Constitutive activation of the angiotensin II type 1 receptor alters the spatial proximity of transmembrane 7 to the ligand-binding pocket.

Activation of G protein-coupled receptors by agonists involves significant movement of transmembrane domains (TM) following binding of agonist. The underlying structural mechanism by which receptor activation takes place is largely unknown but can be inferred by detecting variability within the environment of the ligand-binding pocket, which constitutes a water-accessible crevice surrounded by the seven TM helices. Using the substituted cysteine accessibility method, we initially identified those residues within the seventh transmembrane domain (TM7) of wild type angiotensin II type 1 (AT1) receptor that contribute to forming the binding site pocket. We have substituted successively TM7 residues ranging from Ile276 to Tyr302 to cysteine. Treatment of A277C, V280C, T282C, A283C, I286C, A291C, and F301C mutant receptors with the charged sulfhydryl-specific alkylating agent MTSEA significantly inhibited ligand binding, which suggests that these residues orient themselves within the water-accessible binding pocket of the AT1 receptor. Interestingly, this pattern of acquired MTSEA sensitivity was greatly reduced for TM7 reporter cysteines engineered in a constitutively active mutant of the AT1 receptor. Our data suggest that upon activation, TM7 of the AT1 receptor goes through a pattern of helical movements that results in its distancing from the binding pocket per se. These studies support accumulating evidence whereby elements of TM7 of class A GPCRs promote activation of the receptor through structural rearrangements.

Activation of G protein-coupled receptors by agonists involves significant movement of transmembrane domains (TM) following binding of agonist. The underlying structural mechanism by which receptor activation takes place is largely unknown but can be inferred by detecting variability within the environment of the ligand-binding pocket, which constitutes a water-accessible crevice surrounded by the seven TM helices. Using the substituted cysteine accessibility method, we initially identified those residues within the seventh transmembrane domain (TM7) of wild type angiotensin II type 1 (AT 1 ) receptor that contribute to forming the binding site pocket. We have substituted successively TM7 residues ranging from Ile 276 to Tyr 302 to cysteine. Treatment of A277C, V280C, T282C, A283C, I286C, A291C, and F301C mutant receptors with the charged sulfhydryl-specific alkylating agent MTSEA significantly inhibited ligand binding, which suggests that these residues orient themselves within the water-accessible binding pocket of the AT 1 receptor. Interestingly, this pattern of acquired MTSEA sensitivity was greatly reduced for TM7 reporter cysteines engineered in a constitutively active mutant of the AT 1 receptor. Our data suggest that upon activation, TM7 of the AT 1 receptor goes through a pattern of helical movements that results in its distancing from the binding pocket per se. These studies support accumulating evidence whereby elements of TM7 of class A GPCRs promote activation of the receptor through structural rearrangements.
G protein-coupled receptors (GPCRs) 1 comprise a large family of cell surface receptors that mediate diverse responses to a large variety of sensory and hormonal signals. As a common characteristic, they all possess seven membrane-spanning domains that constitute structural support for signal transduction. However, despite a low sequence identity, GPCRs are thought to mediate signal transduction by a mechanism involving common structural movements implicating transmembrane domains (TM) (1). This receptor conformational change is suspected to sustain GTP/GDP exchange on specific guanine nucleotide binding proteins (G proteins) leading to activation of intracellular signaling cascades. Indeed, rotation described by rigid body movement of TM helices as well as translation have been described for activation of rhodopsin and other GPCRs (2)(3)(4).
