Analysis of Transmembrane Domains 1 and 4 of the Human Angiotensin II AT1 Receptor by Cysteine-scanning Mutagenesis*

The octapeptide hormone angiotensin II (AngII) exerts a wide variety of cardiovascular effects through the activation of the AT1 receptor, which belongs to the G protein-coupled receptor superfamily. Like other G protein-coupled receptors, the AT1 receptor possesses seven transmembrane domains that provide structural support for the formation of the ligand-binding pocket. Here, we investigated the role of the first and fourth transmembrane domains (TMDs) in the formation of the binding pocket of the human AT1 receptor using the substituted-cysteine accessibility method. Each residue within the Phe-28(1.32)–Ile-53(1.57) fragment of TMD1 and Leu-143(4.40)–Phe-170(4.67) fragment of TMD4 was mutated, one at a time, to a cysteine. The resulting mutant receptors were expressed in COS-7 cells, which were subsequently treated with the charged sulfhydryl-specific alkylating agent methanethiosulfonate ethylammonium (MTSEA). This treatment led to a significant reduction in the binding affinity of TMD1 mutants M30C(1.34)-AT1 and T33C(1.37)-AT1 and TMD4 mutant V169C(4.66)-AT1. Although this reduction in binding of the TMD1 mutants was maintained when examined in a constitutively active receptor (N111G-AT1) background, we found that V169C(4.66)-AT1 remained unaffected when treated with MTSEA compared with untreated in this context. Moreover, the complete loss of binding observed for R167C(4.64)-AT1 was restored upon treatment with MTSEA. Our results suggest that the extracellular portion of TMD1, particularly residues Met-30(1.34) and Thr-33(1.37), as well as residues Arg-167(4.64) and Val-169(4.66) at the junction of TMD4 and the second extracellular loop, are important binding determinants within the AT1 receptor binding pocket but that these TMDs undergo very little movement, if at all, during the activation process.

The octapeptide hormone angiotensin II (AngII) 4 is the active component of the renin-angiotensin system. It exerts a wide variety of physiological effects, including vascular con-traction, aldosterone secretion, neuronal activation, and cardiovascular cell growth and proliferation (1,2). Virtually all the known physiological effects of AngII are produced through the activation of the AT 1 receptor, which belongs to the G proteincoupled receptor (GPCR) superfamily (3,4). The AT 1 receptor belongs to the rhodopsin-like family A of G protein-coupled receptors, which have a seven-transmembrane helix structure, an extracellular N-terminal tail, an intracellular C-terminal domain, and three extracellular and three intracellular loops.
The seven transmembrane domains (TMDs) of GPCRs constitute structural support for signal transduction. Like other family A GPCRs such as rhodopsin and adrenergic receptors, the AT 1 receptor undergoes spontaneous isomerization between its inactive state and its active state (5). Movement of TMD helices through translational or rotational displacement is believed to be essential to achieve the active state (6 -8). These conformational changes would sustain GTP/GDP exchange on specific guanine nucleotide-binding proteins (G proteins) leading to activation of intracellular signaling cascades (5). For the AT 1 receptor, it has been proposed that TMD3, TMD5, TMD6, and TMD7 may participate in the activation process by providing a network of interactions throughout the AngII-binding pocket (9). The dynamics of this network would require that, following agonist binding, novel or existing interactions between the TMDs would either be created or broken, respectively.
Based on homology with the recent high resolution structures of the ␤1 adrenergic, ␤2 adrenergic, and A 2A adenosine receptors (10 -12) it was expected that the binding site of the AT 1 receptor would be formed between its seven, mostly hydrophobic transmembrane domains and would be accessible to charged water-soluble agonists, like AngII. For this receptor, the binding site would thus be contained within a hydrophilic crevice, the binding pocket, extending from the extracellular surface of the receptor to the transmembrane portions. Because all crystallized structures to date suggest that TMD1 and TMD4 are somewhat removed from the binding pocket (13), these TMDs have not yet received extensive consideration as contributors to the formation of the ligand binding pocket. Nonetheless, experiments have revealed the proximity of TMD4 residues to the ligand-binding pocket in the ␤1 and ␤2 adrenergic receptors as well as the dopamine D2 receptor (14,15). Moreover, numerous mutagenesis studies have provided the basis for a model in which an interaction between Asn-111 in TMD3 and Tyr-292 in TMD7 maintains the AT 1 receptor in the inactive conformation. The agonist AngII would disrupt this interaction and promote the active conformational state (16). In support of this model, it was further shown that substitution of Asn-111 for a residue of smaller size (Ala or Gly) confers constitutive activity on the AT 1 receptor (17)(18)(19).
