Abstract
Some recently published in vitro studies with two metabotropic glutamate 2/3 receptor (mGluR2/3) agonists [(−)-2-oxa-4-aminobicyclo[3.1.0] hexane-4,6-dicarboxylic acid (LY379268) and 1S,2S,5R,6S-2-aminobicyclo[3.1.0]hexane-2,6-bicaroxylate monohydrate (LY354740)] suggest that these compounds may also directly interact with dopamine (DA) D2 receptors. The current in vitro and in vivo studies were undertaken to further explore this potential interaction with D2 receptors. LY379268 and LY354740 failed to inhibit D2 binding in both native striatal tissue homogenates and cloned receptors at concentrations up to 10 μM. LY379268 and LY354740 (up to 10 μM) also failed to stimulate [35S]GTPγS binding in D2L- and D2S-expressing clones in the presence of NaCl or N-methyl-d-glucamine. In an in vivo striatal D2 receptor occupancy assay, LY379268 (3–30 mg/kg) or LY354740 (1–10 mg/kg) failed to displace raclopride (3 μg/kg i.v.), whereas aripiprazole (10–60 mg/kg) showed up to 90% striatal D2 receptor occupancy. LY379268 (10 mg/kg) and raclopride (3 mg/kg) blocked d-amphetamine and phencyclidine (PCP)-induced hyperactivity in wild-type mice. However, the effects of LY379268 were lost in mGlu2/3 receptor knockout mice. In DA D2 receptor-deficient mice, LY379268 but not raclopride blocked both PCP and d-amphetamine-evoked hyperactivity. In the striatum and nucleus accumbens, LY379268 (3 and 10 mg/kg) was without effect on the DA synthesis rate in reserpinized rats and also failed to prevent S-(−)-3-(3-hydroxyphenyl)-N-propylpiperidine-induced reductions in DA synthesis rate. Taken together, the current data fail to show evidence of direct DA D2 receptor interactions of LY379268 and LY354740 in vitro or in vivo. Instead, these results provide further evidence for a novel antipsychotic mechanism of action for mGluR2/3 agonists.
Glutamate is the primary excitatory neurotransmitter in the mammalian central nervous system, and the variety of its effects in the brain is regulated in part by the diversity of receptor subtypes that respond to glutamate. These receptors include both ionotropic (ligand-gated ion channels) and metabotropic (G protein-coupled) receptor families (Dingledine et al., 1999; Schoepp, 2001). Although the ionotropic receptors were the first to be discovered and have been of significant interest as potential therapeutic targets, the metabotropic glutamate (mGlu) receptors are now also of great interest for potential pharmacotherapeutic development. There are eight mGlu receptors, which have been divided into three subgroups, group I (mGlu1 and mGlu5), group II (mGlu2 and mGlu3), and group III mGlu (mGlu4/6/7/8) receptors, on the basis of molecular structure, pharmacology, and signal transduction pathways. The development of selective pharmacological tools for various mGlu receptor subtypes has enabled animal studies suggesting that these receptors may provide potential targets for the treatment of a wide variety of psychiatric and neurological disorders in humans (for reviews see, Schoepp, 2001; Niswender et al., 2005; Pilc et al., 2008).
The potential therapeutic promise of mGlu receptor-targeted molecules was recently reinforced by the finding that LY2140023 (the prodrug of the mGlu2/3 agonist LY404039) produced statistically significant improvements in symptoms in a Phase II trial for patients with schizophrenia (Patil et al., 2007). Preclinical data suggest that the antipsychotic actions of mGlu2/3 receptor agonists such as LY404039 are mechanistically distinct from present antipsychotic drugs and mediated via the selective activation of mGlu2 receptors. For example, the ability of mGlu2/3 receptor agonists to attenuate phencyclidine (PCP) and amphetamine (AMP)-evoked hyperactivity in mouse models of psychosis is blocked by the mGlu2/3 receptor antagonist LY341495 (Cartmell et al., 1999; Rorick-Kehn et al., 2007). Studies carried out in transgenic mice reveal that the antipsychotic actions of the mGlu2/3 receptor agonists LY404039 or LY379268 are lost in mice with targeted deletions of mGlu2, mGlu2/3, but not mGlu3 receptors (Patil et al., 2007; Fell et al., 2008; Woolley et al., 2008), whereas antipsychotic drugs (olanzapine, risperidone, and clozapine) block PCP-induced locomotor activity in both the wild-type and mGlu2/3 receptor knockout animals.
However, two different mGlu2/3 receptor agonists, LY354740 and LY379268, were recently suggested to possess partial agonist activity at both the dopamine (DA) D2 long (D2L) and D2 short (D2S) isoforms of the dopamine D2 receptor (Seeman and Guan, 2008; Seeman et al., 2008). The primary in vitro evidence used to support this claim was a relatively high affinity for displacing [3H]domperidone binding in either rat striatal tissue or cloned human D2L and D2S receptors, and an apparent partial agonist activity as measured by the stimulation of [35S]GTPγS binding in cells expressing either cloned dopamine D2L or D2S receptors. Ki values for LY379268 at D2 receptors were reported to be in the range of 5 to 30 nM and between 20 and 50 nM for LY354740. Based on the structural similarities of LY404039 to LY379268 and LY354740, Seeman (2008) raised the possibility that the clinical activity of LY2140023 might be due to the actions of LY404039 at D2 receptors rather than mGlu2/3 receptors.
