Dopaminergic, serotonergic, and adrenergic receptors are targets for therapeutic actions in schizophrenia. Dopamine D2 receptor partial agonists such as aripiprazole represent a treatment option for patients with this severe disorder. The ineffectiveness of terguride, another D2 receptor partial agonist, in treating schizophrenia was recently attributed to its considerably high intrinsic activity at D2 receptors. In this study, we used functional assays for recombinant D2 receptors and native 5-hydroxytryptamine 2A (5-HT2A), α2C-adrenergic, and histamine H1 receptors to compare the pharmacological properties of terguride and three of its halogenated derivatives (2-chloro-, 2-bromo-, 2-iodoterguride) with those of aripiprazole. Subsequently, we studied the antidopaminergic effects of 2-bromoterguride using amphetamine-induced locomotion (AIL). Its influence on spontaneous behavior was tested in the open field. Extrapyramidal side effect (EPS) liability was evaluated by catalepsy test. In a guanosine 5′-O-(3-[35S]thio)triphosphate ([35S]GTPγS) binding assay, 2-chloro-, 2-bromo-, and 2-iodoterguride produced intrinsic activities at human D2short (hD2S) receptors that were half as high as the intrinsic activity for terguride; aripiprazole lacked agonist activity. 2-Bromoterguride and aripiprazole activated D2S receptor-mediated inhibition of cAMP accumulation to the same extent; intrinsic activity was half as high as that of terguride. All compounds tested behaved as antagonists at human D2long/Gαo (hD2L/Gαo) receptors. Compared with aripiprazole, terguride and its derivatives displayed higher affinity at porcine 5-HT2A receptors and α2C-adrenoceptors and lower affinity at H1 receptors. 2-Bromoterguride inhibited AIL and did not induce catalepsy in rats. Because of its in vitro and in vivo properties, 2-bromoterguride may be a strong candidate for the treatment of schizophrenia with a lower risk to induce EPS.
Standard, or typical, antipsychotic drugs such as haloperidol display prominent antagonist potency at dopamine D2 receptors, with substantial risk of extrapyramidal side effects (EPS) (Baldessarini and Tarazi, 2006). In contrast, atypical antipsychotic drugs, of which clozapine is the prototype, produce fewer EPS than does haloperidol (Leucht et al., 2009). Clozapine, olanzapine, and risperidone have been shown to be more efficacious than typical antipsychotic drugs in the treatment of positive symptoms (delusions, hallucinations) and negative symptoms (affective flattening, alogia, avolition, anhedonia) (Leucht et al., 2009). However, clozapine can induce life-threatening hematologic side effects (agranulocytosis, leukopenia), and clozapine, olanzapine, and risperidone cause considerable weight gain (Kroeze et al., 2003; Roth et al., 2004). Atypical antipsychotics have different pharmacologic profiles at D2 receptors, 5-hydroxytryptamine (serotonin) (5-HT) receptors (5-HT1A, 5-HT2A, 5-HT2C, 5-HT6, 5-HT7), α-adrenoceptors (α1A, α1B, α2C), and histamine H1 receptors (Kroeze et al., 2003; Roth et al., 2004; Meltzer et al., 2012).
The development of aripiprazole, a D2 receptor partial agonist, represents a treatment option for patients with schizophrenia (Burris et al., 2002; Shapiro et al., 2003). As expected for an atypical antipsychotic agent, aripiprazole also exhibited high affinity for multiple serotonin receptors and α-adrenoceptors (Kroeze et al., 2003). Consequently, it was hypothesized that “the balance of partial agonism and antagonism at a multiplicity of receptors is responsible for its efficacy in schizophrenia and related disorders” (Roth et al., 2004). Terguride [1,1-diethyl-3-(6-methyl-8α-ergolinyl)urea] fulfills the criteria of a nonselective drug that interacts with a multiplicity of receptors playing a role in the pathophysiology and treatment of schizophrenia (Roth et al., 2004). Indeed, terguride acts as a D2 receptor partial agonist of appreciable intrinsic activity and as an antagonist at a number of serotonin receptor and α1- and α2-adrenoceptor subtypes (Newman-Tancredi et al., 2002a,b). Terguride was tested in schizophrenia but was found to be clinically efficacious only in reducing negative symptoms (Olbrich and Schanz, 1988, 1991). The relatively high intrinsic activity for terguride at D2S receptors, which was twice as high as that of aripiprazole (Tadori et al., 2005), has recently been suggested to be responsible for the insufficient clinical efficacy of terguride (Natesan et al., 2011). Accordingly, it seems likely that a potential antipsychotic drug should exert high affinity but low intrinsic activity at D2 receptors (Natesan et al., 2011) and, consistent with the pharmacological properties of atypical antipsychotic drugs, affinity for other biogenic amine receptors that play a role in schizophrenia.
There is still a need for better antipsychotics with improved efficacy and reduced adverse effects to treat the various symptoms of schizophrenia. Against this background, we studied the pharmacology of 2-chloroterguride, 2-bromoterguride, and 2-iodoterguride in comparison with terguride and aripiprazole using 1) competition binding assays for recombinant human hD2short (hD2S) and hD2long (hD2L) receptors, stably expressed in Chinese hamster ovary (CHO) cells, to estimate drug affinities; 2) guanosine 5′-O-(3-[35S]thio)triphosphate ([35S]GTPγS) binding assays for hD2S and hD2L receptors to measure G protein activation; 3) hD2S receptor-mediated inhibition of forskolin-stimulated cAMP accumulation; 4) porcine coronary arteries to measure 5-HT–induced contraction mediated by 5-HT2A receptors; 5) porcine pulmonary arteries to measure 5-bromo-6-(2-imidazolin-2-ylamino)quinoxaline (UK-14304)–induced contraction mediated by α2C-adrenoceptors; and 6) porcine pulmonary veins to measure histamine-induced contraction via activation of histamine H1 receptors. In addition, we examined 7) the effect of 2-bromoterguride on spontaneous behavior, 8) its antidopaminergic efficacy using amphetamine-induced locomotion (AIL), and (9) its EPS liability using the catalepsy test in rats. We selected 2-bromoterguride for our in vivo studies because this drug is the dihydro derivative of bromerguride (2-bromolisuride), a drug that has shown atypical antipsychotic properties (Löschmann et al., 1992).
