- Email: [email protected]
Evidence for Multiple Rat VPAC1 Receptor States with Different Affinities for Agonists
Cell. Signal. Vol. 11, No. 9, pp. 691–696, 1999 Copyright 1999 Elsevier Science Inc.
ISSN 0898-6568/99 $ – see front matter PII S0898-6568(99)00041-8
Evidence for Multiple Rat VPAC1 Receptor States with Different Affinities for Agonists Rebeca Busto,† Ma-Guillerma Juarranz,‡ Salvatore De Maria,§ Patrick Robberecht‡* and Magali Waelbroeck‡ †On leave from the Departmento de Bioquı´mica y Biologı´a Molecular, Universidad de Alcala`, Madrid, Spain; ‡Department of Biochemistry and Nutrition, Faculty of Medicine, Universite´ Libre de Bruxelles, CP 611, 808 Route de Lennik, B-1070 Brussels, Belgium; and §on leave from the Centro di Recerca Interdipartimentale di Scienze Computazionali e Biotecnologice (CRISCEB), Napoli, Italia
ABSTRACT. We compare the binding properties of [125I-VIP] and [125I]-Ro 25 1553 to VPAC1 receptors, expressed in stably transfected CHO cells. [125I]-VIP labelled two VPAC1 receptor states, while [125I]-Ro 25 1553 labelled selectively a limited number of high-affinity receptors. This high-affinity state probably corresponds to an agonist-receptor-Gs ternary complex as its properties (guanyl nucleotides, EC50 values and maximal effect) were affected by cholera toxin pre-treatment. Both high- and low-affinity receptors participated in the adenylate cyclase activation. This suggested that agonists activate not only low-affinity uncoupled receptors by facilitating the ternary complex formation, but also activated the high-affinity ternary complex by accelerating the GTP binding to emptied, receptor-bound G proteins. cell signal 11;9:691–696, 1999. 1999 Elsevier Science Inc. KEY WORDS. Vasoactive intestinal peptide, Ro 25 1553, VPAC1 receptor, Agonist, Antagonist, G protein, Ternary complex
INTRODUCTION Two receptors for Vasoactive Intestinal Peptide (VIP) have been cloned in several mammalian species, and are now known as VPAC1 and VPAC2 receptors [1]. These two receptors have similar high affinities for their natural ligands, VIP, PACAP-27, and PACAP-38 (Pituitary Adenylate Cyclase Activating Peptide, 1-27 or 1-38) [1]. We previously demonstrated that the cyclic peptide, Ro 25 1553, is highly selective for VPAC2 receptors [2], and developed selective VPAC1 receptor agonists and antagonists [3, 4]. These peptides can be used, in combination with molecular biology techniques, to identify the VIP receptor subtypes that are expressed in mammalian tissues [5] or cells lines [6] and are responsible for functional responses [7]. We previously observed that although [125I]-Ro 25 1553 is clearly VPAC2selective [2], it did label some receptors in stably transfected Chinese Hamster Ovary (CHO) cells expressing VPAC1 receptors [5]. We compared in this work the properties of the VPAC1 binding sites labelled by [125I]-Ro 25 1553 and [125I]VIP in stably transfected CHO cells. MATERIALS AND METHODS Peptide Synthesis All the peptides were synthesised by solid phase methodology using the Fmoc strategy [3], and purified by HPLC re*Author to whom all correspondence should be addressed. Tel: 132-2555-6229; fax: 132-2-555-6230; e-mail: [email protected]
verse phase and by ion exchange chromatography [2–4]. The peptide purity (.97%) was assessed by capillary electrophoresis and the conformity was verified by electrospray mass spectrometry and sequencing [2–4]. Cell Cultures, Cholera Toxin Treatment, and Membrane Preparations Transfected Chinese Hamster Ovary (CHO) cells, stably expressing the recombinant rat VPAC1 receptor (810 6 80 pmoles/mg of membrane protein), were cultured as previously described [8]. Some cells (as indicated) were pretreated 4h with 10 mg/ml cholera toxin and washed twice with fresh medium immediately before harvesting. For membrane preparations, confluent cells were harvested with a rubber policeman and pelleted by low speed centrifugation; the supernatant was discarded and the cells lysed in 1 mM NaHCO3 solution and immediately frozen in liquid nitrogen. After thawing, the lysate was first centrifuged at 800 3 g for 10 min. The supernatant was further centrifuged at 20,000 3 g for 10 min. The pellet, resuspended in 1 mM NaHCO3, was used immediately as a crude membrane preparation. Receptor Identification VIP and Ro 25 1553 were radioiodinated as previously described [5] by the iodogen method, purified by absorption on a Sep-Pak cartridge, and eluted with 50% acetonitrile in
692
R. Busto et al.
FIGURE 1. VIP and Ro 25 1553 competition curves at VPAC1
receptors. Specific [125I]-VIP binding to stably transfected CHO cells expressing VPAC1 receptors was measured in the absence or presence of the indicated VIP (closed circles) or Ro 25 1553 (closed squares) concentrations. Average of 3–7 competition curves in duplicate.
