Determining the environment of the ligand binding pocket of the human angiotensin II type I (hAT1) receptor using the methionine proximity assay.

The peptide hormone angiotensin II (AngII) binds to the AT0 (angiotensin type 1) receptor within the transmembrane domains in an extended conformation, and its C-terminal residue interacts with transmembrane domain VII at Phe-293/Asn-294. The molecular environment of this binding pocket remains to be elucidated. The preferential binding of benzophenone photolabels to methionine residues in the target structure has enabled us to design an experimental approach called the methionine proximity assay, which is based on systematic mutagenesis and photolabeling to determine the molecular environment of this binding pocket. A series of 44 transmembrane domain III, VI, and VII X --> Met mutants photolabeled either with 125I-[Sar1,p'-benzoyl-L-Phe8]AngII or with 125I-[Sar1,p''-methoxy-p'-benzoyl-L-Phe8]AngII were purified and digested with cyanogen bromide. Several mutants produced digestion patterns different from that observed with wild type human AT1, indicating that they had a new receptor contact with position 8 of AngII. The following residues form this binding pocket: L112M and Y113M in transmembrane domain (TMD) III; F249M, W253M, H256M, and T260M in TMD VI; and F293M, N294M, N295M, C296M, and L297M in TMD VII. Homology modeling and incorporation of these contacts allowed us to develop an evidence-based molecular model of interactions with human AT1 that is very similar to the rhodopsin-retinal interaction.

The octapeptide hormone angiotensin II (AngII) 1 (Fig. 1A) is the active component of the renin-angiotensin system. Virtually all known physiological effects of AngII are produced through the activation of the hAT 1 receptor, which belongs to the class A rhodopsin-like family of the heptahelical G proteincoupled receptor (GPCR) superfamily (1,2). Elucidating the stereochemistry of the ligand-receptor interaction is vital for understanding the mechanism of ligand binding, GPCR activation, and, eventually, rational drug design.
In the past, much effort was devoted to identifying the domains or individual residues of a given receptor that may interact with its ligand. Most experiments to address ligandreceptor interactions were performed with series of receptor mutants to identify specific residues critical to ligand binding (3)(4)(5). It is, however, speculative to deduce precise structures of ligand-receptor interactions through mutagenesis studies alone. More direct approaches have therefore been used to study ligand-receptor interactions. Among these is photoaffinity labeling, which allows covalent incorporation of the ligand within its binding site, presumably at the contact area of the photolabel in the receptor. This ligand-receptor contact can be identified by specific enzymatic or chemical digestion of the labeled receptor (6) or by mass spectrometry (7). The binding pockets within the transmembrane domains of several bioamine receptors have been identified using this kind of approach. The adenosine A 1 receptor (8) and the ␤ 2 adrenergic receptor (9,10) are typical examples. Peptidergic receptors such as hAT 1 and hAT 2 (11,12), neurokinin receptors (13), and several other receptors from the secretin GPCR family B (14) have been also studied using this approach. We previously identified ligandcontact points within the second extracellular loop (ECL) and the seventh transmembrane domain (TMD) of the hAT 1 receptor (12,15,16). Although photoaffinity labeling has been widely used to study peptidergic GPCR binding pockets, generally only a single contact point between a given ligand and its cognate receptor has been identified. The resulting information does not, however, induce sufficient restrictions to generate credible GPCR structures in the ligand-bound state using homology modeling.
Labeling studies using benzophenone residues have identified many ligand-receptor contact points with a surprisingly high ratio of methionine contacts (17)(18)(19). Despite the fact that methionine represents a small proportion of the proteinogenic □ S The on-line version of this article (available at http://www.jbc.org) contains supplemental material in the form of a sequence alignment between the hAT 1 receptor and the bovine rhodopsin and a model of the [Sar 1 ,Bpa 8 ]AngII-liganded hAT 1 receptor with Protein Data Bank information as well.
