Differential Effects of the 5-Hydroxytryptamine (5-HT)_1A Receptor Inverse Agonists Rec 27/0224 and Rec 27/0074 on Electrophysiological Responses to 5-HT_1A Receptor Activation in Rat Dorsal Raphe Nucleus and Hippocampus in Vitro

Impairment of 5-hydroxytryptamine (5-HT) neurotransmission is involved in major neuropsychiatric pathological states (e.g., depression, anxiety, schizophrenia, and Parkinson's disease). Considerable effort has been consistently made to understand the functioning of the 5-HT system and to develop new selective drugs for 5-HT receptor subtypes.

The 5-HT_1A receptor has been extensively studied because of the early discovery of a selective agonist (Gozlan et al., 1983) and receptor cloning (Kobilka et al., 1987; Albert et al., 1990). 5-HT_1A receptors are G protein-coupled receptors (GPCR) expressed throughout the central nervous system (Pompeiano et al., 1992). Stimulation of 5-HT_1A receptors activates Gi/o proteins, leading to at least two different cellular responses: inhibition of adenylate cyclase and opening of inwardly rectifying K⁺ channels (Andrade and Nicoll, 1987; Fargin et al., 1989, Penington et al., 1993). 5-HT_1A receptors mediate the hyperpolarization of neurons in almost all the brain regions (see Barnes and Sharp, 1999), thereby exerting an inhibitory action on cell discharge, including the rhythmic firing of serotonergic cells in the raphe nuclei.

The 5-HT_1A receptors expressed in raphe serotonergic cells (5-HT_1A autoreceptors) display differential characteristics when compared with 5-HT_1A receptors from other, particularly cortical, regions (5-HT_1A heteroreceptors). For example, 5-HT_1A receptor agonist-evoked hyperpolarization of dorsal raphe nucleus (DRN) serotonergic neurons, but not of CA1 pyramidal cells, strongly desensitizes following treatment with antidepressant drugs, namely, selective serotonin reuptake inhibitors (SSRI) (Le Poul et al., 2000). It has been suggested that coupling to specific G proteins may be responsible for region-specific differences in 5-HT_1A receptor desensitization (Li et al., 1997; Le Poul et al., 2000; Mannoury La Cour et al., 2001; see Hensler, 2003). Furthermore, raphe serotonergic cells have a 5-HT_1A receptor reserve (Cox et al., 1993) and respond to weak partial agonists with a higher efficacy than cortical neurons. Thus, 8-OH-DPAT acts as a full agonist in the DRN (Williams et al., 1988) but behaves as a partial agonist in the CA1 hippocampal region (Andrade and Nicoll, 1987; Beck et al., 1992). The classic criteria of agonism and antagonism based on the theory of occupancy are clearly inadequate to explain these cell-type specific effects of drugs acting at 5-HT_1A receptors, and more complex models of ligand-GPCR interaction should be applied.

According to the extended ternary complex model (Samama et al., 1993), GPCR exist in at least two states, inactive (R) and active (R*), and ligands that preferentially bind to one of these states modify their equilibrium. Inverse agonists preferentially bind to R and when coapplied with agonists compete for the receptor's binding site, exerting variable degrees of antagonism on the agonists' effect (Christopoulos and Kenakin, 2002; Kenakin, 2004). In addition, some inverse agonists may promote the change in GPCR conformation from R* to R through an allosteric transition (see Neubig et al., 2003). Recently, it was found that as much as 80% of the receptor ligands acting as antagonists show inverse agonist activity when tested in recombinant receptor systems (Kenakin, 2004). Consistently, the guanosine 5′-O-(3-[³⁵S]thiotriphosphate ([³⁵S]GTPγS) binding assays in cells transfected with 5-HT_1A receptors revealed that many of the “classic,” nonselective 5-HT_1A receptor antagonists (e.g., methiothepin, spiperone) are in fact inverse agonists (e.g., Newman-Tancredi et al., 1997). However, several newer compounds, such as WAY-100635 (Newman-Tancredi et al., 1997; Testa et al., 1999) and Rec 15/3079 (Leonardi et al., 2001), have been found to be neutral antagonists. Although the [³⁵S]GTPγS binding assay may disclose the inverse agonist nature of ligands, the functional consequences of inverse agonism in native tissue not overexpressing receptors remain largely speculative.

In the present work, we characterized the pharmacological properties of two novel compounds synthesized toward the 5-HT_1A receptors, Rec 27/0074 and Rec 27/0224. The [³⁵S]GTPγS binding was used as a parameter for studying the functional consequences of drug binding at the 5-HT_1A receptor level, whereas electrophysiological responses in brain slices were used to characterize cell responses to drug actions on serotonergic cells in the DRN and on pyramidal neurons in the CA1 region of the hippocampus.