0.1% trifluoroacetic acid. Binding was performed at 37 8C in 20 mM Tris-maleate buffer (pH 7.4) enriched with 2 mM MgCl2, 0.1 mg/ml bacitracin, and 1% bovine serum albumin. Non-specific binding was defined as tracer binding in the presence of 1 mM VIP. The incubations were terminated by filtration through glass-fibre GF/C filters presoaked for 24 h in 0.1% polyethyleneimine, and the filters were rinsed three times with a 20 mM (pH 7.4) sodium phosphate buffer containing 1% bovine serum albumin [8]. Adenylate Cyclase Activity Adenylate cyclase activity was determined by the Salomon et al. procedure [9]. Membranes (3–15 mg protein) were incubated in a total volume of 60 ml containing 0.5 mM [a32 P]-ATP, 10 mM GTP, 5 mM MgCl2, 0.5 mM EDTA, 1 FIGURE 3. Guanyl nucleotides competition curves. Specific
[125I]-Ro 25 1553 (closed squares) and [125I]-VIP (closed circles) binding was measured in the absence or presence of the indicated concentrations of GTP (top panel), GDP (centre panel) or GTPgS (bottom panel). Average of 3 competition curves in duplicate.
mM cAMP, 1 mM theophylline, 10 mM phospho(enol) pyruvate, 3 mg/ml pyruvate kinase, and 30 mM Tris-HCl at a final pH of 7.5. The reaction was terminated after a 15 min incubation at 378C by addition of 0.5 ml of a 0.5% sodium dodecyl-sulfate solution containing 0.5 mM ATP, 0.5 mM cAMP, and 20,000 cpm [3H]-cAMP. cAMP was separated from ATP by two successive chromatographies on Dowex 50WX8 and neutral alumina. FIGURE 2. [125I]-VIP and [125I]-Ro 25 1553 saturation curves.
[125I]-VIP (closed circles) and [125I]-Ro 25 1553 (closed squares) binding to stably transfected CHO cells was measured at increasing tracer concentrations (0.1 to 3 nM), and the results presented according to Scatchard. Average of 4 saturation curves in duplicate.
RESULTS AND DISCUSSION Despite the low-affinity of the cyclic VIP analogue Ro 25 1553 for the rat VPAC1 receptor, as judged by its low capac-
Multiple VPAC1 Receptor States
693
TABLE 1. IC50 values of [125I]-Ro 25 1553 and [125I]-VIP binding for different peptides on
VPAC1 receptors expressed in CHO cells
VIP Helodermin (1-37) Secretin VIP (2-28) VIP (3-28) VIP (6-28) [D-Phe4]-VIP [D-Ala4]-VIP PACAP (1-23) [R16]-PACAP (1-23) VPAC1 agonist ([K15, R16]-VIP(1-7)GRF(8-27)) VPAC1 antagonist ([D-Phe2, K15, R16]-VIP(1-7)GRF(8-27)) Ro 25 1553
ity to inhibit 125I-VIP binding (Fig. 1) and by its low Kact value for adenylate cyclase activation (not shown), its iodinated derivative [125I]-Ro 25 1553 labelled specifically membranes prepared from CHO cells expressing the recombinant VPAC1 rat receptor. There was no detectable [125IVIP and [125I]-Ro 25 1553 binding to non transfected CHO cells or to CHO cells transfected with a non-related receptor (data not shown). The following experiments were performed to compare the binding properties of [125I]-VIP and [125I]-Ro 25 1553 on the VPAC1 receptor: 1. [125I]-Ro 25 1553 and [125I]-VIP labelled an apparently homogeneous receptor population (Fig. 2). Analysis of the saturation curves suggested that [125I]-Ro 25 1553 recognised a lower number of sites than [125I]-VIP. Both tracers had the same high-affinity for their respective binding sites. 2. [125I]-Ro 25 1553 binding was almost completely inhibited in the presence of guanyl nucleotides GTP, GDP, and GTPgS (IC50 values of 80, 100 and 10 nM, respectively). [125I]-VIP binding was only partially inhibited (at most 40% inhibition) at somewhat higher nucleotide concentrations (IC50 values of 200, 300 and 20 nM for GTP, GDP and GTPgS, respectively) (Fig. 3).
(n)
[125I]-Ro 25 1553 IC50 value (nM)
[125I]-VIP IC50 value (nM)
(7) (2) (2) (2) (2) (2) (2) (2) (2) (2) (5)
0.17 2.5 5 12 50 47 0.5 0.5 1.14 0.3 0.12
1.1 60 200 110 450 200 7 2.5 35.8 1.5 1.1
(2)
5
20
(3)
3
100
3. The unlabelled peptides competition curves (using both tracers) were analysed by non-linear regression (Table 1 and Fig. 4). All the molecules tested—agonists and antagonists—had significantly (3- to 50-fold) higher affinities for [125I]-Ro 25 1553 than for [125I]-VIP labelled sites. Like [125I]-VIP, [125I]-Ro 25 1553 labelled sites with similar affinities for VIP and for VPAC1 agonist and a 20- to 100-fold lower affinity for unlabelled Ro 25 1553 (Fig. 4). These binding profiles confirmed that both tracers labelled exclusively VPAC1 receptors in these cells. Thus, [125I]-Ro 25 1553 labelled a fraction of the VPAC1 receptors, characterised by their higher affinity for the agonists tested and by their ability to convert to a low-affinity state in the presence of guanyl nucleotides. 4. Assuming that [125I]-VIP labelled all VPAC1 receptors in the CHO cells, these results suggested that the [125I]-VIP labelled binding sites were heterogeneous. We therefore re-analysed the [125I]-VIP competition curves according to a two sites binding model. Fitting of each individual competition curves was not significantly improved under the assumption that two receptor populations coexisted in the CHO cell membranes (not shown), but the highand the low-affinity IC50 values obtained in the absence and presence of GTP were very similar. We therefore de-
TABLE 2. pIC50 (2LogIC50) values of the unlabeled peptides, interacting with [125I]-Ro 25 1553
binding sites or with the high- and low-affinity [125I]-VIP binding sites
VIP Ro 25 1553 VPAC1 agonist (Control)a (1GTP)a a
[125I]-Ro 25 1553: high affinity pIC50
[125I]-VIP: high affinity pIC50
9.80 6 0.06 8.40 6 0.13 9.85 6 0.05 85-100%a 12 6 8%a
9.10 6 0.08 7.48 6 0.05 9.31 6 0.06 67 6 15%a 33 6 5%a
[125I]-VIP: low affinity pIC50 7.81 6.08 7.63 33 34
6 6 6 6 6
0.23 0.16 0.08 15%a 8%a
Tracer binding to this receptor state, expressed as a percentage of the total specific tracer binding measured in the absence of GTP and agonist.