‡ Submitted to fulfill the requirements of a Ph.D. thesis at the Université de Sherbrooke.
§ Holder of a scholarship from the Fonds Québécois de la recherche sur la nature et les technologies.
¶ Senior scholar of the Fonds de la recherche en Santé du Québec. amino acids in most receptors, p-benzoyl-L-phenylalanine (Bpa)-containing peptide labels (Fig. 1A) have been shown to incorporate into Met residues at a disproportionate frequency (13). Previous photochemical studies of the benzophenone radical have indicated that it exhibits strong selectivity for thioether groups (20,21). This selectivity would explain the high ratio of methionine insertion into target proteins through benzophenone photoaffinity labeling. This property can be exploited to introduce Met residues into target structures as "bait" with the goal of identifying other receptor residues that are in close proximity to the ligand label. The immediate molecular environment of this ligand residue can thus be determined. We used this strategy to investigate the binding environment of the C-terminal residue of AngII within the hAT 1 receptor. We used two benzophenone-containing labeling peptides, [Sar 1 ,Bpa 8 ]AngII, a well characterized neutral antagonist on the hAT 1 receptor (22), and [Sar 1 , pЈ-MeO-Bpa 8 ]AngII, which is also an antagonist. For the receptor, a systematic Met mutagenesis strategy was applied to TMD III, TMD VI, and TMD VII to identify other receptor contacts and thus define the binding environment of the C-terminal residue of AngII.

EXPERIMENTAL PROCEDURES
Materials-Bovine serum albumin, bacitracin, soybean trypsin inhibitor, and CNBr were from Sigma-Aldrich. Acetonitrile was from Fisher Scientific. Culture media were from Invitrogen. FuGENE 6 transfection reagent and the protease inhibitor mixture were purchased from Roche Diagnostics. X-ray films (Kodak Biomax® MS) with intensifying screens from Fischer Scientific were used to visualize CNBr digestion fragments. Photolabeling yields were determined using Quantity One® quantitation software from Bio-Rad.
Numbering of Residues-The residues of the human AT 1 receptor were given two numbering schemes. First, residues were numbered based on their positions in the human AT 1 receptor sequence. Second, residues were numbered based on their position relative to the most conserved residue in their respective TMDs within class A GPCRs (23). By definition, the most conserved residue was assigned the position index 50, with incremental numbering of downstream residues and decremental numbering of upstream residues, respectively. This indexing simplifies the identification of aligned residues in different GPCRs.
Oligodeoxynucleotide Site-directed Mutagenesis-Site-directed mutagenesis was performed on the wt-hAT 1 receptor using the overlap PCR method described elsewhere (24). Mutant receptors were subcloned into HindIII-XbaI sites of the mammalian expression vector pcDNA3.1. Site-directed mutations were confirmed by manual and automated DNA sequencing.
Synthesis and Radioiodination of Photoligands-[Sar 1 ,Bpa 8 ]AngII was prepared according to Bosse et al. (22). [Sar 1 ,pЈ-MeO-Bpa 8 ]AngII: pЈ-methoxy-p-methyl benzophenone was prepared according to Horner (25). Photobromination, resin alkylation, and peptide syntheses were carried out as described previously (22). The peptides were purified by reversed phase chromatography, which also permitted the separation of diastereomer peptides. Peptide purity as assessed by high performance liquid chromatography was at least 95%. The correct stereochemistry was assigned through comparison by high performance liquid chromatography with [Sar 1 ,L-Bpa 8 ]AngII and [Sar 1 ,D-Bpa 8 ]AngII made with Boc-L-Bpa and Boc-D-Bpa. Purified peptides were analyzed by matrixassisted laser desorption ionization time-of-flight mass spectrometry (Tofspec2, Micromass). All 125 I-AngII peptides (ϳ1500 Ci/mmol) were prepared using Iodogen® (Perbio Science, Erembodegem, Belgium) as described by Fraker and Speck (26), except that an acetic acid buffer (pH 5.4) was used. The radiolabeled peptides were purified by high performance liquid chromatography on a C-18 column (Waters) with a 20 -40% acetonitrile gradient in 0.05% aqueous trifluoroacetic acid. The specific radioactivity of the radiolabeled peptides was determined by self-displacement and saturation binding analysis.