Materials and Methods
For in vitro organ-bath studies, pig hearts and lungs were used, which were obtained from the Lehr- und Versuchsanstalt für Tierzucht und Tierhaltung (Teltow-Ruhlsdorf, Germany). For in vivo studies, a total of 84 male Sprague-Dawley rats (Élevage Janvier, Le Genest Saint Isle, France) aged 10–11 weeks were used. Animals were housed in Makrolon cages (type IV; n = 3–4) under standard conditions (room temperature: 22 ± 2°C; relative humidity: 55 ± 10%) on a 12-hour light/dark schedule (lights on at 6:00 AM). They received free access to standard laboratory chow (ssniff, Soest, Germany) and tap water. All experiments were performed in accordance with the guidelines of the German Animal Protection Law and were approved by the Berlin State Authority (“Landesamt für Gesundheit und Soziales”).
UK-14304 was a gift from Allergan Pharmaceuticals (Westport, County Mayo, Ireland). The following drugs were purchased: d-amphetamine from Berlin-Chemie (Berlin, Germany), aripiprazole from Toronto Research Chemicals (Toronto, ON, Canada), cocaine hydrochloride and histamine dihydrochloride from Merck (Darmstadt, Germany), and 5-hydroxytryptamine creatinine sulfate from Agros Organics (Geel, Belgium). 4-(3-Butoxy-4-methoxybenzyl)imidazolin-2-one, Cremophor EL, 3-isobutyl-1-methylxanthine (IBMX), indomethacin, Nω-nitro-l-arginine methyl ester (l-NAME), mepyramine hydrogen maleate, norepinephrine bitartrate, prazosin hydrochloride, propranolol hydrochloride, and quinpirole hydrochloride were obtained from Sigma-Aldrich (Taufkirchen, Germany). Terguride, 2-chloroterguride, 2-bromoterguride, and 2-iodoterguride were obtained from Alfarma SRO (Cernosice, Czech Republic). (S)-(−)-1,4-Dihydro-2,6-dimethyl-5-nitro-4-[2-(trifluoromethyl)phenyl]-3-pyridinecarboxylic acid methyl ester [(S)-(–)-Bay K 8644] was purchased from Tocris Biosciences (Bristol, UK), and ketanserin tartrate and haloperidol were from Janssen Pharmaceuticals (Beerse, Belgium).
For the in vitro experiments, the drugs were dissolved in distilled water, dimethyl sulfoxide (ergot derivatives; aripiprazole in assays using recombinant receptors), 50% (v/v) ethanol (indomethacin and prazosin), or a mixture of 50% (v/v) ethanol and an equimolar amount of 1 N HCl (aripiprazole in assays using native receptors) to a 1–30 mM stock solution. Stock solutions were stored at −18°C and were freshly diluted in distilled water, phosphate-buffered saline (PBS), or HEPES buffer before the beginning of the experiment. The final concentrations of ethanol and dimethylsulfoxide did not exceed 0.1 and 0.01%, respectively.
For the in vivo experiments, aripiprazole and 2-bromoterguride were made soluble in 15% Cremophor EL, and haloperidol and d-amphetamine were dissolved in 0.9% saline. The appropriate dose of 2-bromoterguride was determined in pilot experiments.
Groups of three rats were treated with 0.1, 0.3, and 1.0 mg/kg 2-bromoterguride to analyze the influence on spontaneous and cataleptic behavior. As 1.0 mg/kg 2-bromoterguride produced a profound sedative effect on spontaneous locomotion, all subsequent experiments were performed with 0.1 and 0.3 mg/kg 2-bromoterguride. Drugs were freshly prepared before being injected intraperitoneally (injection volume: 1.0 ml/kg body weight).
Radioligand Binding Studies on hD2S and hD2L Receptors.
Competition binding experiments were performed as described previously elsewhere (Hübner et al., 2000) using membrane preparations from CHO cells stably expressing either hD2S or hD2L receptors (Hayes et al., 1992). Assays were run with membranes at protein concentration per well of 2 µg/ml for D2S and 4 µg/ml for D2L and [3H]spiperone (specific activity 84 Ci/mol; PerkinElmer, Rodgau, Germany) at final concentrations of 0.1 and 0.2 nM for D2S and D2L receptors, respectively. The KD values for D2S and D2L receptors were 0.040 and 0.12 nM; the Bmax values were 3470 fmol/mg and 1310 fmol/mg, respectively. Membranes, radioligand, and test compound were incubated for 60 minutes at 37°C in binding buffer (50 mM Tris-HCl, 1 mM EDTA, 5 mM MgCl2, 100 µg/ml bacitracin, and 5 µg/ml soybean trypsin inhibitor at pH 7.4). Incubations were terminated by rapid filtration through Whatman GF/C filters presoaked with 0.3% polyethylenimine. The filters were rinsed 5 times with ice-cold Tris-NaCl buffer (50 mM Tris-HCl, 120 mM NaCl at pH 7.4). After 3 hours of drying at 60°C, the filters were counted for radioactivity using scintillation spectrometry. Nonspecific binding was determined in the presence of 10 µM haloperidol. Specific binding was about 90% of the total binding. Protein concentration was determined by the method of Lowry using bovine serum albumin as a standard.
GTPγS Binding Studies on hD2S Receptors.
