Role of Aromaticity of Agonist Switches of Angiotensin II in the Activation of the AT1 Receptor*

We have shown previously that the octapeptide angiotensin II (Ang II) activates the AT1 receptor through an induced-fit mechanism (Noda, K., Feng, Y. H., Liu, X. P., Saad, Y., Husain, A., and Karnik, S. S. (1996)Biochemistry 35, 16435–16442). In this activation process, interactions between Tyr4 and Phe8 of Ang II with Asn111 and His256 of the AT1receptor, respectively, are essential for agonism. Here we show that aromaticity, primarily, and size, secondarily, of the Tyr4side chain are important in activating the receptor. Activation analysis of AT1 receptor position 111 mutants by various Ang II position 4 analogues suggests that an amino-aromatic bonding interaction operates between the residue Asn111 of the AT1 receptor and Tyr4 of Ang II. Degree and potency of AT1 receptor activation by Ang II can be recreated by a reciprocal exchange of aromatic and amide groups between positions 4 and 111 of Ang II and the AT1 receptor, respectively. In several other bonding combinations, set up between Ang II position 4 analogues and receptor mutants, the gain of affinity is not accompanied by gain of function. Activation analysis of position 256 receptor mutants by Ang II position 8 analogues suggests that aromaticity of Phe8 and His256 side chains is crucial for receptor activation; however, a stacked rather than an amino-aromatic interaction appears to operate at this switch locus. Interaction between these residues, unlike the Tyr4:Asn111 interaction, plays an insignificant role in ligand docking.

We have shown previously that the octapeptide angiotensin II (Ang II) activates the AT 1 receptor through an induced-fit mechanism (Noda, K., Feng 8 and His 256 side chains is crucial for receptor activation; however, a stacked rather than an amino-aromatic interaction appears to operate at this switch locus. Interaction between these residues, unlike the Tyr 4 :Asn 111 interaction, plays an insignificant role in ligand docking. A central question in characterizing the biochemistry and action of the octapeptide hormone angiotensin II (Ang II) 1 is the mechanism by which the Tyr 4 and Phe 8 residues mediate the biological functions of Ang II. In vivo, Ang II is an important regulator of mean arterial pressure, water-electrolyte balance, and cardiovascular homeostasis. A clear understanding of the molecular mechanism of Ang II activation of cell surface receptors is necessary for the development of therapeutic agents to treat disorders such as high blood pressure and cardiac hypertrophy. A recently proposed model, based on NMR constraints of locked Ang II analogues, suggests that a very large pharmacophore spanning the entire structure of Ang II is presented to the receptor. Every other Ang II residue is involved in receptor contact. Only the Tyr 4 and Phe 8 side chains are considered agonist "switches" because analogues of Ang II function as agonists in vivo if the position 4 residue is tyrosine and the position 8 residue is a phenylalanine. Modifications of Tyr 4 and Phe 8 in Ang II give rise to antagonists in vivo that display high affinity for the Ang II receptor but at higher concentrations elicit partial receptor agonism (1)(2)(3)(4)(5).
Ang II type I (AT 1 ) and type II (AT 2 ) receptors belonging to the G-protein-coupled receptor (GPCR) superfamily are mediators of Ang II effects. AT 1 receptor is necessary and sufficient for regulating blood pressure, and activates intracellular inositol phosphate (IP) production via coupling to a pertussis toxin-insensitive G protein (3). Ang II-binding pocket consists of transmembrane domain and the extracellular loops. Two salt-bridge interactions, one between the ␣-COOgroup of Phe 8 of Ang II and Lys 199 of the AT 1 receptor and the other between the Arg 2 of Ang II and the Asp 281 of the AT 1 receptor, have been assigned. These salt-bridge interactions are not critical for receptor activation. Thus, a charge-separation mechanism described for the light activation of rhodopsin and the agonist activation of structurally related monoamine receptors is not a valid paradigm in the Ang II activation of the AT 1 receptor. Instead, the interactions of the Tyr 4 and Phe 8 residues of Ang II initiate the AT 1 receptor activation process (6 -10).
