Abstract
Although most antidepressants suppress serotonin (5-HT) and/or noradrenaline reuptake, blockade of 5-HT2C receptors and α2-adrenoceptors likewise enhances monoaminergic transmission. These sites are targeted by the urea derivative N- [4-methoxy-3-(4-methylpiperazin-1-yl)phenyl]-1,2-dihydro-3-H-benzo[e]indole-3-carboxamide (S32212). S32212 was devoid of affinity for monoamine reuptake sites, yet displayed pronounced affinity (pKi, 8.2) for constitutively active human 5-HT2CINI (h5-HT2CINI) receptors, behaving as an inverse agonist in reducing basal Gαq activation, [3H]inositol-phosphate production, and the spontaneous association of h5-HT2CINI-Renilla luciferase receptors with β-arrestin2-yellow fluorescent protein. Furthermore, upon 18-h pretreatment, S32212 enhanced the plasma membrane expression of h5-HT2CINI receptors as visualized by confocal microscopy and quantified by enzyme-linked immunosorbent assay. Its actions were prevented by the neutral antagonist 6-chloro-5-methyl-N-[6-(2-methylpyridin-3-yloxy)pyridin-3-yl]indoline-1-carboxamide (SB242,084), which also impeded the induction by long-term exposure to S32212 of otherwise absent Ca2+ mobilization in mouse cortical neurones. In vivo, S32212 blunted the inhibitory influence of the 5-HT2C agonist 2-(3-chlorobenzyloxy)-6-(1-piperazinyl)pyrazine (CP809,101) on ventrotegmental dopaminergic neurones. S32212 also blocked 5-HT-induced Gαq and phospholipase C activation at the h5-HT2A and, less potently, h5-HT2B receptors and suppressed the discriminative stimulus properties of the 5-HT2A agonist 1-(2,5-dimethoxy-4-iodophenyl)-2-aminopropane in rats. S32212 manifested marked affinity for human α2A- (pKi 7.2), α2B- (pKi 8.2), and α2C- (pKi 7.4) adrenoceptors, at which it abolished noradrenaline-induced recruitment of Gαi3, Gαo, adenylyl cyclase, and extracellular-regulated kinase1/2. Moreover, S32212 dose-dependently abolished the discriminative stimulus effects of the α2-adrenoceptor agonist (S)-spiro[(1-oxa-2-amino-3-azacyclopent-2-ene)-4,2′-(1′,2′,3′,4′-tetrahydronaphthalene)] (S18616). Finally, S32212 displayed negligible affinity for α1A-adrenoceptors, histamine H1 receptors, and muscarinic M1 receptors. In conclusion, S32212 behaves as an inverse agonist at h5-HT2C receptors and as an antagonist at human α2-adrenoceptors (and h5-HT2A receptors). Its promising profile in preclinical models potentially relevant to the treatment of depression is described in J Pharmacol Exp Ther 340:765–780, 2012.
Introduction
Major depression is a common, serious, and heterogeneous disorder for which current treatment is inadequate (Sartorius et al., 2007; Wittchen et al., 2011). In the search for more effective drugs, recent years have witnessed a focus on drugs interacting selectively with one specific target. Despite encouraging observations with antagonists of N-methyl-d-aspartate and nicotinic α4β2 receptors (Rosenzweig-Lipson et al., 2007; Rozas, 2009; Mineur and Picciotto, 2010; Carr and Lucki, 2011; Skolnick et al., 2011), no new agents have as yet been launched. Accordingly, in addition to the association of mechanistically distinct antidepressants, there is a rekindling of interest in agents that interact with several therapeutically relevant mechanisms (Morphy and Rankovic, 2005; Millan, 2006; Wong et al., 2008; Blier et al., 2009). Such selectively nonselective, multitarget, or designed multiple ligands may permit more effective, rapid, and broader control of the cardinal symptoms of depression, as well as poorly treated comorbid symptoms such as cognitive impairment, pain, anxiety, sexual dysfunction, and insomnia (Morphy and Rankovic, 2005; Millan, 2006; Wong et al., 2008; Blier et al., 2009; Millan et al., 2012). One example is agomelatine, which acts “synergistically” as a combined melatonin (MT)1/MT2 agonist and serotonin (5-HT)2C receptor antagonist; agomelatine improves depressed mood in rodents and patients, while also expressing anxiolytic properties, favoring sleep, and countering circadian desynchronization (Kasper and Hamon, 2009; de Bodinat et al., 2010).
Although the melatonergic properties of agomelatine are a fundamental component of its mechanism of action, 5-HT2C antagonist actions are also of importance in several respects. Blockade of tonically and/or constitutively active 5-HT2C receptors on GABAergic interneurones inhibitory to monoaminergic pathways disinhibits the corticolimbic release of dopamine and nonadrenaline (NA), thereby alleviating depressed mood (Gobert et al., 2000; Millan et al., 2000b; Aloyo et al., 2009; Leggio et al., 2009; Di Giovanni et al., 2010). Although the influence of 5-HT2C receptors is not necessarily unitary (Millan, 2005; Rosenzweig-Lipson et al., 2007; Siuciak et al., 2007; Dekeyne et al., 2008; Carr and Lucki, 2011), their antagonism also improves mood via other mechanisms, such as enhanced hippocampal neurogenesis (Dekeyne et al., 2008; Soumier et al., 2009). Moreover, blockade of 5-HT2C receptors is associated with anxiolytic properties (Millan, 2005; Heisler et al., 2007; Dekeyne et al., 2008) and the promotion of sexual function (Hull et al., 2004; Millan, 2006; Kennedy and Rizvi, 2009). 5-HT2C antagonists also favor restorative slow-wave sleep (Smith et al., 2002; Landholt and Wehrle, 2009; de Bodinat et al., 2010); although their precise role compared with their 5-HT2A counterparts remains uncertain (Millan, 2006; Morairty et al., 2008). Although data are limited, blockade of 5-HT2A receptors may, by analogy to antagonism of 5-HT2C receptors, exert a positive influence on mood, as well as on certain domains of cognitive performance (Millan, 2006; Berg et al., 2008; Landholt and Wehrle, 2009; Pandey et el., 2010; Carr and Lucki, 2011).
The above comments underpin interest in 5-HT2C receptors as targets for improving the treatment of depressed states, and it should, in principle, be possible to associate their blockade with complementary properties other than melatonin agonism. In this regard, α2-adrenoceptors (ARs) are of particular interest inasmuch as α2-AR autoreceptors localized on cell bodies and terminals exert a tonic inhibitory influence on the activity of adrenergic pathways throughout corticolimbic structures (Millan et al., 2000b; Invernizzi and Garattini, 2004). α2-AR heteroreceptors also exert a tonic inhibitory control over mesocortical dopaminergic projections and, although not tonically active, ascending serotonergic pathways (Millan et al., 2000b; Invernizzi and Garattini, 2004). Correspondingly, α2-AR blockade is associated with modest antidepressant actions in rodents and humans, and it enhances the speed of onset and/or efficacy of several antidepressant agents (Millan et al., 2000a; Invernizzi and Garattini, 2004; Sanacora et al., 2004; Yanpallewar et al., 2010; Serres et al., 2011). By analogy to 5-HT2C receptors, blockade of α2-ARs promotes sexual function in depressed patients and counters the loss of libido and ejaculatory performance provoked by selective reuptake inhibitors and other classes of drugs that suppress 5-HT reuptake (Millan, 2006; Viitamaa et al., 2006). Although the role of α2-ARs in cognitive processes is complex, blockade of α2-ARs inhibitory to cholinergic input to the frontal cortex and hippocampus enhances certain mnemonic domains, such as attention and cognitive flexibility (Lapiz and Morilak, 2006; Millan, 2006; Dekeyne et al., 2012). This is important in view of the currently inadequate treatment of the cognitive impairment of depressed states that comprises the most frequent residual symptom in “remitted” patients after treatment (Preiss et al., 2009; Marazziti et al., 2010).
