Abstract
S 16924 showed a pattern of interaction at multiple (>20) native, rodent and cloned, human (h) monoaminergic receptors similar to that of clozapine and different to that of haloperidol. Notably, like clozapine, the affinity of S 16924 for hD2 and hD3 receptors was modest, and it showed 5-fold higher affinity for hD4 receptors. At each of these sites, using a [35S]GTPγS binding procedure, S 16924, clozapine and haloperidol behaved as antagonists. In distinction to haloperidol, S 16924 shared the marked affinity of clozapine for h5-HT2Aand h5-HT2C receptors. However, an important difference to clozapine (and haloperidol) was the high affinity of S 16924 for h5-HT1A receptors. At these sites, using a [35S]GTPγS binding model, both S 16924 and clozapine behaved as partial agonists, whereas haloperidol was inactive. In vivo, the agonist properties of S 16924 at 5-HT1Aautoreceptors were revealed by its ability to potently inhibit the firing of raphe-localized serotoninergic neurones, an action reversed by the selective 5-HT1A receptor antagonist, WAY 100,635. In contrast, clozapine and haloperidol only weakly inhibited raphe firing, and their actions were resistant to WAY 100,635. Similarly, S 16924 more potently inhibited striatal turnover of 5-HT than either clozapine or haloperidol. Reflecting its modest affinity for D2 (and D3) autoreceptors, S 16924 only weakly blocked the inhibitory influence of the dopaminergic agonist, apomorphine, upon the firing rate of ventrotegmental area-localized dopaminergic neurones. Further, S 16924 only weakly increased striatal, mesolimbic and mesocortical turnover of dopamine (DA). Clozapine was, similarly, weakly active in these models, whereas haloperidol, in line with its higher affinity at D2 (and D3) receptors, was potently active. In the frontal cortex (FCX) of freely moving rats, S 16924 dose-dependently reduced dialysate levels of 5-HT, whereas those of DA and NAD were dose-dependently increased in the same samples. In contrast, although S 16924 also suppressed 5-HT levels in the striatum and nucleus accumbens, DA levels therein were unaffected. Clozapine mimicked this selective increase in DA levels in the FCX as compared to striatum and accumbens. In contrast, haloperidol modestly increased DA levels in the FCX, striatum and accumbens to the same extent. In distinction to S 16924, clozapine and haloperidol exerted little influence upon 5-HT levels. Finally, the influence of S 16924 upon FCX levels of 5-HT, DA (and NAD) was attenuated by WAY 100,635. In conclusion, S 16924 possesses a profile of interaction at multiple monoaminergic receptors comparable to that of clozapine and distinct to that of haloperidol. In addition, S 16924 is a potent, partial agonist at 5-HT1A receptors. Correspondingly, acute administration of S 16924 decreases cerebral serotoninergic transmission and selectively reinforces frontocortical as compared to subcortical dopaminergic transmission. In line with these actions, S 16924 shows a distinctive profile of activity in functional (behavioral) models of potential antipsychotic activity (companion paper).
Classical neuroleptics, such as haloperidol, control the positive symptoms of schizophrenia (hallucinations, delusions, etc.) via the blockade of limbic D2 receptors targeted by hyperactive mesolimbic dopaminergic pathways (Holcomb et al., 1996; Kahn and Davis, 1995). However, neuroleptics are poorly effective against negative-cognitive symptoms, such as mutism and blunted affect. These symptoms reflect a disruption in the activity of mesocortical dopaminergic pathways and, more generally, a perturbation in the function of the prefrontal cortex and FCX, commonly termed “hypofrontality” (Jentsch et al., 1997; Knable and Weinberger, 1997). Indeed, neuroleptics may exacerbate negative symptoms by blocking FCX-localized D2 receptors and provoking an extrapyramidal syndrome. An additional disadvantage of neuroleptics is that a substantial population of patients do not respond satisfactorily to their administration (Kane and Freeman, 1994). Further, neuroleptics induce a pronounced hyperprolactinemia and associated endocrinological disorders by antagonism of tonically active D2 receptors on hypophyseal lactotrophs, and a marked extrapyramidal motor syndrome due to blockade of D2receptors in the basal ganglia (Cunningham-Owens, 1996). Finally, long-term treatment with neuroleptics may ultimately result in the emergence of tardive dyskinesias, an irreversible motor problem likely related to striatal D2 receptor blockade, although its precise origin remains uncertain (Cunningham-Owens, 1996).
The above observations suggest that the improved treatment of schizophrenia requires drugs with characteristics different to those of typical neuroleptics and targeting sites other than, or in addition to, D2 receptors. In this respect, the dibenzodiazepine, clozapine, has attracted enormous interest inasmuch as this “atypical” antipsychotic manifests only modest affinity for D2 receptors yet is effective in a subpopulation of neuroleptic-resistant patients, improves negative symptomology, presents a benign extrapyramidal potential and does not elicit tardive dyskynesia (Kane and Freeman, 1994; Meltzer, 1995). An ongoing challenge is to identify the key receptorial interactions underlying the superior clinical profile of clozapine and, in this regard, numerous hypotheses have been formulated (Brunello et al., 1995; Kinon and Lieberman, 1996). These include: equilibrated antagonist activity at D1 and D2 receptors (Gerlach and Hansen, 1992); preferential antagonist activity at D4vs. D2 receptors (Seeman et al., 1997) and pronounced antagonist activity at adrenergic (AR) receptors (Baldessarini et al., 1992). In addition, a convincing body of evidence points to the importance of 5-HT2A and, possibly, 5-HT2C receptors: these are concentrated in corticolimbic regions and the basal ganglia, are involved in the modulation of mood and motor behavior and modulate the activity of dopaminergic pathways (Brunello et al., 1995;Casey, 1993; Kelland and Chiodo, 1996; Kennett et al., 1997;Roth and Meltzer, 1995; Schmidt and Fadayel, 1995). Thus, clozapine has marked affinity for 5-HT2C receptors, blockade of which facilitates mesocortical dopaminergic transmission (Gobert et al., 1998 and unpublished observations; Kennett et al., 1997; Pessia et al., 1994). Further, a preferential blockade of 5-HT2Avs. D2 receptors by antipsychotic drugs, such as clozapine, has been convincingly correlated with a low propensity to elicit an extrapyramidal motor syndrome (Meltzer, 1995; Roth and Meltzer, 1995; Wadenberg, 1996) and changes in the levels of 5-HT2A receptors have been documented in the FCX of schizophrenic patients (Burnet et al., 1996; Gurevich and Joyce, 1997). Alterations in 5-HT1A receptor levels have also been documented in schizophrenia and, more recently, a potential significance of actions at 5-HT1A receptors in the treatment of schizophrenia has been evoked (Burnet et al., 1996; Simpson et al., 1996) (see “Discussion”).
