Binding of [³H]PD 128907, a Putatively Selective Ligand for the D₃ Dopamine Receptor, in Rat Brain: A Receptor Binding and Quantitative Autoradiographic Study

Abstract

[³H]PD 128907 has been proposed as a selective ligand for the D₃ dopamine receptor. This study characterizes the binding of this radioligand in rat brain using in vitro radioligand binding and autoradiographic methods. In radioligand binding studies, [³H]PD 128907 exhibited 0.3 nmol/L affinity for a single, low density site in ventral striatal membranes. The pharmacological profile for [³H]PD 128907 was similar to that of [³H](+)-7-OH-DPAT with the rank order of potency for dopamine agonists being PD 128907 ≈ 7-OH-DPAT ≈ quinpirole ≥ dopamine; for antagonists, spiperone > (+)-butaclamol ≈ domperidone ≥ haloperidol > SCH 23390. Guanyl nucleotides had no effect on the binding of either ligand. These observations indicate labeling of a dopaminergic site with characteristics consistent with the D₃ receptor. In autoradiographic studies, highest densities of [³H]PD 128907–labeled sites were observed in islands of Calleja followed by the nucleus accumbens, nucleus of the horizontal limb of the diagonal band, the molecular layer of cerebellar lobule X, and the ventral caudate/putamen.

Main

[³H]PD 128907 has been identified as a putatively selective ligand for the D₃ dopamine receptor (Akunne et al. 1995). A member of the D₂ receptor family, the D₃ site is of particular interest because its mRNA is expressed preferentially in brain regions such as the nucleus accumbens, olfactory tubercle, and islands of Calleja (Bouthenet et al. 1991; Sokoloff et al. 1990). These brain regions are terminal fields of the mesolimbic dopamine projection that is hypothesized to be involved in producing psychotic symptoms in schizophrenia (Stevens 1973). Furthermore, unlike mRNA for the D₂ receptor, relatively little D₃ mRNA is expressed in the caudate/putamen (Bouthenet et al. 1991; Sokoloff et al. 1990). This suggests that the D₃ site may be a target for novel antipsychotic drugs that might be free of extrapyramidal effects. It has also been suggested that the D₃ receptor may play a role in the reinforcing properties of cocaine (Caine and Koob 1993) and might thus represent a potential target in the treatment of drug abuse.

Since the cloning of the D₃ receptor, several radioligands thought to be selective for this site have been synthesized including [³H]PD 128907 (R-(+)-trans-3,4,4a, 10b-tetrahydro-4-propyl-2H,5H-[1]benzopyrano[3,4-b]-1,4-oxazin-9-ol]) (Akunne et al. 1995) and [³H]7-OH-DPAT (7-hydroxy-diphenylaminotetralin) (Lévesque et al. 1992). However, there is some controversy over the selectivity of [³H]7-OH-DPAT for the D₃ receptors over the D₂ (Gonzales and Sibley 1995). Interaction of the ligand with the sigma site has also been reported (Wallace and Booze 1995). Although confusion over the selectivity of [³H]7-OH-DPAT may be due to the use of different assay protocols, the conflicting reports have limited the utility of this ligand as a tool for the study of D₃ sites in brain. On the other hand, PD 128907 produced behavioral and neurochemical effects in brain suggestive of activity at the D₃ site (Pugsley et al. 1995). Likewise, [³H]PD 128907 exhibited significant selectivity in transfected cell lines (Akunne et al. 1995). However, the utility of this ligand in rat brain has not yet been demonstrated.

Because comprehensive study of a novel receptor system in brain requires the ability to selectively visualize the receptor protein, as well as receptor mRNA, the utility of [³H]PD 128907 as a D₃-selective agent was evaluated. The pharmacological profile, guanyl nucleotide regulation, and regional distribution of [³H]PD 128907–labeled sites in rat brain are described and assessed using receptor binding and quantitative autoradiographic methods.

