Identification of angiotensin II-binding domains in the rat AT2 receptor with photolabile angiotensin analogs.

To identify binding domains between angiotensin II (AngII) and its type 2 receptor (AT2), two different radiolabeled photoreactive analogs were prepared by replacing either the first or the last amino acid in the peptide with p-benzoyl-L-phenylalanine (Bpa). Digestion of photolabeled receptors with kallikrein revealed that the two photoreactive analogs label the amino-terminal part of the receptor within the first 182 amino acids. Digestion of 125I-[Bpa1]AngII·AT2 receptor complex with endoproteinase Lys-C produced a glycoprotein of 80 kDa. Deglycosylation of this 80-kDa product decreased its apparent molecular mass to 4.6 kDa and further cleavage of this 4.6-kDa product with V8 protease decreased its molecular mass to 3.6 kDa, circumscribing the labeling site of 125I-[Bpa1]AngII within amino acids 3-30 of AT2 receptor. Treatment of 125I-[Bpa8]AngII·AT2 receptor complex with cyanogen bromide produced two major receptor fragments of 3.6 and 2.6 kDa. Cyanogen bromide hydrolysis of a mutant AT2 receptor produced two major fragments of 12.6 kDa and 2.6 kDa defining the labeling site of 125I-[Bpa8]AngII within residues 129-138 of AT2 receptor. Our results indicate that the amino-terminal tail of the AT2 receptor interacts with the amino-terminal end of AngII, whereas the inner half of the third transmembrane domain of AT2 receptor interacts with the carboxyl-terminal end of AngII.

The octapeptide angiotensin II (AngII) 1 recognizes two distinct types of receptors on target cells: the type 1 receptor (AT 1 ) and the type 2 receptor (AT 2 ). The AT 1 receptor mediates all the known physiological actions of AngII including regulation of blood pressure and water and electrolyte balance (1). The functional roles of the AT 2 receptor are not well defined yet but recent studies suggest that it could act as a physiological antagonist of AT 1 -mediated pressor effect and also regulate central nervous system functions related to locomotion and exploratory behavior (2,3). Other studies also suggest that the AT 2 receptor inhibits cell proliferation and induces cell death (4,5). AT 1 and AT 2 receptors have been cloned from several species. They are members of the G protein-coupled receptor superfamily, which is characterized by seven putative transmembrane helices. AT 1 and AT 2 receptors display a low degree (33%) of amino acid sequence similarity (6 -9).
The elucidation of primary structures of numerous G protein-coupled receptors has prompted investigators to look for and identify domains in receptors directly involved in ligand binding. Most of this work was done on members of the ␤-adrenergic receptor family which bind bioamines (Ͻ0.2 kDa) in the outer third of the plasma membrane between transmembrane helices (10). Ligand-binding domains have also been identified for larger agonist (Ͼ10 kDa) such as thyroid-stimulating hormone and follicule-stimulating hormone which bind to the extracellular amino-terminal region of their G proteincoupled receptors (11). The ligand-binding domains of G protein-coupled receptors for small bioactive peptides (0.5-5 kDa) have not been fully characterized. Recent studies suggest, however, that certain proximal loop regions as well as transmembrane regions may be important binding determinants (12)(13)(14)(15)(16).
