Evidence for a Direct Interaction between the Penultimate Aspartic Acid of Cholecystokinin and Histidine 207, Located in the Second Extracellular Loop of the Cholecystokinin B Receptor*

Recently, we reported that the mutation of His207 to Phe located in the second extracellular loop of the cholecystokinin B receptor strongly affected cholecystokinin (CCK) binding (Silvente-Poirot, S., Escrieut, C., and Wank, S. A. (1998) Mol. Pharmacol. 54, 364–371). To characterize the functional group in CCK that interacts with His207, we first substituted His207 to Ala. This mutation decreased the affinity and the potency of CCK to produce total inositol phosphates 302-fold and 456-fold without affecting the expression of the mutant receptor. The screening ofl-alanine-modified CCK peptides to bind and activate the wild type and mutant receptors allowed the identification of the interaction of the C-terminal Asp8 of CCK with His207. The H207A-CCKBR mutant, unlike the wild type receptor, was insensitive to substitution of Asp8 of CCK to other amino acid residues. This interaction was further confirmed by mutating His207 to Asp. The affinity of CCK for the H207D-CCKBR mutant was 100-fold lower than for the H207A-CCKBR mutant, consistent with an electrostatic repulsion between the negative charges of the two interacting aspartic acids. Peptides with neutral amino acids in position eight of CCK reversed this effect and displayed a gain of affinity for the H207D mutant compared with CCK. To date, this is the first report concerning the identification of a direct contact point between the CCKB receptor and CCK.

Cholecystokinin (CCK) 1 is found throughout the gastrointestinal tract and the central nervous system where it functions both as a hormone and a neurotransmitter (1). CCK exists physiologically in multiple forms processed from a 115-amino acid preprohormone. Post-translational processing of CCK involved sulfation of tyrosine at position seven from the C terminus and amidation of the C-terminal phenylalanine. In the gastrointestinal system, CCK regulates motility, pancreatic exocrine secretions and growth, gastric emptying, and inhibition of gastric acid secretion. In the nervous system, CCK is implicated in anxiogenesis, satiety, analgesia, and regulation of dopamine release. The actions of CCK are mediated by two pharmacologically distinct receptor subtypes, the CCKA and the CCKB receptors. The longer sulfated forms (CCK-58, CCK-39, CCK-33, and CCK-8) bind to the CCKAR and CCKBR with similar nanomolar affinities. The nonsulfated peptides as well as shorter C-terminal fragments like CCK-4 and CCK-5 still bind with nanomolar affinities the CCKBR but display very low affinities for the CCKAR. Gastrin, a related family peptide that shares with CCK the C-terminal pentapeptide also discriminates the two subtypes by presenting similar affinity as CCK-8 for the CCKBR but micromolar affinity for CCKAR. Studies using synthesized CCK fragments have shown that the Cterminal sulfated and amidated octapeptide Asp-Tyr(SO 3 H)-Met-Gly-Trp-Met-Asp-Phe-NH 2 retains the full spectrum of biological activity (2)(3)(4)(5). The cloning of the cDNA coding for these receptors has shown that CCKA and CCKB receptors belong to the superfamily of seven transmembrane G-protein-coupled receptor and that they have approximately 50% homology (6,7). Due to the important physiological roles of CCK acting through these two receptor subtypes and, therefore, its possible implication in associated disorders, there has been considerable interest in the identification of ligands that selectively activate or inhibit the CCKAR and CCKBR. However, to date none of these compounds have been used as a therapeutic agent (8 -13). The characterization of the interactions between the key pharmacophores of CCK and their partners in the receptor would certainly help in a more rational approach to the design of new molecules or the modification of preexisting molecules to optimize their properties. Despite numerous studies concerning the identification of amino acids involved in the binding site of agonist and antagonist ligands for the different classes of G-protein-coupled receptors (14,15), very few studies concerned the identification of the functional groups in the ligand that directly interact with the identified amino acid in the receptor. However such studies are essential to model the binding pocket accurately. One way to address this issue is through the use of complementary substitutions in the ligand and the receptor as previously reported for G protein-coupled receptors for neurotransmitters and peptide receptors (16 -20). We successfully used this approach for the CCKA receptor subtype to demonstrate that two amino acids Trp 39 and Gln 40 , located in the N-terminal domain of the CCKAR, interact with the residues Arg 1 and Asp 2 of CCK (21). Recently, we identified an-other interaction between Met 195 , located in the second extracellular loop of CCKAR, and the aromatic ring of sulfated tyrosine of CCK (Tyr 3 ), an important pharmacophore for conferring high affinity and activity of the peptide (22). To date, no such study has been reported for the CCKBR. We demonstrated by mutational analysis of the CCKBR that the extracellular domains of the CCKBR are also implicated in the high affinity binding of CCK and gastrin (23,24). We showed that the mutation of the residue His 207 , located in the second extracellular loop of the CCKBR induced a dramatic decrease in CCK binding (24). Regarding the important role of the second extracellular loop for the high affinity binding of CCK in the CCKAR, the present study was undertaken to determine the amino acid within CCK that interacts with His 207 in the CCKBR. The screening of modified CCK peptides for their binding and biological activity toward the wild type receptor and a variety of His 207 -CCKB receptor mutants permitted us to demonstrate that His 207 directly interacts with the C-terminal aspartic acid of CCK, an essential residue for this family of peptides.
