Báo cáo khoa học: An arginyl in the N-terminus of the V1a vasopressin receptor is part of the conformational switch controlling activation by agonist

doi:10.1046/j.1432-1033.2003.03865.x

Eur. J. Biochem. 270, 4681–4688 (2003) (cid:2) FEBS 2003

An arginyl in the N-terminus of the V1a vasopressin receptor is part of the conformational switch controlling activation by agonist

Stuart R. Hawtin1,*, Victoria J. Wesley1, John Simms1, Rosemary A. Parslow1, Alice Miles1, Kim McEwan1, Mary Keen2 and Mark Wheatley1 1School of Biosciences and 2Department of Pharmacology, Division of Neuroscience, The Medical School, University of Birmingham, Edgbaston, Birmingham, UK

wildtype binding for both peptide and nonpeptide antago- nists. The ratio of Ki to EC50, an indicator of efficacy, was increased for all substitutions. Consequently, although [R46X]V1aR constructs have a lower affinity for agonist, once AVP has bound all 19 are more likely than the wildtype V1aR to become activated. Therefore, in the wildtype V1aR, Arg46 constrains the inactive conformation of the receptor. On binding AVP this constraint is alleviated, promoting the transition to active V1aR. Our findings explain why arginyl is conserved at this locus throughout the evolutionary lineage of the neurohypophysial peptide hormone receptor family of G-protein coupled receptors.

Keywords: GPCR; vasopressin; ligand binding; cell signa- ling; peptide hormone.

Defining how the agonist–receptor interaction differs from that of the antagonist–receptor and understanding the mechanisms of receptor activation are fundamental issues in cell signalling. The V1a vasopressin receptor (V1aR) is a member of a family of related G-protein coupled receptors that are activated by neurohypophysial peptide hormones, including vasopressin (AVP). It has recently been reported that an arginyl in the distal N-terminus of the V1aR is critical for binding agonists but not antagonists. To determine specific features required at this locus to support high affinity agonist binding and second messenger generation, Arg46 was substituted by all other 19 encoded amino acids. Our data establish that there is an absolute requirement for arginyl, as none of the [R46X]V1aR mutant constructs sup- ported high affinity agonist binding and all 19 had defective signalling. In contrast, all of the mutant receptors possessed

include the isotocin receptor, the mesotocin receptor, the oxytocin receptor (OTR), the vasotocin receptor and three subtypes of vasopressin receptor: V1a, V1b and V2 (V1aR, V1bR and V2R respectively). In addition to the characteristic architecture of GPCRs [2], the neurohypophysial peptide hormone receptors exhibit certain conserved sequence motifs and share related pharmacologies (reviewed in [3–6]). This has allowed these receptors to be classified as a subfamily of GPCRs.

The neurohypophysial hormones are nonapeptides, ami- dated at the C-terminus and with an intramolecular disulphide bond between positions one and six, which creates a 20-membered ring and a tripeptide tail. Peptides of the vasopressin (AVP)-like family possess a positively charged residue at position eight whereas those of the oxytocin-like family have a neutral residue. Lower verte- brates possess vasotocin (AVT; [Ile3]vasopressin) rather than AVP, whereas isotocin ([Ser4,Ile8]oxytocin) and meso- tocin ([Ile8]oxytocin) are the evolutionary precursors of oxytocin [1]. The receptors that mediate the effects of these hormones are G-protein coupled receptors (GPCRs) and

Correspondence to M. Wheatley, School of Biosciences, University of Birmingham, Edgbaston, Birmingham, B15 2TT, UK. Fax: + 44 121 414 5925, Tel.: + 44 121 414 3981, E-mail: m.wheatley@bham.ac.uk Abbreviations: AVP, [Arg8]vasopressin; AVT, [Arg8]vasotocin; GPCR, G-protein coupled receptor; InsP, inositol phosphate; InsP3, inositol trisphosphate; OTR, oxytocin receptor; [R46X]V1aR, Arg fi mutant vasopressin V1a receptor at position 46; V1aR, vasopressin V1a receptor; V1bR, vasopressin V1b receptor; V2R, vasopressin V2 receptor. *Present address: Institute of Cell Signalling, Medical School, Queen’s Medical Centre, Nottingham, UK. (Received 13 May 2003, revised 25 September 2003, accepted 3 October 2003)

The V1aR mediates a plethora of responses to AVP in addition to the well-characterized vasopressor effect [3]. Consequently, this receptor subtype is widely distributed and generates nearly all of the physiological actions of AVP with the notable exceptions of antidiuresis (V2R) and adrenocorticotropic hormone secretion (V1bR). This has been the stimulus for the development of a range of V1aR antagonists, initially peptides [7] and more recently non- peptides [8,9]. Although agonists and antagonists exhibit competitive binding to the V1aR, only agonists promote the active receptor conformation with subsequent second messenger generation. Understanding how the agonist– receptor interaction differs from that of the antagonist– receptor and defining the mechanisms of receptor activation are fundamental issues in cell signalling.

