Characterization of [³H]LS-3-134, a Novel Arylamide Phenylpiperazine D3 Dopamine Receptor Selective Radioligand

Abstract

LS-3-134 is a substituted N-phenylpiperazine derivative that has been reported to exhibit a) high-affinity binding (Ki value 0.2 nM) at human D3 dopamine receptors, b) >100-fold D3 vs. D2 dopamine receptor subtype binding selectivity and c) low-affinity binding (Ki values >5,000 nM) at sigma 1 and sigma 2 receptors. Based upon a forskolin-dependent activation of the adenylyl cyclase inhibition assay, LS-3-134 is a weak partial agonist at both D2 and D3 dopamine receptor subtypes (29% and 35% of full agonist activity, respectively). In this study, [³H]-labeled LS-3-134 was prepared and evaluated to further characterize its use as a D3 dopamine receptor selective radioligand. Kinetic and equilibrium radioligand binding studies were performed. This radioligand rapidly reaches equilibrium (10-15 min at 37°C) and binds with high affinity to both human (Kd = 0.06 ± 0.01 nM) and rat (Kd = 0.2 ± 0.02 nM) D3 receptors expressed in HEK-293 cells. Direct and competitive radioligand binding studies using rat caudate and nucleus accumbens tissue indicate that [³H]LS-3-134 selectively binds a homogeneous population of binding sites with a dopamine D3 receptor pharmacological profile. Based upon these studies we propose that [³H]LS-3-134 represents a novel D3 dopamine receptor selective radioligand that can be used for studying the expression and regulation of the D3 dopamine receptor subtype.

Introduction

There are three dopaminergic pathways in the brain: the nigrostriatal pathway, the mesocorticolimbic pathway and the tuberoinfundibular pathway (Fuxe et al., 1977). These pathways are involved in movement coordination, cognition, emotion, affect, memory, reward mechanisms and the regulation of prolactin secretion by the pituitary (Beaulieu and Gainetdinov, 2011; Nakajima et al., 2013). Multiple neurological and neuropsychiatric disorders, including Parkinson’s Disease, Tourette’s Syndrome, tardive dyskinesia, schizophrenia and schizoaffective disorders, appear to be related to disturbances of the dopaminergic systems (Nord and Farde, 2011; Denys et al., 2013; Bezard et al., 2003; Gottwald and Aminoff, 2008).

The dopamine receptor subtypes are members of the G protein coupled receptor (GPCR) protein superfamily. Based upon similarities in primary structure and pharmacologic properties, there are two types of dopamine receptors, D1-like and D2-like dopamine receptors. Agonist binding to the D1-like receptor subtypes, which includes the D1 (D1a) and the D5 (D1b) dopamine receptors, leads to a) Gα_olf/s protein coupling and the stimulation of adenylyl cyclase (Wollemann, 1980; Monsma et al., 1990; Corvol et al., 2007) and b) the binding of arrestin (Kim et al., 2004; Urs et al., 2011).

Agonist binding at D2-like (D2, D3 and D4) receptors leads to the activation of several signaling pathways including a) inhibition of adenylyl cyclase activity (Missale et al., 1998), b) activation of G protein-coupled inwardly rectifying potassium channels (GIRKs) (Kuzhikandathil et al., 1998), c) mobilization of intracellular Ca²⁺ stores (Sternweis and Smrcka, 1992), d) activation of a Ca²⁺-dependent protein phosphatase (calcineurin), e) Ca²⁺-dependent mitogen-activated protein (MAP) kinase (Nishi et al., 1997; Yan et al., 1999) and f) Akt/glycogen synthase kinase-3β (GSK-3β) signaling through the scaffolding protein β-arrestin 2 (Park et al., 2011).

D2 and D3 dopamine receptors are structurally and pharmacologically similar (Sokoloff et al., 1990; Boundy et al., 1993). Both receptor subtypes are found in the central nervous system, with highest densities in the striatum, nucleus accumbens, olfactory tubercles, islands of Calleja and basal ganglia (Levant, 1997; Missale et al., 1998).

The D3 receptor was found to share 50% amino acid sequence homology with the D2 receptor subtype (Sokoloff et al., 1990). The amino acid sequence homology for the helical transmembrane spanning (TMS) segments of the D2 and D3 dopamine receptor subtypes was found to be 75-80%. Since the TMS regions are involved in the construction of the orthosteric binding site, it was not surprising that the pharmacologic profiles of D2 and D3 receptors were initially found to be similar (Sokoloff et al., 1990; Boundy et al., 1993; Chio et al., 1994; Chien et al., 2010). There is a high degree of amino acid sequence homology between rat, mouse and human D3 (hD3) dopamine receptors (Giros et al., 1990; Park et al., 1995).

The hD3 dopamine receptor is the only member of the dopamine receptor subtypes for which high resolution x-ray diffraction analysis is available (Chien et al., 2010). A series of computer-based molecular modeling and ligand docking studies have provided insight into how the arylamide substituted phenylpiperazines bind to the D2 and D3 receptors and the molecular basis for how D3 vs. D2 binding selectivity is achieved (Wang et al., 2010; Chien et al., 2010; Newman et al., 2012).

Because of the high level of amino acid homology within the TMS segments between the D2 and D3 dopamine receptor subtypes, it has been a difficult task to identify D3 receptor subtype selective compounds that could be used as imaging agents to study the regulation of D3 receptor expression in neurological and neuropsychiatric disorders. The development of high-affinity D3 versus D2 and D2 versus D3 receptor selective compounds could provide both pharmacotherapeutic and imaging tools to precisely define the role of the D2-like receptors in a) neuropsychiatric disorders such as schizophrenia and Tourette’s syndrome (Emilien et al., 1999; Caccia et al., 2013), b) motor disorders such as Parkinson’s Disease and L-dopa-induced dyskinesia (Bezard et al., 2003; Sanchez-Pernaute et al., 2007; Kumar et al., 2009) and c) the abuse of cocaine and methamphetamine (Caine and Koob, 1993; Cheung et al., 2012; Cheung et al., 2013).

In 2011, Tu and colleagues reported on the synthesis and pharmacological characterization of a series of arylamide phenylpiperazines that contain a fluoro group, including N-[(4-[4-[2-(2-fluoroethoxy)phenyl]-1-piperazinyl]butyl]-4-(3-thienyl)benzamide (LS-3-134) (Tu et al., 2011a). LS-3-134 has a) high-affinity binding (Ki value 0.2 nM) at hD3 dopamine receptors, b) >100-fold D3 vs. D2 dopamine receptor subtype binding selectivity and c) low-affinity binding (Ki values >5,000 nM) at sigma 1 and sigma 2 receptors (Tu et al., 2011a). Forskolin-dependent adenylyl cyclase activation inhibition studies indicated that LS-3-134 is a weak partial agonist at both D2 and D3 dopamine receptor subtypes (29% and 35% of full agonist activity, respectively) (Tu et al., 2011a).