694
R. Busto et al.
FIGURE 4. Selective agonists competition curves. VIP (closed cir-
cles), Ro 25 1553 (closed squares) and VPAC1 agonist [K15,R16]VIP(1-7)GRF(8-27), closed triangles) competition curves were obtained using [125I]-VIP (top panel) or [125I]-Ro 25 1553 (bottom panel) as tracers. Average of 3–7 competition curves in duplicate.
cided to analyse simultaneously the [125I]-VIP competition curves obtained for each agonist in the absence and presence of GTP (in parallel experiments), under the assumption that GTP affected the total [125I]-VIP binding to the high- and low-affinity sites, but did not affect the two corresponding agonist IC50 values. Thanks to the larger data size, the pooled competition curves were significantly better fitted under the assumption that two (not one) receptor states coexisted in the CHO cell membranes (Table 2 and Fig. 5). The inhibitory effect of GTP on specific [125I]-VIP binding could be explained exclusively by its effect on tracer binding to the high-affinity sites: Total tracer binding to the low-affinity sites was not significantly affected by GTP addition (Table 2, bottom). [125I]-Ro 25 1553 binding in the presence of GTP was too low to allow competition curves analysis. The agonist IC50 values for high-affinity [125I]-VIP bind-
FIGURE 5. Comparison of “one site” and “two sites” models.
[125I]-VIP competition curves were obtained in the absence and in the presence of 100 mM GTP (Average of 3 competition curves in duplicate.) The dashed lines represent the best fit competition curves obtained under the assumption that [125I]-VIP labelled a single binding site, with a different affinity in the absence and presence of GTP. The full lines represent the best fit competition curves obtained under the assumption that [125I]VIP labelled two binding sites populations, with different affinities for agonists, and that GTP affected the high and low-affinity binding sites proportions, but not their affinities for the agonists.
ing inhibition were higher than the [125I]-Ro 25 1553 IC50 values. Our results suggested that unlabelled as well as labelled VIP had very high affinities (<0.1–0.2 nM) for the high-affinity sites. At the 0.1 to 0.3 nM concentration, which we routinely use for [125I]-VIP competi-
Multiple VPAC1 Receptor States
695
FIGURE 6. Predicted saturation curves. Theoretical saturation
curves were calculated under the assumption that two different tracers were used at concentrations between 0.1 and 3 nM. Tracer 1 (closed circles) was assumed to recognise 10% of the receptors with a high-affinity (0.2 nM) and the remaining 90% with a low-affinity (10 nM). Tracer 2 (closed squares) was assumed to recognise the same binding sites, with KD values of 3 and 800 nM, respectively.
tion curves, this tracer occupied a significant proportion of the high (but not low) affinity receptors: The [125I]VIP competition curves were more shallow and shifted to lower concentrations when the tracer concentration was decreased (not shown). 5. If radioiodination of VIP and Ro 25 1553 did not affect their binding properties, both tracers must label preferentially the binding sites for which they have the highest affinity. The competition curves suggested that [125I]-VIP and [125I]-Ro 25 1553 labelled <10% of the VPAC1 receptors with KD values of <0.2 and 3 nM, respectively, and had KD values of <10 and 800 nM, respectively, for the remaining 90% of the VPAC1 receptors. In order to verify that, despite this high selectivity, the existence of two receptor states might be overlooked in saturation studies [10], we calculated the theoretical “expected” Scatchard plots. The predicted [125I]-VIP Scatchard plot was, as expected, curvilinear (Fig. 6). The curvature, however, was not statistically significant and would easily be overlooked in actual binding studies due to experimental errors. The [125I]-Ro 25 1553 Scatchard plot was linear, since this tracer is not expected to label low-affinity sites at the concentrations used. Both predicted KD values were in reasonable agreement with experimental data. The predicted BMax values depended on the tracer chosen, and were always significantly lower than the real (total) receptor density. Taken together, our results suggested that radioiodination did not markedly affect the binding properties of either VIP or Ro 25 1553; that both agonists recognised preferentially a small fraction of the VPAC1 receptors expressed in CHO cells; and that saturation curves using [125I]-VIP or [125I]-Ro 25 1553 markedly underestimated the receptor density.
FIGURE 7. Effect of Cholera Toxin on the guanyl nucleotide
competition curves. Specific [125I]-Ro 25 1553 binding was evaluated in the absence and presence of the indicated GTP (top panel), GDP (centre panel) or GTPgS (bottom panel) concentrations, using membranes from control cells (closed squares), or from cells pre-treated 4h with cholera toxin (open squares). Average of 3 competition curves in duplicate.