The substituted-cysteine accessibility method (SCAM) (20 -22) is an ingenious approach for systematically identifying residues in a transmembrane domain (TMD) that contribute to the binding pocket of a GPCR. MTS reagents react 10 9 faster with ionized thiolates (S Ϫ ) than with un-ionized thiols (SH) (23), and ionization of cysteine occurs to a significant extent only in the aqueous medium (24). Thus, in TMDs, the sulfhydryl of a cysteine residue, which is introduced by mutagenesis one at a time, facing toward the binding pocket should react much faster with charged sulfhydryl-specific reagents such as positively charged methanethiosulfonate ethylammonium (MTSEA) than sulfhydryls facing toward the interior of the protein or the lipid bilayer (25). We use two criteria for identifying an engineered cysteine as forming the surface of the binding site crevice: (i) the reaction with a methanethiosulfonate reagent alters binding irreversibly; (ii) this reaction is retarded by the presence of ligand. We previously used this approach to identify the residues in TMD2, TMD3, TMD5, TMD6, and TMD7 that form the surface of the binding site pocket in the wild-type AT 1 receptor and in the constitutively active N111G-AT 1 receptor (26 -30). Here, we report the application of SCAM to probe both TMD1 and TMD4 in the wild-type and constitutively active AT 1 receptor backgrounds.
Numbering of Residues in TMDs-Residues in TMD1 and TMD4 of the human AT 1 receptor were given two numbering schemes. First, residues were numbered according to their positions in the human AT 1 receptor sequence. Second, residues were also indexed according to their position relative to the most conserved residue in the TMD in which they are located (32). By definition, the most conserved residue was assigned the position index "50," e.g. in TMD4, Trp-153 is the most conserved residue and was designated Trp-153 (4.50) , whereas the downstream residue was designated Leu-154 (4.51) and the upstream residue was designated Ile-152 (4.49) . 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 recep-  tor with the overlap PCR method (Expand high fidelity PCR system, Roche Applied Science). Briefly, forward and reverse oligonucleotides were constructed to introduce cysteine mutations between Phe-28 (1.32) and Ile-53 (1.57) for TMD1 and between Leu-143 (4.40) and Phe-170 (4.67) for TMD4, as well as to substitute residue Arg-167 (4.64) for either Lys, His, Asp, Glu, or Ile. PCR products were subcloned into the HindIII-XbaI sites of the mammalian expression vector pcDNA3.1. Site-directed mutations were then confirmed by automated DNA sequencing by aligning the AT 1 sequence with multiAlin (33). Cell Culture and Transfections-COS-7 cells were grown in Dulbecco's modified Eagle's medium 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 the cells reached ϳ90% confluency, they were transfected with 4 g of plasmid DNA and 15 l of Lipofectamine TM 2000 according to the manufacturer. After 24 h, transfected cells were trypsinized, distributed into 12-well plates, and grown for an additional 24 h in complete Dulbecco's modified Eagle's medium containing 100 IU/ml penicillin and 100 g/ml streptomycin, before MTSEA treatment or binding assays.