Notwithstanding the improved antipsychotic profile of clozapine, it cannot be considered as an ideal antipsychotic agent. First, there remains a population of patients irresponsive to clozapine, and its impact upon primary negative symptoms may be limited (Kane and Freeman, 1994; Meltzer, 1995). Second, clozapine provokes autonomic and cardiovascular side-effects via actions at nonmonoaminergic receptors, notably histaminic and muscarinic sites. In addition, a minority (∼5%) of patients display seizures, probably due to interference with central GABAergic and glutamatergic transmission (Cunningham-Owens, 1996). Third, the chemical structure of clozapine is associated with a potentially fatal agranulocytosis in 1 to 2% of patients treated (Liu and Uetrecht, 1995).
In the light of the above observations, it would clearly be of interest to develop antipsychotic agents possessing the beneficial properties of clozapine yet lacking its disadvantages. In our efforts to identify such antipsychotic drugs, we have characterized a novel benzodiozopyrrolidine, S 16924 (fig. 1). Herein, the receptorial profile of S 16924 is characterized, together with its modulation of dopaminergic, serotoninergic and adrenergic transmission in cortical, limbic and striatal regions. In the following article, the putative antipsychotic as compared to extrapyramidal properties of S 16924 are documented.
Methods
Binding.
Competition binding studies were performed at multiple dopaminergic, serotonergic and AR receptor types, as well as at DA, 5-HT and NAD reuptake sites. Assay conditions are summarized in tables 1, 2and 3 (see also Millan et al., 1995). Isotherms were analyzed by nonlinear regression, using the program “PRISM” (Graphpad Software Inc., San Diego, CA) to yield Inhibitory Concentration (IC)50 values.Ki s were derived from IC50 values according to the Cheng-Prusoff equation: Ki = IC50/(1 + L/Kd ) where L is the concentration of radioligand and Kd is the dissociation constant of the radioligand.
Measurement of agonist efficacy and antagonist potency at hD2, hD3, hD4 and h5-HT1A receptors.
Receptor-linked G-protein activation at hD2, hD3, hD4 and h5-HT1A receptors was determined by measuring the stimulation of [35S]-GTPγS (1000 Ci/mmol; NEN, Les Ulis, France) binding as described in Newman-Tancredi et al.(1997). Briefly, CHO membranes (50 μg protein) expressing the respective receptors were incubated (20 min, 22°C) with agonists and/or antagonists in a buffer containing HEPES 20 mM (pH 7.4), GDP (3 μM), MgCl2 (3 mM for hD3 and hD4, 10 mM for hD2), NaCl (100 mM for hD3 and hD4, 150 mM for hD2) and [35S]-GTPγS (0.1 nM for hD2 and hD4, 1 nM for hD3). Nonspecific binding was defined with GTPγS (10 μM). Agonist efficacy was expressed relative to that of DA or 5-HT (=100%) which were tested at maximally effective concentrations in each experiment. For antagonist studies, membranes were preincubated with antagonist and a single concentration of agonist for 30 min before the addition of [35S]-GTPγS. For concentration-response curves of the inhibition of DA-stimulated [35S]-GTPγS binding, Kb values were calculated as described in Newman-Tancredi et al. (1997). Experiments were terminated by rapid filtration through Whatman GF/B filters using a Brandel cell harvester. Radioactivity retained on the filters was determined by liquid scintillation counting. Protein concentration was determined colorimetrically using a bicinchoninic acid assay kit (Sigma Chimie, St Quentin-Fallavier, France).
In vivo studies.
Male Wistar rats (Iffa Credo, L’arbresle, France, 220–240 g body weight) were housed in sawdust-lined cages with free access to food and water. Laboratory temperature was 21 ± 1.0°C and humidity 60 ± 5%. There was a 12 hr/12 hr light-dark cycle with lights on at 7:30.
Influence upon striatal DA and 5-HT turnover.
As described in detail previously (Gobert et al., 1995a), the influence of drugs upon striatal DA and 5-HT turnover in rats was evaluated by measuring the levels of the DA precursor, DOPA, and of the 5-HT precursor, 5-HTP in the striatum 60 min after s.c. injection of drugs and 30 min after injection of the decarboxylase inhibitor, NSD 1015 (100 mg/kg, s.c.). Tissues were homogenized in 500 μl of 0.1 M HClO4 containing 0.5% Na2S2O5 and 0.5% EDTA and centrifuged at 15,000 × g for 15 min at 4°C. Supernatants were diluted in the mobile phase. HPLC analysis followed by electrochemical detection was used for determination of tissue levels of DOPA and 5-HTP. The column characteristics and elution phases were as follows: column, Hypersil ODS (5 μm), C18, 150 × 4.6 mm maintained at 25°C; mobile phase, KH2PO4 (100 mM), EDTA (0.1 mM), sodium octylsulfonate (0.5 mM), methanol (5%) adjusted to pH 3.15 with PO4H3. The flow rate was 1 ml/min. Electrochemical detection was performed using a Waters M460 detector with a working potential of 850 mV against an Ag/AgCl reference. Levels of DOPA and 5-HTP were expressed relative to control (vehicle) values (=0%). Data were analyzed by ANOVA followed by Dunnett’s test.
