Introduction

Top Summary Introduction Materials & Methods Results Discussion References

The effects of CCK₈, the carboxyl-terminal octapeptide fragment of cholecystokinin, are mediated by two distinct receptors: the CCK_A receptors, mainly found in the gastrointestinal system, and the CCK_B receptors, predominant in the central nervous system (see review in Wank, 1995). Both receptors, cloned from various species, belong to the GPCR superfamily, characterized by seven membrane-spanning segments and a extracellular NH₂-terminal domain. On binding of CCK₈, the CCK_B receptor activates PLC, triggering an increase in IP production (Jagerschmidt et al., 1995; Wank, 1995). The brain CCK_B receptor is involved in several important adaptational processes. Thus, selective CCK_B antagonists were shown capable to block the panic attacks induced by administration of CCK₄ (Bradwejn et al., 1992) and to be endowed with antidepressant-like properties in rodents (Derrien et al., 1994). Furthermore, due to the negative feedback control achieved by CCK₈ via CCK_B receptor activation on the opioid system, the CCK_B antagonists were shown to produce a strong potentiation of the antinociceptive effects induced by opioids (Maldonado et al., 1993). On the other hand, potent and selective CCK_B receptor agonists such as BC 264 (Charpentier et al., 1988) were shown to improve vigilance and memory processes (see review in Daugé and Roques, 1995). All these data account for the great interest devoted to the design of selective CCK_B ligands.

Although numerous results support the localization of binding sites for small amine ligands (e.g., histamine, 5-hydroxytryptamine, dopamine) within the TM domains of GPCRs (Schwartz and Rosenkilde, 1996), several peptide ligands were reported to interact with both TM and extracellular domains of these receptors. Thus, for instance, it has been shown by point mutagenesis and chimeric studies that opioid receptors interact with their ligands at multiple sites, both extracellular and intramembranous (Befort et al., 1996; Pepin et al., 1997). Likewise, numerous studies have provided evidence for an important role of residues in the amino-terminal domain, the extracellular loops, and the transmembrane helices for the ability of several receptors to bind their ligands, such as neurokinin receptor (Fong et al., 1992), V_1a vasopressin receptor (Chini et al., 1995), oxytocin receptor (Chini et al., 1996), somatostatin receptor (Kaupman et al., 1995), and vasoactive intestinal peptide receptor (Du et al., 1997).

Few studies have been devoted to the CCK_B receptor. Using site-directed mutagenesis, eight residues located in the TM helices of the human CCK_B receptor (Kopin et al., 1995) and His381 located in the TM-VII of the rat CCK_B receptor (Jagerschmidt et al., 1996) have been suggested to be involved in CCK_B versus CCK_A antagonist selectivity. In addition, a segment of five amino acids in the second extracellular loop of the CCK_B receptor was shown to be essential for the high affinity of the natural peptide agonist gastrin, suggesting that determinants of the binding site of the CCK_B receptor are situated, at least partially, within the extracellular domains (Silvente-Poirot and Wank, 1996). This result has been confirmed in the case of the CCK_A receptor, in which two amino acids in the amino-terminal region have been identified as critical components of the agonist binding site (Kennedy et al., 1997). Moreover, we recently reported that the Asp100 residue, located within the second TM segment of the rat CCK_B receptor and highly conserved among the GPCR superfamily, was involved in agonist-induced IP production (Jagerschmidt et al., 1995). All these findings were in good agreement with docking experiments performed with a three-dimensional model of the CCK_B receptor (Jagerschmidt et al., 1995, 1996).

On the other hand, several aromatic residues were shown to be highly conserved within the transmembrane domains of GPCRs, in particular, in helices IV, V, and VI (Probst et al., 1992; Underwood et al., 1994; Befort et al., 1996) (Table 1). These conserved aromatic residues were proposed to play an important role in the spatial organization of the binding site (Underwood et al., 1994). Moreover, these amino acids could be involved in the signal transduction mechanism occurring after agonist-induced receptor activation as described in the case of the neurokinin type 1 receptor (Huang et al., 1994), the 5-hydroxytryptamine₂ receptor (Choudhary et al., 1993), and the angiotensin II type AT₁ receptor (Marie et al., 1994).