The findings reported by Seeman and co-workers (2008) of dopamine D2 receptor affinity and partial agonist activity of LY379268 and LY354740 were surprising, because previous studies had not suggested any direct interactions with dopaminergic receptors. Thus, the current work was undertaken as an attempt to replicate the in vitro findings of Seeman and co-workers (2008) and to look for the ability of either LY379268 and LY354740 to 1) displace [3H]domperidone binding to rat striatal tissue or membranes containing human D2L or D2S receptor isoforms and 2) stimulate either D2L or D2S receptor activity as measured by receptor-stimulated binding of [35S]GTPγS in membranes expressing either human isoform. In addition, we conducted a series of behavioral and neurochemical studies that were designed to evaluate the possible interaction of mGlu2/3 receptor agonists with D2 receptors in vivo. Thus, we assessed in vivo striatal D2 receptor occupancy in the rat for LY379268, LY354740 monohydrate, and the partial D2 receptor agonist aripiprazole by using an LC/MS/MS-based method (Barth et al., 2006). Furthermore, we compared LY379268 to the D2/D3 antagonist raclopride for reversal of PCP and d-amphetamine-induced hyperlocomotion in wild-type mice or mice deficient of mGlu2/3 receptors or DA D2 receptors. Finally, brain DA synthesis rate in reserpinized rats (Svensson et al., 1991) was used as a sensitive in vivo neurochemical assay to study potential agonist/antagonist properties of mGlu2/3 receptor agonists at D2 receptors.
Materials and Methods
In Vitro Experiments.
Materials.
Domperidone, S-(−)-sulpiride, 3-hydroxytyramine (dopamine), raclopride, 3-(3-hydroxyphenyl)-N-propylpiperidine (3-PPP), trizma base, N-methyl-d-glucamine (NMDG), and polyethylenimine were purchased from Sigma-Aldrich (St. Louis, MO). Magnesium chloride hexahydrate, EDTA, and sodium chloride were acquired from Mallinkrodt Chemical (Paris, KY). Aripiprazole, LY354740, and LY379268 were synthesized at the Lilly Research Laboratories (Indianapolis, IN). [3H]Domperidone was custom made (50 Ci/mmol) and purchased from PerkinElmer Life and Analytical Sciences (Waltham, MA). [35S]GTPγS (1250 Ci/mmol) was obtained from PerkinElmer Life and Analytical Sciences. Membranes containing human D2 dopamine receptor (short variant, D2S) transfected into Chinese hamster ovary cells were purchased from PerkinElmer Life and Analytical Sciences. Membranes containing human D2 dopamine receptor (long variant, D2L) transfected into Chem-1 cells (Millipore Bioscience Research Reagents, Temecula, CA) were obtained from Millipore Corporation (Billerica, MA). Wheat germ agglutinin PVT SPA beads were purchased from GE Healthcare (Chalfont St. Giles, UK).
[3H]Domperidone Binding Assay.
Membrane preparation.
Rat striata were purchased from Zivic Laboratories (Pittsburgh, PA). Rat striatal homogenates were prepared as described previously (Wainscott et al., 2005). Rat striatal membranes were thawed, diluted to 0.085 mg/tube in binding buffer (13 mM MgCl2, 1.67 mM EDTA, 133 mM NaCl, and 67 mM Tris-HCl, pH 7.4), and homogenized with a polytron. hD2L and hD2S membranes were thawed, diluted to 0.0022 and 0.0085 mg/tube in binding buffer, respectively, and then homogenized with a polytron.
[3H]Domperidone binding.
The domperidone binding assays were adapted from that originally described by Grigoriadis and Seeman (1986). Binding assays were performed in duplicate in 0.4-ml total volume. Membrane suspension (100 μl) and 100 μl of drug dilution were added to 200 μl of binding buffer containing [3H]domperidone (2 nM) in minitubes. Minitubes were incubated at room temperature for 60 min. The incubation was terminated by rapid filtration through Whatman hydrophilic GF/C, 96-well filter plates that had been presoaked in 0.5% polyethylenimine. The filter plate was washed three times with ice-cold 50 mM Tris-HCl, pH 7.4. Nonspecific binding was defined by sulpiride (10 μM). The amount of [3H]domperidone trapped on the filters was determined by liquid scintillation spectrometry. The actual free radioligand concentration was determined by sampling the supernatant of identical tubes in which the bound radioligand was separated from the free radioligand by centrifugation.
D2 Receptor-Stimulated [35S]GTPγS Binding Assay.
Compounds were diluted in serial half-log dilutions in either deionized H2O for agonist studies or 1 μM dopamine for antagonist studies. Nonspecific binding was defined in the presence of deionized H2O for agonist studies or 10 μM raclopride for antagonist studies, with maximal binding defined by 10 μM dopamine for agonist studies and 1 μM dopamine for antagonist experiments. All tests were performed in duplicate.
[35S]GTPγS binding was modified from published conditions (Sim et al., 1995; Thomas et al., 1995). Assays were performed in Tris-HCl, MgCl2, EDTA, and either NaCl or NMDG to enhance partial agonist activity (Lin et al., 2006) with an adaptation to scintillation proximity technology. Final concentrations of each component were 50, 10, and 0.5 mM, respectively, for Tris-HCl, MgCl2, and EDTA and 100 mM for NaCl or NMDG. [35S]GTPγS was diluted in assay buffer containing GDP, pargyline, and sodium ascorbate [10 μM, 10 μM, and 0.1% with [35S]GTPγS (0.2 nM) final concentrations, respectively]. Wheat germ agglutinin PVT SPA beads were diluted in assay buffer at 500 mg/25 ml, mixed well, and added to each well at 1 mg bead/well.
Membranes containing human D2 dopamine receptor (short variant, D2S) transfected into Chinese hamster ovary cells were thawed quickly, and 1.7 mg of protein were diluted in 20 ml of assay buffer; contents were homogenized and added to the assay plates at 50 μl/well. Membranes containing human D2 dopamine receptor (long variant, D2L) transfected into Chem-1 cells (Millipore Bioscience Research Reagents) were thawed, and 3 mg of cells were homogenized in 20 ml of buffer and added to the assay plates at 50 μl/well.
Plates were sealed with clear plate sealers (PerkinElmer Life and Analytical Sciences) and allowed to incubate at room temperature for 2 h before being counted on 35S settings on the PerkinElmer 1450 Microbeta Trilux Scintillation and Luminescence Counter. Data were calculated with Microsoft Excel to convert points to percent-specific binding then analyzed with GraphPad Prism (GraphPad Software Inc., San Diego, CA) by using nonlinear regression analysis to fit curves and generate EC50 and Emax values for compounds run in agonist mode and IC50 and Emin values for compounds run in antagonist mode.