The localization of ligand-binding domains in the G proteincoupled receptor family has been mostly studied using approaches such as, site-directed mutagenesis, deletion analysis, and construction of chimeric receptors (17). Since these mutations may affect hormone binding indirectly by altering the conformation of a receptor or its expression at the plasma membrane, a more direct approach for the identification of the AT 2 receptor ligand-binding domains should be envisaged. We previously reported the covalent labeling of the AT 2 receptor with the photoreactive AngII analog [Bpa 8 ]AngII (18 -20). In the present study, another high affinity photoreactive analog was prepared by replacing the amino-terminal end of AngII with Bpa. These two photoreactive AngII analogs were used to label the AT 2 receptor of PC-12 cells. The peptide-binding domains of the receptor were identified with each ligand after targetted enzymatic and chemical fragmentation. culture media were obtained from Life Technologies, Inc. [Bpa 1 ]AngII and [Bpa 8 ]AngII were synthesized in our laboratories by the solid phase method and purified by high performance liquid chromatography as described (21). 125 I-AngII, 125 I-[Bpa 1 ]AngII, and 125 I-[Bpa 8 ]AngII (specific radioactivities ϳ1000 Ci/mmol) were prepared with IODO-GEN as described by Fraker and Speck (22). Briefly, 50 l of the peptide solution (0.2 mM) was incubated with 5 g of IODO-GEN (Pierce Chemical Co.), 1 mM Na 125 I (2200 Ci/mmol), 10 l of acetic acid (2 M), and 30 l of water for 30 min at room temperature. The labeled peptides were purified by high performance liquid chromatography on a C-18 column (10 m) (Allteck Associates Inc.; number 29004) with a 20 -40% acetonitrile gradient. The specific radioactivity of the labeled hormones was determined by self-displacement and saturation binding analysis.
Site-directed Mutagenesis-pcDNA1 containing the rat AT 2 receptor cDNA clone was digested with HindIII and XbaI endonucleases and cloned into M13mp18 also digested with HindIII and XbaI. The codon change in the rat AT 2 cDNA was made by site-directed mutagenesis using an in vitro mutagenesis kit (Sculptor, Amersham). One oligonucleotide was constructed to induce a mutation at methionine 116. The mutagenic primer is listed (altered nucleotide is underlined): methionine 116 to leucine (rAT 2 M116L): 5-GGCTCTTTGGACCTGTGCTGTG-CAAAGTGT-3. After confirmation of site-directed mutation by DNA sequencing, the rAT 2 M116L gene was excised from the M13mp18RF form by digestion with HindIII and XbaI and subcloned into the multiple cloning site of pcDNA3 that had been digested by these same restriction enzymes.
Transfection of COS-7 Cells-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. Cells were seeded into 75-cm 2 culture dishes at a density of 25,000 cells/cm 2 . When cells reached ϳ50% confluency, they were washed once with serum-free DMEM and transfected with 4 g of plasmid DNA and 30 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 medium containing 100 IU/ml penicillin and 100 g/ml streptomycin. Transfected cells were grown for 48 -72 h before photoaffinity labeling and binding assays.
Cell Culture-PC-12 cells between the second and the fifteenth passages were used in all experiments. Cells were seeded into 75-cm 2 culture dishes at a density of 25,000 cells/cm 2 and grown in DMEM with 10% (v/v) fetal bovine serum, 50 IU/ml penicillin, 60 g/ml streptomycin, and 2 mM L-glutamine. Cells were kept in culture at 37°C, in a CO 2 incubator, and the medium was changed daily or every other day depending on the state of confluency.
Binding Assays-Cell membrane preparation and binding assays were performed as described previously (19,20).
Photoaffinity Labeling-Transfected COS-7 cells and PC-12 cells (ϳ1 ϫ 10 7 ) were incubated in 4 ml of binding medium containing the photoreactive radioligands (6 nM), in the presence of L-158,809 (1 M) (an AT 1 selective non peptide analog). After 45 min at room temperature, cells were washed with 20 ml of ice-cold binding medium (without bovine serum albumin) and irradiated for 60 min at 0°C under filtered UV light (365 nm) (mercury vapor lamp serial number JC-Par-38 from Westinghouse and Raymaster black light filters number 5873 from Gates and Co. Inc., Long Island, NY). Cells were then gently scraped with a rubber policeman and centrifuged (200 ϫ g) for 10 min at 4°C. The pellet was solubilized in a buffer containing 100 mM Na 2 HPO 4 , pH 8.5, 25 mM EDTA, 0.1 mg/ml soybean trypsin inhibitor, and 1% (v/v) Nonidet P-40. After centrifugation (13,000 ϫ g for 10 min at 4°C), the supernatants were kept at Ϫ80°C until further analysis.