Influence on cerebral DA turnover.
As described in detail previously (Millan et al., 1995), the ratio of DOPAC to DA levels was determined in various cerebral tissues 30 min after s.c. administration of drugs. Levels were determined by HPLC/electrochemical detection as described above. DOPAC/DA ratios were expressed relative to control (vehicle) values (=0%). Data were analyzed by ANOVA followed by Dunnett’s test.
Influence upon the electrical activity of dopaminergic neurones.
As previously described (Lejeune et al., 1997), rats were anesthetized with chloral hydrate (400 mg/kg, i.p.), the femoral vein was catheterized and rats were placed in a stereotaxic apparatus. A tungsten micro-electrode was lowered into the VTA. Dopaminergic neurones were identified as previously and baseline recording performed over 5 min. Drugs were dissolved in sterile water and injected i.v. in a volume of 0.5 ml/kg, followed by a 0.1 ml saline flush. Compounds were administered cumulatively i.v. at intervals of 2 to 5 min. In antagonist studies, they were administered (single dose) 2 min after the injection of apomorphine (0.031 mg/kg, i.v.). Data acquisition and analysis were performed using Spike 2 software (C.E.D., Cambridge, England) and results are expressed as firing rate (60-sec bins at time of peak drug action) as a percentage of baseline pre-injection values (=0%).
Influence on the electrical activity of serotoninergic neurones.
For evaluation of the influence of drugs upon the activity of serotoninergic neurones of the DRN, an identical protocol was used as described above for dopaminergic neurones (Lejeune et al., 1994, 1997). Drugs were administered in cumulative doses i.v. at intervals of 2 to 5 min. In antagonist studies, drugs were administered at a single dose followed, 2 min later, by a single injection of WAY 100,635 (0.031 mg/kg, i.v.) or (−) tertatotol (2.0 mg/kg, i.v.).
Determination of extracellular levels of DA, 5-HT and NAD in the FCX, accumbens and striatum.
The procedure used has been described in detail elsewhere (Gobert et al., 1995b, 1998). Under pentobarbital anesthesia (60 mg/kg, i.p.), rats were placed in a stereotaxic apparatus and a guide cannula implanted in the FCX or in both the accumbens and the contralateral striatum. Five days later, a Cuprophan CMA/11 probe of 4 mm length (FCX and striatum) or of 2 mm length (accumbens) and 0.24 mm o.d. was lowered into position and perfused at 1 μl/min with a phosphate-buffered Ringer solution (147.2 mM NaCl, 4 mM KCl and 2.3 mM CaCl2, pH 7.3). Two hours later, dialysis was commenced and samples taken every 20 min. Three basal samples were taken, then the drug was injected. Samples were then taken for another 3 hr. For interaction studies, WAY 100,635 (0.16 mg/kg, s.c.) was injected followed, 20 min later, by either S 16924 (2.5) or clozapine (2.5). Levels of DA, 5-HT and NAD were simultaneously quantified in individual samples using HPLC and coulometric detection with the following conditions: 20-μl dialysate samples were diluted with 20 μl of mobile phase (NaH2PO4: 75 mM, EDTA: 20 μM, sodium decanesulphonate: 1 mM, methanol: 17.5%, triethylamine 0.01%, pH 5.70) and 33-μl samples were analyzed by HPLC with a column (hypersil ODS 5 μm, C18, 150 × 4.6 mm) maintained at 43°C for separation and a coulometric detector (ESA 5014, Coulochem II) for quantification. The first electrode of the detector was set at −70 mV (reduction) and the second at +280 mV (oxidation). The mobile phase was delivered at a flow rate of 2 ml/min. The assay sensitivity was between 0.1 and 0.2 pg/sample for DA, NAD and 5-HT. Drug effects were expressed as a percentage of basal values (=0%). Data were analyzed by ANOVA with sampling time as the repeated within-subject factor.
Drugs.
All drugs were dissolved in sterile water, if necessary with a few drops of lactic acid. The pH was adjusted to as close to neutrality as possible (>5.0). Drugs were injected s.c. unless otherwise specified. Drug sources and salts were as follows. Clozapine and apomorphine HCl (Research Biochemicals International, Natick, MA). S 16924 HCl, WAY 100,635 HCl and haloperidol were synthetized by O. Muller and G. Lavielle (Servier).
Results
Patterns of displacement.
At each of the receptors presented in tables 1 to 3, S 16924, clozapine and haloperidol presented monophasic isotherms for displacement of the respective radioligands (slope factors not significantly differing from unity) (not shown). In figure 2, the overall receptor profiles of S 16924, haloperidol and clozapine at several key receptor types are depicted. It may be seen that the profiles of S 16924 and clozapine corresponded closely, whereas that of haloperidol was markedly different.
Affinities of S 16924, clozapine and haloperidol at multiple dopaminergic receptors (table 1).
