Abstract
In hippocampal membranes, the selective 5-hydroxytryptamine (5-HT1A) receptor agonists 8-hydroxy-dipropylaminotetralin (8-OH-DPAT) andN,N-dipropyl-5-carboxamidotryptamine (N,N-DP-5-CT) stimulated guanosine-5′-O-(3-thio)triphosphate ([35S]GTPγS) binding by 130 to 140%; binding stimulated by nonselective agonists (5-HT and 5-CT) was ∼30% greater. However, the selective 5-HT1A receptor antagonistN-[2-[4-(2-methoxyphenyl)-1-piperazinyl]ethyl]-N-2-pyridinyl-cyclohexanecarboxamide (WAY100,635) completely abolished the increases produced by 8-OH-DPAT and N,N-DP-5-CT but only eliminated 70% of that elicited by 5-CT. The rank potency order of the tested agonists was identical with their rank order of affinity for 5-HT1A receptors [5-CT ≅ N,N-DP-5-CT > R-(+)-8-OH-DPAT > 5-HT > ipsapirone]. Racemic 8-OH-DPAT and the partial agonist ipsapirone exhibited lower intrinsic activity thanR-(+)-8-OH-DPAT. R-(+)-8-OH-DPAT also stimulated [35S]GTPγS binding in cortex, but not in striatum, which lacks 5-HT1A receptors. Partial irreversible inactivation of 5-HT1A receptors, in vitro with phenoxybenzamine (0.3 or 1 μM) or in vivo withN-ethoxycarbonyl-2-ethoxy-1,2-dihydroquinoline (1 mg/kg), reduced the maximal response produced byR-(+)-8-OH-DPAT but did not alter its EC50. In autoradiographic sections, R-(+)-8-OH-DPAT stimulated [35S]GTPγS binding in 5-HT1A receptor-rich regions (dorsal hippocampus, 123%; lateral septum, 111%; midhippocampus, 110%; dorsal raphe nucleus, 83%; medial prefrontal cortex, ∼60%). The EC50 ofR-(+)-8-OH-DPAT did not vary significantly among brain regions (46–96 nM). Partial irreversible blockade of 5-HT1A receptors in brain sections (phenoxybenzamine, 10 μM) reduced the maximal response without altering the EC50 in both the hippocampus and dorsal raphe. Despite prior evidence that dorsal raphe somatodendritic 5-HT1Aautoreceptors exhibit high receptor/effector coupling efficiency (receptor reserve) compared with postsynaptic receptors in hippocampus, there was no evidence of a difference at the level of receptor/G protein coupling.
Of the ∼15 serotonin receptors that have been identified to date (Boess and Martin, 1994), the 5-hydroxytryptamine (5-HT)1A receptor has arguably received the most attention, primarily because selective 5-HT1Areceptor agonists (buspirone, ipsapirone, and gepirone) display anxiolytic/antidepressant effects, and because considerable evidence has accumulated that its function is altered after repeated treatment with anxiolytic and antidepressant drugs (Blier and De Montigny, 1994). Very recent studies in 5-HT1A receptor knockout mice (Heisler et al., 1998; Parks et al., 1998; Ramboz et al., 1998) have directly demonstrated the importance of this receptor in animal models of anxiety and depression: the mutant animals exhibited both increased anxiety and behaviors similar to those observed after antidepressant treatment. The 5-HT1A receptor is located both presynaptically on 5-HT cell bodies (as somatodendritic autoreceptors) in the dorsal and median raphe nuclei, and postsynaptically primarily in the limbic system (hippocampus, lateral septum, entorhinal and medial prefrontal cortex) (Vergé et al., 1986; Pompeiano et al., 1992; Kia et al., 1996). [Terminal autoreceptors are not 5-HT1A (Vergé et al., 1985) but 5-HT1B/D (Sari et al., 1997)]. Somatodendritic autoreceptor stimulation inhibits the firing of 5-HT neurons (Sprouse and Aghajanian, 1988) via membrane hyperpolarization consequent to activation of a pertussis toxin-sensitive G protein-coupled K+ conductance (Innis and Aghajanian, 1987); the decrease in impulse flow results in a decrease in 5-HT synthesis (Hjorth and Magnusson, 1988; Meller et al., 1990) and release (Hjorth and Sharp, 1991) in terminal areas innervated by these neurons. Some postsynaptic 5-HT1A receptors (e.g., in hippocampus) also mediate a hyperpolarizing response by increasing a K+ conductance (Beck et al., 1992), whereas others (whose physiological function is unclear) are coupled to inhibition of forskolin-stimulated adenylyl cyclase activity (Bockaert et al., 1987; Yocca et al., 1992).