Rhodopsin belongs to class A within the GPCR superfamily. The recent elucidation of the three-dimensional structure of rhodopsin has shed much light into how photoisomerization is coupled to conformational change, which leads to G protein signalization (5,6). Also belonging to class A, the AT 1 receptor binds the octapeptide hormone angiotensin II (AngII) and activates the G protein G q/11 . This leads to the increase of intracellular Ca 2ϩ levels following hydrolysis by phospholipase C of membranous phosphatidylinositol-bis-phosphate into diacylglycerol and inositol 1,4,5-trisphosphate. Most of the physiological actions of AngII on cardiovascular, endocrine, and neuronal systems are mediated by the AT 1 receptor (7). However, the molecular mechanisms by which the AT 1 receptor activates those pathways remain elusive. It has been proposed that TM3, TM5, TM6, and TM7 might participate in this activation process by providing a network of interactions through the AngIIbinding pocket (8). The dynamics of this network is suspected to be modified following agonist binding, thereby forcing the receptor to form new interactions between TMs. Another way in which an alteration in interactions of TM residues can yield interesting phenotypes is the existence of constitutively active receptors. For instance, it is well documented that N111G mutants within TM3 of the AT 1 receptor confer a high level of constitutive activity (9), higher than N111A mutants. This mutation may release receptor constraints between TMs, thereby breaking the existing network involving multiple intramolecular interactions. Indeed, a rigid body movement of TM2 has been described recently for the AT 1 receptor following such mutations at Asn 111 (10). However, because TM7 bears multiple activation determinants and motifs, such as the 298 NPXXY 302 motif (11), which has been implicated in receptormediated G protein activation, their role in the N111G receptor mutant is quite intriguing.
Like in many GPCRs, the binding site of the AT 1 receptor is formed among its seven, mostly hydrophobic transmembrane segments and is accessible to charged water-soluble agonists, like AngII. Thus, for this receptor, the binding site is contained within a water-accessible crevice, the binding pocket, extending from the extracellular surface of the receptor into the transmembrane domain (12). The surface of this crevice is formed by residues that can contact specific determinants on ligands and by other residues that may play a structural role as well as affect binding indirectly. Here, using the substituted cysteine accessibility method (SCAM), we probed the AT 1 receptor-binding pocket (13).

EXPERIMENTAL PROCEDURES
Materials-Bovine serum albumin, bacitracin, and soybean trypsin inhibitor were from Sigma. The sulfhydryl-specific alkylating reagent used was CH 3 SO 2 -SCH 2 CH 2 NH 3 ϩ , MTSEA, which was purchased from Toronto Research Chemicals, Inc. (Toronto, Canada). The cDNA clone of the human AT 1 receptor subcloned in the mammalian expression vector pcDNA3 was kindly provided by Dr. Sylvain Meloche (Université de Montréal). LipofectAMINE and culture media were obtained from Invitrogen. 125 I-[Sar 1 ,Ile 8 ]AngII (specific radioactivity, ϳ2000 Ci/mmol) was prepared with Iodo-GEN® (Pierce) according to the method of Fraker and Speck (14) and as previously reported (11).
Numbering of Residues in Transmembrane Domains-Residues in the seventh transmembrane domain of the human AT 1 receptor are given two numbering schemes. First, residues are numbered according to their positions in the human AT 1 receptor sequence. Second, residues are also indexed relative to the most conserved residue in the TM in which it is located (N1,N2) (15). N1 refers to the TM number, and N2 refers to the position of the residue relative to the most conserved one with numbers decreasing toward the N terminus and increasing toward the C terminus. By definition, the most conserved residue is assigned the position index 50, e.g. Pro 299(7.50) and therefore Asn 298(7.49) and Leu 300 (7.51) . This indexing simplifies the identification of aligned residues in different GPCRs.
Oligodeoxynucleotide Site-directed Mutagenesis-Site-directed mutagenesis was performed on the wild type AT 1 receptor with the overlap PCR method described elsewhere (16). Mutant receptors were subcloned into HindIII-XbaI sites of the mammalian expression vector pcDNA3. Site-directed mutations were then confirmed by manual and automated DNA sequencing.
Cell Culture and Transfections-COS-7 cells were grown in Dulbecco's modified Eagle's medium (DMEM) containing 2 mM L-glutamine and 10% (v/v) fetal bovine serum. The cells were seeded into 100-mm culture dishes at a density of 2 ϫ 10 6 cells/dish. When cells reached ϳ70% confluency, they were washed with serum-free DMEM and transfected with 2 g of plasmid DNA and 15 l of LipofectAMINE in 8 ml of serum-free DMEM. The cells were incubated for 5 h at 37°C, and the media were replaced with a complete DMEM containing 100 IU/ml penicillin and 100 g/ml streptomycin. Transfected cells were trypsinized after 24 h, plated into 12-wells plates, and grown for an additional 24 h before MTS treatment. Transfected cells in 100-mm culture dishes were grown for 48 h before binding assays.
