Two Basic Residues of the h-VPAC 1 Receptor Second Transmembrane Helix Are Essential for Ligand Binding and Signal Transduction*

We mutated the vasoactive intestinal peptide (VIP) Asp 3 residue and two VPAC 1 receptor second transmem- brane helix basic residues (Arg 188 and Lys 195 ). VIP had a lower affinity for R188Q, R188L, K195Q, and K195I VPAC 1 receptors than for VPAC 1 receptors. [Asn 3 ] VIP and [Gln 3 ] VIP had lower affinities than VIP for VPAC 1 receptors but higher affinities for the mutant receptors; the two basic amino acids facilitated the introduction of the negatively charged aspartate inside the transmembrane domain. The resulting interaction was necessary for receptor activation. 1/[Asn 3 ] VIP and [Gln 3 ] VIP were partial agonists at VPAC 1 receptors; 2/VIP did not fully activate the K195Q, K195I, R188Q, and R188L VPAC 1 receptors; a VIP analogue ([Arg 16 ] VIP) was more efficient than VIP at the four mutated receptors; and [Asn 3 ] VIP and [Gln 3 ] VIP were more efficient than VIP at the R188Q and R188L VPAC 1 receptors; 3/the [Asp 3 ] nega- tive charge did not contribute to the recognition of the VIP 1 antagonist, [AcHis 1 ,D-Phe 2 ,Lys 15 ,Arg 16 ,Leu 27 ] VIP (1–7)/growth

The neuropeptides vasoactive intestinal polypeptide (VIP) 1 and pituitary adenylate cyclase-activating polypeptide (PACAP) contribute to the regulation of intestinal secretion and motility, of the vascular tone, of the exocrine and endocrine secretions, of immunological responses, and to the development of the central nervous system (1)(2)(3). The effects of VIP are mediated through interaction with two receptor subclasses named the VPAC 1 and VPAC 2 receptors; the effects of PACAP are also mediated through interactions with the same receptors, as well as through a selective receptor named PAC 1 (3,4). VPAC 1 , VPAC 2 , and PAC 1 receptors are encoded by different genes and expressed in different cell populations in both the central nervous system and peripheral tissues (3,5,6). They are preferentially coupled to G␣ s proteins that stimulate adenylate cyclase activity. The PAC 1 and VPAC 1 receptors may stimulate, in addition, inositol trisphosphate synthesis and calcium mobilization (7,8). This effect is however detected only at high VPAC 1 receptor expression levels (8). VIP and PACAP receptors are members of a large family of G protein-coupled receptors, often referred as the GPCR-B family (4,9), that includes the secretin, glucagon, glucagon-like peptide-1, calcitonin, parathyroid hormone, and growth hormone releasing factor (GRF) receptors. The VIP, PACAP, secretin, and GRF receptors constitute a subfamily based on the homology of the ligands and of the receptors. Each receptor recognizes its own cognate ligand with a high affinity but recognizes at least one other parent peptide with a comparable or a lower affinity (4). Because of the sequence homology of the ligands and the receptors, the information obtained on one receptor-ligand pair can be anticipated to be relevant also in the other systems.
The positioning of the ligand on the receptors is still poorly understood. Investigations of chimeric receptors and mutants have indicated that the large amino-terminal domain (10 -13) structured by disulfide bridges (14 -16) makes a key contribution to ligand recognition that several other highly conserved residues play a role in the general structure (17,18) and that creating constitutively active receptors through mutations in the intracellular part of the receptor is possible (19).
The amino-terminal part of the ligand is necessary for high affinity binding and for second messenger activation; its deletion in VIP, PACAP, and secretin reduced both the affinity and the intrinsic activity of the peptide (20). The identification of the receptor residues interacting with the amino terminus of the ligand is a prerequisite to model the active form of the receptor and conceive new ligands, preferably non-peptidic, that could be of therapeutic interest. We focused in this work on the human VPAC 1 receptors and investigated the contribution of two basic residues located in the second transmembrane helix to ligand recognition and adenylate cyclase activation.
We obtained evidence that both basic residues are important for recognition of the Asp 3 of VIP and stabilization of the active VIP-receptor complex conformation. Our results also suggested that a second binding mode, that does not involve recognition of the VIP Asp 3 residue and does not induce receptor activation, is also possible. Four Mutated Receptors, R188L, R188Q, K195Q,  and K195I, and their Stable Expression in Chinese Hamster Ovary (CHO) Cells-The human VPAC 1 receptor cDNA was cloned by PCR according to the previously reported sequence (21), using specific primers. Generation of the four mutated receptors was achieved using the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla CA) essentially according to the manufacturer's instructions. Briefly, the human VPAC 1 receptor-coding region, inserted into the mammalian expression vector pcDNA3.1 (Invitrogen Corp.), was submitted to 22 cycles of PCR (95°C for 30 s, 54°C for 1 min, and 68°C for 14 min) in a 50-l reaction volume. The forward and reverse primers were complementary and contained the desired nucleotide changes, flanked on either side by 15 perfectly matched nucleotides (only the forward primers are shown): Arg 188 to Leu, CATATCCTTCATCCTGCTGGCTGCC-GCTGTCTTC; Arg 188 to Gln, CATATCCTTCATCCTGCAGGCTGCCG-CTGTCTTC; Lys 195 to Ile, GCCGCTGTCTTCATCATAGACTTGGCCC-TCTTC; Lys 195 to Gln, GCCGCTGTCTTCATCCAAGACTTGGCCCTC-TTC.