Whereas haloperidol displayed potent affinity at native rat and cloned hD2 receptors, S 16924 mimicked the modest affinity of clozapine at these sites. Similarly, the affinity of S 16924 and clozapine at cloned rat and hD3 receptors was modest in contrast to the marked affinity of haloperidol for these sites. Haloperidol manifested about 5-fold lower affinity for hD4 (the hD4.4 isoform) as compared to hD2 receptors whereas clozapine displayed a mild preference (about 2-fold) for hD4 sites. This preferential affinity for hD4vs.hD2 sites was more pronounced for S 16924 (about 5-fold selectivity). Compared with D2 receptors, the affinity of haloperidol was markedly lower at both native D1 and cloned, hD1 receptors. In distinction, both S 16924 and clozapine presented comparable and modest affinity for native D1 and cloned hD1 receptors as compared with D2 receptors. Haloperidol, S 16924 and clozapine all showed similar affinities for cloned hD5 receptors as compared to hD1 receptors. The affinity of S 16924 at DA reuptake sites was negligible.
Affinities of S 16924, clozapine and haloperidol at multiple serotoninergic receptors (table 2).
Haloperidol showed negligible affinity at native rat 5-HT1A and cloned h5-HT1A receptors although the modest affinity of clozapine at these sites was similar to its affinity at D2 receptors (table 1). In contrast, S 16924 showed pronounced affinity at both rat 5-HT1A and h5-HT1A receptors that was ≈20-fold superior to its affinity at D2 sites. Haloperidol showed negligible affinity for 5-HT1B sites, at which clozapine displayed very weak and S 16924 only low affinity. The affinity of haloperidol at native 5-HT2A and cloned h5-HT2A receptors was weak vs. its affinity at D2 sites, and haloperidol displayed negligible affinity for native 5-HT2C and cloned h5-HT2C sites as well as for h5-HT2B sites. In distinction, clozapine and S 16924 both showed markedly higher affinity at native, 5-HT2A and cloned h5-HT2Avs. D2 receptors. Similarly, in contrast to haloperidol, both S 16924 and clozapine manifested marked affinity for native 5-HT2C and cloned h5-HT2C receptors. Interestingly, for all ligands, affinities were higher at cloned, h5HT2Cvs.native, porcine 5-HT2C sites, and this difference was significant (P < .05) for S 16924 and clozapine. Whether this observation reflects a species difference, or a difference between native, tissue vs. cloned, transfected receptors, remains to be elucidated. S 16924 and haloperidol also showed pronounced affinity for h5-HT2B sites. At 5-HT3 receptors, clozapine displayed modest affinity whereas neither haloperidol nor S 16924 displayed significant affinity. The affinity of all drugs for 5-HT4 and 5-HT5A sites was low. However, both S 16924 and clozapine, in contrast to haloperidol, showed significant affinity at 5-HT6 and 5-HT7 sites. S 16924 did not manifest significant affinity for 5-HT reuptake sites.
Influence of S 16924, clozapine and haloperidol at multiple adrenergic receptors (table 3).
S 16924, clozapine and haloperidol all shared potent affinity for native α1-AR as well as α1A- and α1B-AR receptors. However, when expressed relative to their affinity at D2 receptors, S 16924 and clozapine, but not haloperidol, revealed a marked preference for α1-, α1A- and α1B-AR sites in each case. S 16924 and clozapine showed modest affinity at both native α2A- and cloned hα2A-AR receptors. Further, they also showed modest affinity for cloned hα2B- and hα2C-AR receptors. The affinity of haloperidol for each of these α2-AR receptor types was negligible. S 16924 did not show significant affinity for β1- and β2-AR receptors or for NAD reuptake sites.
Influence of S 16924 as compared to clozapine and haloperidol upon [35S]-GTPγS binding at hD2, hD3and hD4 receptors.
Dopamine elicited a concentration-dependent increase in [35S]-GTPγS binding to cloned hD2, hD3 and hD4receptors with Effective Concentration (EC)50 values of 353 ± 52, 15.6 ± 3.9 and 109 ± 15 nM, respectively. In contrast, neither S 16924, clozapine nor haloperidol stimulated binding at these receptors (fig. 3 and not shown). Indeed, they all behaved as antagonists at hD2, hD3 and hD4 receptors, concentration-dependently inhibiting the stimulation of [35S]-GTPγS binding induced by DA (3, 1 and 1 μM respectively) (fig. 3 and not shown). Kb values calculated for S 16924 were: hD2, 34.2 ± 3.7 nM; hD3, 79.8 ± 7.2 nM and hD4, 5.0 ± 1.8 nM. Kb values calculated for clozapine were: hD2, 71.7 ± 11.1 nM; hD3, 251 ± 80 and hD4, 48.4 ± 3.7 nM. Kb values calculated for haloperidol were: hD2, 0.58 ± 0.10; hD3, 28.8 ± 11.3 and hD4, 1.37 ± 0.18 nM.
Influence of S 16924 as compared to clozapine and haloperidol upon [35S]-GTPγS binding at h5-HT1Areceptors.
Serotonin concentration-dependently increased [35S]-GTPγS binding at h5-HT1A receptors with an EC50 of 16.8 ± 3.9 (fig.4). Even at a very high concentration (10 μM), haloperidol failed to stimulate [35S]-GTPγS binding (not shown). However, clozapine (EC50 of 1740 ± 736 nM) stimulated binding to 43.8 ± 3.1% of levels attained with 5-HT (defined as 100%) (not shown). S 16924 stimulated [35S]-GTPγS binding by 54.1 ± 11.3%, but with considerably greater potency than clozapine: the EC50 for S 16924 was 11.3 ± 0.4 nM (fig. 4). S 16924-stimulated [35S]-GTPγS binding was inhibited by the selective 5-HT1A antagonist, WAY 100,635, with an IC50 of 3.18 ± 0.53 (fig. 4). WAY 100,635 was inactive alone (not shown).
Influence of S 16924 as compared to clozapine and haloperidol upon cerebral turnover of DA and 5-HT.
