Abstract
Previous studies have shown that human 5-hydroxytryptamine (5-HT)1A receptors stably expressed in transfected cell lines show constitutive G-protein activity, as revealed by the inhibitory effect of inverse agonists, such as spiperone, on basal guanosine-5′-O-(3-[35S]thio)-triphosphate ([35S]GTPγS) binding. In the present study, we evaluated the constitutive activity of native rat 5-HT1A receptors in hippocampal membranes. Using anti-Gαo-antibody capture coupled to scintillation proximity assay under low sodium (30 mM) conditions, we observed high basal [35S]GTPγS binding to Gαo subunits (defined as 100%). Under these conditions, 5-HT and the prototypic selective 5-HT1A agonist (+)8-hydroxy-2-(di-n-propylamino)tetralin [(+)-8-OH-DPAT] both stimulated [35S]GTPγS binding to Gαo to a similar extent, raising binding to approximately 130% of basal with pEC50 values of 7.91 and 7.87, respectively. The 5-HT1A-selective neutral antagonist [O-methyl-3H]-N-(2-(4-(2-methoxyphenyl)-1-piperazinyl)ethyl)-N-(2-pyridinyl)cyclohexanecarboxamide trihydrochloride (WAY100,635) could block these effects in a competitive manner with pKb values (5-HT, 9.57; (+)-8-OH-DPAT, 9.52) that are consistent with its pKi value at r5-HT1A receptors (9.33). In this native receptor system, spiperone and methiothepin reduced basal [35S]GTPγS binding to Gαo in a concentration-dependent manner to 90% of basal with pIC50 values of 7.37 and 7.98, respectively. The inhibition of basal [35S]GTPγS binding induced by maximally effective concentrations of spiperone (10 μM) or methiothepin (1 μM) was antagonized by WAY100,635 in a concentration-dependent manner (pKb, 9.52 and 8.87, respectively), thus indicating that this inverse agonism was mediated by 5-HT1A receptors. These data provide the first demonstration that native rat serotonin 5-HT1A receptors can exhibit constitutive activity in vitro.
Constitutive activity of G-protein-coupled receptors (GPCR) provides a mechanistic basis for inverse agonism, a phenomenon observed in heterologous expression systems in which some pharmacological agents are able to inhibit basal activity, as measured in second messenger system assays (de Ligt et al., 2000; Kenakin, 2004). This phenomenon has been described for several GPCRs, and a number of clinically relevant drugs have been shown to act as inverse agonists on some GPCR (Milligan, 2003a; Kenakin, 2004). Constitutive activity has been described in a number of recombinant systems and in “physiological” peripheral tissue preparations (de Ligt et al., 2000) but only rarely in native brain tissue; among the few exceptions are serotonin 5-HT2C receptors (De Deurwaerdere et al., 2004) and H3 histaminergic receptors (Morisset et al., 2000). Moreover, there is evidence that mutated GPCRs with elevated constitutive activity are associated with some human diseases (de Ligt et al., 2000; Seifert and Wenzel-Seifert, 2002; Kenakin, 2004). A better characterization of this phenomenon may thus improve our understanding of the mechanisms of action of clinically important drugs and may also help improve future drug development.
Serotonin 5-HT1A receptors, which are members of the GPCR family, are key targets for the treatment of mood disorders, including anxiety and depression (Barnes and Sharp, 1999) and may also improve the outcome of psychotic disorders (Millan, 2000; Meltzer et al., 2003; Newman-Tancredi et al., 2005). Several reports have shown that human (h)5-HT1A receptors expressed in stably transfected cell lines showed constitutive activity, as revealed by the inverse agonist properties of spiperone and methiothepin in that heterologous expression system (Newman-Tancredi et al., 1997a,b; Stanton and Beer, 1997; McLoughlin and Strange, 2000). The effects of spiperone were concentration-dependent and could be blocked by the selective neutral 5-HT1A receptor antagonist WAY100,635. Likewise, h5-HT1A receptor constitutive activity could be demonstrated on a specific coupling to GTP-binding protein Gz when this particular G-protein was coexpressed with h5-HT1A receptor in Spodoptera frugiperda (Sf9) cell line (Barr and Manning, 1997) and on Gαi3 activation in Chinese hamster ovary cells expressing h5-HT1A receptors (Newman-Tancredi et al., 2002). Constitutive activity of 5-HT1A receptors could also be demonstrated when the receptor was coupled to Gαi1 in human embryonic kidney cells using a fusion protein paradigm (Milligan et al., 2001). In light of these observations, several groups have evaluated constitutive activity of 5-HT1A receptors in brain tissue environment (Odagaki and Fuxe, 1995; Alper and Nelson, 1998; Newman-Tancredi et al., 2003b; Odagaki and Toyoshima, 2005a,b). However, constitutive activity of native 5-HT1A receptors has not been described to date.