Construction of the
Following PCR, 10 l were analyzed by agarose gel electrophoresis, and the remaining 40 l were digested for at least 2 h by 1 l of DpnI restriction enzyme (Stratagene, La Jolla, CA) to remove the parental methylated DNA. The digested PCR products were transformed into TOP10 One Shot competent Escherichia coli bacterial cells (Invitrogen Corp.). Of several colonies verified by agarose gel electrophoresis of miniprep plasmid DNA (22), three were retained and further purified on Qiaquick PCR purification spin columns, and the mutations were checked for by DNA sequencing on an ABI automated sequencing apparatus, using the BigDye Terminator sequencing prism kit from ABI (Perkin-Elmer). Plasmid DNA from one clone for each mutation, containing the correct nucleotide substitutions, was prepared using a midiprep endotoxin-free kit (Stratagene, La Jolla, CA), the complete nucleotide sequence of the receptor coding region was verified by DNA sequencing, and 20 g were electroporated (Electroporator II; Invitrogen Corp.) into wild type CHO-K1 cells. Selection was carried out in culture medium (50% HamF12, 50% Dulbecco's modified Eagle's medium, 10% fetal calf serum, 1% penicillin (10 milliunits/ml), 1% streptomycin (10 g/ml), 1% L-glutamine (200 mM; Life Technologies LTD, Paisley, United Kingdom), supplemented with 600 g of geneticin (Gly 418 )/ml culture medium. After 10 to 15 days of selection, isolated colonies were transferred to 24-well microtiter plates and grown until confluence, trypsinized, and further expanded in 6-well microtiter plates, from which cells were scraped and membranes prepared for screening by an adenylate cyclase activity assay in the presence of 10 M VIP.
Membrane Preparation-Membranes were prepared from scraped cells lysed in 1 mM NaHCO 3 by immediate freezing in liquid nitrogen. After thawing, the lysate was first centrifuged at 4°C for 10 min at 400 ϫ g, and the supernatant was further centrifuged at 20,000 ϫ g for 10 min. The resulting pellet, resuspended in 1 mM NaHCO 3 , was used immediately as a crude membrane fraction.
Radioiodination of Three Different Tracers and Binding Studies-Binding studies were performed as described (23) using 125 I-VIP, 125 I-[Gln 3 ] VIP, or 125 I-VIP 1 antagonist. The three tracers were radiolabeled similarly and had comparable specific radioactivity (6,23). In all cases, the nonspecific binding was defined as residual binding in the presence of the corresponding unlabeled peptide (1 M). Binding was performed at 20°C in a total volume of 120 l containing 20 mM Tris-maleate, 2 mM MgCl 2 , 0.1 mg/ml bacitracin, 1% bovine serum albumin (pH 7.4) buffer. 3 to 30 g of protein were used per assay. Bound and free radioactivity were separated by filtration through glass-fiber GF/C filters presoaked for 24 h in 0.01% polyethyleneimine and rinsed three times with a 20 mM (pH 7.4) sodium phosphate buffer containing 1% bovine serum albumin.
Adenylate Cyclase Activation-Adenylate cyclase activity was determined by the procedure of Salomon et al. (24), as described previously. Membrane proteins (3-15 g) were incubated in a total volume of 60 l containing 0.5 mM [␣ 32 P]ATP, 10 M GTP, 5 mM MgCl 2 , 0.5 mM EGTA, 1 mM cAMP, 1 mM theophylline, 10 mM phospho(enol)pyruvate, 30 g/ml pyruvate kinase, and 30 mM Tris-HCl at a final pH value of 7.8. The reaction was initiated by membranes addition and was terminated after a 15-min incubation at 37°C by adding 0.5 ml of a 0.5% sodium dodecyl sulfate solution containing 0.5 mM ATP, 0.5 mM cAMP, and 20,000 cpm [ 3 H]cAMP. cAMP was separated from ATP by two succes-sive chromatographies on Dowex 50W ϫ 8 and neutral alumina. . The peptide purity (Ͼ97%) was assessed by capillary electrophoresis, and the conformity was verified by electrospray mass spectrometry.