Studies assessing 5-HT1A receptor function were complicated by findings that the potency and intrinsic activity of various 5-HT1A receptor drugs differed when examined in models of pre- and postsynaptic receptor activation; these drugs were more potent and efficacious at somatodendritic autoreceptors in the dorsal raphe than at postsynaptic receptors in, e.g., the hippocampus (for review, see Meller et al., 1990). In a series of studies using a variety of pre- and postsynaptic models of 5-HT1A receptor function, we showed that this was due to differences in receptor/effector coupling efficiency (Meller et al., 1990, 1992; Yocca et al., 1992; Cox et al., 1993), methodologically defined as differences in receptor reserve (Furchgott and Bursztyn, 1967): somatodendritic autoreceptors displayed a large receptor reserve, whereas postsynaptic receptors did not. From a pharmacological perspective, this served to explain how a particular drug, such as the weak partial agonist BMY 7378, could demonstrate apparent full agonism in the dorsal raphe yet act as an antagonist at postsynaptic receptor sites in the hippocampus (Meller et al., 1990). Although these studies proved valuable in defining the pharmacological activity of 5-HT1A receptor drugs for eliciting various functional responses, they provided only an overall assessment of the efficiency of the signal transduction cascade between receptor occupation and response at pre- and postsynaptic 5-HT1A receptor sites.
A direct measure of the initial, activation step of receptor/G protein coupling can be obtained from agonist-stimulated guanosine-5′-O-(3-thio)triphosphate ([35S]GTPγS) binding to receptors (Lazareno and Birdsall, 1993; Wieland and Jakobs, 1994). Recently, agonist-stimulated binding of [35S]GTPγS to 5-HT1A receptors has been demonstrated in hippocampal membranes (Sim et al., 1997; Alper and Nelson, 1998), cloned cell membranes (Newman-Tancredi et al., 1997), and brain sections by autoradiography (Sim et al., 1997; Waeber and Moskowitz, 1997; Dupuis et al., 1998b). In the present study, this technique was used to determine whether previously described regional differences in receptor/effector-coupling efficiency are demonstrable at the level of receptor/G protein coupling. Two consequences would be expected if this were the case. First, the potency of a full agonist for stimulating [35S]GTPγS binding should be greater (i.e., lower EC50) at somatodendritic autoreceptors in the dorsal raphe than at postsynaptic receptors in the hippocampus. Second, partial irreversible blockade of 5-HT1Areceptors should shift the EC50 for the agonist to the right in the dorsal raphe (Meller et al., 1990) but only reduce the maximum response (without altering the EC50) in the hippocampus (Yocca et al., 1992). The present results demonstrate that neither of these expectations was fulfilled.
Experimental Procedures
Animals.
Male Sprague-Dawley rats (200–250 g; Taconic Farms, Germantown, NY) were maintained on a 12-h light/dark cycle and housed four per cage with food and water ad libitum.
Materials.
5-HT hydrochloride,R-(+)-8-hydroxy-dipropylaminotetralin hydrobromide (8-OH-DPAT), (±)-8-OH-DPAT, 5-carboxamidotryptamine (5-CT) maleate,N,N-dipropyl-5-carboxamidotryptamine (N,N-DP-5-CT) maleate, WAY100,635 {N-[2-[4-(2-methoxyphenyl)-1- piperazinyl]ethyl]-N-2-pyridinyl-cyclohexanecarboxamide} maleate, phentolamine mesylate, and phenoxybenzamine hydrochloride (PBZ) were obtained from Research Biochemicals (Natick, MA). EEDQ (N-ethoxycarbonyl-2-ethoxy-1,2-dihydroquinoline) was purchased from Aldrich Chemicals (Milwaukee, WI), adenosine deaminase, GDP, and GTPγS were purchased from Sigma Chemical Co. (St. Louis, MO), and [35S]GTPγS (1250 Ci/mmol) was obtained from NEN (Boston, MA). SuperFrost/Plus slides were purchased from Fisher Scientific Co. (Pittsburgh, PA).