Partial Purification of the Labeled Complex-The solubilized photolabeled receptors were diluted with an equal volume of 2 ϫ loading 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 2 h at 37°C. SDS-polyacrylamide gel electrophoresis (PAGE) was performed as described by Laemmli (23) using 1.5-mm 8% gels. The gel was then dried and exposed to x-ray film (Kodak XAR-5) with an intensifying screen. The labeled proteins were isolated from the preparative gel using a passive elution protocol similar to that described by Blanton and Cohen (24). After autoradiography, radioactive bands were excised from dried gels and rehydrated with appropriate digestion buffer. The gel slices were macerated and eluted with 2 ml of buffer for 3-4 days at 4°C under gentle agitation. Under these conditions we repeatedly recovered at least 80% of the initial radioactivity. The eluted proteins were filtered (Acrodisc 0.22 m; Gelman Sciences) and the gel slices were washed with 10 ml of digestion buffer. The whole eluate (12-15 ml) was concentrated approximately 150 times using Centriprep-10 and Centricon-10 (Amicon) and the partially purified proteins were aliquoted in fractions of 2-4 ϫ 10 5 cpm, lyophilized, and kept at Ϫ80°C.
Endoglycosidase Digestion-Partially purified photolabeled receptors were resuspended in digestion buffers containing 0.2% (v/v) Nonidet P-40. PNGase-F (33-100 units/ml) was added and samples were incubated for different periods of time as indicated.
Protease Digestion-The partially purified photolabeled receptors (10,000 -300,000 cpm) were resuspended in 25 l of digestion buffer containing 100 mM NH 4 HCO 3 , pH 8.0, and 0.1% (w/v) SDS. Under these conditions V8 protease is known to cleave at the carboxyl-terminal side of glutamate residues. Samples were incubated for 3-4 days at room temperature with the indicated amounts of V8 protease. Partially purified photolabeled receptors (10,000 -300,000 cpm) were also digested for 18 -24 h at 37°C with indicated amounts of endo Lys-C in 25 l of digestion buffer containing 25 mM Tris-HCl, pH 8.5, 1 mM EDTA, and 0.1% (w/v) SDS. For kallikrein digestions, photolabeled receptors (ϳ20 -50 g of proteins) were resuspended in 25 l-50 l of digestion buffer containing 100 mM Na 2 HPO 4 , pH 8.5, 25 mM EDTA, 1 mg/ml soybean trypsin inhibitor, 0.1 mg/ml L-1-chloro-3-(4-tosylamido)-4-phenyl-2-butanon and 0.2% (v/v) Nonidet P-40. Indicated amounts of kallikrein were added and samples were incubated for 2 h at 37°C. All digestions were terminated by adding an equal volume of 2 ϫ loading buffer (previously described) and boiling the samples for 3 min. When subsequent digestions were needed, products of the first digestion were identified by SDS-PAGE and autoradiography, and recovered from non-fixed gels by passive elution for 12-18 h at 37°C with 200 -400 l of digestion buffer. Extracted proteins were lyophilized and submitted to chemical or enzymic digestion as described.
Chemical Digestion-For CNBr hydrolysis, partially purified photolabeled receptors (60,000 -450,000 cpm) were resuspended in 50 l of 70% (v/v) trifluoroacetic acid and CNBr (50 l) was added to a final concentration of 100 mg/ml. Samples were incubated at room temperature, in the dark, for 24 -36 h. Reactions were terminated by adding 500 l of water. Samples were lyophilized, resuspended in denaturating buffer, and analyzed by SDS-PAGE.