Materials and Methods

All the animal manipulations were performed according to the European Community guidelines for animal care (DL 116/92, application of the European Communities Council Directive 86/609/EEC) and approved by the Committee for Animal Care and Experimental Use of the University of Florence.

Radioligand Binding Assays. Radioligand binding studies for a number of receptors were performed using experimental procedures previously described in detail. Binding to 5-HT_1A serotonin receptors was performed on membrane homogenates of HeLa cells stably transfected with human 5-HT_1A serotonin receptors (Testa et al., 1999) using [³H]8-OH-DPAT as the radioligand. Binding to human cloned α₁-adrenoceptors was performed on membrane homogenates of Chinese hamster ovary cells transfected with DNA expressing the gene encoding each α₁-adrenoceptor subtype (Testa et al., 1995) and labeled with [³H]prazosin. Binding studies on native α₂-adrenoceptors and D₂ dopamine receptors (Leonardi et al., 1994) were carried out in membranes of rat cerebral cortex taken from male Sprague-Dawley rats (200–300 g; Charles River Italica, Calco, Italy) using [³H]rauwolscine or [³H]spiperone, respectively.

Binding of Rec 27/0224 and Rec 27/0074 to rat 5-HT_1B, bovine 5-HT_1D, and human recombinant 5-HT_2A, 5-HT_2B, 5-HT_2C, 5-HT₃, 5-HT_4c, 5-HT_4d, 5-HT_4e, 5-HT_5A, 5-HT₆, and 5-HT₇ serotonin receptor subtypes was carried out by Cerep at their Poitiers Laboratories (Celle l'Evescault, France) using standardized operating procedures (ISO 9001:2000) fully described and available online at the web site (http://www.cerep.fr/Cerep/Users/pages/catalog/assay/catalog.asp).

[³⁵S]GTPγS Binding at 5-HT_1A Receptors. The effects of the different compounds tested on [³⁵S]GTPγS binding were evaluated as previously described (Testa et al., 1999) with minor modifications. Cell membranes from HeLa cells transfected with human cloned 5-HT_1A receptors were resuspended in buffer containing 20 mM HEPES, 3 mM MgCl₂, and 120 mM NaCl, pH 7.4. The membranes were incubated with 10 μM GDP and decreasing concentrations of test drugs (from 10 μM to 0.01 nM) or decreasing concentrations of 5-HT (from 100 μM to 0.1 nM, reference curve; data not shown) for 20 min at 30°C in a final volume of about 0.25 ml. [³⁵S]GTPγS (200–250 pM in 10 μl) was added to the samples and incubated for an additional 30 min at 30°C. Nonspecific binding was determined in the presence of 10 μM GTPγS. The incubation was stopped by the addition of ice-cold HEPES buffer and rapid filtration through Unifilter GF/C filters using a Filtermate cell harvester (PerkinElmer Life and Analytical Sciences, Boston, MA). The filters were washed four times with a total of 1.2 ml of the same buffer. Radioactivity was counted by liquid scintillation spectrometry with an efficiency >90% (TopCount Packard; PerkinElmer Life and Analytical Sciences).

The inhibition of specific binding by the compounds was analyzed to estimate the IC₅₀ value by the nonlinear curve-fitting using Prism 3.02 software (GraphPad Software Inc., San Diego, CA). The IC₅₀ value was converted to K_i using the equation by Cheng and Prusoff (1973). Similar fitting procedures were used for the data obtained with the [³⁵S]GTPγS binding.

Electrophysiological Recordings. Methods for obtaining slices and recordings in the two preparations were essentially those previously reported in detail (Corradetti et al., 1996, 1998) with minor modifications.

Male Wistar rats (Harlan Italy, Udine, Italy) were anesthetized with ether and decapitated with a guillotine. The brain was quickly removed and cooled in partially frozen oxygenated artificial cerebrospinal fluid (aCSF) and was bubbled with a 95% O₂/5% CO₂ gas mixture.

Preparation of Slices from DRN and Hippocampus. Using a vibratome (T1000; DSK, Nagoya, Japan), a block of tissue containing the dorsal raphe was cut into sections (350–400 μm thick) while immersed in ice-cold aCSF with the following composition: 120 mM NaCl, 3.5 mM KCl, 1.2 mM NaH₂PO₄, 1.3 mM MgCl₂, 2 mM CaCl₂, 25 mM NaHCO₃, and 11 mM d-glucose, pH 7.3. The hippocampi were rapidly isolated and placed on ice-cold oxygenated aCSF with the following composition: 124 mM NaCl, 3.0 mM KCl, 1.25 mM NaH₂PO₄, 1.4 mM MgSO₄, 2 mM CaCl₂, 25 mM NaHCO₃, and 11 mM d-glucose, pH 7.4. Hippocampal slices from the dorsal hippocampus (400 μm thick) were cut using a McIlwain tissue chopper (Mickle Laboratory Engineering Co. Ltd., Gomshall, UK).