Previously published positron emission tomography (PET) imaging studies using [¹⁸F]-labeled LS-3-134 indicated that it is capable of imaging D3 dopamine receptors in nonhuman primate brain, in the absence of D2 receptor subtype labeling (Mach et al., 2011). This finding is in contrast to previously reported results obtained using a variety of other D3 vs. D2 receptor subtype selective ligands including the “D3-preferring” radiotracer [¹¹C](+)-PHNO, which has been shown to label both D2 and D3 receptors in PET imaging studies in humans (Tziortzi et al., 2011; Payer et al., 2014a,b). Our studies indicate that [¹⁸F]LS-3-134 has a potential advantage over [¹¹C](+)-PHNO in that it is capable of selectively imaging D3 receptors, without the confound of D2 receptor subtype labeling. Prior to initiating the costly and time-consuming steps required to achieve FDA approval for translational imaging studies in humans, a more detailed analysis of the in vitro binding properties of LS-3-134 is warranted. Therefore, the molecular pharmacological studies detailed in this report were performed to further characterize the D3 vs. D2 dopamine receptor binding selectivity of LS-3-134 by preparing a [³H]-radiolabeled analog and using classical equilibrium and kinetic radioligand binding techniques to monitor the interaction of this radioligand with D2 and D3 dopamine receptors expressed in transfected cell lines and rat brain tissue.

Materials and methods

Test compounds

8-triazaspiro(4.5)decan-4-one,8-(3-(p-fluorobenzoyl)propyl)-1-phenyl-3 hydrochloride (spiperone), 1H-Benzo(6,7)cyclohepta(1,2,3-de)pyrido(2,1-a)isoquinolin-3-ol, 2,3,4,4a,8,9,13b,14-octahydro-3-(1,1-dimethylethyl)-(3S-(3-alpha,4a-alpha,13b-beta)) ((+)-butaclamol) hydrochloride and S-(−)-3-chloro-5-ethyl-N-[(1-ethyl-2-pyrrolidinyl)methyl]-6-hydroxy-2-methoxybenzamide hydrochloride (eticlopride), were purchased from Sigma-Aldrich. The synthesis and purification of the SV series of compounds, including 1-((5-methoxy-1H-indol-3-yl)methyl)-4-(4-(methylthio)phenyl)piperidin-4-ol (SV 293), have been previously reported (Vangveravong et al., 2006; Vangveravong et al., 2011). The synthesis and purification of the 4-(dimethylamino)-N-(4-(4-(2-methoxyphenyl)piperazin-1-yl)butyl)benzamide (WC-10) and N-[(4-[4-[2-(2-fluoroethoxy)phenyl]-1-piperazinyl]butyl]-4-(3-thienyl)benzamide (LS-3-134) have been previously reported (Tu et al., 2011a).

LS-3-134 was radiolabeled with tritium ([³H]LS-3-134) by American Radiolabeled Chemicals, Inc. (St. Louis, MO) to a specific activity of 60 Ci/mmol. [¹²⁵I]IABN was prepared and purified as previously described (Luedtke et al., 2000). [³H]raclopride was obtained commercially (NEN-PerkinElmer).

Membrane preparations

Stably transfected HEK-293 cells expressing human D3 dopamine receptors (hD3 HEK-293), rat D3 dopamine receptors (rD3 HEK-293), human D2 dopamine receptors (hD2 HEK-293) were prepared using either a pIRES-neo expression vector (human D2 and D3 DNA insert) or a pcDNA1 (rat D3 DNA insert). Transfected cells were maintained in tissue culture using DMEM-high glucose media with 10% fetal calf serum (Invitrogen) (with penicillin (100 units/ml) and streptomyocin (100 ug/ml) media containing the 400 ug/ml G418 for HEK cells expressing human D2 or D3 dopamine receptors or 2 ug/ml puromycin for cells expressing rat D3 dopamine receptors.

For the tissue preparation of transfected HEK cells were grown to approximately 70-80% confluence, harvested and re-suspended in cold (4°C) 50 mM Tris-HCl/150 mM NaCl/10 mM EDTA buffer, pH 7.5 (homogenization buffer). After centrifugation (4°C) the cell pellet was homogenized using a Polytron homogenizer (Brinkmann Instruments, Westbury, NY (setting 6)). The cell homogenate was centrifuged at 12,000 × g at 4°C for 15 minutes. The pellet was resuspended in cold homogenization buffer, kept on ice, aliquoted (1.0 ml) into 1.5 ml plastic microfuge vials (1.0 ml/T75 flask (BD Falcon) and kept at −80°C.

For equilibrium binding studies (saturation direct binding studies or competitive binding studies) HEK cell homogenates were suspended in homogenization buffer and incubated with radioligand, in the presence or absence of inhibitor at 37°C for either a) 30 minutes when using [³H]LS-3-134 and [³H]raclopride (300 ul total volume) or b) 60 minutes when using [¹²⁵I]IABN (total volume = 150 ul).

Frozen intact brains from Sprague Dawley rats (4 month old males) were purchased from BioChemed (Winchester, VA). The striatum and nucleus accumbens were dissected on ice and kept in cold (0°C) buffer until homogenization using a Polytron homogenizer. The pellet was washed with homogenization buffer four times and collected by centrifugation at 12,000 × g for 15 minutes at 4°C. The membrane pellet was suspended in homogenization buffer, aliquoted (1.0 ml) into 1.5 ml microfuge tubes (striatum or nucleus accumbens from 1 rat brain/tube) and kept at −80°C. Just prior to the binding experiments endogenous dopamine the frozen homogenates were resuspended with an equal volume of homogenization buffer and incubated at 37°C for 30 minutes to further promote the dissociation of any bound endogenous dopamine (Leff and Creese, 1985). Membrane homogenates were then put on ice and centrifuged at 12,000 × g for 20 minutes (4°C). The pellet was resuspended in cold (0°C) homogenization buffer.

Radioligand Binding Studies

Kinetic analysis

The association rates (Kon) for the binding of [³H]LS-3-134 to rat or human D3 dopamine receptors expressed in HEK-293 cells were determined by measuring the amount of ligand bound specifically as a function of time (0, 1, 2.5, 5, 7.5, 10, 15, 20, 30, 40 50 and 60 min time points) at a constant concentration of radioligand (approximately equal to the Kd value) and receptor. The bound specific (Bs) value was defined as the difference between bound total (Bt) and bound nonspecific (Bns), where the bound nonspecific was obtained by the addition of 5 μM final concentration of (+)-butaclamol in the assay. For each kinetic analysis at least three independent experiments were performed. Each time point was assayed by removing duplicate aliquots of 200 μL from the reaction mixtures (Bt or Bns) and filtering the membranes using glass fiber filters.

The association kinetic data were fit to the follow equation:

$$\mathbf{Y}={\mathbf{Ymax}}^{\ast }(1-\mathbf{exp}(-1^{\ast }{Kob}^{\ast }\mathbf{t}))$$

where Y is ligand bound, Ymax is equal to the maximum ligand bound under the experimental conditions, t is time (minutes) and Kob is the observed association rate. For these studies it is assumed that the Kon, the association rate constant (M⁻¹ min⁻¹), is:

$$Kon=(Kob-Koff)∕[\mathbf{L}]$$

where L is defined as the radioligand concentration and Koff is obtained from dissociation studies.