Analysis of Products of Proteolysis and Chemical Cleavage-The products of proteolysis and chemical cleavage were analyzed by SDS-PAGE using 16.5% acrylamide Tris-Tricine gels (Bio-Rad) followed by autoradiography on x-ray films (Kodak XAR-5). 14 C-Labeled low molecular protein standards (Life Technologies, Inc.) were used to determine apparent molecular masses. Running conditions, fixation, and coloration of gels were performed according to the manufacturer's instructions. Fig. 1 shows the primary structures of AngII and photoreactive AngII analogs used in this study. Asp 1 (19,20). The labeling of the AT 2 receptor by the two photoreactive analogs was completely abolished by PD123319 (10 M) (an AT 2 receptor selective ligand) (lanes 2 and 5) and by AngII (1 M) (lanes 3 and 6) thereby confirming the specificity and the selectivity of the labeling. Although both photoreactive analogs successfully labeled the AT 2 receptor, determination of covalent incorporation yields (calculated from the ratio of total radioactivity found in isolated bands to total specific binding observed before photolysis) revealed that 125 I-[Bpa 1 ]AngII, with a ϳ10% yield of covalent incorporation, was approximately six times less effective than 125 I-[Bpa 8 ]AngII (ϳ60% yield of covalent incorporation). These differences are well illustrated in Fig. 2 8 ]AngII⅐AT 2 complex migrated with a molecular mass of 70 kDa (Fig. 4a, lane 1). This receptor fragment was the final digestion product since prolonged incubation with kallikrein (Fig. 4a, lanes 3 and 5) did not reveal any lower molecular weight fragment. As shown in Fig. 3, the AT 2 receptor is N-glycosylated exclusively in its amino-terminal extracellular tail (6,7). The relatively high molecular mass of the 70-kDa digestion product and its glycoprotein-like migration behavior (broad band) suggest that it corresponds to the labeled 1-182 fragment of the AT 2 receptor.

Specificity of Photoreactive Analogs in Binding Experiments and Photoaffinity Labeling-
To confirm the location of the 70-kDa receptor fragment, kallikrein-treated 125 I-[Bpa 8 ]AngII⅐AT 2 complex was deglycosylated with PNGase-F. Under these conditions a labeled fragment of 18 kDa was produced (Fig. 4b,  ]AngII, AT 2 receptor was partially purified and digested with endo Lys-C which cleaves on the carboxyl-terminal side of lysine residues. The digestion product migrated as a broad band of ϳ80 kDa (Fig. 5a, lane 2). Again the high M r and the broadness of the band suggested a glycoprotein nature. After extraction of the endo Lys-C digestion product, treatment with PNGase-F resulted in a 4.6-kDa fragment which migrated as a sharp band suggesting that it was completely deglycosylated (Fig. 5c, lane 2). Interestingly, similar results were obtained with a simplified protocol where the photolabeled AT 2 receptor was simultaneously digested with endo Lys-C and PNGase-F (Fig. 5b, lane 2). Since as previously mentioned, the AT 2 receptor is N-glycosylated exclusively in its amino-terminal ectodomain, the cleavage probably occurred at one of two lysine residues, located at positions 38 and 42 (Fig.  3). Cleavage at these residues should produce either a 5.2-kDa fragment (including the photolabel of 1.3 kDa) that corresponds FIG. 1. A, amino acid sequence of AngII and photoreactive AngII analogs. Residues represented in bold characters correspond to amino acid modifications. B, mechanism of covalent modification following photoactivation. The absorption of a photon at ϳ350 nm by the Bpa moiety (1) results in the promotion of one-electron from a nonbonding sp 2 -like n-orbital on oxygen to an antibonding * -orbital of the carbonyl group (2). In the diradicaloid triplet state (2), the electron-deficient oxygen n-orbital is electrophilic and therefore interacts with weak C-H bonds (2) to produce benzpinacol-type compounds (3) (31).

FIG. 2. Photoaffinity labeling of AT 2 receptor from PC-12 cells.
After photolabeling, cellular proteins were solubilized, denatured, and submitted to SDS-PAGE on a 8% acrylamide separating gel (40 g of protein/lane) followed by autoradiography. to the labeled 3-38 peptide or a 5.6-kDa fragment that corresponds to the labeled 3-42 peptide.