MATERIALS AND METHODS

[³H]PD 128907 Binding Assays

The [³H]PD 128907 binding assays were performed according to a modification of the methods of Akunne et al. (1995). Discrete brain regions were isolated by freehand dissection on ice from the brains of adult male Sprague-Dawley rats (Harlan Bioproducts, Indianapolis, IN). Brain tissue was homogenized with a PRO250 Homogenizer (setting 4 of 6) for 10 s in 20 volumes of assay buffer (50 mmol/L Tris, 1 mmol/L EDTA; pH 7.4). The crude homogenate was centrifuged twice at 48,000 × g for 15 min, resuspending the pellet in 20 volumes of assay buffer each time. The final pellet from membrane preparation was resuspended in buffer to yield a final concentration of 10 mg original wet weight (o.w.w.)/ml. Binding assays were performed in duplicate in disposable polystyrene tubes. The final assay volume was 0.5 ml. For binding site saturation studies, 8–10 concentrations of (+)-[N-propyl-2,3-³H]PD 128907 (116 Ci/mmol; Amersham, Arlington Heights, IL) were used. For all other studies, the final concentration of [³H]PD 128907 was ∼0.3 nmol/L. Unless otherwise noted, ventral striatal (nucleus accumbens and olfactory tubercle) membranes were incubated with [³H]PD 128907 and various concentrations of competing drugs (RBI, Natick, MA) or 5′-guanylyl-imidodiphosphate (Gpp(NH)p) (Sigma, St. Louis, MO). Binding was initiated by the addition of membrane homogenate. Nonspecific binding was defined by 10 μmol/L quinpirole. Preliminary experiments indicated the optimal incubation time at 23°C to be 3 h. The reaction was terminated by rapid filtration through Whatman GF/B filters pretreated with 0.5% polyethyleneimine using a Brandel cell harvester. Filters were washed 3 times with 3 ml ice-cold buffer (50 mmol/L Tris; pH 7.4 at 23°C), and placed in scintillation vials. After the addition of Beckman Ready Protein+ scintillation cocktail, vials were shaken, allowed to equilibrate for 2 h, and radioactivity quantitated using a Beckman 6500 scintillation counter. Protein concentrations were determined using the BCA method (Pierce, Rockford, IL). Specific binding of [³H]PD 128907 is expressed as fmol/mg protein. Data from saturation and competition experiments were analyzed using the nonlinear least-squares curve-fitting program LIGAND. Results are expressed as the mean ± SE.

[³H](+)-7-OH-DPAT Binding Assays

Assays were performed as described for [³H]PD 128907 (see above) with the following exceptions. Membranes were prepared in assay buffer (50 mmol/L HEPES, 1 mmol/L EDTA; pH 7.4, at 4°C) to yield a final concentration of 7.5 mg o.w.w./ml. The concentration of [³H] (+)-7-OH-DPAT (139 Ci/mmol; Amersham, Arlington Heights, IL) was ∼0.2 nmol/L for single point assays. Nonspecific binding was defined by 10 μmol/L quinpirole. Preliminary experiments indicated the optimal incubation time at 23°C to be 1 h. Bound ligand was separated from free by rapid filtration through untreated Whatman GF/B filters. The wash buffer was 50 mmol/L Tris, 120 mmol/L NaCl; pH 7.4 at 23°C.

Receptor Autoradiography with [³H]PD 128907

Adult male Sprague-Dawley (200–300 g; Charles River Laboratories, Wilmington, DE) were killed by decapitation, brains rapidly removed, frozen in isopentane, and stored at −70°C until sectioning. Sagittal and coronal brain sections (20 μm) were cut on a cryostat, thaw-mounted onto chrome-alum/gelatin-coated slides and stored at −70°C until use.

[³H]PD 128907 autoradiography was performed according to a modification of methods previously described (Levant et al. 1993). Optimal assay conditions were determined in preliminary wash-out and association experiments. Slide-mounted brain sections were brought to room temperature and allowed to dry thoroughly. Duplicate sections from the same animal were used for each data point. Slides were incubated with ∼0.7 nmol/L [³H]PD 128907 in assay buffer (50 mmol/L Tris, 1 mmol/L EDTA, pH 7.4 at 23°C) for 2 h at 23°C. Nonspecific binding was defined in the presence of 1 μmol/L spiperone. After incubation, slides were dipped in ice-cold assay buffer, washed for two consecutive 2-min periods in ice-cold assay buffer, dipped in ice-cold deionized H₂O, and dried under a cool air stream. Radiolabeled sections were subsequently apposed to ³H-Hyperfilm (Amersham, Arlington Heights, IL) with 20 μm [³H]methylmethacrylate autoradiographic standards (Amersham, Arlington Heights, IL) for a period of 12 weeks. ³H-Hyperfilm was developed according to the manufacturer's instructions. Brain sections were then stained with cresyl violet.