In Vivo Experiments.
Animals.
All experiments were conducted in accordance with the National Institutes of Health Guide for Care and Use of Laboratory Animals (http://www.nap.edu/readingroom/books/labrats/) and were approved by the Eli Lilly Institutional Animal Care and Use Committee. Eli Lilly is an American Association of Laboratory Animal Care-accredited facility. Male Sprague-Dawley rats (Harlan, Indianapolis, IN) weighing between 200 to 350 g were tested in the receptor occupancy and the dopamine synthesis rate assays. The sample sizes for each experiment were as follows: receptor occupancy, n = 4 to 5 per group; dopamine synthesis rate n = 6 to 12/group. Rats were group housed with standard laboratory chow and water available ad libitum. Rats were maintained on a 12-h light-dark cycle (lights on at 6:00 AM, lights off at 6:00 PM). Male mice (mGlu2/3 double knockouts, DA D2 knockouts, and their respective wild-type animals) weighing approximately 25 to 30 g (Taconic Farms, Germantown, NY) were tested in PCP or AMP-induced hyperlocomotion experiments (n = 8 per group). Mice were housed in groups by genotype under a 12-h light-dark cycle (lights on at 6:00 AM, lights off at 6:00 PM) with food and water freely available.
Materials.
LY354740 monohydrate, LY379268, and aripiprazole were synthesized at Lilly Research Laboratories. d-Amphetamine sulfate, PCP hydrochloride, quinpirole, S-(−)-3-PPP HCl, and reserpine were obtained from Sigma-Aldrich. Raclopride was purchased from Tocris Bioscience (Ellisville, MO) and dissolved in 0.9% NaCl by the drop-wise addition of 8.5% lactic acid. LY354740 monohydrate and LY379268 were dissolved in sterile 0.9% NaCl by the dropwise addition of 5 N NaOH. d-Amphetamine sulfate and phencyclidine hydrochloride were dissolved in 0.9% saline. Quinpirole and S-(−)-3-PPP were dissolved in sterile water. Aripiprazole and reserpine were dissolved in minimal quantities of dimethyl sulfoxide.
Receptor Occupancy Assay.
In vivo D2 occupancy and analysis of unlabeled raclopride levels by LC/MS/MS.
Radioactivity raclopride was used as an occupancy tracer in the LC/MS/MS-based experiments to assess in vivo D2 occupancy and was administered at an intravenous dose of 3 μg/kg dissolved in sterile water (Barth et al., 2006). To examine D2 dose-occupancy relationships LY379268, LY354740 monohydrate, and aripiprazole were administered to groups of 4 to 5 rats pretreated with test compound or its vehicle. One hour later (90 min for aripiprazole), animals were briefly restrained and administered a low intravenous dose of unlabeled raclopride tracer via the lateral tail vein. LY379268 was administered at doses of 1, 3, and 10 mg/kg i.p. LY354740 monohydrate was administered at doses of 10, 30, and 60 mg/kg i.p. Aripiprazole was administered p.o. at doses of 10, 30, and 60 mg/kg. Fifteen minutes after the intravenous administration of the raclopride tracer rats were sacrificed by cervical dislocation, and rat brain striatum and cerebellum tissue samples were dissected, weighed, and kept on ice. Brain tissues samples were processed, and raclopride measurements were made by using a liquid chromatograph [high-performance liquid chromatography (HPLC); model 1100, Agilent Technologies, Wilmington, DE] with triple quadrapole mass spectral detection. The HPLC system used a C18 column (Zorbax Eclipse, 2.1 × 50 mm, 3.5-μm particle size; Agilent Technologies) with an aqueous mobile phase consisting of 32% acetonitrile with 0.1% formic acid. Raclopride was quantified after elution from the HPLC column by using an API 4000 triple quad mass spectrometer (Applied Biosystems, Foster City, CA) in positive electrospray mode using multiple reaction monitoring methods to observe the transition from parent to daughter ions with mass-to-charge ratios of 347.1 and 112.3, respectively. Chromatographic assays were calibrated by using a standard curve generated by extracting a series of brain tissue samples from nontreated animals to which known quantities of raclopride had been added. Please refer to Chernet et al. (2005) and Barth et al. (2006) for more detailed methods.
Receptor occupancy calculation
Receptor occupancy calculations were made for each animal by using the widely used ratio method (Farde et al., 1988; Kapur, 2000; Wadenberg et al., 2000) and the following equation: 100 × (1 − [(Ratiot − 1)/(Ratioc − 1)]) = %Occupancy.
The “Ratiot” represents the ratio of raclopride concentrations measured in the striatum to those measured in the cerebellum in individual animals pretreated with antipsychotic compounds. The “Ratioc” represents the average ratio of raclopride levels measured in the striatum to that measured in the cerebellum for the vehicle-pretreated group.
Mouse Locomotor Activity Assessment.
Generation of mGlu2/3 receptor knockout mice.
mGlu2/3 receptor knockout mice (Taconic Farms) were generated by homologous recombination as described in detail by Fell et al. (2008). The targeting construct was injected into the R1 line of mouse embryonic stem cells. The recombinant embryonic stem cells were injected into murine C57BL/6 blastocysts, and chimeric males were mated with ICR(CD-1) females. Male offspring carrying the null allele were backcrossed for three generations (N3) with ICR(CD-1) females. N3 offspring were then interbred, and siblings homozygous for the null allele were used as founders that provided all mGlu2 receptor knockout mice and mGlu3 receptor knockout mice used here. Males homozygous for the null allele of mGlu3 were intercrossed with females homozygous for the null allele of mGlu2 to produce an intercross colony heterozygous for both mGlu2 and mGlu3 null alleles. Breeding of siblings from the intercross colony produced an intracross colony, and progeny were genotyped to identify homozygotes for both the mGlu2 and mGlu3 null alleles. Founder animals were selected to generate colonies of double knockout mice and wild-type mice that were used in all subsequent experiments.