Binding Experiments-Cell membrane preparation and binding assays were performed as described previously (12). COS-7 cells were grown for 48 h post-transfection in 100-mm culture dishes, washed once with PBS, and subjected to one freeze-thaw cycle. Broken cells were then gently scraped in washing buffer (25 mM Tris-HCl, pH 7.4, 100 mM NaCl, 5 mM MgCl 2 ), centrifuged at 2500 ϫ g for 15 min at 4°C, and resuspended in binding buffer (25 mM Tris-HCl, pH 7.4, 100 mM NaCl, 5 mM MgCl 2 , 0.1% (w/v) bovine serum albumin, 0.01% (w/v) bacitracin). Broken cells (20 -40 g of protein) were incubated for 1 h at room temperature with 0.05 nM of 125 I-[Sar 1 ,Ile 8 ]AngII (2000 Ci/mmol) and increasing concentrations of [Sar 1 ,Ile 8 ]AngII. Bound radioactivity was separated from free ligand by filtration through GF/C filters presoaked for at least 3 h in binding buffer. Nonspecific binding was measured in the presence of 1 M unlabeled [Sar 1 ,Ile 8 ]AngII. Receptor-bound radioactivity was evaluated by ␥ counting.
MTS Treatment-Experiments were performed using a procedure modified from Javitch et al. (17). 48 h hours after transfection, cells plated into 12-well plates were washed with PBS and incubated with freshly prepared MTSEA at the stated concentrations (typically from 0.5 to 6 mM) in a final volume of 0.2 ml at room temperature for 3 min. The reaction was stopped by washing the cells with ice-cold PBS. The intact cells were then incubated in binding medium (DMEM, 25 mM HEPES, 0.1% bovine serum albumin, pH 7.4) containing 0.05 nM 125 I-[Sar 1 ,Ile 8 ]AngII for 1.5 h at room temperature. After washing with ice-cold PBS, the cells were lysed with 0.1 N NaOH, and the radioactivity was evaluated by ␥ counting. The percentage of fractional inhibition was calculated as [1 Ϫ (specific binding after the MTS reagent/specific binding without the reagent)] ϫ 100%.
Measurement of the Second Order Rate Constant for MTSEA Reaction-The second order rate constant (k) for the reaction of MTSEA with each susceptible receptor mutants was estimated by determining the extent of reaction after a fixed period of time (3 min) with five concentrations of MTSEA (typically from 0.003 to 6 mM) (all in excess over the reactive sulfhydryls). The fraction of initial binding Y was fit to (1 Ϫ plateau) e Ϫkct ϩ plateau, where plateau is the fraction of residual binding at saturating concentrations of MTSEA, k is the second order rate constant (in M Ϫ1 s Ϫ1 ), c is the concentration of MTSEA ([M]), and t is the time (180 s). Fitting was achieved using the EnzFitter software from BioSoft (Ferguson, MO).
Protection against MTSEA Reaction by [Sar 1 ,Ile 8 ]AngII-Transfected cells plated into 12-well plates were washed once with PBS and incubated in the presence or the absence of 100 nM [Sar 1 ,Ile 8 ]AngII for 1 h at 16°C to avoid internalization of membrane expressed receptors. The cells were washed to remove excess ligand and then treated with a concentration of MTSEA that was just sufficient to achieve maximal inhibition of binding to each receptor. The cells were washed three times with ice-cold PBS and were submitted to an acidic wash (150 mM NaCl, 50 mM acetic acid, pH 3.0) to remove bound ligand. The cells were then incubated in binding medium (DMEM, 25 mM HEPES, 0.1% bovine serum albumin, pH 7.4) containing 0.05 nM 125 I-[Sar 1 ,Ile 8 ]AngII for 3h at 16°C. After washing with ice-cold PBS, the cells were lysed with 0.1 N NaOH, and the radioactivity was evaluated by ␥ counting. The percentage of protection was calculated as [(inhibition in the presence of [Sar 1 ,Ile 8 ]AngII)/(inhibition in the absence of ligand)] ϫ 100%.