After sectioning, slices were kept in oxygenated aCSF for at least 1 h at room temperature (20–23°C). A single slice was then placed on a nylon mesh and completely submerged in a small chamber and superfused with oxygenated aCSF (32–34°C) at a constant flow rate of 2 to 3 ml/min. The drugs were administered through a three-way tap, and a complete exchange of the chamber volume occurred in 1 min. In some experiments, the DRN slices were incubated in Rec 27/0224 for >240 min in oxygenated aCSF (32–34°C), whereas the controls were incubated in similar conditions in the presence of the solvent used for solubilizing Rec 27/0224.

Extracellular Recording of Serotonergic Cell Firing in DRN Slices. Extracellular recordings were made with glass microelectrodes filled with 2 M NaCl (12–15 MΩ). Firing was facilitated by adding the α-adrenoceptor agonist phenylephrine (3 μM) to the superfusing aCSF (Vandermaelen and Aghajanian, 1983). Cells were identified as serotonergic neurons according to the following criteria: biphasic or triphasic action potentials of 2- to 3-ms duration, slow (0.5–2.5 Hz) and regular pattern of discharge, and inhibition of cell firing produced by 5-HT_1A receptor activation. Electric signals were fed into a high-input impedance amplifier (NL 102G; Digitimer Ltd., Hertfordshire, UK), an oscilloscope, and an electronic rate meter (D130; Digitimer Ltd.); triggered by individual neuronal action potentials; and connected to an A/D converter and a personal computer. Using dedicated software, the integrated firing rate was recorded, computed, and displayed on a chart recorder as consecutive 10-s samples.

Intracellular Recording from CA1 Pyramidal Cells. CA1 pyramidal neurons were recorded in current-clamp mode with 3 M KCl-filled (35–50 MΩ) or 2 M potassium methylsulfate-filled (45–80 MΩ) electrodes. Electrical signals were amplified with an Axoclamp 2A (Axon Instruments Inc., Union City, CA) and displayed on an oscilloscope and chart recorder (2800; Gould Instrument Systems Inc., Cleveland, OH). Traces were stored on a digital tape (sampling frequency 48 kHz, DTR 1200; Bio-Logic, Claix, France) and on a computer using pClamp software (pClamp 6.02; Axon Instruments) for off-line analysis.

Several criteria were used for choosing cells for the experiments as follows: stable resting membrane potential (r.m.p.) of at least –60 mV and with no spontaneous action potential firing; no sudden decreases in the cell membrane input resistance (R_in), indicating cell damage; and constant amplitude of the spike (≥80 mV) obtained by direct activation of the cell during the control period before tetrodotoxin application. When the cells appeared to have reached stable membrane potentials, pulses of hyperpolarizing current (200–400 pA, 400 ms, 0.05–0.1 Hz) were delivered through the recording electrode to monitor changes in R_in during drug application.

Concentration-Response Curves and Data Analysis. 5-Carboxamidoytryptamine (5-CT) was used for to evoke 5-HT_1A receptor-mediated responses in CA1 pyramidal cells and DRN serotonergic cells because it has a fast and almost full agonist action in CA1, whereas other 5-HT_1A receptor agonists (e.g., 8-OH-DPAT; Beck et al., 1992) produce much slower and/or partial responses in this region.

The lack of selectivity of 5-CT for 5-HT_1A versus 5-HT₇ or 5-HT_1B receptors should not be a drawback in our experimental conditions because both the responses evoked by 5-CT, cell membrane hyperpolarization in the CA1 and inhibition of cell firing in the DRN, depend solely on 5-HT_1A receptor activation. In fact, effect of 5-CT is fully blocked by WAY-100635 in both regions (Corradetti et al., 1996, 1998). Equiactive 5-CT concentrations producing about 90% of the maximum response in CA1 pyramidal cells (100 nM; Beck et al., 1992; Corradetti et al., 1996) and in the DRN (30 nM; Williams et al., 1988; Corradetti et al., 1998) were used for testing the effects of Rec 27/0224 and Rec 27/0074 on electrophysiological responses to 5-HT_1A receptor activation.