The above observations suggest that combined antagonism of 5-HT2C (and/or 5-HT2A) receptors and α2-ARs might offer an effective and broad-based platform for the control of depressed states, if shorn of the unwanted blockade of histamine H1 receptors, α1A-ARs, and muscarinic M1 receptors seen with the tetracyclic agents mirtazapine and mianserin (Anttila and Leinonen, 2001; Millan, 2006; Carr and Lucki, 2011). Accordingly, the present work describe a novel urea derivative, N- [4-methoxy-3-(4-methylpiperazin-1-yl)phenyl]-1,2-dihydro-3-H-benzo[e]indole-3-carboxamide (S32212) (Fig. 1), which behaves as an inverse agonist at 5-HT2C receptors and as an antagonist at α2-ARs and possesses 5-HT2A antagonist properties. This article discusses the receptor binding and functional profile of S32212 in diverse cellular responses, whereas Dekeyne et al. (2012) characterize the actions of S32212 in neurochemical and behavioral paradigms related to its potential influence on depressed mood, anxiety, cognitive function, sleep, and sexual function.
Materials and Methods
Evaluation of Affinities at 5-HT2CINI Receptors, α2-Receptor Subtypes, and Other Binding Sites
As described previously (Millan et al., 2002; Dekeyne et al., 2008), HEK-293 cell membranes expressing recombinant human 5-HT2CINI receptors, and CHO cell membranes stably expressing h5-HT2CVSV, h5-HT2B, or h5-HT2A receptors (10–20 μg of protein), were incubated with [3H]mesulergine (1 nM), or [3H]ketanserin (0.5 nM) for h5-HT2A receptors, at 37°C for 120 min in a buffer containing 20 mM HEPES, pH 7.4, 2 mM EDTA, and 0.1% ascorbic acid. Nonspecific binding was defined with mianserin (1 μM) for h5-HT2CINI, h5-HT2CVSV, and h5-HT2A receptors or 5-HT (10 μM) for h5-HT2B receptors. Affinities at hα2A, hα2B-ARs, and hα2C-ARs stably expressed in CHO cells were also determined as described previously (Millan et al., 2000a) by using membranes (10 μg of protein) incubated with [3H]{2-(2-methoxy-1,4-benzodioxan-2-yl)-2-imidazoline} (RX821,002) (1 nM for hα2A- and hα2C-ARs and 2 nM for hα2B-ARs) and using phentolamine (10 μM) to define nonspecific binding. Standard protocols were used for the determination of affinities at other classes of 5-HT receptors, α1-ARs, histamine H1 and muscarinic M1 receptors, etc. (see Table 1 for summary of key details). Isotherms were analyzed by nonlinear regression using Prism (GraphPad Software Inc., San Diego, CA) to generate half-maximal inhibitory concentrations (IC50) transformed into Ki values according to Cheng and Prusoff (1973): Ki = IC50/(1 + L/Kd), where L is the radioligand concentration and Kd is the dissociation constant.
Inverse Agonist Properties at Constitutively Active h5-HT2CINI Receptors
Antibody-Capture/Scintillation Proximity Assays of Coupling to Gαq.
As described previously (Cussac et al., 2002), transfected HEK-293 cells were homogenized by using a Polytron homogenizer (Kinematica, Liittau-Lucerne, Switzerland), and homogenates were centrifuged at 20,000g for 20 min at 4°C. The pellet (membrane fraction) was resuspended in the assay buffer containing HEPES (20 mM; pH 7.4), MgCl2 (50 mM), NaCl (150 mM), and GDP (0.1 μM) and then incubated with drugs and 0.2 nM [35S]GTPγS for 60 min at room temperature. The reaction was stopped by solubilizing cell membranes with the detergent nonyl phenoxypolyethoxylethanol (0.3% v/v final) and gently agitating them for 30 min. Rabbit anti-Gαq/11 polyclonal antibodies (Tebu-Bio, Le Perray-en-Yvelines, France) were then added, and plates were incubated for 1 h to allow antibody-Gα complexes to form. At the end of incubation, SPA beads coated with a secondary anti-rabbit antibody (GE Healthcare, Chalfont St. Giles, Buckinghamshire, UK) were added, and incubation was performed with gentle agitation overnight before centrifugation and counting of radioactivity on a TopCount microplate scintillation counter (PerkinElmer Life and Analytical Sciences, Waltham, MA). Isotherms were analyzed by nonlinear regression using Prism to yield IC50/EC50 values as outlined above.
Measurement of Inositol Phosphate Production.
As described previously (Chanrion et al., 2008), transfected HEK-293 cells grown in 96-well plates (0.25 × 106 cells/well) were labeled overnight with 0.5 μCi/well [3H]myo-inositol (10–20 Ci/mmol; GE Healthcare). Cells were washed twice in Locke's solution containing 150 mM NaCl, 20 mM HEPES, 4.2 mM KCl, 0.5 mM MgCl2, 1.8 mM CaCl2, and 33 mM glucose, incubated in Locke's solution supplemented with 10 mM LiCl for 10 min, and then exposed to drugs for 30 min. IP generation was terminated by the addition of 0.1 M formic acid. Supernatants were recovered, and total [3H]IPs were purified in 96-well plates by ion exchange chromatography using a DOWEX AGI·X8 resin (Bio-Rad Laboratories, Hercules, CA). The [3H]IPs were then eluted with a solution of 10 M ammonium formate/0.1 M formic acid. Radioactivity was determined by scintillation counting. Results are expressed as the amount of [3H]IP produced in comparison to radioactivity present in the 10% Triton X-100/0.1 M NaOH-solubilized membrane fraction. Isotherms were analyzed by nonlinear regression using Prism to yield IC50/EC50 values as outlined above.
Confocal Microscopy and ELISA Determination of h5-HT2CINI Receptor Cell Surface Expression.
As described previously (Chanrion et al., 2008), HEK-293 cells transfected with 1 μg of FLAG-h5-HT2CINI receptor cDNA plus 1 μg of Gαq were grown on glass coverslips. Four hours after transfection, cells were exposed for 18 h to drugs. They were then washed in PBS and fixed in 4% (w/v) paraformaldehyde in PBS for 15 min at room temperature. After three washes with glycine (0.1 M), cells were permeabilized with 0.1% (w/v) Triton X-100 for 5 min. They were then incubated with PBS containing 10% BSA for 30 min at 37°C and overnight at 4°C with a rabbit anti-FLAG antibody (1:1000; Sigma, St. Louis, MO), in PBS supplemented with 3% BSA. Cells were washed three times with PBS + 10% BSA and incubated for 1 h at room temperature with a Cy3-labeled anti-rabbit antibody (1:2000 dilution in PBS + 3% BSA; Invitrogen, Paisley, UK). After three washes, coverslips were mounted on glass slides in Mowiol 4.88 (Calbiochem, San Diego, CA). Confocal laser scanning microscopy was performed by using a 1024 Bio-Rad Laboratories confocal system. Series of optical sections were collected with a step of 0.40 μm and scanned at 1024 × 1024-pixel resolution. Quantification of receptor cell surface expression was performed by ELISA under nonpermeabilized conditions. After 18-h drug incubation, cells grown in 96-well plates were fixed with 4% paraformaldehyde for 20 min at room temperature. After two washes, cells were incubated in PBS containing 1% fetal calf serum for 30 min then with a horseradish peroxidase-conjugated anti-FLAG monoclonal antibody (1:5000; Sigma) for 30 min. After five washes, the chromogenic substrate (Supersignal ELISA Femto; Thermo Fisher Scientific, Waltham, MA) was added, and immunoreactivity was detected at 492 nm with a Wallac Victor2 luminescence counter (PerkinElmer Life and Analytical Sciences). Control experiments were performed by omitting the primary antibody or using cells transfected with empty vectors. Values were also normalized with respect to the total amount of protein. For each data point, four determinations were averaged, and results were analyzed by ANOVA followed by a Student Newman-Keuls test.
Bioluminescence Resonance Energy Transfer Studies of h5-HT2CINI Receptor β-Arrestin Association.