The N-terminus of the V1aR provides agonist-specific binding epitopes, as truncation of the distal segment of the V1aR N-terminus prevents high affinity AVP binding but does not affect antagonist binding [10]. A similar

4682 S. R. Hawtin et al. (Eur. J. Biochem. 270)

(cid:2) FEBS 2003

by automated fluorescent sequencing (Alta Bioscience, Birmingham, UK).

Cell culture and transfection

HEK 293T cells were routinely cultured in Dulbecco’s modified Eagles medium supplemented with 10% (v/v) fetal bovine serum, penicillin (100 IUÆmL)1) and strepto- mycin (100 lgÆmL)1) in humidified 5% (v/v) CO2 in air at 37 (cid:4)C. Cells were seeded at a density of approximately 5 · 105 cells per 100 mm dish and transfected after 48 h using a calcium phosphate precipitation protocol with 10 lg DNA per dish.

Radioligand binding assays

situation has been reported for other members of the neurohypophysial hormone receptor family; for example, the distal N-terminus is required for agonist binding to the OTR [11,12] and also to the vasotocin receptor [13], suggesting a common role for the N-terminal domain in agonist binding throughout this GPCR subclass. This is supported by the observation that disruption of AVP binding to a truncated V1aR could be functionally rescued by a chimeric construct in which the N-terminus of the OTR is replaced with the corresponding sequence of the V1aR [10]. The role of the N-terminus has been addressed recently by alanine-scanning mutagenesis of the N-terminus of the V1aR; this revealed that a single residue (Arg46) located in the distal segment of this domain is critical for high affinity agonist binding but not antag- onist binding [14]. The corresponding residue in the OTR is also an arginyl (Arg34) and furthermore this arginyl is required for high affinity agonist binding to the OTR [15]. Indeed, an arginyl is completely conserved at this locus in all members of the neurohypophysial peptide hormone receptor family cloned to date, suggesting that this residue fulfils an important common role required specifically for agonist binding throughout this subfamily of GPCRs.

The aim of this study was to understand the structural requirements at this key locus of the V1aR which endow high affinity agonist binding, by systematically mutating the critical arginyl to all the other 19 amino acids encoded in mRNA. In addition, this study identifies Arg46 as part of the conformational switch mechanism which controls conversion of inactive V1aR to active receptor in response to AVP.

A washed cell membrane preparation of HEK 293T cells, transfected with the appropriate receptor construct, was prepared as previously described [16] and the protein concentration determined using the BCA protein assay kit (Pierce Chemical Co., Tattenhall, Cheshire, UK) with bovine serum albumin as the standard. Radioligand binding assays were performed as described previously [17] using either the natural agonist [Phe3-3,4,5-3H]AVP (68.5 CiÆ mmol)1; DuPont NEN, Stevenage, Herts, UK) or the V1aR-selective peptide antagonist [Phe3-3,4,5-3H] d(CH2)5- Tyr(Me)2AVP (99 CiÆmmol)1; DuPont NEN) [18] as the tracer ligand. Binding data were analyzed by nonlinear regression to fit theoretical Langmuir binding isotherms to the experimental data using the FIG. P program (Biosoft, Cambridge, UK). Individual IC50 values obtained for competing ligands were corrected for radioligand occu- pancy as described [19] using the radioligand affinity (Ki) experimentally determined for each individual construct.

Experimental procedures

Materials

AVP was purchased from Sigma. The cyclic antagonist, 1-(b-mercapto-b,b-cyclopentamethylenepropionic acid), 2- (O-methyl)tyrosine AVP [d(CH2)5Tyr(Me)2AVP] and linear antagonist, phenylacetyl-D-Tyr(Me)2Arg6Tyr(NH2)9AVP were from Bachem (St Helens, UK). The nonpeptide antagonist (SR 49059) was provided by Sanofi Recherche (Toulouse, France). Cell culture media, buffers and supple- ments were purchased from Gibco (Uxbridge, UK). Restriction enzymes Eco81I, Pfl23II and SdaI were obtained from MBI fermentas (Sunderland, UK).