In the CA1 region, the effect of 5-CT was measured as the maximal hyperpolarizing response obtained with drug application. In pilot experiments and from previous experience (Corradetti et al., 1996, 1998), the application of 5-CT for 3 to 5 min elicited a steady-state response.

To obtain the IC₅₀ of Rec 27/0074 or Rec27/0224 for a given concentration of 5-CT, the effect of 5-CT was tested in control aCSF and in the presence of various concentrations of the drug (≥30 min) and gradually increased using a cumulative protocol. The magnitude of the response obtained in the presence of each concentration of Rec 27/0074 or Rec 27/0224, normalized versus that obtained in control aCSF, was fitted to a hyperbolic function: E = E_max/[1 + (IC₅₀/[D])ⁿ], where E was the response produced by the drug under study at the concentration [D], E_max was the maximal response, and n was the slope index. A nonlinear regression fitting was carried out using Prism 3.02.

In the DRN, a baseline firing rate of at least 5 min (5–15 min) was recorded before the application of 5-CT that was superfused until the maximal response was obtained (2.5–3.5 min). In each experiment, drug application time was kept constant, and the interval for measurement of the drug effect, chosen based on time elapsed after 5-CT application (corrected for dead space time) to 50% recovery of the firing rate, was kept constant for the following applications in the presence of the various drugs under study. This procedure was favored versus the measurement of peak responses to 5-CT because it also took the duration of 5-CT effects into consideration and allowed for better measurements of smaller responses.

The number of action potentials discharged during control 5-CT application was measured and compared with that expected in the same interval of time calculated based on the average firing rate recorded for 2 min immediately preceding 5-CT application. All the responses to 5-CT obtained in the presence of the inverse agonists were normalized versus the response to 5-CT in the control. In experiments in which citalopram effects were tested, steady-state responses were measured at the last minute of drug application (10 min) for each concentration.

Drugs. The compounds coded as Rec and WAY-100635 were synthesized in the Recordati Chemical Department (Recordati S.p.A, Milan, Italy) according to the methods described by Leonardi et al. (2002) for Rec 27/0074, Rec 27/0224, and Rec 15/3079 and by Cliffe and Mansell (2000) for WAY-100635. [³H]8-OH-DPAT, [³H]prazosin, [³H]rauwolscine, [³H]spiperone, and [³⁵S]GTPγS were obtained from PerkinElmer Life and Analytical Sciences. For receptor binding studies, the compounds were dissolved in absolute alcohol or deionized water according to their solubility.

Spiperone hydrochloride and 5-CT maleate were from Sigma/RBI (Natick MA); GDP, GTPγS, and phenylephrine hydrochloride were from Sigma-Aldrich (Milan, Italy); and tetrodotoxin was from Alomone Labs (Jerusalem, Israel). Citalopram was from Recordati S.p.A.

In the electrophysiological experiments, Rec 27/0224 was dissolved (25 mM) in dimethylformamide and stored in a refrigerator at 4°C. Aliquots (10 μl) were taken the day of the experiment and added with 240 μl of Tween 80/H₂O (1:1000 v/v) to obtain an intermediate stock solution of 1 mM, which was further diluted >1000-fold in aCSF and used for the experiments. In control experiments (n = 3), the application (90 min) of the solvent at a concentration 3-fold higher than that used for the present experiments did not affect the response to 5-CT.

Statistical Analysis. Numerical data are given as the mean ± S.E.M. Wilcoxon or Mann-Whitney tests were used for statistical analysis as appropriate; a value of P < 0.05 was considered statistically significant.

Results

The activity at 5-HT_1A receptors of Rec 27/0224 and Rec 27/0074 was investigated using radioligand and [³⁵S]GTPγS binding methods. In these assays, spiperone, a known inverse agonist at 5-HT_1A receptors (Newman-Tancredi, 1997), was used for comparison.

Radioligand Binding Studies. Rec 27/0224 and Rec 27/0074 showed high-affinity binding at human 5-HT_1A receptors expressed in HeLa cells with K_i values for displacing 1.04 nM (pK_i = 8.98) and 0.9 nM (pK_i = 9.05) of [³H]8-OH-DPAT, respectively (Table 1). Both compounds were highly selective for 5-HT_1A receptors showing a moderate affinity for h5-HT_2B (pK_i = 7.0) and h5-HT₇ (pK_i = 7.15), and pK_i values ≤ 6.5 for rat 5-HT_1B, bovine 5-HT_1D, and human recombinant 5-HT_2A, 5-HT_2B, 5-HT_2C, 5-HT₃, 5-HT_4c, 5-HT_4d, 5-HT_4e, 5-HT_5A, 5-HT₆, and 5-HT₇ serotonin receptor subtypes. Rec 27/0224 was 200- to 10,000-fold more selective for 5-HT_1A receptors versus human α_1a-, α_1b-, and α_1d-adrenoceptors, rat α₂-adrenoceptors, and D₂ receptors. Rec 27/0074 showed a lower (4–10-fold) selectivity versus human α₁-adrenoceptor subtypes. For comparison, binding data for spiperone and for two structurally related 5-HT_1A receptor antagonists, WAY-100635 and Rec 15/3079, are included in Table 1.