As described previously (Millan et al., 2011), HEK-293 cells were seeded in six-well plates then transiently transfected with 50 ng of h5-HT2CINI-Renilla luciferase (Rluc) cDNA and 500 ng of β-arrestin2-YFP. Twenty four hours after transfection, cells were transferred to 96-well plates and subjected to BRET analysis 48 h after transfection. Coelenterazine H substrate (Interchim, Montluçon, France) was added at a final concentration of 5 μM for 10 min at room temperature before drug addition and BRET measurements. Readings were performed by using a lumino/fluorometer (Mithras; Berthold Technologies, Bad Wildbad, Germany) permitting sequential integration of luminescence signals with two filter settings (Rluc filter, 485 ± 10 nm; YFP filter, 530 ± 12.5 nm). Emission signals at 530 nm were divided by emission signals at 485 nm. The difference between the ratio obtained with cotransfected Rluc and YFP fusion proteins and that obtained with Rluc fusion protein alone was defined as the BRET ratio. Results were expressed in milliBRET units (with 1 milliBRET unit corresponding to the BRET ratio values multiplied by 1000).
Calcium Imaging in Primary Cultures of Mouse Cortical Neurones.
As described previously (Chanrion et al., 2008), cortical neurones, grown in Lab-Tek II chamber slides and pretreated or not for 18 h with drug, were loaded with Fura-2 acetoxymethyl ester (Invitrogen, Carlsbad, CA) at a final concentration of 12.5 μM for 30 min at 37°C in Locke's solution. Drugs were also included into the loading medium. Cells were rinsed three times in Locke's solution and incubated for 30 min in dye-free Locke's solution in the absence of drugs. Lab-Tek slides were then placed on the stage of an IX70 microscope (Olympus, Tokyo, Japan) and superfused with Locke's solution. Imaging of intracellular Ca2+ changes in individual cells was accomplished by ratiometric imaging of Fura-2 fluorescence at 340- and 380-nm excitation by using the MetaFluor Imaging system (Molecular Devices, Sunnyvale, CA). Fluorescence was excited by illumination via a 20× water immersion objective with rapid light wavelength switching provided by a DG4 filter wheel (Sutter Instrument Company, Novato, CA) and detected by a charge-coupled device camera. Images were obtained for 30 s before the application of 5-HT to establish a stable baseline. Ca2+ responses obtained in representative fields of cells (50–80 cells/field) from at least three experiments done on different sets of neurones were compiled.
Antagonist Properties at h5-HT2CVSV, h5-HT2B, and h5-HT2A Receptors
Antibody-Capture/Scintillation Proximity Assays Studies of Coupling to Gαq.
In antagonist experiments, h5-HT2CVSV, h5-HT2A, or h5-HT2B receptor membranes were preincubated on 96-well plates with S32212 30 min before the addition of 5-HT (10 nM for h5-HT2CVSV and 100 nM for h5-HT2A and h5-HT2B) and [35S]GTPγS (0.2 nM; PerkinElmer Life and Analytical Sciences). For each receptor subtype, the inhibition of 5-HT-induced Gαq activation by S32212 was determined by using a similar SPA procedure to that described above for h5-HT2CINII receptors. Isotherms were analyzed by nonlinear regression using Prism to yield EC50 and IC50 values. The pKB values of S32212 for the inhibition of 5-HT-stimulated [35S]GTPγS binding were calculated according to Cheng and Prusoff (1973): KB = IC50/[1 + (agonist/EC50)].
Evaluation of Phospholipase C Activity: [3H]PI Depletion Assay.
As described previously (Cussac et al., 2002), CHO cells stably expressing h5-HT2A, h5-HT2B, and h5-HT2CVSV receptors were grown in adherent culture in 225-cm2 flasks with UltraCHO medium (Lonza Verviers SPRL, Verviers, Belgium) containing sodium pyruvate (1 mM), dialyzed fetal calf serum (0.1%), and geneticin (G-418; 400 μg/ml). At confluence, cells were labeled with 2 μCi/ml of [3H]myo-inositol (10–20 Ci/mmol; GE Healthcare) for 24 h in serum-free UltraCHO medium. Adherent cells were rinsed twice in Krebs-LiCl buffer, scraped from the flask, and washed again by slow centrifugation. Cells were resuspended in Krebs-LiCl and left to stand for 15 min at 37°C before use or kept at −80°C in Krebs-LiCl/10% dimethyl sulfoxide until assay. For determination of [3H]PI depletion, incubation of cells loaded with [3H]myo-inositol was performed in 96-well plates with 5-HT (1 μM for h5-HT2A and 30 nM for h5-HT2B and h5-HT2CVSV) at 37°C for 20 min (h5-HT2C) or 30 min (h5-HT2A and h5-HT2B) in a 0.4-ml final volume. Assays were stopped with 0.4 ml of methanol/HCl (88 ml of 100% methanol + 12 ml of 1 N HCl). The 96-well plates were sonicated for 2 min, and membranes were recovered by using a Filter Harvester (PerkinElmer Life and Analytical Sciences). Radioactivity retained on filters was quantified by liquid scintillation. Isotherms were analyzed by nonlinear regression using Prism, and the pKB value of S32212 was calculated as described above.
Antagonist Actions at hα2A-, hα2B-, and hα2C-ARs
Determination of [35S]GTPγS Binding.
[35S]GTPγS binding was determined as documented in detail elsewhere (Audinot et al., 2002). In brief, membranes of CHO cells expressing hα2A-, hα2B-, or hα2C-AR (1.9 pmol/mg protein) were incubated for 60 min at 22°C with S32212 and/or NA (10 μM) and [35S]GTPγS (0.1 nM) in a buffer containing 20 mM HEPES, pH 7.4, 100 mM NaCl, 3 μM GDP, and 3 mM MgSO4. Incubations were terminated by rapid filtration through Whatman GF/B filters using a Filter Harvester (PerkinElmer Life and Analytical Sciences). Radioactivity retained on the filters was quantified by liquid scintillation counting. Isotherms were analyzed by nonlinear regression using Prism, and the pKB value of S32212 was calculated as described above.
SPA Studies of Coupling to Gαi3 and Gαo.
The coupling of hα2A- and hα2C-ARs to Gαi3 was evaluated by using SPA procedures essentially as documented in detail elsewhere (Cussac et al., 2002). In brief, membranes were preincubated with S32212 or assay buffer for 30 min on 96-well plates before the addition of NA (0.3 μM for hα2A-AR and 10 μM for hα2C-AR) and [35S]GTPγS (0.2 nM) for 1 h at 22°C. The buffer contained 20 mM HEPES, pH 7.4, 0.3 μM GDP, 3 mM MgCl2, and 150 mM NaCl. The reaction was stopped by solubilizing membranes with Nonidet P-40 (0.27% final concentration). After gentle agitation for 30 min, 10 μl of mouse monoclonal anti-Gαi1/3 antibodies or anti-Gαo antibodies (Enzo Life Sciences, Inc., Farmingdale, NY) were then added, and plates were incubated for 1 h. At the end of the incubation period, SPA beads coated with anti-mouse secondary antibodies (GE Healthcare) were added and incubated with gentle agitation overnight before counting radioactivity on a TopCount microplate scintillation counter (PerkinElmer Life and Analytical Sciences). Nonspecific binding was defined with 10 μM GTPγS. Isotherms were analyzed by nonlinear regression using Prism, and pKB values were calculated as described above.
Measurement of cAMP Formation at hα2A-and hα2C-ARs by Homogeneous Time-Resolved Fluorescence.