Synthesis of Modified CCK Peptides-The L-alanine-modified peptides were synthesized by solid phase synthesis on a Pioneer Perseptive Biosystems apparatus using Fmoc-peptide amide linker-polyethylene glycol flow continuous resin. Fmoc amino acid side chains were protected with the acid-labile protecting groups (Asp: t-butyl; Tyr: t-butyl; Trp, t-butyloxycarbonyl). Coupling reactions were carried out with a 4-fold excess of Fmoc-protected amino acid residues, reagent HBTU (2-(1H-benzotriazol-1-yl)-1,1,3,3,-tetramethyluronium hexafluorophosphate), and diisopropylethylamine for 1 h. At the end of the synthesis, the Fmoc-protecting group was removed, and the peptidyl resin was treated at room temperature for 2 h with 10 ml of reagent K solution (trifluoroacetic acid/water/phenol/thioanisole/ethanedithiol 82.5/5/5/5/ 2.5). The peptides were isolated by precipitation with diethyl ether, solubilized in a mixture of acetonitrile/water/trifluoroacetic acid (50/50/ 0.1, v/v/v), and lyophilized. These peptides were purified by reverse phase HPLC using a Waters Delta Prep 4000 instrument on a Delta-Pak C18 column (15 m, 100 ϫ 150 mm) with UV detection at 214 nm at a flow rate of 50 ml/min of water/trifluoroacetic acid (0.1%) and acetonitrile/trifluoroacetic acid (0.1%). The purity of the peptides was checked by analytical reverse phase HPLC with a Beckman instrument on a Delta-Pak C18 analytical column. The peptide structure was assessed by mass spectroscopy on a Platform II (Micromass, Manchester, UK) quadrupole mass spectrometer fitted with an electrospray interface.
Construction of Mutant Receptor cDNAs-Mutant receptor cDNAs were constructed by oligonucleotide-directed mutagenesis (Quick-Change TM site-directed mutagenesis kit, Stratagene, France) using the rat CCKBR cDNAs as template. Oligonucleotides were designed to include a silent restriction site to facilitate analysis of mutant constructs by restriction endonuclease digestion. Mutations were confirmed by DNA sequencing using an automated sequencer (Applied Biosystems).
Transfection of Wild Type and Mutant Receptor cDNAs into Mammalian Cells-COS-7 cells were grown in Dulbecco's modified Eagle's medium supplemented with 5% fetal calf serum. Two g of the CCKBR and mutant receptor cDNAs subcloned in pCDL-SRa were transiently transfected into COS-7 cells using the DEAE/dextran method as described previously (23). 24 h after transfection, the transfected cells were transferred to 24-well culture plates, seeded at a density of approximately 1 ϫ 10 5 cells/well, and assayed for 125 I-BH-CCK-9 binding or inositol phosphate hydrolysis.