View this table: [in this window] [in a new window]   TABLE 1 Sequence comparison of TM-V and TM-VI helices within various GPCRs The predicted TM helices of the rat CCKB receptor are compared with the corresponding sequences of other representative members of the GPCRs. Numbers above the sequences refer to amino acid positions within the rat CCKB receptor (NK-1, neurokinin-1 receptor; BK, bradykinin receptor; AT-1, angiotensin receptor type 1; 5HT2, 5-hydroxytryptamine receptor type 2; M1, muscarinic receptor type 1; 2-AR, 2 adrenergic receptor). The single-letter amino acid code is used. The highly conserved aromatic residues are indicated in bold.

To gain further insights into the mechanism of ligand recognition and transduction process for the CCK_B receptor, we investigated the role of the amino-terminal region and of three conserved aromatic residues located on TM-V (Phe227) and TM-VI (Phe347 and Trp351) helices of the rat CCK_B receptor (Table 1 and Fig 1). Using site-directed mutagenesis, the aromatic residues were replaced by alanine, and a amino-terminally truncated (1,53) CCK_B receptor was constructed. The WT, the truncated, and the F227A, F347A, and W351A mutated rat CCK_B receptors were transiently expressed in Cos-7 cells, and radioligand binding experiments and IP assays were performed. This enabled us to demonstrate that Trp351 interacts with the CCK agonists and antagonists, whereas Phe227 and Phe347 are more important for CCK antagonists binding. Moreover, the amino-terminal region of the CCK_B receptor seems not to be involved in CCK ligands recognition. On the other hand, at least one residue (Phe347) localized in the sixth TM domain was identified as essential for the signal transduction process.

View larger version (51K): [in this window] [in a new window]   Fig. 1.   Schematic representation of the rat CCKB receptor showing the postulated transmembrane topology. The phenylalanine (F227, F347) and tryptophan (W351) residues are highlighted; these residues were individually converted to alanine, as described in Experimental Procedures. Boxed, seven TM domains.

	Materials and Methods

Top Summary Introduction Materials & Methods Results Discussion References

Reagents. CCK₄ was purchased from Bachem (Buhendrof, Switzerland). CCK₈, pBC 264, L-365,260, L-364,718, and PD-134,308 (Fig. 2) were synthesized in the laboratory according to reported procedures (Evans et al., 1986; Charpentier et al., 1988; Lotti and Chang, 1989; Hughes et al., 1990). Radiolabeled compounds, [-³³P]dATP (specific activity, 1000-3000 Ci/mmol), [³H]pCCK₈ (specific activity, 60-90 Ci/mmol), and myo-[2-³H]inositol (specific activity, 60-90 Ci/mmol) were purchased from Amersham (Les Ulis, France). [³H]pBC 264 (specific activity, 20-30 Ci/mmol) was from NEN (Les Ulis, France). Cell culture reagents were from GIBCO-BRL (Cergy, France), and the AG1-X8 Dowex resin was from BioRad (Ivry/Seine, France).

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Structures of CCK agonists and antagonists.

Site-directed mutagenesis. The cDNA of the rat CCK_B receptor was obtained as described previously (Jagerschmidt et al., 1995). For construction of the amino-terminally truncated (1-53) CCK_B receptor, the PvuII site at residue 53 was used for restriction endonuclease cleavage. Three oligonucleotides were designed to replace the codon for phenylalanine (TTC) located at amino acid positions 227 and 347 with a codon for alanine (5'-CTGCTTTTGGCCTTCATCCCG-3' and 5'-GTTTTGCTTGCCTTCCTGTGT-3', respectively) and to replace the codon for tryptophan (TGG) located at amino acid position 351 with a codon for alanine (5'-TTCCTGTGTGCGCTGCCAGTG-3'). Double-strand mutagenesis was carried out as described previously (Jagerschmidt et al., 1995). Authenticity of the mutations was confirmed by sequencing the constructions over the entire protein-coding region with the Sequenase Version 2.0 DNA sequencing kit (United States Biochemical, Cleveland, OH) and [-³³P]dATP.