Molecular Modeling-All of the calculations were performed on a Silicon Graphics Octane2 work station. Theoretical structure of the hAT1 receptor was generated by homology modeling based on the crystal structure of the bovine rhodopsin (Protein Data Bank code 1F88). A pair-wise sequence alignment between the two primary structures was performed using the program Homology of Insight II (Accelrys Inc., San Diego, CA). After identification of the structurally conserved regions (transmembrane domains), these coordinates were transferred to the sequence of the hAT1 receptor. The structures of the loops of hAT1 were modeled by using the loop generation program of Insight II. The potential energy of the model structure of hAT1 was minimized by using the molecular modeling package of Insight II/ Discover with consistent valence forcefield (18). Two disulfide bridges, Cys 18 -Cys 275 and Cys 101 -Cys 180 , were used as distance restraints. A distance-dependent dielectric constant of 4 was used with simple harmonic potential for bond length energy. No cross-term energies were included, and the peptide bonds were forced to planarity. 1 Sensitivity to MTSEA Added Extracellularly-To verify the contribution of the 10 endogenous cysteines ( Fig. 1) to ligand binding, we initially performed the SCAM approach on the wild type AT 1 receptor. Whole cells transiently expressing the AT 1 receptor were treated with varying amounts of the sulfhydryl-specific alkylating reagent MTSEA. Fig. 2 shows that treatment with various concentrations of MTSEA had very little effect on the binding properties of the AT 1 receptor. No more than a 10% reduction in binding was observed, showing the relatively low contribution of the endogenous cysteines of the receptor to the binding pocket. Alkylation of the AT 1 receptor with other charged MTS reagents (methanethiosulfonylethyl-trimethylammonium and methanethiosulfonylethyl-sulfonate) also did not significantly affect ligand binding (data not shown).

Assessment of hAT
Binding Properties of hAT 1 Mutant Receptors Bearing Cysteines in TM7-To establish whether residues of TM7 of the AT 1 receptor orient themselves into the binding pocket and hence could potentially affect ligand binding, we generated a series of mutants whereby residues Ile 276 (7.27) to Tyr 302(7.53) were replaced by cysteines. Each mutant receptor was transiently expressed in COS-7 cells. To assess the conservation of global conformation of these receptors after such substitutions, pharmacological parameters describing the equilibrium binding of the radiolabeled antagonist 125 I-[Sar 1 ,Ile 8 ]AngII such as K d and B max were determined ( Table I). All of the mutant receptors exhibited high affinity binding for 125 I-[Sar 1 ,Ile 8 ]AngII except for D278C (7.29) , D281C (7.32) , M284C (7.35) , and P285C (7.36) , which did not demonstrate any detectable binding, and N295C (7.46) , which showed a moderate 13-fold increase in K d (Table I). Those receptors where binding was undetected were not used for SCAM analysis. B max values for all detectable receptors ranged from 0.2-to 1.6-fold that of the wild type receptor (Table I).
Effect of Extracellularly Added MTSEA on Binding Properties of Mutant Receptors-To verify the location of reporter cysteines introduced into TM7 of the receptor toward the binding pocket, all detectable receptor mutants were treated with varying concentrations of MTSEA. MTSEA at 0.5 mM significantly reduced ligand binding to 5 of the 20 cysteine substitution mutants tested (Fig. 3A). Increasing the MTSEA concen-tration to 2 mM significantly reduced binding to seven mutant receptors (Fig. 3B). Higher MTSEA concentrations (6 mM) did not significantly modify the binding properties of the unaf- Bold open circles represent Cys residues that are though to be linked via a disulfide bridge, and black closed circles correspond to Cys residues whose side chains do not participate in disulfide bonding. TM7 residues being the object of this study are located between Ile 276 and Tyr 302 inclusively. Potential N-glycosylation sites are represented by italic numbers.   To gain more insight about the location in the binding pocket of the reporter cysteines accessible to MTSEA, receptor mutants were preincubated with the ligand prior to MTSEA treatment. The cells were then submitted to an acid wash to dissociate the bound ligand, and receptors were then assayed for binding with the radiolabeled antagonist. Fig. 4 shows how preincubating with the antagonist [Sar 1 ,Ile 8 ]AngII essentially protected the mutant receptors A277C (7.28) , V280C (7.31) , T282C (7.33) , A283C (7.34) , and I286C (7.37) from the effect of MTSEA with protection levels ranging from 65 to 90% (Fig. 4). More moderate protection from MTSEA was obtained for mutant receptors A291C (7.42) and F301C (7.52) , which exhibited a 30% protection level.