Binding of 125 I-BH-CCK-9 and [ 3 H]L-365,260 to Transfected COS-7 Cells-Twenty-four h after the transfer of transfected cells to 24-well plates, the cells were washed once with cold phosphate-buffered saline, pH 7.4, containing 0.1% bovine serum albumin and incubated in Dulbecco's modified Eagle's medium containing 0.1% bovine serum albumin for 60 min at 37°C with either 50 pM (WT-CCKBR) or 500 pM (H207A-CCKBR) of 125 I-BH-(Thr,Nle)CCK1-9s with and without increasing concentrations of unlabeled peptides. Cell-associated 125 I-BH-(Thr,Nle)CCK1-9s was separated from free radioligand by washing 2ϫ with phosphate-buffered saline containing 2% bovine serum albumin. Cell-associated 125 I-BH-(Thr,Nle)CCK1-9s was collected with 0.5 ml of 0.1 N NaOH added to each well, and radioactivity was detected in a ␥ counter. Nonspecific binding (determined in presence of 1 M CCK-8) was always less than 10% that of total binding. Plasma membranes from transfected COS cells were prepared as described in Kennedy et al. (21) and incubated (50 -100 g) with 500 pM [ 3 H]L-365,260 for 45 mn at 37°C in the presence of increasing L-365,260 concentrations (27). After incubation, the membranes were washed and filtered through Whatman GF/C glass fiber filters over a vacuum-filtering manifold. Filters were counted in a ␤ counter. Binding assays were performed in duplicate in at least three separate experiments. Binding data were determined using the nonlinear, least squares curve-fitting computer program, Ligand (28) 11.1 mM glucose, and 0.5% bovine serum albumin), incubated 1 h at 37°C with PI buffer containing the indicated concentrations of peptides. The reaction was stopped with 1 ml of methanol/HCl added to each well, and the content was transferred to a AG 1-X8 (formate form) column (Bio-Rad). Each column was washed twice with 3 ml of water followed by 2 ml of 5 mM sodium tetraborate, 60 mM sodium formate. Total inositol phosphates were eluted from the column with 2 ml of 1 M ammonium formate, 100 mM formic acid. myo-[ 3 H]inositol phosphate ␤ radioactivity was detected in a liquid scintillation counter (Packard Instrument Co.). EC 50 was calculated using Graph-Pad Prism program software.

Effect of the H2O7A Mutation on Binding and Biological
Activity of the CCKBR-In a recent work, we have shown (24) that the mutation of His 207 in the CCKBR to phenylalanine (the equivalent amino acid in the CCKAR) resulted in a marked decrease in CCK affinity. Upon transient expression of the H207F-CCKBR mutant into COS cells, there was no detectable binding of radiolabeled CCK, and there was a 3,044-fold reduction in CCK-stimulated inositol phosphate production compared with the WT-CCKBR. All together these results suggested that His 207 was crucial for conferring high affinity to CCK (24). The present study was undertaken to identify the amino acid residue within CCK that interacts with His 207 .
We first tried a potentially less drastic exchange by mutating His 207 to Ala. This mutation resulted in undetectable binding of 125 I-BH-(Thr,Nle)CCK1-9s to the H207A-CCKBR mutant transiently expressed in COS-7 cells using standard radioligand concentration (50 pM). By contrast, significant 125 I-BH-(Thr,Nle)CCK1-9s binding was observed in the presence of 500 pM radioligand. Under these conditions, Scatchard analysis of CCK octapeptide (CCK2-9s) competition binding to the H207A-CCKBR mutant revealed a single class of binding sites with an affinity of 232 (Ϯ12) nM and a maximal binding capacity of 4.35 (Ϯ1.52) pmol/10 6 cells. Compared with the WT-CCKBR, which had an affinity of 0.51 (Ϯ0.15) nM and a maximal binding capacity of 2.13 (Ϯ1.62) pmol/10 6 cells, the H207A-CCKBR mutant displayed a decrease in CCK2-9s affinity of 456-fold despite a similar level of cell surface expression. This mutant also retained similar efficacy compared with WT-CCKBR (11fold over basal) for CCK2-9s-stimulated increase in total IP production, although the EC 50 was 302-fold higher than the WT-CCKBR (EC 50 : 145 (Ϯ79) nM versus 0.48 (Ϯ0.3) nM). The decrease in CCK2-9s potency to stimulate IP production correlates well with the decrease in CCK2-9s affinity observed for the H207A-CCKBR mutant (302-fold versus 456-fold). On the basis of the marked shift observed in CCK2-9s affinity and potency when His 207 was mutated to Ala, we hypothesized that His 207 might interact with a crucial amino acid of CCK.