Recent advances in [35S]GTPγS binding assays have improved the selectivity of this technique for studying GPCR (Milligan, 2003b). Scintillation proximity assay (SPA) can be coupled with Gα-selective antibodies in [35S]GTPγS binding assays, allowing detection of the response of a single Gα-protein subtype to pharmacological manipulation within a homogenate (Newman-Tancredi, 2003; Wu and Liu, 2005). The rat hippocampus contains high densities of 5-HT1A receptors (Barnes and Sharp, 1999) that may be essentially coupled to Gαo (Mannoury la Cour et al., 2006). In the present study, we characterized the effect of selected 5-HT1A receptor drugs on a SPA-based Gαo-selective immunocapture assay using two agonists [serotonin and (+)-8-OH-DPAT] and two inverse agonists (spiperone and methiothepin). These results indicate that 5-HT1A receptors constitutively activate Gαo subunits in rat brain tissue homogenate.
Materials and Methods
Animal Handling. Animals were handled and cared for in accordance with the Guide for the Care and Use of Laboratory Animals (Institute of Laboratory Animal Resources, 1996) and the European Directive EEC/86/609, and the protocol was carried out in compliance with French regulations and with local Ethics Committee Guidelines for animal research.
Receptor Affinities in Rat Hippocampal Membranes. Frozen hippocampi of male rats [Ico: OFA SD (I.O.P.S. Caw); Iffa Credo, L'Arbresele, France] weighing 240 to 260 g were used for binding studies and were stored at –80°C before use. The membrane preparations and binding assays were performed in duplicate as described previously (Assié et al., 1999). In brief, membranes were homogenized in 20 volumes of ice-cold Tris-HCl buffer (50 mM, pH 7.4 at 25°C), washed with two cycles of centrifugation/resuspension, followed by a 10-min incubation at 37°C, and ending with a final cycle of centrifugation/resuspension. Membranes (750 μg of tissue/well) were incubated at 23°C in microplates with 96 deep wells, each containing 1 nM [3H]8-OH-DPAT, and increasing concentrations of drugs ranging from 10–11Mto10–5M. After a 2-h incubation period, the microplate contents were rapidly filtered under vacuum through GF/B unifilter microplates with two rapid 2-ml washes with ice-cold Tris-HCl buffer. Nonspecific binding was defined with 10 μM 5-HT, whereas total binding was defined in the absence of any displacer. The radioactivity retained on the filters was measured by scintillation spectroscopy (TopCount; PerkinElmer Life and Analytical Sciences, Boston, MA) in 40 μl of scintillation fluid (Microscint 20; PerkinElmer Life and Analytical Sciences). The results of displacement experiments were analyzed using the nonlinear regression program Prism ver. 4.03 (GraphPad Software, San Diego, CA), which allowed calculation of an IC50. Ki were calculated using the Cheng-Prusoff equation (Cheng and Prusoff, 1973): Ki = IC50/(1 – ([ligand]/KDligand)) where [ligand] is the measured concentration of [3H](+)-8-OH-DPAT in the assay and KDligand is the affinity of [3H](+)-8-OH-DPAT in that assay (0.75 ± 0.08 nM). Density of 5-HT1A receptors in that assay was 14.3 ± 0.8 fmol/mg of fresh tissue (corresponding to approximately 300 fmol/mg of protein).