Altered Accessibility to TM7 Reporter Cysteines of a Constitutively Active AT 1 Receptor-We made use of the constitutively active receptor, hAT 1 -N111G (19), to assess and map the potentially altered accessibility of the engineered cysteines for MTSEA. As can be seen in Fig. 5, sensitivity of this receptor to 2 mM MTSEA is similar to that of the wild type receptor with a reduction of only 7% of total binding. Higher concentrations of  MTSEA decreased ligand binding up to 10% of the untreated receptors (Fig. 5).
As a first step in evaluating the putative altered phenotype of TM7 engineered cysteine on ligand binding following MT-SEA treatment, we determined the pharmacological properties of the 25 receptor mutants bearing cysteines. Within the hAT 1 -N111G receptor background, 19 mutants conserved high binding affinities for the ligand with expression levels varying from 0.4 to 1.4 times that of the hAT 1 -N111G receptor (Table III). We did not detect binding to the I276C (7.27) , D278C (7.29) , M284C (7.35) , P285C (7.36) , I288C (7.39) , and I290C (7.41) receptors engineered in the hAT 1 -N111G receptor background. Surprisingly, the D281C (7.32) receptor mutant exhibited a regain of high affinity binding (0.6 nM) when engineered in the hAT 1 -N111G receptor background as compared with that of the wild type receptor background in which there was no detectable binding (see Table I versus Table III).
Mutant receptors were then treated with increasing concentrations of MTSEA and assessed for binding with the radiolabeled antagonist 125 I-[Sar 1 ,Ile 8 ]AngII. Interestingly, all of the mutant receptors were insensitive to MTSEA, except for the N111G-A291C (7.42) receptor mutant. Those mutants that were sensitive to MTSEA treatment in the WT background (A277C (7.28) , V280C (7.31) , T282C (7.33) , A283C (7.34) , I286C (7.37) , and F301C (7.52) ) conserved a constitutive activity in the hAT 1 -N111G background based on the production of inositol phosphates under basal conditions (data not shown). Thus, the Cys substitutions in the N111G background did not abrogate their intrinsic capacity of activating phospholipase C in an agonistindependent manner. The N111G-A291C (7.42) receptor exhibited 66% binding inhibition following treatment with 2 mM MTSEA (Figs. 5 and 6). This mutant also exhibited an elevated susceptibility to MTSEA when compared with wild type receptor as denoted by its second order reaction rate of 36 Ϯ 11 M Ϫ1 s Ϫ1 (6 M Ϫ1 s Ϫ1 for wild type). We also observed that this particular receptor was very effectively protected from reaction with MTSEA with a protection level evaluated at 88% (Fig. 4). Two other receptor mutants, N111G-T282C (7.33) and N111G-I286C (7.37) , exhibited a mild decrease of ligand binding following MTSEA treatment, producing a maximum of 28 and 25%, respectively, of binding inhibition (Figs. 5 and 6). DISCUSSION The aim of this study, which relied on SCAM analysis, was to gain insight into the orientation of the residues of TM7 within the binding pocket of the AT 1 receptor and to assess the struc-  tural mechanisms underlying receptor activation. Changes in ligand binding were monitored in a wild type receptor (ground state) background and compared with a constitutively active receptor. The SCAM method is based on the reactivity of engineered cysteines to MTSEA, a reagent that reacts a billion times faster with ionized cysteines than with the un-ionized thiol (13) and thus will covalently alkylate any cysteine located in a hydrophilic environment. Indeed lipid-exposed, buried, or disulfide-bonded cysteines are unlikely to ionize to a significant extent and hence are assumed to be unaffected by such modification induced by the MTS reagents. Two criteria were used to establish that an engineered cysteine was accessible in the binding site crevice: 1) MTSEA added extracellularly irreversibly inhibited ligand binding and 2) binding activity was protected from MTSEA by pretreatment with an AT 1 receptor ligand.