Drug Treatments and Tissue Dissections.
In some experiments, rats were injected with vehicle or the irreversible receptor antagonist EEDQ (1 mg/kg s.c.) and sacrificed 24 h later. Whole hippocampus, cerebral cortex, and corpus striatum were grossly dissected as described previously (Meller et al., 1990; Meller and Bohmaker, 1996).
[35S]GTPγS Binding in Membranes.
The method is essentially identical with that described by Sim et al. (1997). Brain tissues were homogenized in ice-cold homogenization buffer (50 mM Tris-HCl, 3 mM MgCl2, 1 mM EGTA, pH 7.4), centrifuged at 48,000g for 10 min at 4°C, and pellets were washed once by resuspension. Membrane pellets were resuspended in assay buffer (50 mM Tris-HCl, 3 mM MgCl2, 0.2 mM EGTA, 100 mM NaCl, pH 7.4) and stored in aliquots at −70°C. Thawed aliquots were diluted with assay buffer and preincubated at 30°C for 10 min with adenosine deaminase (10 mU/ml final concentration) to reduce basal binding (Sim et al., 1998), centrifuged, and resuspended in assay buffer at a protein concentration of 100 μg/ml. In some experiments, membranes were exposed to the receptor alkylating PBZ (0.3 or 1 μM final concentration) or vehicle during the 10-min preincubation period and were washed once by resuspension. Membranes (10 μg of protein) were incubated for 1 h at 30°C with 0.05 nM [35S]GTPγS and 20 μM GDP in the presence or absence of various concentrations of drugs (1 ml total volume). Basal binding was measured in the absence of drugs, and nonspecific binding in the presence of 10 μM nonradioactive GTPγS. The reaction was stopped by rapid filtration over Whatman GF/B filters, the filters were washed with 3 × 5 ml of ice-cold wash buffer (50 mM Tris-HCl, pH 7.4) and bound radioactivity was quantitated by liquid scintillation counting.
[35S]GTPγS Autoradiography.
The method is identical with that described by Childers and colleagues (Sim et al., 1997). Harvested brains were frozen by slow immersion in 2-methylbutane maintained at −35°C and stored frozen at −70°C. Coronal 12-μm sections were cut on a cryostat (Reichert-Jung model 2800 Frigocut N) maintained at −17°C, with the atlas of Paxinos and Watson (Paxinos and Watson, 1986) to identify regions of interest. Sections were thaw-mounted on SuperFrost/Plus slides and stored on ice during collection. The slides were dried under vacuum overnight at 4°C, then stored with dessicant at −70°C. Thawed slides were dried in a stream of cool air for 30 min and transferred to five-slide plastic mailers. Sections were equilibrated in assay buffer (50 mM Tris-HCl, 3 mM MgCl2, 0.2 mM EGTA, 100 mM NaCl, pH 7.4) for 10 min at 25°C; in some experiments PBZ (10 μM) or vehicle was present during this equilibration period. Sections were then preincubated in assay buffer containing 2 mM GDP, adenosine deaminase (10 mU/ml final volume) and, where appropriate, antagonist drugs (WAY100,635 or phentolamine), for 15 min at 25°C, followed by incubation (2 h; 25°C) in fresh assay buffer containing 0.04 nM [35S]GTPγS, 2 mM GDP, and adenosine deaminase in the presence or absence of various concentrations of agonist/antagonist drugs. Basal binding was carried out in the absence of any drugs and nonspecific binding in the presence of 10 μM GTPγS. Incubations were terminated by two washes (2 min each) in ice-cold wash buffer (50 mM Tris-HCl, pH 7.0) and a brief rinse in ice-cold distilled water. Sections were dried overnight at room temperature and apposed to autoradiographic film (Hyperfilm βmax; Amersham, Arlington Heights, IL) for 3 to 5 days. Autoradiograms were analyzed with a computerized image analysis system (MCID-M4; Imaging Research, St. Catherines, Ontario, Canada). In a standard assay, various amounts of [35S]GTPγS were added to brain pastes, which were frozen and cut on a cryostat. Brain paste sections were exposed to autoradiographic film together with [14C] microscales (Amersham), which allowed for conversion of optical densities to nCi [35S]/mg tissue equivalent (Sim et al., 1995). However, it was found that radioactivity and optical density were linearly correlated at least up to A1.5; therefore, autoradiographic films (maximum A, <1.0) were routinely subjected to relative quantitation, and data were plotted as percentage of increase above basal.