Autoradiographic images were digitized and quantified using the Macintosh-based video densitometry program NIH “Image” version 1.4. Best-fit curves of optical density generated by the [³H]methylmethacrylate autoradiographic standards resulted when a Rodbard plot was used to describe the relationship between optical density and radioactivity. Brain regions were identified according to the atlas of Paxinos and Watson (1986). Autoradiograms of coronal sections were sampled bilaterally. Measurements represent average pixel optical density by volume analysis. Results, expressed as fmol/mg tissue equivalent were not corrected for differential quenching by white and gray matter. Digitized images of autoradiograms from Image were used to generate the figures presented in this article.

RESULTS

Radioligand-Binding Studies

Preliminary Studies of [³H]PD 128907 Binding in Ventral Striatal Membranes

Preliminary studies were performed to determine the optimal tissue concentration and length of incubation (data not shown). When used at a concentration equal to the K_D value (∼0.3 nmol/L), [³H]PD 128907 binding reached steady state within 3 h at 23°C. At a final tissue concentration of 10 mg o.w.w./ml, less than 3% of the total [³H]PD 128907 was bound at steady state. Specific binding represented 60% of the total [³H]PD 128907 bound.

Saturation Analysis of [³H]PD 128907 and [³H](+)-7-OH-DPAT Binding

Binding site saturation analyses were performed with [³H]PD 128907 and [³H](+)-7-OH-DPAT in ventral striatal (nucleus accumbens and olfactory tubercle) membranes. Analysis by LIGAND for [³H]PD 128907 (0.01–3 nmol/L) revealed a single binding site with a K_D value of 0.30 ± 0.02 nmol/L and a B_max value of 26 ± 2.1 fmol/mg of protein (Figure 1 ). Binding of [³H](+)-7-OH-DPAT (0.01–3 nmol/L) was consistent with interactions at a single site with a K_D value of 0.18 ± 0.01 nmol/L and a B_max value of 36 ± 2.1 fmol/mg of protein (Table 1).

Figure 1

Saturation isotherms of [³H]PD 128907 binding in rat ventral striatal membranes. Membranes were incubated with 11 concentrations of [³H]PD 128907 (0.01–3 nmol/L) as described in Materials and Methods. Nonspecific binding was defined in the presence of 10 μmol/L quinpirole. Three independent determinations yielded a K_D value of 0.30 ± 0.02 nmol/L and a B_max value of 26 ± 2.1 fmol/mg protein as estimated by the nonlinear least-squares curve fitting program LIGAND. Data from all saturation analyses were consistent with the interaction of the radioligand with a single binding site. Representative results are shown.

Table 1 Affinity and Density of Sites Labeled by [³H]PD 128907 and [³H](+)-7-OH-DPAT in Rat Ventral Striatum and Caudate/Putamen

To evaluate the D₂/D₃ selectivity of [³H]PD 128907 and [³H](+)-7-OH-DPAT, saturation studies were performed in caudate/putamen, a brain area that expresses the D₂ receptor predominantly over the other D₂-related subtypes (for review, see Levant 1996). Saturation analysis of [³H]PD 128907 binding in caudate/putamen was performed using significantly higher concentrations of radioligand (0.5–75 nmol/L) than those used to assess putative D₃ binding. These studies revealed no specific binding of [³H]PD 128907 in caudate/putamen above that attributable to the low levels of putative D₃ binding observed in that tissue.