We previously obtained evidence for interaction of the Tyr 4 and Phe 8 side chains of Ang II, respectively, with Asn 111 and His 256 residues of the AT 1 receptor (10,11). Asn 111 also plays a critical role in stabilizing the basal "inactive" conformation of the native AT 1 receptor (11)(12)(13). His 256 is important for coupling agonist occupancy to G-protein activation (10). However, the nature of the bonds between Ang II and the receptor that switch the AT 1 receptor to its active state conformation is not clearly defined. In this report, we examine the hypothesis that amino-aromatic bonding between the agonist "switches" Tyr 4 and Phe 8 of Ang II and the respective agonist switch-binding residues Asn 111 and His 256 of the AT 1 receptor is responsible for initiating receptor activation. The absence of amino-aromatic interaction should primarily affect the receptor activation process. Analogues bearing saturated unnatural amino acid analogue ␤-cyclohexylalanine (Cha) substituents at the X 4 and X 8 positions of Ang II were synthesized and tested with wild-type and AT 1 receptor mutants with amino acid replacements at position Asn 111 or His 256 . Cha does not have a negatively charged planar aromatic ring but through its saturated ring provides nearly the same size and hydrophobicity as the aromatic rings of Tyr and Phe. We show that Cha replacements at either position 4 or position 8 of Ang II principally hinder ligand-dependent activation of the receptor. Analogues of Ang II-Analogues of [Sar 1 ]Ang II were synthesized and reverse phase high performance liquid chromatography purified by the peptide synthesis core facility at the Lerner Research Institute. The accuracy of synthesis was confirmed by electrospray mass spectrometry of the pure analogues using a PE-Sciex model AP1 III spectrometer. Concentration of the peptide in stock solutions was estimated by molarity of individual amino acids determined against known standards on an amino acid analyzer.
To evaluate the role of the aromaticity of Tyr 4 and Phe 8 rings without a change of residue size, ␤-cyclohexylalanine, an unnatural analogue of Phe, was substituted at these positions to obtain [Sar 1 (14), was estimated in each case as described below. The pharmacological studies on the wildtype and mutant receptors using various position 4 analogues is shown in Figs. 1-6.
Molecular Graphics and Estimation of Side Chain Surface Area-To generate the surface area values of Cha, mono-I-Tyr, and di-I-Tyr, we modeled the different molecules using bond lengths and angles extracted from a small molecule structure data base in an interactive graphics program called "O" (15). We then used the program "GRASP" to calculate the surface areas (16). The surface area values for the natural amino acid residues are from Ref. 14. Mutagenesis and Expression of the AT 1 Receptor-The synthetic rat AT 1 receptor gene, cloned in the shuttle expression vector pMT-2, was used for expression and mutagenesis, as described previously (8 -11, 17). Mutants were prepared either by the restriction fragment replacement method or by the polymerase chain reaction method, and DNA sequence analysis was done to confirm the mutations. To express the AT 1 receptor protein, 10 g of purified plasmid DNA/10 7 cells was used in transfection. COS1 cells (American Type Culture Collection, Rockville, MD), cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum, were transfected by the DEAE-dextran method. Transfected cells cultured for 72 h were harvested and cell membranes were prepared by the nitrogen parbomb disruption method. The receptor expression was assessed in each case by immunoblot analysis (data not shown) and by 125 I-[Sar 1 ,Ile 8 ]Ang II saturation binding analysis.
Radioligand Binding Studies-125 I-[Sar 1 ,Ile 8 ]Ang II-binding experiments were carried out under equilibrium conditions, as described previously (8 -11, 17). For competition binding studies, membranes expressing the wild-type receptor or the mutants were incubated at room temperature for 1 h with 300 pM 125 I-[Sar 1 ,Ile 8 ]Ang II and various concentrations of the agonist or antagonist. All binding experiments were carried out at 25°C in a 250-l volume. Nonspecific binding of the radioligand measured in the presence of 10 M 127 I-[Sar 1 ,Ile 8 ]Ang II was Ͻ3-5% of the total binding. After equilibrium was reached, the binding experiments were stopped by filtering the binding mixture through Whatman GF/C glass fiber filters, which were extensively washed further with binding buffer to wash the free radioligand. The bound ligand fraction was determined from the counts per minute (cpm) remaining on the membrane. Equilibrium binding kinetics were determined using the computer program Ligand . The K d values represent the mean Ϯ S.E. of three to five independent determinations.