Data Analysis.
All dose-response curves were fit with the ALLFIT program of De Lean et al. (1978). This iterative program allowed dose-response curves to be simultaneously analyzed for best fit with the four-parameter logistic equation Y = (a − d)/[(1 +X/c)b] + d, where a is the response at zero dose, b is the slope factor, c is the ED50, andd is the response at “infinite” dose. Y is the response elicited by a particular dose X. Fits were obtained with and without constraints being placed on values for these parameters. The program calculated partial F tests, which were used to determine whether a particular set of constraints significantly worsened the fit obtained relative to a standard set of constraints or no constraints at all. Extensive use of the program has been described previously (Meller et al., 1990, 1992; Yocca et al., 1992; Cox et al., 1993).
Results
5-HT1A Receptor Specificity in Hippocampal Membranes.
Dose-response curves for a number of 5-HT agonists were generated in hippocampal membranes (Fig.1). The mean EC50and Emax values are shown in Table1. The rank potency order of the tested agonists was identical with their rank order of affinity (Hibert et al., 1987) for 5-HT1A receptors (5-CT ≅ N,N-DP-5-CT > R-(+)-8-OH-DPAT > 5-HT > ipsapirone). For R-(+)-8-OH-DPAT, the EC50 (28 nM) andEmax (127% stimulation above basal) were essentially identical with those reported previously (Sim et al., 1997). Because the stimulation elicited by the nonspecific 5-HT1 agonists 5-HT and 5-CT was ∼30% greater than that produced by the selective 5-HT1Aagonists R-(+)-8-OH-DPAT and N,N-DP-5-CT, we investigated the possibility that the former drugs were stimulating additional 5-HT1 receptor subtypes (5-HT1B-F) to yield a greater percentage of stimulation. Indeed, the selective 5-HT1Areceptor antagonist WAY100,635 (1 μM) (Forster et al., 1995) completely abolished the stimulation of [35S]GTPγS binding produced by 1 μMR-(+)-8-OH-DPAT or 0.1 μM N,N-DP-5-CT, but only eliminated 70% of the increase produced by 0.3 μM 5-CT (Fig.2), supporting the idea that ∼30% of the stimulation produced by 5-CT was via non-5-HT1A receptor sites. Although WAY100,635 shows some affinity (∼100-fold less) for α1-receptors (Forster et al., 1995), the α-adrenergic antagonist phentolamine (1 μM) did not alterR-(+)-8-OH-DPAT-stimulated [35S]GTPγS binding (data not shown). As expected (Pauwels et al., 1997; Alper and Nelson, 1998), racemic 8-OH-DPAT displayed lower intrinsic activity, and the known partial agonist ipsapirone displayed ∼45% of the intrinsic activity ofR-(+)-8-OH-DPAT; WAY100,635 behaved as a neutral antagonist (Fig. 1; Table 1). Based on these results, R-(+)-8-OH-DPAT was used to stimulate [35S]GTPγS binding in all subsequent studies.
R-(+)-8-OH-DPAT-Stimulated [35S]GTPγS Binding in Cortex and Striatum.