The agonist potencies of the compounds were investigated in a [35S]GTPγS assay using membranes of CHO cells stably expressing hD2S receptors (Görnemann et al., 2008). Homogenates of membranes (Bmax = 3300–4800 fmol/mg) were diluted in HEPES buffer (20 mM HEPES, 10 mM MgCl2, 100 mM NaCl, 40 µg/ml saponin; pH 7.4) and incubated at 37°C with 1 µM GDP (HEPES buffer), the reference agonist quinpirole, and the test compound (ergot derivative) applying 10 different concentrations (0.01–1000 nM) as triplicates at a final volume of 200 µl in 96-well microplates.
In separate experiments, quinpirole (0.1 nM to 1 mM), aripiprazole (1 pM to 10 µM), and quinpirole (5 µM) plus aripiprazole (1 pM to 10 µM) were used. After 30 minutes, 0.1 nM [35S]GTPγS (specific activity 1250 Ci/mmol; PerkinElmer Life and Analytical Sciences, Waltham, MA) were added, and the incubation was continued for a further 30 minutes. The experiment was terminated by rapid filtration through GF/B filters using an automated cell harvester; the filters were washed five times with ice-cold washing buffer (140 mM NaCl, 10 mM KCl, 1.5 mM KH2PO4, 8 mM Na2HPO4, pH 7.4), dried at 60°C for 3 hours, and the trapped radioactivity was counted in a microplate scintillation counter. [35S]GTPγS binding data from each individual experiment were expressed as a percentage of the full response to quinpirole.
GTPγS Binding Studies on hD2L/Gαo Receptors.
Membrane preparations of transiently transfected human embryonic kidney 293 (HEK293) cells that expressed hD2L receptors coupled to Gαo were used in a [35S]GTPγS assay, as recently described elsewhere (Tschammer et al., 2011). Receptor expression determined in saturation experiments was 1640–1850 fmol/mg protein. The assay was performed in 96-well plates at a final volume of 200 µl of incubation buffer (20 mM HEPES, 10 mM MgCl2, 100 mM NaCl, and 70 mg/l saponin; pH 7.4). Membranes (30 µg/ml membrane protein), quinpirole (0.1 nM to 1 mM), the test compound (0.01 nM to 100 µM), quinpirole (1 µM) plus the test compound (0.1 pM to 100 µM), and 10 µM GDP were preincubated in the absence of [35S]GTPγS for 30 minutes at 37°C. In additional experiments, quinpirole (0.1 nM to 1 mM), aripiprazole (1 pM to 10 µM), and quinpirole (5 µM) plus aripiprazole (1 pM to 10 µM) were used.
After the addition of 0.10 nM [35S]GTPγS, membranes were incubated for an additional 30 minutes at 37°C. Incubation was terminated by filtration through Whatman GF/B filters soaked with ice-cold PBS. The filter-bound radioactivity was measured as described previously. Three to eight experiments per compound were performed with each concentration in triplicate. [35S]GTPγS binding data from each individual experiment were expressed as a percentage of the full response to quinpirole.
Inhibition of cAMP Accumulation.
Inhibition of forskolin-stimulated cAMP accumulation mediated via hD2S receptors was measured using the bioluminescence-based cAMP-Glo assay (Promega, Madison, WI) according to the manufacturer’s instructions. In brief, CHO cells stably expressing hD2S receptors were seeded into a half-area 96-well plate (5000 cells/well) 24 hours before the experiment. Cells were washed with PBS (pH 7.4) to remove traces of serum and then incubated with the test compound (10 pM to 100 µM) dissolved in serum-free medium containing forskolin (20 µM), 3-isobutyl-1-methylxanthine (IBMX; 500 µM), and 4-(3-butoxy-4-methoxybenzyl)imidazolin-2-one (Ro 20-1724; 100 µM). After 15 minutes of incubation at 25°C, cells were lysed with cAMP-Glo lysis buffer (for 15 minutes), followed by the kinase reaction performed by the addition of cAMP-Glo reaction buffer containing protein kinase A (20 minutes). The final step was the addition of an equal volume of kinase-Glo reagent and the measurement of bioluminescence on a Victor3V microplate reader (PerkinElmer Life and Analytical Sciences).
Four to seven experiments per test compound were performed with each concentration in triplicate. The effects on cAMP accumulation were expressed as a percentage of the full response to quinpirole.
Tissue Bath Studies.
Pig hearts and lungs were placed in ice-cold oxygenated Krebs-Henseleit solution (KHS; 95:5% O2/CO2) of the following composition: 118 mM NaCl, 4.7 mM KCl, 1.6 mM CaCl2 (2.5 mM for pulmonary arteries and veins), 1.2 mM MgSO4, 1.2 mM KH2PO4, 25 mM NaHCO3, and 10 mM d-glucose (pH 7.4). Coronary arteries (left anterior descending and left circumflex) were removed from the hearts, and small branches of pulmonary arteries and veins were dissected from the lungs. The tissues were cleaned of fat and adhering tissue. The pulmonary arteries and veins were stored overnight at 4°C in previously gassed KHS containing indomethacin (30 µM). Preliminary experiments have shown that tissue storing overnight does not impair the contractility of the smooth muscle.
On the next day, the vessels were cut into rings (3–4 mm long and 2–3 mm inner diameter). Vascular rings were horizontally suspended between two L-shaped stainless steel hooks (300 µm diameter). The tissues were mounted in water-jacketed 20-ml organ chambers and constantly exposed to oxygenated KHS (pH 7.4, 37°C). Preparations were connected to an isometric force transducer (FMI TIM-1020; FMI Föhr Medical Instruments, Seeheim-Jugenheim, Germany) attached to a TSE 4711 transducer coupler (TSE Systems, Bad Homburg, Germany) and a Siemens C 1016 compensograph (Siemens AG, Erlangen, Germany) for the continuous recording of changes in tension.
Porcine Coronary Arteries (Functional 5-HT2A Receptor Assay).
