A Domain for G Protein Coupling in Carboxyl-terminal Tail of Rat Angiotensin II Receptor Type 1A*

To delineate domains essential for G q protein cou- pling in the C-terminal region (C-tail) of rat angiotensin II (Ang II) receptor type 1A (AT 1A ), we modified the putative cytosolic regions of the receptor by truncation or alanine substitution and determined resultant changes in the guanosine 5 (cid:42) -3- O -(thio)triphosphate (GTP (cid:103) S) effect on Ang II binding and inositol trisphosphate production by the agonist. Independently, we studied the effect of synthetic C-tail peptides (P-5) and its alanine substitution analogs on [ 35 S]GTP (cid:103) S binding to G q . Effects of GTP (cid:103) S on Ang II binding (shift to a low affinity form) and inositol trisphosphate production in the deletional mutant receptor 1–317 AT 1A was similar to wild type AT 1A , whereas in shorter C-terminal deletion mutants 1–309, 1–311, 1–312, 1–313 AT 1A , and substitutional mutants Y312A, F313A, and L314A these activities were markedly reduced. The binding of [ 35 S]GTP (cid:103) S to G q was promoted by NaCl, and 20 (cid:109) M GTP. After a 50- (cid:109) l aliquot of the reaction mixture was rapidly filtered through a nitrocellulose filter (pore size, 0.45 (cid:109) m) and washed three times with the stopping buffer, the filter was counted in a liquid scintillation counter. The maximal binding of [ 35 S]GTP (cid:103) S to G q was measured in the presence of 1 (cid:109) M GTP (cid:103) S and 25 m M Mg 2 (cid:49) at room temperature by the method of Northup et al. (21) as a positive control. Statistical Analysis— The results of experiments with the synthetic peptide study was examined by unpaired Student’s t test. p values less than 0.05 were considered significant.

receptor essential for intracellular signal transduction appear to be the first intracellular loop (ICL1) and the C-terminal regions of the second intracellular loop (ICL2) and the third intracellular loop (ICL3) (11). Four isoforms of prostaglandin E receptor subtype EP3, which differ only at their C-terminal tails and are produced by alternative splicing, couple to different G proteins. Thus the C-terminal tail of EP3 determines G protein specificity (12).
Studies by Wang et al. (13) using chimeras of human AT 1 and AT 2 suggested that the N-terminal portion of ICL3 was important for G q coupling. We reported observations suggesting that the acidic-arginine-aromatic (DRY) triplet of ICL2, the C-terminal portion of ICL2, the C-terminal region of ICL3, and the cytosolic C-terminal tail region were involved in G protein coupling. Our data from transient transfection of the AT 1A receptor in COS7 cells showed that the last 50 amino acid residues (beyond Phe 309 ) were also important for G q coupling (14). Thomas et al. (15) reported that truncation of the last 45 amino acid residues of the rat AT 1A beyond Leu 314 was not important for efficient coupling to the G protein. Thus, we focused on the amino acid sequence between Lys 310 and Leu 314 , constructed five deletional mutants and four substitutional mutants of rat AT 1A , and examined their InsP 3 production and ligand binding. We also synthesized nine peptides based on the amino acid sequence of the cytosolic region and examined their G q activation with the aim of defining a region in the C-tail essential for the coupling to G q in rat AT 1A .

MATERIALS AND METHODS
Mutagenesis-The entire coding region of rat kidney AT 1A was cloned into EcoRI site of a plasmid pUC19 (16). A KpnI-EcoRI fragment was subcloned into polylinker sites of the plasmid vector pBluescript II KSϩ, and single-stranded DNA was prepared using helper phage R 408 (Stratagene). Site-directed mutagenesis was performed by the procedure of Kunkel (17). Sites of truncation and substitution are shown in Fig. 1. The mutated DNA sequences were confirmed by Sanger's dideoxynucleotide sequencing method (18). The mutated AT 1A cDNA was excised with enzymes BamHI and XhoI and introduced into the expression vector pCDNA1.
Stable Expression of Wild Type AT 1A and Its Mutants in CHO-K1 Cells-Forty g of plasmid constructs containing the wild type or mutated rat AT 1A cDNA were co-transfected with 1 g of pSV-G1-Neo (Green Cross Corp.) into 5 ϫ 10 6 of Chinese hamster ovary (CHO-K1) cells in 500 l of phosphate buffer using a gene pulser (Bio-Rad). Native CHO-K1 cells do not express Ang II receptor.