6. It is usually assumed that agonists stabilise ternary complexes involving the G protein responsible for second messenger activation, and that GDP and GTP destabilise the ternary complex and thereby inhibit agonists binding. In order to test the hypothesis that [125I]-Ro 25 1553 labelled selectively VPAC1 receptor-GS complexes, we investigated the effect of a short-term cholera toxin
696
R. Busto et al. TABLE 3. Comparison of the agonists binding and functional properties
VIP Ro 25 1553 VPAC1 agonist
High affinity IC50
Low affinity IC50
EC50
0.16 nM 4.0 nM 0.14 nM
15 nM 830 nM 23 nM
2.0 nM 100 nM 12 nM
treatment on this tracer’s binding properties. As shown in Figure 7, in membranes from cholera toxin treated cells, the guanyl-nucleotide resistant [125I]-Ro 25 1553 binding increased significantly from 11 6 4 to 25 6 3%, and the GTP and GTPgS concentrations necessary for half maximal effect (EC50) increased significantly from 30 6 5 to 100 6 20 mM (GTP), and also from 5 6 1 to 40 6 5 mM (GTPgS), respectively. These results supported the hypothesis that [125I]-Ro 25 1553 labelled selectively Gs-coupled receptors, affected by cholera toxin. 7. The Kact values of adenylate cyclase activation (EC50 in Table 3) were measured in CHO cell membranes expressing the rat VPAC1 receptor at low densities in order to avoid spare receptors. They did not agree with the [125I]-Ro 25 1553 binding data: They were closer to the “average” affinities, the agonist concentrations necessary to inhibit [125I]-VIP binding by 50% (IC50 in Table 3). This is important from a theoretical point of view: If G protein activation is proportional to the density of ternary complex, the agonists Kact must be equivalent to their affinity for the ternary complex [11]), that is, to their [125I]-Ro 25 1553 IC50 values (Table 3). In contrast, if agonists activate uncoupled receptors (by facilitating G protein recognition and GDP release), their EC50 values should reflect the occupation of the “low-affinity” uncoupled receptor site [12]. The agonists Kact value can be used as a diagnostic tool, to evaluate the extent of receptor precoupling [12]. From the data presented in Table 3, comparing the EC50 values with the IC50 values on both receptor states, it is clear that the agonist EC50 values were intermediate between the high- and low-affinity KD values. This result suggested that both receptor states are important for adenylate cyclase activation: VIP not only activates low-affinity uncoupled receptors (by facilitating the GDP release and the formation of a ternary complex involving the agonist, receptor, and empty G protein), but it also activates the ternary complex, probably by accelerating GTP binding to the receptor bound, empty G protein. CONCLUSION [125I]-Ro 25 1553 labelled selectively Gs-coupled VPAC1 receptors, while [125I]-VIP recognised high-affinity (Gs cou-
pled) and low-affinity (uncoupled) VPAC1 receptors. Both receptor states are involved in adenylate cyclase activation in response to agonists. These findings were supported by grants n8 3.4513.95 from F.R.S.M.; by an “Action de Recherche Concerte´e” from the Communaute´ Francaic¸e de Belgique; and by a “Interuniversity Poles of Attraction Programme—Belgian State, Minister’s Office—Federal Office for Scientific, Technical and Cultural Affairs”. R. Busto was supported by a Fellowship from the “Comunidad de Madrid” (Spain), M. G. Juarranz is a post-doctoral fellow from Marie Curie Research Training Grant (European Comission).
References 1. Harmar A.J., Arimura A., Gozes I., Journot L., Laburthe M., Pisegna J.R., Rawlings S.R., Robberecht P., Said S.I., Sreedharan S.P., Wank S.A. and Waschek J.A. (1998) Pharmacol. Rev. 50, 265–270. 2. Gourlet P., Vertongen P., Vandermeers A., VandermeersPiret M.-C., Rathe´ J., De Neef P. and Robberecht P. (1997) The long-acting vasoactive intestinal polypeptide agonist RO 25 1553 is highly selective of the VIP2 receptor subclass. Peptides 18, 403–408. 3. Gourlet P., Vandermeers A., Vertongen P., Rathe´ J., De Neef P., Cnudde J., Waelbroeck M. and Robberecht P. (1997) Development of high-affinity selective VIP1 receptor agonists. Peptides 18, 1539–1545. 4. Gourlet P., De Neef P., Cnudde J., Waelbroeck M. and Robberecht P. (1997) In Vitro properties of a high-affinity selective antagonist of the VIP1 receptor. Peptides 18, 1555–1560. 5. Vertongen P., Schiffman S.N., Gourlet P. and Robberecht P. (1997) Peptides 18, 1547–1554. 6. Simonneaux V., Kienlen-Campard P., Loeffler J.P., Basille M., Gonzalez B.J., Vaudry H., Robberecht P. and Pe´vet P. (1998) Pharmacological, molecular and functional characterization of vasoactive intestinal polypeptide/pituitary adenylate cyclase-activating polypeptide receptors in the rat pineal gland. Neuroscience 85, 887–896. 7. Robberecht P., De Neef P. and Lefebvre R.A. Influence of selective VIP receptor agonists in the rat gastric fundus. (1998) Eur. J. Pharmacol. 359, 77–80. 8. Ciccarelli E., Vilardaga J. P., De Neef P., Paolo E., Waelbroek M., Bollen A. and Robberecht P. (1994) Regul. Pept. 54, 397– 407. 9. Salomon Y., Londos C. and Rodbell M. (1974) Anal. Biochem. 58, 541–548. 10. Swillens S., Waelbroeck M. and Champeil, P. (1995) Trends in Pharmacol. Sci. 16, 151–155. 11. Colquhoun, D. (1998) Br. J. Pharmacol. 125, 924–947. 12. Waelbroeck M., Boufrahi L. and Swillens S. (1997) J. Theor. Biol. 187, 15–37.