Data Analysis-All competition curves and dose-effect curves were analyzed by a non-linear regression program (Graph Pad Prism, San Diego, CA).  (Table II). The VIP 1 antagonist, PG 97 269 (23), did not stimulate adenylate cyclase at any concentration but inhibited competitively the effect of VIP with a K i value of 2-3 nM.

Interaction of VIP Analogues with the Human Wild Type
Analysis of the Mutated R188Q, R188L, K195Q, and K195I VPAC 1 Receptors-The clones transfected with the DNA coding for the mutated receptors were selected on the basis of the ability of 10 M VIP membranes to activate adenylate cyclase.
The selected clones were then tested for 125 I-VIP binding. Binding was non-significant in all the clones tested. As expected from the binding data, the EC 50 values for VIP were 30 to 100-fold higher than at VPAC 1 receptors (Fig. 1). There was no noticeable difference between the R188L and R188Q receptors on one hand and the K195I and K195Q on the other hand.
We then tested the ability of VIP analogues to stimulate adenylate cyclase. [Arg 16 ] VIP, [Asn 3 ] VIP, and [Gln 3 ] VIP were more potent and more efficient than VIP at the R188L-and R188Q-VPAC 1 receptors, and [Glu 3 ] VIP and the VIP 1 agonist behaved as partial agonists on both mutant receptors ( Fig. 1  and results not shown). The results are compared in Table II with the data obtained at wild type receptors with the same peptides.
The VIP 1 antagonists' affinity for the mutated receptors was evaluated by comparing VIP, [Arg 16 ] VIP, or VIP 1 agonist doseeffect curves in the absence and presence of 0.1, 0.3, or 1.0 M antagonist ( Fig. 2 and results not shown). The VIP dose-effect curves were shifted dose-dependently to higher concentrations in the presence of antagonist, as expected for K i values Ϸ 2-3 nM at the five receptors ( Fig. 2 and results not shown).
Taken together, these results suggested that the affinity of [Gln 3 ] VIP and of the VIP 1 antagonist for the mutant receptors might be sufficient to allow binding studies using these radiolabeled peptides. This was indeed the case. The binding experiments were performed at 20°C as tracer binding was unstable at 37°C (probably because of a receptor instability, as already noted for chimeric receptors (10,25)). The binding of both tracers was rapid (equilibrium was achieved within 20 min) and reversible (data not shown). The peptides' IC 50 values (reported in Table I) did not depend on the tracer used. Representative competition curves are shown in Fig. 3. The R188Q and K195Q receptor concentrations were comparable (within 2-fold) to the VPAC 1 receptor concentration.
We were surprised that, even though the presence of a negative charge in position 3 was clearly deleterious for recognition of the R188Q, R188L, K195Q, and K195I mutant VPAC 1 receptors by agonists, the VIP 1 antagonist (that also possesses an Asp 3 residue) retained a high affinity for the mutated receptors. We therefore synthesized and tested two additional peptides with an Asn 3 residue, the Asn 3 VIP 1 agonist and Asn 3 VIP 1 antagonist. As shown in Tables I and II, replacing the Asp 3 by an Asn 3 residue reduced the affinity and efficacy of the VIP 1 agonist at wild type receptors but increased its affinity and efficacy on the mutated R188Q receptor. In contrast, replacing Asp 3 by an Asn 3 residue in the VIP 1 antagonist did not affect its affinity for the receptors studied in this work; the modified peptide always behaved as a high affinity antagonist (Fig. 4). DISCUSSION VIP receptors belong to a G protein-coupled receptor family that does not share any sequence homology with rhodopsin or with the ␤-adrenergic receptor family. Previous work with chimeric receptors (see Refs. 10 -13 and 29 -34) and cross-linking studies (35)(36)(37) led to a general VIP positioning model; the amino-terminal and carboxyl-terminal sequences of VIP and of its analogues seemed to interact with the "7TM" (transmem-   In view of the significant sequence homologies between secretin, VIP, PACAP, GRF, glucagon, glucagon-like peptide , and gastric inhibitory peptide on the one hand and between their receptors on the other hand, we hypothesized that these different peptides recognize "equivalent" binding sites in their respective receptors. The observation that all the aforementioned peptides except glucagon possess an acidic residue (Asp or Glu) in position 3, and that all the receptors (except glucagon's) possess aligned basic Arg and Lys residues at the extracellular end of TM2, led us to suspect that these residues might interact in the peptide-receptor complex.
Our preliminary results confirmed that, as in secretin receptors (17,38), the VPAC 1 receptor Arg 188 and Lys 195 were important for VIP recognition; we were unable to obtain significant 125 I-VIP binding at the mutated receptors, and very high VIP concentrations were necessary to activate the adenylate cyclase. The following results suggested in addition that VIP was unable to stabilize the mutated receptors' active conformations sufficiently to ensure full receptor activation: [Asn 3 ] VIP, [Gln 3 ] VIP, and [Arg 16 ] VIP were more efficient than VIP at the R188Q and R188L VPAC 1 receptors, and [Arg 16 ] VIP was more efficient than VIP at the K195Q and K195I VPAC 1 receptors.