Functional Studies—[³⁵S]GTPγS Binding. The functional intrinsic activity of Rec 27/0224, Rec 27/0074, and spiperone was measured with [³⁵S]GTPγS binding studies carried out in HeLa cells stably expressing 5-HT_1A receptors. All the compounds concentration-dependently inhibited basal binding of [³⁵S]GTPγS, an effect typical of inverse agonism (Fig. 1).

The pEC₅₀ and maximal inhibition of [³⁵S]GTPγS binding, calculated from concentration-binding curves, for Rec 27/0224 was 8.58 [95% confidence limit (CL), 8.80–8.36] and 44.8 ± 1.7%, respectively. This effect of Rec 27/0224 was comparable with that of spiperone, for which a pEC₅₀ value of 8.17 (95% CL, 8.21–8.12) and a maximum [³⁵S]GTPγS binding inhibition value of 40.2 ± 0.4% were found (data not shown). Rec 27/0074 produced a smaller (25 ± 2.5%) inhibition of [³⁵S]GTPγS binding (pEC₅₀, 8.86; 95% CL, 8.96–8.77).

Functional Studies—Electrophysiology. Experiments were carried out in 83 DRN and 32 CA1 hippocampal slices obtained from 74 rats.

Effects of Rec 27/0224 and Rec 27/0074 on 5-HT_1A Receptor-Mediated Hyperpolarization in CA1 Pyramidal Cells. To investigate the effects of inverse agonists at the level of postsynaptic cells, we tested Rec 27/0224 and Rec 27/0074 on the 5-HT_1A receptor-mediated hyperpolarization produced by 5-CT in hippocampal CA1 pyramidal neurons. Intracellular recordings from CA1 pyramidal cells (r.m.p., –63.5 ± 0.9 mV; R_in, 48.2 ± 1.8 MΩ; n = 32) were performed in the presence of tetrodotoxin (1 μM) to functionally isolate postsynaptic effects of the drugs. Rec 27/0224 and Rec 27/0074 (1–300 nM) did not significantly affect cell r.m.p. and R_in (Wilcoxon test), although a slight depolarizing trend was observed in the presence of Rec 27/0224. After 60- to 90-min application of 300 nM Rec 27/0224, or at the end of cumulative concentration-response curves, the change in membrane potential (ΔV_m) was 1.8 ± 0.9 mV, whereas R_in was 103 ± 3% of the control (n = 9). Similarly, a ΔV_m of 0.5 ± 0.9 V_m and a R_in of 102 ± 4% (n = 5) were found for 300 nM Rec 27/0074. The lack of hyperpolarizing effects of the two compounds indicates the absence of agonist action at 5-HT_1A receptors. Both compounds decreased 5-HT_1A receptor-mediated hyperpolarization produced by 5-CT (100 nM) in a concentration-dependent manner (Fig. 2) but were unable to completely block 5-CT responses within the range of concentrations tested (1–300 nM).

The IC₅₀ of Rec 27/0224, obtained from the curve relating the effects of 5-CT in the presence of increasing concentrations of Rec 27/0224, was 18.0 nM (95% CL, 3.2–90.2 nM), and the maximal block of 5-CT-induced hyperpolarization was 53.4% (95% CL, 38.4–76.4%). In five of these experiments, concentration-response curves could also be obtained for individual cells, leading to an average IC₅₀ value of 18.2 ± 14.5 nM and a calculated maximal effect of 54.5 ± 4.5%. In additional experiments, a single concentration of Rec 27/0224 (300 nM), corresponding to the highest used in cumulative curves, was applied for 60 to 90 min. The resulting block of 5-CT-induced hyperpolarization (47.2 ± 15.8%; n = 4) was not statistically different (Mann-Whitney test) from that observed with concentration-response experiments. The other inverse agonist, Rec 27/0074, appeared more potent in antagonizing 5-CT-induced hyperpolarization of CA1 pyramidal cells with an IC₅₀ of 0.8 nM (95% CL, 0.5–1.3 nM) and maximal block of 74.1% (95% CL, 64.2–84.0%).