ISSN 0898-6568/99 $ – see front matter PII S0898-6568(99)00041-8
Evidence for Multiple Rat VPAC1 Receptor States with Different Affinities for Agonists Rebeca Busto,† Ma-Guillerma Juarranz,‡ Salvatore De Maria,§ Patrick Robberecht‡* and Magali Waelbroeck‡ †On leave from the Departmento de Bioquı´mica y Biologı´a Molecular, Universidad de Alcala`, Madrid, Spain; ‡Department of Biochemistry and Nutrition, Faculty of Medicine, Universite´ Libre de Bruxelles, CP 611, 808 Route de Lennik, B-1070 Brussels, Belgium; and §on leave from the Centro di Recerca Interdipartimentale di Scienze Computazionali e Biotecnologice (CRISCEB), Napoli, Italia
ABSTRACT. We compare the binding properties of [125I-VIP] and [125I]-Ro 25 1553 to VPAC1 receptors, expressed in stably transfected CHO cells. [125I]-VIP labelled two VPAC1 receptor states, while [125I]-Ro 25 1553 labelled selectively a limited number of high-affinity receptors. This high-affinity state probably corresponds to an agonist-receptor-Gs ternary complex as its properties (guanyl nucleotides, EC50 values and maximal effect) were affected by cholera toxin pre-treatment. Both high- and low-affinity receptors participated in the adenylate cyclase activation. This suggested that agonists activate not only low-affinity uncoupled receptors by facilitating the ternary complex formation, but also activated the high-affinity ternary complex by accelerating the GTP binding to emptied, receptor-bound G proteins. cell signal 11;9:691–696, 1999. 1999 Elsevier Science Inc. KEY WORDS. Vasoactive intestinal peptide, Ro 25 1553, VPAC1 receptor, Agonist, Antagonist, G protein, Ternary complex
INTRODUCTION Two receptors for Vasoactive Intestinal Peptide (VIP) have been cloned in several mammalian species, and are now known as VPAC1 and VPAC2 receptors [1]. These two receptors have similar high affinities for their natural ligands, VIP, PACAP-27, and PACAP-38 (Pituitary Adenylate Cyclase Activating Peptide, 1-27 or 1-38) [1]. We previously demonstrated that the cyclic peptide, Ro 25 1553, is highly selective for VPAC2 receptors [2], and developed selective VPAC1 receptor agonists and antagonists [3, 4]. These peptides can be used, in combination with molecular biology techniques, to identify the VIP receptor subtypes that are expressed in mammalian tissues [5] or cells lines [6] and are responsible for functional responses [7]. We previously observed that although [125I]-Ro 25 1553 is clearly VPAC2selective [2], it did label some receptors in stably transfected Chinese Hamster Ovary (CHO) cells expressing VPAC1 receptors [5]. We compared in this work the properties of the VPAC1 binding sites labelled by [125I]-Ro 25 1553 and [125I]VIP in stably transfected CHO cells. MATERIALS AND METHODS Peptide Synthesis All the peptides were synthesised by solid phase methodology using the Fmoc strategy [3], and purified by HPLC re*Author to whom all correspondence should be addressed. Tel: 132-2555-6229; fax: 132-2-555-6230; e-mail: [email protected]
verse phase and by ion exchange chromatography [2–4]. The peptide purity (.97%) was assessed by capillary electrophoresis and the conformity was verified by electrospray mass spectrometry and sequencing [2–4]. Cell Cultures, Cholera Toxin Treatment, and Membrane Preparations Transfected Chinese Hamster Ovary (CHO) cells, stably expressing the recombinant rat VPAC1 receptor (810 6 80 pmoles/mg of membrane protein), were cultured as previously described [8]. Some cells (as indicated) were pretreated 4h with 10 mg/ml cholera toxin and washed twice with fresh medium immediately before harvesting. For membrane preparations, confluent cells were harvested with a rubber policeman and pelleted by low speed centrifugation; the supernatant was discarded and the cells lysed in 1 mM NaHCO3 solution and immediately frozen in liquid nitrogen. After thawing, the lysate was first centrifuged at 800 3 g for 10 min. The supernatant was further centrifuged at 20,000 3 g for 10 min. The pellet, resuspended in 1 mM NaHCO3, was used immediately as a crude membrane preparation. Receptor Identification VIP and Ro 25 1553 were radioiodinated as previously described [5] by the iodogen method, purified by absorption on a Sep-Pak cartridge, and eluted with 50% acetonitrile in
692
R. Busto et al.
FIGURE 1. VIP and Ro 25 1553 competition curves at VPAC1
receptors. Specific [125I]-VIP binding to stably transfected CHO cells expressing VPAC1 receptors was measured in the absence or presence of the indicated VIP (closed circles) or Ro 25 1553 (closed squares) concentrations. Average of 3–7 competition curves in duplicate.