The rate of dissociation (Koff) of [³H]LS-3-134 at D3 dopamine receptors expressed in HEK-293 cells was determined by allowing the receptor-radioligand mixture to reach equilibrium (30 min at 37°C) and then adding (at time = 0) a high concentration of non-radiolabeled competitive inhibitor (5 μM (+)-butaclamol final concentration). For these studies the final radioligand concentration was approximately equal to the Kd value. Control samples containing no competitive inhibitor were included at each time point to correct for the stability of the receptor preparation as a function of time. Specific binding was determined as a function of time (0, 2.5, 5, 7.5, 10, 15, 20, 30, 45, 60, 80, 100, 120, 150, 180, 210 and 240 minutes) over a 2 hour time period. The value for Koff was determined by the slope of the simple first order equation:

$$\ln [\mathbf{B}]∕[\mathbf{Bo}]=-Kofft$$

where Bo is the concentration of receptor-ligand complex at t = 0 (McGonigle and Molinoff, 1989).

Saturation isotherm

The binding properties of [³H]LS-3-134 were characterized using a filtration-binding assay using a) human and rat D3 stably expressed dopamine receptors in HEK-293 cells and b) striatum and nucleus accumbens rat tissue. Membrane homogenates were suspended in 50 mM Tris–HCl/150 mM NaCl/10 mM EDTA buffer, pH=7.5 and incubated with increasing concentrations of [³H]LS-3-134, where the radioligand concentration varied from approximately 1–4 times the Kd value. Borosilicate 16 × 100 mm glass tubes were used (Fisherbrand disposable Culture Tubes). [³H]LS-3-134 binding studies were performed at 37°C for 30 min, using 2.5 μM (+)-butaclamol to define the non-specific binding. Binding was terminated by the addition of cold wash buffer (10mM Tris–HCl/150mM NaCl, pH = 7.5) and filtration over a glass-fiber filter (Pall A/B filters, #66198). After the addition of Cytoscint scintillation cocktail (Acros Organics), radioactivity was quantitated using a 1414 Wallac Guardian (Perkin Elmer) scintillation counter. The protein concentration of the membranes was determined using a BCA reagent (Pierce) and bovine serum albumin was used as the protein standard.

Estimates of the equilibrium dissociation constant (Kd) and maximum number of occupied binding sites (Bmax) were determined by one-site fit nonlinear regression analysis, where the Bmax value is reported as fmol/mg protein. The equilibrium binding constant was fit to:

$$\mathbf{B}={\mathbf{Bmax}}^{\ast }[\mathbf{L}]∕(\mathbf{Kd}+[\mathbf{L}])$$

where B is ligand bound, Bmax is the maximum ligand bound (obtained by extrapolation to high concentrations of radioligand), Kd is the equilibrium dissociation binding constant, which is defined as the radioligand concentration needed to achieve a half-maximum binding at equilibrium. Mean Kd values ± S.E.M. are reported for at least three independent experiments. The data were linearly transformed as described by Scatchard (1949).

D2 and D3 dopamine receptor combination experiments

The binding studies using hD2 and hD3 dopamine receptors expressed in stably transfected HEK-293 cells were performed using [³H]LS-3-134, [³H]raclopride or [¹²⁵I]IABN at a concentration approximately equal to the Kd value for binding at human D3 receptor. A constant concentration of radioligand and a constant amount of hD3 HEK-293 cell homogenate was incubated with varying amounts of hD2 cell homogenate, for a final volume of 300 μL. To determine the level of nonspecific binding, binding studies were performed in presence of 2.5 μM (+)-butaclamol. After incubation at 37°C for 30 minutes the binding was terminated by the addition of cold wash buffer (10mM Tris–HCl/150mM NaCl, pH = 7.5) and filtration over a glass-fiber filter (Pall A/B filters, #66198). A 1414 Wallac Guardian (Perkin Elmer) scintillation counter was used to measure the radioactivity of [³H]LS-3-134 and [³H]raclopride. A 2470 Wallac Wizard² (Perkin Elmer) gamma counter was used to measure the radioactivity of [¹²⁵I]IABN. The protein concentration of the membranes was determined by using a BCA reagent (Pierce).

Competitive radioligand binding studies

Competitive radioligand studies were performed to determine the concentration of inhibitor that inhibits 50% of the specific binding of the radioligand (IC50 value) using a) human and rat D3 dopamine receptors transfected in HEK-293 cells and b) rat striatal tissue. The final radioligand concentration was approximately equal to the Kd value (Table II). For each competition curve, triplicates were performed using two concentrations of inhibitor per decade over 5 orders of magnitude. Binding was terminated by the addition of cold wash buffer (10mM Tris–HCl/150mM NaCl, pH = 7.5) and filtration over a glass-fiber filter (Pall A/B filters, #66198). A 1414 Wallac Guardian (Perkin Elmer) scintillation counter was used to measure the radioactivity of [³H]LS-3-134.

Table II.

Summary of the equilibrium dissociation constants for [³H]LS-3-134.

Tissue	Kd (nM)	Bmax (fmol/mg protein)	n
human D3 HEK 293	0.056 ± 0.012**	2780 ± 634	5
rat D3 HEK 293	0.223 ± 0.016	2441 ± 401	4
rat striatum	0.300 ± 0.071	54.4 ± 13	5
rat N. accumbens	0.522 ± 0.16	99.8 ± 16	4

Kd values were determined by one-site fit nonlinear regression analysis of binding experiments and are shown as the mean ± S.E.M. of n independent determinations. Statistical difference was found when hD3 HEK-293 was compared to rD3 HEK-293 cells

^**p<0.001, one-way ANOVA)

The competition curves were modeled for a single site using the following equation:

$$\mathbf{Bs}=\mathbf{Bo}-(\mathbf{Bo}+\mathbf{L})∕(\mathrm{IC}50+\mathbf{L})$$

where Bs is the amount of ligand bound to receptor and Bo is the amount of ligand bound to receptor in the absence of competitive inhibitor. L is the concentration of the competitive inhibitor. The IC50 value is the concentration of competitive inhibitor that inhibits 50% of the total specific binding. IC50 values were determined by using nonlinear regression analysis with Table Curve 2D v5.01 (Jandel, SYSTAT Software). The values for Bns and Bo were constrained using experimentally derived values. The IC50 values were converted to equilibrium dissociation constants (Ki values) using the Cheng and Prusoff (1973) correction. Mean Ki values ± S.E.M. are reported for at least three independent experiments.

Protein concentrations

The amount of protein used in the equilibrium and kinetic binding assays varied depending on which tissue was used because of differences in receptor affinity and expression density. For direct and competitive binding studies a final assay volume of 300 μl was used with the following amounts of tissue protein: a) human D3 receptor HEK cells, 3.5-6.5 μg total protein (0.012 to 0.22 μg protein/ul), b) rat D3 HEK cells, 15 to 20 μg total protein (0.05 to 0.067 μg protein/μl), c) rat striatum, 230-320 μg total protein (0.75 to 1.10 μg protein/ul) and d) rat nucleus accumbens, 190 to 220 ug total protein (0.63 to 0.73 μg protein/μl). For kinetic studies, similar protein concentrations (μg/μl) were used.