The measurement of cAMP responses to NA stimulation was performed by homogeneous time-resolved fluorescence using a competitive immunoassay between native cAMP produced by cells and the cAMP labeled with the dye d2 as described previously (Gabriel et al., 2003). In brief, CHO cells stably expressing human α2A and α2C adrenoceptors were grown to 80% confluence. Cell suspension (5 μl; 10,000 cells per well) was added in 384-well culture plates and preincubated with increasing concentrations of S32212 (2.5 μl) diluted in a PBS buffer containing 3-isobutyl-1-methylxanthine (1 mM) and forskolin (10 μM) for 5 min. At the end of the incubation period, NA (30 nM) was added to each well and incubated with cells for 30 min at room temperature. The reaction was then stopped by adding 5 μl of cAMP d2 conjugate and 5 μl of anti-cAMP antibodies labeled with Lumi4-Europium cryptate diluted in lysis buffer. After 1 h of incubation, the plates were read at 665/620 nm by using an EnVision multilabel plate reader (PerkinElmer Life and Analytical Sciences). The results calculated from the 665/620-nm ratio were expressed as a percentage of the inhibition by NA of forskolin-induced cAMP production.
Evaluation of Extracellular-Regulated Kinase Phosphorylation by Immunoblotting.
CHO cells expressing hα2A-ARs or hα2C-ARs were grown in six-well plates until confluence. Cells were then washed twice with serum-free medium and incubated overnight in the medium. Cells were preincubated for 30 min with S32212 or buffer, and then stimulated with NA (1 μM) for 5 min. Phosphorylated ERK1/2 was measured in cell extracts by immunoblotting using a monoclonal antibody against phosphorylated pp42 mitogen-activated protein kinase (ERK2) and pp44 mitogen-activated protein kinase (ERK1) (Cell Signaling Technology, Danvers, MA) as described previously (Cussac et al., 1999). The signal was quantified by measurement of optical density and expressed as a percentage of the maximal effect induced by NA (1 μM). Isotherms were analyzed by nonlinear regression using Prism, and pKB values were calculated as described above.
Actions of S32212 at 5-HT2C Receptors, 5-HT2A Receptors, and α2-ARs In Vivo
Animals.
These studies used male Wistar rats and male C57BL/6 mice (Charles River Laboratories, L'Arbresle, France), weighing 200 to 250 g and 22 to 25 g upon arrival, respectively. They were housed in sawdust-lined cages with, unless otherwise specified below, unrestricted access to standard chow and water. There was a 12-h light/dark cycle with lights on at 7:30 AM. Laboratory temperature and humidity were 21 ± 0.5°C and 60 ± 5%, respectively. Animals were adapted to laboratory conditions for at least 1 week before testing. All procedures conformed to international European ethical standards (86/609-EEC) and the French National Committee (décret 87/848) for the care and use of laboratory animals.
Electrical Activity of Ventral Tegmental Area Dopaminergic Cell Bodies.
The technique used has been detailed previously (Millan et al., 2000a). In brief, rats (275–325 g) were anesthetized with chloral hydrate (400 mg/kg i.p.) and, after cannulation of the saphenous vein, they were placed in a stereotaxic apparatus. A tungsten microelectrode was slowly lowered into the VTA (from bregma and sinus surface, anteroposterior −5.5, lateral ± 0.9, dorsoventral −7.2/8.5 mm). After amplification and analog to digital conversion of electrical activity, data were recorded by using Spike2 software (CED, Cambridge, UK). Dopaminergic neurones were identified by their wave form and spontaneous firing pattern (≈5.0 Hz). One cell was recorded per animal. Spontaneous firing rate was recorded for 5 min before intravenous administration of vehicle. Three minutes later, vehicle or 2-(3-chlorobenzyloxy)-6-(1-piperazinyl)pyrazine (CP809,101) (1 mg/kg i.v.) were administered. Three minutes later, S33212 (intravenously in a volume of 0.5 ml/kg) or vehicle was administered. Drug effects were characterized over 60-s periods and expressed as percentage of change from spontaneous firing rate (defined as 0%). They were analyzed by two-way ANOVA followed by Newman-Keuls test.
Induction of Penile Ejections in Rats by the Selective 5-HT2C Agonist CP809,101.
The procedure was adapted from Millan et al. (1997). Rats weighing 120 to 140 g were individually placed in transparent, Plexiglas observation cages after S32212 or vehicle treatment. Thirty minutes later, animals received CP809,101 (0.63 mg/kg s.c.) (Siuciak et al., 2007), and erections were counted over 30 min. In a complementary experiment, the α2-AR antagonist RX821,002 (2.5 mg/kg s.c.) or vehicle was administered 15 min before the 5-HT2C antagonist 6-chloro-5-methyl-N-[6-(2-methylpyridin-3-yloxy)pyridin-3-yl]indoline-1-carboxamide (SB242,084) (0.63 mg/kg i.p) or vehicle and 30 min before CP809,101 (0.63 mg/kg s.c.).
5-Hydroxytryptophan-Induced Head Twitches in Mice.
The 5-HT precursor 5-HTP was injected at a dose of 200 g/kg i.p., then mice were placed individually in Plexiglas observation cylinders. Five minutes after 5-HTP injection, the number of head twitches made in 5 min was counted. S32212 or vehicle was given subcutaneously 30 min before 5-HTP.
Discriminative Stimulus Properties of the 5-HT2A Agonist 1-(2,5-Dimethoxy-4-Iodophenyl)-2-Aminopropane and the α2-AR Agonist (S)-Spiro[(1-Oxa-2-Amino-3-Azacyclopent-2-ene)-4,2′-(1′,2′,3′,4′-Tetrahydronaphthalene)] in Rats.
Using procedures detailed previously (Schreiber et al., 1994), singly housed rats with restricted access to food (10–11 g/day) were trained to discriminate 1-(2,5-dimethoxy-4-iodophenyl)-2-aminopropane (DOI) (0.63 mg/kg i.p.) from vehicle or (S)-spiro[(1-oxa-2-amino-3-azacyclopent-2-ene)-4,2′-(1′,2′,3′,4′-tetrahydronaphthalene)] (S18616) (0.01 mg/kg s.c.) from vehicle, using two-lever, fixed-ratio 10, food-reinforced procedures. Each 15-min daily session (5 days/week) commenced 15 min after vehicle or the training drug administration. After completion of training, test sessions were performed twice a week with training sessions on the other days. In test sessions, S32212 was injected 30 min before the training drug. The lever selected was that on which 10 responses were emitted first.
Data Analyses and Behavioral Procedures.
In drug discrimination assays, lever selection data (percentage of rats selecting the training drug lever) were compared by Fisher's exact probability tests to control values (100%). In addition, to control for a potentially disruptive influence on motor performance, response rates (the total number of responses on both levers) were compared with the preceding drug training session by use of paired t tests. For the other behavioral procedures, dose effects were analyzed by one-way ANOVA followed by Dunnett's test.
Plasmids, Drugs, and Structures.
The construct encoding the pCMV/Flag-h5-HT2CINI was described previously (Chanrion et al., 2008). The plasmid encoding Gαq (Gαq/prK5) was kindly provided by Dr J. P. Pin (Institut de Génomique Fonctionnelle, Montpellier, France). The plasmid encoding pcDNA3.1(+)-h5-HT2CINI-RedLuc was cloned internally (Servier, Neuilly-sur-Seine, France), and the pcDNA3.1-β-arrestine2-YFP plasmid was generated as described previously (Millan et al., 2011). For in vivo studies, S32212 was used either subcutaneously (dissolved in sterile water) or intraperitoneally (suspended in water with a few drops of Tween 80). Doses were in terms of the base. Drug salts and sources were as follows: 5-HT (creatinine sulfate) was purchased from Sigma. (−)Noradrenaline ditartrate, DOI, and RX821,002 were obtained from Sigma/RBI (Natick, MA). S32212 HCl, SB242,084, S18616, and CP809,101 were synthesized by Servier.
Results
Binding Profile of S32212 at h5-HT2CINI Receptors and hα2-AR Subtypes Compared with Other Sites.