Transfected CHO-K1 cells were cultured for 2 days in 10-cm dishes in Ham's F12 medium (Life Technologies, Inc.) containing 10% fetal calf serum. Then the culture medium was changed to selection medium containing 400 g/ml Geneticin (G418, Life Technologies, Inc.). When individual colonies emerged 10 -14 days after the transfection, 60 sufficiently separated colonies were isolated and inoculated into 200 l of selection medium in 96-well plates. Each of these colonies was scaled up independently to 24-well plates, and the binding assay was performed using 125 I-Ang II (NEN Life Science Products). The binding assay was * This work was supported by Scientific Grant B-06454298 from the Ministry of Education, Science and Culture and the Uehara Memorial Foundation, Japan, and by USPHS Research Grants HL-14192, HL-35323, and HL58205 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
repeated three times for each clone. Two colonies were selected after the binding assay, and 300 single cells isolated from the colonies were cultured in the selection medium in 96-well plates. Two or three weeks later, each of these clones was scaled up independently to 24-well plates, and binding assay was performed again. The clone expressing the highest specific binding of 125 I-Ang II was selected.
Binding Assay-AT 1A -expressing CHO-K1 cells were grown in Ham's F12 medium with 10% fetal calf serum in 24-well plates. They were washed with Hank's balanced salt solution and incubated for 90 min at 37°C in 250 l of Ham's F12 medium with 2% fetal calf serum. Varying concentrations of [ 125 I-Sar 1 ,Ile 8 ]Ang II (NEN Life Science Products) from 0.3 to 10 nM were incubated in this medium for determination of the specific binding. After the incubation, cells were immediately placed on ice, washed three times with ice-cold Hanks' balanced salt solution, and then solubilized with 250 l of 0.5 N NaOH. Radioactivity was measured by a gamma counter. Inositol Trisphosphate (InsP 3 ) Determination-CHO-K1 cells transfected with mutated AT 1A cDNA were grown in 35-mm dishes. Confluent cells were washed with 1.0 ml of 20 mM HEPES buffer, preincubated in 0.5 ml of 20 mM HEPES buffer containing 0.1% bovine serum albumin (BSA) for 20 min at 37°C, then in 0.5 ml of HEPES buffer with 0.1% BSA and 10 mM LiCl for 10 min. Then the cells were incubated in the same HEPES buffer with or without 1 M Ang II for 10 s at 37°C. InsP 3 was extracted with a 1.5-ml mixture of chloroform/methanol, 12 N HCl (1:2:0.05, v/v). A 0.4-ml mixture of chloroform and distilled water (1:1, v/v) was added to the extract and centrifuged. The supernatant was washed with 0.8 ml of chloroform and centrifuged. The supernatant was dried in a Speedvac. The dried extract was redissolved with 150 l of distilled water, sonicated for 30 min, and centrifuged. InsP 3 in 100 l of the supernatant was measured by competitive receptor binding using an InsP 3 assay kit (NEN Life Science Products).
Effect of GTP␥S on Ang II Binding-Transfected CHO-K1 cells were grown in 10-cm dishes, washed with Hanks' balanced salt solution, scraped, and collected by centrifugation at 1500 ϫ g for 5 min. The plasma membrane fraction was prepared by a published method (19). Membranes obtained were suspended at a protein concentration of 250 g/ml in 50 mM Tris buffer (pH 7.4) containing 200 mM NaCl, 10 mM MgCl 2 , 1 mM EDTA, 0.1% BSA, and 100 g/ml phenylmethanesulfonyl fluoride and used as the membrane preparation.
Dose response was determined as follows. Suspended membranes were incubated with 0.1 nM 125 I-Ang II at 25°C for 60 min in the presence of varying concentrations of GTP␥S. For studying the time course of ligand binding, membranes were incubated with 0.1 nM 125 I-Ang II for 60 min at 37°C to attain binding equilibrium and unlabeled Ang II (1 M) or GTP␥S (10 M) or both of them were added to the mixture and incubated for another 60 min. The membrane-bound radioligand was separated from the free radioligand by filtration over glass filters (GF/B) using a cell harvester (Millipore). Radioactivity was measured in a gamma counter.
Synthetic Peptides and Heterotrimeric G q -The amino acid sequences of the peptides used in this study are shown in Fig. 1. They were synthesized by the solid-phase method and purified to 95-99% homogeneity by high performance liquid chromatography using a Nucleosil 5 C18 column eluted with a linear concentration gradient (0 -60%) of CH 3 CN containing 0.1% trifluoroacetic acid. The lyophilized synthetic peptide was dissolved in water. Heterotrimeric forms of G q proteins from bovine liver were purified to homogeneity as published (20).