Binding Experiments-COS-7 cells were grown for 36 h after transfection in 100-mm culture dishes, washed once with PBS, and subjected to one freeze-thaw cycle. Broken cells were then gently scraped into 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% bovine serum albumin, 0.01% bacitracin, 0.01% soybean trypsin inhibitor). Saturation binding experiments were done by incubating broken cells (20 -  The numbers indicate the positions of cysteines and other residues in the receptor. The gray closed circles represent cysteine residues that are thought to be linked via disulfide bridges, and the black closed circles represent cysteine residues whose side chains do not form a disulfide bridge. Mutated TMD1 residues are located between Phe-28 (1.32) and Ile-53 (1.57) inclusively while mutated TMD4 residues are located between Leu-143 (4.41) and Phe-170 (4.66) inclusively. Potential N-glycosylation sites (Asn-4, Asn-176, and Asn-188) are indicated. Asn-111 in TMD3 is also shown in gray.
bound radioactivity was evaluated by ␥ counting. Results are presented as means Ϯ S.D. Binding data (B max and K d ) were analyzed with Prism version 5.0 for Windows (GraphPad Software, San Diego CA), using a one-site binding hyperbola nonlinear regression analysis.
Treatment with MTSEA-The MTSEA treatment was performed according to the procedure of Javitch et al. (21), with minor modifications. Two days after transfection, cells grown in 12-well plates were washed with PBS and incubated for 3 min at room temperature with freshly prepared MTSEA at the desired concentrations (typically from 0.5 to 6 mM) in a final volume of 200 l. The reaction was stopped by washing the cells with ice-cold PBS. Intact cells were then incubated in binding medium (Dulbecco's modified Eagle's medium, 25 mM HEPES, 0.1% bovine serum albumin, pH 7.4) containing 0.05 nM 125 I [Sar-1,Ile-8]AngII for 90 min at room temperature. After wash-ing with ice-cold PBS, the cells were lysed with 0.1 N NaOH, and the radioactivity was evaluated by ␥ counting. The percentage of fractional binding inhibition was calculated as (1 Ϫ (specific binding after the MTSEA treatment/specific binding without the treatment)) ϫ 100.
Protection against MTSEA Reaction by [ 1 Receptors Bearing Cysteines in TMD1 and TMD4-To identify the residues in TMD1 and TMD4 that face the binding pocket of the AT 1 receptor, we mutated 26 consecutive residues between Phe-28 (1.32) and Ile-53 (1.57) of TMD1 and 28 consecutive residues between Leu-143 (4.40) and Phe-170 (4.67) of TMD4 to cysteine, one at a time. Each mutant receptor was transiently expressed in COS-7 cells. To assess the conservation of global conformation of these receptors after such substitution, pharmacological parameters describing the equilibrium binding of the radiolabeled ligand 125 I-[Sar-1,Ile-8]AngII such as K d and B max were determined (Tables 1  and 2  The vertical line represents an arbitrary threshold used to identify cysteine-sensitive mutants. It was set at a value corresponding to binding inhibition 20% greater than the value for the N111G-AT 1 receptor. The white bars indicate mutant receptors for which binding activities were not appreciably reduced when compared with the wild-type receptor after treatment with MTSEA. The black bars indicate mutant receptors for which binding activities were reduced after treatment with MTSEA. Each bar represents the means Ϯ S.D. of data from at least three independent experiments. F170C (4.67) -AT 1 did not display any detectable binding activity and were therefore not used for SCAM analysis.

Binding Properties of Mutant AT
Effect of Extracellularly Added MTSEA on Binding Properties of Mutant Receptors in the Wild-type Background-To verify whether the reporter cysteines introduced into either TMD1 or TMD4 were oriented toward the binding pocket, mutant receptors were treated with concentrations of MTSEA varying between 0.5 and 6 mM. As reported previously (26), the various concentrations of MTSEA had very little effect (no more than a 20% reduction at high MTSEA concentrations) on the binding properties of the wild-type AT 1 receptor, which contains 10 endogenous cysteines (Fig. 1). For TMD1 mutants, Fig. 2 shows that MTSEA at 0.5 mM or higher concentration strongly inhibited ligand binding toward mutants M30C (1.34) -AT 1 (binding inhibition of 69%) and T33C (1.37) -AT 1 (binding inhibition of 48%). For TMD4 mutants, Fig. 3 shows that MTSEA at 0.5 mM or higher concentration inhibited ligand binding toward mutant V169C (4.66) -AT 1 (binding inhibition of 34%). The bind-ing properties of all other mutant receptors were not significantly affected by MTSEA treatment.
