Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The peptide hormone angiotensin II (AngII; Asp¹-Arg²-Val³-Tyr⁴-Ile/Val⁵-His⁶-Pro⁷-Phe⁸), activates type 1 angiotensin receptors (AT₁) to maintain arterial blood pressure and cardiovascular homeostasis. AT₁ receptors (AT_1A and AT_1B in rodents) are seven transmembrane-spanning receptors that couple to G_q/11-phospholipase C-1, leading to intracellular inositol trisphosphate (IP₃) production, calcium mobilization, and activation of protein kinase C. In addition, activated AT₁ receptors couple to mitogen-activated protein kinase (MAPK) and receptor- and soluble-tyrosine kinase pathways (de Gasparo et al., 2000). The high-affinity binding (K_D, ~1 nM) of AngII to the AT₁ receptor, and its subsequent activation, results from multiple interactions between discrete amino acid side chains in AngII and specific residues on the receptor positioned by the arrangement of the seven transmembrane domains. Specifically, two key pairings, Arg² (AngII) with Asp²⁸¹ (receptor) and the -carboxyl of Phe⁸ (AngII) with Lys¹⁹⁹ (receptor), convey high-affinity binding but seem less important for receptor coupling to IP₃ generation (Feng et al., 1995; Noda et al., 1995). This initial docking positions the aromatic side chains of Tyr⁴ and Phe⁸ (AngII) to engage, respectively, Asn¹¹¹ and His²⁵⁶ on the receptor, which instigates receptor activation, coupling to G protein(s) and subsequent phospholipase C-mediated signals (Noda et al., 1995; Noda et al., 1996), but contributes a smaller amount to binding affinity.

The ligand-mediated activation of AT₁ receptors is accompanied by ancillary regulatory events, including receptor phosphorylation, arrestin binding and receptor internalization (Thomas, 1999; Hunyady et al., 2000; Oakley et al., 2001). These processes are initiated soon after ligand-receptor interaction and may contribute to rapid receptor desensitization. The AT₁ receptor is phosphorylated by both G protein-coupled receptor (GPCR) kinases and protein kinase C on specific carboxyl-terminal serine and threonine residues (Thomas, 1999; Hunyady et al., 2000). Phosphorylation by GPCR kinases presumably produces a high-affinity binding site for arrestins, which sterically hinder further G protein coupling and act as adapters for the cellular internalization machinery (Lefkowitz, 1998). Alternatively, arrestins may act as scaffolds to recruit additional signaling molecules, including tyrosine kinases and MAPKs, to GPCRs (Pierce et al., 2001). Phosphorylation of the AT_1A receptor carboxyl terminus has been implicated in rapid agonist-induced endocytosis (Smith et al., 1998; Thomas et al., 1998), presumably via arrestin recruitment (Oakley et al., 2001).

Recent evidence indicates that multiple conformations exist for the AT_1A receptor, some coupling the receptor to signaling pathways and others directing receptor phosphorylation or internalization (Thomas et al., 2000). The transition of the AT_1A receptor through these various states may be induced or stabilized by specific AngII side chains during the docking of ligand onto receptor, because separate residues within the AngII peptide confer high-affinity binding and IP₃ signaling capacity (Feng et al., 1995; Noda et al., 1995; Noda et al., 1996). The phenomenon of ligand-specific receptor states has been referred to as agonist-receptor trafficking (Kenakin, 1995) or biased agonism (Jarpe et al., 1998). This raises a question: do specific points of contact between AngII and the AT_1A receptor induce/stabilize discrete receptor states that dictate receptor processes, such as internalization, phosphorylation, and MAPK activation?

In this study, we examined the capacity of AngII and a series of substituted analogs to promote phosphorylation and internalization of the AT_1A receptor and MAPK signaling. We observed a key role for Phe⁸ of AngII in receptor phosphorylation and activation of MAPK, with important differences in tolerance to specific substitutions. In contrast, receptor internalization, as visualized by confocal microscopy of an enhanced green fluorescent protein-tagged AT_1A receptor, was unaffected by substitutions at any position in AngII.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

AngII Analogs. For AngII analogs used in this study, see Table 1.

The AT_1A-Enhanced Green Fluorescent Protein Receptor. An N-terminal hemagglutinin epitope-tagged AT_1A receptor (NHA-AT_1A) (Thomas et al., 1998) was amplified using polymerase chain reaction with T₇ primer (sense), upstream of a unique HindIII site, and an antisense primer (5'-CAGGATCCCCCTCCACCTCAAAACAAGACGCAGG-3'), which removed the stop codon from the receptor and incorporated a BamHI site (underlined). The polymerase chain reaction product was digested with HindIII and BamHI and inserted into the multiple cloning site of the pEGFP-N1 vector to generate an NHA-AT_1A-EGFP fusion construct. The construct was sequenced to confirm the integrity of the AT_1A coding region and the fusion to EGFP.

Cell Culture and Transfection. Chinese hamster ovary (CHO-K1) cells were maintained in -minimum essential medium containing 10% fetal bovine serum supplemented with antibiotics and antimycotics (complete media). Cells were transferred to 12-well plastic culture dishes, grown to 70% confluence, and transiently transfected using LipofectAMINE as described previously (Thomas et al., 1998), with 0.025 µg of NHA-AT_1A or NHA-AT_1A-EGFP receptor plasmid DNA and 0.6 µg of vector plasmid DNA per well. For eight-well slides, cells were seeded at a density of 100,000 to 140,000 cells/well, grown overnight, and transfected with LipofectAMINE (1.4 µl/well), NHA-AT_1A-EGFP receptor plasmid DNA (7 ng/well), and vector DNA (164 ng/well). All transfected cells were grown in complete media for 48 h and serum-starved overnight before experiments.

