Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

The heptadecapeptide FGGFTGARKSARKLANQ, named nociceptin (Meunier et al., 1995) or orphanin FQ (Reinscheid et al., 1995), is the endogenous ligand of the opioid receptor-like 1 (ORL1) receptor, a G-protein-coupled receptor (GPCR) belonging to the opioid receptor family. Although nociceptin/orphanin FQ (noc/OFQ) shows some structural analogy to opioid peptides and acts at the molecular and cellular level in an essentially identical fashion to µ-, -, and -opioids, it does not activate classical opioid receptors. Furthermore, despite the high degree of amino acid and structural conservation between the classical opioid receptors and the ORL1 receptor, no opioid receptor ligands have been found to bind to the ORL1 receptor with high affinity (reviewed in Darland et al., 1998; Taylor and Dickenson, 1998). Noc/OFQ has been implicated in many behavioral and physiological processes, including nociception, allodynia, locomotion, learning, stress response, sexual behavior, feeding, pituitary function, cardiovascular control, and sodium balance, and its pharmacological actions differ considerably from those of the classical opioids (reviewed in Darland et al., 1998; Taylor and Dickenson, 1998).

To further define the physiological roles of noc/OFQ, selective receptor antagonists are needed. To date, three selective peptide ligands showing antagonist properties at the noc/OFQ receptor have been reported. A truncated analog of noc/OFQ, [Phe¹psi(CH₂-NH)Gly²]noc/OFQ(1-13)-amide {[F/G]NC(1-13)NH₂}, was found to act as a competitive antagonist of noc/OFQ actions in the electrically stimulated guinea pig ileum and mouse vas deferens (Guerrini et al., 1998) preparations. Furthermore, this peptide blocked the cardiovascular effects of noc/OFQ (Madeddu et al., 1999), the noc/OFQ-mediated inhibition of the rat rostral ventrolateral medulla neurons (Chu et al., 1999), and the increase in potassium conductance in amygdaloid neurons (Meis and Pape, 1998). However, in other assays, especially for effects on central sites, it was found to act as a partial (Schlicker et al., 1998; Chiou, 1999) or even full agonist (Calo et al., 1998b; Carpenter and Dickenson, 1998; Grisel et al., 1998; Schlicker et al., 1998; Xu et al., 1998; Kapusta et al., 1999; Yakimova and Pierau, 1999; Okawa et al., 1999), although it competitively inhibited the noc/OFQ-stimulated binding of GTPS in brain preparations obtained from both the rat and mouse (Berger et al., 1999b). Another peptide, acetyl-RYYRIK-amide (Ac-RYYRIK-NH₂), originally identified from a combinatorial library of acetylated hexapeptide amides as a partial agonist for the ORL1 receptor transfected into Chinese hamster ovary (CHO) cells (Dooley et al., 1997), was shown to competitively inhibit the noc/OFQ-evoked coupling of its receptor to G-proteins in rat brain preparations, as measured in GTPS binding assays (Berger et al., 1999a). In agreement with this antagonism, Ac-RYYRIK-NH₂ competitively antagonized the chronotropic effect of noc/OFQ in rat cardiomyocytes (Berger et al., 1999a). However, when tested in vivo for effects on spontaneous locomotor activity in mice, it was found to act as a highly potent full agonist for locomotor inhibition (Berger et al., 1999b).