Mutant receptor constructs

eluted with 10 mL of

Mutation of Arg46 to each of the 19 encoded amino acids was made by a PCR approach. Mutant sense oligonucle- otides (5¢-GGGGGCCTTAGGGGACGTAXXXAATGA GGAGCTGG-3¢) contained the appropriate base change (shown as XXX) for each of the corresponding Arg46 fi Xaa46 substitutions, and a unique Eco81I restriction site (bold). The PCR cycling conditions were as follows: denaturing, 94 (cid:4)C (1 min); annealing, 60 (cid:4)C (2 min); exten- sion, 72 (cid:4)C (1 min) for 30 cycles followed by extension at 72 (cid:4)C (7 min). All mutant PCR products were subcloned into the receptor utilizing unique Eco81I and SdaI restriction sites. All receptor constructs were confirmed AVP-induced inositol phosphate production HEK 293T cells were seeded at a density of 2.5 · 105 cells per well in poly D-lysine coated 12 well plates and transfected after 24 h using TransfastTM (Promega). The assay for AVP-induced accumulation of inositol phosphates was based on that described previously [20]. Essentially, at 16 h post-transfection, medium was replaced with inositol-free Dulbecco’s modified Eagles medium containing 1% (v/v) fetal bovine serum and 2 lCiÆmL)1 myo-[2-3H]inositol (22.0 CiÆmmol)1; DuPont NEN) for 24 h. Cells were washed twice with NaCl/Pi and incubated in inositol-free medium containing 10 mM LiCl for 30 min, after which AVP was added at the concentrations indicated for a further 30 min. Incuba- tions were terminated by adding 0.5 mL of 5% (w/v) perchloric acid containing 1 mM EDTA and 1 mgÆmL)1 phytic acid hydrolysate (final concentrations). Samples were neutralized with 1.2 M KOH, 10 mM EDTA, 50 mM Hepes on ice for 1 h, insoluble material was sedimented at 12 000 g for 5 min and supernatants were loaded onto 0.8 mL AG1-X8 (formate form; Bio-Rad Laboratories, Hemel Hempstead, UK). A mixed inositol fraction containing mono-, bis- and tris-phosphates (InsP–InsP3) was 850 mM NH4COOH containing 0.1 M HCOOH as described [21] and quantified by scintillation counting.

(cid:2) FEBS 2003

The N-terminal switch of the V1aR (Eur. J. Biochem. 270) 4683

Results

The effect of systematic substitution of Arg46 on agonist binding

the 19 mutant receptor constructs decreasing between 700- fold and 3000-fold compared to the wildtype (Table 1). As an example, competition binding curves for one of the 19 engineered mutant receptors are presented in Fig. 2A, which shows AVP and the nonpeptide antagonist SR 49059 binding to wildtype V1aR and the mutant construct [R46A]V1aR. The decrease in agonist affinity observed when Arg46 was mutated was apparent in both the presence and absence of 5¢-guanylylimidodiphosphate (Fig. 2A). Consequently, the decrease in agonist affinity did not merely reflect uncoupling of the receptor–G-protein com- plex. No residue other than the wildtype arginine could support high affinity agonist binding. The inability to bind agonist was universal and was observed regardless of the physico-chemical characteristic of the residue side-chain, be it positively charged, negatively charged, polar, aromatic or aliphatic (Table 1). Addition of guanidinium ion (up to 10 mM) to the [R46A]V1aR construct did not rescue agonist binding (data not shown).

The effect of systematic substitution of Arg46 on intracellular signalling

Fig. 1. Schematic diagram of the V1aR. The V1aR is illustrated as seven a-helical transmembrane domains traversing the lipid bilayer. The enlarged circle shows the amino acid sequence of the distal N-terminus of the V1aR and indicates the position of Arg46 that was subjected to systematic investigation in this study. Glycosylation sites which have been shown to be modified by oligosaccharide [29] are indicated by branched structures.