Inositol Phosphate Formation Studies-COS1 cells (cultured in 60-mm Petri dishes), 24 h after transfection, were labeled for 24 h with [ 3 H]myoinositol (1.5 Ci/Petri dish), specific activity 22 Ci/mmol (Amersham), at 37°C in DMEM containing 10% bovine calf serum. On the day of the functional assay (i.e. 48 h after transfection), the labeled cells were washed with serum-free medium three times and incubated with DMEM containing 10 mM LiCl for 20 min; agonists were added and incubation continued for another 45 min at 37°C. At the end of incu-bation, the medium was removed, and total soluble IP was extracted from the cells by the perchloric acid extraction method, as described previously (8 -11). The amount of [ 3 H]IP eluted from the column was counted and a concentration-response curve generated using iterative nonlinear regression analysis (see Refs. 8 -11 and 17 for additional details).

RESULTS
The Experimental System-Transiently transfected COS1 cell model system was used for analysis as described previously (8 -11, 17). Immunoblotting experiments (data not shown) indicated that the expression of the mutant AT 1 receptors described in this report were Ϯ20% of the level of the wild-type receptor expression. This level of variability in receptor polypeptide expression did not cause significant variation in cell surface receptor numbers (Ϸ1.4 -1.6 ϫ 10 5 sites/cell), which was determined from acid-labile 125 I-[Sar 1 ,Ile 8 ]Ang II binding in intact cells. The B max estimated varied Ϸ2-fold (see Table I). Statistical analysis of 125 I-[Sar 1 ,Ile 8 ]Ang II binding kinetics was best fit to a one-site model, which indicated that a homogeneous population of wild-type and mutant receptors were produced in COS1 cells. Affinity of the wild-type AT 1 receptor to the radioligand was 0.37 Ϯ 0.02 nM. The K d values for the agonist [Sar 1 ]Ang II and the native hormone Ang II were 0.33 Ϯ 0.02 and 1.48 Ϯ 0.05 nM, respectively. The measured affinities did not change significantly in the presence of analogues of GTP, since our membrane preparations have been EDTA washed to uncouple G-proteins. Hence, the K d values in Table I represent the intrinsic affinity of the receptor in the absence of G-protein coupling. The expressed AT 1 receptor bound the nonpeptide antagonist losartan with high affinity (K d , 10 Ϯ 2 nM) and did not bind the AT 2 receptor-selective antagonist PD123319 (K d Ͼ10 Ϫ5 M). The ability of the AT 1 receptor to activate IP production in COS1 cells is shown in Fig. 1.
[Sar 1 ]Ang II-stimulated IP production in COS1 cells varied with the plasmid DNA transfected (see Fig. 1A). Expression plasmid concentrations, Ͼ4 g of DNA/dish, did not further increase the maximal IP produced, as well as cell surface receptor number (see above). All IP measurements shown in this report were carried out at 10 g of expression plasmid DNA/ dish. Therefore, the maximal IP values given in each case truly represent the magnitude of signal transduction and not the differences in cell surface receptor numbers. The basal IP production without [  Ang II is also consistent with size and aromaticity constraints. The binding energy contributions for the interactions were calculated as described in legend to Table II. The energy of interactions indicates that the aromatic ring in Tyr 4 contributed Ϸ2.8 kcal mol Ϫ1 , whereas the hydroxyl group in Tyr 4 contributed Ϸ0.4 kcal mol Ϫ1 and the surface area (based on comparison between Cha 4 and Ile 4 ) accounted for Ϸ0.6 -1.0 kcal mol Ϫ1 (see Table II).