We have recently reported that the new noc/OFQ analog, [Nphe¹]NC(1-13)-amide {[Nphe¹]NC(1-13)NH₂}, behaves as a pure, selective, and competitive noc/OFQ receptor antagonist (Calo et al., 2000). [Nphe¹]NC(1-13)NH₂ competitively antagonized the in vitro inhibitory effects of noc/OFQ on electrically evoked contractions in several isolated tissues (Calo et al., 2000) and on forskolin-stimulated cAMP accumulation in CHO cells expressing the human ORL1 receptor (Hashimoto et al., 2000). In these cells and in the mouse colon (Rizzi et al., 1999), [Nphe¹]NC(1-13)NH₂ antagonized the effects of ORL1 receptor ligands, including [F/G]NC(1-13)NH₂, which acted as a full agonist in these preparations. Furthermore, [Nphe¹]NC(1-13)NH₂ prevented the pronociceptive and antimorphine actions of i.c.v. noc/OFQ in the mouse tail withdrawal assay (Calo et al., 2000) and the stimulatory effect of i.c.v. noc/OFQ on food intake in the rat (Polidori et al., 2000). [Nphe¹]NC(1-13)NH₂ represents the first selective and competitive noc/OFQ receptor antagonist devoid of any residual agonist activity that is also capable of antagonizing the central effects of noc/OFQ that are mimicked by [F/G]NC(1-13)NH₂ or Ac-RYYRIK-NH₂.

For a GPCR, e.g., the noc/OFQ receptor, the model of receptor G-protein coupling implicates that antagonism of this coupling should lead to antagonism of the biological activity evoked by receptor activation. The aim of this study was to investigate the effect of the noc/OFQ antagonist [Nphe¹]NC(1-13)NH₂ on the receptor G-protein coupling measured using GTPS binding in rat and mouse brain preparations and to compare its action with that of the mixed agonist/antagonist peptides.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials and Buffer. The noc/OFQ receptor peptide ligands Ac-RYYRIK-NH₂, [F/G]NC(1-13)NH₂, and [Nphe¹]NC(1-13)NH₂ were synthesized in our institutes as previously described (Calo et al., 1998a). The selective agonists for the µ- and -opioid receptor, [D-Ala²,N-Me-Phe⁴,Gly⁵-ol]-enkephalin and [D-Pen²,D-Pen⁵]-enkephalin, respectively, were obtained from Sigma (Deisenhofen, Germany) and the -opioid receptor agonist U-50,488 was from Research Biochemicals International (Natick, MA). GTP³⁵S (1250 Ci/mmol) was purchased from NEN (Boston, MA). GDP was obtained from Sigma. Bacitracin, obtained from MERCK (Darmstadt, Germany), was heated for 1 h at 70°C in water to inactivate any enzymatic activity before use. Tris/HCl (50 mM, pH 7.4) containing 0.2 mM EGTA and further additions as specified was used as buffer in all experiments.

GTP³⁵S Binding to Rat Brain Membranes and Coronal Sections of Mouse Brain. The cerebral cortex or whole brain (without cerebellum) of male Wistar rats (approximately 250 g) was homogenized with an Ultra-Turrax homogenizer (Janke & Kunkel, Staufen, Germany) in buffer, and the total membrane fraction was obtained by centrifugation for 20 min at 26,000g at 4°C. Approximately 20 µg of membrane protein was incubated with 50 to 150 pM GTP³⁵S in the presence and absence of the indicated peptides at 25°C or 30°C for 2 h in a total volume of 500 µl of buffer supplemented with 1 mg/ml BSA, 0.15 mM bacitracin, and NaCl, GDP, and MgCl₂ at the indicated concentrations. The reaction was terminated by filtration through Whatman GF/B filters using a Brandel harvester (Gaithersburg, MD), and the filters were counted for ³⁵S activity. The concentration-response curves of noc/OFQ in the absence and presence of [Nphe¹]NC(1-13)NH₂ were fitted by nonlinear regression using the program PRISM 2.0 (GraphPad Software Inc., San Diego, CA), and from the EC₅₀ values, Schild plots were constructed.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Optimum Conditions for GTP³⁵S Binding to Brain Membranes and Sections. Noc/OFQ stimulation of GTP³⁵S binding to rat brain cortex membranes was systematically optimized by examining the effects of factors known to be critical in receptor-evoked GTP binding (Sim et al., 1995; Traynor and Nahorski, 1995). Both basal and noc/OFQ- (1 µM) stimulated GTP³⁵S binding were strongly decreased by increasing concentrations of GDP (Fig. 1a). Because the decrease in basal binding was slightly higher than that in stimulated binding, the stimulated-to-basal ratio, i.e., the stimulation factor, was markedly increased at higher GDP concentrations (Fig. 1b). Similarly, NaCl decreased both basal and stimulated binding and produced an increase in the stimulation factor. This salt effect was not specific for Na⁺ because similar data were obtained with KCl (data not shown). In contrast to the effects of GDP, Na⁺, and K⁺, MgCl₂ increased basal and stimulated binding with optimum stimulation occurring at approximately 1 mM. A free Mg²⁺ concentration of 0.3 mM was an absolute requirement for basal and noc/OFQ-stimulated binding because Mg²⁺ chelators prevented binding (data not shown). In agreement with the above data in membranes, the stimulation factor for noc/OFQ-evoked GTP³⁵S binding to mouse brain sections was markedly increased by GDP (data not shown), as already reported for µ-opioid receptor-stimulated binding to rat brain sections (Sim et al., 1995).