Cell Cultures and Transfection of COS-7 Cells-COS-7 cells were grown in Dulbecco's modified Eagle's medium supplemented with 2 mM L-glutamine, 10% (v/v) fetal bovine serum, 100 IU/ml penicillin, and 100 g/ml streptomycin. The cells were incubated at 37°C in a 5% CO 2 atmosphere. Cells were transfected at ϳ70% confluency with FuGENE 6 transfection reagent as per the manufacturer's instructions. Thirtysix hours after the initiation of transfection, the cells were washed once with phosphate-buffered saline (137 mM NaCl, 0.9 mM MgCl 2 , 3.5 mM KCl, 0.9 mM CaCl 2 , 8.7 mM Na 2 HPO 4 , and 3.5 mM NaHPO 4 ) and immediately stored at Ϫ80°C until used.
Binding Studies and Photoaffinity Labeling-Frozen transfected COS-7 cells were thawed for 1 min at 37°C. The broken cells were then gently scraped, resuspended in 10 ml of washing buffer (25 mM Tris-HCl (pH 7.4), 100 mM NaCl, and 5 mM MgCl 2 ), and centrifuged (500 ϫ g for 10 min at 4°C). The pellet was dispersed in binding buffer (25 mM Tris-HCl (pH 7.4), 100 mM NaCl, 5 mM MgCl 2 , and 0.1% (w/v) bovine serum albumin). For binding studies, the broken cell suspension (50 -80 g of protein) was incubated for 60 min at room temperature in the presence of 0.1 nM 125 I-[Sar 1 ,Ile 8 ]AngII (1,500 Ci/mmol) with increasing concentrations of test peptide (15 concentration points in duplicate from 10 Ϫ12 to 10 Ϫ5 M with half-log increases. Bound radioactivity was separated from free ligand by filtration at 4°C through GF/C filters pre-soaked in binding buffer. Receptor-bound radioactivity was evaluated by ␥-counting. Results are presented as means Ϯ S.D. Binding data were analyzed with the Kell program (Biosoft, Ferguson, MO), which uses a weighted nonlinear curve-fitting routine. Maximal binding capacities were determined by approximation using the formula (B/T⅐IC 50 ) (27) from the displacement studies. For photolabeling studies, the broken cell suspension (1 mg of protein) was incubated for 90 min at room temperature in the presence of 3 nM 125 I-[Sar 1 , Bpa 8 ]AngII or 125 I-[Sar 1 ,pЈ-MeO-Bpa 8 ]AngII. After centrifugation at 500 ϫ g, the pelleted broken cells were washed once and resuspended in 0.5 ml of ice-cold washing buffer and then irradiated for 60 min on ice under filtered (Raymaster black light filters, catalog number 5873, Gates and Co. Inc., Franklin Square, NY.) UV light (365 nm) (100 watt mercury vapor lamp, serial number JC-Par-38, Westinghouse). After centrifugation (2,500 ϫ g for 10 min at 4°C), the pellet was solubilized for 30 min at 4°C in modified radioimmune precipitation assay buffer (50 mM Tris-HCl (pH 8), 150 mM NaCl, 0.5% (w/v) deoxycholate, 0.1% (w/v) SDS, and 1% (v/v) Nonidet P-40 supplemented with a protease inhibitor mixture (Complete EDTA-free) (Roche Diagnostics). The cell lysate was centrifuged (15,000 ϫ g for 25 min at 4°C) to remove insoluble material, and the supernatant was kept at Ϫ20°C until used.