Saturation analysis of [³H](+)-7-OH-DPAT in caudate/putamen was also performed using significantly higher concentrations of radioligand (0.5–75 nmol/L) than those used to assess putative D₃ binding. In addition to the very low levels of binding attributable to the putative D₃ site, these studies indicated the labeling of a single site with a K_D value of 23 ± 1.5 nmol/L and a B_max value of 218 ± 5.7 fmol/mg protein.

Pharmacological Profile of [³H]PD 128907 and [³H](+)-7-OH-DPAT Binding Sites

The pharmacological profiles of [³H]PD 128907– and [³H](+)-7-OH-DPAT–labeled sites were determined in ventral striatal membranes using a variety of dopaminergic drugs including those shown to possess significant D₂/D₃ selectivity in previous studies (Sokoloff et al. 1990; Lévesque et al. 1992; Levant and De Souza 1993; Akunne et al. 1995) (Table 2). Dopamine agonists exhibited the following rank order of potency in competition with [³H]PD 128907: (±)-7-OH-DPAT ≈ (+)-PD 128907 ≈ quinpirole > dopamine. A similar rank order of potency was observed in competition with [³H](+)-7-OH-DPAT: (+)-PD 128907 ≥ (±)-7-OH-DPAT ≈ quinpirole ≥ dopamine. In competition with [³H]PD 128907, dopamine antagonists inhibited binding with the following rank order of potency: spiperone > (+)-butaclamol ≥ haloperidol ≈ domperidone > SCH 23390 (Figure 2 ). Antagonists inhibited binding of [³H](+)-7-OH-DPAT with a similar rank order: spiperone > (+)-butaclamol ≥ domperidone > haloperidol > SCH 23390. Overall, dopamine agonists and antagonists exhibited highly similar rank order of potencies in competition with [³H]PD 128907 and [³H](+)-7-OH-DPAT and had a correlation coefficient of 0.707 (Table 2).

Table 2 Pharmacological Profile of Sites Labeled by [³H]PD 128907 and [³H](+)-7-OH-DPAT

Figure 2

Competition of [³H]PD 128907 binding by dopamine agonists and antagonists. Rat ventral striatal membranes were incubated with ∼0.3 nmol/L [³H]PD 128907 and up to 11 concentrations of competing drug (10⁻¹¹ to 10⁻⁵ mol/L) as described in Materials and Methods. Data were analyzed and competition curves generated using the nonlinear least-squares curve fitting program LIGAND. Representative results are shown. The quantitative data from these studies are summarized in Table 1. Dopamine agonists: ▴, PD 128907; ▪, 7-OH-DPAT; •, quinpirole; ♦, dopamine. Dopamine antagonists: ▪, spiperone; •, (+)butaclamol; ▴, domperidone; ♦, haloperidol; □, SCH 23390.

Because previous studies have suggested an interaction of [³H]7-OH-DPAT with the sigma site (Wallace and Booze 1995), the affinity of sigma₁ ligand N-allylnormetazocine for [³H]PD 128907– and [³H](+)-7-OH-DPAT–labeled sites was determined. N-allylnormetazocine had negligible affinity in competition for either putatively D₃-selective ligand.

In all instances, competition data were consistent with interactions at a single binding site as analyzed by LIGAND. Because the density of [³H]PD 128907– and [³H](+)-7-OH-DPAT–labeled sites in other brain regions was quite low (see below), pharmacological characterization of sites in other brain areas was not undertaken.

Guanyl Nucleotide Regulation of [³H]PD 128907 and [³H](+)-7-OH-DPAT Binding

In the presence of the nonhydrolyzable GTP analog, Gpp(NH)p (0.1 mmol/L), 99 ± 6% of total specific [³H]PD 128907 binding was observed in ventral striatal membranes. Similarly, 96 ± 4% of total specific [³H](+)-7-OH-DPAT binding was observed in the presence of 0.1 mmol/L Gpp(NH)p. Because the density of [³H]PD 128907– and [³H](+)-7-OH-DPAT–labeled sites in other brain regions was quite low (see below), assessment of the guanyl nucleotide regulation of sites in other brain areas was not undertaken.