Resting tension was adjusted to 20 mN at the beginning of the experiment. The tissues were stabilized for 60 minutes with replacement of the bathing medium after 30 minutes. During the ensuing equilibration period (160 minutes), the vessels were stimulated twice with KCl (50 mM). The rings were rinsed with KHS for 5 minutes to wash out KCl. l-NAME (100 µM) was added to each tissue bath to inhibit endothelial nitric-oxide synthase. After an additional 30 minutes, the rings were stimulated with 5-HT (1 µM). The rings were rinsed with KHS for 10 minutes to wash out the 5-HT.
A single cumulative concentration-response curve (CRC) to 5-HT was constructed in the absence or presence of antagonist. Antagonists were added to the bathing medium 60 minutes before the construction of the 5-HT curve. Contractile effects were expressed as a percentage of the 5-HT–induced precontraction. All experiments were performed in the continuous presence of prazosin (0.1 µM), cocaine (10 µM), and indomethacin (5 µM) to block α1-adrenoceptors and to inhibit neuronal uptake of 5-HT and vascular eicosanoid production by cyclooxygenase, respectively.
Porcine Pulmonary Arteries (Functional α2C-Adrenoceptor Assay).
Arterial rings were stabilized for 90 minutes with bath fluid replacements every 30 minutes. During a subsequent equilibration period of 180 minutes, vessels were stimulated 4 times with 1 µM norepinephrine (NE) with 6 minutes washings after each contractile challenge. This procedure was considered to yield stable and reproducible contractions. The tension was repeatedly readjusted to 20 mN and remained unchanged after the fourth NE (1 µM) stimulation. A single CRC to UK-14304 was constructed on each arterial ring moderately precontracted with (S)-(–)-Bay K 8644 (L-type Ca2+ channel activator) in the absence or presence of antagonist, as described previously elsewhere (Jantschak and Pertz, 2012). Antagonists were added 2 hours before the construction of the CRC to UK-14304.
Contractile effects were expressed as a percentage of the fourth NE-induced contraction. Cocaine (10 µM), prazosin (0.03 µM), and propranolol (1 µM) were continuously present in the bath fluid to block neuronal uptake of NE and α1- and β-adrenoceptors.
Porcine Pulmonary Veins (Functional Histamine H1 Receptor Assay).
Resting tension was adjusted to 10 mN at the beginning of the experiment. The tissues were stabilized for 60 minutes with replacement of the bathing medium after 30 minutes. During the ensuing equilibration period (100 minutes), the vessels were stimulated 3 times with histamine (3 µM). The rings were rinsed with KHS for 8 minutes to wash out the histamine. A single cumulative CRC to histamine was constructed on each venous ring in the absence or presence of antagonist. Antagonists were added 30 minutes before the construction of the histamine curve.
Contractile effects were expressed as a percentage of the third histamine-induced contraction. Cocaine (10 µM) and propranolol (1 µM) were continuously present in the bath fluid to block neuronal histamine uptake and β-adrenoceptors.
In Vivo Experimental Procedure.
Rats were handled daily for 1 week before testing. All experiments were conducted during the light phase between 8:00 AM and 1:00 PM in a noise-proof chamber with a light intensity of 100 lux in the center of the floor. Test naive animals were used to examine drug-induced effects on spontaneous behavior (n = 30) and AIL (n = 48), respectively. For the catalepsy test, 38 of these rats were tested a second time at random with at least 1 week in between to allow clearance of the drugs.
Rats were tested at two different designs: placing both forepaws on a 9 cm block (block test) or placing the rat on an inclined grid (60°; grid test). Latency to voluntarily remove one paw from the block or a directed movement on the grid was considered to be the end of the test (descent latency). The cutoff time for both tests was set to 150 seconds. The animals were first examined 3 times on the block immediately followed by three test sessions on the inclined grid. Each test was performed at 15, 30, 60, 90, 120, and 150 minutes after the injection of 2-bromoterguride (0.1 and 0.3 mg/kg), haloperidol (0.5 mg/kg), or vehicle (15% Cremophor EL). The average descent latency was recorded for each rat at each time point. Catalepsy was scored from 0 to 5, according to the estimated time (in minutes): 0 = 0.00–0.08; 1 = 0.09–0.35; 2 = 0.36–0.80; 3 = 0.81–1.42; 4 = 1.43–2.24; 5 ≥ 2.25 (Ahlenius and Hillegaart, 1986). An animal was considered cataleptic with a score ≥2 (Wadenberg et al., 2001).
AIL and Spontaneous Behavior in the Open Field.
The effect of 2-bromoterguride on AIL was evaluated in two open-field boxes (50 × 50 × 32 cm; gray PVC; in-house production). The distance traveled (in centimeters) was recorded and analyzed by a computer-based system (VideoMot2; TSE Systems, Bad Homburg, Germany). Animals were administered with either 2-bromoterguride (0.1 and 0.3 mg/kg), aripiprazole (3.0 mg/kg), or vehicle (15% Cremophor EL) and were placed into the open field for a 30-minute habituation period. Subsequently, amphetamine (1.5 mg/kg) or vehicle (0.9% saline) was injected, and the animals were immediately returned to the open field for a 120-minute recording period.
For testing spontaneous behavior of 2-bromoterguride in the open field, the drug was injected, and then the rat was placed back into the home cage. After 30 minutes, the rats were placed in the open-field arenas, and their behavior (locomotor activity [centimeters], rearings [counts]) was measured for 30 minutes.
Data Presentation and Analysis.
Data are presented as mean values ± S.E.M. for n individual experiments or n animals. Data were analyzed and presented using the nonlinear curve-fitting program GraphPad Prism 5.0 (GraphPad Software, Inc., San Diego, CA) or SigmaPlot 11 (Systat Software, Erkrath, Germany), which allowed estimation of IC50, Ki (according to the Cheng-Prusoff equation), EC50, and Emax values (percentage of the maximal response to a reference compound). Ki and EC50 are presented as pKi (negative logarithm of Ki) and pEC50 (negative logarithm of EC50), respectively. Antagonist affinities (pKB = −logKB) for inhibition of quinpirole-stimulated [35S]GTPγS binding were calculated according to Lazareno and Birdsall (1993): KB = IC50/[2 + ([Af]/EC50)nH]1/nH − 1, where IC50 is the inhibitory concentration50 of the antagonist, [Af] is the fixed quinpirole concentration, EC50 is the effective concentration50 of quinpirole alone, and nH is the Hill coefficient of the quinpirole stimulation isotherm.