The specificity of 5-HT1A-receptor-stimulated [35S]GTPγS binding byR-(+)-8-OH-DPAT was further substantiated by detection of agonist-stimulated binding in the cerebral cortex but not in the striatum (Fig. 3), which is known to be devoid of 5-HT1A receptors (Pompeiano et al., 1992). As in the hippocampus, the stimulation in the cortex was completely blocked in the presence of 1 μM WAY100,635 but not by 1 μM phentolamine (data not shown).
Receptor/G Protein Coupling Efficiency for Agonist-Stimulated [35S]GTPγS Binding in Hippocampus.
Dose-response curves were generated for R-(+)-8-OH-DPAT-stimulated binding of [35S]GTPγS in hippocampal membranes after partial irreversible receptor blockade either in vitro by treatment with PBZ (Fig. 4A) or in vivo by treatment with EEDQ (Fig. 4B). Either treatment reduced the maximal response without altering the EC50, similar to the results obtained for inhibition of forskolin-stimulated adenylyl cyclase in the hippocampus after partial irreversible receptor blockade (Yocca et al., 1992). Thus, as expected, 5-HT1Areceptor/G protein coupling in the hippocampus exhibited low efficiency (i.e., no receptor reserve).
Autoradiographic Analysis ofR-(+)-8-OH-DPAT-Stimulated [35S]GTPγS Binding in Different Brain Regions.
Figure5 shows stimulated binding of [35S]GTPγS in the dorsal hippocampus as a function of R-(+)-8-OH-DPAT concentration. Highest stimulated binding was observed in the molecular layer of the dentate gyrus and in strata oriens and radiatum, with lower levels in CA2 and CA3. This laminar and subfield distribution corresponds exactly to that of 5-HT1A receptor-binding sites (Pompeiano et al., 1992). Incubation with the specific 5-HT1A receptor antagonist WAY100,635 (1 μM) abolished the increase produced by 1 μM R-(+)-8-OH-DPAT, but the α-adrenergic antagonist phentolamine (1 μM) had no effect (data not shown). Stimulated binding also was observed in hypothalamic and amygdaloid nuclei, but did not appear to be completely eliminated by WAY 100,635 and was not characterized further. Figure6 shows the dose dependence forR-(+)-8-OH-DPAT-stimulated binding in the dorsal raphe, which was likewise abolished by 1 μM WAY100,635 but not by phentolamine (data not shown). Typical sections depicting nonspecific, basal and 1 or 10 μM R-(+)-8-OH-DPAT-stimulated binding in the mid-hippocampal formation, the lateral septum, and the medial prefrontal cortex are shown in Fig. 7. The medial prefrontal cortex exhibited the lowest maximal stimulation (∼60%) of the brain regions examined. Consistent with the results in membranes (Fig. 3), there was a low level of stimulated binding in the cortex (Figs. 5-7), but not in the striatum (Fig. 7, middle).
Dose-response analyses for R-(+)-8-OH-DPAT-stimulated [35S]GTPγS binding were carried out in sections from four brain regions and are shown in Fig.8. The EC50 forR-(+)-8-OH-DPAT was similar in all brain regions, varying over an ∼2-fold range (46–96 nM), which ALLFIT analysis indicated was not significantly different (see legend to Fig. 8). However, both the slope factor and the maximal response were significantly lower in the dorsal raphe than in the other regions. These results were interesting, given that receptor/effector coupling efficiency was previously demonstrated to be high in the dorsal raphe and low in hippocampus (Meller et al., 1990; Yocca et al., 1992). Consequently, a lower EC50 (greater potency) forR-(+)-8-OH-DPAT was anticipated in the dorsal raphe in comparison to the hippocampus. The lower intrinsic activity ofR-(+)-8-OH-DPAT in the dorsal raphe (∼80% stimulation versus ∼120% in the other regions) is also at variance with expectation.