Statistical Analysis

Statistical analysis involved one-way ANOVA or unpaired t-test, when appropriate. The statistical significance was set up at p < 0.05. ANOVA analysis was followed by Tukey’s multiple comparison tests as appropriate. The statistical and data analysis for Scatchard or kinetic studies were performed using Prism 6.0a (GraphPad Software, San Diego, CA) and the competitive binding results were analyzed using Table Curve 2D v5.01 (Jandel, SYSTAT Software).

Results

Pharmacological properties of LS-3-134

LS-3-134 is a thiophen-containing arylamide with a saturated four-carbon chain adjacent to an ortho substituted fluoroethoxy phenylpiperazine (Figure 1). In previous studies we reported that LS-3-134 exhibits a) high-affinity binding (Ki value 0.2 nM) at hD3 dopamine receptors, b) >150-fold D3 vs. D2 dopamine receptor subtype binding selectivity, c) >1446-fold selectivity for D3 vs. D4 (Table I) and d) low-affinity binding (Ki values >5,000 nM) at sigma 1 and sigma 2 receptors (data not shown). LS-3-134 is a weak partial agonist at both D2 and D3 dopamine receptor subtypes (29% and 35% of full agonist activity, respectively) determined by forskolin-dependent adenylyl cyclase stimulation studies (Tu et al., 2011a).

Figure 1. Synthesis of [³H]LS-3-134.

The synthesis of [³H]LS-3-134 was accomplished via catalytic hydrogenation of the corresponding double bond precursor, LS-3-48, to give the tritiated ligand in specific activity of ~60 Ci/mmol (custom synthesis by American Radiolabeled Chemicals, St. Louis, MO).

Table I.

Summary of the Ki values.

	Ki values (nM)			Binding Selectivity
Compound	hD2	hD3	hD4	D2:D3	D4:D3
LS-3-134	27.7 ± 5.4	0.17 ± 0.01	246 ± 13.3	163	1446
Eticlopride	0.18 ± 0.01	0.15 ± 0.01	107 ± 8.6	1.2	713
WC 10	34.4 □ 3.3	0.80 □ 0.10	896 □ 193	43	1120
				D3:D2	D4:D2
Spiperone	0.06 ± 0.006	0.33 ± 0.02	0.45 ± 0.01	5.5	7.5
SV 293	5.50 ± 0.06	580 ± 64.7	567 ± 98.7	105	103

Data was obtained using human D2-like receptors stably expressed in HEK cells using [¹²⁵I]IABN at the radioligand. Data are presented as the mean ± the S.E.M. for n = 3 (Tu et al., 2011a). The [¹²⁵I]IABN Kd values used to calculate Ki values for human D2, D3 and D4 receptors were 0.03 nM, 0.04 and 0.6 nM, respectively.

Kinetic binding studies of [³H]LS-3-134 using D3 dopamine receptors expressed in HEK-293 cells

In preliminary studies we found that there was a small but consistent difference in the binding affinity for [³H]LS-3-134 for hD3 receptors expressed in HEK cells and binding sites in rat brain tissue (Table II). To better define the extent and basis of this difference, we performed both equilibrium and kinetic binding analysis in rat and human D3 receptors stably expressed in HEK cells.

Association studies were performed to examine the rate of [³H]LS-3-134 binding to rD3 or hD3 receptors expressed in HEK-293 cells. The time to equilibrium for the binding of [³H]LS-3-134 to human and rat D3 dopamine receptors expressed in HEK-293 cells was reached within approximately 10 to 15 minutes and the binding was stable over a 60 minute time period (Figure 2A). The calculated association constants (Kon) differed by approximately 5-fold for hD3 and rD3 receptors (1.05 ± 0.12 × 10⁹ M⁻¹min⁻¹ vs. 0.19 ± 0.04 × 10⁹ M⁻¹ min⁻¹, respectively, **p<0.005, n=3).

Figure 2. Kinetic studies: Association and dissociation of [³H]LS-3-134 in human and rat D3 HEK-293 cells.

Kinetic analysis for the binding of [³H]LS-3-134 using either rat (■) or human (●) D3 dopamine receptors expressed in HEK-293 cells is shown. Representative kinetic studies are shown for (A) a binding association (cpm) of specific bound radioligand per 200 μl of reaction mixture as a function of time, with the line representing a software-assisted fit of the data to a one-site fit and (B) a dissociation experiment as the percent decrease of specific bound radioligand as a function of time after the addition of (+)-butaclamol, where the line represents a fit of the data to a dissociation one–phase decay. For both of these studies the radioligand concentration was approximately equal to the Kd value (Table II) with a) human D3 receptor HEK cells used at 0.012 to 0.22 ug protein/ul and b) rat D3 HEK cells used at 0.05 to 0.067 ug protein/ul.

Further studies indicated that the dissociation half-life (t1/2) of the ligand binding was approximately 2-fold different for human compared to rat D3 dopamine receptors expressed in stably transfected HEK-293 cells (33.0 ± 5.3 min^-1 vs. 17.1 ± 3.2 min⁻¹, respectively *p < 0.05, n=3) (Figure 2B). The dissociation constants (Koff) were statistically different for human (0.021 ± 0.004 min⁻¹) compared to rat D3 dopamine receptor expressed in transfected HEK-293 transfected cells (0.039 ± 0.006 min⁻¹, *p<0.05).

The Kd values calculated from the kinetic studies (Koff /Kon) were 0.020 ± 0.01 nM for hD3 dopamine receptors and 0.20 ± 0.04 nM for rD3 dopamine receptors expressed in HEK-293 cells, indicating a 10-fold difference in binding affinity for the hD3 and rD3 dopamine receptors. These Kd values are in reasonable agreement with the value used for the calculation of the Ki value (0.17 ± 0.01 nM) originally determined from competitive radioligand binding studies using nonradioactive LS-3-134 as the competitive inhibitor of the binding of [¹²⁵I]IABN to hD3 dopamine receptors expressed in HEK-293 cells (Tu et al., 2011a) (Table I).

Equilibrium binding studies of [³H]LS-3-134 using D3 dopamine receptors expressed in HEK-293 cells

The kinetic studies provided a time to equilibrium and suggested that the affinity of [³H]LS-3-134 was slightly higher for hD3 receptors than for rD3 receptors. Subsequently, equilibrium saturation isotherm studies were initiated to determine the Kd values for the binding of [³H]LS-3-134 at hD3 and rD3 dopamine receptors stably expressed in HEK cells (Table II, Figure 3), using an incubation of 30 min at 37°C to ensure that equilibrium was achieved. Equilibrium binding studies indicated that the affinity of the radioligand was slightly higher for hD3 dopamine receptors (0.06 ± 0.01 nM) compared to the rat receptor ortholog (0.22 ± 0.02 nM). As would be anticipated for transfected cell lines, Scatchard transformation of the binding date resulted in linear coefficients >0.95, indicating that [³H]LS-3-134 was binding to a homogeneous population of binding sites (Figure 3).

Figure 3. Direct binding of [³H]LS-3-134 in rat and human D3 dopamine receptors expressed in HEK-293 cells.