Regional Distribution of [³H]PD 128907 and [³H](+)-7-OH-DPAT Binding

Preliminary studies of [³H]PD 128907 and [³H](+)-7-OH-DPAT binding in caudate/putamen and vestibulocerebellum revealed interactions with a single binding site of similar K_D value to those observed in ventral striatum and of relative density similar to that described below (data not shown). Accordingly, single-point binding analysis was used to examine the regional distribution of the binding sites for these radioligands.

Sites labeled by [³H]PD 128907 and [³H](+)-7-OH-DPAT exhibited similar regional distributions in rat brain (Table 3). Binding of both radioligands was highest in ventral striatum (nucleus accumbens and olfactory tubercles), followed by the vestibulocerebellum. Moderate levels of [³H]PD 128907 and [³H](+)-7-OH-DPAT were observed in the caudate/putamen, hypothalamus, and substantia nigra/ventral tegmental area. Moderately low to low densities of sites were detected in frontal cortex, amygdala, hippocampus, and dorsal vermis. Overall, the regional distributions of sites labeled by these radioligands was highly correlated (r² = 0.900).

Table 3 Regional Distribution of [³H]PD 128907 and [³H](+)-7-OH-DPAT Binding in Rat Brain Membranes: Comparison with the Distribution of [³H] Quinpirole–Labeled Sites

The regional distribution of [³H]PD 128907– and [³H](+)-7-OH-DPAT–labeled sites was compared with sites labeled by the D₂/D₃ agonist ligand [³H]quinpirole. The density of [³H]quinpirole-labeled sites in brain areas such as the caudate/putamen, ventral striatum, and substantia/nigra/ventral tegmental area was significantly higher than those labeled by either [³H]PD 128907 or [³H](+)-7-OH-DPAT. The correlation coefficients for the regional distribution of [³H]quinpirole–labeled sites with those labeled by [³H]PD 128907 and [³H]7-OH-DPAT were 0.138 and 0.292, respectively.

Autoradiographic Studies

Substrate Specificity of [³H]PD 128907 Binding in Slide-Mounted Brain Sections

The relative potencies of several compounds at [³H]PD 128907 binding sites in sagittal sections of rat brain are depicted in Figure 3 These data are quantified in Figure 4 . Overall, comparable substrate specificity of [³H]PD 128907 binding was obtained in autoradiographic studies in nucleus accumbens and the molecular layer of cerebellar lobule X (Figure 4). (±)-7-OH-DPAT (10 μmol/L), spiperone (1 μmol/L), and (+)-butaclamol (1 μmol/L) each inhibited [³H]PD 128907 to identical baseline levels in both brain areas examined. The potencies of (±)-7-OH-DPAT and SCH 23390 were similar to that observed in radioligand binding studies with IC₅₀ values of roughly 0.5 nmol/L and 1 μmol/L, respectively. (−)-Butaclamol (1 μmol/L) and Gpp(NH)p (1 mmol/L) failed to inhibit [³H]PD 128907 in either nucleus accumbens or cerebellar lobule X.

Figure 3

Substrate specificity of [³H]PD 128907–labeled sites in rat brain. Slide-mounted sagittal sections (lateral 1.20–1.40 mm) were incubated with 0.7 nmol/L [³H]PD 128907 in the presence of absence of various compounds as indicated. (A) total binding, (B) (±)-7-OH-DPAT (0.5 nmol/L), (C) (±)-7-OH-DPAT (10 nmol/L), (D) spiperone (1 μmol/L), (E) (+)-butaclamol (1 μmol/L), (F) (−)-butaclamol (1 μmol/L), (G) SCH 23390 (1 μmol/L), (H) Gpp(NH)p (1 mmol/L). Acb = nucleus accumbens, lob X = cerebellar lobule X. Quantified data from these autoradiograms is shown in Figure 4.

Figure 4

Quantification of [³H]PD 128907–labeled sites in specific brain regions in the presence of various compounds. Film images in Figure 3 were digitized and quantified as described in Materials and Methods. Each data point represents the mean ± SEM obtained from four slides (two sections per slide). TOT = total binding; 7-OH = (±)-7-OH-DPAT; Spip = spiperone; (+)But = (+)-butaclamol; (−)But = (−)-buta-clamol; SCH = SCH 23390; Gpp = Gpp(NH)p.