To further evaluate these findings, 5-HT1Areceptor/G protein coupling efficiency was directly assessed by generating dose-response curves before and after partial irreversible receptor blockade with PBZ. Preliminary experiments (data not shown) established that a 10-min preincubation with 10 μM PBZ was sufficient to reduce the maximal response to R-(+)-8-OH-DPAT ∼50%. Also, the effect of PBZ was completely prevented in the presence of 1 μM WAY100,635; thus, potential inactivation of other receptor sites by PBZ did not affect R-(+)-8-OH-DPAT-stimulated binding. (Analogous results were obtained previously in vivo with a 5-HT1A receptor drug to prevent the effects of EEDQ; Meller et al., 1990). As expected, partial irreversible inactivation of 5-HT1A receptors in the dorsal hippocampus and the mid-hippocampal formation decreased the maximal stimulation of [35S]GTPγS binding byR-(+)-8-OH-DPAT without altering the EC50 for the agonist (Fig.9). This is indicative of low receptor/G protein coupling efficiency and is consistent with the low overall receptor/effector coupling efficiency observed at postsynaptic sites in hippocampus with the adenylate cyclase assay (Yocca et al., 1992). However, PBZ treatment similarly reduced the maximalR-(+)-8-OH-DPAT-stimulated binding of [35S]GTPγS in the dorsal raphe nucleus without a change in the EC50, indicating that receptor/G protein coupling (as measured with this technique) is also of low efficiency at somatodendritic autoreceptors. The EC50 values for R-(+)-8-OH-DPAT in all three brain regions were not significantly different (see legend to Fig. 9). In contrast, overall receptor/effector coupling efficiency at the somatodendritic 5-HT1A autoreceptors in the dorsal raphe nucleus is high, as shown previously in biochemical and electrophysiological studies (Meller et al., 1990; Cox et al., 1993).
Discussion
The results obtained in the present study in regard to the distribution of 5-HT1A receptor-stimulated [35S]GTPγS binding sites in brain, and the intrinsic activity and potency of agonists, are in general agreement with those described previously (Sim et al., 1997; Waeber and Moskowitz, 1997; Alper and Nelson, 1998; Dupuis et al., 1998b). For example, in hippocampal membranes, the intrinsic activity (127%) and EC50 (28 nM) values forR-(+)-8-OH-DPAT-stimulated binding were essentially identical with those reported by Sim et al. (1997). The finding that the nonselective agonists 5-HT and 5-CT stimulated ∼30% more binding than the selective agonist R-(+)-8-OH-DPAT (Fig. 1), which was incompletely abolished by the selective antagonist WAY100,635 (Fig.2), was also similar to that reported previously (Alper and Nelson, 1998). Although the identity of 5-HT1 receptor subtypes that may be mediating the increased stimulation produced by 5-HT and 5-CT were not investigated in the present study, attempts by others to establish that stimulation of [35S]GTPγS binding occurs via other 5-HT receptor subtypes, with purported selective agonists or antagonists, have yielded mixed results. In autoradiographic studies in guinea pig brain, a selective 5-HT1B/1D antagonist blocked 5-CT-stimulated binding in the substantia nigra (rich in 5-HT1B/1D receptors) but not in other (5-HT1A receptor-rich) areas (Waeber and Moskowitz, 1997). Agonists with high affinity, and purported selectivity, for 5-HT1B/1D receptors stimulated binding not only in the substantia nigra but also in areas enriched in 5-HT1A receptors (hippocampus, lateral septum); the stimulation in the latter areas (but not in substantia nigra) was abolished by a 5-HT1A receptor antagonist (Waeber and Moskowitz, 1997). Attempts to visualize stimulated binding via 5-HT1F receptors in areas enriched in this receptor subtype (e.g., claustrum), with drugs with high affinity for the receptor, were unsuccessful; again, however, 5-HT1A receptor-stimulated response was easily detected (Waeber and Moskowitz, 1997). Others were unable to demonstrate 5-HT1B/1D receptor-mediated increased binding in guinea pig brain sections by autoradiography; indeed, no matter what the selectivity profile of the drug used, only 5-HT1A receptor-stimulated responses appeared to be elicited (Dupuis et al., 1998a,b). Thus, although the demonstration of an increased binding of [35S]GTPγS via various 5-HT receptor subtypes continues to be problematic,R-(+)-8-OH-DPAT-stimulated binding via 5-HT1A receptors was both highly specific and generally robust. The agonist-stimulated binding was completely abolished by the highly selective antagonist WAY100,635, but was unaffected by phentolamine, an antagonist of the only other receptor (α1-adrenergic) for which WAY100,635 has moderate affinity (Forster et al., 1995).