Antagonist affinities (apparent pA2) for inhibition of UK-14304, 5-HT, and histamine-induced contractions were calculated from a single concentration of antagonist using the following equation: pA2 = −log[B] + log(r − 1), where [B] is the molar concentration of the antagonist, and r the concentration ratio of agonist EC50 determined in the presence and absence of the antagonist.
For noncompetitive antagonists, a pD′2 value was calculated according to van Rossum (1963). pD′2 was defined as the negative logarithm of the molar concentration of antagonist that caused a 50% depression of the maximal response to the agonist: pD′2 = −log[B] + log(Emax/Emax* − 1), where Emax is the maximal response to the agonist in the absence of antagonist, and Emax* is the maximal response to the agonist in the presence of an antagonist. When a partial agonist was tested as an antagonist (terguride against histamine), a pKP value was calculated by comparing equiactive molar concentrations of the full agonist A (histamine) in the absence and presence of the partial agonist P (terguride) according to the equation: [A] = m · [A]* + b with m = 1/[1 + (1 − εP/εA) · [P]/KP], where [A] is the molar concentration of A in the absence of P, [A]* is the molar concentration of A in the presence of P, m is the slope of a weighted regression line of [A] versus [A]*, b is the ordinate intercept, and [P] is the molar concentration of P. If εP < < εA, pKP = −logKP can be calculated from: log[(1/m) − 1] = log[P] −logKP (Marano and Kaumann, 1976). Student's t test was used to assess the differences between two mean values.
Spontaneous behavior was analyzed using a one-way analysis of variance (ANOVA) followed by the Holm-Sidak method. Treatment effects on AIL were analyzed by two-way repeated measures ANOVA followed by the Holm-Sidak method with respect to amphetamine treatment. Catalepsy data were evaluated by Kruskal-Wallis one-way ANOVA on ranks followed by Dunn’s test. For reasons of comparability, all data are presented as mean ± S.E.M. P < 0.05 was considered statistically significant.
Dopamine hD2S and hD2L Receptor Binding.
Terguride and its 2-halogenated derivatives showed subnanomolar affinities at hD2S and hD2L receptors stably expressed in CHO cells (Table 1). Affinities of 2-chloroterguride, 2-bromoterguride, and 2-iodoterguride were not different, neither at D2S nor at D2L receptors (P > 0.05). 2-Chloroterguride and 2-bromoterguride had a slight but significant preference for D2S over D2L receptors (P < 0.05).
Effects on GTPγS Binding at hD2S Receptors Stably Expressed in CHO Cells.
Terguride and its 2-halogenated derivatives behaved as low-efficacy partial agonists compared with the full agonist quinpirole as measured by hD2S receptor-mediated incorporation of [35S]GTPγS (Fig. 1A; Table 2). Terguride showed the highest intrinsic activity, which was 2- to 3-fold higher than the efficacies of 2-chloroterguride, 2-bromoterguride, or 2-iodoterguride. Aripiprazole showed no agonist activity but behaved as an antagonist of quinpirole-induced [35S]GTPγS incorporation (Fig. 1B). The antagonist potency for aripiprazole (pKB 9.70 ± 0.18; n = 7) was in the same range as the affinity (pKi 9.23) reported by Lawler et al. (1999).
Effects on cAMP Accumulation at hD2S Receptors Stably Expressed in CHO Cells.
Selected compounds (quinpirole, terguride, 2-bromoterguride, and aripiprazole) were tested for their ability to inhibit cAMP production. Terguride acted as a partial agonist with appreciable intrinsic activity (0.57 versus 1.0 for quinpirole). In contrast, intrinsic activity for 2-bromoterguride was half as high (0.28) as that for terguride. Intrinsic activity for 2-bromoterguride and aripiprazole was the same (Fig. 2; Table 2).
Effects on GTPγS Binding at hD2L/Gαo Receptors Stably Expressed in HEK293 Cells.
Dopamine D2L receptors are coupled to their effectors by Gi and Go proteins (Cordeaux et al., 2001). We measured hD2L receptor-mediated incorporation of [35S]GTPγS in the presence of Gαo. Using this system, terguride and its 2-halogenated derivatives were devoid of agonist activity but behaved as antagonists of quinpirole-induced [35S]GTPγS incorporation (Fig. 3; Table 2). Antagonist potencies (pKB) estimated in these experiments were in good agreement with affinities (pKi) from our binding studies. Aripiprazole had the same pharmacological properties as the ergot derivatives, with no agonist activity but inhibition of quinpirole-induced [35S]GTPγS incorporation (Fig. 3E). Antagonist potency (pKB 9.32 ± 0.23; n = 6, see Table 2) for aripiprazole was in line with the affinity (pKi 9.28) reported in the literature (Lawler et al., 1999).
Effects at Smooth Muscle 5-HT2A Receptors in Porcine Coronary Arteries.
The tissue is endowed with contractile 5-HT2A receptors (Cushing and Cohen, 1993). 5-HT (3 nM to 10 µM) induced a concentration-dependent contraction in coronary arterial rings (pEC50 6.77 ± 0.03; n = 14) that was surmountably blocked by the 5-HT2A receptor antagonist ketanserin (10 nM; apparent pA2 8.99 ± 0.02; n = 6; data not shown). 5-HT–induced contractions were antagonized by terguride and its 2-halogenated derivatives (Fig. 4A; Table 3). Terguride (1 nM) produced only a slight dextral shift of the CRC to 5-HT (concentration ratio r < 2, n = 6) but induced a marked depression of the maximal response from 122 ± 8 to 44 ± 5% (n = 6). The noncompetitive antagonist parameter pD′2 for terguride is shown in Table 3.
