Activation of the AT1 Angiotensin Receptor Is Dependent on Adjacent Apolar Residues in the Carboxyl Terminus of the Third Cytoplasmic Loop*

The C-terminal region of the third intracellular loop of the AT1 angiotensin receptor (AT1-R) is an important determinant of G protein coupling. The roles of individual residues in agonist-induced activation of Gq/11-dependent phosphoinositide hydrolysis were determined by mutational analysis of the amino acids in this region. Functional studies on mutant receptors transiently expressed in COS-7 cells showed that alanine substitutions of the amino acids in positions 232–240 of the third loop had no major effect on signal generation. However, deletion mutations that removed Ile238or affected its position relative to transmembrane helix VI significantly impaired angiotensin II-induced inositol phosphate responses. Substitution of Ile238 with an acidic residue abolished the ability of the receptor to mediate inositol phosphate production, whereas its replacement with basic or polar residues reduced the amplitude of inositol phosphate responses. Substitutions of Phe239 with polar residues had relatively minor effects on inositol phosphate signal generation, but its replacement by aspartic acid reduced, and by positively charged residues (Lys, Arg) significantly increased, angiotensin II-induced inositol phosphate responses. The internalization kinetics of the Ile238 and Phe239 mutant receptors were impaired in parallel with the reduction in their signaling responses. These findings have identified Ile238 and Phe239 as the critical residues in the C-terminal region of the third intracellular loop of the AT1-R for receptor activation. They also suggest that an apolar amino acid corresponding to Ile238 of the AT1-R is a general requirement for activation of other G protein-coupled receptors by their agonist ligands.

The AT 1 angiotensin receptor (AT 1 -R) 1 is a member of the GPCR superfamily and mediates the physiological actions of the octapeptide hormone, Ang II, on cardiovascular regulation and salt/water balance. The binding of Ang II to the AT 1 -R initiates conformational changes that lead to activation of its cognate G protein(s), predominantly G q/11 , in numerous Ang II target tissues. The subsequent stimulation of phospholipase C causes the generation of inositol 1,4,5-trisphosphate and diacylglycerol, leading to the elevation of [Ca 2ϩ ] and activation of protein kinase C (1-3). The AT 1 -R is also coupled to other signaling pathways, including the Ras-mediated activation of mitogen-activated protein kinase and cell growth responses (4,5).
Although GPCRs exhibit a remarkable degree of structural and functional diversity, all share a common architecture in which seven transmembrane helices are linked by alternating intracellular and extracellular loops. Agonist binding to the extracellular loops, and/or the exofacial regions of the transmembrane domains, induces conformational changes in the transmembrane helices and intracellular regions that promote coupling of the activated receptor to its cognate G protein(s). The receptor domains that are involved in G protein coupling have been analyzed in many GPCRs. Studies using chimeric ␣ 2 /␤ 2 -, ␣ 1B /␣ 2 -, and ␣ 1/ ␤ 2 -adrenergic receptors and chimeric M 2 /M 3 muscarinic receptors have shown that modification of IL3 influences G protein selectivity, suggesting that IL3 has a critical role in receptor-G protein interaction (6 -9). The importance of the C-terminal portion of IL3 in receptor conformation has been further indicated by the ability of mutations of specific residues in this region to cause constitutive activation of several adrenergic receptors (6 -8).
Studies on spin-labeled double cysteine mutants of rhodopsin have indicated that rigid body motion of helix VI compared with helix VII is an important event in receptor activation (10) and that movement of the C-terminal region of IL3 occurs in constitutively active receptors (11). In the ␤ 2 -adrenergic receptor, substitution of the C-terminal region of IL3 with those of the ␣ 1B -or ␣ 2A -adrenergic receptors abolished the ability to couple to G s (9,12). In M 2 , M 3 , and M 5 muscarinic receptors, this region has been implicated in agonist-induced coupling to G q/11 or G i , respectively (13)(14)(15). Studies on chimeric human AT 1 -and AT 2 -Rs have demonstrated that the membrane proximal portions of IL3 are important for coupling to G q (16). Also, the substitution of all charged residues in the C-terminal region of IL3 markedly reduces Ang II-induced inositol phosphate signal generation (17). Furthermore, exchange of this region of the AT 1 -R with the corresponding sequence of the ␤ 2 -adrenergic receptor caused coupling of the receptor to G s as well as G q during agonist stimulation (18). Other studies have demonstrated that a 16-amino acid peptide comprising the N-terminal portion of IL3 of the AT 1 -R stimulates the binding of GTP␥S to G q (19).
