at D₁, D₂, D₃ and D₅ receptor subtypes was 2 µM. The binding profile of NGD 94-1 at a wide variety of neurotransmitter receptors and modulatory sites is summarized in the present study. In addition, measures of in vitro functional activity of NGD 94-1 at the D_4.2 receptor subtype suggest antagonism by NGD 94-1. Pharmacological characterization of the binding of tritiated NGD 94-1 to cloned human D₄ receptors is also reported.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Chemicals. [³H]2-Phenyl-4(5)-[4-(2-pyrimidinyl)-piperazin-1-yl)-methyl]-imidazole dimaleate ([³H]NGD 94-1) was custom-labeled by ChemSyn Science Laboratories (Lenexa, KS; 37-44 Ci/mmol). The following radioligands were purchased from NEN-DuPont: [³H]YM 09151-2, [³H]SCH 23390, [³H]prazosin, [³H]ketanserin, [³H]8-OH-DPAT (dipropylaminotetralin), [³H]DTG (ditolylguanidine), [³H]TCP (thienylcyclohexylpiperdine) and [¹²⁵I]peptide YY ([¹²⁵I]PYY). [³H]Mesulergine and [³H]RX 781094 were purchased from Amersham (Arlington Heights, IL). Olanzepine was a generous gift from Eli Lilly & Company (Indianapolis, IN); risperidone was a generous gift from Janssen (Beerse, Belgium); raclopride was a generous gift from Astra (Moindal, Sweden). The chemical structure of NGD 94-1 is shown in figure 1 and was synthesized in house by Neurogen chemists. Compounds under investigation were dissolved in ethanol, dimethyl sulfoxide or deionized water, and subsequent dilutions were made with either assay buffer or deionized water (depending on the assay).

View larger version (11K): [in this window] [in a new window]   Fig. 1.   Structure of NGD 94-1.

Cloning of dopamine receptors. The human D_4.2 minigene in the expression vector pCD-PS (Van Tol et al., 1991) was obtained from the Oregon Health Sciences University. To improve expression levels, the two introns in the D_4.2 receptor coding sequence were removed. An intronless synthetic NotI/KasI D_4.2 fragment was prepared by ligation of four pairs of overlapping oligos representing the intronless sequence between the restriction sites. This ligation was performed initially using T4 DNA ligase (Boehringer Mannheim, Indianapolis, IN). An aliquot of this ligation reaction was then used as template in a ligase chain reaction amplification using Pfu ligase (Stratagene, La Jolla, CA). A fragment of the apparent appropriate size was excised from an agarose electrophoresis gel and purified using -agarase (FMC Corp., Rockland, ME). The corresponding intron-containing NotI/KasI fragment of the D_4.2 minigene was excised using a KasI digest followed by a partial digest with NotI to liberate the vector and flanking D_4.2 coding sequence as an intact fragment. The synthetic intronless fragment was then ligated into the vector/flanking-sequence fragment. The resultant plasmid, pD4, was confirmed by sequencing to be the full-length intronless human D_4.2 coding sequence in pCD-PS.

Expression of dopamine clones. COS-1 and CHO-K1 cells were purchased from ATCC (Rockville, MD). For transient expression of recombinant receptors, COS-1 cells cultured in 175-cm² T-flasks in DMEM containing 10 mM HEPES and 5% FBS were transfected using lipofectin. After 3 days, the cells were detached using a nonenzymatic cell dissociation solution, pelleted by centrifugation, washed once with PBS, centrifuged and the final pellets stored at 80°C. Stable cell lines expressing recombinant receptors were isolated by calcium phosphate transfection (Graham and van der Eb, 1973) of CHO-K1 cells maintained in Ham's F-12 medium containing 10 mM HEPES and 10% FBS. Transfectant clones were selected in the presence of G418 (550 mg/ml). For stable lines expressing the D_4.2 receptor, pD4 (D₄) was cotransfected with pSV2Neo (Clontech, Palo Alto, CA) to confer G418 resistance to the transfected cells.