To further define the 125 I-[Bpa 1 ]AngII-binding domain, the 4.6-kDa fragment obtained after co-digestion with endo Lys-C and PNGase-F was submitted to digestion with V8 protease. Fig. 6, lane 3, shows that, under these conditions, the 4.6-kDa fragment was converted to a 3.6-kDa fragment. The only site of cleavage for V8 protease within amino acids 3-42 is after glutamate 30 (Fig. 3). Cleavage of the 4.6-kDa fragment at this site should produce either a labeled 2.6-kDa fragment (31-42 peptide ϩ photolabel), a labeled 2.2-kDa fragment (31-38 peptide ϩ photolabel), or a labeled 4.3-kDa fragment (3-30 peptide ϩ photolabel) (Fig. 3). Based on the relatively high molecular mass of the digestion product (3.6 kDa), the 3-30 peptide is clearly a better candidate than the 31-42 or 31-38 peptides for the binding domain of 125 I-[Bpa 1 ]AngII. This conclusion is fur-ther strengthened by experiments in which photolabeled-AT 2 receptor was co-digested with PNGase-F and V8 protease. Under these conditions, a 4.1-kDa fragment was obtained, locating the binding domain within the first 30 amino acids of the AT 2 receptor (results not shown). Together these results show that 125 I-[Bpa 1 ]AngII is labeling a site within residues 3-30 of the extracellular amino-terminal tail of the AT 2 receptor (Fig. 10).
Mapping the 125 6). Samples were run on a 8% acrylamide separating gel followed by autoradiography. Protein standards of the indicated molecular masses (kDa) were run in parallel. These results are representative of at least three separate experiments. B, 125 I-[Bpa 8 ]AngII-labeled AT 2 receptor (50 g of membrane protein) (lane 1) was incubated for 2 h at 37°C in the absence (lanes 2 and 3) or presence (lanes 4 and 5) of tissue kallikrein (50 g) before digestion with PNGase-F (33 units/ml) for 2 h at 37°C (lanes 3 and 5). Samples were run on a 12% acrylamide separating gel followed by autoradiography. 14 C-Labeled protein standards of the indicated molecular masses (kDa) were run in parallel. These results are representative of three separate experiments.

FIG. 5. Endo Lys-C digestion of 125 I-[Bpa 1 ]AngII-labeled AT 2 receptor.
A, partially purified photolabeled AT 2 receptor (24,000 cpm) was incubated in the absence (lane 1) or presence (lane 2) of endo Lys-C (2.5 g) at 37°C for 22 h. Samples were run on a 10% acrylamide separating gel followed by autoradiography. Protein standards of the indicated molecular masses (kDa) were run in parallel. These results are representative of three separate experiments. B, photolabeled receptor (47,000 cpm) was incubated with PNGase-F (100 units/ml) at room temperature for 24 h. The sample was aliquoted in two fractions one of which received digestion buffer (lane 1) and the other received 1.8 g of endo Lys-C (lane 2). Incubation was prolonged for 22 h at 37°C. Samples were run on a 16.5% acrylamide Tris-Tricine separating gel followed by autoradiography. 14 C-Labeled protein standards of the indicated molecular masses (kDa) were run in parallel. These results are representative of three separate experiments. C, photolabeled receptor (100,000 cpm) was incubated with endo Lys-C (1 g) for 20 h at 37°C. The sample was run on a 8% acrylamide separating gel. The 80-kDa labeled receptor fragment was located and recovered by passive elution from gel slices. The receptor fragment (40,000 cpm) was incubated in the absence (lane 1) or presence (lane 2) of PNGase-F (40 units/ml) for 2 h at 37°C. Samples were run on a 16.5% acrylamide Tris-Tricine separating gel followed by autoradiography. 14 C-Labeled protein standards of the indicated molecular masses (kDa) were run in parallel. These results are representative of two separate experiments.