The structural requirements of the residue at position 46 in the N-terminus of the V1aR (Fig. 1) were investigated, with respect to supporting high affinity agonist binding, by site- directed mutagenesis. The wildtype Arg46 was systematic- ally substituted by all the other 19 encoded amino acids. These receptor constructs were then characterized pharma- cologically by radioligand binding assay after expression in HEK 293T cells and compared to wildtype V1aR. The wildtype and mutant receptors were expressed at the same level of approximately 1–2 pmolÆmg)1 protein. Four differ- ent classes of ligand were available to probe the ligand binding site: (a) peptide agonist; (b) cyclic peptide antagonist possessing a disulphide bond, 20-membered ring and short peptide tail; (c) linear peptide antagonist and (d) nonpeptide antagonist. The binding affinities of the cyclic peptide antagonist [d(CH2)5Tyr(Me)2AVP] [18], the linear peptide antagonist [phenylacetyl-D-Tyr(Me)2Arg6Tyr(NH2)9AVP] [22] and the nonpeptide antagonist (SR 49059) [23] for all the 19 mutant receptors engineered, were comparable to wildtype V1aR (Table 1). This was important, as it allowed accurate quantification of the pharmacological characteris- tics of each mutant construct by radioligand binding studies using [3H]d(CH2)5Tyr(Me)2AVP as the tracer ligand. Furthermore, normal antagonist binding established that substitution of the arginine had not resulted in a distorted or misfolded receptor protein. In contrast to the three classes of antagonist, the binding of AVP was dramatically affected by the substitution of Arg46, with the affinity of AVP for all

The stimulation of accumulated InsPs by increasing concentrations of AVP, was assayed for each of the 19 mutant constructs and the dose–response characteristics compared to the wildtype V1aR. None of the 19 encoded amino acids could substitute the Arg46 and retain the wildtype V1aR second messenger generation. Dose– response curves for AVP-stimulated accumulation of InsP–InsP3 by the wildtype V1aR and a representative mutant construct ([R46A]V1aR) are presented in Fig. 2B. The degree of perturbation of second messenger generation was dictated by the identity of the residue present at position 46, with the EC50 for AVP-induced InsP–InsP3 accumula- tion increasing from between 10-fold to 600-fold compared to the wildtype V1aR (Fig. 3B). The acidic residues at position 46 in [R46E]V1aR and [R46D]V1aR were the most detrimental to signalling. This effect was due predominantly to the negative charge, as removal of this charge in the constructs [R46Q]V1aR and [R46N]V1aR increased the Emax and resulted in EC50 values that were 28 fold and 14 fold lower than their respective acids (Table 1). Histidine was the least disruptive substitution, as the EC50 of [R46H]V1aR was only 10 fold higher than wildtype V1aR, although the Emax was depressed. Lysine could not replace arginine because [R46K]V1aR was no better than [R46P]V1aR, [R46V]V1aR or [R46A]V1aR at signalling. [R46M]V1aR, [R46C]V1aR and the hydroxyl-containing [R46S]V1aR and [R46T]V1aR were all similarly impaired with EC50 values increased approximately 100 fold over the wildtype. The addition of hydroxyl to the aromatic ring of phenylalanine had a neutral effect as the sensitivity of [R46F]V1aR and [R46Y]V1aR to AVP was almost identical. The residues Trp and Gly represent the two extremes of side-chain size, however, both [R46W]V1aR and [R46G]V1aR exhibited a similar increase in the EC50 of AVP-induced InsP produc- tion of (cid:1) 150 fold. A similar perturbation was observed with the branched aliphatic substitutions in [R46L]V1aR and [R46I]V1aR. The wildtype V1aR and the 19 mutant V1aR constructs are arranged in rank order with respect to binding in Fig. 3A and with respect to second messenger

4684 S. R. Hawtin et al. (Eur. J. Biochem. 270)

(cid:2) FEBS 2003

Table 1. Pharmacological profile of [R46X]V1aRs. Mutant V1aRs were expressed in HEK 293T cells and characterized pharmacologically. Dis- sociation constants (Ki) were calculated from IC50 values after correcting for the radioligand occupancy as described in Experimental procedures. EC50 and Emax values relate to AVP-induced accumulation of InsP–InsP3 in cells expressing wildtype (WT) or mutant receptors. Values shown are the mean ± SEM of three separate experiments performed in triplicate. Data for R46A, R46K, R46L and R46E are taken from [14]. CA, cyclic peptide antagonist; LA, linear peptide antagonist; SR 49059, nonpeptide antagonist.