Radioligand binding assays. Cos-7 cells, which do not express CCK_B receptor, were grown onto 24-well plates and transfected, using the calcium phosphate precipitation method, with 0.5 µg of plasmid DNA/10⁵ cells as described previously (Jagerschmidt et al., 1996). Then, 8 µg of plasmid containing either the WT or truncated receptor cDNA also were transfected into Cos-7 (10⁵ cells) using the calcium phosphate precipitation method as described by Graham and van der Erb (1973). At 48 hr after the transfections, the binding assays were performed directly on cells in Dulbecco's modified Eagle's medium containing 5 mM MgCl₂ and 0.2 mg/ml bacitracin. Each assay (90 min, 25°) was performed in a final volume of 0.5 ml. For saturation binding experiments, the concentration of the [³H]pCCK₈ or [³H]pBC 264 varied from 50 to 6000 pM. For competition experiments, a fixed concentration of 500 pM of the radioligand [³H]pCCK₈ or [³H]pBC 264 were used in the presence of various concentrations of the competitor. Nonspecific binding was determined by using 1 µM CCK₈. Incubations were stopped by removing the media. The cells were harvested and the radioactivity was counted as described previously (Jagerschmidt et al., 1995). Parameters describing [³H]pCCK₈ saturation binding (i.e., K_d and B_max) were determined using the computer program EBDA. K_i values were determined by using the Cheng-Prussof equation: K_i = IC₅₀/[1 + (radioligand concentration/K_d value of the radioligand)].

IP assays. Cos-7 cells transiently expressing WT and mutated CCK_B receptors were assayed for CCK₈-stimulated PI hydrolysis, as described previously (Jagerschmidt et al., 1995). Briefly, transfected cells were grown in the presence of 1 µCi/ml myo-[2-³H]inositol for 16 hr at 37°. Cells then were treated with 10 mM LiCl for 30 min at 37° and various concentrations of CCK₈ were added to the cells. After 45 min at 37°, the incubation medium was removed, and the cells were washed twice with 1 ml of PBS. The reaction was stopped by adding 400 µl of ice-cold 75% methanol and 300 µl of 0.12% Triton. The cells were scraped, and the suspension was subjected to chloroform extraction. Then, 0.5 ml of the aqueous phase was added to 4.5 ml of H₂O. The solution was loaded onto a 0.5-ml column containing AG1-X8 Dowex anion exchange resin. The column was washed with 1 ml of distilled water followed by 5 ml of 5 mM sodium borate/60 mM sodium formate. Total [³H]IP then were eluted into scintillation vials with 5 ml of 1 M ammonium formate/0.1 M formic acid. After the addition of scintillation mixture, radioactivity was counted.

Statistical analysis. The results presented are the mean ± standard error. Data were analyzed using one-way analysis of variance and paired Student's t tests. Differences were considered statistically significant at p < 0.05.

	Results

Top Summary Introduction Materials & Methods Results Discussion References

Identification of aromatic residues of the rat CCK_B receptor important for the binding of CCK ligands. The WT and mutant CCK_B receptors were expressed transiently in Cos-7 cells. To determine whether the mutated receptors were expressed correctly by the cells and displayed CCK affinities comparable to that of the WT receptor, saturation studies were first performed. Except for cells expressing the W351A mutated receptor, the binding of [³H]pCCK₈ to transfected cells was specific and saturable. In the case of the W351A mutant receptor, the binding experiments performed with [³H]pCCK₈ were difficult to interpret due to high nonspecific binding. Moreover, the absence of saturation at a concentration of 6 nM of this radioligand suggested a decrease in its affinity (data not shown). To overcome this problem, we used the CCK₈-derived peptidomimetic [³H]pBC264, a selective CCK_B receptor agonist endowed with both a higher affinity and a lower nonspecific binding than [³H]pCCK₈ (Charpentier et al., 1988; Durieux et al., 1992). With this radioligand, specific and saturable binding was observed for both WT and W351A receptors. [³H]pBC264 was then chosen for further characterization of the W351A mutant receptor. On the other hand, in the experimental conditions used (0.5 µg of DNA transfected), the B_max value for the amino-terminal truncated CCK_B receptor (1,53) was strongly decreased. Thus, to facilitate the determination of the binding characteristics of this mutant, the amount of DNA used in the transfection was increased (8 µg).