0.1% trifluoroacetic acid. Binding was performed at 37 8C in 20 mM Tris-maleate buffer (pH 7.4) enriched with 2 mM MgCl2, 0.1 mg/ml bacitracin, and 1% bovine serum albumin. Non-specific binding was defined as tracer binding in the presence of 1 mM VIP. The incubations were terminated by filtration through glass-fibre GF/C filters presoaked for 24 h in 0.1% polyethyleneimine, and the filters were rinsed three times with a 20 mM (pH 7.4) sodium phosphate buffer containing 1% bovine serum albumin [8]. Adenylate Cyclase Activity Adenylate cyclase activity was determined by the Salomon et al. procedure [9]. Membranes (3–15 mg protein) were incubated in a total volume of 60 ml containing 0.5 mM [a32 P]-ATP, 10 mM GTP, 5 mM MgCl2, 0.5 mM EDTA, 1 FIGURE 3. Guanyl nucleotides competition curves. Specific
[125I]-Ro 25 1553 (closed squares) and [125I]-VIP (closed circles) binding was measured in the absence or presence of the indicated concentrations of GTP (top panel), GDP (centre panel) or GTPgS (bottom panel). Average of 3 competition curves in duplicate.
mM cAMP, 1 mM theophylline, 10 mM phospho(enol) pyruvate, 3 mg/ml pyruvate kinase, and 30 mM Tris-HCl at a final pH of 7.5. The reaction was terminated after a 15 min incubation at 378C by addition of 0.5 ml of a 0.5% sodium dodecyl-sulfate solution containing 0.5 mM ATP, 0.5 mM cAMP, and 20,000 cpm [3H]-cAMP. cAMP was separated from ATP by two successive chromatographies on Dowex 50WX8 and neutral alumina. FIGURE 2. [125I]-VIP and [125I]-Ro 25 1553 saturation curves.
[125I]-VIP (closed circles) and [125I]-Ro 25 1553 (closed squares) binding to stably transfected CHO cells was measured at increasing tracer concentrations (0.1 to 3 nM), and the results presented according to Scatchard. Average of 4 saturation curves in duplicate.
RESULTS AND DISCUSSION Despite the low-affinity of the cyclic VIP analogue Ro 25 1553 for the rat VPAC1 receptor, as judged by its low capac-
Multiple VPAC1 Receptor States
693
TABLE 1. IC50 values of [125I]-Ro 25 1553 and [125I]-VIP binding for different peptides on
VPAC1 receptors expressed in CHO cells
VIP Helodermin (1-37) Secretin VIP (2-28) VIP (3-28) VIP (6-28) [D-Phe4]-VIP [D-Ala4]-VIP PACAP (1-23) [R16]-PACAP (1-23) VPAC1 agonist ([K15, R16]-VIP(1-7)GRF(8-27)) VPAC1 antagonist ([D-Phe2, K15, R16]-VIP(1-7)GRF(8-27)) Ro 25 1553
ity to inhibit 125I-VIP binding (Fig. 1) and by its low Kact value for adenylate cyclase activation (not shown), its iodinated derivative [125I]-Ro 25 1553 labelled specifically membranes prepared from CHO cells expressing the recombinant VPAC1 rat receptor. There was no detectable [125IVIP and [125I]-Ro 25 1553 binding to non transfected CHO cells or to CHO cells transfected with a non-related receptor (data not shown). The following experiments were performed to compare the binding properties of [125I]-VIP and [125I]-Ro 25 1553 on the VPAC1 receptor: 1. [125I]-Ro 25 1553 and [125I]-VIP labelled an apparently homogeneous receptor population (Fig. 2). Analysis of the saturation curves suggested that [125I]-Ro 25 1553 recognised a lower number of sites than [125I]-VIP. Both tracers had the same high-affinity for their respective binding sites. 2. [125I]-Ro 25 1553 binding was almost completely inhibited in the presence of guanyl nucleotides GTP, GDP, and GTPgS (IC50 values of 80, 100 and 10 nM, respectively). [125I]-VIP binding was only partially inhibited (at most 40% inhibition) at somewhat higher nucleotide concentrations (IC50 values of 200, 300 and 20 nM for GTP, GDP and GTPgS, respectively) (Fig. 3).
(n)
[125I]-Ro 25 1553 IC50 value (nM)
[125I]-VIP IC50 value (nM)
(7) (2) (2) (2) (2) (2) (2) (2) (2) (2) (5)
0.17 2.5 5 12 50 47 0.5 0.5 1.14 0.3 0.12
1.1 60 200 110 450 200 7 2.5 35.8 1.5 1.1
(2)
5
20
(3)
3
100
3. The unlabelled peptides competition curves (using both tracers) were analysed by non-linear regression (Table 1 and Fig. 4). All the molecules tested—agonists and antagonists—had significantly (3- to 50-fold) higher affinities for [125I]-Ro 25 1553 than for [125I]-VIP labelled sites. Like [125I]-VIP, [125I]-Ro 25 1553 labelled sites with similar affinities for VIP and for VPAC1 agonist and a 20- to 100-fold lower affinity for unlabelled Ro 25 1553 (Fig. 4). These binding profiles confirmed that both tracers labelled exclusively VPAC1 receptors in these cells. Thus, [125I]-Ro 25 1553 labelled a fraction of the VPAC1 receptors, characterised by their higher affinity for the agonists tested and by their ability to convert to a low-affinity state in the presence of guanyl nucleotides. 4. Assuming that [125I]-VIP labelled all VPAC1 receptors in the CHO cells, these results suggested that the [125I]-VIP labelled binding sites were heterogeneous. We therefore re-analysed the [125I]-VIP competition curves according to a two sites binding model. Fitting of each individual competition curves was not significantly improved under the assumption that two receptor populations coexisted in the CHO cell membranes (not shown), but the highand the low-affinity IC50 values obtained in the absence and presence of GTP were very similar. We therefore de-
TABLE 2. pIC50 (2LogIC50) values of the unlabeled peptides, interacting with [125I]-Ro 25 1553
binding sites or with the high- and low-affinity [125I]-VIP binding sites
VIP Ro 25 1553 VPAC1 agonist (Control)a (1GTP)a a
[125I]-Ro 25 1553: high affinity pIC50
[125I]-VIP: high affinity pIC50
9.80 6 0.06 8.40 6 0.13 9.85 6 0.05 85-100%a 12 6 8%a
9.10 6 0.08 7.48 6 0.05 9.31 6 0.06 67 6 15%a 33 6 5%a
[125I]-VIP: low affinity pIC50 7.81 6.08 7.63 33 34
6 6 6 6 6
0.23 0.16 0.08 15%a 8%a
Tracer binding to this receptor state, expressed as a percentage of the total specific tracer binding measured in the absence of GTP and agonist.