S32212 possessed marked affinity for 5-HT2CINI receptors, displaying a pKi of 8.18 ± 0.07 for displacement of [3H]mesulergine binding (Fig. 2A; Table 1): its affinity was similarly high for 5-HT2CVSV receptors (8.05 ± 0.03) (Fig. 3A; Table 1). Pronounced affinity was also seen for h5-HT2A receptors (8.24 ± 0.01), whereas its affinity at h5-HT2B sites (6.98 ± 0.01) was approximately 20-fold lower than for 5-HT2CINI sites (Table 1). S32212 exhibited marked affinities at hα2A-, hα2B-, and hα2C-ARs (pKi 7.21 ± 0.08, 8.24 ± 0.18, and 7.44 ± 0.01, respectively) compared with lower affinities for hα1-AR and hβ-AR subtypes. Compared with h5-HT2CINI receptors, modest affinities were seen for most other classes of 5-HT receptor, h5-HT1A, 5-HT1B, 5-HT1D, 5-HT4, h5-HT5A, h5-HT6, and h5-HT7, whereas negligible affinity was found at 5-HT3 receptors. The affinities of S32212 for dopamine receptor subtypes, monoamine transporters, and monoamine oxidase A and B were low. Affinities were likewise low for histamine hH1 and hH3 sites, although modest affinities were seen for hH2 receptors. S32212 displayed low affinities for muscarinic hM1 and hM2 receptors and for a broad range (approximately 80) of other receptors, enzymes, and channels assays (extended CEREP and “in house” screens; see CEREP web site, www.cerep.fr, and Table 1 for methodological details). It is noteworthy that S32212 displayed negligible (pKi <5.0) affinities for the following sites: opiate μ, δ, and κ, adenosine (A1, A2 adenosine), angiotensin I, benzodiazepine, bradykinin B2, calcitonin gene-related peptide, cannabinoid (CB1, CB), cholecystokinin (A, B), choline uptake, endothelin (A, B), GABA (A, B), glutamate (α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid, N-methyl-d-aspartate), imidazoline (I2), MT1 and MT2, neurokinin (NK1, NK2), neuropeptide Y (Y1, Y5), nicotinic, prostanoid (TXA2/PGH2), σ1 and σ2, estrogen, progesterone, and testosterone receptors. It also showed low affinity (pKi < 5.0) for Ca2+ channels (diltiazem site), Na+ channels (batrachotoxin site), and K+ channels (ATP, Ca2+, and voltage dependent), and for a variety of enzymes (acetylcholinesterase, adenyl cyclase, cyclooxygenase 1, guanylyl cyclase, 5-lipooxygenase, nitric-oxide synthase, phospholipase A2, phospholipase C, phosphodiesterase I and III, protein kinase C, and sodium-potassium ATPase).
Inverse Agonist Properties of S32212 at Constitutively Active h5-HT2CINI Receptors Coupled to Gαq and Phospholipase C.
In HEK-293 cells expressing 5-HT2CINI receptors, [35S]GTPγS binding to Gαq was concentration-dependently induced by 5-HT with a pEC50 value of 6.96 ± 0.10, whereas the neutral antagonist SB242,084 was inactive. S32212 suppressed basal [35S]GTPγS binding by 24 ± 2% with a potency (pEC50) of 7.14 ± 0.15 (Fig. 2B). Likewise, 5-HT concentration-dependently increased IP accumulation with a pEC50 value of 6.86 ± 0.16, SB242,084 was inactive, and S32212 concentration-dependently decreased basal IP accumulation with a pEC50 of 7.73 ± 0.22 (Fig. 2C).
Antagonist Properties of S32212 at Constitutively Silent h5-HT2CVSV Receptors Coupled to Gαq and Phospholipase C.
In cells expressing constitutively silent h5-HT2CVSV receptors, S32212 concentration-dependently inhibited the enhancement of [35S]GTPγS binding to Gαq induced by 5-HT (10 nM) with a pKb of 7.42 ± 0.19 (Fig. 3B). S32212 showed no intrinsic efficacy alone. Similar data were obtained for the activation of PLC: 5-HT-induced [3H]PI depletion was concentration-dependently blocked by S32212 (inactive alone) with a pKb of 7.77 ± 0.15 (Fig. 3C).
Inverse Agonist Properties of S32212 at h5-HT2CINI Receptors Recruiting β-Arrestin2.
5-HT concentration-dependently increased the basal BRET signal for h5-HT2CINI-Rluc receptor association with β-arrestin2-YFP with a pEC50 of 6.75 ± 0.22. By contrast, S32212 concentration-dependently inhibited the 5-HT-triggered BRET signal with a pKB of 6.91 ± 0.12 (data not shown). S32212 alone suppressed the BRET signal, and the specificity of its effect was demonstrated by its reversal upon pretreatment (15 min) with SB242,084, which was itself inactive alone (Fig. 4).
Inverse Agonist Properties of S32212 for Enhancing Cell Surface Expression of h5-HT2CINI Receptors.
Treatment of HEK-293 cells expressing h5-HT2CINI receptors with S32212 for 18 h induced a marked and concentration-dependent redistribution of h5-HT2CINI receptors to the cell surface, as quantified by ELISA. This action was abolished by SB242,084, which alone did not affect the subcellular distribution of the receptor (Fig. 5, B and C). Accordingly, as visualized by confocal microscopy, in the presence of S32212, h5-HT2C receptors were essentially absent from the cytosol and detected mainly at the plasma membrane. This effect was suppressed in the presence of SB242,084, which was itself inactive (Fig. 5A).
Inverse Agonist Properties of S32212 for Enhancing 5-HT-Induced Ca2+ Mobilization in Cortical Neurones.
Application of 5-HT (up to 1 μM) did not induce a detectable increase in cytosolic Ca2+ concentrations in cultured cortical neurones under control conditions (Fig. 6A). By contrast, application of 5-HT after pretreatment with S32212 (1 μM) for 18 h elicited a clear cytosolic Ca2+ elevation (Fig. 6B). When S32212 was applied for 18 h in the presence of SB242,084 its induction of the response to 5-HT was abolished, with SB242,084 being inactive alone (data not shown; Chanrion et al., 2008). Furthermore, SB242,084 administered acutely also blocked the Ca2+ response to 5-HT in cells pretreated with S32212, indicating that the 5-HT-elicited Ca2+ response was mediated by 5-HT2C receptors.
Antagonist Properties of S32212 at h5-HT2A and h5-HT2B Receptors: Coupling to Gαq and Phospholipase C.
In cells expressing h5-HT2A and h5-HT2B receptors, S32212 concentration-dependently blocked 5-HT-induced stimulation of [35S]GTPγS binding to Gαq with pKB values of 7.72 ± 0.07 and 6.57 ± 0.10, respectively (Fig. 7). No effect of S32212 was observed when applied alone. Further demonstrating its antagonist properties at h5-HT2A and h5-HT2B receptors, S32212 concentration-dependently blocked 5-HT-induced-[3H]PI depletion with pKB values of 8.06 ± 0.05 and 7.38 ± 0.10, respectively. When applied alone, S32212 had no effect.
Antagonist Properties of S32212 at hα2A, hα2B, and hα2C-ARs: Influence on [35S]GTPγS Binding.
S32212 concentration-dependently blocked the induction by NA (3 μM) of [35S]GTPγS binding to the overall population of G proteins at hα2A-, hα2B-, and hα2C-ARs with pKB values of 6.84 ± 0.13, 7.68 ± 0.2, and 6.87 ± 0.08, respectively (data not shown). Alone S32212 was inactive (data not shown).
Antagonist Properties of S32212 at hα2A- and hα2C-ARs: Coupling to Gαi3 and Gαo, cAMP Production, and ERK Phosphorylation.