As determined by the ratio of tissue levels of DA to those of its metabolite, DOPAC, haloperidol potently and markedly enhanced DA turnover in projection areas of mesocortical (FCX), mesolimbic (olfactory tubercles and nucleus accumbens) and nigrostriatal (striatum) pathways (fig.5). In contrast, S 16924 and clozapine only weakly and less markedly increased DA synthesis in each of these regions (fig. 5). Similarly, on determination of levels of the DA precursor, DOPA, after pretreatment with the decarboxylase inhibitor, NSD 1015 (100 mg/kg, s.c.), haloperidol elicited a potent and pronounced induction in striatal DA synthesis whereas S 16924 and clozapine were only weakly active (fig. 5). Haloperidol failed to modify striatal levels of the 5-HT precursor, 5-HTP, an index of 5-HT synthesis, whereas striatal levels of 5-HTP were potently and markedly decreased by S 16924 and slightly depressed by clozapine (fig. 5).
Influence of S 16924 as compared to clozapine and haloperidol upon the electrical activity of VTA-localized dopaminergic neurones.
The dopaminergic agonist, apomorphine (0.031 mg/kg, i.v.), markedly reduced the firing rate of dopaminergic neurones in the VTA (fig.6). This action was dose-dependently inhibited by haloperidol and, less potently, by S 16924 and clozapine (fig. 6). ID50s (95% CLs) were as follows: 0.004 (0.002–0.006), 0.18 (0.12–0.19) and 0.22 (0.15–0.34), respectively. Administered alone, haloperidol and, less potently, clozapine and S 16924 slightly but dose-dependently and significantly increased firing rate (fig. 6).
Influence of S 16924 as compared to clozapine and haloperidol upon the electrical activity of DRN-localized serotoninergic neurones.
S 16924 potently and dose-dependently inhibited the firing of DRN-localized serotoninergic neurones with an ID50 (95% CLs) of 0.02 (0.01–0.03) (fig. 7). Clozapine also inhibited DRN firing over a higher dose-range: ID50 (95% CLs) = 0.08 (0.02–0.3). Haloperidol was also effective, although only at high doses: ID50 (95% CLs) = 0.4 (0.2–0.9) (fig. 7). The inhibitory influence of S 16924 was blocked by WAY 100,635 and a further 5-HT1A antagonist, (−)-tertatolol, neither of which significantly modified firing rate upon administration alone (fig. 7). The ac- tions of clozapine and haloperidol were not affected by WAY 100,635 or (−)-tertatotol (fig.7).
Influence of S 16924 as compared to clozapine and haloperidol upon extracellular levels of DA, 5-HT and NAD in the FCX, nucleus accumbens and striatum of freely-moving rats.
Haloperidol (0.63) elicited a modest increase in dialysate levels of DA in the FCX and a similar, though more sustained, elevation in extracellular levels of DA levels in the nucleus accumbens and striatum (fig.8). In contrast S 16924 (2.5) and clozapine (2.5) increased dialysate levels of DA in the FCX without markedly modifying those of DA in either the accumbens or striatum (fig. 8). This facilitatory influence of S 16924 on FCX levels of DA was expressed dose-dependently and, in parallel, it dose-dependently increased and decreased FCX levels of NAD and 5-HT, respectively (fig.9). S 16924 also markedly decreased 5-HT levels in both accumbens and striatum (fig.10). Clozapine also increased dialysate levels of NAD in the FCX (fig. 11) without significantly affecting levels of 5-HT either in this structure or in the accumbens, although it significantly decreased 5-HT levels in striatum (fig. 10). Haloperidol similarly provoked a modest increase in levels of NAD in FCX (fig. 11) without modifying levels of 5-HT in the FCX, accumbens or striatum (fig. 10). The influence of S 16924 upon FCX levels of DA and 5-HT was significantly inhibited by WAY 100,635 (fig.12). In contrast, WAY 100,635 did not significantly modify the influence of clozapine upon FCX levels of DA, 5-HT or NAD (fig. 13).
Discussion
S 16924 displays a profile of interaction at multiple dopaminergic, serotoninergic and adrenergic receptors similar to that of the “atypical” antipsychotic, clozapine and different to that of the neuroleptic, haloperidol. In particular, it shares the more marked activity of clozapine at hD4, 5-HT2A, 5-HT2C and α1-AR as compared to D2 receptors. In addition, in contrast to both clozapine and haloperidol, S 16924 possesses potent, partial agonist properties at 5-HT1A receptors. This distinctive binding profile of S 16924 is reflected in vivo by its ability, upon acute administration, to inhibit serotoninergic transmission and to preferentially reinforce frontocortical vs. subcortical dopaminergic transmission.
Antagonist properties at cloned hD2, hD3and hD4 receptors.