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Competition of CCK ligands toward the binding of [3H]pBC264 to WT and W351A rat CCKB receptors. Cos-7 cells transiently expressing WT () or W351A mutant () receptors were tested for competition experiments using pBC 264, CCK4 and PD-134,308 as described in Experimental Procedures. Each point is the average of triplicate determinations. Data, which are representative of three or four separate experiments, are depicted with standard error values in Table 4.

Functionality of the mutant receptors. Rat CCK_B receptors expressed in Cos-7 cells display agonist-mediated dose-dependent increases in PI hydrolysis (Jagerschmidt et al., 1995; Wank, 1995). To determine the functionality of each of the mutant receptors, the CCK₈-induced accumulation of IP in transfected Cos-7 cells was measured.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   CCK8 stimulation of IP formation in Cos-7 cells expressing WT (), F227A (), F347A (), and W351A () CCKB receptors. Cells were transfected using 8 µg of DNA/105 cells, and assays were performed as described in Experimental Procedures. Data points represent mean values of triplicate determinations in three experiments. There was no CCK8-stimulated IP formation in nontransfected Cos-7 cells. For comparison, the maximum amount of [3H]IP accumulation induced by CCK8 activation of WT CCKB receptor, in a typical experiment, was 1700 cpm, with a basal accumulation of 250 cpm.

View larger version (19K): [in this window] [in a new window]   Fig. 5.   CCK8 stimulation of IP formation in Cos-7 cells expressing WT () and mutated () CCKB receptors. After transfection, Cos-7 cells were stimulated using 0.1 µM CCK8, and assays were performed as described in Experimental Procedures. Histograms, mean values of triplicate determinations in three experiments for cells expressing 100-200 (A) and 700-800 (B) fmol receptor/106 cells. Hatched bars, CCK8 stimulation of IP formation in nontransfected Cos-7 cells.

	Discussion

Top Summary Introduction Materials & Methods Results Discussion References

Numerous studies have examined peptide ligand binding domains of GPCRs showing a considerable diversity in both the number and location of ligand interacting determinants. Several aromatic residues located in the TM helices are conserved among most of the GPCRs (Probst et al., 1992). This is the case for the Ar-X₂-Pro-X₇-Ar motif (where Ar is aromatic and X is any residue), which is very often found in the TM-V domain, and for Ar-X₃-Ar-X-Pro, which is found in the TM-VI helix of most of the GPCRs (Underwood et al., 1994). These motifs also were found in the CCK_B receptor, and the high conservation of these constituting amino acids (Table 1) suggests that they could have a critical role in the structural organization and/or in the functioning of GPCRs. Furthermore, the importance of the external amino-terminal region of several GPCRs for the binding of peptide ligands has been reported (Hjorth et al., 1994; Hong et al., 1994). Thus, Miyake (1995) has shown that a splice variant of the CCK_B receptor, in which the amino-terminal extracellular domain and almost all the residues of the first transmembrane helix were absent, displayed altered binding and functional properties. Therefore, in the current study, we evaluated the involvement in ligand binding and signal transduction of three aromatic residues belonging to the highly conserved domains (Phe227 in TM-V, Phe347 and Trp351 in TM-VI) and of the external amino-terminal region of the CCK_B receptor (Fig. 1).

Exchange of Phe227 for alanine produced only a slight decrease in the affinity of the selective CCK_A antagonist L-364,718, without effects on CCK₈ and CCK_B antagonists affinities and second messenger production. A recent study on the human NK₁ receptor demonstrated that the exchange of a tyrosine localized in the same TM-V domain (Tyr205) (Table 1) for alanine resulted in a decrease of substance P affinity, associated with a similar loss in the biological potency illustrated by a parallel reduction in IP production (Huang et al., 1994). Together, these results emphasize the diversity of GPCRs binding sites.