On the basis of MTSEA insensitivity of the wild type receptor, it can be suggested that either endogenous cysteines are not alkylated by MTSEA or alkylation of reactive cysteines does not affect ligand binding. Four cysteine residues of the AT 1 receptor are suspected to participate in disulfide bridges linking the N-terminal to ECL3 and ECL1 to ECL2 (20) and thus would not be alkylated. Other cysteines found within various TMs would either face the lipid environment of the cell membrane and be unavailable or would simply not be a deter-minant of the binding pocket conformation. Interestingly, a recent report has revealed that exposition of the AT 1 receptor to MTSEA was able to inhibit ligand binding (10), whereas we did not observe significant binding loss. This discrepancy could be explained by the fact that prior to MTSEA treatment, Miura and Karnik (10) used a cell disruption method, which exposes both the intracellular and the extracellular domains of the receptor to the reagent. Using membrane preparations we also obtained binding inhibition after MTSEA treatment (data not shown). However, in the present study, our methodological approach of adding MTSEA on whole adherent cells expressing the AT 1 receptor will essentially expose only the extracellular ligand-accessible side of the receptor to MTSEA.
The 25 individual mutants incorporating reporter cysteines spanned residues 276 -302. Molecular modeling of AT 1 receptor TM7 by homology to the rhodopsin high resolution crystal structure, suggests that residues 276 -281 are located in an extracellular domain as part of ECL3. MTSEA-sensitive Ala 277(7.28) is positioned in a flexible domain rigidified by a disulfide bridge between the N-terminal tail and ECL3, whereas Val 280(7.31) sits just at the border of TM7, which extends down to Leu 305(7.56) (Fig. 7). Thr 282 (7.33) and Ile 286(7.37) lie on the same helix face with appreciable exposition to a potential hydrophilic pocket, an exposition that seems to be facilitated by the presence of a helical deviation around Phe 293(7.44) and Asn 294(7.45) . Ala 291(7.42) and Phe 301(7.52) lie parallel to each other on the same helix face mostly facing the same area of the binding pocket. Ala 291(7.42) is predicted to face the protein interior, a feature that seems to be highly conserved among GPCRs, because this position also points inside the binding pocket of the D2 dopamine receptor, M3 muscarinic receptor, A2 adenosine receptor, and rhodopsin (21)(22)(23)(24). Phe 301(7.52) is part of a classical ␣-helical segment underlying the conserved NPXXY motif following the helical deviation.
The modeling also shows a regular accessibility pattern exposing one face of the helix to the binding pocket while leaving the other facing away (Fig. 7). Therefore, residues lying on the MTSEA-accessible face are more likely to be involved in ligand binding. Indeed, Phe 293(7.44) , a residue that constitutes a crosslinking site for a C-terminally modified analogue of AngII (25), points toward the same water-accessible crevice identified here (Fig. 7). Also, Phe 301(7.52) has been suggested to form the binding pocket for nonpeptidic antagonists and peptide agonists (26). In contrast, residues that were not identified as being MTSEA-sensitive, such as Asn 295(7.46) (27) and Tyr 292(7.43) (28), have been shown to affect receptor activation rather than peptide binding and would therefore lie on the opposite face of the helix (Fig. 7). Furthermore, endogenous TM7 Cys residues (Cys 289 (7.40) and Cys 296(7.47) ) are located on the opposite, lipidexposed face of the helix (Fig. 7), a feature that could be responsible for their poor reactivity with MTSEA as demonstrated in the present study.