(cid:2) FEBS 2003

The N-terminal switch of the V1aR (Eur. J. Biochem. 270) 4685

Fig. 3. Effect of substitution of Arg46 of the V1aR by all other encoded amino acids. Wildtype V1aR (wt) and [R46X]V1aR constructs were expressed in HEK 293T cells and characterized pharmacologically. (A) Rank order of [R46X]V1aR constructs with respect to the binding affinity (pKi) of AVP and (B) rank order of [R46X]V1aR constructs with respect to pEC50 value for AVP-induced accumulation of mono-, bis- and tris-phosphates (InsP–InsP3). Data shown are the mean ± SEM of three separate experiments each performed in triplicate. Basal values (mean ± SEM) were 1108 ± 248, 956 ± 206, 958 ± 245, 1240 ± 207, 1064 ± 166, 996 ± 175, 968 ± 139, 1076 ± 213, 1024 ± 197, 1184 ± 117, 1228 ± 230, 1045 ± 205, 1252 ± 241, 1073 ± 186, 978 ± 169, 1037 ± 233, 1214 ± 206, 1008 ± 185, 1163 ± 202 and 1236 ± 242 d.p.m. for wildtype V1aR, R46H, R46Q, R46N, R46P, R46V, R46F, R46Y, R46A, R46K, R46C, R46S, R46M, R46G, R46I, R46T, R46W, R46L, R46D and R46E mutant receptors, respectively. Mock-transfected cells did not bind ligand or exhibit AVP-induced accumulation of inositol phosphates.

Fig. 2. Pharmacological characterization of [R46A]V1aR. (A) Compe- tition radioligand binding studies with agonist AVP in the absence (h,j) or presence (n,m) of 10)4 M 5¢-guanylylimidodiphosphate, or nonpeptide antagonist SR 49059 (s,d) were performed using a membrane preparation of HEK 293T cells transiently transfected with wildtype V1aR (open symbols) or [R46A]V1aR (filled symbols). Data are the mean ± SEM of three separate experiments each performed in triplicate. Values are expressed as percent specific binding, where nonspecific binding was defined by d(CH2)5Tyr(Me)2AVP (10 lM). A theoretical Langmuir binding isotherm has been fitted to the experi- mental data as described in Experimental procedures. (B) AVP- induced accumulation of mono-, bis- and tris-phosphates (InsP–InsP3) in HEK 293T cells transfected with wildtype V1aR (h) or [R46A]V1aR mutant (j). Data are the mean ± SEM of three separate experiments each performed in triplicate. Values are expressed as percent maximum stimulation induced by AVP at the stated concentrations.

generation in Fig. 3B. Collectively, these results established that Arg46 has a critical role in agonist activation of the V1aR.

Discussion

Receptors which mediate the effects of the neurohypophy- sial peptide hormone family are structurally and pharma- cologically related and as such form a subclass of GPCRs (reviewed in [3–6]). It has been shown recently that the N-terminus of the V1aR provides agonist-specific binding epitopes. Consequently, truncation of the V1aR N-terminus prevented high affinity AVP binding but did not affect antagonist binding [10]. A single arginyl is conserved in the distal N-terminus of all V1aRs and OTRs cloned to date, suggesting functional importance (Fig. 1). This has now been confirmed experimentally, as substitution of this residue by alanine in the rat V1aR and the human OTR

4686 S. R. Hawtin et al. (Eur. J. Biochem. 270)

(cid:2) FEBS 2003

Fig. 4. Lack of correlation between PEC50 values and PKi values for [R46X]V1aR constructs. Each square represents one of the 19 different [R46X]V1aR mutant constructs; m, wildtype V1aR.

profoundly and selectively disrupted agonist binding and signalling [14,15].

and functional EC50 values to markedly different extents. For example, substitution by His and Trp in [R46H]V1aR and [R46W]V1aR resulted in a decrease in affinity of approximately 700-fold compared to the wildtype V1aR, but the increase in EC50 was 10-fold and 600-fold respectively (Table 1). Indeed, there is no correlation between the Ki and EC50 values for the mutant receptors (Fig. 4).

Fig. 5. Agonist-induced activation of the V1aR is dictated by the residue at position 46. Wildtype V1aR (wt) and [R46X]V1aR constructs were expressed in HEK 293T cells and characterized pharmacologically. In each case the binding affinity (pKi) of AVP and the pEC50 value for AVP-induced accumulation of mono-, bis- and tris-phosphates (InsP– InsP3) was determined from three separate experiments each per- formed in triplicate. The pEC50 ) pKi value was calculated and the [R46X] substitutions presented in rank order.