Partial Purification of the Labeled Complex-The solubilized photolabeled receptor complexes were diluted in an equal volume of 2ϫ Laemmli buffer (120 mM Tris-HCl (pH 6.8), 20% (v/v) glycerol, 4% (w/v) SDS, 200 mM dithiothreitol, and 0.05% (w/v) bromphenol blue) and incubated for 60 min at 37°C. SDS-PAGE was performed as described previously (28)  The membranes were then exposed to UV light and solubilized. Photolabeled and solubilized receptor material was submitted to SDS-PAGE using 7.5% separating gels. The apparent molecular masses of the protein standards are indicated.
was then cut into slices, and the radioactive content was measured using a ␥-counter. The labeled receptor was passively eluted from the gel slices into fresh electrophoresis buffer (25 mM Trizma Tris base (pH 8.3), 250 mM glycine, and 0.1% (w/v) SDS) for 3-4 days at 4°C with gentle agitation as described by Blanton and Cohen (29). The eluate (ϳ40 ml) was concentrated to a final volume of 0.100 -0.250 ml using an Amicon-10 filter (Millipore) and stored at Ϫ20°C.
CNBr Hydrolysis-The partially purified photolabeled receptor (3,500 -10,000 cpm) was diluted in a 3:5 mixture of 30% trifluoroacetic acid and CNBr dissolved in 100% acetonitrile to obtain a final concentration of 50 mg/ml. Samples were incubated at room temperature in the dark for 16 -18 h. One milliliter of water was added to terminate the reaction. The samples were lyophilized and resuspended in Laemmli buffer (1ϫ final), loaded at 2,500 -3,500 cpm on 16.5% SDS-polyacrylamide Tris-Tricine gels (Bio-Rad), and revealed by autoradiography on x-ray films (Kodak Biomax® MS). 14 C-Labeled low molecular protein standards (Invitrogen) were used to determine the apparent molecular masses. Running conditions and fixation procedures were performed according to the manufacturer's instructions.
Inositol Phosphate Production-COS-7 cells were seeded in six-well plates, transfected, and labeled for 24 h in serum-free, inositol-free Dulbecco's modified Eagle's medium containing 10 Ci/ml myo-[ 3 H]inositol (Amersham Biosciences). Cells were washed twice with phosphate-buffered saline containing 0.1% (w/v) dextrose and then incubated in stimulation buffer (Dulbecco's modified Eagle's medium containing 25 mM Hepes, 10 mM LiCl, and 0.1% bovine serum albumin, pH 7.4) for 30 min at 37°C. Inositol phosphate production was induced with 100 nM AngII for 10 min at 37°C in stimulation buffer. Incubations were terminated by the addition of ice-cold perchloric acid (5% (v/v) final concentration). Water-soluble inositol phosphates were then extracted with an equal volume of a 1:1 (v/v) mixture of 1,1,2-trichlorotrifluoroethane and tri-n-octylamine. The samples were mixed vigorously and centrifuged at 2500 ϫ g for 30 min. The upper phase containing the inositol phosphates was applied to an AG1-X8 resin column (Bio-Rad). The inositol phosphates were eluted sequentially by the addition of an ammonium formate/formic acid solution of increasing ionic strength. Fractions containing inositol phosphates were collected and measured in a liquid scintillation counter.  (6.50) , and Pro-303-Pro-299 (7.50) . The coordinates of the assigned structurally conserved regions were then transferred to the sequence of hAT 1 , followed by ECL2 and ECL3 generation with the data base in HOMOLOGY. The conserved disulfide bond between ECL1 and ECL2 was then added to hAT 1 , and the potential energy was minimized sequentially using Discover with a consistent valence force field (31). In the first step of the minimization, all of the heavy atoms were fixed and all atoms, except those of the TMD backbones, were free to move. In the final step, all of the atoms were unrestrained.