2-Iodo-, 2-bromo-, and 2-chloroterguride (3 nM each) induced dextral shifts of the 5-HT curve with concentration ratios r > 2 (apparent pA2 values, see Table 3) and reduced Emax from 122 ± 8 to 62 ± 9, 50 ± 3, and 38 ± 5%, respectively (Fig. 4A). In contrast to the insurmountable antagonist properties of 2-chloro-, 2-bromo-, and 2-iodoterguride, aripiprazole inhibited the contractile 5-HT effect in a surmountable manner (Fig. 4B), albeit with 40- to 135-fold lower antagonist potency than the ergolines. The apparent pA2 of 7.12 for 0.5 µM aripiprazole (see Table 3) was similar to the pKi (7.46) at cloned human 5-HT2A receptors (Shapiro et al., 2003).
Effects at Smooth Muscle α2C-Adrenoceptors in Porcine Pulmonary Arteries.
This tissue is endowed with contractile α2C-adrenoceptors (Jantschak and Pertz, 2012). The α2-adrenoceptor agonist UK-14304 (0.3 nM to 3 µM) induced a concentration-dependent contraction in pulmonary arterial rings after moderate precontraction with (S)-(–)-Bay K 8644 (L-type Ca2+ channel activator). The pEC50 for UK-14304 was 7.70 ± 0.09 (n = 10). Terguride and its 2-halogenated derivatives were potent antagonists of the contractile UK-14304 response (Fig. 5A; Table 3). 2-Chloroterguride, 2-bromoterguride, and 2-iodoterguride showed higher antagonist potency compared with the antagonist effect of the parent compound terguride (Table 3). The effects of 2-chloroterguride, 2-bromoterguride, and 2-iodoterguride were nearly identical. Aripiprazole (0.5 µM) also blocked UK-14304-induced contractions, albeit with 400- to 1800-fold lower affinity than terguride and its derivatives (Fig. 5B, Table 3). The apparent pA2 of 7.20 for aripiprazole (see Table 3) was in line with the pKi (7.42) at cloned human α2C-adrenoceptors (Shapiro et al., 2003).
Effects at Smooth Muscle Histamine H1 Receptors in Porcine Pulmonary Veins.
Histamine (10 nM to 30 µM) induced a concentration-dependent contractile response in pulmonary venous rings. The pEC50 was 6.58 ± 0.05 (n = 8). The effects of histamine were surmountably blocked by mepyramine (H1 receptor antagonist; Fig. 6D). The apparent pA2 of 8.70 ± 0.08 (n = 5) for mepyramine argues for an involvement of H1 receptors in the contractile response to histamine in this tissue. Terguride (1 µM) behaved as a low-efficacy partial agonist; the slight contractile response to terguride (5 ± 1%) was abolished by mepyramine (1 µM). As expected for a partial agonist, terguride (1 µM) inhibited the contractile histamine response (Fig. 6A). The halogenated derivatives of terguride were devoid of partial agonist effects and antagonized the histamine-induced contraction (Fig. 6B; Table 3). Aripiprazole also inhibited the histamine response but with an affinity that was 20- to 80-fold higher than that of terguride and its derivatives (Fig. 6C; Table 3). The apparent pA2 of 7.82 for aripiprazole (see Table 3) was in line with the pKi (7.60) at cloned human H1 receptors (Shapiro et al., 2003).
2-Bromoterguride did not induce catalepsy in the block or grid test. All estimated mean catalepsy scores of 2-bromoterguride were <2, the level at which catalepsy is considered to begin (Wadenberg et al., 2001) (Fig. 7). However, haloperidol (0.5 mg/kg) induced cataleptic behavior both in block and grid test and prolonged descent latency compared with vehicle and/or 2-bromoterguride. The results were: block test: haloperidol versus 0.1 mg/kg 2-bromoterguride at 60 minutes (H = 9.8, P = 0.02, df = 3); haloperidol versus vehicle and 0.1 mg/kg 2-bromoterguride at 90 minutes (H = 23.3, P < 0.001, df = 3), 120 minutes (H = 20.1, P < 0.001, df = 3), and 150 minutes (H = 15.6, P < 0.001, df = 3); grid test: haloperidol versus vehicle, 0.1 and 0.3 mg/kg 2-bromoterguride at 30 minutes (H = 16.6, P < 0.001, df = 3); haloperidol versus vehicle and 0.1 mg/kg 2-bromoterguride at 60 minutes (H = 24.3, P < 0.001, df = 3), 90 minutes (H = 25.8, P < 0.001, df = 3), 120 minutes (H = 23.0, P < 0.001, df = 3), and 150 minutes (H = 19.9, P < 0.001, df = 3) (Fig. 7).
Effects of 2-Bromoterguride on Spontaneous Behavior in the Open Field.
Locomotor activity was affected by 2-bromoterguride treatment. Both 2-bromoterguride doses, 0.1 and 0.3 mg/kg, decreased spontaneous locomotion (F2,27 = 25.2; P < 0.001; Fig. 8) and rearings (F2,27 = 19.0; P < 0.001; vehicle: 131 ± 24; 0.1 mg/kg 2-bromoterguride: 86 ± 13; 0.3 mg/kg 2-bromoterguride: 46 ± 14).
Effects of 2-Bromoterguride on AIL.
Amphetamine-treated rats showed hyperlocomotion compared with the vehicle group. 2-Bromoterguride (0.1 and 0.3 mg/kg) and aripiprazole inhibited amphetamine-induced locomotion. A statistically significant main effect of treatment (F4,473 = 16.0; P < 0.001), time (F11,473 = 24.9; P < 0.001), and interaction of the factors (F44,473 = 4.3; P < 0.001) was found (Fig. 9).