Effects of Rec 27/0224 and Rec 27/0074 on the 5-HT_1A Receptor-Mediated Response of DRN Serotonergic Cells. Extracellular recordings of DRN serotonergic cell firing were used to investigate the effects of inverse agonists on 5-HT_1A autoreceptors. In these experiments, multiple applications of the 5-HT_1A receptor agonist 5-CT were used to reversibly inhibit cell firing in control conditions and in the presence of a single concentration of an inverse agonist. Application of Rec 27/0224 or Rec 27/0074 (10–300 nM), per se, increased the cell firing rate in about 80% of the recordings with no change or small (<10%) decreases in the remaining cells.

It is noteworthy that the firing rate of serotonergic cells is dependent on the activation of α₁-adrenoceptors, for which Rec 27/0224 and Rec 27/0074 have a moderate affinity (Table 1). However, because in our experimental conditions α₁-adenoceptors are maximally activated by phenylephrine and the two compounds did not decrease the cell firing rate, the contribution of α₁-adrenoceptor antagonism to their effects can be ruled out.

The increase in cell firing rate often developed slowly, reaching a steady state within 30 to 40 min of application. At the highest concentration tested (300 nM, 60 min), Rec 27/0224 and Rec 27/0074 increased the firing rate of serotonergic cells by 21.0 ± 6% (n = 6) and 22.5 ± 8.3% (n = 7), respectively.

Rec 27/0224 antagonized the 5-CT-mediated inhibition of cell firing in a time- and concentration-dependent manner. The threshold concentration of Rec 27/0224 for observing a depression in the response to 30 nM 5-CT was 10 nM (Fig. 3A). The effect of Rec 27/0224 was detectable after 30 min of application and increased during the 1st h. To assess whether the effect of Rec 27/0224 had reached a steady state within 1 h of drug application, in a separate set of experiments, the slices were preincubated for 240 to 280 min in the presence of the compound. As shown in Fig. 3, B and C, the preincubation with 10 nM Rec 27/0224 decreased the response to 5-CT by 27.8 ± 8.9% (n = 9), a reduction no different from that obtained at 60 min of superfusion (24.7 ± 7.1%; n = 6; Mann-Whitney test). The antagonism of responses to 5-CT developed with a similar time course with higher concentrations of Rec 27/0224 (Fig. 4A). These results show that Rec 27/0224 had reached a steady state within 60 min of application. Furthermore, repetitive applications of 5-CT were not responsible for the slow time course of Rec 27/0224 effects. Indeed, a single application of 5-CT after superfusion of 100 nM Rec 27/0224 for 60 min (Fig. 4B) produced a response very similar to that observed in experiments in which 5-CT was applied three times during Rec 27/0224 application (29.3 ± 5.4, n = 5 and 29.4 ± 7.4, n = 5, respectively; Mann-Whitney test).

Concentration-response relationships for the inhibition of 5-CT responses by Rec 27/0224 and Rec 27/0074 at an application of 60 min are shown in Fig. 4C. The IC₅₀ values obtained from curves for Rec 27/0224 and Rec 27/0074 were 34.9 nM (95% CL, 22.5–54.3 nM) and 16.5 nM (95% CL, 11.9–22.8 nM), respectively. In contrast with the results obtained in CA1 pyramidal cells, in serotonergic cells, both compounds apparently fully antagonized the 5-CT-evoked inhibition of the firing rate.

To compare the effects of inverse agonists in the two regions with those of orthosteric antagonists, we reanalyzed raw data obtained with WAY-100635 in our previous work (Corradetti et al., 1996, 1998). The IC₅₀ values of WAY-100635 were 0.98 ± 0.09 nM (n = 5) and 3.27 ± 03 (n = 6) for the effect of 100 nM 5-CT on pyramidal cells and 30 nM 5-CT on serotonergic cells, respectively. Therefore, it appears that inverse agonists, particularly Rec 27/0224 and the neutral antagonist WAY-100635, have opposite region-potency relationships. In a final set of experiments, we investigated whether inverse agonists could block the effects of endogenous 5-HT on the firing rate of serotonergic cells (Fig. 5).

To activate 5-HT_1A receptors with endogenous 5-HT and thereby inhibit serotonergic cell firing, we used citalopram, which elicits an increase in extracellular 5-HT levels through the selective block of 5-HT transporters (Arborelius et al., 1995).