Data in Table II indicate that AT 1 receptor activation is specifically hindered by Cha 4 substitution in Ang II. The [Sar 1 ,Cha 4 ]Ang II analogue partially activated (48 Ϯ 5%) the receptor function at concentration 1320-fold Ͼ K d (see Fig. 1  group of Tyr 4 in Ang II is crucial for activation (see below). Taken together, these observations indicate that activation of receptor by [X 4 ]Ang II analogues is not directly dependent on their binding to the AT 1 receptor (Fig. 2B).

Effect of Size Substitution of Asn 111 on Interaction with Different [X 4 ]Ang II Analogues-Interaction of the [X 4 ]
Ang II analogues with the residue 111 mutant receptors suggests that the interaction between the Asn 111 of the AT 1 receptor and the Tyr 4 of Ang II is a side chain size-dependent amino-aromatic bonding, which is essential for effective hormone-receptor coupling (Fig. 3). Amide group bearing substitution mutants N111Q and N111H bound [Sar 1 ]Ang II and [Sar 1 ,Phe 4 ]Ang II with 1.3-fold better affinity when compared with the wild-type AT 1 receptor. These mutants also bound [Sar 1 ,Cha 4 ]Ang II and [Sar 1 ,Ile 4 ]Ang II analogues with slightly improved affinity over that of the wild-type receptor. The N111K mutant affinity for these analogues is comparable to that exhibited by the wildtype receptor (Fig. 3).
In contrast, the affinity of [Sar 1 ]Ang II was decreased 1.6-, 5.1-, and 3.3-fold, respectively, in the N111I, N111F, and N111Y mutants. Substitution of a larger residue at position 111 that did not provide an amide group cost 0.2-1.2 kcal mol Ϫ1 more energy in binding Tyr 4 . The [Sar 1 ,Cha 4 ]Ang II and [Sar 1 ,Ile 4 ]Ang II analogues interacted with these three mutants with significantly higher affinity, perhaps through hydrophobic interaction restoring specificity. Most importantly, magnitude of increased affinity of the N111I mutant was consistent with an interaction dependent on size and hydrophobicity. These results further substantiate our previous proposal that Asn 111 in the middle of transmembrane helix III of the AT 1 receptor interacts with Tyr 4 of Ang II (11).
The gain of affinity, however, is not accompanied by gain of function in most bonding combinations set up between Ang II position 4 analogues and receptor mutants (see Fig. 3) to evaluate the critical requirement for amino-aromatic interaction in signal transduction. The maximal IP production stimulated by [Sar 1 ]Ang II and [Sar 1 ,Xaa 4 ]Ang II in most of the mutants shown in Fig. 3 was Ϸ50% reduced in comparison to the wildtype AT 1 receptor. In several combinations the binding interaction was preserved, for instance Tyr 4 -Gln 111 , Tyr 4 -His 111 , Tyr 4 -Lys 111 , Tyr 4 -Ile 111 , Phe 4 -Gln 111 , Phe 4 -His 111 , and Phe 4 -Ile 111 , but the activation was inadequate (see Fig. 3). The K d and EC 50 in these ligand-receptor combinations were close to (Ͻ2-fold) that of a wild-type situation, but the maximal IP response was impaired. Thus, the increased size of the residue 111 affected receptor activation much more than the binding affinity (K d ) and potency (EC 50 ). An exception is the Phe 4 -Tyr 111 interaction, in which the measured K d , EC 50 , and maximal IP values are comparable to that of Phe 4 -Asn 111 interaction. The reasons for this are unknown.