Structure-Function Studies of CCK in COS Cells
Expressing the WT-CCKBR-Since no previous CCK structure-function studies have been reported for the CCKBR in any expression systems, we first determined to what extent each amino acid of CCK contributes to its affinity and potency for stimulation of total IP in the WT-CCKBR transiently expressed in COS-7 cells. For that, we tested different fragments of CCK and synthesized CCK octapeptides modified by replacing each single amino acid by an L-alanine. We expected similar affinity and potency decreases resulting from alanine substitutions in the receptor and in the ligand. In this first screening, the L-alaninemodified peptides were synthesized in a nonsulfated form for the convenience of the synthesis. As shown in Table I, the N-terminal extension of CCK has a small effect on the affinity and activity of CCK peptides. CCK1-9s and CCK3-9s have similar affinities and potencies for the WT-CCKBR than CCK2-9s, indicating that Arg 1 and Asp 2 do not contribute to the activity and affinity of CCK. The absence of sulfate on the Tyr 3 decreases by 5.6-fold the affinity and by 4-fold the potency of CCK2-9, whereas the complete exchange of the sulfated Tyr 3 by an Ala induces a 28-fold and 16-fold decrease in affinity and potency, respectively. These data are close to that found with the C-terminal tetrapeptide (CCK6-9), which displays a 16-fold and 12-fold decrease in affinity and potency, respectively, compared with CCK2-9s. In addition, the exchange of Thr 4 and Gly 5 to Ala had no further additional effect on affinity and potency, indicating that only the presence of the sulfated Tyr 3 slightly increases the affinity and potency of CCK peptides compared with CCK6-9. Together these data suggest that the important shift observed in both affinity and potency of CCK2-9s for the H207A-CCKBR mutant could not be due to an interaction of His 207 with one of the N-terminal amino acids, 1 to 5, of CCK. We then tested the contribution of the last four C-terminal amino acids. As shown in Table I, the exchange of Trp 6 by Ala induces a dramatic decrease in both the affinity and potency. It was difficult to determine an accurate K i be-cause [Ala 6 ]CCK2-9 bound the WT-CCKBR with a very low affinity. Consistent with this result, the EC 50 of [Ala 6 ]CCK2-9 was reduced 13,975-fold. These results are not in favor of an interaction of Trp 6 with His 207 , because Trp 6 substitution results in a large decrease in the affinity and potency of the modified peptide for the WT-CCKBR, which does not correlate with that of CCK2-9s for the H207A-CCKBR mutant. The affinities of [Ala 7 ]CCK2-9 and [Ala 9 ]CCK2-9 were reduced 5,224-and 5,008-fold, respectively, which correlated with the 4,629-and 6,837-fold decrease in potencies observed for these two peptides, whereas the exchange of Asp 8 by Ala induces a less important decrease in the affinity and potency of [Ala 8 ]CCK2-9 with a shift of 1,191-and 935-fold, respectively ( Table I). Considering that these peptides are not sulfated, the decrease in the affinities and potencies of [Ala 7 ]CCK2-9, [Ala 8 ]CCK2-9, and [Ala 9 ]CCK2-9 for the WT-CCKBR are closed to the 456-and 302-fold reduction in CCK2-9s affinity and potency observed for the H207A-CCKBR mutant, suggesting that these residues could interact with His 207 .