Superfusion with increasing concentrations of citalopram caused a concentration-dependent inhibition of serotonergic cells with an IC₅₀ value of 59.9 nM (95% CL, 58.7–61.1 nM; n = 6; Fig. 5B). Pretreatment of DRN slices with Rec 27/0224 (100 nM; 60 min; n = 5) potently antagonized citalopram effects (Fig. 5, A and B). Rec 27/0074 exerted a similar effect on citalopram-elicited inhibition of serotonergic cell firing (Fig. 5C). Therefore, the two inverse agonists were able to antagonize the inhibitory effect of endogenous 5-HT on the firing of serotonergic cells.

Discussion

Our data show that two selective 5-HT_1A receptor inverse agonists display differential antagonism of electrophysiological responses at native 5-HT_1A receptors expressed by serotonergic cells of DRN and CA1 pyramidal neurons. These results provide the first electrophysiological characterization of the functional effects of 5-HT_1A receptor inverse agonists in brain slices and show that a region-specific block of 5-HT_1A receptor-mediated responses can be achieved using this class of 5-HT_1A receptor ligands. Rec 27/0224 and Rec 27/0074 are high-affinity ligands at 5-HT_1A receptors with a high degree of selectivity for this 5-HT receptor subtype (Table 1).

Both compounds have significant inverse agonist activity in [³⁵S]GTPγS assays, qualitatively similar to that of spiperone, a recognized 5-HT_1A receptor inverse agonist (Newman-Tancredi et al., 1997). In the same conditions, a structurally related compound, Rec 15/3079, acted as a neutral antagonist in [³⁵S]GTPγS binding assays (Leonardi et al., 2001), as well as WAY-100635 (Newman-Tancredi et al., 1997; Testa et al., 1999). The binding affinities of Rec 27/0224 and Rec 27/0074 at 5-HT_1A receptors were similar, whereas their efficacy in inhibiting [³⁵S]GTPγS binding differed significantly, indicating that other intrinsic properties of the two compounds, and not affinity for 5-HT_1A receptors, were responsible for their differences in efficacy.

One of the functional consequences of inverse agonism at a given receptor is antagonism of the responses elicited by agonists. Accordingly, Rec 27/0224 and Rec 27/0074 behaved as antagonists in electrophysiological studies. However, comparison of their effects with those of the orthosteric antagonist WAY-100635 revealed peculiar pharmacological characteristics of inverse agonists.

Characteristics of 5-HT_1A Receptor Antagonism by Inverse Agonists. In a substantial fraction of the recorded serotonergic cells, the firing rate was increased by Rec 27/0224 and Rec 27/0074. This effect developed slowly and had a variable magnitude. However, the increase in firing rate by Rec 27/0224 and Rec 27/0074 cannot directly be interpreted as a phenotypic manifestation of inverse agonism in electrophysiological assays, analog to the decrease in [³⁵S]GTPγS binding. In fact, WAY-100635, a neutral antagonist at 5-HT_1A receptors, also produces variable increases in the cell firing rate. This effect has been ascribed to relief from the inhibition of cell firing exerted by endogenous 5-HT at 5-HT_1A receptors (Craven et al., 1994; Gartside et al., 1995; Corradetti et al., 1996; Fornal et al., 1996; Mundey et al., 1996), and the variability of responses likely results from different levels of extracellular 5-HT in individual in vitro preparations. Notably, the effect of WAY-100635 was much faster than that of inverse agonists.

A characteristic that more clearly differentiates Rec 27/0224 and Rec 27/0074 from neutral antagonists is the slow kinetics to reach steady-state antagonism of 5-HT_1A receptor-mediated responses. In DRN, the fact that the effect of 10 nM Rec 27/0224 reached its plateau at 60 min and progressed no further at 240 min allows the conclusion that the effect was concentration-dependent, but that the mechanisms of action of the drug needed time to fully develop its effects. A low diffusion rate in the tissue was unlikely because Rec 27/0224 is much more lipophilic than Rec 27/0074, but both needed >40 min to reach a steady-state response. The possibility that slow cellular mechanisms were involved (e.g., disappearance of receptors through internalization or persistent uncoupling from G proteins) is also unlikely because, in that case, prolonged application of the drugs would have produced continuous increases in the drug effect.

In contrast, the relatively long time required by Rec 27/0224 and Rec 27/0074 for achieving their full effects could be explained by an allosteric action typical of inverse agonists, comprising a slow shift of the receptors toward inactive conformation(s). Several further differences characterize Rec 27/ 0224 and Rec 27/0074 effects compared with those of neutral antagonists.

For instance, in the same experimental conditions used in the present work, WAY-100635 fully blocked the effects of 5-CT in the CA1 pyramidal cells at lower concentrations than those needed in DRN serotonergic cells (Corradetti et al., 1996, 1998). In contrast, inverse agonists, and in particular Rec 27/0224, were more effective in the DRN than in the hippocampus.