When VIP and its receptor are free, the VIP Asp 3 and the receptor Arg 188 and Lys 195 side chains probably form dipole-ion interactions with surrounding water molecules. These favorable interactions are disrupted upon ligand binding; they must be compensated by ligand-receptor interactions to allow high affinity binding. The "uncharged VIP analogues" [Asn 3 ] VIP and [Gln 3 ] VIP had a higher affinity than VIP for the "uncharged receptor mutants" (R188Q, R188L, K195Q, and K195I VPAC 1 receptors), suggesting that the VIP Asp 3 and the receptor Arg 188 and Lys 195 side chains were in close proximity in the agonist-receptor complex. The affinity loss that we observed upon replacement of the VIP Asp 3 , VPAC 1 receptor Arg 188 , and VPAC 1 receptor Lys 195 (30-to 100-fold) was however comparatively small and did not support the hypothesis that the Asp negative charge is close enough to the receptor Arg and Lys positive charges to form ionic bonds (40). It is more likely that Asp 3 , Arg 188 , and Lys 195 formed strong hydrogen bonds (i.e. dipole-dipole or ion-dipole interactions); the two receptor basic residues probably participated in the formation of an electrophilic pocket that recognized the negatively charged VIP Asp 3 side chain.
[Asn 3 ] VIP and [Gln 3 ] VIP behaved as partial agonists at wild type receptors but were more efficient than VIP at the R188Q and R188L mutant receptors. The incomplete activation of the mutant receptors by VIP might be caused by difficulties in burying the anionic Asp 3 side chain deep enough in the (uncharged) mutant receptors' binding site. This hypothesis is indirectly supported by the observation that replacing the VIP Asp 3 or the VPAC 1 receptor Arg 188 or Lys 195 with uncharged amino acids affected the recognition of the VIP 1 agonist (an efficient partial agonist) somewhat less than the recognition of VIP and did not affect binding of the VIP 1 antagonist, a compound that does not induce receptor activation. Due perhaps to steric hindrance between the receptor and the large D-Phe 2 side chain, the VIP 1 antagonist Asp 3 was apparently unable to enter the agonist binding pocket and to trigger receptor activation.
The parathyroid hormone (PTH) receptor belongs to the same receptor family as the VPAC 1 receptor. A molecular model of the PTH-receptor interaction has been developed, based on experimental data from cross-linking studies, spectroscopic investigations of the hormone and receptor fragments, and theoretical structure predictions (39). According to this model, PTH recognizes extracellular receptor domains (the amino-terminal domain and extracellular loops) but penetrates very little if at all inside the compact transmembrane helices bundle. It is tempting to suggest that, like PTH, VIP initially recognizes an extracellular binding site. In a second step, driven and stabilized i.e. by the Asp 3 -Lys 195 /Arg 188 interactions, a transmembrane binding pocket opens and recognizes the agonists' amino-terminal amino acids, and the receptor activates intracellular G proteins. "Too large" amino-terminal VIP amino acids (D-Phe 2 instead of Ser 2 and pCl-Phe 6 instead of Phe 6 ) might prevent the recognition of this activated receptor conformation by steric hindrance; [D-Phe 2 ] and [pCl-Phe 6 ] VIP or VIP/GRF analogues usually behave as VIP antagonists (27,28).
The location of the VIP and secretin "Asp 3 binding site" to transmembrane helix 2 was somewhat unexpected; indeed, in the rhodopsin-like ␤-adrenergic G protein-coupled receptor family, the ligand binding pocket is lined by TMs 3 to 7 and does not involve TMs 1 and 2. It is important to note in this respect that most of the "signature" amino acids that define the ␤-adrenergic G protein-coupled receptor family, including the proline residues that participate in the formation of the agonist binding pocket, are absent from the secretin receptor family (including VPAC 1 receptors). Our results suggest that (in contrast with the G protein binding site that appears to involve the same intracellular loops in both receptor families) the agonist binding site was located in very different transmembrane regions in the secretinand ␤-adrenergic-receptor families. Further studies will be needed to extend this observation and allow the construction of an activated agonist-receptor complex model.
To conclude, our present results suggested that the VIP Asp 3 side chain fitted inside the transmembrane helix bundle, in close proximity to TM2 Lys 195 and Arg 188 . This interaction was essential for receptor activation. The VIP 1 antagonist Asp 3 residue did not recognize the same binding pocket perhaps because of unfavorable coulombic interactions between the D-Phe 2 side chain and the receptor.