Demonstration That His 207 of the CCKBR Interacts with Asp 8 of CCK-We subsequently tested the equivalent sulfated compounds [Ala 7 ]CCK2-9s, [Ala 8 ]CCK2-9s, and [Ala 9 ]CCK2-9s on the WT-CCKBR and H207A-CCKBR to determine if these residues interact with His 207 . We hypothesized that removing the functional group in CCK that interacts with His 207 should not further affect the affinity and potency of the modified peptide for the H207A-CCKBR mutant compared with CCK2-9s, because the mutation should have already disrupted the interaction. On the contrary, exchanging a functional group that does not interact with His 207 should induce an additional decrease in the affinity and potency of the modified peptide for the H207A-CCKBR mutant. As shown in Table II, [Ala 7 ]CCK2-9s displayed a 321-and 268-fold lower affinity and potency for the H207A-CCKBR mutant compared with CCK2-9s. The decrease in the affinity and potency of [Ala 7 ]CCK2-9s for the WT-CCKBR was in the same range, 686-and 650-fold, respectively. No inhibition of 125 I-BH-(Thr,Nle)CCK1-9s binding and no production of IP at a concentration of 10 Ϫ4 M were observed with [Ala 9 ]CCK2-9s on the H207A-CCKBR mutant, indicating a dramatic shift in the affinity and potency. This peptide displayed a 4897-and 5025-fold lower affinity and potency than CCK2-9s on the WT-CCKBR. Together, these results indicate that Met 7 and Phe 9 modifications of CCK and H207A mutation effects are additive and support that Met 7 and Phe 9 do not TABLE I Affinities and potencies of CCK peptides, CCK fragments, and L-alanine-modified CCK peptides for the WT-CCKBR transiently expressed in COS cells K i values were calculated from competition curves of 125 I-BH-(Thr,Nle)CCK1-9s binding by the indicated peptides. EC 50 values were calculated from dose-response curves of total IP production stimulated by the indicated peptides. Results are expressed as mean Ϯ S.E. of three to five separate experiments from different batches of transfected cells. The factor F CCK2-9s represents the effect of the mutation on the affinity or on the potency of the peptides tested relative to CCK2-9s. The efficacies of the peptides tested to stimulate total inositol phosphate production are expressed as the percent of the maximal increase obtained using 1 ϫ 10 Ϫ6 M CCK2-9s and referred as E max . ND, not determined.

Binding
Inositol phosphate production interact with His 207 . In contrast to that observed with [Ala 7 ]CCK2-9s and [Ala 9 ]CCK2-9s, the affinity and potency of [Ala 8 ]CCK2-9s for the H207A-CCKBR mutant were similar to that CCK2-9s but were decreased 460-and 151-fold, respectively, for the WT-CCKBR (Fig. 1, right, and  Fig. 1 and Table II). To confirm this interaction, these data were analyzed according to the mutant cycle methodology of Hidalgo and MacKinnon (29) and Schreiber and Fersht (30). ⍀ values, defined as ⍀ ϭ K i (WT:WT) ϫ K i (mut:mut)/K i (WT: mut) ϫ K i (mut:WT), determined if two modified or mutated residues are independent of one another (⍀ will be unity) or if these residues interact (⍀ will deviate from unity). As shown in Table II We subsequently tested the effect of exchanging Asp 8 of CCK2-9s with other amino acid residues on the ability of the modified peptide to bind and stimulate total IP production to both wild type and mutated receptor (Table III). For the WT-CCKBR, the exchange of Asp 8 to Glu reduced the affinity and the potency of [Glu 8 ]CCK2-9s, 199-and 148-fold, respectively. These values are similar to that obtained when Asp 8 was exchanged to Ala, indicating that the length of the side chain at this position is as crucial as the presence of the carboxylic group. Similarly, the exchange of Asp 8 by Phe in CCK2-9s reduced the affinity and the potency of the modified peptide [Phe 8 ]CCK2-9s to the same extent (Table III). CCK peptide with an Asp 8 to Leu substitution ([Leu 8 ]CCK2-9s) bound to the WT-CCKBR with an affinity lower than the other peptides modified at position 8. Moreover, this peptide did not stimulate IP production, even at a concentration of 10 Ϫ4 M, indicating that the substitution of Asp 8 by Leu affects the agonist activity of the peptide. This result is in accordance with the importance of the length of the side chain at this position. All these modified peptides had similar affinities and potencies compared with CCK2-9s for the H207A-CCKBR mutant (except [Leu 8 ]CCK2-9s, which did not stimulate IP even at 10 Ϫ4 M as with the WT-CCKBR), indicating that unlike the WT-CCKBR, the H207A-CCKBR mutant is insensitive to the changes introduced at position 8 of CCK (Table III). It should be noted that all these modified peptides had a 1.6-to 2-fold lower efficacy for stimulating total IP production in the H207A-CCKBR mutant compared with the WT-CCKBR. This is most likely due to the additive effects of the changes introduced both in the mutated receptor and in the modified peptides, whereas CCK2-9s had the same efficacy on both wild type and mutated receptor.