Initial experiments to determine whether differences in 5-HT1A receptor/G protein coupling efficiency are demonstrable at pre- versus postsynaptic sites focused on assessing the receptor reserve for R-(+)-8-OH-DPAT-stimulated [35S]GTPγS binding in membranes. As expected, in hippocampal membranes partial irreversible 5-HT1A receptor blockade, by treatment with either PBZ in vitro or EEDQ in vivo, reduced the maximum response to the agonist but did not alter its EC50 (Fig. 4), indicative of low receptor/G protein coupling efficiency. This finding is consistent with the low overall receptor/effector coupling efficiency observed at postsynaptic sites in hippocampus with the adenylyl cyclase assay (Yocca et al., 1992). Attempts to evaluate coupling efficiency in membranes of dorsal raphe tissue obtained by micropunch were hampered by the need to pool tissue from many animals, the labor-intensiveness of this task, the relatively low yield of protein recovered, and, not least, by an unexpectedly low percentage of stimulation even in control tissue. This is probably due to the difficulty of obtaining, even by micropunch, dorsal raphe uncontaminated by adjacent tissue. These difficulties rendered the generation of dose-response curves in PBZ-treated membranes highly problematic. Consequently, further experiments were carried out using quantitative autoradiographic analysis of brain sections, where the signal was fairly robust even in the dorsal raphe nucleus and dose-dependent stimulation could be readily ascertained (Figs. 5, 6, and 8).
The EC50 for R-(+)-8-OH-DPAT did not differ significantly between the dorsal raphe and either the dorsal or mid-hippocampus (Fig. 8), in contrast to many previous studies, using behavioral and electrophysiological paradigms, that established that 5-HT1A agonists such as 8-OH-DPAT are more potent in the former region than the latter (see references cited in Meller et al., 1990). The lack of a regional difference in receptor/G protein coupling efficiency indicated by this finding was clearly corroborated by partial irreversible receptor blockade experiments. With brain sections that visualized the dorsal raphe nucleus and the hippocampus at two different levels, PBZ treatment significantly reduced the maximal response to R-(+)-8-OH-DPAT in each region without significantly affecting its EC50 (Fig. 9). The results in the dorsal raphe using this technique stand in marked contrast to a previous study assessing overall receptor/effector coupling efficiency at somatodendritic autoreceptors in vivo, where partial irreversible 5-HT1A receptor blockade shifted the dose response for 8-OH-DPAT >8-fold to the right (Meller et al., 1990).
The absence of an observable difference in the efficiency of 5-HT1A receptor/G protein coupling in the dorsal raphe and the hippocampus may be due to one or more factors. First, and most parsimoniously, the high receptor/effector coupling efficiency observed in vivo at raphe nucleus somatodendritic 5-HT1A autoreceptors may reflect overall amplification of the individual steps in the signal transduction cascade that is not observed at the level of the activation step of receptor/G protein coupling. Differences in 5-HT1A receptor density, receptor/G protein ratio (i.e., stoichiometry), G protein subunits available for coupling (and their attendant variables such as kinetics and/or equilibria of receptor/G protein binding), and effectors may all contribute to the overall efficiency of receptor/effector coupling (Newman-Tancredi et al., 1997; Pauwels et al., 1997; Selley et al., 1998). [Receptor densities in the two regions are similar, however (Li et al., 1997)]. Interestingly, results paralleling those obtained herein have been reported for the μ-opiate receptor. Classical in vivo receptor inactivation experiments demonstrated that μ-opiate receptor-mediated antinociception displays high receptor/effector coupling efficiency (receptor reserve) (Zernig et al., 1995), whereas μ-opiate receptor-mediated stimulation of [35S]GTPγS binding in thalamic membranes exhibited an absence of receptor reserve (using an indirect method of assessment) (Selley et al., 1998).