Generation of dopamine D2 receptor knockout mice.
Mice with selective deletions of dopamine D2L and D2S receptors (Taconic Farms) were generated by homologous recombination as previously described in Kelly et al. (1997). The D2 dopamine receptor targeting vector was electroporated into the D3 ES cell line (Doetschman et al., 1985), and correctly targeted ES cells were injected into embryonic day 3.5 B6 blastocysts, as described previously (Rubinstein et al., 1996). Mice used in this study were congenic N5 generated by backcrossing to the B6 strain for five generations. Wild-type C57BL/6J mice were obtained from Taconic Farms.
Assessment of locomotor behavior.
Behaviors were monitored in transparent, plastic shoebox cages with dimensions of 45 × 25 × 20 cm, a 1-cm depth of wood chips as bedding, and a plastic cage top. These items were placed in a rectangular frame containing a grid of 12 photocell beams in an 8 × 4 configuration (Kinder Scientific, Poway, CA) that was positioned 2.5 cm from the floor of the cage for the detection of body movements (ambulations) and recorded by computer for analysis. Mice were initially acclimated to a plastic shoebox cage for 15 min before being given an intraperitoneal injection of LY379268 or raclopride or sterile water. After an additional 30 min, the mice were administered an intraperitoneal injection of sterile water or PCP. Motor activity was monitored for 60 min after the injection of PCP. The same test paradigm was followed for AMP treatment. All drugs were mixed fresh before use and administered intraperitoneally. Mice were dosed with a volume of 10 ml/kg. Software analysis of beam breaks, under the definitions of Hamilton Kinder, resulted in the measurement of three different parameters: ambulations (pattern of beam breaks indicating that the animal has relocated its entire body), distance moved (cm−1), and time at rest (total seconds in a 60-min session in which no new beams were broken, measured at 1-s intervals). The doses of LY379268 (10 mg/kg) and raclopride (3 mg/kg) used in these studies were selected based on the results of preliminary studies carried out in our laboratories. Preliminary dose-response studies revealed that raclopride (1–10 mg/kg) and LY379268 (1–30 mg/kg) significantly reversed PCP (7.5 mg/kg) and AMP (5 mg/kg)-evoked behavioral activation in C57BL/6 and (ICR)CD-1 mice.
Dopamine Synthesis Rate in Reserpinized Rats.
Eighteen hours before sacrifice, male rats were given vehicle (2 ml/kg sterile saline) or reserpine (5 mg/kg s.c.) and returned to the home cage. The next day, rats were administered vehicle (2 ml/kg sterile saline), 3 mg/kg LY379268, 10 mg/kg LY379268, 10 mg/kg aripiprazole, 30 mg/kg S-(−)-3-PPP, or 1 mg/kg quinpirole. All drugs were administered 30 min before NSD1015 (100 mg/kg s.c.) except aripiprazole, which was dosed 60 min before NSD1015. Thirty minutes after the administration of NSD1015 (100 mg/kg), the animals were moved to an adjacent room before being sacrificed by decapitation and their brains were removed. Each brain was dissected, and the striatum and nucleus accumbens were removed and frozen on dry ice. In a second experiment, we assessed the ability of LY379268 (10 mg/kg) or haloperidol (0.5 mg/kg) to block S-(−)-3-PPP (30 mg/kg)-mediated reductions in the dopamine synthesis rate. As found in the previous experiment, male rats were administered 5 mg/kg reserpine s.c.18 h before sacrifice and returned to the home cage. The next day, rats were dosed with vehicle (2 ml/kg sterile saline), 10 mg/kg LY379268, 0.5 mg/kg haloperidol, 30 mg/kg S-(−)-3-PPP, or the combination of LY379268/S-(−)-3-PPP or haloperidol/S-(−)-3-PPP. All drugs were administered 30 min before NSD1015 (100 mg/kg s.c.). Thirty minutes after the administration of NSD1015 (100 mg/kg), the animals were moved to an adjacent room before being sacrificed by decapitation and the brain was removed for dissection. Each brain was dissected, and the striatum and nucleus accumbens were removed and frozen on dry ice. All drugs were mixed fresh before use and administered subcutaneously and dosed with a volume of 2 ml/kg.
Tissue samples were weighed individually and stored at −80°C in plastic tubes containing 1 ml of radioactivity HCl (0.01 N) and an antioxidant (0.5 mg/ml l-cysteine) until analyzed for DOPA levels. Immediately before analysis, samples were thawed at room temperature and ultrasonicated. After sonication, 100 μl of perchloric acid (1.5 M) was added, and the samples were vortexed and stored at 4°C for 60 min. The samples were then centrifuged for 5 min at 12,000 rpm, and the DOPA content of the supernatant was analyzed by HPLC with electrochemical detection. A BDS-Hypersil 5-μm C18 analytical column (2 × 150 mm; Thermo Fisher Scientific, Waltham, MA) was used. The mobile phase consisted of 75 mM sodium phosphate monobasic, 50 mg/l 1-octanesulfonic acid sodium salt, 0.5 mM EDTA, and 4% methanol, pH 2.6 (adjusted with phosphoric acid). The flow rate for the analytical column was 0.4 ml/min, which was maintained at 40°C with a column heater. A BAS LC4C electrochemical detector (BAS Bioanalytical Systems, West Lafayette, IN) with a glassy carbon electrode (E = 0.75 V) was used to detect DOPA at a range setting of 5 nA; 20 μl was injected onto the column. The data were collected by using an EZChrom chromatography data system (Scientific Software, San Ramon, CA) running on a Hewlett Packard computer, which calculated peak heights and sample concentrations.
Statistical Analysis.
Homologous competition of unlabeled domperidone with [3H]domperidone binding was used for calculation of the Kd and Bmax using GraphPad Prism, and the IC50 values from competition curves were determined by nonlinear regression analysis with Excel. Mean and S.E.M. were calculated for animals in each dose group. Neurochemical and behavioral data were analyzed by one-way analysis of variance, and then post hoc comparisons were made by a Bonferroni corrected t test.