Although these findings have clearly implicated the C-terminal region of IL3 in the coupling of the AT 1 -R to G q , the amino acids that are critical for this interaction have not been identified. To resolve this question, agonist-induced functional responses of mutant AT 1a -Rs bearing replacements and deletions of amino acids in this region were determined in transiently transfected COS-7 cells.

EXPERIMENTAL PROCEDURES
Materials-DMEM, fetal bovine serum, and antibiotic solutions were from Biofluids (Rockville, MD). Ang II was from Sigma. 125  Mutagenesis and Transient Expression of Mutant Rat AT 1a -Rs-Mutations were created using the Mutagene kit (Bio-Rad) in the sequence of the rat AT 1a -R cDNA subcloned into the mammalian expression vector, pcDNAI/Amp (Invitrogen, San Diego, CA), and mutated sequences were verified by dideoxy sequencing as described previously (20). Rat AT 1a replacement mutant receptors are indicated by the single letter code of the original amino acid followed by the position number and the single letter code of the exchanged amino acid. Deletion mutant AT 1a receptors are indicated by Del followed by the single letter code of the deleted amino acid and its position number. COS-7 cells were seeded at 5 ϫ 10 4 cells/well in 24-well culture plates in 1 ml of DMEM containing 10% (v/v) fetal bovine serum, 100 IU/ml penicillin, and 100 g/ml streptomycin (COS-7 medium) and cultured for 3 days before transfection using 0.5 ml of OptiMEM containing 8 g/ml of Lipo-fectAMINE and the required DNA (usually 2 g/ml) for 6 h at 37°C. Cells were cultured for 2 additional days in fresh COS-7 medium prior to use.
[Sar 1 ,Ile 8 ]Ang II Binding to Intact Cells-To determine the expression level and structural integrity of the mutant receptor, the number of Ang II binding sites was determined by incubating the transfected cells with [ 125 I-Sar 1 ,Ile 8 ]Ang II (0.05-0.1 Ci/sample) and increasing concentrations of unlabeled [Sar 1 ,Ile 8 ]Ang II in DMEM for 6 h at 4°C. The bound radioactivity associated with the cells was measured by ␥-spectrometry, and the binding curves were analyzed with the LIGAND computer program using a one-site model as described previously (21).
AT 1a -R Internalization Assay-Internalization of mutant and wildtype AT 1a -Rs were measured using 125 I-Ang II at 37°C by ␥-spectrometry, and values of the endocytotic rate constants (K e ) were calculated as described previously (21). The percent of internalized ligand at each time point was calculated from the ratio of the acid-resistant specific binding to the total (acid-released ϩ acid-resistant) specific binding.
Inositol Phosphate Measurements-Transfected COS-7 cells in 24well plates were labeled by incubation for 24 h in inositol-free DMEM containing 0.1% (w/v) bovine serum albumin, 2.5% (v/v) fetal bovine serum, antibiotics, and 20 Ci/ml myo-[2-3 H]inositol. After washing and preincubation with 10 mM LiCl for 30 min, 1 M Ang II was added for a further 20 min, and incubations were stopped with perchloric acid.
Inositol phosphates were extracted, the combined inositol bisphosphates ϩ inositol trisphosphates fractions were separated using Bio-Rad AG 1-X8 columns, and their radioactivities were determined by liquid scintillation counting as described (22). At the expression levels used in this study, there was a linear relationship between cell surface receptor expression and the magnitude of agonist-stimulated inositol phosphate production (21).

RESULTS
Binding Properties of Mutant AT 1 -Rs-To identify residue(s) in the C-terminal region of IL3 of the AT 1a -R that are required for receptor activation, we analyzed a series of mutant receptors containing replacements or deletions of residues in the Lys 232 -Phe 239 segment (Fig. 1). All mutants retained high binding affinity for [ 125 I-Sar 1 ,Ile 8 ]Ang II (Table I), indicating their retention of the normal global conformation of the wild-type receptor. However, the mutant receptors displayed considerable variation in their expression levels compared with the wild-type receptor, and the expression of the Del N235 mutant receptor was markedly reduced (Table I).