Membrane preparation. For the cloned dopamine receptor assays, pellets containing cloned membranes were thawed on ice and resuspended in ice-cold 50 mM Tris buffer (pH 7.4 at 25°C) containing 120 mM NaCl, 1 mM EDTA and 5 mM MgCl₂. All subsequent work was performed on ice. The membranes were homogenized using a Brinkman Polytron (10 sec, setting 5). The homogenate was centrifuged at 48000 × g and 4°C for 10 min. The pellet was resuspended in fresh buffer and the centrifugation was repeated. The pellet was again resuspended in fresh buffer and centrifuged a final time at 48000 × g and 4°C for 10 min. The pellet was resuspended to the appropriate final protein concentration with 50 mM Tris buffer (pH 7.4 at 25°C) containing 120 mM NaCl. The protein content was determined using the Bio-Rad assay (Hercules, CA), with bovine plasma -globulin as the standard. For the receptor binding assays using brain tissue, the tissue of choice was dissected on ice from male Sprague-Dawley rat brains (fresh/frozen, stored at 20°C; PelFreez, Rogers, AR), and the tissue preparation was performed as described in the references (see table 1). For the NPY₁ binding assay, SK-N-MC cells (ATCC; Rockville, MD) were plated into 24-well plates. When confluent, the intact cells were used in the binding assay as previously described (Gordon et al., 1990).

Radioligand binding assays to determine selectivity. The procedures for the radioligand binding assays were performed using the tissue and assay conditions as summarized in table 1. All assays were validated for specificity using appropriate reference standards. NGD 94-1 was tested in each receptor binding assay at concentrations ranging from 10 pM to 10 µM.

[³H]NGD 94-1 binding assay. Each sample was tested in triplicate in a final volume of 1.0 ml in 12 × 75 mm polypropylene test tubes containing 1.0 nM [³H]NGD 94-1 and cloned human D₄ membranes (40 mg protein stable expression) in 50 mM Tris buffer (pH 7.4 at 25°C) containing 120 mM NaCl. Nonspecific binding was defined with 1 µM spiperone. After a 90-min incubation for [³H]NGD 94-1 at 25°C in a shaking waterbath, the samples were rapidly filtered through Whatman GF/C filters. The filters were rinsed with two 5-ml washes of chilled assay buffer. After equilibration of the filters in 3 ml of Ultima Gold scintillation fluid, radioactivity bound to the filters was quantitated with liquid scintillation spectroscopy at an efficiency of 55-65%. For saturation analysis, 10-12 concentrations of [³H]NGD 94-1 (0.04-13.0 nM) were tested in duplicate for binding at cloned human D_4.2 receptors in four independent experiments. For analysis of binding parameters at human D_4.4 and D_4.7 receptors using [³H]NGD 94-1, the methodology used was identical with that used for human D_4.2 receptors described above. [³H]YM 09151-2 binding was assessed in CHO membranes expressing the human D_4.2 receptor using 10-12 concentrations (0.03-13.0 nM) after a 60-min incubation at 25°C with 1 µM spiperone as the nonspecific ligand.

Adenylate cyclase assay. The dopamine D₄ receptor is a G-protein-coupled receptor and is negatively linked to adenylate cyclase, so that agonists at D₄ receptors will inhibit forskolin-stimulated cAMP production (Asghari et al., 1995). Total cell cAMP content was measured by a modification of the method of Lobaugh and Blackshear (1990). CHO cells stably expressing the cloned human D_4.2 receptor were plated in 24-well plates 1 to 2 days before the assay, and were grown to confluence in Ham's media with 10% FBS. On the day of the assay, each well was washed three times with serum-free Ham's media containing 0.1 M HEPES. After a 30-min preincubation with drug and 50 µM IBMX (phosphodiesterase inhibitor) at 37°C, cells were incubated with 5 µM forskolin for 15 min at 37°C. To stop the reaction, the plates were washed three times with cold PBS. Each well was incubated with 0.1 mM HCl for 20 min at room temperature. An aliquot of each sample was transferred to 12 × 75 mm polypropylene tubes, and the acid was neutralized with the addition of a solution containing 0.1 mM HEPES-0.1 mM K₂CO₃. The remaining acid was removed from the plates, and the cells were lysed with 0.5% Triton X-100. The protein content of each well was then determined with the Bio-Rad protein assay (Hercules, CA). The cAMP content in the neutralized extracts was determined with a cAMP RIA kit (NEN-DuPont, Boston, MA). The samples were quantitated using a gamma counter with an efficiency of 80 to 85%.