S 16924 displayed, as with clozapine, only modest affinity at hD2 and hD3receptors. In previous studies, clozapine has shown a mild, although variable and radioligand-dependent, preference for hD4vs. hD2 receptors (Newman-Tancredi et al., 1997; Seeman et al., 1997), a finding confirmed herein, and S 16924 also displayed higher affinity at hD4than hD2 receptors. Human D2, hD3and hD4 receptors all couple via G proteins to (specific isoforms) of adenylyl cyclase and other intracellular transduction systems (Levant, 1997; Newman-Tancredi et al., 1997). Thus, activation and blockade of hD2, hD3 and hD4 sites can be evaluated by the binding of [35S]-GTPγS, which interacts with “activated” G proteins after their ligand-induced dissociation from the corresponding receptor. Using this approach, we recently demonstrated that clozapine and haloperidol behave as antagonists at hD4 receptors (Newman-Tancredi et al., 1997) and S 16924 also behaved as a potent antagonist at hD4 sites. Our findings also show that S 16924, haloperidol and clozapine behave as antagonists at hD2 and hD3 receptors. The antagonist properties of S 16924 at hD2 receptors are of significance inasmuch as blockade of postsynaptic D2 receptors in limbic structures, by counteracting the hyperactivity of mesolimbic dopaminergic pathways, may reduce the positive symptoms of schizophrenia (Kahn and Davis, 1995). Whether antagonism of postsynaptic D3 receptors, which are enriched in limbic tissue, is of importance to the actions of antipsychotic drugs is still under debate (Levant, 1997). The more pronounced activity of S 16924 at hD4vs. hD2 receptors is of interest inasmuch as such a preference has been proposed to account for the superior antipsychotic profile of clozapine vs. haloperidol (Seeman et al., 1997). This contention has, however, been challenged (Roth et al., 1995) and putative alterations in levels of mRNA encoding D4 receptors in schizophrenics are controversial (Marzella et al., 1997; Mulcrone and Kerwin, 1996; Seeman et al., 1997). Further, the selective D4 receptor antagonist, L 745,870, was not effective in controlling psychosis in a clinical study (Kramer et al., 1997). In addition, the preclinical profiles of L 745,870 and other selective D4 antagonists provide little evidence that antagonism of D4 receptors controls the positive symptoms of schizophrenia (Bristow et al., 1997). Nevertheless, D4 receptor blockade may improve the cognitive-attentional symptoms of schizophrenia (Tallman, 1997; Millan MJ and Dekeyne A, unpublished observations).
Interaction with D1 and D5 receptors.
The similar affinity of S 16924 and clozapine at D1vs. D2 receptors contrasts to the preference of haloperidol for the latter. This observation is of interest inasmuch as 1) a dysequilibrium in the activity of striatal populations of D1 and D2 receptors may contribute to extrapyramidal, side-effects and 2) actions at D1 receptors may contribute to antipsychotic properties (Gerlach and Hansen, 1992) (companion paper). D5 receptors present marked similarities to D1 receptors, and clozapine and haloperidol possess similar affinity at hD5vs. hD1sites (Sunahara et al., 1991). Similarly, the modest affinity of S 16924 at hD5 receptors was comparable to its affinity at hD1 receptors. Although D5receptors are differentially localized to D1 receptors, their functional significance remains unknown and certain actions ascribed to D1 sites may actually be mediated by D5 receptors (Bergson et al., 1995;Meador-Woodruff et al., 1996).
Antagonist actions at D2 and D3 receptorsin vivo: modulation of cerebral DA synthesis.
The activity of dopaminergic pathways is tonically inhibited by D2 (and D3) receptors localized on their dendrites and terminals and, possibly, by postsynaptic populations of D3 sites acting via a feedback loop (Gobert et al., 1995b; Koeltzow et al., 1998; Tepper et al., 1997). Correspondingly, S 16924, clozapine and, more potently, haloperidol elevated DA synthesis in regions innervated by mesocortical (FCX), mesolimbic (olfactory tubercles and accumbens) and nigrostriatal projections (striatum) (Gobert et al., 1995aand b; Kahn and Davis, 1995). Interestingly, the magnitude of the increases in DA turnover evoked by S 16924 and clozapine were less marked than for haloperidol. One possible explanation is that elevations in DA turnover reflect inverse agonist actions at D2 autoreceptors rather than blockade of tonic DA activity (Nilsson et al., 1996). However, as with haloperidol, clozapine possesses negative efficacy at hD2 sites (Hall and Strange, 1997). An alternative explanation is that the serotoninergic and/or adrenergic actions of S 16924 and clozapine (see below) may intervene to moderate their influence upon DA synthesis. Irrespective of the underlying mechanisms, the finding that DA synthesis was little perturbed by S 16924 in the striatum is of importance inasmuch as extrapyramidal motor effects are correlated with an elevation of striatal DA synthesis (Lucas et al., 1997) (companion paper).
Partial agonist actions at 5-HT1A receptors in vivo.
Whereas S 16924 preferentially interacted at 5-HT1Avs. D2 receptors, clozapine interacted with equivalent potency, and haloperidol interacted exclusively with D2vs. 5-HT1Areceptors. In a [35S]GTPγS binding model, S 16924 behaved as a partial agonist with an efficacy equivalent to that of clozapine (Newman-Tancredi et al., 1996). This cellular model of 5-HT1A receptor stimulation possesses a sensitivity comparable to that of postsynaptic 5-HT1Areceptors (Newman-Tancredi et al., 1997; Lejeune et al., 1997), at which S 16924 behaves as a partial agonist in vivo (companion paper). Inhibitory 5-HT1Aautoreceptors on serotoninergic cell bodies are more sensitive than their postsynaptic counterparts (Meller et al., 1990;Newman-Tancredi et al., 1997). Correspondingly, S 16924 markedly reduced striatal and accumbens release of 5-HT and it reduced striatal turnover of 5-HT at doses substantially lower than those enhancing striatal DA turnover. These data, underpinned by the WAY 100,635-reversible inhibitory influence of S 16924 upon DRN firing rate and FCX dialysate levels of 5-HT, suggest that S 16924 acutely inhibits serotoninergic transmission via agonist actions at 5-HT1Aautoreceptors. This activity is of particular significance in several respects. First, stimulation of 5-HT1A receptors facilitates the activity of mesocortical dopaminergic (and adrenergic) pathways (Lejeune et al., 1997; Millan et al., 1997) (see below). Second, an inhibition in 5-HT release is associated with a reduction in anxious states (Coplan et al., 1995;Meller et al., 1990) and S 16924 possesses anxiolytic properties (Dekeyne A, and Millan MJ, unpublished observations). Third, activation of 5-HT1A autoreceptors may counter the induction of extrapyramidal motor symptoms due to striatal D2 receptor blockade (Lucas et al., 1997) and S 16924 does not elicit catalepsy in rats (companion paper). In contrast to S 16924, clozapine only modestly inhibited striatal release and turnover of 5-HT and failed to modify dialysate levels of 5-HT in the accumbens or FCX. Further, the inhibitory influence of clozapine on DRN firing is mediated by its antagonist properties at α1-AR receptors, a mechanism that may also intervene in the weak reduction of DRN firing by haloperidol (Lejeune et al., 1994).