Inositol Phosphate Responses of Alanine Replacement and Deletion Mutant AT 1 -Rs-The ability of each mutant receptor to couple to G q was evaluated by measuring the Ang II-induced stimulation of inositol phosphate production in transfected COS-7 cells. Alanine substitutions of the AT 1a -R caused no major changes of the basal inositol phosphate levels (Fig. 2). The Ang II-stimulated inositol phosphate responses were significantly reduced in the D236A and D237A mutant receptor (Fig. 2, upper panel), but the expression levels of these mutant receptors were also lower (Table I). When the inositol phosphate responses were normalized to the number of expressed receptors, it was evident that the replacement of each residue in the Lys 232 -Phe 239 segment by alanine had no significant effect on the ability of the receptor to couple to G q and initiate signal transduction (Fig. 2, lower panel). Double alanine substitution of Lys 232 and Arg 234 had no major effect on receptor expression (Table I) or inositol phosphate signal generation (Fig. 2). Triple alanine substitution of Asn 235 , Asp 236 , and Asp 237 markedly decreased the expression level (Table I) Table I). These data indicate that negatively charged amino acids in this region are required for normal receptor expression and argue against the role of charged residues in inositol phosphate signal generation.
The C terminus of IL3 of GPCRs has been predicted to  (40). The most conserved residues in GPCRs are shown in white letters on a black background. Glycosylation sites of the receptor and amino acids in the third intracellular loop of the AT 1 -R that were mutated in this study are also indicated. assume an ␣-helical structure that is continuous with the sixth transmembrane helix (23). Because deletions of individual residues within a helix change the orientation of the remaining amino acids, such mutations are likely to modify or abolish functionally important contacts by the constituent residues of the helix. When single amino acids in the C-terminal region of IL3 were deleted to identify functionally important residues, only the removal of Asn 235 reduced the expression of the AT 1a -R to such an extent that normalization of its inositol phosphate response was not reliable. Deletion of residues Arg 240 , Phe 239 , and Ile 238 , adjacent to the sixth intramembrane helix, markedly impaired Ang II-induced inositol phosphate generation (Fig. 3). However, deletion of Asp 237 , Arg 234 , or Pro 233 , residues that are more distant from helix 6, had no effect on the inositol phosphate responses (Fig. 3). These findings suggest that the residue located in position 238 is an important structural determinant of Ang II-induced signal generation.
Inositol Phosphate Responses of Ile 238 and Phe 239 Substitution Mutant AT 1 -Rs-In adrenergic receptors, substitutions of amino acids in the location corresponding to position 239 of the AT 1a -R cause constitutive activation of the mutant receptors (24). A more detailed analysis of the structural requirements in positions 238 and 239 was performed to study the role of this region in AT 1a -R activation. Replacement of Phe 239 with other apolar residues such as alanine and leucine or with tyrosine, a polar residue, had little or no effect on inositol phosphate signal generation (Fig. 4). However replacement with asparagine or serine significantly impaired the signaling ability of the AT 1 -R. Replacement of Phe 239 with the acidic residue, aspartate, had the most profound effect and reduced inositol phosphate production by about 90%. In contrast, replacement of Phe 239 with basic residues (lysine or arginine) significantly enhanced inositol phosphate responses. None of the tested mutants had a major effect on inositol phosphate levels in unstimulated cells, indicating that, in contrast to the adrenergic receptor(s), replacement of the amino acid in this position with other residues does not cause constitutive activation of the AT 1a -R (Fig. 4, Table I).
Replacement of Ile 238 with polar or charged amino acids also affected the signal generation efficacy of the AT 1 -R (Fig. 5). Substitution of Ile 238 with apolar alanine had no effect on inositol phosphate responses, but replacement with polar serine impaired inositol phosphate production when the data were normalized to the expression levels of the receptor (Fig. 5, lower  panel). Replacement of Ile 238 with aspartate considerably increased the expression of the receptor in COS-7 cells (Table I). However, despite the increased expression level of the I238D receptor its inositol phosphate responses were significantly impaired and when normalized for expression were reduced by ϳ85%. Substitution of Ile 238 with lysine, a basic residue, had no effect on the expression level of the receptor, but the inositol phosphate responses of the I238K receptor were significantly reduced (Fig. 5, Table I).   (Table I) and are shown as a percentage of the wild-type response. The data represent mean Ϯ S.E. from three independent experiments each performed in duplicate. InsP 2 , inositol bisphosphates; InsP 3 , inositol trisphosphates.