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Dependence on GDP concentration of the binding of 44 pM GTP35S to rat cortex membranes (20 µg) in the absence (basal) and presence (stimulated) of 1 µM noc/OFQ at 30°C for 1 and 3 h (50 mM Tris, pH 7.4, 100 mM NaCl, 3 mM MgCl2, 0.2 mM EGTA, 1 mg/ml BSA, 0.15 mM bacitracin). From the amount of GTP35S bound (a) the stimulation factor was calculated as stimulated/basal (b). Data in a represent the mean ± S. D. of triplicate determinations in a single experiment. Several additional experiments at various conditions (25 and 30°C, 1-3 h) showed similar results.

Influence of [Nphe¹]NC(1-13)NH₂ on GTP³⁵S Binding. As determined from the above results, the remainder of the studies used the optimized medium components of 1 mM MgCl₂, 100 to 200 µM GDP, and 100 mM NaCl in the studies with membranes. However, using brain sections, the GDP concentration was increased to concentrations as high as 1 mM. Noc/OFQ stimulated GTP³⁵S binding to rat cortex membranes with an EC₅₀ of 17.7 ± 1.15 nM (S.E., n = 4). Increasing concentrations of [Nphe¹]NC(1-13)NH₂ produced a rightward shift of the concentration-response curves of noc/OFQ without changing the maximum response (Fig. 2a), indicating competitive antagonism. From Schild plots with a mean slope of 0.89 (Fig. 2b), a pA₂ value of 7.76 ± 0.10 (S.E., n = 4) and a Schild constant of 19.0 nM were calculated. Furthermore, [Nphe¹]NC(1-13)NH₂ also inhibited the noc/OFQ-stimulated binding in sections obtained from mouse brain (Fig. 3).

View larger version (12K): [in this window] [in a new window]   Fig. 2.   Effect of increasing concentrations of [Nphe1]NC(1-13)NH2 on the concentration-response curves for noc/OFQ-stimulated binding of 74 pM GTP35S to 20 µg of rat cortex membranes in incubations for 2 h under optimized conditions (50 mM Tris, pH 7.4, 100 µM GDP, 100 mM NaCl, 1 mM MgCl2, 0.2 mM EGTA, 1 mg/ml BSA, 0.15 mM bacitracin) at 25°C (a) and the corresponding Schild plot (b). Values in a are expressed as difference of binding in the presence and absence of noc/OFQ (i.e., net binding) obtained from triplicate samples in a single experiment. Four independent experiments were performed yielding a pA2 of 7.76 ± 0.10 (S.E.).