GTP³⁵S binding. Agonist-induced GTP³⁵S binding by G-protein-coupled receptors provides a functional measure of G-protein activation. This assay has been widely used for many G-protein-coupled receptors and offers the possibility to distinguish agonists from antagonists and to determine potency and efficacy of agonists (and partial agonists) for a given G-protein-coupled receptor (Thomas et al., 1995; O'Boyle and Lawler, 1995). Moreover, inverse agonist activity can be measured with this assay (Thomas et al., 1995). GTP³⁵S binding activity was measured by a modification of a previously described method (Wieland and Jakobs, 1994). CHO cells stably expressing the human D_4.2 receptor were grown to confluence in Ham's media supplemented with 10% fetal calf serum, harvested and then stored as pellets at 80°C. Thawed cells were homogenized using a Polytron (30 sec, setting 5) in 50 mM Tris, pH 7.4, 10 mM MgCl₂ and 2 mM EGTA. Membrane homogenates were centrifuged at 14,000 × g for 10 min and the pellet was washed one time in cold PBS. The final pellet was resuspended in homogenization buffer and stored at 80°C. On the day of the assay, thawed membrane homogenates were resuspended in assay buffer (50 mM Tris, pH 7.4, 120 mM NaCl, 10 mM MgCl₂, 2 mM EGTA, 0.1% BSA, 0.1 mM bacitracin, 100 KIU/ml Aprotinin, 5 µM GDP) and added to reaction tubes at a concentration of 25 µg/0.200 ml. Reactions were initiated by the addition of 100 pM GTP³⁵S and of individual compounds ranging in concentration from 0.1 nM to 10 µM. After a 30-min incubation at 27°C with mild shaking, the reaction was terminated by vacuum filtration over GF/C filters with ice-cold wash buffer (50 mM Tris, pH 7.4, 5 mM MgCl₂). Bound GTP³⁵S was determined by liquid scintillation spectrometry. Nonspecific binding was defined by 10 µM GTPS and represented less than 10% of total binding.

GTP shift. G-protein-coupled receptors exist in both a high-affinity agonist and a low-affinity agonist state. G-protein activation by GTP has been shown to define the low-affinity agonist state (Zahniser and Molinoff, 1978; Grigoriadis and Seeman, 1985). NGD 94-1 binding was characterized at human D_4.2 receptors in the absence and presence of 200 µM GTP. The binding of NGD 94-1 and of reference compounds (dopamine, haloperidol and ()-eticlopride) was analyzed in the presence and absence of 200 µM GTP at membranes expressing human D_4.2 receptors with 0.1 nM [³H]YM 09151-2. Nonspecific binding was defined by 1 µM spiperone. The reaction was terminated by rapid vacuum filtration through Whatman GF/C filters after a 120-min incubation at room temperature. After equilibration of the filters in 3 ml of Ultima Gold scintillation fluid, radioactivity bound to the filters was quantitated with liquid scintillation spectroscopy at an efficiency of 55 to 65%. [³H]NGD 94-1 binding at cloned D_4.2 receptors was also examined in the absence and presence of 200 µM GTP with the assay conditions described above for [³H]NGD 94-1.

Data analysis. Binding data were analyzed by the nonlinear curve-fitting program RS/1 (BBN Software Products Corp., Cambridge MA) or SigmaPlot (Jandel Scientific, San Rafael, CA). Kinetic data were converted to a K_d value by the following equation (Bylund and Yamamura, 1990): K_d = k₁/k₊₁, such that k₊₁ = (k_obs  k₁)/[L], where [L] is the radioligand concentration. Calculated IC₅₀ values were converted to K_i values by the Cheng-Prusoff correction (Cheng and Prusoff, 1973) with the following equation: K_i = IC₅₀/(1 + [L]/K_d), where [L] is the radioligand concentration and K_d is the dissociation constant for the radioligand, as determined by saturation analysis. Analysis of variance analysis of the cyclase data was performed with StatView (BrainPower Inc., Calabasas, CA) to detect any significant differences between treatments.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Selectivity of NGD 94-1 for the D_4.2 receptor. For the cloned dopamine receptor binding assays, no specific binding was observed with either wild-type COS-1 and CHO-K1 cell membranes or vectors without the appropriate dopamine receptor DNA (data not shown). However, dopamine receptor subtypes were appropriately expressed as measured by RNA expression (slot blot) and by restriction enzyme digestion. NGD 94-1 had the highest affinity (K_i value = 3.6 ± 0.6 nM) for the cloned human D_4.2 receptor, as compared with the other dopamine receptors (table 2). This affinity of NGD 94-1 for the cloned human D_4.2 receptor was more than 600 times higher than that observed for both primate and rat D₂ receptors. NGD 94-1 was inactive at D₁, D₃ and D₅ receptors, because NGD 94-1 inhibited specific binding less than 50% up to a concentration of 10 µM. Of the other receptors tested, 5-HT_1A was the only one for which NGD 94-1 had appreciable affinity (K_i value = 180 ± 10 nM).