Internalization of Mutant AT 1a -Rs-Individual replacements of each amino acid in the Lys 232 -Ile 238 segment with alanine had no major effect on the agonist-induced internalization kinetics of the AT 1a -R in COS-7 cells, but replacement of Phe 239 with alanine modestly reduced the internalization rate of the receptor (Fig. 6). Other substitutions of Ile 238 and Phe 239 caused impairment of the internalization kinetics of the mutant receptors (Fig. 7) that paralleled their reductions in inositol phosphate generation (Fig. 2). Replacement of Ile 238 with alanine had no effect on the internalization kinetics, but substitution with charged residues caused a significant decrease in the internalization kinetics of the receptor (Fig. 7 and Table II). Compared with the wild-type AT 1a -R, the K e values of the I238K and I238D mutant receptors were reduced to 45 Ϯ 2% and 10 Ϯ 0.1% of the wild-type value, respectively. The K e values of mutant receptors in which Phe 239 was replaced by other residues (F239K, F239R, F239Y, F239A, and F239L) were 132 Ϯ 1.8%, 79.4 Ϯ 0.7%, 85.8 Ϯ 6.4%, 60.6 Ϯ 8.3%, and 66.3 Ϯ 8.6% of that of the wild-type receptor, respectively. In contrast, the K e values of the F239S, F239N, and F239D mutant receptors were reduced to 4.5 Ϯ 2.1%, 6.1 Ϯ 3.5%, and 2.9 Ϯ 1.0% of that of the wild-type receptor, respectively ( Fig. 7 and Table II). DISCUSSION The N-and C-terminal portions of IL3 have been implicated in G protein coupling to several seven transmembrane domain receptors (25)(26)(27). Discontinuous sequences of IL3, IL2, and the C-terminal cytoplasmic tail have been predicted to comprise an agonist-induced conformation-dependent surface that permits interaction with G proteins (28). Earlier studies using chimeric AT 1 -Rs (16) and substitutions of charged residues in the Cterminal portion of IL3 (17) suggested that this region of the receptor is also required for Ang II-induced signal generation. In addition to activating G protein-dependent second messenger systems, the AT 1a -R also undergoes phosphorylation and rapid internalization after agonist stimulation (29 -32). Substitution of amino acids 234 -240 of the AT 1 -R with the corresponding sequence of the ␤ 2 receptor impaired its internalization kinetics, suggesting that this region has a role in the agonist-induced internalization of the receptor (18). However, alanine substitutions of individual amino acids in the C-terminal region of IL3 did not affect the internalization kinetics of the AT 1 -R, with the exception of the F239A mutation, which reduced the internalization kinetics in a moderate extent.
Studies on the effects of replacing the distal segment of IL3 of the AT 1a -R with the corresponding regions of the ␣ 1 -adrenergic, ␤ 2 -adrenergic, and AT 2 receptors revealed that the RNDDIFR sequence (from Arg 234 to Arg 240 ) of the AT 1a -R has a major role in G protein coupling and receptor signaling (18). markedly attenuated inositol phosphate production. Because the C-terminal region of IL3 is believed to form an extension of the ␣-helical structure of the sixth transmembrane helix, this effect of amino acid deletions could result from changes in the position of Ile 238 relative to the rest of the helix. The lack of effect of alanine substitutions raises the possibility that, similar to the role of Leu 222 in the N-terminal region of IL3 (33,34), apolar amino acids in this region are favorable for signal generation.
Although substitutions with alanine or other apolar amino acids were tolerated, replacements of Ile 238 with polar or charged amino acids significantly impaired the Ang II-induced inositol phosphate responses of the AT 1a -R. These data suggest that the presence of a hydrophobic amino acid at position 238 is required for AT 1 -R activation. It has been reported that the four amino acid sequence, VTIL (Val 385 , Thr 386 , Ile 389 , and Leu 390 ) located at the boundary of IL3 and helix 6, determines the G i coupling specificity of the M 2 muscarinic receptor. Also, this region of the M 2 receptor has been proposed to interact with the C-terminal domain of G␣ i/o (15,35). The corresponding AALS motif (Ala 488 , Ala 489 , Leu 492 , and Ser 493 ) of the G qcoupled M 3 muscarinic receptor together with IL2 and the N-terminal part of IL3 were suggested to be directly involved in coupling to G q (36). The position of Ile 238 of the AT 1 -R corresponds to the position of Val 385 of the M 2 receptor and Ala 488 of the M 3 receptor. As in the muscarinic receptors, this region appears to be an important determinant of G protein activation of the AT 1 -R.
In addition to their effects on inositol phosphate signal gen-eration, mutations of Ile 238 also impaired the internalization kinetics of the AT 1a -R. Although the agonist-induced internalization of GPCRs often occurs in parallel with G protein activation, mutational analysis of the AT 1 and other receptors has shown that the structural requirements of these processes are distinct, and AT 1 -R internalization can occur in the absence of inositol phosphate signal generation and G protein coupling (20,31,32,37,38). Because receptor internalization does not depend on G protein-mediated signaling, the inhibitory effects of substituting Ile 238 with polar and charged amino acids on receptor internalization are independent of G protein coupling.