Interestingly, most of the MTSEA-accessible residues that have been identified by our SCAM analysis (A277C (7.28) , V280C (7.31) , T282C (7.33) , A283C (7.34) , and I286C (7.37) ) lie toward the top portion of the transmembrane domain, possibly close to the interface between lipid and the extracellular milieu (Fig. 7). This result can be partly explained by the fact that the residues necessary for an appropriate interaction between the ligand and the receptor are mainly located within this interface (29 -31). Thus, these residues would form the top of the binding pocket of the receptor with the result that their alkylation with MTSEA may produce a steric hindrance, thereby impeding efficient binding of the ligand. After identifying A291C (7.42) as a sensitive residue, we then obtained a stretch of nine unaffected residues before acquiring sensitivity with F301C (7.52) . A possible explanation for the poor MTSEA-mediated effect on ligand binding between Ala 291(7.42) and Phe 301(7.52) comes from molecular modeling of the AT 1 receptor (Fig. 7). When comparing TM7 to the accessibility pattern brought by the SCAM study, those nine unaffected residues perfectly superimpose with a conserved helical deviation at the center of the transmembrane domain. This helix deformation is consistent with observations made for rhodopsin (24) and has already been mapped by SCAM studies in TM7 of the D2 dopamine receptor (21). This particular structure probably lengthens the spatial proximity between residues Tyr 292(7.43) -Leu 300(7.51) and the ligand-containing binding crevice. It is therefore possible that 1) residues in the Tyr 292(7.43) -Leu 300(7.51) stretch do not react with MTSEA because of their relative inaccessibility from the binding pocket or 2) residues Tyr 292(7.43) -Leu 300(7.51) react with MTSEA but do not contribute in forming the binding pocket to disturb ligand binding. However, the first hypothesis seems more probable because we had previously shown that Phe 8 of AngII interacts with Phe 293(7.44) and Asn 294(7.45) of the AT 1 receptor (25), therefore supporting the contention that the C terminus of AngII dives deep inside the receptor-binding pocket to initiate agonist-mediated receptor activation. Our data using protection assays and reaction rates of MT-SEA also support the notion that specific residues within TM7 contribute to forming the binding pocket. Indeed, the most reactive and possibly very accessible Cys residues (A277C (7.28) , V280C (7.31) , and T282C (7.33) ) were highly protected by the presence of the ligand, whereas the A291C (7.42) and F301C (7.52) mutants, although less reactive, remained nonetheless protected. However, protection does not mean that a particular residue makes contact with the ligand. We cannot rule out the possibility of indirect effects through propagated structural changes that would make a reporter Cys more or less accessible upon ligand binding.
To further probe into the mechanisms by which receptors undergo structural changes from the inactive to the active state, we took advantage of the constitutively active hAT 1 -N111G receptor. It is believed that the isomerization of conformers toward the active state, which involves transmembrane movement, is stabilized by agonist binding and would be mimicked in part by the constitutively active receptor (1,33). Thus, we verified accessibility of TM7 residues within the structural background of the hAT 1 -N111G receptor to MTSEA and compared the pattern to the one obtained for the wild type receptor (Fig. 8). We found that those Cys residues that were unresponsive to MTSEA (in regard to binding inhibition) kept the same behavior in the N111G mutant receptor (Fig. 8). Interestingly, those Cys residues (A277C (7.28) , V280C (7.31) , T282C (7.33) , A283C (7.34) , I286C (7.37) , and F301C (7.52) ) that have been shown to react with MTSEA in the wild type receptor and therefore affected ligand binding (Fig. 8A), were either much less or no longer reactive when engineered in the hAT 1 -N111G receptor background where constitutive activity was main-FIG. 7. Molecular modeling of hAT 1 receptor TM7 residues by homology to bovine rhodopsin. Schematic model of the AT 1 receptor as generated by homology modeling showing the MTSEA-sensitive AT 1 mutant receptors bearing reporter cysteines in TM7. Residues ranging from Ala 277 (7.28) to Tyr 302(7.53) are shown. White residues represent Cys substitution that has yielded no inhibition following MTSEA treatment, whereas orange residues correspond to MTSEA sensitive substitutions identified by our SCAM analysis. Green residues represent a segment of the AT 1 receptor encompassing helix 8. tained (Fig. 8B). This led us to hypothesize that either accessibility of TM7 residues or their spatial proximity within the binding pocket has been drastically altered in the presence of the mutation of residue 111, which is found in TM3 (Fig. 9). This would imply that, in this model of receptor activation and as a direct consequence of transmembrane movement (translation), certain residues of TM7 would distance themselves from the binding pocket (Fig. 9).
Moreover, we observed that the N111G-A291C (7.42) mutant exhibited a greatly induced reaction rate with MTSEA and an increased extent of binding inhibition after MTSEA reaction compared with the same mutant engineered in the wild type receptor background (Fig. 3 versus Fig. 6). A possible explanation might be that some helix rotation has occurred concomitantly with the proposed translation, therefore making the Cys 291(7.42) side chain more exposed to the binding pocket in the N111G receptor. A similar translation/rotation movement has been documented for a constitutively active ␤ 2 -adrenergic receptor and light-activated rhodopsin whereby a significant translation of TM3 occurs away from TM6 on the cytoplasmic receptor interface while both TMs rotate on the extracellular interface (4, 34 -36). More recently, using the SCAM approach, a rigid body rotation of TM2 has been described for the constitutively activated AT 1 receptor bearing mutations at residue Asn 111(3.35) (10).