694
R. Busto et al.
FIGURE 4. Selective agonists competition curves. VIP (closed cir-
cles), Ro 25 1553 (closed squares) and VPAC1 agonist [K15,R16]VIP(1-7)GRF(8-27), closed triangles) competition curves were obtained using [125I]-VIP (top panel) or [125I]-Ro 25 1553 (bottom panel) as tracers. Average of 3–7 competition curves in duplicate.
cided to analyse simultaneously the [125I]-VIP competition curves obtained for each agonist in the absence and presence of GTP (in parallel experiments), under the assumption that GTP affected the total [125I]-VIP binding to the high- and low-affinity sites, but did not affect the two corresponding agonist IC50 values. Thanks to the larger data size, the pooled competition curves were significantly better fitted under the assumption that two (not one) receptor states coexisted in the CHO cell membranes (Table 2 and Fig. 5). The inhibitory effect of GTP on specific [125I]-VIP binding could be explained exclusively by its effect on tracer binding to the high-affinity sites: Total tracer binding to the low-affinity sites was not significantly affected by GTP addition (Table 2, bottom). [125I]-Ro 25 1553 binding in the presence of GTP was too low to allow competition curves analysis. The agonist IC50 values for high-affinity [125I]-VIP bind-
FIGURE 5. Comparison of “one site” and “two sites” models.
[125I]-VIP competition curves were obtained in the absence and in the presence of 100 mM GTP (Average of 3 competition curves in duplicate.) The dashed lines represent the best fit competition curves obtained under the assumption that [125I]-VIP labelled a single binding site, with a different affinity in the absence and presence of GTP. The full lines represent the best fit competition curves obtained under the assumption that [125I]VIP labelled two binding sites populations, with different affinities for agonists, and that GTP affected the high and low-affinity binding sites proportions, but not their affinities for the agonists.
ing inhibition were higher than the [125I]-Ro 25 1553 IC50 values. Our results suggested that unlabelled as well as labelled VIP had very high affinities (<0.1–0.2 nM) for the high-affinity sites. At the 0.1 to 0.3 nM concentration, which we routinely use for [125I]-VIP competi-
Multiple VPAC1 Receptor States
695
FIGURE 6. Predicted saturation curves. Theoretical saturation
curves were calculated under the assumption that two different tracers were used at concentrations between 0.1 and 3 nM. Tracer 1 (closed circles) was assumed to recognise 10% of the receptors with a high-affinity (0.2 nM) and the remaining 90% with a low-affinity (10 nM). Tracer 2 (closed squares) was assumed to recognise the same binding sites, with KD values of 3 and 800 nM, respectively.
tion curves, this tracer occupied a significant proportion of the high (but not low) affinity receptors: The [125I]VIP competition curves were more shallow and shifted to lower concentrations when the tracer concentration was decreased (not shown). 5. If radioiodination of VIP and Ro 25 1553 did not affect their binding properties, both tracers must label preferentially the binding sites for which they have the highest affinity. The competition curves suggested that [125I]-VIP and [125I]-Ro 25 1553 labelled <10% of the VPAC1 receptors with KD values of <0.2 and 3 nM, respectively, and had KD values of <10 and 800 nM, respectively, for the remaining 90% of the VPAC1 receptors. In order to verify that, despite this high selectivity, the existence of two receptor states might be overlooked in saturation studies [10], we calculated the theoretical “expected” Scatchard plots. The predicted [125I]-VIP Scatchard plot was, as expected, curvilinear (Fig. 6). The curvature, however, was not statistically significant and would easily be overlooked in actual binding studies due to experimental errors. The [125I]-Ro 25 1553 Scatchard plot was linear, since this tracer is not expected to label low-affinity sites at the concentrations used. Both predicted KD values were in reasonable agreement with experimental data. The predicted BMax values depended on the tracer chosen, and were always significantly lower than the real (total) receptor density. Taken together, our results suggested that radioiodination did not markedly affect the binding properties of either VIP or Ro 25 1553; that both agonists recognised preferentially a small fraction of the VPAC1 receptors expressed in CHO cells; and that saturation curves using [125I]-VIP or [125I]-Ro 25 1553 markedly underestimated the receptor density.
FIGURE 7. Effect of Cholera Toxin on the guanyl nucleotide
competition curves. Specific [125I]-Ro 25 1553 binding was evaluated in the absence and presence of the indicated GTP (top panel), GDP (centre panel) or GTPgS (bottom panel) concentrations, using membranes from control cells (closed squares), or from cells pre-treated 4h with cholera toxin (open squares). Average of 3 competition curves in duplicate.