Results
Radioligand Binding Studies.
[3H]Domperidone binding to D2 receptors has been reported to be able to distinguish agonists from antagonists, based on agonist production of biphasic displacement curves (Seeman et al., 2003). To test whether LY354740 or LY379268 might display similar properties, [3H]domperidone binding was carried out in membranes from rat striatum and in membranes expressing either the cloned human D2L or D2S receptors.
Seeman et al. (2008) have published that LY354740 and LY379268 displaced [3H]domperidone binding in rat striatal membranes in a biphasic manner and that LY379268 also displaced [3H]domperidone binding from membranes expressing the cloned human D2L receptor in a similar fashion. Such displacement is typical of agonists, with the high-affinity component representing the agonist high-affinity state of the receptor, and the lower affinity component representing the agonist low-affinity state. In agreement with a previous report (Seeman et al., 2003), we found that [3H]domperidone binding can reveal biphasic competition curves with the endogenous agonist dopamine (see Supplemental Fig. 1.) However, this phenomenon seemed to be dependent upon the membrane source. Thus, rat striatal membranes showed a robust biphasic displacement curve, the cloned human D2L-expressing membranes showed a small biphasic curve, and the D2S-expressing membranes showed only a simple monophasic displacement curve to dopamine. In the rat striatal membranes, DA inhibition of [3H]domperidone binding revealed a high-affinity component that constituted approximately 77% of the displaceable binding with an EC50 = 103 nM. In the human D2L-expressing membranes, DA also produced a biphasic curve with the high-affinity component accounting for 35% of the displaceable binding and having an EC50 = 118 nM. DA inhibition of [3H]domperidone binding in the D2S-expressing membranes gave only a monophasic displacement curve with an EC50 = 8821 nM.
LY354740 and LY379268 were both evaluated for interaction with the [3H]domperidone binding in all three D2 receptor-expressing membrane preparations. Figures 1 and 2 show the results of examining cloned human D2L and rat striatal membranes, respectively (data with cloned human D2S is shown in Supplemental Fig. 2). As can be seen from these figures, neither LY354740 nor LY379268 produced significant inhibition of binding in any of the membrane preparations up to the highest concentration tested (10 μM). The D2 receptor antagonist/weak partial agonist aripiprazole, run as a control, exhibited the expected high affinity in all three membrane preparations (EC50 = 5.76 nM, D2S; 3.84 nM, D2L; 3.08 nM, rat striatum), completely inhibiting specific [3H]domperidone binding. Dopamine, the endogenous agonist for the receptor, also completely inhibited [3H]domperidone binding with IC50 values dependent upon the membrane preparation examined. Unlabeled domperidone also inhibited [3H]domperidone binding with the expected high affinity. It is interesting to note that unlabeled domperidone was able to inhibit [3H]domperidone binding in striatal membranes to a lower level than the 10 μM sulpiride used to define nonspecific binding (Fig. 2), suggesting that in addition to D2 receptors [3H]domperidone binds to some non-D2 constituent in the striatum.
[35S]GTPγS Binding Studies.
To examine the possibility that LY354740 or LY379268 might have functional activity at D2 receptors, the compounds were tested in [35S]GTPγS binding assays by using cloned human D2S and D2L isoforms of the receptor. As controls, the endogenous agonist dopamine and the partial agonist 3-PPP were included in the assays. Aripiprazole, a very weak partial agonist, was also included. Aripiprazole, although having high affinity for D2 receptors, has such low efficacy that only certain in vitro assays are able to reveal its partial agonism (Burris et al., 2002), whereas many assays only see antagonism with aripiprazole (Lawler et al., 1999; Jordan et al., 2007). In addition to standard [35S]GTPγS binding assay conditions, assays were also run in parallel with NMDG replacing sodium, a condition that has been shown to enhance the apparent efficacy of D2 receptor partial agonists in the [35S]GTPγS binding assay (Lin et al., 2006).
Figure 3 shows the effects of the aforementioned compounds on the stimulation of [35S]GTPγS binding in cells expressing the cloned human D2L receptor. DA potently stimulated [35S]GTPγS binding through the D2L isoform (EC50 = 149 nM, NaCl; EC50 = 22.2 nM, NMDG). 3-PPP acted as a partial agonist in assays containing NaCl (Emax = 36.3% relative to DA; Fig. 3A), and efficacy was increased when NMDG was substituted for NaCl (Emax = 74.3% relative to DA; Fig. 3B). In this clone-expressing D2L, there was no detectable agonist activity of aripiprazole, LY379268, or LY354740 regardless of the assay conditions.
In the presence of 1 μM DA to activate D2L-mediated [35S]GTPγS binding, both raclopride and aripiprazole acted as potent full antagonists (Fig. 4) regardless of the buffer conditions. In the presence of NaCl 3-PPP showed some weak antagonist activity (Fig. 4A), which was not seen in the presence of NMDG. Neither LY379268 nor LY354740 showed any effects on DA-stimulated [35S]GTPγS binding in the D2L-expressing cells.
The compounds were also evaluated at the cloned human D2S isoform for any potential agonist activity (Supplemental Fig. 3). In the presence of 100 nM NaCl or 100 mM NMDG, dopamine showed potent stimulation of [35S]GTPγS binding (EC50 = 123 nM, NaCl; EC50 = 35.7 nM, NMDG). The partial D2 receptor agonist, 3-PPP, stimulated [35S]GTPγS binding to approximately 45% compared with dopamine in the NaCl-containing buffer, whereas aripiprazole, LY279368, and LY354740 showed no agonist activity. When NMDG was substituted for NaCl, the efficacy of 3-PPP increased to approximately 88% compared with dopamine. Aripiprazole showed a slight trend toward agonist activity, but neither LY279368 nor LY354740 showed any agonist stimulation even in the presence of NMDG.