The ratio of EC50 to Ki is an indicator of efficacy, i.e. the likelihood that a receptor will become activated and initiate a functional response once an agonist has bound [26]. This parameter (expressed as pEC50 ) pKi) was increased for all 19 substitutions compared to the wildtype Arg46, with the precise value depending on the identity of the substituent amino acid (Fig. 5). These data establish that the mutant receptors are much less likely than the wildtype to bind AVP, but once AVP has bound all 19 mutant receptors are Having established the functional necessity of Arg46 in the V1aR it was important to identify the features of the arginyl side-chain that supported high affinity agonist binding. A comprehensive approach was undertaken in which each of the 19 alternative amino acids encoded by mRNA were substituted at position 46 and the biological characteristics of the mutant receptors assessed. None of the other amino acids could replace Arg46 whilst still main- taining the wildtype pharmacological characteristics with respect to either agonist binding (Fig. 3A) or second messenger generation (Fig. 3B). This was particularly noticeable for AVP binding, where the affinity (Ki) was almost uniformly impaired irrespective of the nature of the substitution (Fig. 3A). In contrast, second messenger gen- eration was more sensitive to the amino acid at residue 46, with the EC50 value for AVP-induced InsP production varying between 10 fold higher than the wildtype to 600-fold higher (Fig. 3B). Lys46 was not an effective substitute for Arg46. A major feature of the arginyl side-chain is the guanidinium moiety. High concentrations of guanidinium ion (10 mM) however, could not endow the [R46A]V1aR construct with wildtype pharmacology. Therefore free guanidinium could not be co-ordinated within the mutant receptor in the appropriate orientation to recover agonist binding. Although Lys46 is an inadequate substitution, charge is nevertheless a significant aspect of the Arg at this locus in the wildtype V1aR because reversal of the charge in [R46E]V1aR and [R46D]V1aR resulted in the least respon- sive receptors of the 20 studied (Fig. 3B). In addition, removal of this negative charge with [R46Q]V1aR and [R46N]V1aR increased the responsiveness to AVP by 14-fold and 30-fold respectively (Table 1, Fig. 3B). While differences in the observed Emax values may reflect differ- ences in transfection efficiencies, it is also possible that a number of different active conformations have been gener- ated, some better able to activate the G-protein Gq than others.

Trp and Gly represent the two extremes of side-chain size, with accessible surface areas of 210 A˚ and 33 A˚ respectively [24], but [R46W]V1aR and [R46G]V1aR had very similar EC50 values for InsP generation (Table 1). Therefore, residue 46 is not spatially restricted within the receptor architecture. Generation of an intracellular signal requires the receptor to adopt an active conformation in addition to binding the agonist. Ground-state (R) and active (R*) conformations of GPCRs exist in equilibrium. Agonists have a higher affinity for the active receptor (R*) which stabilizes this conformation and subsequently establishes productive R*–G-protein coupling (reviewed in [25]). All of the Arg46 substitutions in this study produced a marked reduction in the ability of AVP to both bind to the receptor (Fig. 3A) and to initiate a second messenger response (Fig. 3B). It might be supposed that the impaired signalling simply reflected a reduced ability of the mutant receptors to assume an active conformation. Such a situation would clearly lead to a reduction in signalling response and would also produce a decrease in agonist affinity, as agonists have a low affinity for the ground state of the receptor. However, close inspection of the data reveals that this is not the case. Different amino acid substitutions affect binding affinity

(cid:2) FEBS 2003

The N-terminal switch of the V1aR (Eur. J. Biochem. 270) 4687

& Boer, G.J., eds), pp. 335–368. Plenum Press, New York, NY, USA.

8. Pettibone, D.J. & Freidinger, R.M. (1997) Discovery and devel- opment of non-peptide antagonists of peptide hormone receptors. Biochem. Soc. Trans. 25, 1051–1057.

9. Freidinger, R.M. & Pettibone, D.J. (1997) Small molecule ligands for oxytocin and vasopressin receptors. Med. Res. Rev. 17, 1–16. 10. Hawtin, S.R., Wesley, V.J., Parslow, R.A., Patel, S. & Wheatley, M. (2000) Critical role of a subdomain of the N-terminus of the V1a vasopressin receptor for binding agonists; functional rescue by the oxytocin N-terminus. Biochemistry 39, 13524–13533.