Scintillation Proximity Assay. Hippocampi from male rats [Ico: OFA SD (I.O.P.S. Caw); Iffa Credo] weighing 240 to 260 g were dissected on a cold plate and frozen at –80°C until used. Tissues were homogenized in ice-cold HEPES-DTT buffer (20 mM HEPES, pH 7.0, 0.2 mM EDTA, and 0.2 mM DTT) with 1 mM GTP using a Polytron homogenizer (Kinematica, Basel, Switzerland), GTP favoring endogenous ligand dissociation. The homogenate was incubated at 35°C for 15 min. Membranes were washed by two cycles of centrifugation at 20,000g for 15 min at 4°C and resuspension in ice-cold HEPES-DTT with 1 mM GDP. Final pellet was resuspended in 300 volumes (based on tissue weight) HEPES-DTT buffer containing 5 mM MgCl2, 100 mM NaCl (Gα-selective antibodies assay) or 30 mM NaCl (pharmacological assay), and 50 μM GDP (HEPES-DTT assay buffer).
All reactions were performed at room temperature on 96-well plates in a final volume of 200 μl of HEPES-DTT assay buffer. Membranes (8 μg of protein/well) were preincubated for 30 min at room temperature with drugs (agonist ± antagonist), buffer (to define basal) or 10 μM GTPγS (to define nonspecific). At the end of this preincubation, 0.4 nM [35S]GTPγS was added, and the membranes were incubated for 60 min. Incubation was stopped by adding Nonidet P-40, and the plate was agitated for another 30 min before addition of 0.2 μg of anti-Gα-selective antibodies to each well. The antibodies used were rabbit polyclonal anti-Gαq/11, anti-Gαi3, and anti-Gαs/olf from Santa Cruz Biotechnology (Santa Cruz, CA) or mouse monoclonal anti-Gαo and anti-Gαi1 from BIOMOL Research Laboratories (Plymouth Meeting, PA). Anti-Gαq/11, anti-Gαi1, and anti-Gαs/olf have been characterized previously by Western blot using purified recombinant Gαi1,Gαi2,Gαi3,Gαo,Gαq/11, and Gαs/olf protein subtypes (Cussac et al., 2002, 2004) and are highly selective for their respective G-protein subtypes. As described by the manufacturer, we observed cross-reactivity of anti-Gαi3 with Gαi1 and Gαi2, with a stronger reaction on Gαi3 (data not shown). Specificity of anti-Gαo was assessed by western blot using 50 ng of purified recombinant Gαi1,Gαi2,Gαi3,Gαo,Gαq/11, and Gαs/olf protein subtypes as described previously (Cussac et al., 2002, 2004). Anti-Gαo are highly selective for Gαo (Fig. 1A). Note that the basal signal for each antibody will depend on the affinity and selectivity of these antibodies for the G-protein subtype(s) and the revelation system (affinity of secondary antibodies) and basal signal therefore cannot be compared between antibodies.
The primary antibodies were left to react for 60 min under agitation before adding 50 μl of the secondary antibodies (anti-rabbit or anti-mouse coupled to scintillation proximity assay beads; GE Healthcare, Little Chalfont, Buckinghamshire, UK), diluted according to manufacturer's recommendations. The secondary antibodies were left to react for another 60 min, and the plate was centrifuged at 1000g for 15 min at room temperature to take the complex down at the bottom of the well, and radioactivity was immediately measured on a TopCount radioactivity counter (PerkinElmer Life and Analytical Sciences). Raw disintegrations-per-minute data were transformed by subtracting nonspecific and normalizing to percentage of basal [35S]GTPγS binding. All pharmacological parameters were derived from sigmoid nonlinear regression using Prism version 4.03. For antagonism by WAY100,635, pKb were calculated using the Cheng-Prusoff equation (Kb = IC50/(1 – ([Ago]/EC50Ago))), where [Ago] is the fixed concentration of agonist used for the antagonism assay and EC50Ago is the EC50 of that agonist when tested alone.
Results
The Gα protein subtypes activated by 5-HT1A in rat hippocampus were evaluated using SPA-based antibody capture assay with a set of antibodies selective for different Gα-proteins (see Materials and Methods for details). Table 1 lists the basal [35S]GTPγS binding (in disintegrations per minute) observed with the various antibodies, and the absolute (disintegrations per minute) and relative (percentage of basal) changes elicited by 10 μM (+)8-OH-DPAT. The maximal basal [35S]GTPγS binding was observed with the anti-Gαo-selective antibodies (Fig. 1A), and these antibodies also showed the maximal absolute (+4607 DPM) and relative (+71%) increases after (+)8-OH-DPAT treatment (Table 1). A moderate (+46%) increase in relative [35S]GTPγS binding was also observed with anti-Gαi3 antibodies (Table 1), but absolute [35S]GTPγS binding represented only a minor fraction (+371 DPM, or 8% of the response observed with Gαo on a disintegrations-per-minute basis; Table 1). Because of the cross-reactivity of anti-Gαi3 antibodies with Gαi1 and Gαi2, this increase may be due predominantly to Gαi2 and/or Gαi3 proteins, because no change of [35S]GTPγS binding was seen with Gαi1-selective antibodies (Table 1), thus demonstrating that 5-HT1A receptors do not couple to Gαi1 in rat hippocampus. We therefore used the Gαo-selective antibodies for investigating constitutive activity of 5-HT1A receptors in rat hippocampal membrane preparations.