Serotonin 5-HT2A and 5-HT2C receptors.
A preferential blockade of 5-HT2Avs.D2 receptors has been associated with a reduced propensity to elicit extrapyramidal side-effects and, possibly, an improved efficacy in the control of resistant patients and negative-cognitive symptoms (Roth and Meltzer, 1995; Schmidt and Fadayel, 1995). Thus, it is of significance that, in analogy to clozapine (Canton et al., 1994; Roth and Meltzer, 1995), S 16924 showed more pronounced affinity at 5-HT2A than D2 receptors. Indeed, antagonism of 5-HT2A receptors is an important, clozapine-like feature of the pharmacology of S 16924 (companion paper). The higher affinity of clozapine at 5-HT2Cvs. D2 sites (Canton et al., 1994;Roth and Meltzer, 1995) was similarly mimicked by S 16924. Although the significance of 5-HT2C receptor blockade has been questioned as regards a reduced propensity to elicit extrapyramidal symptoms (Roth and Meltzer, 1995), there are several further, potentially important consequences of 5-HT2C receptor blockade. First, antagonism of 5-HT2C receptors markedly facilitates mesocortical dopaminergic transmission (Gobert et al., 1998; Kelland and Chiodo, 1996; Pessia et al., 1994). Second, 5-HT2C receptor antagonists display anxiolytic properties (Kennett et al., 1997). Third, based on studies of transgenic mice lacking 5-HT2C receptors, it has been suggested that the weight gain provoked by antipsychotics reflects 5-HT2C receptor blockade (Cunningham-Owens, 1996;Tecott et al., 1995). Nevertheless, certain antipsychotics, such as risperidone, elicit weight gain despite low affinity at 5-HT2C receptors (Cunningham-Owens, 1996). Thus, other mechanisms, such as histamine1 receptor blockade, may also be involved (Cunningham-Owens, 1996). An interesting question concerns the functional significance of the combined blockade of 5-HT2A/2C receptors and activation of 5-HT1Areceptors, as shown by S 16924. Serotoninergic transmission is inhibited by 5-HT1A autoreceptors, and postsynaptic 5-HT1Avs. 5-HT2A/2C receptors exert an opposite influence on cellular transduction mechanisms, resulting in neuronal hyperpolarization and excitation, respectively. Thus, these properties may, as suggested previously, act synergistically (Millanet al., 1992): for example, in enhancing mesocortical dopaminergic transmission (see below). It would be of interest to perform long-term studies of the antipsychotic and other actions of the parallel activation and blockade of 5-HT1A and 5-HT2A/2C receptors, respectively.
5-HT6 and 5-HT7 receptors.
S 16924 displayed significant affinity at 5-HT6 receptors and it has been proposed that an action of clozapine at these sites may contribute to its atypical profile (Monsma et al., 1993). Although Roth et al. (1994) suggested that relatively high affinity at 5-HT6vs. D2 receptors may not be a distinguishing feature of “atypical” antipsychotics, the preferential corticolimbic localization of 5-HT6receptors is of pertinence regarding the negative and cognitive-attentional symptoms of schizophrenia (Sleight et al., 1997). S 16924 also mimicked the high affinity of clozapine at 5-HT7 receptors. These are enriched in several limbic and cortical regions and have been implicated in depressive states, which can aggravate the negative symptoms of schizophrenia (Sleightet al., 1997; although see Gobbi et al., 1996). Moreover, sleep cycles are disrupted in schizophrenics and 5-HT7 receptors are concentrated in the suprachiasmatic nucleus wherein they fulfill an important role in controlling circadian rhythms (Lovenberg et al., 1993).
Interaction at α1-AR receptors.
A perturbation of adrenergic transmission is related to positive crises in schizophrenic patients and to an intensification of negative symptoms and the risk of relapse after treatment withdrawal (Mass et al., 1993). S 16924 mimicked the pronounced affinity of clozapine at α1-ARs (as well as α1A- and α1B-ARs) (table 3), which are enriched in the thalamus, hippocampus, FCX and other structures implicated in the control of mood and in the pathophysiology of psychiatric disorders (Baldessariniet al., 1992). Several lines of evidence suggest that α1-AR blockade may afford advantages in the treatment of schizophrenia. First, blockade of (limbic or cortical) α1-AR receptors inhibits the induction of locomotion by psychostimulants (Blanc et al., 1994; Prinssen et al., 1994; Svensson et al., 1995). Second, coadministration of α1-AR antagonists with haloperidol results in a clozapine-like, preferential inhibition of the activity of mesolimbic vs. nigrostriatal dopaminergic pathways (Laneet al., 1988). Third, blockade of thalamic α1-AR receptors may improve the gating of sensory information to the cortex, a process that is defective in psychotic patients (Goldberg and Gold, 1995). Although peripheral α1-AR blockade is associated with orthostatic hypotension, drug titration circumvents this effect, to which tolerance may develop (Cunningham-Owens, 1996).
Modulation of the electrical activity of VTA dopaminergic neurones.