11. Postina, R., Kojro, E. & Fahrenholz, F. (1996) Separate agonist and peptide antagonist binding sites of the oxytocin receptor defined by their transfer into the V2 vasopressin receptor. J. Biol. Chem. 271, 31593–31601.

12. Hawtin, S.R., Howard, H.C. & Wheatley, M. (2001) Identification of an extracellular segment of the oxytocin receptor providing agonist-specific binding epitopes. Biochem. J. 354, 465–472. 13. Hausmann, H., Richters, A., Kreienkamp, H.-J., Meyerhof, W., Mattes, H., Lederis, K., Zwiers, H. & Richter, D. (1996) Muta- tional analysis and molecular modeling of the nonapeptide hor- mone binding domains of the [Arg8]vasotocin receptor. Proc. Natl Acad. Sci. USA 93, 6907–6912.

14. Hawtin, S.R., Wesley, V.J., Parslow, R.A., Patel, S. & Wheatley, M. (2002) A single residue (Arg46) located within the N-terminus of the V1a vasopressin receptor is critical for binding vasopressin but not peptide or non-peptide antagonists. Mol. Endocrinol. 16, 600–609.

15. Wesley, V.J., Hawtin, S.R., Howard, H.C. & Wheatley, M. (2002) Agonist-specific, high affinity binding epitopes are contributed by an arginine in the N-terminus of the human oxytocin receptor. Biochemistry 41, 5086–5092.

more likely than the wildtype V1aR to become activated. This implies that arginyl at this locus is a constraining residue which contributes to maintaining the conformational switch of the V1aR in the off-state. The unique ability of arginyl to fulfil this constraining function in the N-terminus of the V1aR is similar to the unique constraining role of alanyl previously reported for Ala293 in the distal i3-loop of the a1b-adrenergic receptor [27]. Although substitution of Arg46 facilitated agonist-induced transition of the V1aR to the active state, signalling was nevertheless still dependent on agonist because an increase in basal signalling was not observed. Furthermore, the marked decrease in agonist affinity with the [R46X]V1aR constructs implies that Arg46 has an additional function facilitating high affinity agonist binding. Consequently, mutation of Arg46 had the dual effect of (a) decreasing agonist affinity and (b) promoting the agonist-induced active conformation due to the loss of a stabilizing constraint on the ground state of the receptor. Therefore, the agonist had a lower affinity than the wildtype V1aR but once the agonist bound, the receptor was more likely to signal. This dual role suggests that binding of the agonist releases the Arg46-mediated constraint on the ground state of the receptor, thereby promoting agonist- induced activation. Interestingly, (cid:2)dual-role(cid:3) residues were also identified in transmembrane helix VII of the M1 muscarinic acetylcholine receptor (mAChR) recently. In this mAChR however, these residues stabilize the ground state of the receptor but also subsequently stabilize the G-protein interaction of the active receptor [28].

16. Wheatley, M., Howl, J., Yarwood, N.J., Davies, A.R.L. & Parslow, R.A. (1997) Preparation of a membrane fraction for receptor studies and solubilisation of receptor proteins with retention of biological activity. Methods Mol. Biol. 73, 305–332. 17. Howl, J., Langel, U¨ ., Hawtin, S.R., Valkna, A., Yarwood, N.J., Saar, K. & Wheatley, M. (1997) Chimeric strategies for the rational design of bioactive analogs of small peptide hormones. FASEB J. 11, 582–590.

In summary, Arg46 of the V1aR is required for high affinity agonist binding and signalling. It is one of the constraining residues which maintain the V1aR in the inactive conformation and as such is part of the receptor activation switch. There is an absolute requirement for arginyl at position 46 for these functions.

Acknowledgements

We are grateful to Dr Claudine Serradeil-Le Gal (Sanofi Recherche, France) for providing a sample of SR 49059. This work was supported by a grant to M.W. from the Biotechnology and Biological Sciences Research Council.

References

18. Kruszynski, M., Lammek, B., Manning, M., Seto, J., Haldar, J. & Sawyer, W.H. (1980) 1-(b-mercapto-b,b-cyclopentamethylene- propionic acid), 2-(O-methyl) tyrosine arginine vasopressin and 1-(b-mercapto-b,b-cyclopentamethylenepropionic acid) arginine vasopressin, two highly potent antagonists of the vasopressor response to arginine vasopressin. J. Med. Chem. 23, 364–368. 19. Cheng, Y. & Prusoff, W.H. (1973) Relationship between the inhibition constsant (Ki) and concentration of inhibitor which causes 50 % inhibition (IC50) of an enzyme reaction. Biochem. Pharmacol. 22, 3099–3108.