To allow ligand [Sar 1 ,Bpa 8 ]AngII to be incorporated into the receptor, a cavity has to be generated. This was accomplished by a slight rotation (7°) of the angle of Asn-231. The Bpa molecule was modeled with INSIGHTII BUILDER and was placed between TMD III, TMD VI, and TMD VII as suggested by the photolabeling results. A first minimization of the complex between hAT 1 and Bpa (the coulombic terms were turned off) was performed using restraints (2 Å Ͻ d Ͻ 7Å) between the C␤ atoms of the photolabeled Met residues and the ketone oxygen of Bpa. The backbone atoms of hAT 1 were held in their position during this step. An extended Ang-(1-7) peptide (Sar-Arg-Val-Tyr-Val-His-Pro) was then appended to the N-terminal of Bpa. A subsequent minimization step was performed until the maximum derivative was Ͻ0.1 kcal/mol. The complex was further refined by adding a second disulfide bond between Cys-18 and Cys-274 (3) and by relaxing the ECL and performing a final overall minimization step.

Site-directed Mutagenesis and Photoaffinity
Labeling of hAT 1 -To identify the receptor residues that participate in the ligand binding pocket of the C-terminal amino acid of AngII, 44 X 3 Met mutations were induced in TMD III, TMD VI, and TMD VII of wt-hAT 1 (Fig. 2). Each mutant receptor was transiently expressed in COS-7 cells. Membranes containing wt-hAT 1 or two selected mutants, F293M (7.44) and N294M ( were photolabeled with 3 nM 125 I-[Sar 1 ,Bpa 8 ]AngII or 125 I-[Sar 1 ,pЈ-MeO-Bpa 8 ]AngII. They produced a broad band migrating diffusely between 75 and 180 kDa on SDS-polyacrylamide gels (Fig. 1B, lanes 1-3). Labeling was completely prevented when the experiments were carried out in the presence of 10 M AngII (Fig. 1B, lanes 4 -6) 1 and all the mutants. Covalent incorporation yields were calculated from the ratio of the total radioactivity in the 75-180 kDa bands versus the total specific binding observed before photolysis.
Pharmacological Properties of the Met Mutant Receptors-To assess the pharmacological profiles of the mutant receptors, K D and B max were determined using binding studies (Table I).

Generation of Liganded Receptor Structures by Molecular
Modeling-Sequence alignment (supplemental data, available on the on-line version of this article) between the hAT 1 receptor and the bovine rhodopsin used to identify and assign the structurally conserved regions had all the strictly conserved residues of class A GPCR aligned. The lengths of all TMDs and loops of hAT 1 were identical to those of the template structure, with the exception of the extracellular loops between the transmembrane regions TM IV and TM V (ECL2) and TM VI and TM VII (ECL3). There were no gaps in any of the TMDs. Pro and Gly, which can induce kinks in the TMD, were found at the same positions in bovine rhodopsin and hAT 1 , with the exception of Pro-285, which was at position 7.36 instead of 7.38 as in bovine rhodopsin. It was not possible to satisfy all the Bpa-C␤ constraints (2 Å Ͻ d Ͻ 7Å) with a single structure. The exclusion of the faintly Met-labeled mutant T260M, however, allowed for a structure with all ketone-sulfur distances in the 8-Å range. The molecular model of the [Sar 1 ,Bpa 8 ]AngII-liganded hAT 1 receptor is presented in Fig. 7 and in the supplemental material, available in the on-line version of this article. DISCUSSION In the present study, a new photoaffinity scanning approach called MPA was used as a strategy to probe the molecular binding environment of the hAT 1 receptor. In wt-hAT 1 , non-Met contact is made by [Sar 1 ,Bpa 8 ]AngII at positions Phe-293 (7.44) and Asn-294 (7.45) in TMD VII (16). Other non-Met contacts using the Bpa moiety have also been reported on other GPCRs (14,32,33) and are logically the consequence of the absence of Met residues in the binding locus environment. If a Met residue is introduced into the hAT 1 structure in sufficient proximity to the photoactive residue of the bound ligand, then part or all of the labeling should occur at this introduced Met residue because of the Met-selective nature of the photogenerated benzophenone radical. CNBr digestion of the covalent complex should thus generate a new fragment and produce a different SDS-PAGE profile. The main usefulness of the MPA approach is that receptor-ligand contacts can be directly and immediately determined without lengthy, multiple purification steps, protein digestions, or other manipulations.