[Bpa 8 ]AngII-photolabeled AT 2 receptor was partially purified and digested with V8 protease. The patterns of fragmentation were clearly distinct from those obtained with 125 I-[Bpa 1 ]AngII⅐AT 2 complex. Digestion with V8 protease produced 28-and 15.2-kDa fragments (Fig. 7, lane 2). Deglycosylation had no effect on the mobility of the 15.2-kDa fragment indicating that it was not a glycosylated fragment (result not shown). Knowing that the binding domain of 125 I-[Bpa 8 ]AngII is located within amino acids 1-182 (Fig. 4) and that the 15.2-kDa digestion product is not glycosylated, the fragment located between alanine 46 and glutamate 188 (estimated molecular mass of the peptide ϩ photolabel: 17.8 kDa) is the best candidate for the binding domain of 125 I-[Bpa 8 ]AngII (Fig. 3). Prolonged incubations in the presence of PNGase-F reduced the proportion of the 28-kDa fragment and increased the proportion of a fragment migrating close to the 15.2-kDa fragment, suggesting that the 28-kDa fragment is the glycosylated 31-188 peptide of AT 2 receptor, containing a putative site of glycosylation at asparagine 34 (Fig. 3).
To further define the 125 I-[Bpa 8 ]AngII-binding domain, photolabeled AT 2 receptor was submitted to hydrolysis with CNBr which cleaves specifically at the carboxyl-terminal side of methionine residues. Fig. 8, lane 2, shows that, under these conditions, two major digestion products of 3.6 and 2.6 kDa were obtained. These results suggest that the photolabel binds to 117-138 peptide (estimated molecular mass of 3.7 kDa, including the photolabel) in the third transmembrane domain of AT 2 . Hydrolysis at methionine 128 of this peptide would produce two fragments of comparable apparent sizes (estimated size of 2.6 and 2.4 kDa for 117-128 peptide and 129 -138 peptide, respectively). To confirm this hypothesis and to identify the 2.6-kDa fragment, we generated a mutant of the rat AT 2 receptor in which methionine 116 was replaced by a leucine, therefore abolishing that putative CNBr cleavage site. In competitive FIG. 7. V8 protease digestion of 125 I-[Bpa 8 ]AngII-labeled AT 2 receptor. Partially purified photolabeled AT 2 receptor (250,000 cpm) was incubated in the absence (lane 1) or presence (lane 2) of V8 protease (50 g) for 4 days at room temperature. Samples were run on a 16.5% acrylamide Tris-Tricine separating gel followed by autoradiography. 14 C-Labeled protein standards of the indicated molecular masses (kDa) were run in parallel. These results are representative of at least three separate experiments.  1 and 3) of CNBr (100 mg/ml) for 24 h at room temperature in the dark. Samples were run on a 16.5% acrylamide Tris-Tricine separating gel followed by autoradiography. 14 C-Labeled protein standards of the indicated molecular masses (kDa) were run in parallel. AngII, partially purified, and submitted to hydrolysis by CNBr. Fig. 9, lane 1, shows that CNBr hydrolysis of the wild-type AT 2 receptor produced the previously described 3.6-and 2.6-kDa fragments. CNBr hydrolysis of the rAT 2 M116L receptor still produced the 2.6-kDa fragment and a longer 12.6-kDa fragment (Fig. 9, lane 3). Under these conditions the only possibility is that the 2.6-kDa fragment is the 129 -138 peptide located in the inner half of the third transmembrane domain of mutant AT 2 receptor. Indeed, if labeling had occurred between leucine 116 and methionine 128 in the mutant receptor, exclusively higher molecular mass labeled receptor fragments (13)(14)(15) would have been produced. The 12.6-kDa fragment produced by CNBr hydrolysis of rAT 2 M116L most probably corresponds to the mutant receptor 54 -138 peptide (estimated molecular mass of the peptide ϩ photolabel: 10.9 kDa). The incomplete CNBr hydrolysis of both native and mutant receptors suggest that cleavage at methionine 128 is impaired. This may result from reduced solubility of the protein in strong dissociating agents (like trifluoroacetic acid), sterical masking of the methionine residue by surrounding amino acids, or oxidation of the methionine residue occurring during protein manipulations and/or acid hydrolysis (25). Together these results show that 125 I-[Bpa 8 ]AngII is labeling a site within residues 129 -138 of AT 2 receptor (Fig. 10). DISCUSSION Our results indicate that, upon binding, the amino-terminal tail of the AT 2 receptor interacts with the amino-terminal end of AngII whereas the inner half of the third transmembrane domain of AT 2 receptor interacts with the carboxyl-terminal end of AngII. To our knowledge, this is the first study providing data on the binding domains of