6. It is usually assumed that agonists stabilise ternary complexes involving the G protein responsible for second messenger activation, and that GDP and GTP destabilise the ternary complex and thereby inhibit agonists binding. In order to test the hypothesis that [125I]-Ro 25 1553 labelled selectively VPAC1 receptor-GS complexes, we investigated the effect of a short-term cholera toxin
696
R. Busto et al. TABLE 3. Comparison of the agonists binding and functional properties
VIP Ro 25 1553 VPAC1 agonist
High affinity IC50
Low affinity IC50
EC50
0.16 nM 4.0 nM 0.14 nM
15 nM 830 nM 23 nM
2.0 nM 100 nM 12 nM
treatment on this tracer’s binding properties. As shown in Figure 7, in membranes from cholera toxin treated cells, the guanyl-nucleotide resistant [125I]-Ro 25 1553 binding increased significantly from 11 6 4 to 25 6 3%, and the GTP and GTPgS concentrations necessary for half maximal effect (EC50) increased significantly from 30 6 5 to 100 6 20 mM (GTP), and also from 5 6 1 to 40 6 5 mM (GTPgS), respectively. These results supported the hypothesis that [125I]-Ro 25 1553 labelled selectively Gs-coupled receptors, affected by cholera toxin. 7. The Kact values of adenylate cyclase activation (EC50 in Table 3) were measured in CHO cell membranes expressing the rat VPAC1 receptor at low densities in order to avoid spare receptors. They did not agree with the [125I]-Ro 25 1553 binding data: They were closer to the “average” affinities, the agonist concentrations necessary to inhibit [125I]-VIP binding by 50% (IC50 in Table 3). This is important from a theoretical point of view: If G protein activation is proportional to the density of ternary complex, the agonists Kact must be equivalent to their affinity for the ternary complex [11]), that is, to their [125I]-Ro 25 1553 IC50 values (Table 3). In contrast, if agonists activate uncoupled receptors (by facilitating G protein recognition and GDP release), their EC50 values should reflect the occupation of the “low-affinity” uncoupled receptor site [12]. The agonists Kact value can be used as a diagnostic tool, to evaluate the extent of receptor precoupling [12]. From the data presented in Table 3, comparing the EC50 values with the IC50 values on both receptor states, it is clear that the agonist EC50 values were intermediate between the high- and low-affinity KD values. This result suggested that both receptor states are important for adenylate cyclase activation: VIP not only activates low-affinity uncoupled receptors (by facilitating the GDP release and the formation of a ternary complex involving the agonist, receptor, and empty G protein), but it also activates the ternary complex, probably by accelerating GTP binding to the receptor bound, empty G protein. CONCLUSION [125I]-Ro 25 1553 labelled selectively Gs-coupled VPAC1 receptors, while [125I]-VIP recognised high-affinity (Gs cou-
pled) and low-affinity (uncoupled) VPAC1 receptors. Both receptor states are involved in adenylate cyclase activation in response to agonists. These findings were supported by grants n8 3.4513.95 from F.R.S.M.; by an “Action de Recherche Concerte´e” from the Communaute´ Francaic¸e de Belgique; and by a “Interuniversity Poles of Attraction Programme—Belgian State, Minister’s Office—Federal Office for Scientific, Technical and Cultural Affairs”. R. Busto was supported by a Fellowship from the “Comunidad de Madrid” (Spain), M. G. Juarranz is a post-doctoral fellow from Marie Curie Research Training Grant (European Comission).
References 1. Harmar A.J., Arimura A., Gozes I., Journot L., Laburthe M., Pisegna J.R., Rawlings S.R., Robberecht P., Said S.I., Sreedharan S.P., Wank S.A. and Waschek J.A. (1998) Pharmacol. Rev. 50, 265–270. 2. Gourlet P., Vertongen P., Vandermeers A., VandermeersPiret M.-C., Rathe´ J., De Neef P. and Robberecht P. (1997) The long-acting vasoactive intestinal polypeptide agonist RO 25 1553 is highly selective of the VIP2 receptor subclass. Peptides 18, 403–408. 3. Gourlet P., Vandermeers A., Vertongen P., Rathe´ J., De Neef P., Cnudde J., Waelbroeck M. and Robberecht P. (1997) Development of high-affinity selective VIP1 receptor agonists. Peptides 18, 1539–1545. 4. Gourlet P., De Neef P., Cnudde J., Waelbroeck M. and Robberecht P. (1997) In Vitro properties of a high-affinity selective antagonist of the VIP1 receptor. Peptides 18, 1555–1560. 5. Vertongen P., Schiffman S.N., Gourlet P. and Robberecht P. (1997) Peptides 18, 1547–1554. 6. Simonneaux V., Kienlen-Campard P., Loeffler J.P., Basille M., Gonzalez B.J., Vaudry H., Robberecht P. and Pe´vet P. (1998) Pharmacological, molecular and functional characterization of vasoactive intestinal polypeptide/pituitary adenylate cyclase-activating polypeptide receptors in the rat pineal gland. Neuroscience 85, 887–896. 7. Robberecht P., De Neef P. and Lefebvre R.A. Influence of selective VIP receptor agonists in the rat gastric fundus. (1998) Eur. J. Pharmacol. 359, 77–80. 8. Ciccarelli E., Vilardaga J. P., De Neef P., Paolo E., Waelbroek M., Bollen A. and Robberecht P. (1994) Regul. Pept. 54, 397– 407. 9. Salomon Y., Londos C. and Rodbell M. (1974) Anal. Biochem. 58, 541–548. 10. Swillens S., Waelbroeck M. and Champeil, P. (1995) Trends in Pharmacol. Sci. 16, 151–155. 11. Colquhoun, D. (1998) Br. J. Pharmacol. 125, 924–947. 12. Waelbroeck M., Boufrahi L. and Swillens S. (1997) J. Theor. Biol. 187, 15–37.