In cloned D2S receptors, the antagonists aripiprazole and raclopride both potently inhibited the ability of 1 μM DA to stimulate [35S]GTPγS binding in the presence of either NaCl or NMDG in the buffer. Neither 3-PPP, LY379268, nor LY354740 produced measurable antagonism of DA-mediated stimulation of [35S]GTPγS binding in the buffer containing NMDG (Supplemental Fig. 4). In the NaCl-containing buffer, 3-PPP showed a trend toward inhibition of DA-mediated stimulation of [35S]GTPγS binding, with a moderately good fit by nonlinear regression (R2 = 0.5153). However, the LY379268 and LY354740 data points in the NaCl-containing buffer showed only poor fits by nonlinear regression (R2 = 0.1208 and 0.1914, respectively) with the 95% confidence limits for the curve top estimates overlapping those of the bottom estimates, indicating no significant effect of these compounds on DA-stimulated binding.
In Vivo D2 Receptor Occupancy by LC/MS/MS.
The mGlu2/3 receptor agonists LY379268 and LY354740 monohydrate did not exhibit significant DA D2 occupancy in the rat 60 min after intraperitoneal administration (Table 1). However, aripiprazole occupied D2 receptors in vivo in a dose-dependent fashion after oral gavage (Table 1). High levels of DA D2 receptor occupancy, 55, 89, and 93%, were achieved after aripiprazole at doses of 10, 30, and 60 mg/kg 90 min after oral administration.
Effects on LY379268 and Raclopride on PCP or AMP-Induced Hyperlocomotor Activity in mGlu2/3 Receptor Knockout Mice.
LY379268 (10 mg/kg) and raclopride (3 mg/kg) were tested for their ability to reverse PCP (7.5 mg/kg) and amphetamine (5 mg/kg)-evoked behaviors in mGlu2/3 receptor-deficient mice and their respective wild-type controls. Basal locomotor activity, PCP (7.5 mg/kg), or amphetamine (5 mg/kg)-induced behavioral activation did not show statistically significant differences between the wild-type or mGlu2/3 receptor knockout animals (Fig. 5, top). Whereas LY379268 (10 mg/kg) produced a highly significant reversal of PCP (7.5 mg/kg)-induced ambulations in wild-type mice (P < 0.05), the compound was unable to block PCP-evoked increases in ambulatory activity in mGlu2/3 receptor-deficient mice (Fig. 5, top). The selective dopamine D2/D3 receptor antagonist raclopride attenuated PCP-evoked locomotor activation at 3 mg/kg in both the mGlu2/3 receptor-deficient mice and their wild-type controls (P < 0.05 in both cases). Pretreatment with the mGlu2/3 receptor agonist LY379268 (10 mg/kg) or raclopride (3 mg/kg) significantly blocked amphetamine-induced hyperlocomotion (Fig. 5, bottom) in the wild-type animals (P < 0.05 in both cases). However, LY379268 was without effect on amphetamine-induced hyperactivity in the mGlu2/3 receptor-deficient mice. In contrast, raclopride was able to fully block amphetamine-induced hyperactivation in mGlu2/3 receptor-deficient mice (P < 0.05).
Effects on LY379268 and Raclopride on PCP or AMP-Induced Hyperlocomotor Activity in DA D2 Receptor Knockout Mice.
Basal locomotor activity levels were not significantly different between the D2 receptor knockout mice and the wild-type controls; however, the hyperlocomotor response to both PCP (7.5 mg/kg) and AMP (5 mg/kg) was significantly attenuated in the D2 receptor knockout mice compared with the corresponding wild-type animals (Fig. 6). Pretreatment with the mGlu2/3 receptor agonist LY379268 (10 mg/kg) significantly reversed PCP-evoked increase in locomotor behavior in mice lacking dopamine D2 receptors and their respective wild-type control animals. However, raclopride (3 mg/kg) was only effective at reversing PCP-evoked hyperlocomotor activity in the wild-type control animals. Whereas LY379268 (10 mg/kg) produced a highly significant reversal of AMP (5 mg/kg)-induced ambulations in both the wild-type (P < 0.05) and D2 receptor knockout mice (P < 0.05), the selective D2/D3 receptor antagonist raclopride was only able to block AMP-evoked increases in ambulatory activity in the wild-type animals. Raclopride was without effect on the locomotor-stimulating effects of AMP (5 mg/kg) in D2 receptor-deficient mice.
Effect of LY379268 and DA Agonists on the DA Synthesis Rate in the Striatum and Nucleus Accumbens.
The administration of reserpine (5 mg/kg) to male rats significantly increased the DA synthesis rate, measured in vivo as DOPA accumulation, approximately 2-fold in the striatum (Fig. 7, top) and the nucleus accumbens (Fig. 7, bottom). Consistent with previous reports, the DA agonist quinpirole (1 mg/kg) and the partial DA agonist S-(−)-3-PPP (30 mg/kg) significantly decreased DOPA accumulation in the striatum (79 and 67%, respectively; P < 0.01) and the nucleus accumbens (62 and 53.6%, respectively; P < 0.01). The weak partial DA agonist aripiprazole also significantly decreased DOPA accumulation in both the striatum (31%; P < 0.01) and nucleus accumbens (30%; P < 0.01); however, aripiprazole-mediated reductions in DOPA accumulation were smaller than those seen with S-(−)-3-PPP (30 mg/kg) or quinpirole (1 mg/kg). In marked contrast to the full DA or partial DA agonists, the selective mGlu2/3 receptor agonist LY379268 was without effect on DOPA accumulation in either the striatum or nucleus accumbens at 3 or 10 mg/kg.
Effect of LY379268 or Haloperidol on Partial DA Agonist-Mediated Reductions in the DA Synthesis Rate in the Striatum and Nucleus Accumbens.