Representative saturation isotherm analyses of the binding of [³H]LS-3-134 to either rat (■) or human (●) D3 receptors in transfected HEK-293 cells are shown. (A) Direct binding data is expressed as the Bound Specific (fmoles/mg of protein) as a function of radioligand concentration. (B) Linear transformation (Scatchard plot) of the direct binding of [³H]LS-3-134 of the data shown in A. For this experiment, Kd = 0.055 nM and Bmax = 1983 fmol/mg protein for human D3 HEK-293 cells (r² = 0.97) and Kd = 0.356 nM and Bmax = 1730 fmol/mg protein for rat D3 HEK-293 cells (r²=0.98).

Therefore, the results of both our kinetic and equilibrium binding studies indicate a modestly higher affinity for the binding of [³H]LS-3-134 to human, compared to the rat, D3 dopamine receptor (Table II).

Effect of NaCl on the binding of [³H]LS-3-134 to D3 dopamine receptors

Previous studies have shown that for some D2-like dopamine receptor selective radioligands, including [¹²⁵I]epidepride and [¹²⁵I]NCQ 298, the binding affinity at D2-like dopamine receptors is influenced by the presence or absence of NaCl in the assay (Neve, 1991; Hall et al., 1991). Therefore, the effect of NaCl in the binding buffer was investigated. It was found that the Kd and Bmax values were essentially the same using 0 or 150 mM NaCl (Figure 4, Table III).

Figure 4. Binding of [³H]LS-3-134 to human D3 dopamine receptors expressed in HEK-293 cells in the presence or absence of NaCl.

Direct binding analysis of [³H]LS-3-134 using either 0 mM NaCl (■) or 150 mM NaCl (●) in the assay buffer (performed in parallel at 37°C for 30 min). (A) Nonlinear representation of the data expressed as the Bound Specific (fmoles/mg of protein) as a function of radioligand concentration is shown. (B) Linear transformation (Scatchard plot) of the binding of [³H]LS-3-134 of the data is shown. The values calculated for this representative experiment were Kd = 0.039nM and Bmax = 2696 for NaCl free with a r² = 0.97 and Kd = 0.041nM and Bmax = 2976 for NaCl at 150 mM with a r² = 0.95.

Table III.

Summary of the equilibrium dissociation constants of [³H]LS-3-134 for human D3 dopamine receptors expressed in HEK-293 cells in the presence or absence of NaCl.

[NaCl]	Kd (nM)	Bmax (fmol/mg protein)	n
0 mMolar	0.041 ± 0.006	2314 ± 581	3
150 mMolar	0.056 ± 0.012	2780 ± 634	5

Kd and Bmax values were determined by one-site fit nonlinear regression analysis of radioligand binding experiments. Data is presented as the mean ± S.E.M. of independent experiments (n).

[³H]LS-3-134 binding sites in rat brain tissue

Equilibrium binding studies using [³H]LS-3-134 and rat striatum and nucleus accumbens were then performed using the buffer and experimental conditions (37°C for 30 min) used for transfected HEK-293 cells. The equilibrium dissociation constants were comparable to the values obtained for rD3 receptors expressed in transfected HEK-293 cells (Table II). A one-way ANOVA indicates that the Kd values for hD3 HEK-293 cells are statistically different from a) rD3 receptors expressed in HEK-293 cells (p<0.001) and b) binding sites in rat striatum and nucleus accumbens rat tissue (p<0.05). However, there was no statistical difference between the Kd values obtained for rD3 dopamine receptors expressed in HEK-293 cells and [³H]LS-3-134 binding sites observed in rat brain tissue. Representative linearly transformed plots of the binding of [³H]LS-3-134 to rD3 dopamine receptors expressed in HEK-293 cells, rat striatum and rat nucleus accumbens are shown in (Figure 5). These Scatchard plots exhibited linear correlation coefficients (r² values) >0.95, suggesting that in rat CNS tissue [³H]LS-3-134 was binding to a homogeneous population of binding sites.

Figure 5. Direct binding of [³H]LS-3-134 to rat D3 dopamine receptors expressed in HEK-293 cells, striatum and nucleus accumbens.

Scatchard transformation of direct [³H]LS-3-134 binding studies for (A) HEK-293 transfected cells expressing rat D3 dopamine receptors (Bmax = 4039 fmoles/mg protein), (B) rat striatum (Bmax = 70.68 fmoles/mg protein) and (C) rat nucleus accumbens rat tissue (Bmax = 95.90 fmoles/mg protein). The Scatchard plots are from a single representative experiment of each tissue. Mean Kd and Bmax values ± S.E.M. and the number of independent experiments (n) is presented in Table II.

Competitive radioligand binding studies using [³H]LS-3-134, human and rat D3 dopamine receptors expressed in HEK-293 cells and rat striatum tissue

Competitive radioligand binding studies were performed using human or rat D3 HEK-293 cells and rat striatal tissue membranes using a radioligand concentration approximately equal to the Kd value. The range of the concentration inhibitor varied depending upon the affinity of the inhibitor. The Ki values calculated for these competitive radioligand binding studies are presented in Table IV.

Table IV.

Summary of the Ki values obtained for [³H]LS-3-134 competitive binding studies using rat and human D3 dopamine receptors expressed in HEK-293 cells and in rat striatal tissue.

	Ki values (nM)
	[125I]IABN hD3-HEK-293	[3H]LS-3-134
Compound		hD3-HEK 293	rD3-HEK 293	rat striatum
eticlopride	0.15 ± 0.01	0.46 ± 0.09	0.12 ± 0.05	0.37 ± 0.09
spiperone	0.33 ± 0.02	0.25 ± 0.05	0.54 ± 0.07	2.63 ± 0.19
WC 10	0.80 ± 0.10	1.25 ± 0.36	5.71 ± 0.81	2.46 ± 0.19
(+)-butaclamol	4.31 ± 0.80	1.07 ± 0.60	0.55 ± 0.12	0.12 ± 0.05
SV 293	580 ± 64.7	324 ± 39.2	892 ± 44.2	323 ± 49.1
(−)-butaclamol	>10,000	>10,000	>10,000	>10,000

Data are presented as the mean ± S.E.M. The mean is of n ≥ 3 for all compounds. The Ki values were calculated from IC₅₀ values using the Cheng & Prusoff correction (1973) using the Kd values for [³H]LS-3-134 presented in Table II.

Initially, we studied the displacement of [³H]LS-3-134 by stereoisomers (+)-butaclamol and (−)-butaclamol (Figure 6). The Ki value for the displacement of [³H]LS-3-134 by (+)-butaclamol from human D3 receptors expressed in HEK cells was similar to values obtained using [¹²⁵I]IABN and (−)-butaclamol did not exhibit an appreciable displacement of [³H]LS-3-134 or [¹²⁵I]IABN (Table IV) in any of the tissues assessed. This stereoselective inhibition is consistent with radioligand binding at dopamine receptor subtypes. In addition, (+)-butaclamol has often been used the define the nonspecific (non-receptor) binding for D2-like dopamine receptors because it generally exhibits a plateau of inhibition (e.g., lack of displacement of nonspecific binding) at higher concentrations. A plateau was observed for both rD3-HEK cells and rat striatal tissue at concentration of 10⁻⁷ to 10⁻⁸ molar (Figure 6). Furthermore, the competition data presented in Figure 6 is consistent with the radioligand and (+)-butaclamol binding to a homogeneous population of binding sites in both rD3 HEK cells (Figure 6A) and rat striatal tissue (Figure 6B) because a nonlinear correlation coefficient (r²) of >0.95 was found for the composite inhibition curves. Based upon those observations, (+)-butaclamol was used to define non-specific binding in the subsequent competitive binding assays.