Reversal of Amino-aromatic Interaction-An efficient functional interaction is reproduced by substitution of an Asn residue at position 4 in [Sar 1 ]Ang II and a Phe residue at position 111 in the receptor (see Fig. 4, also see Table II). The binding affinity of [Sar 1 ,Asn 4 ]Ang II for the wild-type AT 1 receptor is 230-fold reduced. High affinity for [Sar 1 ,Asn 4 ]Ang II was regained in the N111F (K d ϭ 0.49 Ϯ 0.05 nM) and N111Y (K d ϭ  Fig. 1 for cpm values) was taken as 100%. In each experiment, this control was included for comparison and values are normalized for variations in IP formation in basal and mock-transfected cells. Relative activation value close to 100% indicates that activation of the mutant by the analogue and [Sar 1 ]Ang II are nearly identical. Note that substitution with smaller residues, such as Gly. Ala, Ser, and Cys for Asn 111 , gives rise to a partially activated receptor that binds almost all Ang II analogues with high affinity, and maximal stimulation of these mutant receptors requires Ang II, but not its agonist side chains. Hence, those mutations are not included in the current analysis (32).  290,198, and 2980 nM, respectively. These observations suggest that providing an electronegative group at the X 4 position is insufficient. The fact that Asn 111 /Phe 4 and Phe 111 /Asn 4 combinations produced identical affinity and potency demonstrates that an amino-aromatic interaction is essential for the fidelity of AT 1 receptor-Ang II coupling. The wild-type receptor interaction with Ang II yields Ϸ20% more IP, suggesting that the critical interaction provided by the Tyr 4 hydroxyl group is not mimicked in Asn 111 /Phe 4 and Phe 111 /Asn 4 combinations. If conventional electrostatic/hydrogen bonding occurred between Asn 111 and Tyr 4 , our analysis would have been expected to restore affinity and potency (equal to that of the wild-type receptor) in several different combinations.
Pharmacological Characterization of Phe 8 -modified [Sar 1 ] Ang II Analogues-Substitution of Cha 8 for Phe 8 did not significantly affect the binding affinity, but reduced the level of AT 1 receptor activation (Table I, Fig. 5). The maximal IP response elicited by [Sar 1 ,Cha 8 ]Ang II was reduced by 60 Ϯ 5%, which appears to be caused by diminished AT 1 receptor activation by the bound analogue. Size alteration at the Phe 8 position has a nearly insignificant effect on binding. The change of K d resulting from the substitution of Phe 8 with Gly 8 , Ala 8 , Thr 8 , and Ile 8 was within 6-fold. Replacement with Glu 8 and Trp 8 produced lower affinity, 89 Ϯ 9 and 13.4 Ϯ 2 nM, respectively, indicating that charged groups and large hydrophobic groups are not accommodated at the putative Phe 8 binding site on the receptor. Preservation of the aromatic character, as in [Sar 1 ,Trp 8 ]Ang II, yielded nearly full receptor activation, suggesting that the reduction of binding affinity at the Phe 8 binding pocket did not affect the ability to activate. Thus, the aromaticity of Phe 8 is required for receptor activation, but essentially plays no role in the receptor-binding step.
Effect of Different Substitutions of His 256 on Activation by Ang II-Ang II-Phe 8 -mediated functional activation is specifically and uniquely dependent on His 256 (10). Substitution of His 256 did not affect Ang II binding affinity, but caused receptor to lose the ability to be activated by Ang II. We examined which residue combination would mimic the type of functional interaction between Phe 8 and His 256 , employing three substitution mutants of His 256 (Fig. 6). The binding affinity of H256A, H256Q, and H256Y mutants for [Sar 1 ]Ang II was within 2-fold of the binding affinity for the wild-type. The receptor activation was reduced Ϸ60% in the H256Q and Ϸ70% in the H256A mutant receptors. In contrast, substitution of His 256 with a Tyr (H256Y) led to a Ϸ6% reduction of maximal IP response. Activation by the [Sar 1 Cha 8 ]Ang II was relatively less affected by various mutations (see Fig. 6). The [Sar 1 Cha 8 ]Ang II evoked Ϸ40% response in the wild-type receptor, Ϸ52% in the H256Y mutant, Ϸ35% in the H256Q mutant, and Ϸ28% in the H256A mutant receptor. The most surprising outcome in this analysis was the ability of a tyrosine substituted for His 256 to display full activation, which suggests that a stacking interaction involving the aromatic rings of His 256 and Phe 8 might be responsible for AT 1 receptor activation.