In SPAs performed on CHO cells transfected with hα2A-ARs, S32212 blocked NA-induced binding of [35S]-GTPγS to Gαi3 and Gαo with pKB values of 6.28 ± 0.13 and 7.21 ± 0.17, respectively; S32212 was inactive alone (Fig. 8, A and B). Likewise in CHO cells transfected with hα2A-ARs, the induction by NA (1 μM) of ERK1/2 phosphorylation was blocked by S32212 (pKB 6.44 ± 0.08), and it was inactive alone (Fig. 8D). Similar antagonist profiles for S32212 were obtained in CHO cells transfected with hα2C-ARs. S32212 antagonized the NA-induced binding of [35S]-GTPγS to Gαi3 and Gαo (pKB 7.18 ± 0.04 and 7.55 ± 0.08, respectively) and was inactive alone (Fig. 8, E and F). Likewise, in cells expressing hα2C-ARs, the increase in ERK1/2 phosphorylation elicited by NA (1 μM) was blocked by S32212 (pKB 6.25 ± 0.08), which was inactive alone (Fig. 8H). In CHO cells expressing hα2A- and hα2C-ARs, S32212 blocked the NA-induced inhibition of forskoline (10 μM)-stimulated cAMP production (Fig. 8, C and G), and it was inactive alone (data not shown).
Blockade by S32212 of the Inhibitory Influence of the 5-HT2C Agonist CP809,101 on the Firing Rate of VTA-Localized Dopaminergic Cells.
Two-way ANOVA performed on the influence of CP809,101 in the absence or presence of S32212 (Fig. 9A) showed a significant effect of CP809,101 (F1,11 = 10.1; p < 0.01) and a significant S32212 × CP809,101 interaction (F1,22 = 8.0; p < 0.01). The selective 5-HT2C agonist CP809,101 (1 mg/kg i.v.) significantly decreased the firing rate of dopaminergic cells of the VTA. The time course (Fig. 9B) indicated a maximal inhibition of 52.4 ± 15.0% versus spontaneous firing rate with a rapid onset of action (100 s after injection). S32212 (1 mg/kg i.v.), which did not significantly modify the firing rate of dopaminergic cells by itself, completely prevented the inhibitory influence of CP809,101 (Fig. 9).
Influence of S32212 on CP809,101-Induced Penile Erections.
Penile erections provoked by CP809,101 (0.63 mg/kg s.c), were dose-dependently blocked by the selective 5-HT2C antagonist SB242,084 (0.04–2.5 mg/kg i.p.; data not shown). For S32212, ANOVA was as follows: F3,35 = 4.1; P < 0.05. S32212 displayed a significant effect only at an intermediate dose (2.5 mg/kg s.c.) (Fig. 10A). Suggesting that its antagonist properties at α2-ARs may interfere with its reduction of CP809,101-induced penile erections, the α2-AR antagonist RX821,002 prevented the blockade by SB242,084 of penile erections provoked by CP809,101 (Fig. 10B). Two-way ANOVA was as follows: effect of SB242,084, F1,32 = 8.9, P < 0.01; effect of RX821,002, F1,32 = 3.4, P > 0.05; and interaction, F1,32 = 0.8, P > 0.05.
Blockade by S32212 of 5-HTP-Induced Head Twitches and Discriminative Stimulus Properties of DOI.
S32212 dose-dependently blocked the induction of head twitches in mice by 5-HTP (200 g/kg i.p.) (Fig. 10C). ANOVA was as follows: F4,25 = 14.3; P < 0.001. Further demonstrating antagonist properties at 5-HT2A sites, S32212 dose-dependently blocked the discriminative stimulus properties of the 5-HT2A agonist DOI (0.63 mg/kg i.p.) (Fig. 10D). Over the dose range tested (0.04–10.0 mg/kg i.p.), there was no influence of S32212 on response rate versus the preceding drug training session (paired t tests, P > 0.05; data not shown).
Blockade by S32212 of the Discriminative Stimulus Properties of the α2-AR Agonist S18616.
S32212 dose-dependently blocked the discriminative stimulus properties of the selective α2-AR agonist S18616 (0.01 mg/kg s.c.) (Fig. 10E), without any significant influence on response rate over the dose range tested (0.31–2.5 mg/kg s.c.) (data not shown).
Discussion
Distinctive Multitarget Binding Profile of S32212.
S32212 displayed high affinity for 5-HT2CINI/VSV receptors, the blockade of which should afford antidepressant activity and other useful properties (see Introduction of Dekeyne et al., 2012). It also recognized 5-HT2A receptors, antagonism of which may provide complementary benefits, whereas its affinity for closely related 5-HT2B receptors was less marked. Compared with 5-HT2CINI receptors, the affinity of S32212 was substantially (∼100-fold) lower at 5-HT3, 5-HT4, 5-HT5, and 5-HT7 receptors, whereas it displayed affinities 20- to 80-fold inferior for 5-HT1A, 5-HT1B/5-HT1D, and 5-HT6 receptors. Highly sensitive 5-HT1A receptors operate as inhibitory autoreceptors on serotonergic perikarya, and their recruitment is associated with anxiolytic properties (Millan et al., 2000b, 2008; Millan, 2006; Carr and Lucki, 2011; Yamamura et al., 2011). However, S32212 affected neither the firing rate of dorsal raphe nucleus-localized serotonergic neurones nor the release of 5-HT in corticolimbic structures (Dekeyne et al., 2012), so 5-HT1A autoreceptors are unlikely to be implicated in its anxiolytic actions. Stimulation of postsynaptic 5-HT1A receptors counters depressed states, and their indirect recruitment has been implicated in the antidepressant actions of mirtazapine (Anttila and Leinonen, 2001; Szegedi and Schwertfeger, 2005; Millan, 2006; Carr and Lucki, 2011; Yamamura et al., 2011). However, postsynaptic 5-HT1A receptors are less sensitive than their presynaptic counterparts (Millan et al., 2008), and S32212 did not evoke prototypical responses such as hypothermia, so they are unlikely to be involved in its antidepressant effects. Activation of postsynaptic 5-HT1B sites has also been proposed as a mechanism of antidepressant action (Millan 2006; Carr and Lucki, 2011), but S32212 showed only weak partial agonist actions at h5-HT1B receptors (EC50 ≈8 μM for induction of cAMP formation; C. Mannoury la Cour, unpublished observation), so they are unlikely to be involved in its effects. Finally, antagonism of 5-HT6 receptors is associated with anxiolytic and antidepressant properties and an enhancement of cognition (Fone, 2008; Carr and Lucki, 2011). However, antagonist actions of S32212 at 5-HT6 receptors (IC50 ≈6 μM for inhibition of 5-HT-induced cAMP formation; C. Mannoury la Cour, unpublished observation) are weak. Collectively, the modest actions of S32212 at several classes of 5-HT receptor other than 5-HT2C (or 5-HT2A) sites, although putatively favorable, are unlikely to play a major role in its mood-improving and other useful properties.
S32212 recognized all three isoforms of α2-AR, blockade of which should positively affect mood, cognition, and other functions (below and Dekeyne et al., 2012). Conversely, its affinity was low for α1-ARs, the blockade of which (most importantly of the α1A-subtype) induces sedation, orthostatic hypotension, and cardiovascular-autonomic side effects (Szegedi and Schwertfeger, 2005; Millan, 2006; Sartorius et al., 2007). This is an important distinction to the high potencies of several antidepressants at α1A-ARs, notably tricyclics such as desipramine and the tetracyclic agents mirtazapine and mianserin (Anttila and Leinonen, 2001; Szegedi and Schwertfeger, 2005; Millan 2006).
Sedation, somnolence, weight gain, and compromised cognition are risks associated with the blockade of histamine H1 receptors, as illustrated by mirtazapine, the 5-HT2C antagonist/5-HT reuptake inhibitors, trazodone, and tricyclics such as amitriptyline (Anttila and Leinonen, 2001; Szegedi and Schwertfeger, 2005; Millan, 2006; Sartorius et al., 2007). It is thus important that S32212 had low affinity for H1 receptors. We were surprised to find that it showed modest affinity for histamine H2 receptors (Table 1), a clinically validated mechanism for moderating the increased risk for ulcers associated with stress-related disorders such as depression (Harty and Ancha, 2006). However, any functional relevance of H2 blockade by S32212 to treatment of the excess gastric acid secretion provoked by chronic stress remains to be evaluated. Finally, the negligible affinity of S32212 for muscarinic M1 receptors deserves emphasis inasmuch as their blockade by mirtazapine, mianserin, and tricyclics can trigger adverse autonomic side effects such as dry mouth, gastrointestinal disturbances, and dizziness (Szegedi and Schwertfeger, 2005; Millan, 2006; Sartorius et al., 2007).