The present finding of no difference in the efficiency of 5-HT1A receptor-stimulated G protein activation between the hippocampus and dorsal raphe supports the indirect findings of a previous study in guinea pig brain sections (Dupuis et al., 1998b). These authors found no significant difference in the relative potency and intrinsic activity of several 5-HT1Aagonists, including racemic 8-OH-DPAT, for stimulating [35S]GTPγS binding in hippocampus and dorsal raphe and postulated that the high GDP concentrations used may have masked differences in receptor/G protein coupling efficiency (although a differential receptor reserve in these brain regions in this species has not been established). This postulated masking effect may relate to the necessity of using high concentrations of GDP to promote guanine nucleotide exchange and optimize detection of agonist-stimulated [35S]GTPγS binding. It has been repeatedly noted that in membranes and brain sections GDP concentrations of 1 to 50 μM and 1 to 2 mM, respectively, are generally required to maximize the percentage of stimulation of binding by agonists of various receptors (Sim et al., 1995). Moreover, many investigators have noted that intrinsic activity differences among agonists are amplified as the concentration of GDP is increased; concomitantly, agonist potencies are reduced (Pauwels et al., 1997;Alper and Nelson, 1998; Selley et al., 1997, 1998). This is analogous to situations where overall receptor/effector coupling efficiency is low (no receptor reserve), and differences in intrinsic activity (maximal tissue response) are directly related to the efficacy of the agonists. In contrast, when receptor/effector coupling efficiency is high (large receptor reserve), even partial agonists display full intrinsic activity, and the potency of agonists is increased. The suggested mechanism (Selley et al., 1997) posits that high efficacy (i.e., full) agonists are inherently better able to overcome the effect of excess GDP to stabilize the inactive (GDP-bound) form of the G protein, and thus at high GDP concentrations the intrinsic activity differences between full and partial agonists are amplified. This hypothesis is supported by the observation that partial agonists are less able to promote the release of prebound GDP and thus display lower intrinsic activity for stimulating [35S]GTPγS binding (Lorenzen et al., 1996). However, there is no basis for expecting that high GDP concentrations would differentially affect the potency of a particular agonist [e.g., R-(+)-8-OH-DPAT] in different brain regions. Thus, although high GDP concentrations may reduce the potency of R-(+)-8-OH-DPAT in both brain regions, a true difference in potency (reflecting a difference in receptor/G protein coupling efficiency) should still be discernable. The analysis of receptor/G protein coupling by this technique may, for other unknown reasons, be inherently limited to the detection of low-efficiency coupling for this activation step in the signal transduction cascade, although it may be well suited for defining differences in intrinsic efficacy among agonists (Selley et al., 1998). Further studies (perhaps with different techniques) are needed to determine the contribution coupling efficiency at individual steps of the signal transduction pathway makes to overall assessments of receptor/effector coupling efficiency.
Acknowledgments
We thank Drs. Steven Childers and Laura Sim for helpful discussions in the establishment of the autoradiographic techniques.
Footnotes
-
Send reprint requests to: Dr. Emanuel Meller, Department of Psychiatry, New York University Medical Center, 550 First Ave., New York, NY 10016. E-mail: emanuel.meller{at}med.nyu.edu
-
↵1 This work was supported by U.S. Public Health Service Grant NS 23618.
- Abbreviations:
- 5-HT
- 5-hydroxytryptamine
- GTPγS
- guanosine-5′-O-(3-thio)triphosphate
- 8-OH-DPAT
- 8-hydroxy-dipropylaminotetralin hydrobromide
- 5-CT
- 5-carboxamidotryptamine
- N,N-DP-5-CT
- N,N-dipropyl-5-carboxamidotryptamine
- WAY100,635
- N-[2-[4-(2-methoxyphenyl)-1-piperazinyl]ethyl]-N-2-pyridinyl-cyclohexanecarboxamide maleate
- PBZ
- phenoxybenzamine hydrochloride
- EEDQ
- N-ethoxycarbonyl-2-ethoxy-1,2-dihydroquinoline
- Received July 16, 1999.
- Accepted November 3, 1999.
- The American Society for Pharmacology and Experimental Therapeutics