Low sodium concentrations (30 mM) and relatively high GDP concentrations (50 μM) were used to detect 5-HT1A receptor constitutive activity in this system. Under those conditions, basal [35S]GTPγS labeling of Gαo was approximately 25,000 dpm, and this basal value was defined as 100% (Figs. 1 and 2). Under these assay conditions, serotonin, the endogenous ligand for 5-HT1A receptors, and the prototypical selective 5-HT1A receptor agonist (+)8-OH-DPAT both increased [35S]GTPγS binding to Gαo in a concentration-dependent manner (Fig. 1 and Table 2). The selective neutral 5-HT1A antagonist WAY100,635 competitively reversed this effect; increasing concentration of the antagonist reversed the effects of maximally effective concentrations (1 μM) of serotonin or (+)8-OH-DPAT (Fig. 2 and Table 2).
In this hippocampal tissue homogenate preparation, spiperone and methiothepin showed inverse agonist properties, reducing basal [35S]GTPγS binding to Gαo in a concentration-dependent manner (Fig. 1 and Table 2). WAY100,635 partly reversed the maximal inhibition of basal binding induced by a fixed concentration of spiperone (10 μM) or methiothepin (1 μM) (Fig. 2 and Table 2). In comparison, when a higher concentration of NaCl (100 mM) was used, spiperone and methiothepin inhibited [35S]GTPγS binding only at very high concentrations (> 1 μM; data not shown).
Discussion
Human 5-HT1A receptors expressed in recombinant cell lines have been shown to possess constitutive activity, as revealed by spiperone and methiothepin's inverse agonist properties (see Introduction). Likewise, constitutive activity of h5-HT1A could be demonstrated when the receptor was coexpressed with GTP-binding protein Gz in Sf9 cells (Barr and Manning, 1997) and in human embryonic kidney cells expressing a fusion protein combining 5-HT1A to Gαi1 (Milligan et al., 2001), thus suggesting that constitutive activity is an inherent property of 5-HT1A receptors. Nevertheless, multiple factors can influence detection of constitutive activity. For example, the amplitude of response to inverse agonist depends on GPCR/G-protein stoichiometry (i.e., an increase in 5-HT1A/G-protein ratio increasing the amplitude of inverse agonist response) (Newman-Tancredi et al., 1997b). It is important therefore to assess constitutive activity of GPCR in an environment in which the “natural” GPCR/G-protein ratio is expressed.
The present study is, to our knowledge, the first to demonstrate constitutive activity of native rat 5-HT1A receptors. Previous studies using classic [35S]GTPγS binding assay failed to detect inverse agonist properties of spiperone and methiothepin in rat brain tissue (Alper and Nelson, 1998; Newman-Tancredi et al., 2003b; Odagaki and Toyoshima, 2005a,b). Alper and Nelson (1998) did report that methiothepin, but not spiperone, diminished G-protein activation at high concentrations (10 μM), but this effect was not reversed by the selective neutral antagonist WAY100,635, indicating that it was not related to 5-HT1A receptors. In contrast, using SPA-based Gαo-selective immunocapture on [35S]GTPγS binding assay, the present study demonstrates inverse agonism at native r5-HT1A receptors with both spiperone and methiothepin that was antagonized by WAY100,635. Rat hippocampal 5-HT1A receptors show a strong coupling to Gαo (Mannoury la Cour et al., 2006). Indeed, in the present study, comparison of antibodies selective for various Gα proteins using SPA-based immunocapture assay showed that the most prominent response of rat hippocampal 5-HT1A receptors to (+)8-OH-DPAT was activation of Gαo. The SPA-based Gαo-selective immunocapture approach used here thus permitted characterization of the inverse agonist properties of spiperone and methiothepin on 5-HT1A receptors by selectively detecting the G-protein showing the best coupling to 5-HT1A receptors in this tissue. Under the present assay conditions, WAY100,635 had no activity of its own at concentrations up to 10 μM, but it antagonized the effect of the agonists and inverse agonists in a competitive manner. Calculated pKb for this antagonism are consistent with pKi values derived from competition binding of WAY100,635 to rat 5-HT1A receptors (Newman-Tancredi et al., 2005) and from pKb previously published on similar assays (Odagaki and Toyoshima, 2005a,b). These data therefore confirm that the decreases in Gαo labeling induced by spiperone and methiothepin are indeed mediated by interaction at r5-HT1A receptors.