Although it cannot be excluded that Ile 238 directly couples the receptor to G protein activation and internalization, these data suggest that Ile 238 is required for a step in receptor activation that is proximal to receptor G protein coupling. The most likely mechanism is the participation of Ile 238 in an intramolecular interaction that is required for the active conformation of the receptor. Such a mechanism is consistent with a recent report on the mapping of G protein binding domains of the AT 1a -R (39). In contrast to peptides derived from the N-terminal region of IL2 and IL3 and the juxtamembranous region of the cytoplasmic tail, peptides derived from the C-terminal region of IL3 did not compete for receptor G protein coupling (39). The proposed participation of Ile 238 in an intramolecular interaction is also in agreement with the GPCR model of Baldwin (40), which predicts that the side chain of the amino acid in the position of Ile 238 faces helix 7 and thus points away from the putative G protein binding region (28,40).
In view of the requirement for an apolar amino acid in the position of Ile 238 in the AT 1 -R, the properties of the amino acids in this location in other members of GPCR family I were examined. Among 171 GPCR sequences from the data base of comprehensive alignment of G protein-coupled receptor sequences compiled by the Molecular Recognition Section of the National Institutes of Health, only 22 receptors contain polar amino acids at this position. These residues were 15 threonines, 4 serines, 1 asparagine, and 2 histidines. All other members of the GPCR family I contain an apolar residue at this position. The frequent occurrence of a corresponding apolar amino acid in GPCRs raises the possibility of a general requirement for such residues in GPCR activation by agonist ligands. The position of Phe 239 in the AT 1 -R has a different amino acid requirement compared with that of position 238. The location of this amino acid corresponds to Thr 386 in the VTIL motif of the M 2 muscarinic receptor and Ala 489 in the AALS motif of the M 1 , M 3 , and M 5 muscarinic receptors. In addition, Phe 239 corresponds to a position that frequently causes constitutive activation if substituted in adrenergic receptors (24). However, a completely different mechanism of constitutive activation has been reported for the AT 1 -R (41). In contrast to Ile 238 , replacement of Phe 239 with polar residues had a variable effect on AT 1 -R activation. Substitutions of apolar amino acids or tyrosine caused only minor changes in signal generation and internalization of the receptor, but substitution of asparagine or serine decreased these responses. Although substitution of a negatively charged aspartic acid for Phe 239 severely impaired inositol phosphate generation and receptor internalization, replacement by positively charged residues increased these responses. Recent studies have predicted that AT 1 -R activation involves an outward movement of the cytoplasmic end of the sixth transmembrane helix, accompanied by a clockwise rotation (as viewed from the cell interior) by ϳ30° (42). A possible explanation for the opposite effects of basic and acidic amino acids is that positively charged residues facilitate this movement, whereas the presence of a negative charge interferes with it. Similar to the results obtained with mutations of Ile 238 , substitution of Phe 239 caused a parallel impairment of inositol phosphate signal generation and receptor internalization. These data suggest the Phe 239 , like Ile 238 , has a role in the formation of the active conformation of the receptor. This conclusion is in accordance with the results of studies on rhodopsin-transducin interaction that showed that substitution of the amino acids (Val 250 and Thr 251 ) of rhodopsin that correspond to Ile 238 and Phe 239 of the AT 1a -R interfere with the formation of the active (M II) state of the photoreceptor and thus indirectly affect its interaction with transducin (43).
In summary, this study demonstrates the importance of two specific residues, Ile 238 and Phe 239 , in the C terminus of IL3 in the mechanism of activation of the AT 1 -R. The parallel impairment of receptor signaling and internalization suggests that these adjacent apolar residues may participate in intramolecular interaction(s) that are required for receptor activation. The exact mechanism of this interaction and its role in other GPCRs requires further investigation. However, based on the present findings, and other evidence that implicates this region of GPCRs in signal transduction, it is possible that these amino acids and their counterparts have a general role in receptor activation by agonist ligands. FIG. 7. Internalization kinetics of Ile 238 and Phe 239 replacement mutant AT 1a -Rs. COS-7 cells expressing the wild-type (WT) and Ile 238 (A) or Phe 238 (B) replacement mutant receptors were incubated with 125 I-Ang II at 37°C for the indicated times. Acid-resistant and acid-sensitive binding were determined as described in Fig. 6, and the internalized (acid-resistant) binding was expressed as percent of the total binding at each time point. The data represent mean Ϯ S.E. from three independent experiments each performed in duplicate.