The mutation of Phe347 for alanine in the rat CCK_B receptor did not alter the binding of CCK agonists. This result suggests that in the rat CCK_B receptor, the removal of the aromatic ring of residue 347 located in the TM-VI does not affect the binding site for agonists. Furthermore, the mutant rat CCK_B receptor (F347A) exhibits a slightly improved affinity for the CCK_B antagonist L-365,260, whereas affinity for the benzodiazepine-derived antagonist L-364,718 remains unchanged (Table 3). This could be explained by differences in the spatial orientation of the two antagonists within the receptor binding site. Accordingly, in a previous study, we have shown that the two structurally close benzodiazepine-derived compounds L-365,260 and L-364,718 recognize some specific residues into the ligand binding pocket of the CCK_B receptor (Jagerschmidt et al., 1996). Moreover, mutation F347A of the CCK_B receptor, while slightly affecting the binding of CCK antagonists, did not affect the binding of CCK agonists, suggesting that both classes of molecules do not interact with the same region of the receptor. Differences between chemical determinants for agonists and antagonists binding has already been shown for peptide receptors, including the CCK_B receptor (Beinborn et al., 1993; Mantamadiotis and Baldwin, 1994).

Studies on the functionality of the CCK_B receptor showed a different pattern of responses in the case of F347A and F227A mutants. Thus, mutation of Phe347 disrupts PI signaling, indicating a complete loss of the transduction process associated with agonists binding. In contrast, at comparable expression levels, both efficacy and potency values were found to be very close for the WT and F227A receptors, indicating that the Phe227 residue of the rat CCK_B receptor is not involved in the signal transduction mechanism. The loss in pharmacological activity of CCK₈ with the F347A mutant receptor was not accompanied by a similar loss in affinity for CCK₈. Moreover, the observed lack of functional activity cannot be attributed to differences in receptor expression level; for similar amounts of F347A and WT receptors, only the latter induced phosphatidylinositol hydrolysis (Fig. 4). This observation suggests that mutation of Phe347 affects a region involved in the agonist-induced second messenger production without affecting the overall integrity of the receptor. It is interesting to observe that the amino acids involved in G protein activation and located in TM VI and VII domains of GPCRs are very often tyrosine (Wess et al., 1992; Marie et al., 1994), threonine (Wess et al., 1992), or proline (Wess et al., 1993) residues. Nevertheless in the 5-hydroxytryptamine₂ receptor, a phenylalanine residue also has been shown to be implicated in second messenger production (Choudhary et al., 1993). Recently, we have shown that the Asp100 residue of the rat CCK_B receptor is involved in signal transduction. It was hypothesized that Asp100 points in a direction of a cluster of basic amino acids (Lys333/Lys334/Arg335) located in the third intracellular loop of the receptor at the bottom of the TM-VI domain (Jagerschmidt et al., 1995). This was confirmed by results reported by Wang (1997) showing that these three basic amino acids play a critical role in CCK_B receptor activation of G_q proteins. Thus, although not directly involved in the binding of CCK_B ligands as shown by the lack of change in binding affinities after its replacement by alanine, Phe347, which belongs to the TM-VI domain, could be a residue implicated in transduction processes by playing a key role in agonist-induced changes in receptor conformation triggering G protein stimulation. The same explanation could be proposed if we take into account the allosteric model of receptor. In both cases, the exchange of Phe347 by alanine could produce a conformational change in the sequence containing the basic triplet Lys333/Lys334/Arg335, located just beneath TM VI, in the third intracellular loop, known to be involved in G protein coupling (Schwartz and Rosenkilde, 1996).

To further explore which amino acids are critical for ligand binding, we have shown by site-directed mutagenesis that Trp351, another conserved aromatic amino acid, when mutated to alanine, leads to a loss in affinity for CCK₈ and CCK₄ and for the CCK₄-based antagonist, PD-134,308 (6-, 10-, and 10-fold, respectively). These results indicate that the Trp351 residue seems to be involved in the agonist binding site of the rat CCK_B receptor. This observation and the fact that the binding of the pseudopeptide agonist pBC264 and of nonpeptide ligands, such as benzodiazepine-derived antagonists L-365,260 and L-364,718, was unaffected by the mutation, suggest that replacement of Trp351 by alanine directly affects a region of the binding site for CCK on the receptor without affecting the overall integrity of the receptor. These results emphasize previous results obtained with the NK₁ (Fong et al., 1994), CCK_A (Kennedy et al., 1997), and CCK_B (Jagerschmidt et al., 1996) receptors, suggesting that within the receptor binding site, the residues involved in ligand recognition are different, with molecules belonging to different chemical classes, a result also found in -opioid receptors (Befort et al., 1996). A Trp residue, located at an equivalent position in the TM-VI domain of M₃ muscarinic receptor, also was shown to be involved in both agonist and antagonist binding (Wess et al., 1993), demonstrating the critical role of this highly conserved Trp residue. In good agreement with the current experimental data, molecular modeling studies have suggested that this Trp residue, in concert with several other aromatic amino acids, may be directly involved in ligand/receptor recognition (Hibert et al., 1991; Underwood et al., 1994). Furthermore, the mutation of the Trp351 residue in the rat CCK_B receptor, while strongly affecting the binding of CCK₄ and CCK₄-derived ligand (PD-134, 308), affects to a lesser extent the binding of CCK₈, suggesting the existence of specific interaction or interactions between the carboxyl-terminal part of CCK₈ [Asp-Tyr(SO₃H)-Met-Gly] and the CCK_B receptor, stabilizing the ligand/receptor complex. On the other hand, the reduction in affinity for CCK₈ of the mutant W351A receptor resulted in a loss in the biological efficacy of CCK₈ to produce IP. However, as for the WT receptor, the efficacy of CCK₈ to stimulate PI hydrolysis (EC₅₀ = 8.7 nM) was in the same range as its affinity (K_i = 8.14 nM) for the sulfated octapeptide. Moreover, in the absence of stimulation, no IP release was observed, with the W351A mutant showing that this receptor is not constitutively activated.