View larger version (59K): [in this window] [in a new window]   Fig. 3.   Effect of [Nphe1]NC(1-13)NH2 on the stimulation of the binding of approximately 80 pM GTP35S by 1 µM noc/OFQ to 10-µm coronal sections obtained from mouse brain. Sections were incubated under optimized conditions (50 mM Tris, pH 7.4, 100 mM NaCl, 3 mM MgCl2, 1 mM GDP, 0.2 mM EGTA) for 2 h at 25°C in the absence (upper row) and presence of noc/OFQ (middle row) and in the presence of noc/OFQ and 10 µM [Nphe1]NC(1-13)NH2 at the same time (lower row). Autoradiograms representative of sections at the level of frontal cortex (left sections) and striatum (right sections) are shown.

View larger version (30K): [in this window] [in a new window]   Fig. 4.   Effect of [Nphe1]NC(1-13)NH2 (10 µM) on the stimulation of GTP35S (80 pM) binding by opioid agonists to whole rat brain (without cerebellum) membranes (20.5 µg). Incubations were for 2 h at 25°C in 50 mM Tris, pH 7.4, 200 µM GDP, 100 mM NaCl, 1 mM MgCl2, 0.2 mM EGTA, 1 mg/ml BSA, 0.15 mM bacitracin with agonists (1 µM) specific for the noc/OFQ, µ- {[D-Ala2,N-Me-Phe4,Gly5-ol]-enkephalin (DAMGO)}, - {[D-Pen2,D-Pen5]-enkephalin (DPDPE)}, and -opioid receptors (U50,488) in the absence and presence of [Nphe1]NC(1-13)NH2. ***P < .001 versus samples without [Nphe1]NC(1-13)NH2; in all other cases there were no significant differences. Three independent experiments resulting in the same significancies were performed.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Influence of noc/OFQ (), [Nphe1]NC(1-13)NH2 (), [F/G]NC(1-13)NH2 (), and Ac-RYYRIK-NH2 () on the binding of approximately 100 pM GTP35S to 20.5 µg of rat cortex membranes. Incubations were for 2 h at 25°C (50 mM Tris, pH 7.4, 100 mM NaCl, 1 mM MgCl2, 0.2 mM EGTA, 1 mg/ml BSA, 0.15 mM bacitracin) at 100 µM GDP (a) and at 5 µM GDP (b). Values are expressed as difference of binding in the presence and absence of the peptides (net binding) obtained from triplicate samples in a representative experiment from four similar experiments. From these and further experiments, the activities relative to noc/OFQ of Ac-RYYRIK-NH2, [F/G]NC(1-13)NH2, and [Nphe1]NC(1-13)NH2 at 100 µM GDP were found to be [% ± S.E.M. (n)] 6.83 ± 1.42 (5), 15.59 ± 2.82 (5), and 1.70 ± 1.21 (18), respectively. At 5 µM GDP values of 24.0 ± 2.84 (6), 27.96 ± 4.66 (6), and 8.11 ± 3.86 (6) were obtained.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

Noc/OFQ was found to stimulate the binding of GTP³⁵S to rat and mouse brain preparations (Sim et al., 1996; Shimohira et al., 1997; Albrecht et al., 1998) and cells transfected with the ORL1 receptor (Adapa and Toll, 1997; Dooley et al., 1997) as anticipated for GPCRs. In this study, [Nphe¹]NC(1-13)NH₂ was shown to inhibit noc/OFQ-stimulated binding of GTP³⁵S to rat cortex membranes without changing the maximum agonist stimulation (Fig. 2), indicating a competitive antagonism of noc/OFQ receptor G-protein coupling. In addition, [Nphe¹]NC(1-13)NH₂ also inhibited noc/OFQ-stimulated GTP³⁵S binding to mouse brain sections (Fig. 3). In rat brain membranes, [Nphe¹]NC(1-13)NH₂ did not significantly inhibit the GTP³⁵S binding stimulated by µ-, -, and -opioid receptor agonists (Fig. 4). These data confirm the selectivity of [Nphe¹]NC(1-13)NH₂ for the noc/OFQ receptor, as earlier demonstrated in direct receptor binding studies or testing the peptide against the effects elicited by opioid agonists in electrically stimulated isolated tissues (Calo et al., 2000), and, in addition, are in agreement with the antagonism of the compound in vitro (Rizzi et al., 1999; Calo et al., 2000; Hashimoto et al., 2000) and in vivo (Calo et al., 2000; Polidori et al., 2000).