The amplitude of the inverse agonism response observed with spiperone and methiothepin is relatively modest (10% reduction in basal labeling of Gαo or 33% of the amplitude of agonists responses under those conditions). Two factors may contribute to the moderate inverse agonism response of these drugs on 5-HT1A receptors. First of all, both drugs may be partial inverse agonists on 5-HT1A receptors. Although no efficacy data are available for these two drugs as inverse agonists on rat 5-HT1A receptors, this efficacy may be assessed in recombinant systems expressing these receptors alone using a [35S]GTPγS homologous displacement protocol with unlabeled GTPγS (Audinot et al., 2000; Rouleau et al., 2002). Second, because this assay is performed on tissue, several Gαo-coupled receptors may be constitutively active, thus limiting the contribution of 5-HT1A to a fraction of the overall constitutive activity present in that tissue. Indeed, reversal of the response of spiperone and methiothepin by WAY100,635 is only partial (two thirds of effect), suggesting that other constitutively active receptors may be involved in the effect of these two drugs (see below).
The demonstration that native r5-HT1A receptors show constitutive activity in vitro suggests that this phenomenon also occurs in vivo. Constitutive activity has also been demonstrated in rat brain tissues for serotonin 5-HT2C receptors (De Deurwaerdere et al., 2004) and H3 histaminergic receptors (Morisset et al., 2000), and certain human pathological conditions, including metabolic diseases and some cancers, are believed to be associated with abnormal levels of GPCR constitutive activity (Kenakin, 2004). It is also possible that certain neuropsychiatric diseases may be associated with abnormal constitutive activity of some GPCRs, and several clinically relevant drugs have been shown to have inverse agonist properties on GPCRs in recombinant systems (Milligan, 2003a; Kenakin, 2004). The present study assessed the level of constitutive activity of 5-HT1A receptors in hippocampal membranes prepared from brains of untreated normal animals. In view of the importance of 5-HT1A receptors in neuropsychiatric disorders (Millan, 2000; Meltzer et al., 2003; Newman-Tancredi et al., 2005), it will be of interest to evaluate whether constitutive activity of central 5-HT1A receptors may be affected by pharmacological or pathophysiological factors. A better understanding of the factors affecting 5-HT1A receptors constitutive activity in vivo may thus help improve therapeutic approaches toward neuropsychiatric disorders.
The pharmacological behavior of inverse agonists in vivo may have other physiological consequences. For example, inverse agonists may produce more prominent receptor upregulation than neutral antagonists (Adan and Kas, 2003), and inverse agonists may lead to increases in G-protein expression levels (Kenakin, 2004). A recent electrophysiological study assessed the effect of two selective 5-HT1A inverse agonists (Rec 27/0224 and Rec 27/0074) on hippocampal and dorsal raphe neurons (Corradetti et al., 2005). In this physiological assay, the two inverse agonists fully antagonized the effect of 5-CT applications on dorsal raphe neurons but with a much slower time course to reach steady state than would have been expected from their binding affinities at 5-HT1A receptors, or compared with the time course to reach steady state with the neutral antagonist WAY100,635. Corradetti et al. (2005) suggested that this phenomenon may be explained by a slow allosteric shift of the receptor toward an inactive state. Moreover, in contrast to WAY100,635, which could fully antagonize 5-CT-induced hyperpolarization of hippocampal CA1 neurons, these two compounds showed only partial antagonism, suggesting that these drugs may behave differentially on neuronal populations expressing either pre- or postsynaptic 5-HT1A receptors. These observations are consistent with the differential coupling of the pre-(dorsal raphe) and postsynaptic (hippocampus) 5-HT1A receptors suggested by various observations (Hensler, 2002; Mannoury la Cour et al., 2006) as differential response to allosteric effects of inverse agonists may be expected on differentially coupled receptor.