More interestingly, inverse agonists did not fully antagonize the 5-HT_1A receptor-mediated cell membrane hyperpolarization in CA1 pyramidal cells. In individual experiments, a substantial fraction (20–50%) of the response to 5-CT was left unaffected by both compounds at concentrations fully effective in serotonergic cells. A similar region-specific antagonism of 5-HT_1A electrophysiological responses was found in vivo using the inverse agonist spiperone (Blier et al., 1993).

Considering that Rec 27/0224 and Rec 27/0074 can completely displace 8-OH-DPAT in binding assays, the expected effect of these inverse agonists in electrophysiological experiments in the hippocampus was full antagonism of 5-CT actions. However, the binding studies were done on cell fragments, whereas the functional studies were performed on intact cells. The partial antagonism of electrophysiological responses in hippocampus by Rec 27/0224 and Rec 27/0074 suggests the existence of two functional subpopulations of 5-HT_1A receptors in intact CA1 pyramidal cells. One possibility is that in these neurons 5-HT_1A receptors can take on a conformation fully responsive to agonists but not to inverse agonists. An alternative explanation, although thermodynamically unfavorable, is that these compounds induce conformational changes of 5-HT_1A receptors, leading to forms with a low affinity for inverse agonists but still responsive to agonists.

Molecularly identical 5-HT_1A receptors may take on conformations differentially responsive to ligands when located in specific subcellular microenvironment (e.g., somatic versus dendritic spines or trunk) or in different neurons. This difference can be explained based on the expected properties of the interaction of inverse agonists with GPCR. In fact, the shift between active and inactive receptor conformations that an inverse agonist ligand can produce is constrained by equilibrium constants proper of that GPCR. On the other hand, these constants depend on the microenvironment surrounding the GPCR (Christopoulos and Kenakin, 2002). This encompasses the type of G protein coupled, the “tightness” of coupling, the number of receptors, and/or phosphorylation, etc. and therefore may be different for the same receptor protein in different neurons. Thus, coupling of 5-HT_1A receptors to different G proteins in CA1 pyramidal cells and serotonergic cells may per se explain differential receptor behavior in the two brain regions, as suggested by Hensler (2003). It has consistently been reported (Mannoury La Cour et al., 2001) that 5-HT_1A receptors are coupled to Gαo proteins in the hippocampus and to Gαi3 proteins in the DRN. Interestingly, inverse agonists have been found to inhibit Gαi3 activation by 5-HT_1A receptors expressed in Chinese hamster ovary cells (Newman-Tancredi et al., 2002).

In summary, regardless of the exact molecular mechanism of action of these compounds in the hippocampus, the pharmacological properties of Rec 27/0224 and Rec 27/0074 support the notion that these ligands acted through a mechanism different from orthosteric antagonism and that a region-specific block of 5-HT_1A receptors can be achieved by these ligands.

Possible Therapeutic Implications of Region-Specific Antagonism at 5-HT_1A Receptors. Administration of SSRI causes an immediate increase in brain serotonin level, which appears to be necessary for the subsequent development of the antidepressant action. The increase in serotonin level at the same time activates 5-HT_1A somatodendritic autoreceptors, leading to inhibition of serotonergic cell firing. This negative feedback is believed to contribute to the slow onset of the therapeutic effect of SSRI (2–3 weeks). 5-HT_1A receptor antagonists block the inhibition of serotonergic cell discharge caused by the SSRI-elicited increase in 5-HT levels (Arborelius et al., 1995; Romero and Artigas, 1997), suggesting the potential utility of 5-HT_1A receptor antagonists in hastening the therapeutic effect of antidepressant drugs (see Artigas et al., 1996). For this purpose, an ideal antagonist should selectively block raphe 5-HT_1A autoreceptors without affecting postsynaptic 5-HT_1A heteroreceptors. However, the currently available antagonists appear devoid of regional selectivity (Corradetti et al., 1996, 1998).

The observed block of the citalopram-induced inhibition of DRN cell firing by Rec 27/0224 and Rec 27/0074 and their limited effects at hippocampal 5-HT_1A receptors suggest that inverse agonists may accelerate the onset of SSRI therapeutic actions in vivo (Gartside et al., 1995; see Artigas et al., 1996) with minimal interference at postsynaptic 5-HT_1A receptors.

In conclusion, we show that two 5-HT_1A receptor inverse agonists display a considerable degree of region selectivity and are effective in blocking the effects of SSRI in the raphe. As to whether these characteristics are relevant to therapeutic effects needs to be seen in in vivo testing.