Two other noc/OFQ receptor-specific peptide ligands, Ac-RYYRIK-NH₂ and [F/G]NC(1-13)NH₂, were previously found to antagonize noc/OFQ-mediated GTP³⁵S binding to rat cortex membranes and rat and mouse brain sections (Berger et al., 1999a,b). At first glance, no marked differences in this antagonism between Ac-RYYRIK-NH₂, [F/G]NC(1-13)NH₂, and [Nphe¹]NC(1-13)NH₂ were observed except for the lower antagonist potency of the latter compound as shown by the Schild constants of 6.58 nM (Berger et al., 1999a), 3.83 nM (Berger et al., 1999b), and 19.0 nM (Fig. 2), respectively. According to the model of coupling of a GPCR to G-proteins, antagonists of receptor-evoked GTP binding would be expected to inhibit all subsequent signaling steps and effects. Contrary to expectation, [F/G]NC(1-13)NH₂ behaved as a partial or full noc/OFQ receptor agonist in a number of assays in vitro (Schlicker et al., 1998; Chiou, 1999; Okawa et al., 1999) as well as in vivo (Calo et al., 1998b; Carpenter and Dickenson, 1998; Grisel et al., 1998; Xu et al., 1998; Kapusta et al., 1999; Yakimova and Pierau, 1999), in addition to exhibiting antagonist behavior for several actions of noc/OFQ at some central (Meis and Pape, 1998; Chu et al., 1999) and peripheral (Guerrini et al., 1998; Madeddu et al., 1999) sites. Similarly, Ac-RYYRIK-NH₂ produced agonist effects for inhibition of locomotoric activity in mice (Berger et al., 1999b), but antagonist effects on noc/OFQ-evoked increase in beat frequency of rat myocardiocytes (Berger et al., 1999a). Of these three peptides described, only [Nphe¹]NC(1-13)NH₂ acted as a pure antagonist in vitro and in vivo.

The main drawback in the quantitative determination of receptor G-protein coupling using the GTP³⁵S binding assays is the low amount of receptor-stimulated tracer binding compared with basal. Optimum conditions for high stimulation factor of GTP³⁵S binding in membranes required high GDP concentrations of approximately 100 µM GDP (Fig. 1). However, for the activation or inhibition of the signal transduction cascade downstream of G-proteins, the absolute amount of activated G-proteins, i.e., of G-protein-bound GTP, rather than the relative increase in binding elicited by the receptor agonist is important. Lowering the GDP concentration from 100 to 5 µM increased this amount and showed that [F/G]NC(1-13)NH₂ and Ac-RYYRIK-NH₂ were able to stimulate net GTP binding to nearly 30% of that evoked by noc/OFQ. In contrast, [Nphe¹]NC(1-13)NH₂ did not stimulate binding (Fig. 5). These data indicate that at low occupation of the nucleotide binding site by GDP in the inactivated G-protein, the relative efficacy of low-efficacy agonists in the GTPS binding assay is enhanced. Similar results were obtained with cannabinoid (Griffin et al., 1999) and µ-opioid receptor ligands (Selley et al., 1997).

The mechanism underlying the dependence of relative efficacy in G-protein activation on GDP concentration remains to be established. It is possible that the G-protein receptor complexes produced by low- and high-efficacy agonists differ kinetically in that low-efficacy agonists activate empty G-protein nucleotide-binding sites more effectively than sites occupied by GDP, from where GDP has to dissociate before GTP can be bound. This would agree with the interpretation of cannabinoid receptor-stimulated GTP³⁵S binding curves in the presence of different GDP concentrations that cannabinoid agonists decreased the affinity of GDP to G-proteins in proportion to their efficacies (Breivogel et al., 1998). Furthermore, because 100 µM GDP was found to decrease the affinity of noc/OFQ in receptor binding studies 2- to 3-fold (data not shown), GDP could exert an allosteric influence on the activation of the G-protein differently for agonists of different efficacy, even at equilibrium.