Kinetics of [³H]NGD 94-1 binding

¹

1

¹

Saturation analysis of [³H]NGD 94-1 binding. A representative saturation curve for [³H]NGD 94-1 binding at human D_4.2 receptors stably expressed in CHO cells is shown in figure 2. The average K_d and B_max values from four independent experiments were 2.24 ± 0.29 nM and 4047 ± 295 fmol/mg protein, respectively. The linear Rosenthal plot (see inset of fig. 2) further demonstrates the one-component binding to cloned human D_4.2 receptors observed with [³H]NGD 94-1. A comparison of [³H]NGD 94-1 binding to human D_4.2, D_4.4 and D_4.7 receptor variants showed high-affinity, saturable binding that was similar between the three receptor variants. The affinities (K_d) of [³H]NGD 94-1 for the D_4.2, D_4.4 and D_4.7 receptor variants were 1.4 nM, 1.7 nM and 1.1 nM, respectively.

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Representative saturation curve for [3H]NGD 94-1 binding to cloned human D4.2 receptors stably expressed in CHO cells. Each concentration was tested in duplicate by use of 10 to 12 concentrations of [3H]NGD 94-1 (0.04-13 nM). The average Kd and Bmax values, as determined by computer analysis of the saturation isotherm data for four independent experiments, were 2.24 ± 0.29 nM and 4047 ± 295 fmol/mg protein, respectively. The inset shows the corresponding linear Rosenthal plot of the data.

Pharmacology. The affinities of a variety of dopaminergic standards, as well as some newer generation antipsychotic agents, were assessed in the [³H]NGD 94-1 binding assay. Representative inhibition curves are shown in figure 3 and a complete summary of K_i (nM) values are listed in table 3. Spiperone and haloperidol had high affinity (K_i values of 0.31 and 3.3 nM, respectively) for the D_4.2 receptor labeled by [³H]NGD 94-1, as did NGD 94-1 itself (2.2 nM). Clozapine had a K_i value of 46 nM at the D_4.2 receptor subtype. Raclopride and (±)7-OH-DPAT, which are reported to have selective affinities for D₂ and D₃ receptors, as well as (+)SCH 23390, which is selective for D₁ and D₅ receptors, displayed weak D_4.2 receptor binding affinities, consistent with the expected D₄ pharmacology (K_i = 2860 ± 250, 520 ± 41 and 2650 ± 220 nM, respectively). The Hill coefficients of the [³H]NGD 94-1 competition curves were approximately 1.0. [³H]NGD 94-1 binding data were also compared with [³H]YM 09151-2 binding data (table 3). The excellent correlation (r² = 0.987) between D_4.2 affinities with the two different ligands demonstrates the overlap in binding of a nonselective (YM 09151-2) and selective (NGD 94-1) ligand in a pure population of D_4.2 receptors. D₄ receptor affinity measured by [³H]NGD 94-1 was also shown to be similar among the D_4.2, D_4.4 and D_4.7 receptor variants (table 4).

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Representative inhibition curves for NGD 94-1, spiperone, clozapine, (±)7-OH-DPAT, (+)SCH 23390 and raclopride in the [3H]NGD 94-1 binding assay with cloned human D4.2 receptors stably expressed in CHO cells. Each point represents the average of triplicate determinations. Best fit curves were drawn with Sigmaplot (Jandel Scientific) and correspond to the nonlinear regression analysis performed with RS/1.

In vitro functional activity. The functional activity of NGD 94-1 was assessed by measuring the ability of NGD 94-1 to block both agonist-induced inhibition of cAMP production and the agonist-induced binding of GTP³⁵S. The basal amount of cAMP produced by the CHO cells expressing human D_4.2 receptors was very low (figure 4). NGD 94-1 (10 µM) and the dopamine receptor full agonist quinpirole (1 µM) had no effect on the basal amount of cAMP formation. The addition of 5 µM forskolin, which directly activates adenylate cyclase, caused a 50-fold increase in cAMP levels in this preparation. At a saturating concentration of 1 µM, quinpirole significantly inhibited forskolin-stimulated cAMP production by approximately 60% (P <.05). NGD 94-1 alone did not inhibit forskolin-stimulated cAMP production, but completely reversed the quinpirole-induced inhibition of cAMP production (P <.05). These data suggest antagonism by NGD 94-1 at the D_4.2 receptor in this receptor population (fig. 4).