FIG. 1. Competition of 125 I-BH-(Thr,Nle)CCK1-9s binding (left) and stimulation of total IP production (right) by CCK2-9s (OE) and [Ala 8 ]CCK2-9s (f) on the WT-CCKBR (broken line) and H207A-CCKBR (solid line) mutant transiently expressed in COS cells.
Binding is expressed as the percent of specifically bound 125 I-BH-(Thr,Nle)CCK1-9s. Inositol phosphate production is expressed as percent of maximum stimulation. The data represent the mean of three to five experiments performed in duplicate.

TABLE II
Affinities and potencies of CCK2-9s and modified CCK2-9s peptides for the WT-CCKBR and H207A-CCKBR mutant transiently expressed in COS cells K i values were calculated from competition curves of 125 I-BH-(Thr,Nle)CCK1-9s binding by the indicated peptides. EC 50 values were calculated from total IP production dose-response curves stimulated by the indicated peptides. Results are expressed as mean Ϯ S.E. of three to five separate experiments from different batches of transfected cells. The factor F CCK2-9s represents the effect of the mutation on the affinity or on the potency of the peptides tested relative to CCK2-9s. The efficacies of the peptides tested to stimulate total inositol phosphate production are expressed as the percent of the maximal increase obtained using 1 ϫ 10 Ϫ6 M CCK2-9s on the WT-CCKBR and referred to as E max . The F WT factor represents the effect of the mutation on the affinity or on the potency of the peptides tested relative to the wild type receptor. ND, not determined. The ⍀ factor was calculated according to the equation described under "Results." For values less than unity, reciprocals are given for ease of comparison.  (22), we tested a CCK peptide modified at position 3, [Ala 3 ]CCK2-9, to further demonstrate that the effect of substituting His 207 on CCK affinity and potency was specific for Asp 8 . The effect of mutating His 207 and substituting Tyr 3 of CCK was additive. [Ala 3 ]CCK2-9 displayed an affinity and a potency for the H207A-CCKBR mutant that were further reduced 11-and 13-fold compared with CCK2-9s. Moreover, the affinity and potency of [Ala 3 ]CCK2-9 was reduced 100-and 246-fold compared with the WT-CCKBR (Table III).
To further demonstrate the interaction of Asp 8 of CCK with His 207 in the CCKBR, we mutated His 207 to Asp. If Asp 8 interacts with the residue 207 in the receptor, the introduction of a negative charge at position 207 should further decrease CCK2-9s affinity and potency compared with the H207A mutant receptor because of the repulsive effect between the two negatively charged residues. After transient expression of the H207D-CCKBR mutant in COS cells, 125 I-BH-(Thr,Nle)-CCK1-9s binding was undetectable even with 500 pM radioligand. However, Scatchard analysis of the CCKBR nonpeptide antagonist, [ 3 H]L365,260, binding on transfected COS-7 membranes revealed a similar level of expression of the H207D-CCKBR mutant (B max ϭ 3.9 Ϯ 1.3 fmol/g of protein) compared with WT-CCKBR (B max ϭ 2.4 Ϯ 1.5 fmol/g of protein). These data suggest that the loss of 125 I-BH-(Thr,Nle)CCK1-9s binding was caused by an important decrease in CCK affinity and was not due to a loss in receptor expression. Consistent with the important shift in CCK affinity observed, the potency of CCK2-9s for the H207D-CCKBR mutant was further decreased 112-fold (maximal efficacy of 67%) compared with the H207A-CCKBR mutant (Table IV and Fig. 2). Moreover, the [Glu 8 ]CCK2-9s peptide weakly stimulated total IP production at 10 Ϫ4 M (around 9%) at the H207D-CCKBR mutant, whereas [Glu 8 ]CCK2-9s maintained similar affinity and potency on the H207A-CCKBR mutant than CCK2-9s (Table IV). When [Ala 8 ]CCK2-9s was tested for its ability to stimulate total IP production through the H207D-CCKBR mutant, we observed a 12-fold increase in [Ala 8 ]CCK2-9s potency relative to CCK2-9s (Table IV and Fig. 2). A similar increase (27-fold) in potency was also observed with [Phe 8 ]CCK2-9s (Table IV). Thus, a gain of function was observed when the repulsive force between Asp 207 and Asp 8 was removed by replacing Asp 8 by a neutral amino acid. DISCUSSION Use of mutational studies to identify receptor amino acids important for ligand binding cannot by themselves determine whether the mutated amino acids have a direct versus an indirect effect. The effect of a mutation may directly influence the binding pocket by substituting a residue that directly binds the ligand. Alternatively, the mutation may indirectly affect the binding pocket by causing a conformational change that disrupts a distal binding site. One way to address this question is to characterize the functional group in the ligand that interacts with the mutated receptor amino acid by using modified ligands. We have recently shown by mutational analysis of the CCKBR that one residue, His 207 , located in the second extracellular loop of the CCKBR when mutated, affects CCK2-9s