Kage et al. (18) observed that when the ⑀-methyl group of the Met side chain is labeled by Bpa, CNBr hydrolysis releases the labeling ligand as a thiocyanomethyl derivative (Fig. 6) that can be detected by mass spectroscopy (18,34) or, as in our study, by SDS-PAGE (16,35). This observation led to the confirmation of suspected contact sites in several peptidergic receptors (15,16,35). The Met-selective labeling mechanism of the benzophenone radical is based on the formation of a chargetransfer complex between the photogenerated radical and the sulfur atom in the thioether (20,21). The following step, that is, insertion of the ketone radical into an adjacent C-H bond, can occur either in the ␥-methylene or the ⑀-methyl group, with only the latter leading to ligand release upon CNBr cleavage (Fig. 6). Exclusive ␥-labeling has been confirmed by mass spectrometry, which has shown that different rearrangements take place but that neither protein cleavage nor ligand release occur upon CNBr digestion (36). The cause of exclusive ␥-labeling instead of the anticipated ⑀-labeling is probably due to the sterical relations of the non-covalent charge- transfer complex. Independently of the ␥ or ⑀ insertion, labeling outside the normally labeled wt-hAT 1 -(285-334) fragment is revealed by new fragments or ligand release following CNBr digestion.
Most CNBr digestion profiles of labeled Met mutants indicate essentially ⑀-methyl labeling of the introduced Met residues with ensuing ligand release. The exclusive ␥-methylene incorporation by mutants F249M and W253M suggests that the benzophenone moiety is in close proximity to the TMD VI backbone. Such a favorable environment is present when the hydrophobic benzophenone residue is embedded in the hydrophobic aromatic cluster motif (6.40) IVLFFFFSWL (6.49) of TMD VI as seen in Fig. 7.
Studies using the substituted cysteine accessibility method (SCAM) on hAT 1 suggest that AngII inserts into a binding pocket where TMD III and TMD VII participate (35,38). The present study identified TMD III, TMD VI, and TMD VII as part of the ligand binding pocket, which is also the case for bioamine GPCRs (39 -41) and rhodopsin (42)(43)(44). This is further proof of a highly conserved feature among this very large family of receptors. Interestingly, in TMD III of hAT 1 the experimentally determined contact points were Leu-112 (3.36) and Tyr-113 (3.37) . In bovine rhodopsin, the corresponding resi-  8 ]AngII (␤) Cells transfected with the appropriate receptor were assayed as described under "Experimental Procedures." Binding affinities (K D ) and maximal binding capacities (B max ) are expressed as the means Ϯ S.D. of the values obtained from n independent experiments. Mutants A104M (3.28) , S105M (3.29) , and D263M (6.58) did not have any detectable binding activity.