Mutant receptors were subsequently treated with increasing concentrations of MTSEA and assessed for binding with 125 I-[Sar-1,Ile-8]AngII. Like the wild-type receptor, the N111G-AT 1 receptor was relatively insensitive to a 3-min treatment with MTSEA concentrations ranging from 0.5 to 2 mM, again indicating the relatively low contribution of the endogenous cysteines in the binding site pocket (Fig. 4). For TMD1 mutants in the N111G background, Fig. 4 shows that treatment with 0.5 mM MTSEA strongly inhibited the 125 I-[Sar-1,Ile-8]AngII binding properties of the M30C (1.34) -N111G-AT 1 mutant (binding inhibition of 64%). At 2 mM MTSEA, the binding properties of T33C (1.37) -N111G-AT 1 were significantly inhibited (binding inhibition of 45%). The L48C (1.52) -AT 1 mutant receptor, which displayed sensitivity to 2 mM MTSEA in the ground state, did not demonstrate any sensitivity to MTSEA in the constitutively active state. For TMD4 mutants in the N111G background, all mutant receptors were insensitive to MTSEA treatment (Fig. 5). It is interesting to note that V169C (4.66) , which was sensitive to MTSEA in the inactive ground state receptor background, lost its sensitivity in the N111G receptor background.
Contribution of Position Arg-167 (4.64) to AT 1 Receptor Binding-To assess the contribution of the positively charged side chain of Arg-167 (4.64) in 125 I-[Sar-1,Ile-8]AngII binding to the AT 1 receptor, we tested the binding affinities of the R167H, R167K, R167D, R167E, and R167I mutant receptors. We found that, with the exception of the R167K mutant, all mutant receptors did not display any binding activity (Table 5). In view of the similarity of the structure of MTSEA (SCH 2 CH 2 NH 3 ϩ ) to that of the arginine side chain, we treated both the R167C-AT 1 and R167C-N111G-AT 1 mutant receptors with 0.5 mM MTSEA before proceeding with the binding assay (Table 5). We found that binding affinities were restored by the MTSEA treatment for both mutants, although the ground state receptor still displayed a 6-fold reduction of binding when compared with the wild-type hAT 1 receptor (Table 5).
Protection Assay-To confirm that reporter cysteines accessible to MTSEA are located within the binding pocket, receptor mutants were saturated with the competitive ligand [Sar-1,Ile-8]AngII prior to MTSEA treatment. Cells were then washed with an acid buffer to dissociate bound ligand. The receptors were then assayed for binding with the radiolabeled competitive ligand. Fig. 6 shows how preincubating with the competitive ligand [Sar-1,Ile-8]AngII protected mutant receptors M30C (1.34) and T33C (1.37) from the inhibitory effect of MTSEA, in both the wild-type and N111G backgrounds, with pro-tection levels ranging from 45% to 60%. However, we found that preincubation with [Sar-1,Ile-8]Ang II weakly protected the MTSEA-sensitive V169C (4.66) mutant receptor (13% inhibition).

DISCUSSION
The rationale of this study, which relied on SCAM analyses, was to gain insight into the orientation of both TMD1 and TMD4 of the hAT 1 receptor by identifying the residues accessible to MTSEA within the binding site pocket. Mapping these residues in the ground state receptor and the constitutively active N111G background allowed us to measure relative changes in the position of certain residues, thus providing valuable information with which to infer structural changes underlying the activation of the AT 1 receptor. The SCAM method is based on the reactivity of engineered cysteines to MTSEA, a reagent that reacts 10 9 times faster with ionized cysteines than with the un-ionized thiol (23) 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 MTSEA.