TABLE IV
Potencies of CCK2-9s and modified CCK2-9s peptides for the H207A-CCKBR and H207D-CCKBR mutants transiently expressed in COS cells EC 50 values were calculated from total IP production dose-response curves stimulated by the indicated peptides. Results are expressed as mean Ϯ S.E. of three to five separate experiments from different batches of transfected cells. The factor F CCK2-9s represents the effect of the mutation on the affinity or on the potency of the peptides tested relative to CCK2-9s. The efficacies of the indicated peptides to stimulate total inositol phosphate production are expressed as the percent of the maximal increase obtained using 1 ϫ 10 Ϫ6 M CCK2-9s on the WT-CCKBR and referred to as E max . ND, not determined.  III Affinities and potencies of CCK2-9s and modified CCK2-9s peptides for the WT-CCKBR and H207A-CCKBR mutant transiently expressed in COS cells K i values were calculated from competition curves of 125 I-BH-(Thr,Nle)CCK1-9s binding by the indicated peptides. EC 50 values were calculated from total IP production dose-response curves stimulated by the indicated peptides. Results are expressed as mean Ϯ S.E. of three to five separate experiments from different batches of transfected cells. The factor F CCK2-9s represents the effect of the mutation on the affinity or on the potency of the peptides tested relative to CCK2-9s. The efficacies of the indicated peptides to stimulate total inositol phosphate production are expressed as the percent of the maximal increase obtained using 1 ϫ 10 Ϫ6 M CCK2-9s on the WT-CCKBR and referred to as E max . ND, not determined.

WT-CCKBR
H207A-CCKBR WT-CCKBR H207A-CCKBR high affinity binding (24). Regarding the direct role of the second extracellular loop of the CCKAR for the high affinity binding of CCK (22), the aim of the present study was to characterize the functional group in CCK that interacts with His 207 in the CCKBR and thus to determine if His 207 is likely to be directly involved in the CCK binding pocket. We first showed that the mutation of His 207 to Ala decreased 456-fold the affinity of CCK2-9s, and 302-fold, its potency to stimulate total IP production compared with the wild type CCKB receptor. The similar expression of the mutated receptor and wild type receptor at the cell surface as well as the similar biological efficacy of CCK2-9s for the two receptors indicated that the mutation did not introduce a gross conformational change in the receptor. The mutation of His 207 to Ala, albeit less drastic than to Phe, still affected the affinity and potency of CCK2-9s, suggesting that this residue might interact with an important amino acid of CCK. In a preliminary approach to identifying the putative amino acid of CCK interacting with His 207 , we first screened L-alanine-substituted analogues of CCK octapeptide at each position for their ability to bind and stimulate IP production at the wild type receptor. We expected similar affinity and potency decreases resulting from Ala substitutions in the receptor and in the ligand. This strategy permitted us to identify three residues at the C terminus of CCK, Met 7 , Asp 8 , and Phe 9 , as being putative partners of His 207 , since the affinities and potencies of the corresponding modified peptides for the WT-CCKBR were close to that of CCK2-9s for the H207A-CCKBR mutant. Among these residues, Asp 8 was identified as the only residue interacting with His 207 , since the H207A-CCKBR mutant displayed similar affinities and potencies for [Ala 8 ]CCK2-9s and CCK2-9s, whereas its affinity and potency for [Ala 7 ]CCK2-9s and [Phe 9 ]CCK2-9s were importantly decreased. The interaction between Asp 8 and His 207 was confirmed by the fact that the H207A-CCKBR mutant was unable to distinguish between the native ligand and a series of analogues modified at position 8, whereas the WT-CCKBR was highly sensitive to changes introduced at this position. Moreover the peptides modified at position 8 displayed similar affinities and potencies for the wild type and mutated receptor, whereas the analogues modified at positions 7 and 9 bound to and activated H207A-CCKBR mutant with decreased affinities and potencies relative to the wild type receptor.