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Complete blockade of quinpirole-induced inhibition of cAMP formation by NGD 94-1 in CHO cells expressing cloned human D4.2 receptors. Data are expressed as percent forskolin-stimulated cAMP production and represent the average of four independent experiments. Concentrations used are as follows: 5 µM forskolin (F), 1 µM quinpirole (Q), 10 µM NGD 94-1. * Indicates significantly (P <.05) different from F and F+NGD 94-1. +Indicates significantly (P <.05) different from F+Q.

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Agonist/antagonist effects on GTP35S binding. Representative data are expressed as the percent of the maximal quinpirole response. Each drug concentration was assayed two to three times. (A) GTP35S binding in the presence of increasing concentrations of quinpirole, NGD 94-1, haloperidol and ()-PPP. (B) Dose-dependent reversal of 100 nM quinpirole-stimulated GTP35S binding by NGD 94-1 and haloperidol.

G-Protein activation by GTP. The data in figure 6 show that 200 µM GTP effectively uncoupled the receptor from its G-protein(s). GTP treatment converted cloned human D_4.2 receptors from a high-affinity state (K_i = 728 ± 147 nM) to a low-affinity state (K_i = 1317 ± 147 nM) for dopamine. In contrast, the binding affinity of NGD 94-1 for D_4.2 receptors was not shifted by GTP. Likewise, the binding affinities of the dopamine receptor antagonists ()-eticlopride and haloperidol were also unaffected by GTP (table 5). Saturation analysis of [³H]NGD 94-1 binding to cloned human D_4.2 receptors in the presence and absence of GTP also showed neither a shift in affinity (K_d) nor maximum binding (B_max) by GTP (data not shown).

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Representative inhibition curves for NGD 94-1 and dopamine in the absence and presence of 200 µM GTP using cloned human D4.2 receptors stably expressed in CHO cells. Each point represents the average of three experiments. Receptor binding was assessed in the presence of 0.1 nM [3H]YM 09151-2.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

A considerable amount of basic and clinical research has surrounded the possible involvement of the dopamine D₄ receptor in the pathophysiology of schizophrenia. The pharmacological profile of the D₄ receptor (Van Tol et al., 1991), the localization of D₄ receptor mRNA in the limbic system (Meador-Woodruff et al., 1994) and the possible elevation in density of this receptor (Seeman et al., 1993; Sumiyoshi et al., 1994; Murray et al., 1995; however, see also, Lahti et al., 1996; Reynolds, 1996; Reynolds and Mason, 1995) suggest a role for the D₄ receptor subtype in schizophrenia. However, the lack of D₄-selective compounds has slowed progress in this field.

This paper reports the identification of NGD 94-1, a novel, highly selective antagonist at the dopamine D₄ receptor. NGD 94-1 was shown to bind to cloned human D_4.2 receptors with high affinity (~2 nM) and with greater than 600-fold selectivity over D₁, D₂, D₃ and D₅ receptors and 21 other receptor systems examined (table 2). NGD 94-1 was shown to bind to 5-HT_1A (180 nM) and 5-HT₃ (750 nM) receptors, in which NGD 94-1 was approximately 50- and 200-fold selective for the D_4.2 receptor, respectively. NGD 94-1 shares a common tail with a known anxiolytic, buspirone, which is a relatively high-affinity partial agonist at the 5-HT_1A receptor. This structural feature likely contributes to the moderate 5-HT_1A binding affinity of NGD 94-1.

Tritiated NGD 94-1 displayed binding to the cloned human D_4.2 receptor that was specific, reversible, saturable, protein dependent and of high affinity. The binding of [³H]NGD 94-1 to cloned human D_4.2 receptors expressed in CHO cells displayed a pharmacological profile similar to that observed with [³H]YM 09151-2 at cloned human D_4.2 receptors as well as to previously published D₄ binding affinities. Thus, the binding profile of [³H]NGD 94-1 is clearly similar to D₄ pharmacology, as demonstrated by the high correlation with reported affinity values.