Figure 8 shows the effect of haloperidol (0.5 mg/kg) or LY379268 (10 mg/kg) on S-(−)-3-PPP-mediated reductions in DOPA content. As in the previous study, the administration of S-(−)-3-PPP (30 mg/kg) significantly reduced striatal (Fig. 8, top) and nucleus accumbens (Fig. 8, bottom) DOPA accumulation in reserpinized rats. This reduction was significantly blocked by the coadministration of the D2 receptor antagonist haloperidol (P < 0.01). In contrast, S-(−)-3-PPP (30 mg/kg)-mediated reductions in striatal and accumbens DOPA were unaffected by the coadministered of the selective mGlu2/3 receptor agonist LY379268 (10 mg/kg).
Discussion
The goal of the present work was to understand the interaction of LY379268 and LY354740 with dopamine D2 receptors that was reported by Seeman and co-workers (Seeman and Guan, 2008; Seeman et al., 2008). However, regardless of whether we looked at [3H]domperidone binding to native or cloned D2 receptor isoforms or D2-mediated stimulation of [35S]GTPγS binding, we were unable to see any evidence for interaction of either LY379268 or LY354740 with the dopamine D2 receptor. Likewise, a subsequent series of neurochemical and behavioral assays focused on D2 receptor-mediated actions failed to reveal evidence for a direct interaction of LY379268 or LY354740 with D2 receptors in vivo.
Seeman et al. (2008) reported that LY354740 and LY379268 inhibited [3H]domperidone binding to D2 receptors in rat striatal membranes and cloned human D2L membranes in a clearly biphasic fashion with the high-affinity component giving the following: LY354740, Ki = 24 nM in rat striatal membranes; LY379268, Ki = 21 nM in rat striatal membranes; and Ki = 32 nM in human D2L-expressing membranes. In the current work, the endogenous agonist DA revealed biphasic competition curves with [3H]domperidone as the radioligand. Aripiprazole, a very low efficacy D2 receptor partial agonist (Jordan et al., 2007), produced monophasic competition curves, showing high affinity in all membrane preparations. However, neither LY379268 nor LY354740 gave significant inhibition of [3H]domperidone binding in any of these membrane preparations up to the highest concentration tested (10 μM). It is clear that the [3H]domperidone binding assay as run in the current work was capable of identifying both agonists and antagonists at various forms of the D2 receptor. Yet, no significant effects of either LY379268 or LY354740 on [3H]domperidone binding were seen.
Seeman et al. (2008) also showed that LY354740 and LY379268 acted as partial agonists for stimulating human D2L-mediated [35S]GTPγS binding (approximately 22% stimulation relative to DA). As would be expected of a low-efficacy partial agonist, both LY354740 and LY379268 acted as antagonists for inhibiting human D2L-mediated [35S]GTPγS binding that was stimulated by 1 μM DA: LY354740, IC50 = 400 nM, Ki = 43 nm; LY354740, IC50 = 120 nM, Ki = 27 nM (Seeman et al., 2008). In our assays of D2 receptor-mediated stimulation of [35S]GTPγS binding, the partial agonist 3-PPP was able to stimulate [35S]GTPγS binding in both the cloned human D2S- and D2L-expressing membranes. In parallel assays substituting NMDG for NaCl in the assays, a condition reported to enhance partial agonist efficacy in in vitro assays (Lin et al., 2006), an increase in the efficacy of 3-PPP was seen. However, no agonist activity of either LY354740 or LY379268 was observed in either assay condition. It should be pointed out that aripiprazole, which has been reported to be a D2 receptor partial agonist with very low efficacy, also showed no significant agonist activity in these assays. Depending on the sensitivity of the assay, aripiprazole is reported as an antagonist or a low-efficacy partial agonist (Lawler et al., 1999; Burris et al., 2002; Jordan et al., 2007) in the literature. Thus, if LY354740 or LY379268 were very low-efficacy partial agonists, on the order of aripiprazole, they would not have been picked up in the [35S]GTPγS binding assays used in the current work. However, if either LY354740 or LY379268 had significant affinity for the D2 receptor, they should have been picked up in the antagonist form of the [35S]GTPγS binding assays. Both raclopride and aripiprazole produced clear inhibition of DA-stimulated [35S]GTPγS binding, whereas neither LY354740 nor LY379268 produced any significant inhibition out to the highest concentration tested (10 μM).
Having failed to replicate the in vitro findings of Seeman and co-workers (2008), we designed a series of neurochemical and behavioral studies to examine the possible direct interactions with D2 receptors by mGlu2/3 receptor agonists in vivo. We used an LC/MS/MS-based receptor occupancy method (Chernet et al., 2005; Barth et al., 2006) to determine the interaction of compounds with the D2 receptor in the rat striatum. In agreement with previous studies (Natesan et al., 2006), the partial D2 receptor agonist aripiprazole (10–60 mg/kg) dose-dependently occupied D2 receptors in the rat striatum after oral dosing. In marked contrast to aripiprazole, neither LY379269 nor LY354740 demonstrated D2 receptor occupancy in the rat striatum.
Activation of postsynaptic D2 receptors, especially the D2L isoform, is thought to be relevant for locomotor activity (Usiello et al., 2000; Wang et al., 2000; Xu et al., 2002; Lindgren et al., 2003). Mice with a targeted deletion of D2L receptors show reduced levels of locomotor and rearing behaviors and appear to be less sensitive to the locomotor-suppressing and cataleptic effects of D2 receptor antagonists (Usiello et al., 2000; Wang et al., 2000; Xu et al., 2002). We investigated the relative contribution of mGlu2/3 receptors and D2 receptors in the efficacy of LY379268 and raclopride in the PCP and AMP-evoked hyperlocomotor models of psychosis. Administration of PCP or AMP increased locomotor activity in wild-type animals, mGlu2/3 receptor knockout mice, and D2 receptor knockout mice. The mGlu2/3 receptor agonist LY379268 blocked both PCP and AMP-induced hyperactivity in wild-type mice and mice lacking D2 receptors. However, the efficacy of LY379268 against both psychostimulants was lost in mGlu2/3 receptor-deficient animals. As expected, the selective D2/D3 receptor antagonist raclopride was equally effective at reversing PCP and AMP-induced locomotor effects in both wild-type and mGlu2/3 receptor-deficient mice; however, the effects of raclopride were lost in D2 receptor knockout mice. These data are in agreement with previous studies in mice demonstrating that the effects of mGlu2/3 receptor agonists in psychostimulant models of psychosis are dependent on functional mGlu2 receptors (Patil et al., 2007; Fell et al., 2008; Woolley et al., 2008) and not mediated through a direct interaction with D2L receptors in vivo.