Three technical points should be noted concerning the data in the present study. First, buffer containing low sodium (30 mM) was used because this condition increases GPCR constitutive activity and may therefore improve identification of inverse agonists with [35S]GTPγS binding assays (de Ligt et al., 2000). The role of sodium may be to allosterically stabilize the uncoupled (inactive) conformation of the receptors (de Ligt et al., 2000), thus low sodium concentrations favoring the constitutively active conformation of the receptor. Allosteric modulation of GPCR ligand binding by sodium is associated with a highly conserved aspartate residue located in the second transmembrane domain, near the intracytoplasmic side of GPCR (Horstman et al., 1990), a residue that may also be critical for GPCR coupling to G-proteins (Odagaki and Toyoshima, 2005a). An important issue is clearly the “physiological” concentration of NaCl affecting the levels of GPCR constitutive activity, and this question remains a subject of debate (Newman-Tancredi et al., 2003a).
Second, it is unlikely that the inverse agonism of methiothepin and spiperone observed here would be due to antagonism of residual endogenous serotonin in the tissue because the homogenate was preincubated for 15 min at 35°C in a large excess of buffer containing GTP to favor endogenous ligand dissociation, followed by several membrane washes. Moreover, WAY100,635 had no activity at concentrations up to 10 μM in these assay conditions, providing further indication that no remaining endogenous 5-HT was affecting basal [35S]GTPγS binding.
Third, the inverse agonism elicited by spiperone and methiothepin was not completely reversed by WAY100,635; a residual lowering of Gαo activation persisted even at high antagonist concentrations. This suggests that other receptor systems that couple to Gαo may also be involved. Indeed, both spiperone and methiothepin are known to interact with other receptors at which they have been shown to be inverse agonists. For example, spiperone acts as an inverse agonist on dopamine D2 receptors (Roberts and Strange, 2005) and on α1 adrenoceptors (Rossier et al., 1999), whereas methiothepin has inverse agonist properties on 5-HT1B (Newman-Tancredi et al., 2003a) and 5-HT1D receptors (Audinot et al., 2000). Thus, the inhibition of [35S]GTPγS binding to Gαo remaining with maximal WAY100,635 concentrations may be due to an action at other receptors. The present method targeting Gαo activation provides a strategy to investigate the action of inverse agonists at other receptor subtypes.
In conclusion, these data generated by antibody capture methodology associated with SPA detection provide the first demonstration that native rat 5-HT1A receptors show constitutive activation of Gαo proteins in an hippocampal tissue environment, suggesting that rat 5-HT1A receptors may be constitutively active in vivo at Gαo and/or other G-protein subtypes.
Acknowledgments
We greatly appreciate the excellent technical assistance of Nathalie Danty with the receptor binding assays, and Fabrice Lestienne with the Western blots.
Footnotes
-
This work has been previously presented in abstract form: Martel JC, Ormiere AM, Assie MB, Cussac D, and Newman-Tancredi A (2005) Spiperone and methiothepin act as inverse agonists on native rat hippocampal 5-HT1A receptors coupled to Gαo-protein. Soc Neurosci Abstr31:373.
-
Article, publication date, and citation information can be found at http://molpharm.aspetjournals.org.
-
doi:10.1124/mol.106.029769.
-
ABBREVIATIONS: GPCR, G-protein-coupled receptor; WAY100,635, [O-methyl-3H]-N-(2-(4-(2-methoxyphenyl)-1-piperazinyl)ethyl)-N-(2-pyridinyl)cyclohexanecarboxamide trihydrochloride; [35S]GTPγS, guanosine-5′-O-(3-[35S]thio)-triphosphate; (+)8-OH-DPAT, (+)8-hydroxy-2-(di-n-propylamino)tetralin; SPA, scintillation proximity assay; 5-HT, 5-hydroxytryptamine; DTT, dithiothreitol.
- Received August 10, 2006.
- Accepted December 13, 2006.
- The American Society for Pharmacology and Experimental Therapeutics