A primary goal of our study was to examine the pharmacologic properties of 2-halogenated derivatives of terguride at dopaminergic (D2), serotonergic (5-HT2A), adrenergic (α2C), and histaminergic (H1) receptors because these receptors play a role in the pathophysiology and treatment of schizophrenia (Roth et al., 2004). The effects of terguride, the parent drug of these compounds, at D2, 5-HT2A, α2C, and H1 receptors were fully in agreement with those reported by others (Newman-Tancredi et al., 2002a,b; Pertz et al., 2006; Kekewska et al., 2012).
Striatal dopaminergic hyperfunction plays a causative role in schizophrenia (Howes and Kapur, 2009; Simpson et al., 2010). At therapeutic doses, antipsychotic drugs block striatal D2 receptors. There are two isoforms of the D2 receptor: the short one (D2S) has the function of a somatodendritic autoreceptor of central dopaminergic neurons, and the long one (D2L) plays the role of a postsynaptic receptor in vivo (Khan et al., 1998). D2 receptor partial agonists such as aripiprazole, preclamol, 7-[3-(4-[2,3-dimethylphenyl]piperazinyl)propoxy]-2(1H)-quinolinone (OPC-4392), and terguride can show large variations in intrinsic activity and potency at both isoforms. Depending on the assay used (e.g., adenylyl cyclase or GTPγS binding), the differential activation of signaling pathways (a phenomenon termed functional selectivity), the level of receptor expression, the host cell type, and the drug concentration, these compounds demonstrated partial agonist or antagonist properties at hD2L and hD2S receptors (Lawler et al., 1999; Burris et al., 2002; Newman-Tancredi et al., 2002a; Shapiro et al., 2003; Tadori et al., 2005).
In our studies, aripiprazole was devoid of agonist activity and behaved as an antagonist at hD2S receptors in the GTPγS binding assay (Fig. 1B). However, aripiprazole was a partial agonist of moderate intrinsic activity at hD2S receptors in the adenylyl cyclase assay (i.a. 0.29; see Fig. 2; Table 3). It should be noted that the measurement of cAMP levels is focused on a strongly amplified event downstream of the G protein–coupled receptor, whereas the measurement of GTPγS reflects a comparably lower amplification event close to the receptor (Strange, 2010); as a consequence, a partial agonist in the adenylyl cyclase assay may behave as an antagonist in the GTPγS binding assay. Our findings that aripiprazole caused no agonist effect at hD2L receptors expressed in CHO cells in the GTPγS binding assay is totally in line with the data reported by Shapiro et al. (2003).
The observation in our study that terguride behaved as a silent hD2L receptor antagonist and a hD2S receptor partial agonist is consistent with the pharmacologic profile of this drug in GTPγS binding studies of Newman-Tancredi et al. (2002a). It has been hypothesized that the inefficacy for the treatment of schizophrenia of D2 receptor partial agonists such as terguride, preclamol, or OPC-4392 (Benkert et al., 1995) may be based upon their appreciably high intrinsic activity at D2S receptors (Natesan et al., 2011). In our GTPγS binding studies, we could show that halogenation of terguride in the 2-position reduced the intrinsic activity by 46–68% compared with that of the parent drug terguride (Fig. 1A; Table 2). Moreover, 2-bromoterguride showed an intrinsic activity that was half as high as that of terguride at hD2S receptors in the adenylyl cyclase assay. Most interestingly, 2-bromoterguride mimicked aripiprazole in this assay; intrinsic activity was the same for both drugs (Fig. 2; Table 2).
These data are in agreement with the hypothesis of a reciprocal relationship between intrinsic activity and antipsychotic efficacy of partial agonists in schizophrenia (Natesan et al., 2011). In addition to the effects at hD2S receptors, terguride and its derivatives behaved as potent antagonists at hD2L/Gαo receptors. The hD2L receptor preferentially couples to Gαo over each of the three other Gαi subtypes: Gαi1, Gαi2, and Gαi3 (Gazi et al., 2003). Postsynaptic D2L receptor blockade is the essential antipsychotic mechanism downstream of the primary dopaminergic abnormality (Howes and Kapur, 2009).
A relatively high affinity for 5-HT2A compared with D2 receptors, the so-called 5-HT2A/D2 ratio, has been proposed as a significant but not exclusive criterion for differentiating typical and atypical antipsychotic drugs; a higher affinity for 5-HT2A versus D2 receptors may improve negative symptoms and may have an impact on EPS liability (Meltzer et al., 2012). In our experiments, aripiprazole had a more than 150-fold lower affinity for 5-HT2A receptors compared with D2L/2S receptors. In contrast, 2-halogenated derivatives of terguride possessed a high affinity for both receptors with 4- to 25-fold greater affinity for D2L/2S over 5-HT2A receptors.
In addition to D2 and 5-HT2A receptors, α2-adrenoceptors are targets in the treatment of schizophrenia (Lindström, 2000). In contrast to α2A-adrenoceptors, which are widely distributed in the brain, α2C-adrenoceptors are predominantly located in the striatum (Scheinin et al., 1994; Uhlén et al., 1997; Fagerholm et al., 2008). Animal experiments and clinical studies have demonstrated that the therapeutic effect of an antipsychotic drug can be enhanced by the addition of idazoxan or mirtazapine, drugs that block α2C- and α2A-adrenoceptors (Litman et al., 1996; Hertel et al., 1999; Wadenberg et al., 2007; Joffe et al., 2009; Marcus et al., 2010). Antipsychotics such as clozapine, olanzapine, risperidone, and quetiapine exhibited higher affinities for α2C- over α2A-adrenoceptors (Kalkman and Loetscher, 2003; Kroeze et al., 2003).