GTP␥S Binding Assay-[ 35 S]GTP␥S binding to 10 nM purified heterotrimeric G q promoted by synthetic peptides was measured in 25 mM HEPES-NaOH buffer (pH 7.4) containing 120 M MgCl 2 , 100 M EDTA, and 100 nM [ 35 S]GTP␥S in the absence of phospholipids as described by Okamoto et al. (9). Briefly, G q was incubated at 37°C for 10 min in the absence (control) or presence of a synthetic peptide (100 M). Statistical Analysis-The results of experiments with the synthetic peptide study was examined by unpaired Student's t test. p values less than 0.05 were considered significant. Table I the dissociation constants (K d ) and B max values of the wild type AT 1A and its mutants determined by Scatchard analysis were similar, indicating that the mutants possessed similar ligand binding affinity and sites of comparable magnitude. In this study [ 125 I-Sar 1 ,Ile 8 ]Ang II was used as ligand. Scatchard plots indicated single high affinity sites. When 125 I-Ang II was used as ligand results indicated similar single high affinity sites. The possible presence of low affinity sites was practically undetectable.

Binding Affinity of Mutant Receptors-As shown in
Effects of a Stable GTP Analog-As shown in Fig. 2, the binding of 125 I-Ang II to wild type AT 1A , Mut 318-del, and Mut K310,311Q receptors were dose-dependently decreased by GTP␥S, whereas the effect of GTP␥S (shift from a high affinity state to a low affinity form) was practically abolished in Mut 310-del, Mut 312-del, Mut 313-del, Mut 314-del, Mut Y312A, Mut F313A, and Mut L314A receptors.
Time-related changes in dissociation of 125 I-Ang II from the wild type and mutated receptors are shown in Fig. 3. The binding of 125 I-Ang II to the receptors in membrane preparations reached a plateau in 60 min. In wild type AT 1A and all of its mutants, the receptor-bound 125 I-Ang II was displaced by 1 M unlabeled Ang II to similar extents (the range of half-life time of dissociation was 19. InsP 3 Formation-Binding of Ang II to AT 1A activates a PLC via G q resulting in stimulation of InsP 3 formation. Thus, increased InsP 3 formation by Ang II can be considered to indicate effective coupling to G q of the mutants. In unmutated AT 1A , InsP 3 production was significantly increased from 2.52 Ϯ 0.05 pmol/dish of unstimulated control to 16.55 Ϯ 1.88 pmol/dish at 10 s after Ang II stimulation. Similar results were obtained in Mut 318-del and Mut K310,311Q. By contrast, in Mut 310-del, Mut 312-del, Mut 313-del, Mut 314-del, Mut Y312A, Mut F313A, and Mut L314A responses of InsP 3 to Ang II stimulation were abolished (Fig. 4).
Effects of Synthetic Peptides on G Protein Activation-G q was incubated for 10 min in the presence of [ 35 S]GTP␥S with peptides representing domains in the cytoplasmic segments of native AT 1A (P-1 to P-5) or mutated peptides of P-5. As shown in Fig. 5, Peptides P-3 and P-5 activated G q as well as positive control. The G q -activating function was attenuated to 25% relative to intact P-5 in Mut P-5 (Tyr 312 , Phe 313 , and Leu 314 were replaced by alanine). The uptake of [ 35 S]GTP␥S was significantly lower in Mut Y, Mut F (p Ͻ 0.01) and Mut L (p Ͻ 0.05) than in P-5.

DISCUSSION
The cytoplasmic C-terminal (C-tail) region was shown to play an essential role in agonist-induced receptor internalization (15). However, its role in the G protein-coupled phospholipase activation has been controversial. Now in three independent approaches using five deletion mutants, four alanine substitution mutants, and synthetic peptides with native and mutated amino acid sequences corresponding to an N-terminal region of C-tail, we were able to identify the tripeptide region Tyr 312 - Phe 313 -Leu 314 (Fig. 5) as a domain essential for G q activation.