The importance of the amino-terminal region of the CCK_B receptor was demonstrated by the discovery of a splice variant that was deleted in the amino-terminal extracellular region (1,63) and the upper part of the first transmembrane domain and had different binding properties than the WT CCK_B receptor (Miyake, 1995). Moreover, a segment of five amino acids in the second extracellular loop of the CCK_B receptor was shown to be essential for the high affinity of this receptor for gastrin, suggesting that determinants of the binding site of the CCK_B receptor could be located within the extracellular domains (Silvente-Poirot and Wank, 1996). To further explore this hypothesis, we evaluated the affinity of several CCK ligands for a mutant receptor in which only the external amino-terminal domain has been deleted (1,53). The results obtained show that this truncation did not affect the binding of the CCK ligand, which is in contrast to the 5-fold reduction in affinity observed with the splice variant (1,63) (see above). Thus, it seems that the extracellular amino-terminal region of the CCK_B receptor is not important for ligand/receptor complex formation. In the case of the CCK_A receptor, Kennedy et al. (1997) identified two amino acids located in the amino-terminal region that play a role in agonist binding. The difference between the two CCK receptor types is reflected by the low sequence homology between their amino-terminal domains. The importance of this amino-terminal region has been investigated in several GPCRs, and the results show divergent results. Thus, ligands have been shown to bind to the external amino-terminal domain of several receptors, such as -opioid (Kong et al., 1994) or AT₁ (Hjorth et al., 1994) receptors. In contrast, deletion of the 64 amino-terminal amino acids of the µ-opioid receptor did not affect binding of agonists and antagonists, indicating that the amino-terminal domain does not contribute to ligand binding (Surratt et al., 1994). This also seems to be the case for the CCK_B receptor. Nevertheless, this amino-terminal domain that possesses glycosylation sites seems to be important for the expression of the receptor at the membrane levels, as the truncated mutant shows an expression level 10-fold lower than that of the CCK_B WT receptor.

In conclusion, using site-directed mutagenesis of the CCK_B receptor and analysis of the binding affinity and biological potency of CCK ligands, we have shown that the amino-terminal region is not involved in the formation of the ligand/receptor complex. Moreover, we demonstrated that the highly conserved aromatic amino acid residues in GPCRs, Phe227 and Phe347, do not play an important role in the recognition of the agonists, whereas a loss in the affinity of the antagonists to the mutated receptors was observed. We also have identified, for the first time, one amino acid (Trp351) in the agonist binding site of the receptor that is involved in the binding of the carboxyl-terminal sequence of CCK₈ as illustrated by the similar reduction in affinity for both CCK₈ and CCK₄. With regard to signal transduction, it seems that the Phe347 residue plays a key role in CCK₈-stimulated IP production. The results obtained with the W351A mutant also seem to demonstrate that the Trp351 residue could be involved in receptor/G protein coupling. These findings represent an important step toward the complete delineation of the agonist and antagonist binding sites, as well as the determination of the regions involved in the specificity of receptor/G protein coupling.