DISCUSSION
Like a large variety of hormone GPCRs, the molecular mechanism involved in Ang II-mediated AT 1 receptor activation remains unclear. Some insights into the mechanism have come from structure-function studies and the discovery of mutations that uncouple receptor activation from agonist binding (8 -13, 20 -22). Two most important such mutations in AT 1 receptor affect amide group bearing residues, Asn 111 and His 256 , both interact with the agonist switches, Tyr 4 and Phe 8 of Ang II. Since aromatic rings are able to participate in amino-aromatic bonding (23)(24)(25) in addition to the obvious hydrophobic, hydrogen bonding and van der Waals interactions because of an accessible center of negative charge (from the ␦electrons), our goal was to distinguish potential bonding interactions of Tyr 4 and Phe 8 in Ang II. Because the information available regarding Ang II function was obtained from in vivo functional studies prior to the cloning of the receptor, and also before radioligand binding assay came into routine use (5), our study is justified. Primary comparison involved Cha-substituted Ang II analogues but the aliphatic-substituted analogues served mostly to calibrate size effects. Cha lacks a negative charge, lacks hydrogen bonding potential, and provides hydrophobicity. The volume effect from Cha modification is not significant because the space-filling models indicate that the overall sizes of chair and boat configurations of the cyclohexyl ring and the planar configuration of aromatic ring are substantially similar (23). Since the functional significance of the highly directional and significantly attractive interaction between amide groups of Asn, Gln, His, and Lys and the aromatic rings of Tyr, Phe, and Trp has been questioned in several proteins (23)(24)(25), we evaluated whether the amino-aromatic bonding interactions are essential in Ang II-AT 1 receptor coupling. Our results indicate that the aromaticity of both Tyr 4 and Phe 8 is crucial, but an amino-aromatic bonding operates at the Tyr 4 switch of Ang II and a stacked rather than an amino-aromatic interaction appears to operate at the Phe 8 switch locus.
The complex role of Tyr 4 -Asn 111 amino-aromatic bonding in Ang II-AT 1 receptor coupling is suggested by the observation that both affinity and potency of receptor activation is recreated by reversal of interaction (Fig. 4), but not in several other combinations that restore binding affinity. Functional equivalent of this bonding is recapitulated in Phe 4 -Phe 111 Phe 4 -Tyr 111 , and Tyr 4 -Tyr 111 interactions. An edge-to-face aromaticaromatic interaction that might be operating in all three instances would be similar to an amino-aromatic interaction based on past examples (23,(25)(26)(27). The enthalpic contribution of this interaction to Ang II binding is estimated at 3.2 kcal/mol (see Table II) which is comparable to the estimated energy (3.3 kcal/mol) of amino-aromatic hydrogen bonds in the protein structure data base (26). This interaction is superimposed with size and hydrophobicity constraints. For instance, the reduction in binding affinity resulting from Tyr 4 3 Ile 4 change in [Sar 1 ]Ang II could be overcome by substitution of Asn 111 in the receptor with larger hydrophobic residues Ile, Phe, and Tyr but not by larger hydrophilic residues Lys, Gln, and His (Fig. 3). Increasing the accessible surface area from Ϸ160 Å (Asn 111 ) to Ϸ180 Å (Gln 111 ), and Ϸ195 Å (His 111 ) appeared to increase affinity for [Sar 1 ]Ang II, and [Sar 1 ,Phe 4 ]Ang II. The gain of affinity in both instances, however, is without gain in receptor activation, indicating that Tyr 4 is the most complex switch in Ang II because the receptor-binding and agonism-specifying elements are structurally integrated. The size chiefly influences the K d , but receptor activation requires stringent conservation of size and aromaticity.