Different experimental approaches produce results leading to different and sometimes contradicting conclusions. Wang et al. (13) using chimeric human AT 1 with a grafted AT 2 C-tail that shows G q activation concluded that the major determinant of G q coupling specificity is in the ICL-3, and the C-tail has little role in the activation of PLC. However, we had shown that deletion beyond Phe 309 (Mut 310-del) abolished the G q -coupled inositol 1,4,5-trisphosphate formation (14). Since both of these modifications could introduce additional factors such as conformational changes or the effect of ICL3 not directly related to the action of deleted or replaced residues, multiple approaches had to be taken. Loss of G q coupling in Mut 310-del and complete recovery of the G q activation in Mut 318-del narrowed the G q coupling domain to residues 310 -317 (Figs. 1 and 4) (14). The observation of a robust activity with Mut 315-del by Thomas et al. (15) further narrowed it to a region between residues 310 and 314. Almost complete loss of the activity with Mut 312-del, 313-del, 314-del, and single residue alanine mutation Y312A, F313A, and L314A and the preservation of a full PLC activity with the double mutant K310Q,K311Q indicated that Tyr 312 -Phe 313 -Leu 314 is the essential domain required for PLC activation. Its essential role in G q coupling was also determined by loss of the well known GTP␥S-induced shift to a low affinity state for agonist binding in these mutants (Figs. 2  and 3). Further evidence for the essential role of the tripeptide sequence for the G protein coupling was obtained by a third and completely independent approach in which peptides with the amino acid sequences of the native and alanine-substituted C-tail (residues 307-320) were allowed to interact with purified heterotrimeric G q , and binding to [ 35 S]GTP␥S was examined. Again, alanine substitution of Tyr 312 -Phe 313 -Leu 314 singly or three together significantly reduced GTP␥S binding. It is interesting to note that, whereas the triple mutant Mut P-5 lost almost the entire binding ability, other mutants, particularly Mut L (L314A), retained recognizable binding, although the conformation of the C-tail domain of AT 1A and the shorter synthetic segment may have a different conformation. These results suggest synergism of the three residues and some difference in the role of Tyr 312 and Leu 314 in G q ␣ activation and GTP␥S binding.
The G protein coupling sites seem to vary from receptor to receptor, and no definitive rules or consensus sequences seem to exist. For example the N-formyl peptide receptor uses ICL2 (22), and G protein specificity of PGE 2 receptor isoforms (EP3) is determined by the C-tail region. More than a single domain could participate in the interaction. The possibility of cooperation of ICL3 and amphipathic ␣-helical structure of the N-terminal region of the C-tail has been proposed by Probst et al. (23). The ␤-adrenergic receptor uses the Cterminal region of ICL3 and the N-terminal region of C-tail for G s activation (10).
The present finding that the synthetic 16-mer peptide P-3 with the amino acid sequence of the N-terminal region of ICL3 and the C-tail peptide (P-5) activated purified G q just as well as the C-tail peptide (P-5) (Fig. 5) supports the observation of Hunyady et al. (24) that the deletion of a sequence (215-226) in this domain abolished G q activation. It also supports the observation of Wang et al. (13) that in chimeras of AT 1 and AT 2 , ICL3 plays dominant roles in G q coupling. Shirai et al. (25) showed the same ICL3 domain activates G i1 , G i2 , and G o by using the synthetic peptide P-3. These results indicate that AT 1A may use and require the tripeptide sequence of C-tail in collaboration with ICL3 in G q activation. Interesting information revealed by the activation of G proteins by these peptides are that the same peptides (P-3 and P-5) are capable of activating G i , G o , and G q . Mechanisms by which a receptor selects the type of G proteins are yet to be clarified. On the other hand, peptides with unrelated sequences like P-1, P-2, and P-4 which did not show the activation may be considered as controls and indicate that activation by P-3 and P-5 is specific to their sequences.
B max values of mutated receptors expressed in each cell line were at levels comparable to that of the wild type. Hence, the decrease in InsP 3 formation in Mut 310-del, Mut 312-del, Mut 313-del, Mut 314-del, Mut Y312A, Mut F313A, and Mut L314A should be due to the receptor mutation rather than a decrease in expression of each mutant receptor. Our previous study using substitutional mutations of basic polar amino acid residues in ICL2 and ICL3 indicated that ICL2 and the C-terminal domain of ICL3 would be important for G q coupling (14). These mutations targeted at domains with dense electrical charges probably caused nonspecific conformational changes and led to erroneous results that could be misinterpreted.
Tyr 292 in transmembrane domain 7 was reported to be essential for G protein coupling (26). The conserved sequence NPLFY at the bottom of transmembrane domain 7 was shown to contribute to both agonist binding and signal transduction (24). Thus, the junctional area of AT 1 between transmembrane domain 7 and C-tail seems to play an important role in receptor signaling. This area of AT 1 also contains the sequence KKFKK that was shown to be an unusual G i activator domain of insulin growth factor II receptor (9). However, in AT 1 mutation to Lys-Lys-Phe-Gln 310 -Gln 311 did not have any effect on G q coupling. This observation helped our work in narrowing the G q activating domain to Tyr 312 -Phe 313 -Leu 314 .
In summary, the present study presents evidence that a G q coupling site in the type 1A angiotensin receptor AT 1A should reside between residues 312 and 318 in the C-terminal tail, and the specific sequence Tyr 312 -Phe 313 -Leu 314 is essential for coupling and activation of the G q protein.