Autoradiographic Localization of [³H]PD 128907–Labeled Sites

The distribution of [³H]PD 128907–labeled sites in rat brain is shown in Figure 5 . Quantitative analysis of regional [³H]PD 128907 binding appears in Table 4. Specific [³H]PD 128907 binding was heterogeneously distributed throughout the brain with highest densities (>10 fmol/mg tissue equiv.) observed in the islands of Calleja, followed by the nucleus accumbens. Within the nucleus accumbens, somewhat more [³H]PD 128907 binding was observed in the shell than in the core. Dense binding was also observed in the nucleus of the horizontal limb of the diagonal band. Moderate densities of [³H]PD 128907 binding (5–10 fmol/mg tissue equiv.) were observed in the molecular layer of cerebellar lobule X, olfactory tubercle, ventral pallidum, substantia nigra pars compacta, and caudate/putamen. [³H]PD 128907 binding in the caudate/putamen, however, was not uniform with greater binding observed in the anteroventral and dorsomedial portions of the nucleus with relatively little binding in the dorsal portion. Low to very low densities of [³H]PD 128907 binding (<5 fmol/mg tissue equiv.) were observed in the other brain region examined such as the hypothalamus, thalamus, amygdala, cortical areas excluding piriform, cerebellum excluding the molecular layer of lobule X, olfactory bulb, and hippocampus. No differential labeling of cell layers of the hippocampus and dentate gyrus was observed.

Figure 5

Autoradiographic localization of [³H]PD 128907–labeled sites in rat brain. Data shown represent total binding. Specific autoradiograms are representative of those obtained from each of four individual animals (two sections per slide). The darker areas of the autoradiograms indicate regions of higher receptor density. Slide-mounted brain sections (20 μm) were incubated with ∼0.7 nmol/L [³H]PD 128907 as described in Materials and Methods. Nonspecific binding, as defined by 1 μm spiperone, 1 μm (+)-butaclamol, or 10 μm (±)-7-OH-DPAT is shown in Figure 3. Quantified data from these autoradiograms are summarized in Table 4. Stereotaxic coordinates for each level: (A) Bregma 4.20 mm, (B) Bregma 2.20 mm, (C) Bregma 1.30 mm, (D) Bregma 0.48, (E) −0.30 mm, (F) Bregma −3.60 mm, (G) Bregma −4.80 mm, (H) Bregma −7.64 mm, (I) Lateral 0.40 mm, (J) Lateral 2.40 mm. Abbreviations: Acb = nucleus accumbens, av = anteroventral caudate/putamen, dm = dorsomedial caudate/putamen, ICj = islands of Calleja, HDB = nucleus of the horizontal limb of the diagonal band, VP = ventral pallidum, LPO = lateral preoptic area, LH = lateral hypothalamus, SN = substantia nigra, lob X = cerebellar lobule X.

Table 4 Autoradiographic Distribution of [³H]PD 128907–Labeled Sites Rat Brain

DISCUSSION

Because of the potential therapeutic importance of the dopamine D₃ site (Sokoloff et al. 1990; Caine and Koob 1993), considerable effort has been directed toward understanding its neurobiology. To this end, several radioligands purported to be D₃ selective have been identified including [³H]PD 128907 and [³H]7-OH-DPAT. Before extensive study of this novel receptor in brain may be undertaken, however, the ability of radioligands to distinguish between closely related dopamine receptor subtypes, as well as other receptors, must be demonstrated. The present study assesses the D₃ selectivity of [³H]PD 128907 in rat brain and uses this radioligand to characterize the pharmacological profile, guanyl nucleotide regulation, and regional distribution of putative D₃ sites.