We further demonstrate that the MTSEA-mediated binding inhibition on a receptor mutant for which Phe 301(7.52) has been substituted for Cys is greatly altered in a constitutively active mutant of the AT 1 receptor when compared with the same mutation engineered in the wild type background (Figs. 6 and 3). Phe 301(7.52) is part of a highly conserved motif among the rhodopsin-like family of GPCRs, the 7.49 NPXXY 7.53 motif, involved in activation of numerous GPCRs including the AT 1 receptor (11,(37)(38)(39). However, its precise role in receptor activation is still obscure. It has been suggested that the NPXXY motif could be part of an intramolecular bonding network implicating interhelical interactions between TM2, TM3, and TM7. Indeed, the recently elucidated structure of bovine rhodopsin shows that the complex responsible for those interhelical interactions includes residues Gly 120(3.35) -Asp 83(2.50) -Asn 302(7.49) (24,40). Such a complex is well conserved in the AT 1 receptor, implicating the corresponding residues Asn 111(3.35) -Asp 74(2.50) -Asn 298 (7.49) . Studies have revealed that Asn 111 (3.35) could interact with Tyr 292 (7.43) in the basal state of the receptor and that AngII binding would favor new interactions between Asn 111 (3.35) and Tyr 4 of AngII; Tyr 292(7.43) would then be hydrogen-linked to Asp 74(2.50) in the activated receptor state (41). Furthermore, mutations at Asn 111(3.35) would impede the network between TM3 and TM7, resulting in a more relaxed receptor leading to constitutive activity (9). Our data on the altered behavior of the F301C (7.52) treated with MTSEA would support the idea that the N111G receptor achieves its constitutive activity partly by altering the NPXXY conformation.
Our data comparing the AT 1 receptor ground state versus an activated state strongly point toward an altered spatial proximity of TM7 with regards to the binding pocket which exposes TM7 residues Ala 277(7.28) , Val 280(7.31) , Thr 282(7.33) , Ala 283(7.34) , Ile 286(7.37) , Ala 291(7.42) , and Phe 301(7.52) to a water-accessible crevice. This movement of TM7 for an activated GPCR is reminiscent of recent observations on rhodopsin (5) and is thought to define an important structural mechanism supporting recep- FIG. 9. Proposed structurally related activation mechanisms for the hAT 1 receptor. Extracellular view of AT 1 seven transmembrane domains based on the arrangement described for bovine rhodopsin (24). Residues that affect ligand binding once treated with MTSEA span a fixed area (ϳ180°) of TM7 symbolized by a black-filled halfcircle. The water-accessible crevice forming the binding pocket is suggested to be located between TMs 1, 2, 3, 5, 6, and 7. In the ground state, constraining intramolecular interactions help to keep the receptor in a basal state where functional coupling with G protein is kept at its minimum for the system to remain sensitive to ligand stimulation. Some of those constraining interactions in the case of the AT 1 receptor are thought to be generated between TM2, TM3, and TM7 (32). Following agonist binding, intramolecular interactions are broken, and GPCRs go through a series of conformational rearrangements implicating TMs. In the present study we propose that TM7 could be excluded from the water-filled crevice forming the binding pocket following activation of the AT 1 receptor .   FIG. 8. Helical wheel representation of TM7 reporter Cys residues and their pattern of reactivity to MTSEA. Positions in TM7 of MTSEA reacted Cys residues affecting [Sar 1 ,Ile 8 ]AngII binding are shown in a helical wheel representation viewed from the extracellular side for receptors with no additional mutation in TM3 (A) and for receptors in which Asn 111 (3.35) has been mutated for Gly in addition to reporter Cys in TM7 (B). Black closed circles correspond to Cys that have been shown in this study to inhibit 50% or more of 125 I-[Sar 1 ,Ile 8 ]AngII binding when reacted with MTSEA, whereas gray closed circles indicate those that inhibited ligand binding at 25% or more. White circles indicate those mutant receptors that shown no effect on ligand binding when reacted with MTSEA or positions resulting in low or undetectable binding when substituted for Cys residues. tor activation mediated by AngII. This could be a common feature found in numerous rhodopsin-like GPCRs.