In the initial projection map of rhodopsin (34), both TMD1 and TMD4 appeared as outliers of the binding pocket, with a large lipid-exposed surface and few polar residues (13). Also, all crystallized GPCR structures to date suggest that both these TMDs are somewhat removed from the binding pocket and thus have not been the subject of intense scrutiny. Despite this, a previous report has indicated that important peptide ligand-   JANUARY 22, 2010 • VOLUME 285 • NUMBER 4 binding determinants are located within the AT 1 receptor N terminus, in particular adjacent to the top of TMD1 (35). For a number of class A GPCRs, many site-directed mutagenesis studies had indicated that several residues spanning positions 4.61 to 4.72 contribute to agonist and/or antagonist binding, suggesting a role for the extracellular portion of TMD4 in ligand binding (35)(36)(37)(38)(39)(40)(41)(42)(43)(44). A recent study on rhodopsin using photoreactive chromophores has shown that position 4.58 of TMD4 is directly labeled following photoactivation, implying a role for TMD4 during rhodopsin activation (45).

SCAM Method to Study the AT 1 Ligand-binding Pocket
In this study, a surprising number of mutants in which residues (V26C, I27C, F28C, V29C, I31C, P32C, and L35C) were replaced with Cys did not show any detectable binding toward [Sar-1,Ile-8]Ang II. Although this result suggests that these positions are involved in ligand binding, such a binding mode in which simultaneous direct contact between all these residues and the Ang II ligand occurs is unlikely. Although a previous report (46) proposed that residue Asp-1 of AngII interacts with the N-terminal domain proximal to TMD1, a direct interaction between Asp-1 and a residue of this segment has not yet been directly proven. Therefore, a more plausible explanation for the lack of ligand binding would be that this stretch of residues is not interacting directly with the ligand but instead are involved in proper receptor folding and assembly (47).
As reported previously, the insensitivity of the wild-type receptor to MTSEA suggests either that endogenous cysteines are not alkylated by MTSEA or that their alkylation does not affect the binding of the ligand (26). Our approach of adding the MTSEA reagent to whole adherent cells expressing the AT 1 receptor essentially exposed only the extracellular ligand-accessible side of the receptor to MTSEA. For both TMD1 and TMD4, the MTSEA-accessible residues that we identified with the SCAM approach are located at the top (M30C (1.34) , T33C (1.37) , and V169C (4.66) ) portion of the TMDs. These results suggest that this portion of both TMDs is involved in the interaction with the ligand and are part of the binding pocket in the ground state of the receptor. Indeed, by a mechanism that could be steric, electrostatic, or indirect, the alkylation of these residues with MTSEA hampered the binding of the ligand. We found that the competitive ligand [Sar-1,Ile-8]AngII protected mutants M30C (1.34) and T33C (1.37) from the effect of MTSEA and mutant V169C (4.66) was very weakly protected (Fig. 6). We conclude that Val-169 (4.66) is likely located at the margin of the binding pocket such that the presence of [Sar-1,Ile-8]Ang II in the binding pocket cannot block the access of MTSEA, resulting in weak protection.
To further investigate the mechanism by which the AT 1 receptor undergoes structural changes during the transition from its inactive to its active state, we took advantage of the constitutively active N111G-AT 1 receptor. It is believed that the isomerization of conformers toward the active state, which involves transmembrane movement, is stabilized by the binding of an agonist and would be mimicked, at least in part, by the constitutively active receptor (5,48). Thus, within the structural background of the N111G-AT 1 receptor, we verified the accessibility of both TMD1 and TMD4 residues to MTSEA, and we compared the pattern obtained with that of the wild-type receptor. For TMD1, we found that Cys-substituted mutants M30C (1.34) -N111G-AT 1 and T33C (1.37) -N111G-AT 1 maintained their sensitivity to MTSEA in the constitutively active The vertical line represents an arbitrary threshold used to identify cysteine-sensitive mutants. It was set at a value corresponding to binding inhibition 20% greater than the value for the N111G-AT 1 receptor. The white bars indicate mutant receptors for which binding activities were not appreciably reduced when compared with that of the N111G-AT 1 receptor after treatment with MTSEA. Each bar represents the means Ϯ S.D. of data from at least three independent experiments. receptor background, whereas for TMD4 we found that mutant Val-169 (4.66) -N111G-AT 1 lost its sensitivity to MTSEA (Fig. 5). In the protection assay, the competitive ligand [Sar-1,Ile-8]AngII offered effective protection to both sensitive mutants M30C (1.34) -N111G-AT 1 and T33C (1.37) -N111G-AT 1 against the alkylating effect of MTSEA (Fig. 6). For TMD4, the divergence in the sensitivity of Cyssubstituted mutants in the wildtype background and in the N111G-AT 1 receptor background for position Val-169 (4.66) suggests that the accessibility of this residue and its spatial proximity within the binding pocket was altered due to the single substitution of an asparagine for a glycine at position 111 in TMD3. It should be noted that position 4.66 is located at the junction between TMD4 and the second extracellular loop of AT 1 . Interestingly, the second extracellular loop of rhodopsin has recently been shown to move away from the binding pocket following activation, highlighting a possible role for this ECL in the GPCR activation process (49). However, for TMD1, we detected no such divergence in sensitivities between both receptor backgrounds, which indicates no major structural changes of TMD1 during activation.