Receptor binding analysis revealed that, under the in vitro assay conditions used in this study, [³H]PD 128907 exhibited a binding profile in rat brain consistent with interactions at a single binding site. The pharmacological profile of [³H]PD 128907–labeled sites labeled was similar to that of [³H](+)-7-OH-DPAT, suggesting that these ligands label a common D₂-like binding site. Whereas the pharmacology of these sites was similar to that of the classic D₂ receptor (for review, see Seeman and Grigoriadis 1987; Sibley and Monsma 1992), sites labeled by the putative D₃-selective ligands differ from the D₂ receptor in that [³H]PD 128907– and [³H](+)-7-OH-DPAT–labeled sites are present in lower density, display a different regional distribution, and are insensitive to guanyl nucleotides. Thus, neither ligand appears to interact significantly with the D₂ receptor under the assay conditions used in this study.

Instead, sites labeled by [³H]PD 128907 and [³H](+)-7-OH-DPAT display characteristics similar to those reported for the cloned D₃ receptor in expression systems in terms of pharmacological profile and apparent insensitivity to guanyl nucleotides (Boundy et al. 1993; Freedman et al. 1994; MacKenzie et al. 1994; Sokoloff et al. 1990, 1992). Furthermore, the densities and regional distributions of [³H]7-OH-DPAT– and [³H]PD 128907–labeled sites in rat brain are generally consistent with the distribution of D₃ receptor mRNA (Bouthenet et al. 1991) and the distribution of D₃ receptor binding estimated from experiments using nonselective radioligands (Landwehrmeyer et al. 1992; Levant et al. 1993; Parsons et al. 1993; Booze and Wallace 1995). Taken together, the present data suggest the interaction of these radioligands with a common binding site with the characteristics of the D₃ dopamine receptor.

Thus, in our hands, both [³H]PD 128907 and [³H](+)-7-OH-DPAT label the putative D₃ site in rat brain. Although both ligands exhibited high affinity, they differed in their selectivity between the D₂ and D₃ sites. Saturation analysis using high concentrations of [³H]PD 128907 (0.5–75 nmol/L) in caudate/putamen, a brain area that expresses D₂ receptor predominantly over other D₂-related subtypes (for review, see Levant 1996), indicated no binding above that attributable to the low density of putative D₃ sites present in that tissue. Thus, the affinity of this radioligand for the D₂ receptor is sufficiently low to preclude determination by filtration assay (i.e., K_D > 100 nmol/L) (Bennett and Yamamura 1985). As such, it is concluded that the D₂/D₃ receptor selectivity of this radioligand is significantly greater than 300-fold and that labeling of D₂ receptors is negligible under the in vitro assay conditions used in this study. In contrast, similar saturation studies with [³H](+)-7-OH-DPAT revealed binding to a site in caudate/putamen of appropriate density for the D₂ receptor (Levant 1995; Levant et al. 1992) with a K_D value of 23 nmol/L. Thus, [³H](+)-7-OH-DPAT exhibits only about 100-fold selectivity for the putative D₃ receptor over the D₂ under the in vitro assay conditions used in this study.

Interactions of [³H]7-OH-DPAT with the sigma site have also been reported (Wallace and Booze 1995). The low potency of the sigma₁ ligand (+)N-allylnormetazocine in the radioligand binding studies, however, suggests no such interaction for either [³H](+)-7-OH-DPAT or [³H]PD 128907 under the in vitro conditions used in our radioligand binding studies. Likewise, autoradiographic localization of [³H]PD 128907 binding sites indicated very low density of labeling in regions noted for dense expression of sigma sites such as the cerebellar cortex (Gundlach et al. 1986).