To further evaluate the direct effects of mGlu2/3 receptor agonists at D2 receptors, we used a dopamine-depleted (reserpinized) animal preparation, which is a particularly sensitive assay for in vivo D2 agonist effects and is also sensitive to partial DA agonists with low intrinsic activity due to the development of supersensitive receptors (or increases in receptor reserve) after prolonged synaptic depletion of DA (Carlsson, 1983; Hjorth et al., 1988). Under these conditions, the mGlu2/3 receptor agonist LY379268 was without effect on striatal or accumbens DOPA accumulation at either 3 or 10 mg/kg. In contrast, the partial D2 receptor agonists aripiprazole or S-(−)-3-PPP and the full D2 agonist quinpirole reduced both striatal and accumbens DOPA accumulation, consistent with their intrinsic activity at the D2 receptor (Hjorth et al., 1988; Svensson et al., 1991; Iñiguez et al., 2008). In a second experiment, we assessed the ability of LY379268 or the D2/D3 antagonist haloperidol to block partial D2 agonist-mediated reductions in DOPA accumulation in reserpinized rats. Whereas haloperidol significantly blocked the effects of the partial agonist S-(−)-3-PPP on DOPA accumulation in the striatum and nucleus accumbens, the mGlu2/3 receptor agonist LY379268 was without effect on the reductions in DOPA accumulation. Taken together, these data demonstrate that LY379268 does not have obvious direct agonist (either full or partial agonist) or antagonist effects at the D2S receptor in vivo.
The discrepancies between the findings in the current work and those reported by Seeman and co-workers (Seeman and Guan, 2008; Seeman et al., 2008) are puzzling. In a recent meeting abstract, Zysk et al. (2008) could find no interaction of either LY354740 or LY379268 with D2 receptors measured by [3H]raclopride, [3H]4-propyl-9-hydroxynaphthoxazine, or [3H]domperidone binding assays. In addition, they saw no functional activity at D2 receptors of either LY354740 or LY379268 as measured by GTPγS or CellKey assays (MDS Analytical Technologies, Concord, ON, Canada). In a previous study, Seeman et al. (2005) have reported that other glutamatergic compounds, specifically, PCP, ketamine, and dizocilpine (MK-801), show relatively high affinity for D2 receptors when assayed using [3H]domperidone binding and act as full agonists when measured in vitro using a [35S]GTPγS binding assay. However, Jordan et al. (2006) saw no agonist or antagonist activity of phencyclidine, ketamine, or dizocilpine at cloned D2 receptors measured by either a [35S]GTPγS binding assay or a calcium flux assay. The reasons for the lack of concordance between the findings by Seeman and co-workers and these other laboratories remain to be determined.
In conclusion, the in vitro and in vivo findings presented here clearly demonstrate that mGlu2/3 receptor agonists of the class represented by LY379268 and LY354740 do not directly interact with D2L or D2S receptors. These findings are consistent with the results of a recent clinical trial (Patil et al., 2007) in which patients receiving the mGlu2/3 receptor agonist prodrug LY2140023 monohydrate did not develop typical side effects associated with blockade of D2 receptors such as extrapyramidal side effects, dyskinesia, akathisia, or parkinsonian-like effects of hyperprolactinemia. Taken together, our results provide further evidence for a novel, nondopaminergic antipsychotic mechanism of action for mGlu2/3 receptor agonists.
Acknowledgments
We thank Dr. Malcolm J. Low for providing a breeding pair of the B6.Cg-Drd2tm1Low mutant mice developed in his laboratory at Oregon Health and Science University (Portland, OR).
Footnotes
-
This work was supported by Eli Lilly and Company, Indianapolis, IN.
-
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
doi:10.1124/jpet.109.160598
-
↵ The online version of this article (available at http://jpet.aspetjournals.org) contains supplemental material.
-
ABBREVIATIONS:
- mGlu
- metabotropic glutamate
- LY2140023
- (1R,4S,5S,6S)-2-thiabicyclo[3.1.0]hexane-4,6-dicarboxylic acid, 4-[(2S)-2-amino-4-(methylthio)-1-oxobutyl]amino-, 2,2-dioxide monohydrate
- LY404039
- (−)-(1R,4S,5S,6S)-4-amino-2 sulfonylbicyclo [3.1.0] hexane-4,6-dicarboxylic acid
- PCP
- phencyclidine
- AMP
- amphetamine
- LY341495
- 2S-2-amino-2-(1S,2S-2-carboxycycloprop-1-yl)-3-(xant-9-yl) propanoic acid
- LY379268
- (−)-2-oxa-4-aminobicyclo[3.1.0] hexane-4,6-dicarboxylic acid
- LY354740
- 1S,2S,5R,6S-2-aminobicyclo[3.1.0]hexane-2,6-bicarboxylate monohydrate
- DA
- dopamine
- D2L
- long isoform of the dopamine D2 receptor
- D2S
- short isoform of the dopamine D2 receptor
- LC/MS/MS
- liquid chromatography/tandem mass spectrometry
- 3-PPP
- 3-(3-hydroxyphenyl)-N-propylpiperidine
- NMDG
- N-methyl-d-glucamine
- HPLC
- high-performance liquid chromatography
- NSD1015
- m-hydroxybenzylhydrazine
- MK-801
- (5R,10S)-(+)-5-methyl-10,11-dihydro-5H-dibenzo[a,d]cyclohepten-5,10-imine hydrogen.
- Received August 17, 2009.
- Accepted September 14, 2009.
- © 2009 by The American Society for Pharmacology and Experimental Therapeutics