Figure 6. Butaclamol competition curves using rat striatal tissue and rat D3 dopamine receptors expressed in transfected HEK-293 cells.

Composite butaclamol competitive radioligand binding curves of [³H]LS-3-134 using the stereoisomers (−)-butaclamol (n > 2) or (+)-butaclamol (n = 3) are shown using (A) rD3 HEK-293 transfected cells and (B) rat striatal tissue. For these experiments the inhibitor concentration ranged from 10^-12 to 10^-7 Molar. A one-site fit analysis is shown for all curves. A nonlinear correlation coefficient (r²) equal to 0.98 was obtained for the inhibition of the binding of the radioligand to rat D3 receptors expressed in HEK cells. When rat striatal tissues was used, r² = 0.97.

Two high affinity compounds at D2 and D3 dopamine receptor antagonists, the benzamide eticlopride and the butyrophenone spiperone, were evaluate using rat D3 receptors expressed in HEK cells (Table IV) and striatal tissue (Table IV and Figure 7). Data presented in Table I indicates that eticlopride binds essentially nonselectively to D2 and D3 receptor and spiperone is 5- to 6-fold selective for D2 receptors compared to D3 receptors. Data shown in Table IV indicates that the Ki values determined by the displacement of [¹²⁵I]IABN and [³H]LS-3-134 by eticlopride using hD3 HEK-293 cells were within 3-fold. Similarly, the Ki value determined by the displacement of [³H]LS-3-134 by eticlopride using rD3 HEK-293 cells and rat striatal tissue were within 3-fold. The Ki values determined for the displacement of [¹²⁵I]IABN and [³H]LS-3-134 by spiperone using hD3 HEK-293 cells were also similar. The Ki values determined by the displacement of [³H]LS-3-134 by spiperone using rD3 HEK-293 cells was found to be 4- to 5-fold lower than that observed in rat striatal tissue.

Figure 7. Competition radioligand binding studies using spiperone or eticlopride with rat striatum.

Composite competitive inhibition of radioligand binding of [³H]LS-3-134 binding by D2-like dopamine receptor selective antagonists spiperone (■) and eticlopride (●) using rat striatal tissue. Competition data was analyzed using a one-site model using (+)-butaclamol to define nonspecific binding. The nonlinear correlation coefficient for spiperone (n = 3) was r² = 0.96 and r² = 0.95 for eticlopride (n = 2).

The competition data presented in Figure 7 is consistent with the radioligand binding to a homogeneous population of binding sites in rat striatal tissue since an r² > 0.95 was observed. It should be noted that at the higher concentration of inhibitor we observed a displacement of nonspecific binding as defined by (+)-butaclamol.

We then examined the ability of the D2 versus D3 dopamine receptor selective compound SV 293 (Figure 8) to displace [³H]LS-3-134 radioligand binding. SV 293 is an indole that is a neutral antagonist for adenylyl cyclase inhibition with 100-fold selectivity for hD2 dopamine receptors compared to the hD3 dopamine receptor subtype (Vangveravong et al., 2006; Luedtke et al., 2012). The Ki values determined by the displacement of [³H]LS-3-134 by SV 293 using hD3 HEK-293 cells, rD3 HEK-293 cells and rat striatum were consistently found to be >300 nM.

Figure 8. Competition radioligand binding studies using dopamine D2 and D3 receptor selective compounds SV 293 and WC 10.

Composite competitive inhibition of the binding of [³H]LS-3-134 binding sites by D2 vs. D3 (SV 293, (●)) or D3 vs. D2 (WC 10, (■)) dopamine receptor selective compounds using either rat D3 receptors expressed in HEK cells (Top) or rat striatal tissue (Bottom). For these experiments the inhibitor concentration ranged from 10⁻¹¹ to 10⁻⁵ Molar using 3 points per decade. Competition data was fit to a one-site model using (+)-butaclamol to define nonspecific binding. The nonlinear correlation coefficient for WC 10 (n = 3) and SV 293 (n = 4) using rD3 HEK cells were both r² = 0.99. The nonlinear correlation coefficients for WC 10 (n = 3) and SV 293 (n = 3) inhibition for rat striatal tissue was 0.97 and 0.95, respectively.

The Ki values for WC 10, a D3 vs D2 dopamine receptors selective compound was also determined. WC 10 is a phenylpiperazine which has 43-fold binding selectivity for hD3 vs. hD2 receptors (Table I). Ki values were consistent for the displacement of [¹²⁵I]IABN and [³H]LS-3-134 in hD3 HEK-293 cells (Table IV). Ki values were also consistent for the displacement of [³H]LS-3-134 in rD3 HEK-293 cells and rat striatal tissue (Table IV and Figure 8).

Figure 8 shows a comparison of composite (n > 3) competitive radioligand binding for the inhibition of [³H]LS-3-134 binding by SV 293 and WC 10 using either rD3 HEK cells or rat striatal tissue. When rD3 HEK cells were used, two concentrations of inhibitor was used per decade. For the studies using rat striatal tissue, three concentrations of inhibitor per decade was used. The appropriate rank order of inhibition was observed. Nonlinear one site fit analysis provided r² values of 0.99 for data obtained using rD3 HEK cells and r² > 0.95 for data obtained using rat striatal tissue, which is again consistent with [³H]LS-3-134 binding to a homogeneous population of binding sites in both of these tissue preparations. It should be noted that at the higher concentrations of inhibitor a displacement of nonspecific binding, as defined using (+)-butaclamol, was observed (Figure 8).

Finally, the ability of [³H]LS-3-134 to bind selectively to the D3 vs. D2 dopamine receptor subtypes was further investigated in two ways. First, equilibrium saturation studies were attempted twice using hD2 HEK cells with [³H]LS-3-134. However, the data from both experiments were not interpretable, suggesting low-affinity binding at the D2 receptor subtype.

Second, we performed a series of direct binding experiment using increasing amounts of HEK-293 cell membranes expressing human dopamine D2 receptors in the presence of a constant amount of hD3 dopamine receptors expressed in HEK-293 cells and a constant concentration of radioligand. These studies were performed using [³H]LS-3-134 and two radioligands which bind to D2 and D3 dopamine receptors with essentially equal affinity: a) [¹²⁵I]IABN and b) [³H]raclopride (Figure 9). (+)-Butaclamol was used to define nonspecific binding. As expected, increased specific binding was observed when increasing amounts of D2 dopamine receptors were added to the binding assay when the nonselective radioligands [¹²⁵I]IABN (Figure 9A) and [³H]raclopride (Figure 9B) were used. However, specific binding did not increase when [³H]LS-3-134 was used (Figure 9C).