In rodents, 5 mg/kg clozapine, a dose that showed a pronounced inhibitory effect on conditioned avoidance response, produced ∼45% D2, ∼65% α2A-, and ∼95% α2C-adrenoceptor occupancy (Marcus et al., 2005). This is in line with observations made by Kalkman and Loetscher (2003) who argued that it is not the affinity of an antipsychotic drug for α2C-adrenoceptors per se but the relative affinity α2C/D2 that is important for efficacy in schizophrenia. According to these authors, a ratio α2C/D2 ≥ 1 suggests that α2C-adrenoceptor blockade may contribute to the therapeutic profile of an antipsychotic drug. This is the case for atypical antipsychotics such as clozapine, olanzapine, risperidone, or quetiapine (Kroeze et al., 2003). Our studies demonstrate that 2-chloro-, 2-bromo-, and 2-iodoterguride block α2C-adrenoceptors with subnanomolar affinities (apparent pA2 10.2–10.5) and show ratios of α2C/D2 ≥ 1, thus suggesting advantageous therapeutic properties. Aripiprazole exhibited much lower α2C-adrenoceptor affinity (pA2 7.2) and therefore does not fulfill the criterion of Kalkman and Loetscher (2003).
Affinities for histamine H1 receptors are positively correlated with weight gain among typical and atypical antipsychotic drugs (Kroeze et al., 2003). Accordingly, it has been recommended that novel antipsychotic drugs should possess a low affinity for H1 receptors. In our studies, the affinity of aripiprazole for H1 receptors was higher than that of terguride and its derivatives.
For our in vivo experiments, we used preclinical tests with high predictive validity, which offer valuable clues to the effect of 2-bromoterguride on central dopaminergic mechanisms. In rodents, cataleptic behavior is evident with most typical antipsychotics and is less obvious with atypical antipsychotics, and it is regarded as a predictor of EPS liability (Arnt and Skarsfeldt, 1998; Wadenberg et al., 2001; Natesan et al., 2006). Consistent with other D2 receptor partial agonists (Semba et al., 1995; Natesan et al., 2011), 2-bromoterguride failed to produce catalepsy. Although 2-bromoterguride increased descent latency in the block and grid tests, a state of immobility was not attained. In contrast, catalepsy was induced by the typical antipsychotic drug haloperidol (positive control). In addition to the catalepsy test, we studied the effect of 2-bromoterguride on spontaneous behavior in the open field. 2-Bromoterguride decreased locomotor activity and rearings of rats. This is in agreement with other studies showing an attenuating effect of D2 receptor partial agonists on locomotion and rearings (Svensson et al., 1991; Nordquist et al., 2008). Hence, the possibility that some effects of 2-bromoterguride observed in AIL are affected by these motor effects cannot be excluded. However, the low affinity of 2-bromoterguride for the H1 receptor argues against an involvement of this receptor in the sedative effect.
Suppression of AIL is a widely used animal model for the detection of potential antipsychotic drugs (Koch, 2006). Psychostimulant-induced hyperlocomotion reflects primarily an activation of D2 receptors mainly in the nucleus accumbens (Porsolt et al., 2010), a response that should readily be antagonized by typical and atypical antipsychotics by competing for postsynaptic D2 receptors in this part of the mesolimbic DA system (Geyer and Ellenbroek, 2003). Our observation that 2-bromoterguride (0.1 and 0.3 mg/kg) inhibited AIL clearly suggests the antidopaminergic properties of this drug. Interestingly, a comparable inhibition of AIL was only achieved with a 10-fold higher dose of aripiprazole (3 mg/kg). Furthermore, inhibition of AIL confirms the in vitro results of our study, in which 2-bromoterguride behaved as a potent antagonist at hD2L/Gαo receptors, an effect that is essential for antipsychotic efficacy (Howes and Kapur, 2009).
In conclusion, our in vitro studies show that 2-halogenated derivatives of terguride, particularly 2-bromoterguride, mimic aripiprazole as partial agonists at hD2S receptors and potent antagonists at hD2L receptors. However, antagonist activities at 5-HT2A receptors and α2C-adrenoceptors are higher, whereas affinities for H1 receptors are lower compared with those of aripiprazole. This may contribute to an improved antipsychotic profile of 2-halogenated derivatives of terguride. In vivo, 2-bromoterguride shows antidopaminergic activity in the AIL test and low EPS liability. Further research is necessary to substantiate the antipsychotic effects of 2-bromoterguride. These studies should include further convergent preclinical tests such as the inhibition of conditioned avoidance response and measurement of striatal Fos induction as a potential molecular marker for antipsychotic activity (Natesan et al., 2011).
The authors thank Dr. T. Paulke and M. Uwarow of the Lehr- und Versuchsanstalt für Tierzucht und Tierhaltung (Teltow-Ruhlsdorf, Germany) for providing pig hearts and lungs for the studies, and Sabine Jacobs and Aurica Kaufeld for skillful technical assistance.
Participated in research design: Brosda, Fink, Gmeiner, Jantschak, Pertz.
Conducted experiments: Brosda, Franke, Hübner, Jantschak, Möller, Pertz.
Performed data analysis: Brosda, Pertz.
Wrote or contributed to the writing of the manuscript: Brosda, Hübner, Jantschak, Pertz.
- Received April 25, 2013.
- Accepted July 16, 2013.
F.J. and J.B. contributed equally to this work.
- amphetamine-induced locomotion
- analysis of variance
- Chinese hamster ovary
- concentration-response curve
- hD2S receptor
- human dopamine D2short receptor
- hD2L receptor
- human dopamine D2long receptor
- extrapyramidal side effects
- human embryonic kidney 293
- 5-hydroxytryptamine (serotonin)
- Krebs-Henseleit solution
- Nω-nitro-l-arginine methyl ester
- phosphate-buffered saline
- Ro 20-1724
- (S)-(−)-Bay K 8644
- (S)-(−)-1,4-dihydro-2,6-dimethyl-5-nitro-4-[2-(trifluoromethyl)phenyl]-3-pyridinecarboxylic acid methyl ester
- guanosine 5′-O-(3-[35S]thio)triphosphate
- Copyright © 2013 by The American Society for Pharmacology and Experimental Therapeutics