Inverse Agonist Properties of S32212 at 5-HT2CINI Receptors.
All isoforms of 5-HT2C receptor couple via Gαq to PLC (Berg et al., 2008; Millan et al., 2008; Aloyo et al., 2009). Demonstrating antagonist properties at 5-HT2CVSV receptors, S32212 blocked both 5-HT-induced [35S]GTPγS binding to Gαq and PLC activation, exerting these actions with potencies corresponding well with its affinity. The affinities of 5-HT2C receptor ligands are generally little affected by mRNA editing (Sanders-Bush et al., 2003; Millan et al., 2008), and the potency of S32212 at nonedited 5-HT2CINI receptors was similar to their 5-HT2CVSV counterparts. Conversely, isoforms differ markedly in their spontaneous (ligand-independent) coupling to transduction mechanisms. The higher constitutive activity of 5-HT2CINI versus 5-HT2CVSV receptors is reflected in the marked suppression of basal Gαq and PLC activation by inverse agonists such as SB206,553, actions blocked by the neutral antagonist SB242,084, which is itself ineffective (Schlag et al., 2004; Chanrion et al., 2008; Dekeyne et al., 2008; Aloyo et al., 2009; Millan et al., 2011). S32212 likewise behaved as an inverse agonist in reducing basal Gαq activity and IP production at 5-HT2CINI receptors.
5-HT2CINI receptors also display constitutive activity as regards coupling to β-arrestin, which fulfils a dual role in signaling and receptor endocytosis. For example, together with calmodulin, recruitment of β-arrestin is involved in 5-HT2C receptor-mediated phosphorylation of ERK1/2 (Werry et al., 2005; Labasque et al., 2008; Knauer et al., 2009). Underpinning the inverse agonist profile of S32212, as demonstrated by BRET, SB242,084 reversibly suppressed spontaneous association of C-terminal Luc-tagged 5-HT2C receptors with YFP-tagged β-arrestin2. Paralleling their constitutive association with β-arrestin, a substantial subpopulation of 5-HT2CINI receptors is localized in the cytosol under resting conditions and, in line with their inhibitory influence on β-arrestin association, inverse agonists promote cell surface expression of 5-HT2CINI receptors (Schlag et al., 2004; Chanrion et al., 2008) and, correspondingly, plasma membrane insertion of 5-HT2C receptors was enhanced by S32212 in cells expressing N-terminal Flag-tagged 5-HT2CINI receptors. Underscoring specificity, S32212-induced 5-HT2CINI receptor cell surface expression was abolished by SB242,084.
Primary cultures of mouse cortical neurones express a majority (approximately 70%) of 5-HT2CINI receptors and other constitutively active isoforms (Chanrion et al., 2008; Dekeyne et al., 2008; Millan et al., 2011). Reflecting constitutive internalization and low cell surface receptor density, an acute 5-HT pulse does not normally elicit a Ca2+ signal. Conversely, chronic treatment with inverse agonists drives 5-HT2C receptors to the cell membrane, leading to a Ca2+ signal in response to 5-HT, as determined by Ca2+ imaging (Chanrion et al., 2008; Dekeyne et al., 2008; Millan et al., 2011). By analogy, long-term treatment with S32212 induced functional Ca2+ responses to 5-HT, an action abolished by SB242,084. Hence, inverse agonist properties of S32212 are indeed expressed at native, cerebral populations of constitutively active 5-HT2C receptor.
Collectively, these data provide compelling evidence that S32212 behaves as an inverse agonist. Nonetheless, other transduction signals under the control of 5-HT2CINI receptors can be differentially affected by specific drugs: for example, SB206,553 is a partial agonist for phospholipase A2 (Aloyo et al., 2009). Hence, additional study will be needed to fully characterize the influence of S32212 on intracellular signals controlled by 5-HT2C receptors.
Blockade by S32212 of 5-HT2C Receptors In Vivo.
Excitatory 5-HT2C receptors are expressed on GABAergic interneurones inhibitory to ascending dopaminergic neurones and, as shown herein using the selective 5-HT2C agonist CP809,101 (Siuciak et al., 2007), their activation suppresses the firing rate of dopaminergic cell bodies in the VTA (Gobert et al., 2000; Millan et al., 2000b; Invernizzi et al., 2007; Di Giovanni et al., 2010). Consistent with antagonist properties of at 5-HT2C sites in vivo, S32212 abolished this action of CP809,101. S32212 did not excite VTA-localized dopaminergic cell bodies upon administration alone nor enhance dopamine release in nucleus accumbens or striatum (Dekeyne et al., 2012). This finding warrants further investigation in light of ongoing discussions concerning the relative influence of neutral antagonists versus inverse agonists on dopaminergic pathways (Millan, 2005; Berg et al., 2008; Aloyo et al., 2009; Leggio et al., 2009; Di Giovanni et al., 2010).
Penile erections are induced in rats by the stimulation of 5-HT2C receptors localized on preganglionic and ventral horn motoneurones in the sacral spinal cord (Bancila et al., 1999; Millan, 2006), and this species-specific response to CP809,101 was inhibited by S32212, which did not itself trigger erections. The U-shaped dose-response curve of S32212 differs to selective 5-HT2C antagonists/inverse agonists (Millan et al., 1997; Andersson, 2001) and probably reflects its additional α2-AR antagonist properties because: 1) selective α2-AR antagonists enhance erectile function by actions both at spinal α2-ARs and prejunctional α2-ARs in the penis itself (Andersson, 2001; Yaïci et al., 2002) and 2) the selective α2-AR antagonist RX821,002 prevents blockade of CP809,101-induced penile erections by the 5-HT2C antagonist SB242,084 (Fig. 10). The broader interest of α2-AR and 5-HT2C antagonist properties in the promotion of (supraspinally integrated) sexual motivation and libido is discussed in Dekeyne et al. (2012).
Antagonist Properties at 5-HT2B Receptors.
5-HT2B receptors are closely related to their 5-HT2C counterparts in primary structure, ligand binding profiles, and coupling pathways (Porter et al., 1999; Cussac et al., 2002, 2008; Millan et al., 2008). However, a few drugs selectively discriminate these sites (Porter et al., 1999; Cussac et al., 2002, 2008; Millan et al., 2008), so it is interesting that S32212 displayed approximately 20-fold lower affinity for 5-HT2B versus 5-HT2CINI receptors. S32212 was likewise an antagonist at 5-HT2B sites in blocking 5-HT-induced activation of Gαq and PLC (Millan et al., 2008), but its influence on other cellular signals such as phospholipase A2 and ERK1/2 remains to be characterized (Cussac et al., 2002, 2008; Millan et al., 2008; Aloyo et al., 2009). There is currently no evidence for constitutive activity at 5-HT2B receptors, and it is unlikely that they are involved in the actions of S32212 upon cortical neurones because its induction of 5-HT-elicited Ca2+ responses was occluded by SB242,084, which is selective for 5-HT2C over 5-HT2B receptors (Kennett et al., 1997; Cussac et al., 2002). In addition, CP809,101 has low affinity for 5-HT2B sites that are facilitatory to dopaminergic transmission and locomotion (Siuciak et al., 2007; Doly et al., 2008; Auclair et al., 2010), suggesting that they are not relevant to the influence of S32212 on dopaminergic (and adrenergic) pathways. More generally, any functional relationship of cerebral 5-HT2B receptors to mood and cognition is unclear, and they are unlikely to be involved in the influence of S32212 on depressed states (Dekeyne et al., 2012).
Antagonist Properties at 5-HT2A Receptors.