The human dopamine D₄ receptor contains a polymorphism within the putative third cytoplasmic loop of the protein that is characterized by a 48-base pair repeat that varies in number in different individuals (Van Tol et al., 1992). The number of these repeats has been shown to vary from 2 to as many as 10 (Lichter et al., 1993); however, the principal D₄ receptor variants are D_4.2, D_4.4 and D_4.7 (Rao et al., 1994). The multiple alleles of the dopamine D₄ receptor may have functional consequences for some physiological mechanisms that involve dopamine. Some data suggest that increased allele frequency of the D₄ receptor repeat polymorphism may be associated with a greater probability of psychiatric disorder (Petronis et al., 1995; Nakamura et al., 1995). Likewise, the different allelic forms may bind neuroleptic drugs with different affinities (Van Tol et al., 1992). [³H]NGD 94-1 was shown to bind with equal high affinity across the three different dopamine D_4.2, D_4.4 and D_4.7 cloned receptor variants examined. In addition, [³H]NGD 94-1 showed similar pharmacology across the different receptor variants. These findings suggest that [³H]NGD 94-1 does not differentiate between the human recombinant D_4.2, D_4.4 and D_4.7 dopamine receptor variants. Therefore, the use of NGD 94-1 as a therapeutic for disorders related to the dopamine D₄ receptor system should not be limited by or take advantage of differences in affinity, at least at the D₄ receptor variants examined in this study.

D₄ receptors are coupled to G-proteins and are negatively linked to adenylate cyclase (Asghari et al., 1995). In the present study, NGD 94-1 had no effect on either basal or forskolin-stimulated cAMP production. However, NGD 94-1 did completely block the quinpirole-induced inhibition of forskolin-stimulated cAMP production. These data suggest antagonism (but not partial agonism) by NGD 94-1 at the cloned D_4.2 receptor. Other antagonists with D₄ receptor affinity, including YM-09151-2, haloperidol, and clozapine, have also been shown to block dopamine stimulation of cAMP production at D₄ receptor subtypes expressed in CHO cells (Asghari et al., 1995). As an additional functional measure, the effect of NGD 94-1 on GTP³⁵S binding in membranes from CHO cells stably expressing the human D_4.2 receptor was also examined. This assay addresses receptor activation by directly measuring the rate of nucleotide exchange on G-proteins. The assay is performed by measuring the binding of a nonhydrolyzable GTP analog (i.e., GTP³⁵S) to the alpha subunit of the agonist-activated G-protein, which represents an initial step in a cascade that results in effector responses (such as change in cAMP levels or arachidonic acid release). NGD 94-1 produced a complete reversal of GTP³⁵S binding induced by quinpirole, but was unable on its own to affect GTP³⁵S binding, further supporting antagonism, but not partial agonism, by NGD 94-1 at the D₄ receptor. More detailed analysis awaits the discovery of the endogenous physiological responses to D₄ receptor activity in vivo.

Additional evidence of an antagonist profile for NGD 94-1 at the D₄ receptor is provided by the lack of a GTP shift on NGD 94-1 binding affinity at the cloned D_4.2 receptor. G-protein activation by GTP has been shown to convert D₂ receptors from a high-affinity agonist state to a low-affinity agonist state (Grigoriadis and Seeman, 1985; Lahti et al., 1992). On the other hand, either no shift or a reciprocal affinity shift by GTP for antagonist binding has been reported at D₂ receptors (Lahti et al., 1992). In the present study, G-protein activation by 200 µM GTP converted cloned human D_4.2 receptors from a high-affinity state for the agonist, dopamine, to a low-affinity state for dopamine. In contrast, the binding affinity of NGD 94-1 at the cloned D_4.2 receptor, as well as the affinities for dopamine receptor antagonists, haloperidol and ()-eticlopride, was not shifted by GTP. This lack of GTP shift on NGD 94-1 binding affinity, combined with the antagonist profile of NGD 94-1 shown in functional models described in this report (i.e., cAMP production; GTP³⁵S binding), is consistent with an antagonist profile for NGD 94-1 at the D₄ receptor.

In summary, NGD 94-1 was shown to bind selectively and with high affinity to dopamine D₄ receptors. Furthermore, NGD 94-1 affected functional coupling at the D₄ receptor in a manner consistent with that of an antagonist. Results of the clinical efficacy of NGD 94-1 should help to further define the role of the D₄ receptor in the action of antipsychotic medications. In addition, the radiolabeling of NGD 94-1 for use in noninvasive experimental and clinical diagnostic imaging provides the opportunity to visualize D₄ receptor density and occupation in the brains of normal and schizophrenic patients. Autoradiographic studies of postmortem normal and schizophrenic brain tissue with use of [³H]NGD 94-1 will allow quantitative assessment of these D₄ receptor parameters. The development of additional D₄ receptor-selective compounds, like NGD 94-1, will contribute to the study of psychiatric disorders, including schizophrenia, and add tools to our armamentarium of mechanism-specific therapeutic agents.