1. Acher, R., Chauvet, J. & Chauvet, M.T. (1995) Man and the chimera. Selective versus neutral oxytocin evolution. Adv. Exp. Med. Biol. 395, 615–627.

2. Baldwin, J.C. (1993) The probable arrangement of the helices in

G-protein-coupled receptors. EMBO J. 12, 1693–1703.

3. Howl, J. & Wheatley, M. (1995) Molecular pharmacology of V1a

20. Howl, J., Rudge, S.A., Lavis, R.A., Davies, A.R.L., Parslow, R.A., Hughes, P.J., Kirk, C.J., Michell, R.H. & Wheatley, M. (1995) Rat testicular myoid cells express vasopressin receptors: receptor structure, signal transduction and developmental regulation. Endocrinology 136, 2206–2213.

vasopressin receptors. General Pharmacol. 26, 1143–1152.

4. Hibert, M., Hoflack, J., Trumpp-Kallmeyer, S., Mouillac, B., Chini, B., Mahe, E., Cotte, N., Jard, S., Manning, M. & Barberis, C. (1999) Functional architecture of vasopressin/oxytocin receptors. J. Recept. Signal Transduct. Res. 19, 589–596.

21. Hawtin, S.R., Tobin, A., Patel, S. & Wheatley, M. (2001) Palmi- toylation of the V1a vasopressin receptor reveals different conforma- tional requirements for signaling, agonist-induced receptor phos- phorylation and sequestration. J. Biol. Chem. 276, 38139–38146. 22. Schmidt, A., Audigier, S., Barberis, C., Jard, S., Manning, M., Kolodziejczyk, A.S. & Sawyer, W.H. (1991) A radioiodinated linear vasopressin antagonist. FEBS Lett. 282, 77–81.

23. Serradeil-Le Gal, C., Wagnon, J., Garcia, C., Lacour, C., Guiraudou, P., Christophe, B., Villanova, G., Nisato, D., Maffrand, J.P. & Le Fur, G.P. (1993) Biochemical and pharma- cological properties of SR 49059, a new, potent, non-peptide

5. Gimpl, G. & Fahrenholz, F. (2001) The oxytocin receptor system: Structure, function and regulation. Physiol. Rev. 81, 629–683. 6. Thibonnier, M., Coles, P., Thibonnier, A. & Shoham, M. (2001) The basic and clinical pharmacology of non-peptide vasopressin receptor antagonists. Annu. Rev. Pharmacol. Toxicol. 41, 175–202. 7. Manning, M., Bankowski, K. & Sawyer, W.H. (1987) Selective agonists and antagonists of vasopressin. In Vasopressin (Gash, D.M.

4688 S. R. Hawtin et al. (Eur. J. Biochem. 270)

(cid:2) FEBS 2003

antagonist of rat and human vasopressin V1a receptors. J. Clin. Invest. 92, 224–231.

27. Kjelsberg, M.A., Cotecchia, S., Ostrowski, J., Caron, M.G. & Lefkowitz, R.J. (1992) Constitutive activation of the alpha 1B-adrenergic receptor by all amino acid substitutions at a single site. J. Biol. Chem. 267, 1430–1433.

24. Richardson, J.S. & Richardson, D.C. (1989) Principles and pat- terns of protein conformation. In Prediction of Protein Structure and the Principles of Protein Conformation (Fasman, G.D., ed.), pp. 1–98. Plenum Press, New York, NY, USA.

28. Lu, Z.-L., Saldanha, J.W. & Hulme, E.C. (2001) Transmembrane domains 4 and 7 of the M1 muscarinic acetylcholine receptor are critical for ligand binding and the receptor activation switch. J. Biol. Chem. 276, 34098–34104.

25. Gether, U. & Kobilka, B.K. (1998) G-protein-coupled receptors: II. Mechanism of agonist activation. J. Biol. Chem. 273, 17979– 17982.

26. Christopoulos, A. & El-Fakahany, E.E. (1999) Qualitative and quantitative assessment of relative agonist efficacy. Biochem. Pharmacol. 58, 735–748.

29. Hawtin, S.R., Davies, A.R.L., Matthews, G. & Wheatley, M. (2001) Identification of the glycosylation sites utilized on the V1a vasopressin receptor and assessment of their role in receptor sig- nalling and expression. Biochem. J. 357, 73–81.