By contrast to 5-HT2B receptors, 5-HT2A receptors are enriched in the central nervous system where they fulfill several important roles, both distinct from and similar to their 5-HT2C counterparts (Millan, 2006; Landholt and Wehrle, 2009; Carr and Lucki, 2011). S32212 abolished 5-HT2A receptor-mediated Gαq and PLC activation, which is a major transduction pathway at 5-HT2A receptors (Millan et al., 2008). Although data from other cascades are awaited (Berg et al., 2008; Millan et al., 2008), confirming its antagonist properties at cerebral (frontal cortex) 5-HT2A receptors, S32212 blocked induction of head twitches by the 5-HT precursor 5-HTP (Pandey et al., 2010). The prototypical agonist DOI elicits a well characterized discriminative stimulus mediated by 5-HT2A sites (Schreiber et al., 1994), so its blockade by S32212 can also be attributed to the antagonism of 5-HT2A receptors. As mentioned in the Introduction and discussed in Dekeyne et al. (2012), a potentially beneficial contribution of 5-HT2A receptor blockade to the influence of S32212 on mood and sleep should be borne in mind.
Antagonist Properties at α2-AR Subtypes.
S32212 interacted with all three classes of human α2-AR, the hα2A-ARs subtype being orthologous to rat “α2D-ARs” (Kable et al., 2000) and, in [35S]GTPγS binding studies of the total population of G proteins recruited (Audinot et al., 2002), S32212 behaved as an antagonist at each. By analogy to 5-HT2C receptors, α2-ARs couple to multiple species of G protein, which in turn linked to diverse transduction pathways (Small et al., 2000). Hence, it is important that antagonist properties of S32212 at hα2A and hα2C-ARs, which fulfill major roles in the control of mood, cognition, and monoaminergic transmission (Kable et al., 2000; Millan et al., 2000), were reproduced by using antibodies specific for Gαi3 and Gαο, the major G-protein isoforms recruited by α2-ARs (Wise et al., 1997; Small et al., 2000; Audinot et al., 2002). Furthermore, S32212 also blocked the inhibitory influence of hα2A and hα2C-ARs upon adenylyl cyclase as determined by a homogeneous time-resolved fluorescence measure of cAMP production (Wise et al., 1997; Audinot et al., 2002). Activation of Gαi, together with engagement of β-arrestin and the Ras-Raf pathway, leads to the phosphorylation of ERK1/2, a highly sensitive measure of even weak agonist properties (Wang et al., 2006), and this procedure confirmed the “pure” antagonist properties of S32212 at α2A and α2C-ARs. Although data are limited, it has been proposed that α2-ARs can display constitutive activity (Murrin et al., 2000; Pauwels and Tardif, 2002). Accordingly, in future work, it would be interesting to determine whether S32212 is an inverse agonist or neutral antagonist at α2-AR subtypes.
Irrespective of this issue, cerebral α2-ARs controlling neurotransmitter release show high “spontaneous” activity caused by pronounced, tonic release of NA (Millan et al., 2000b; Aloyo et al., 2009; Di Giovanni et al., 2010), and antagonist actions of S32212 at central α2-ARs were revealed in neurochemical and other studies discussed in Dekeyne et al., 2012. Moreover, consistent with antagonist properties at cerebral α2A-ARs, S32212 blocked a discriminative stimulus elicited by the selective α2-AR agonist S18616 (Dekeyne and Millan, 2006).
Concluding Comments.
To summarize, the novel urea derivative S32212 possesses a multitarget profile distinct from other antidepressant agents. Its inverse agonist properties at 5-HT2C receptors together with blockade of α2-ARs (plus a potential contribution of 5-HT2A receptor antagonism) is consistent with a beneficial influence on mood, cognition, sleep, and sexual function. Moreover, the low affinities of S32212 for α1A-ARs, muscarinic M1 receptors, and histamine H1 receptors suggest, in contrast to several other antidepressants, that its positive actions should be expressed in the absence of autonomic-cardiovascular and other side effects. The actions of S32212 in preclinical models of potential therapeutic utility relevant to the treatment of depression are consistent with this reasoning and are described in Dekeyne et al. (2012).
Authorship Contributions
Conducted experiments: Mannoury la Cour, Chanrion, Dupuis, Di Cara, Audinot, Cussac, Newman-Tancredi, Kamal, Boutin, Jockers, Marin, Bockaert, Muller, Dekeyne, and Lavielle.
Wrote or contributed to the writing of the manuscript: Millan and Dekeyne.
Acknowledgments
We thank Loretta Iob, Dorothée Sicard, Valérie Pasteau, Manuelle Touzard, and Christine Chaput for technical assistance and Marianne Soubeyran for preparation of the manuscript.
Footnotes
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
-
ABBREVIATIONS:
- MT
- melatonin
- 5-HT
- serotonin
- h5-HT
- human 5-HT
- 5-HTP
- 5-hydroxytryptophan
- AR
- adrenoceptor
- h-AR
- human AR
- CHO
- Chinese hamster ovary
- HEK
- human embryonic kidney
- DOI
- 1-(2,5-dimethoxy-4-iodophenyl)-2-aminopropane
- ERK
- extracellular-regulated kinase
- [35S]GTPγS
- guanosine-5′-O-(3-[35S]thio)-triphosphate
- IP
- inositol phosphate
- NA
- noradrenaline
- PI
- phosphatidyl inositol
- PLC
- phospholipase C
- SPA
- scintillation proximity assay
- LiCl
- lithium chloride
- VTA
- ventral tegmental area
- YFP
- yellow fluorescent protein
- ANOVA
- analysis of variance
- ELISA
- enzyme-linked immunosorbent assay
- Rluc
- Renilla luciferase
- BRET
- bioluminescence resonance energy transfer
- PBS
- phosphate-buffered saline
- BSA
- bovine serum albumin
- S32212
- N- [4-methoxy-3-(4-methylpiperazin-1-yl)phenyl]-1,2-dihydro-3-H-benzo[e]indole-3-carboxamide
- SB242,084
- 6-chloro-5-methyl-N-[6-(2-methylpyridin-3-yloxy)pyridin-3-yl]indoline-1-carboxamide
- S18616
- (S)-spiro[(1-oxa-2-amino-3-azacyclopent-2-ene)-4, 2′-(1′,2′,3′,4′-tetrahydronaphthalene)]
- CP809,101
- 2-(3-chlorobenzyloxy)-6-(1-piperazinyl)pyrazine
- RX821,002
- {2-(2-methoxy-1,4-benzodioxan-2-yl)-2-imidazoline}
- WAY100,635
- N-(2-(4-(2-methoxyphenyl)-1-piperazinyl)ethyl)-N-2-pyridinyl-trihydrochloride
- GR125,743
- N-[4-methoxy-3-(4-methylpiperazin-1-yl)phenyl]-3-methyl-4-(4-pyridyl)benzamide
- BRL43,694
- N-(9-methyl-9-azabicyclo[3.3.1.]non-3-yl-1-methyl-1H-indazol-3-carboxamide
- GR113,808
- 1-methyl-1H-indole-3-carboxylic acid 1-[2-(methylsulfonamido)ethyl]piperidin-4-ylmethyl ester
- LSD
- d-lysergic acid diethylamide
- CGP12177
- 4-[3-[(1,1-dimethylethyl)amino]2-hydroxypropoxy]-1,3-dihydro-2H-benzimidazol-2-one
- SCH23390
- (R)-(+)-7-chloro-8-hydroxy-3-methyl-1-phenyl-2,3,4,5-tetrahydro-1H-3-benzazepine
- GBR12935
- 1-(2-diphenylmethoxyethyl)-4-(3-phenylpropyl)piperazine
- Ro41-1049
- N-(2-aminoethyl)-5-(3-fluorophenyl)-4-thiazolecarboxamide
- Ro19-6327
- N-(2-aminoethyl)-5-chloro-pyridine-2-carboxamide
- AFDX384
- N-[2-[2-[(dipropylamino)methyl]-1-piperidinyl]ethyl]-5,6-dihydro-6-oxo-11H-pyrido[2,3-b][1,4]benzodiazepine-11-carboxamide.
- Received September 30, 2011.
- Accepted November 29, 2011.
- Copyright © 2012 by The American Society for Pharmacology and Experimental Therapeutics