TMD VII is the contact for the photolabeling analogue [Sar 1 ,Bpa 8 ]AngII on residues Phe-293 (7.44) and Asn-294 (7.45) (16). In many MPA-positive mutants, concomitant labeling of TMD VII was also observed (e.g. Y113M (3.37) and especially T260M (6.51) ). MPA on TMD VII could not be demonstrated through the formation of new fragments but solely through ligand release, which was evident in most TMD VII mutants. The ligand release profile of the TMD VII Met mutants indicated that there were continuous ligand contacts from Phe-293 (7.44) through Leu-297 (7.48) , with no ligand release in the biologically relevant X 3 Met mutants flanking this sequence (Ile-290 (7.41) , Ala-291 (7.42) , and Leu-300 (7.51) ). Continuous labeling through five residues in TMD VII implied that one and one-half helical turns with residues pointing away from the receptor core may intermittently contact the photoligand. This is in stark contrast to TMD VI, where a clearly defined internal TMD face emerged (Fig. 7). Recent results with SCAM analyses showed that TMD VII can move considerably in hAT 1 , depending on its activation status (35). Additionally, TMD VII is the only TMD with two Pro residues (Pro-285 (7.36) and Pro-299 (7.50) ), which flank the target sequence in hAT 1 . Furthermore, in bovine rhodopsin this segment does not have a typical ␣-helical structure (43). It is therefore reasonable to argue that this part of the TMD structure is somewhat destabilized, permitting outward pointing residues to turn intermittently inwards. The photochemistry of benzophenone has the particular capacity of repeat activation (45), allowing the interception of an intermediate state by a preferred partner (Met selectivity). Labeling is limited to a short segment of TMD VII at a depth comparable with the contacts in TMD III and with the lower most contacts in TMD VI (F249M (6.44) and W253M (6.48) ) (Fig. 7).
During the modeling process, a liganded hAT 1 structure was sought where all experimentally determined contacts were within Յ8 Å of the Bpa ketone oxygen. In the resulting structure, a single Met contact (T260M (6.55) ) was not compatible with a reasonable action radius of the Bpa-Met interaction, and two other residues (H256M (6.51) and Y113M (3.37) ) FIG. 6. Reaction scheme of photoactivated Bpa with a Met residue and the ensuring CNBr cleavage products. The asterisk (*) indicates a radioactive label for autoradiographic detection. The ␥-pathway produces labeled protein fragments only, and the ⑀ pathway produces ligand release. displayed borderline distances of ϳ9 Å. However, these two mutants, and especially T260M (6.55) , displayed concomitant labeling of the TMD VII contact (7.2 and 10 kDa, Fig. 4, lane 5). These patterns are indicative of a preponderant TMD VII interaction and a simultaneous but more occasional interaction with the mutated residues. Considering the location of the residues close to the membrane surface, these contacts are very suggestive of a transient ligand-receptor structure during the ligand binding process.
Of all the Met-mutated receptors, only a few mutants displayed altered binding properties. Ala-scan and SCAM analysis (Cys-scan) produced some binding-impaired mutants (35,38). Our mutagenesis scheme (X 3 Met) was applied mainly in the TMD area, and mutants with impaired ligand binding were only observed at the extracellular membrane interface. Met is an unbranched pseudo-aliphatic amino acid and can thus adapt to a given hydrophobic environment with minimal sterical hindrance. The MPA-positive mutants were assessed for receptor functionality. Nine of eleven mutants displayed AngII-stimulated inositol phosphate production, confirming biologically relevant receptor mutants. The remaining two mutants seemed to interfere with the activation mechanisms. N294M and L297M displayed normal AngII binding (Table I) but no inositol production. The Asn-294 residue is essential for receptor activation (16,46) but not for peptide ligand binding. L297M is more enigmatic, because almost isosteric changes were introduced. This residue is immediately adjacent to the NPXXY motif where a Tyr 3 Phe mutation also leads to AngII binding but a non-activable AT 1 mutant (47,48). This area seems to be ex-quisitely sensitive to small changes. In general, however, the X 3 Met mutagenesis strategy seems to be of little consequence either to receptor structure or function.
In conclusion, MPA made it possible to experimentally determine the binding environment of a given ligand residue. The MPA strategy can be applied to a large variety of receptor structures because X 3 Met mutagenesis appears to be generally of little biological consequence. The determination of these ligand contacts also allowed the construction of an evidencebased model of the hAT 1 receptor and showed that the receptor structure and ligand binding environment were very similar to those of bovine rhodopsin.