Given the fact that [F/G]NC(1-13)NH₂ and Ac-RYYRIK-NH₂, but not [Nphe¹]NC(1-13)NH₂, exhibit partial agonist activity with respect to G-protein coupling in vitro in membranes under certain conditions, it is also suggested that they are likely to do so in intact cellular systems and in vivo. Why these peptides showed biological activities in vitro as well as in vivo ranging from full agonism to pure antagonism and why [Nphe¹]NC(1-13)NH₂ antagonized the noc/OFQ-stimulated binding with relatively high potency (pA₂ 7.76) when compared with the much lower potency for antagonism of the inhibitory effect of noc/OFQ on cAMP accumulation in CHO cells (pA₂ 6.2; Hashimoto et al., 2000), in the mouse colon (pA₂ 6.0; Rizzi et al., 1999), and on electrically evoked contractions in mouse and rat vas deferens and guinea pig ileum (pA₂ 6.0 to 6.4; Calo et al., 2000) require further clarification.

From results obtained with the ORL1 (Toll et al., 1998), 5-HT_1A (Newman-Tancredi et al., 1997), and µ-opioid receptors (Selley et al., 1998) in transfected cells, it can be concluded that high levels of receptor expression promote agonism for GTP binding activities of low-efficacy ligands. The extent of such activity in a native system may, therefore, depend on the ratio between receptors and G-proteins and possibly on other factors, e.g., GDP. Furthermore, if partial agonist activity in GTP binding can lead to full biological activity, as observed with [F/G]NC(1-13)NH₂ and Ac-RYYRIK-NH₂ in some systems, it must be assumed that a certain portion of the GTP binding capacity, stimulated by the full agonist noc/OFQ, is already sufficient to evoke maximal response. Such a receptor G-protein coupling reserve is in line with the higher potency of Ac-RYYRIK-NH₂ and similar hexapeptides for inhibition of adenylate cyclase when compared with their potency for stimulating GTP³⁵S binding in cells transfected with the ORL1 receptor (Dooley et al., 1997). Differences in partial GTP binding activity of a ligand and in coupling reserve between different cells and tissues could then explain how the same ligand exhibits a wide spectrum of effects from full agonism (sufficiently high partial agonism and coupling reserve) to antagonism (low agonism and reserve). For a pure receptor antagonist, this would mean that, as observed with [Nphe¹]NC(1-13)NH₂, the potency for antagonism of receptor-stimulated GTP binding is higher than that measured for downstream events because increasing inhibition of the coupling with increasing antagonist concentrations will lead to inhibition of the biological effect only after part of GTP binding has already been inhibited. Collectively, the screening of noc/OFQ receptor agonists and antagonists at the level of G-protein coupling as measured by receptor-stimulated GTP³⁵S binding may be misleading if not carefully analyzed.

In summary, although [F/G]NC(1-13)NH₂, Ac-RYYRIK-NH₂, and [Nphe¹]NC(1-13)NH₂ all competitively antagonized the noc/OFQ-evoked receptor G-protein coupling as measured in an optimized GTP³⁵S binding assay, [F/G]NC(1-13)NH₂ and Ac-RYYRIK-NH₂ showed partial agonist activity in this assay under certain conditions, whereas [Nphe¹]NC(1-13)NH₂ behaved as a pure antagonist under all conditions. The spectrum of agonist, partial agonist, and antagonist behavior of [F/G]NC(1-13)NH₂ and Ac-RYYRIK-NH₂ in different in vitro and in vivo systems may possibly be explained by differences in efficacy of GTP binding agonism and in coupling reserve between the different systems.