Further support for an interaction of Asp 8 with His 207 was brought by mutating His 207 to Asp. The drastic effect observed on CCK2-9s affinity and potency when His 207 was substituted to Asp compared with Ala (100-fold) is consistent with an electrostatic repulsion between the negative charges of the Asp 8 of the ligand and the one introduced at position 207 in the receptor. In addition, the severe loss of affinity and potency of [Glu 8 ]CCK2-9s for the H207D-CCKBR mutant is in accordance with an increase of the repulsive effect due to an increase of the length of the side chain while maintaining the carboxylic function in position 8. In agreement with this result, when the repulsive force was removed by substituting Asp 8 of the ligand for a neutral amino acid such as Ala or Phe, the potency of the two modified peptides for the mutant H207D were increased accordingly 12-and 27-fold compared with CCK2-9s. This gain of function is a strong evidence for the direct interaction between His 207 of the CCKBR and Asp 8 of CCK. Moreover, since the affinity and potency of the two peptides [Ala 8 ]CCK2-9s and [Phe 8 ]CCK2-9s for the H207A and H207D-CCKBR mutants were similar, it is likely that the repulsive effect observed was caused by the direct repulsion of the negative charges of the aspartic residues of the ligand and the receptor and is not due to a nonspecific interaction of Asp 207 with another residue of the receptor. All together these results argue that the changes introduced in the mutated receptor and in the ligand affect the same bond.
However, the present work does not allow one to conclude about the nature of this interaction. Imidazole-carboxylate interactions such as the one found here between the carboxyl side chain of CCK and His 207 of the CCKBR are often found as an important stabilizing element in protein structures (31)(32)(33), enzyme reactions (34 -36), and ligand-receptor recognition (37,38). Such interactions can be due to a salt bridge between imidazolium and carboxylate ions or to a network of hydrogen bonds linking the imidazole and carboxylate side chain together. Such an interaction might be considered as the initialization step of a proton pumping mechanism, which could be hypothetized in the activation process of G protein-coupled receptors (37). The implication of Asp 8 of CCK in CCKBR activation, in addition to its role in ligand binding, was previously reported for the stimulation of acid secretion from dispersed gastric cells (39). Consistent with this result, we observed in the present paper that the substitution of Asp 8 for Leu completely abolished total inositol phosphate production both in the wild type CCKBR and in the H207A-CCKBR mutant while maintaining micromolar affinities for both receptors.
Direct interactions of peptides with G protein-coupled receptors have been documented in a limited number of cases. However, it appears that the direct involvement of the extracellular domains in the peptide binding pocket is not unique to the CCKA and CCKB receptors subtypes. A few reports have indicated that the extracellular domains of the peptide receptors are direct contact sites for the peptide ligands. For example, it has been reported that in the AT1 angiotensin receptor, Asp 281 of the extracellular loop 3 and His 183 of the extracellular loop 2, respectively, bind Arg 2 and Asp 1 of angiotensin II (19). Similarly, Tyr 115 of the extracellular loop 1 of the vasopressin V1A receptor interacts with Arg 8 of arginine-vasopressin (40).
Despite the fact that CCK presents a similar high affinity for both the CCKA and the CCKB receptor subtypes, the mutagenesis data obtained in this study clearly indicate a different anchoring of CCK in the CCKBR compared with the CCKAR (21,22). These results are also consistent with a recent study in which we showed that the reciprocal mutation in the CCKBR of the two residues corresponding to Trp 39 and Gln 40 in the CCKAR (21) that directly interact with the N terminus of CCK2-9s have no effect on CCK2-9s binding in the CCKB receptor (24).
In conclusion, using site-directed mutagenesis of the CCKB receptor together with the analysis of the binding and biological potency of different modified CCK analogues, we have shown in this study that His 207 is directly involved in the binding pocket of CCK by interacting with the penultimate Asp 8 residue of CCK. This study represents an important step in the molecular characterization of the CCK binding site especially since this is the first report concerning the identification of a direct contact point between the CCKB receptor and CCK. The importance of this study is reinforced by the fact that it involves an important amino acid of CCK.