The binding and agonism-specifying elements are structurally separate in the Phe 8 switch. The ␣-COOin Phe 8 is the docking group and the benzyl-alanyl moiety is the agonist switch, modifications of which generated potent antagonists without compromising the binding affinity (also seen in Fig. 5). Classical structure-activity relationship studies portrayed that any modification of the Phe 8 that prevents the planar arrangement of benzyl side chain over the Pro 7 -Phe 8 amide bond and the ␣-COOgroup will disrupt agonist potential (1,2,5). Such modifications include aliphatic substitutions, and aromatic groups in D-configurations (D-Phe, ␣-MePhe) or with bulky ring substitutions (Trp, Ind) (1-6). These evidences indicate that rigid planar configuration is a restriction in the interactions of Phe 8 switch with the receptor. Proposed stacking interaction between the aromatic rings of His 256 and Phe 8 is consistent with this requirement. When His 256 is replaced with isosteric Gln or of Phe 8 with isosteric Cha, an inefficient hormonereceptor coupling occurs because Gln and Cha do not participate in the planar stacking interaction with an aromatic ring. Since the Tyr 256 mutant is fully active, this suggests that the His 256 -Phe 8 interaction is not an ion-quadripole type of interaction. In the protein structure data bank, stacked geometry is far more abundant for His-Phe interactions than typical aminoaromatic interactions (29). A true stacking interaction (with 1/r 6 dispersion force between two strategically placed planar rings nearly parallel) is a weaker interaction than an aminoaromatic interaction, consistent with the insignificant contribution of Phe 8 to the binding affinity of Ang II. Underwood et al. have previously suggested an aromatic stacking interaction model for AT 1 receptor agonism involving the nonpeptide agonist L-162,313 (30). Evidences suggest that His 256 makes contact with AT 1 receptor specific non peptide antagonists also (10,31). Therefore, we propose that actuation of the His 256 switch of the receptor is a common step for AT 1 receptor activation by both peptide and nonpeptide agonists and the site for antagonist action. Preservation of planar benzyl imidazole or benzyl acrylic acid ring structure is crucial in these ligands, suggesting that a stacking interaction with His 256 is likely in the mechanism of receptor antagonism as well.
Based on these results, we speculate that the functional coupling of Ang II binding to receptor activation requires a structural coupling through stringent bonding between the agonist switches and respective switch-binding residues. In the wild-type receptor, because Asn 111 is smaller, a conformational change might be required to facilitate the amino aromatic Tyr 4 -Asn 111 interaction. The proposed conformational change is the basis for efficient activation of function. Ang II binding, in all likelihood, does not trigger a similar conformational change in the N111Q, N111H, N111K, N111I, and N111F mutants, which forms the basis for defect. Although an aminoaromatic bonding is expected in N111Q mutant (also N111H), the defect in IP formation suggests that one methylene unit increase in size is detrimental and perhaps uncouples binding and activation. In contrast, the stacking Phe 8 -His 256 interaction is not accompanied by a large conformational change, but likely exposes sites within the receptor for the crucial interactions. Numerous studies have described partial agonist effects of modifying Tyr 4 and Phe 8 in Ang II. For the most part these reports indicated that conservation of these two side chains is critical for Ang II function. Our results extend these observations in that highly efficient receptor activation of Ang II analogues is lost, leaving high affinity binding intact, suggesting the analogue-receptor docking occurs, but due to lack of the required bonding interactions, functional response of the receptor is incomplete. This finding has important implications for drug discovery. CONCLUSIONS Different roles played by the agonist switch residues Tyr 4 and Phe 8 of Ang II in the AT 1 receptor activation are pointed out. The conclusion that two critical coupling contacts between Ang II and the AT 1 receptor involve the transmembrane helices III and VI of the receptor suggests that the receptor activation by Ang II may involve motion of these helices, a phenomenon also observed in other GPCRs (28). Conformational changes induced by the two distinctly different types of bonding interactions could orient helices and loops on the receptor so as to enhance the efficiency of receptor-G-protein coupling. Such conformational switching could also serve as the basis for agonist potency because the ability of a full agonist analogue to properly align helices far exceeds that of partial agonist analogues. The working model suggested is based on the assumption that different peptide analogues induce non-identical conformation of the activated receptor that lead to differences in the kinetics and the magnitudes of responses. Whether this induced-fit mechanism is unique to the subfamily of peptide hormone GPCRs or is a more general mechanism that has not been considered in prototypical GPCRs is unclear at present.