the AT 2 receptor. Our results and those of other groups suggest that there may be a common scheme for the binding domains of small peptide hormones in the G-protein-coupled receptor superfamily. For example, it has been shown that charged residues in the second and third extracellular loops of the AT 1 receptor are major docking points for the amino-terminal end (Asp 1 and Arg 2 residues, respectively) of AngII (13,26). On the other hand, the hydrophobic nature of the phenyl group at position 8 in the AngII molecule (the carboxyl-terminal end of the peptide) suggests that its site of interaction with the AT 1 receptor is within the membrane. Actually, recent studies suggest that upon binding, the carboxyl-terminal end of the AngII molecule occupies a space between helices III, V, VI, and VII of the AT 1 receptor (26,27). Similarly, binding domains for the undecapeptide substance-P (SP) were identified in the first and second extracellular loops, the aminoterminal ectodomain and the outer half of the transmembrane helices II and VII of the NK-1 receptor (14,15,28). In an elegant study using photoreactive analogs of SP, Li et al. (15) have shown that the amino-terminal side (the fourth position) of the 125 I-[Tyr 1 ,Bpa 4 ]SP peptide interacts with the extracellular amino-terminal tail of the NK-1 receptor while the COOHterminal side (the ninth position) of the 125 I-[Tyr 1 ,Bpa 9 ]SP peptide interacts with the second extracellular loop. The authors proposed a model in which the carboxyl-terminal end of the peptide positions itself between helices in the outer part of the plasma membrane whereas the amino-terminal portion of the peptide is stabilized by ectodomains of NK-1 receptor. Based on our results, it is tempting to propose that the AngII molecule binds the AT 2 receptor in a similar fashion, with the carboxyl-terminal end sitting deep within the transmembrane domains and the amino-terminal end interacting with the ectodomains of the receptor. The location of the carboxyl-terminal portion of AngII deep in the plasma membrane fits well with known pharmacological properties of AngII. The agonistic nature of AngII is conferred by its carboxyl-terminal phenylalanine residue (29) and intrahelical amino acids are known to play a major role in G-protein coupled receptor activation upon direct interaction with ligands (10). Our results also suggest that the carboxyl-terminal end of AngII interacts with the third transmembrane domain of the AT 2 receptor. Similarly, it has been shown that residues in the third transmembrane domain of the AT 1 receptor are required for high affinity binding (16, FIG. 10. Two-dimensional representation of the primary structure of the rat AT 2 receptor and its AngII-binding domains. The ligand-binding domains are represented by closed circles: the amino-terminal end of AngII interacts with the extracellular amino-terminal tail within residues 3-30; the carboxyl-terminal end of AngII interacts with the third transmembrane domain within residues 129 -138. Solid lines indicate putative disulfide bridges between cysteines 35 and 290 and between cysteines 117 and 195. 30). Our results also raise another interesting point. If the AngII molecule interacts at the same time with residues in the amino-terminal extracellular tail as well as residues deeply located in the third transmembrane domain, one could speculate that the ectodomain must lie near the outer membrane surface. The AngII molecule has an estimated length of about ϳ30 Å (probably less in solution) which is shorter than a transmembrane helix (ϳ40 Å). To account for such a requirement, the putative disulfide bridge located between the aminoterminal tail and the third extracellular loop of the AT 2 receptor may play an important conformational role in bringing the amino-terminal tail in close proximity to the plasma membrane surface (Fig. 10).
In conclusion, we have identified two peptide-binding regions in the AT 2 receptor with the use of highly potent and specific AngII photoreactive analogs. This approach also allowed the determination of the AngII molecule's orientation in its binding pocket. Based on these results, we conclude that AngII interacts with an extracellular segment and a transmembrane helix of AT 2 receptor. This interaction pattern, also found in other G-protein coupled receptors for small bioactive peptides, may correspond to a highly conserved feature among this very large family or receptors. By recovering large amounts of labeled receptor, it will be possible to sequence the fragments and pinpoint the precise interaction sites.