S 16924, clozapine and, more potently, haloperidol blocked suppression of the firing of VTA-localized dopaminergic neurones by apomorphine, consistent with their antagonist properties at D2 (and D3) autoreceptors. In fact, they slightly enhanced firing rate when administered alone. This action of haloperidol may be attributed to the interruption of a tonic, inhibitory tone at D2 receptors (Gobert et al., 1998), a mechanism that may similarly contribute to the actions of higher doses of S 16924 and clozapine. Further, the antagonist actions of S 16924 and clozapine at 5-HT2C receptors, or their partial agonist actions at 5-HT1A receptors, might also be involved in exciting dopaminergic cell bodies in the VTA (Gobertet al., 1998; Kelland and Chiodo, 1996; Pessia et al., 1994).
The influence of S 16924 on frontocortical dopaminergic and adrenergic transmission.
A dysruption in cortical function (“hypofrontality”) is involved in the negative- and cognitive-attentional symptoms of schizophrenia (Andreasen et al., 1992). In this regard, a perturbation of dopaminergic input to this region has been implicated (Jentsch et al., 1997;Knable and Weinberger, 1997). Correspondingly, a potentiation of mesocortical dopaminergic transmission may improve the negative-cognitive symptoms of schizophrenia. In line with previous work (Moghaddam and Bunney, 1990; see Meltzer, 1995), haloperidol elicited only a modest increase in dialysate levels of DA in the FCX, and a marked (and more prolonged) rise was seen in nucleus accumbens and striatum: these actions may reasonably be attributed to its antagonist properties at D2 (and D3) autoreceptors (Gobert et al., 1995b, 1998). This increase in limbic release of DA by haloperidol may contribute to its lack of antipsychotic efficacy in certain “refractory” patients. In contrast to haloperidol, both clozapine and S 16924 markedly increased dialysate levels of DA in the FCX at doses not markedly modifying DA levels in the accumbens or striatum. This suggests that they may correct FCX hypofrontality without perturbing DA levels in limbic or striatal regions. In view of this regional specificity, an action at D2 (D3) autoreceptors is not likely to be the principal mechanism involved in the influence of S 16924 on FCX levels of DA. Notably, dopaminergic neurones in the parabrachial subdivision of the VTA, which project primarily to the FCX, are subject to a more pronounced serotoninergic control than their paranigral counterparts projecting to limbic structures (Lejeune et al., 1997;Svensson et al., 1995) and selective 5-HT1Areceptor agonists increase dialysate levels of DA in the FCX (Kelland and Chiodo, 1996; Lejeune et al., 1997). In line with these observations, WAY 100,635 attenuated the increase in FCX levels of DA elicited by S 16924. WAY 100,635 did not, however, abolish the elevation in DA levels provoked by S 16924 and, as mentioned above, an action of S 16924 at 5-HT2A or 5-HT2C receptors controlling FCX release of DA may also underlie its enhancement of FCX levels of DA. Such mechanisms may also intervene in the elevation in FCX dialysate levels of DA elicited herein by clozapine inasmuch as its actions were insensitive to WAY 100,635 (but see Rollema et al., 1997). The activity of mesocortical adrenergic neurones is subject to an inhibitory α2A-AR autoreceptor mediated-tone as well as a complex pattern of modulatory serotoninergic influence involving (indirect) facilitatory and inhibitory effects mediated via 5-HT1A and 5-HT2A/5-HT2C receptors, respectively (Gobertet al., 1998; Haddjeri et al., 1997; Millanet al., 1997). Inasmuch as the increase in FCX levels of NAD elicited by S 16924 and clozapine was not markedly attenuated by WAY 100,635, activation of 5-HT1A receptors may play a less important role in these actions than blockade of 5-HT2C (or α2-AR) receptors. In any case, adrenergic pathways in the FCX fulfill an important role in mechanisms controlling vigilance and memory formation (Foote and Aston-Jones, 1995) suggesting that the potentiation in mesocortical adrenergic transmission by S 16924 may be of use in improving cognitive-attentional performance.
Conclusions.
For antipsychotic agents, it is their global pattern of interaction at multiple monoaminergic receptor types, and the relationship between the affinity at specific receptor types to that at D2 receptors, which determines their functional activity in vivo (see fig. 2). In this respect, S 16924 displays a profile of action that differs markedly to that of haloperidol and closely resembles that of clozapine, despite their chemical distinctiveness. In addition, the partial agonist properties of S 16924 at 5-HT1A receptors are more pronounced than those of clozapine. This distinctive component of activity underlies the acute inhibition of serotoninergic transmission by S 16924, fulfills an important role in its selective facilitatory influence on mesocortical dopaminergic transmission, and, as described in the companion paper, contributes to its distinctive pattern of functional actions in experimental models of potential antipsychotic and extrapyramidal activity.
Acknowledgments
The authors thank L. Cistarelli, C. Chaput, C. Melon, V. Pasteau, M. Touzard and L. Defaye-Verrièle for technical assistance, as well as C. Langaney-Le Roy and M. Soubeyran for secretarial assistance.
Footnotes
-
Send reprint requests to: Dr. Mark J. Millan, Institut de Recherches Servier, Centre de Recherches de Croissy, Psychopharmacology Department, 125 Chemin de Ronde, 78290—Croissy-sur-Seine, Paris, France.
- Abbreviations:
- AR
- adrenergic
- ANOVA
- analysis of variance
- CHO
- Chinese hamster ovary
- DA
- dopamine
- DOPA
- dihydroxyphenylalanine
- DOPAC
- dihydroxyphenylacetic acid
- DRN
- dorsal raphe nucleus
- EC50
- effective concentration50
- FCX
- frontal cortex
- GTPγS
- guanylyl 5′-[γ-thio]-triphosphate
- 5-HT
- serotonin
- 5-HTP
- 5-hydroxytryptophane
- HPLC
- high-performance liquid chromatography
- NAD
- noradrenaline
- VTA
- ventral tegmental area
- Received February 4, 1998.
- Accepted April 21, 1998.
- The American Society for Pharmacology and Experimental Therapeutics