Figure 9. Binding of [¹²⁵I]IABN, [³H]raclopride and [³H]LS-3-134 to hD3 receptors in the presence and absence of increasing amounts of hD2 dopamine receptors expressed in HEK-293 cells.

This figure shows the comparison of the binding of three D2-like dopamine receptor selective radioligands to D3 dopamine receptors in the presence of increasing amounts of D2 dopamine receptors expressed in HEK-293 cells. The radioligands that were used include: (A) [¹²⁵I]IABN, (B) [³H]raclopride and (C) [³H]LS-3-134. A constant amount of hD3 HEK-293 cell membranes (2 μg of protein) plus increasing amounts of hD2 HEK-293 cell membranes (from 0.5-4 μg of protein) were added to [³H]raclopride and [³H]LS-3-134 assay. For [¹²⁵I]IABN a constant amount of hD3 HEK-293 cell membranes (0.2 μg of protein) plus increasing amounts of hD2 HEK-293 cell membranes (0.005-0.25 μg of protein) were added. The specific binding (bound total – bound nonspecific) is showed for each radioligand, where the nonspecific binding was defined using (+)-butaclamol.

Discussion

It has been challenging to develop D3 dopamine receptor selective radioligands for in vitro or in vivo studies because of the high degree of structural homology between the D2 and D3 dopamine receptor subtypes. Difficulty in identifying in vivo and in vitro D3 vs. D2 dopamine receptor selective imaging agents has been further compounded because of the lower level of expression of the D3 receptor compared to the D2 receptor subtype. The identification of compounds with the following properties would be required to achieve this goal, including a) high-affinity binding at D3 receptor, b) D3 vs. D2 and D4 receptor subtype binding selectivity (likely >100-fold), c) minimal nonspecific and off-target binding. For in vivo imaging studies, the ability to cross the blood brain barrier and to compete with endogenous dopamine are also essential properties.

For the studies described in this report, [³H]LS-3-134 was prepared and evaluated for use as a D3 dopamine receptor selective radioligand using classic kinetic and equilibrium binding techniques (McGonigle and Molinoff, 1989). This arylamide 2-fluoroethoxy substituted phenylpiperazine compound was selected for radiolabeling studies because it was previously reported to exhibit a) high-affinity binding (Ki value 0.2 nM) at hD3 dopamine receptors, b) >150-fold hD3 vs. hD2 dopamine receptor subtype binding selectivity and c) low-affinity binding (Ki values >5,000 nM) at sigma 1 and sigma 2 receptors. Furthermore, based upon adenylyl cyclase inhibition studies, LS-3-134 is a weak partial agonist at both D2 and D3 dopamine receptor subtypes (Tu et al., 2011a).

Bimolecular association rate constants for the binding of soluble proteins with small molecular weight compounds (300-700 Da) can vary from 10⁵ to 10⁹ M⁻ sec⁻, with association rate constants of >10⁷ M⁻¹ s⁻¹ being compatible with diffusion-limited binding (Zhou et al., 1983). Our estimated association rate constants for hD3 (3.2 × 10⁶ M⁻¹sec⁻¹) and rD3 receptors (1.75 × 10⁷ M⁻¹sec⁻¹) differed by approximately 5-fold, with equilibrium reached within 10 to 15 minutes (37°C). The kinetic association studies performed using membranes from transfected HEK-293 cells indicated that the components of this assay remain stable over a 60-minute time course. The dissociation rate constants for hD3 compared to rD3 receptor expressed in transfected HEK-293 cells were found to be similar, differing by only about 2-fold (0.021 ± 0.004 min⁻ vs. 0.039 ± 0.006 min⁻, respectively).

The results from the association and dissociation kinetic studies were consistent with equilibrium direct binding studies, which indicated that [³H]LS-3-134 binds hD3 receptors with a 4-fold higher affinity compared to rD3 dopamine receptors. While this is only a slight difference in affinity, information on species differences provides the neuroscience research community with the pharmacological information for using this radioligand to study the regulation of D3 dopamine receptors in rats and humans.

Although there is high homology between rat and human dopamine receptor subtypes, the greatest homology is generally found in the transmembrane spanning regions and reduced homology is found in the extracellular and intracellular loops. Our previous molecular modeling studies (Wang et al., 2010) indicate that the phenylpiperazine moiety is the orthosteric pharmacophore and the arylamide interacts with portions of extracellualr loops EL1 and EL2. The displacement of water molecules by the aryl moiety has been proposed to influence ligand binding affinity and play a role in D3 vs. D2 receptor binding selectivity (Newman et al., 2012). Species specific differences in primary structure within this portion of the rat and human D3 receptor orthologs are likely responsible for the observed slight difference in affinity of [³H]LS-3-134 for hD3 and rD3 dopamine receptors.

The equilibrium dissociation constants (Kd values) that were obtained for the binding of [³H]LS-3-134 to rD3 receptors expressed in HEK-293 cells were similar to values obtained for the affinity of [³H]LS-3-134 binding sites expressed in rat striatal and nucleus accumbens tissue. The linearity of the Scatchard plots using rat striatal and nucleus accumbens tissue suggest that [³H]LS-3-134 is binding to a homogeneous population of binding sites. In addition, Bmax values (80-100 fmoles/mg protein) were consistent with previous binding and in situ hybridization studies indicating that D3 receptors are expressed at a 5- to 10-fold lower density than the dopamine D2 receptor subtype.

For this report we have attempted to identify and define conditions that might influence the in vitro binding affinity of [³H]LS-3-134 based upon previous literature on the binding of ligands and radioligands to D2-like dopamine receptors. Previous studies by Neve and co-workers (1990) reported that the D2 vs. D3 receptor nonselective radioligand [¹²⁵I]epidepride, binds with different affinity to dopamine D2 receptors in the absence and presence of NaCl. The affinity of [¹²⁵I]epidepride at D2 receptors was >10-fold different (from 20 to 30 pM) in the presence of NaCl compared to the affinity in the absence of NaCl (250 to 500 pM). Those data suggested that the conformation of D2-like dopamine receptors might be regulated by sodium. Subsequent D2 receptor molecular modeling studies (Neve et al., 2001) suggested the existence of a sodium binding site and the involvement of two serine residues. The studies presented in Table III and Figure 4 appear to rule out NaCl concentration as a modulator of binding affinity of [³H]LS-3-134 at the D3 dopamine receptor subtype.

Both direct and competitive radioligand binding studies using rat striatum suggest that [³H]LS-3-134 binds a homogeneous population of binding sites (one-site fit for competition binding data) with a dopamine D3 receptor pharmacological profile. The rank-order-of-binding for a panel of D2-like dopaminergic ligands to rat striatal [³H]LS-3-134 binding sites was found to be consistent with Ki values obtained using rD3 receptors expressed in transfected cells ((+)-butaclamol < eticlopride < WC 10 < SV293 < (−)-butaclamol).