Whereas the present data indicate that selective labeling of putative D₃ sites may be reasonably obtained using either [³H]PD 128907 and [³H]7-OH-DPAT, labeling of the D₂ site by [³H]7-OH-DPAT has been reported (Gonzales and Sibley 1995). One likely factor contributing to the conflicting reports of the D₂/D₃-selectivity of [³H]7-OH-DPAT is the use of different in vitro assay conditions. Assay conditions, particularly buffer composition, have been shown to influence the D₂/D₃ selectivity of this ligand in vitro (Lévesque et al. 1992; Burris et al. 1995). Specifically, the greatest D₂/D₃ selectivity with [³H]7-OH-DPAT was obtained in the absence of Mg²⁺ and the presence of EDTA. Buffer systems excluding Mg²⁺ and containing EDTA were also selected for [³H]PD 128907 (Akunne et al. 1995). We have found that specific binding of [³H]quinpirole in striatal membranes was nearly abolished in a Tris-EDTA buffer system (unpublished observation), in concordance with previous studies indicating that the high affinity agonist state of D₂-like receptors is not favored in the absence of Mg²⁺ (Sibley and Creese 1983). Because selective visualization of the putative D₃ receptor appears to require in vitro assay conditions that disfavor D₂ binding, these conditions may also affect the binding or functional properties of the D₃ receptor. For example, the failure of guanyl nucleotides to inhibit [³H]7-OH-DPAT and [³H]PD 128907 binding could result from a lack of coupling of the receptors to G-proteins. Alternatively, although the binding of agonist ligands such as [³H]PD 128907 and [³H]7-OH-DPAT is presumed to be to receptors in the high affinity agonist state, this observation might also result if the labeled receptors were actually in the low affinity agonist conformation. Accordingly, care must be taken in the interpretation of data obtained with these methods with regard to their physiological implications.

The distribution of putative D₃ sites in brain has been of particular interest and is the primary basis for the hypothesis that receptor may be a suitable target for novel antipsychotics (Sokoloff et al. 1990). The present data confirm previous studies indicating highest densities of putative D₃ sites in the islands of Calleja followed by the nucleus accumbens as assessed by either radioligand binding or autoradiographic methods. Dense labeling by [³H]PD 128907 has also been reported in the nucleus accumbens and islands of Calleja of the human brain (Hall et al. 1996). As predicted by the distribution of D₃ receptor mRNA (Bouthenet et al. 1991; Curran and Watson 1995; Diaz et al. 1995; Landwehrmeyer et al. 1993; Mengod et al. 1992; Sokoloff et al. 1990), several other brain areas, such as the molecular layer of cerebellar lobule X, ventral pallidum, substantia nigra pars compacta, bed nucleus of the stria terminalis, medial and magnocellular preoptic areas, and nucleus of the horizontal limb of the diagonal band also exhibited moderately dense labeling by [³H]PD 128907.

Whereas the density of D₃ sites in caudate/putamen were low overall, the sites are differentially distributed within the nucleus and are moderately dense in the dorsomedial and anteroventral areas. The presence of moderately dense putative D₃ binding in the “limbic” anterovental striatum is consistent with the similar afferent input as the nucleus accumbens from the allo- meso-, and periallocortices as well as the ventral tegmental area, basolateral amygdala, hippocampus, and prefrontal cortex (Heimer et al. 1995). The presence of D₃ sites in the dorsomedial aspect of the caudate, which also receives afferent from the mesocortex (Heimer et al. 1995), was suggested by previous studies using [³H]quinpirole (Levant et al. 1993).

Whereas dense [³H]PD 128907 binding was observed in brain areas such as the islands of Calleja and nucleus accumbens where dense expression of D₃ receptor mRNA has been reported, the observed densities of putative D₃ sites and the reported densities of D₃ mRNA were not directly correlated throughout the brain. For example, very dense expression of D₃ mRNA expression was reported in areas such as cerebellar lobule X and the bed nucleus of the stria terminalis (Bouthenet et al. 1991) but only moderately dense [³H]PD 128907 binding was observed in these regions. Other notable discrepancies in the distributions of D₃ mRNA and binding include the medial geniculate nucleus, pyramidal cell layer of the hippocampus, granular layer of the dentate gyrus, and certain thalamic and amygdaloid nuclei where moderate to dense expression of D₃ mRNA were observed (Bouthenet et al. 1991), but relatively little putative D₃ binding. These discrepancies may be due to the transport of receptors to brain regions remote from cell bodies or the differential regulation of transcription and translation in different brain areas. This issue must be resolved by further study.

In summary, the present data demonstrate the binding of [³H]PD 128907 to dopaminergic binding sites in rat brain that are generally consistent with previous studies of the D₃ receptor with respect to density, pharmacological profile, guanyl nucleotide regulation. The regional distribution of [³H]PD 128907–labeled sites is similar those labeled by other putatively selective D₃ ligands with the highest density in limbic brain areas and is generally similar to the distribution of D₃ mRNA.