A peculiar observation was made regarding the R167C (4.64) -AT 1 mutant receptor. As noted we found that this mutant was unable to bind [Sar-1,Ile-8]AngII either in the inactive or constitutively active state. But because the structure of MTSEA (SCH 2 CH 2 NH 3 ϩ ) is reminiscent of the Arg residue side chain, we thought of testing whether we could rescue binding by treating cells expressing the mutant receptor with the charged MTSEA. Indeed, binding was restored partly (i.e. K d of 5.45 nM) in the "wild-type" background and completely in the constitutively active background state. To determine the contribution of Arg-167 (4.64) in the ligand binding of the AT 1 receptor, we also replaced Arg with His, Lys, Glu, Asp, and Ile and found that, except for the R167K mutant receptor, substitution at this position totally abolished the binding affinity to [Sar-1,Ile-8]Ang II. These results suggest the loss of binding affinity of the R167C (4.64) mutant receptor is due to the loss of an electrostatic charge that contributes to AT 1 binding, either by direct interaction with the ligand or by an indirect effect on receptor conformation. Although the His residue at position 167 may also possess a positive charge under physio- The vertical line represents an arbitrary threshold used to identify cysteine-sensitive mutants. It was set at a value corresponding to binding inhibition 20% greater than the value for the N111G-AT 1 receptor. The white bars indicate mutant receptors for which binding activities were not appreciably reduced when compared with that of the N111G-AT 1 receptor after treatment with MTSEA. Each bar represents the means Ϯ S.D. of data from at least three independent experiments.

TABLE 5 Binding properties of R167C and N111G-R167C mutant receptors
Cells transfected with the appropriate receptor were assayed as described under "Experimental Procedures." Cells expressing the R167C (4.64) -AT 1 and N111G-R167C (4.64) -AT 1 mutant receptors were preincubated for 3 min at room temperature with freshly prepared 0.5 mM MTSEA in a final volume of 2 ml prior to the freeze-thaw cycle. Binding affinities (K d ) and maximal binding capacities (B max ) are expressed as the means Ϯ S.D. of values obtained in n independent experiments performed in triplicate.  logical conditions, it may be the lack of side-chain extension into the binding pocket that would explain the loss of binding affinity for the R167H mutant receptor. Therefore, it is reasonable to infer that residue Arg-167 (4.64) is located in the wateraccessible binding pocket by its accessibility to MTSEA. In support of the importance of this position, one report suggested that the side chain of the Arg-167 (4.64) of TMD4 may be involved in an NH-aromatic interaction with the phenolic side chain of the Tyr-4 residue of AngII (50). However, the report that the replacement of Arg with Gln totally abolished the binding affinity excludes the potential involvement of NH-aromatic interaction (51). The presence of an ionic bond linking residue Arg-167 (4.64) with residue Asp-263 (6.58) of TMD6 was also proposed (9).
In conclusion, we have identified specific residues at the top of both TMD1 and TMD4 that participate in the formation of the ligand-binding pocket of the AT 1 receptor. Our data comparing the ground state versus an activated state of the AT 1 receptor imply that these TMDs undergo little or no movement during the AT 1 activation process. Taken together with crystallographic data of other GPCRs, as well as data using SCAM that has been obtained from our group on the movements of other TMDs of AT 1 , the results presented here will provide a framework in which to describe the dynamics of AT 1 activation.