It is challenging to unequivocally confirm the D2 vs. D3 receptor subtype selectivity directly using the radiolabeled [³H]LS-3-134. The affinity for the binding of LS-3-134 to hD3 receptors expressed in HEK-293 cells was originally determined from competitive binding studies (Ki value = 170 ± 10 pM) using [¹²⁵I]IABN (Tu et al., 2011a). Affinity measurement obtained in present studies from saturation isotherms using radiolabeled [³H]LS-3-134 (Kd value = 56 ± 12 pM) was found to be comparable to the Ki value. Furthermore, the original competition binding studies indicated that LS-3-134 was >150-fold selective for hD3 receptors compared to hD2 receptors (Tu et al., 2011a).

Equilibrium saturation isotherms for the binding of [³H]LS-3-134 to hD2 dopamine receptors using membranes from transfected cell lines were attempted but no specific binding was observed. This is likely due to an increased level of nonspecific binding that was present as higher concentrations of radioligand were required, because of the lower affinity of the radioligand at D2 receptors. As an alternative strategy for examining the issue of the D3 versus D2 receptor binding selectivity, we carried out a set of experiments in which we mixed transfected cells membranes expressing either the D2 or the D3 receptor subtypes. This receptor mixing experiment was performed to mimic the in vivo co-expression of these two receptor subtypes in the same region of the brain. In these experiments the binding of [³H]LS-3-134, [³H]raclopride or [¹²⁵I]IABN was monitored using a constant amount of human D3 receptors with increasing amounts of human D2 receptors. For these studies the concentration of each radioligand was approximately equal to its Kd value for D3 receptor binding, therefore, approximately 50% of the D3 receptors were occupied. As expected for the D2/D3 receptor nonselective radioligands [³H]raclopride and [¹²⁵I]IABN, a gradual increase in specific binding was observed as D2 receptor levels increased. In contrast, when [³H]LS-3-134 was used no increase in specific binding was observed, suggesting that there was minimal to no binding of the radioligand at D2 receptors.

Radioligands that bind nonselectively to D2 and D3 dopamine receptors, including [³H]spiperone, [³H]raclopride, [¹²⁵I]epidepride, [¹²⁵I]NCQ298 and [¹²⁵I]IBZM, have been available for decades. However, there are several properties inherent to the D3 dopamine receptors which have made it problematic for developing a D3 receptor subtype selective radioligand that can be reliably used to quantitate changes in D3 receptor expression. The low level of D3 dopamine receptor expression and the neuroanatomical co-expression of D2 and D3 receptor subtypes have further complicated this effort. Furthermore, as we have found from these studies, there can be subtle species-related differences in radioligand binding affinity.

Based upon the results presented in this report, it appears that [³H]LS-3-134 can be used to selectively label the D3 dopamine receptor subtype in rat brain tissue. The observed a) monophasic equilibrium binding modes found for direct and competition experiments, b) low level of binding site expression (Bmax values) and c) pharmacological rank order of affinity of dopaminergic ligands are consistent with the labeling of a single population of binding sites, which is the dopamine D3 receptor subtype. However, the low level of D3 dopamine receptor expression, in concert with the low specific activity of tritium, necessitates that multiple rat brains are required to perform saturation isotherm studies. The relative receptor density of individual rats might be compared (e.g., for drug treated vs. vehicle controls or pathology vs. normal controls) using a single radioligand concentration, if it is assumed that the affinity of the radioligand is invariant. In this situation that assumption is likely valid because LS-3-134 is a weak partial agonist at D3 dopamine receptors.

This low intrinsic efficacy differentiates [³H]LS-3-134 from the D2/D3 agonist [³H]7-OH-DPAT, which has previously been used extensively to quantitate D3 dopamine receptor expression. In addition, kinetic analyses indicated that although [³H]R(+)-7-OH-DPAT has a lower molecular weight than the arylamide phenylpiperazines a) 1 hr was required to achieve binding equilibrium (suggesting a possible conformational transition), b) two component association and dissociation curves were observed and c) saturation curves and competition curves suggested the presence of multiple binding components (Hillefors and von Euler, 2001). Radiolabeled agonists likely underestimate D3 receptor density if they bind with low affinity to an uncoupled (“low affinity state”) receptor population.

Several previous PET imaging studies of the dopamine D3 receptor have relied on the use of [¹¹C](+)-PHNO, which was initially developed to image the high affinity state of the D2 receptor (^highD2). However, subsequent studies have revealed that (+)-PHNO binds with high affinity to both D2 and D3 receptors. Consequently, [¹¹C](+)-PHNO has been referred to as a “D3 preferring” radiotracer since it is capable of imaging D3 dopamine receptors better than [¹¹C]raclopride, the most commonly used D2/D3 radiotracer for PET imaging studies. However, recent studies have revealed that [¹¹C](+)-PHNO is not capable of providing accurate measures of D3 receptor density in regions of brain where both D2 and D3 receptors are expressed (Tziortzi et al., 2011; Payer et al., 2014a,b). For example, in the human striatal regions [¹¹C](+)-PHNO PET signal primarily represents D2 receptor binding (Erritzoe et al., 2014).

Our initial PET studies with [¹⁸F]LS-3-134 indicate that this radiotracer labels D3 receptors in the absence of D2 receptors in nonhuman primate brain after lorazepam pretreatment to deplete endogenous dopamine, in order to minimize competition between the radiotracer and endogenous dopamine (Mach et al., 2011). In those studies labeling was found to be consistent with known neuroanatomical expression of dopamine D3 receptors. This current study using a [³H]-labeled analog provides further evidence that LS-3-134 has high a) D3 dopamine receptor affinity and b) binding selectivity for D3 versus D2 dopamine receptors. It also provides evidence suggesting that [³H]-labeled LS-3-134 binds to a homogeneous population of binding sites. However, the conditions for in vivo and in vitro binding studies are quite different. For in vivo studies trace labeling and kinetic modeling analysis is used to quantitate binding, while in vitro binding study protocols often include saturation binding performed at equilibrium. Therefore, further molecular pharmacological studies will be required to examine the a) binding selectivity of these radiolabeled analogs of LS-3-134 in both in vivo and in vitro binding scenarios and b) possible contribution off-target binding with varying experimental protocols.

In summary, [³H]LS-3-134 was synthesized and a series of in vitro binding studies were conducted to characterize its pharmacological properties as a selective radioligand for dopamine D3 receptors. Based upon the molecular pharmacological studies described in this report we propose that the [³H]-radiolabeled analog of LS-3-134 is a D3 dopamine receptor selective radioligand that could be used to study the expression and regulation of the D3 dopamine receptor subtype. It also provides further supportive evidence for using [¹⁸F]LS-3-134 as a promising radiotracer for imaging D3 dopamine receptors independently of the D2 receptor subtype in vivo with PET.

Abbreviations

Bmax

maximum binding sites

GPCR

G protein coupled receptor

hD3

human D3 dopamine receptors

IC50

concentration of inhibitor at which 50% inhibition is achieved

Kd

equilibrium dissociation constant determined using a constant amount of receptor and increasing concentrations of radioligand

Ki

equilibrium association constant determined for competition studies

PET

positron emission tomography

rD3

rat D3 dopamine receptors

TMS

transmembrane spanning