Radioligand Binding Assays. Radioligand binding assays on transiently transfected CHO-K1 cells were performed as described previously (Thomas et al., 1995) using the AngII antagonist [¹²⁵I]Sar¹Ile⁸AngII as tracer. Equilibrium binding was for 5 h at 4°C.

Radioligand Internalization Assay. CHO-K1 cells expressing the NHA-AT_1A or NHA-AT_1A-EGFP receptor were incubated for 1, 2, 5, 10, or 20 min at 37°C with [¹²⁵I]AngII. They were then washed; cell surface-bound ligand was stripped by acid-washing and collected. Radioactivity associated with the cells (internalized) was determined as a percentage of the total (acid wash plus cell-associated) (Thomas et al., 1998).

Cytosensor Microphysiometer Signaling Assay. CHO-K1 cells transiently transfected with NHA-AT_1A or NHA-AT_1A-EGFP were plated into 12-mm Transwells (3-µm pore diameter) at a density of 250,000 cells/well and grown overnight. Transwells with spacers and capsule inserts were placed in the sensor chambers and the cells were allowed to equilibrate for 1 to 2 h. When a steady state was achieved, AngII accumulative concentration-response curves were constructed over the concentration range 10¹¹ to 10⁶ M. Responses were measured as the rate of change of pH, both as a voltage change and as a percentage change from baseline voltage (normalized to 100%) using Cytosoft software.

Receptor Internalization Measured by Confocal Microscopy. CHO-K1 cells were seeded in eight-well chamber slides at a density of 100,000 cell/well, transfected as described above, incubated in complete media for 36 h, and serum-deprived overnight. Cells were equilibrated (37°C, 1 h) in 0.4 ml of angiotensin receptor binding buffer (ARBB): 50 mM Tris-HCl, pH 7.4, containing 120 mM NaCl, 4 mM KCl, 5 mM MgCl₂, 1 mM CaCl₂, 10 µg/ml bacitracin, 2 mg/ml D-glucose, and 0.25% bovine serum albumin. Ligands were added to give a final concentration of 100 times their respective published K_D values (see Table 1) and slides were incubated at 37°C for 20 min. For AngII concentration-response experiments, cells were stimulated with 0, 10¹¹, 10¹⁰, 10⁹, 10⁸, and 10⁷ M AngII. Wells were aspirated and cells fixed with 4% paraformaldehyde in sodium phosphate buffer, pH 7.4 (0.5 ml), for 20 min at room temperature, washed with ARBB, incubated in ARBB with 10% fetal bovine serum and 100 mM glycine for 10 min, washed, and incubated with wheat-germ agglutinin/Texas Red (WGA-TR) conjugate (2 µg/ml) for 30 min at room temperature. WGA-TR was crosslinked to cells by incubating in 4% paraformaldehyde for 20 min. Chambers were removed from the slide, washed twice with Hanks' buffered salt solution (HBSS) and mounted in 90% glycerol/10% HBSS. All stimulations were performed coded and protocols withheld until after analysis of confocal imaging.

MAPK Activation. Serum-starved CHO-K1 cells expressing the AT_1A receptor (in 12 well plates) were stimulated for 4 min with AngII or the substituted AngII analogs (Table 1), rapidly washed twice with ice-cold HBSS, and lysed in 250 µl of radioimmunoprecipitation buffer (50 mM Tris, pH 7.4, 100 mM NaCl, 2 mM EDTA, 50 mM sodium fluoride, 10 mM sodium pyrophosphate, 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 1 µg/ml aprotinin, 5 µg/ml leupeptin, and 1 µg/ml pepstatin) for 30 min at 4°C. Cell lysates were centrifuged at 14,000g for 15 min and 10 µl of supernatant was resolved on SDS-PAGE and Western blotted. Western blots were probed with a monoclonal antibody (E10) to phospho-p44/42 MAPK (T202/Y204) to identify the phosphorylated (active) forms of MAPK and developed by enhanced chemiluminescence. Blots were subsequently reprobed with a rabbit polyclonal antibody (SC93) to detect total MAPK. Blots were quantified using Scion Image software.

Receptor Phosphorylation Assay. The procedure for receptor phosphorylation has been described previously (Thomas et al., 1998). ³²P-loaded cells were stimulated (10 min, 37°C) with AngII and the various substituted analogs at a concentration 100 times that of the K_D for the receptor. HA-tagged AT_1A receptors were immunoprecipitated using anti-HA antibody (12CA5) and phosphorylated receptor resolved by 10% SDS-PAGE and a filmless autoradiographic system (Fujix Bio-imaging Analyzer BAS 1000; Fuji, Tokyo, Japan).

Equipment and Reagents. Equipment and reagents used and their source are as follows: CHO-K1 cells (American Type Culture Collection, Manassas, VA); cell culture media, additives, LipofectAMINE, pEGFP-N1 vector (Invitrogen, Melbourne, Australia); cell culture dishes and 12-mm transwells (Costar, Acton, MA); eight-well glass chamber slides (Nalge Nunc International, Naperville, IL); [³²P]orthophosphate (ICN Biomedicals, Costa Mesa, CA); protein A-agarose (Roche Applied Science, Melbourne, Australia); wheat germ agglutinin/Texas Red conjugate (Molecular Probes, Eugene, OR); angiotensin II (Auspep, Melbourne, Australia); [¹²⁵I]AngII and [¹²⁵I]Sar¹Ile⁸AngII (specific activity, ~2000 Ci/mmol; Austin Biomedical Services, Melbourne, Australia); other AngII analogs were synthesized in the Cleveland Clinic Core synthesis facility (Cleveland, OH). The MAPK antibodies E10 and SC93 were from Cell Signaling Technologies (Beverly, MA) and Santa Cruz Biotechnology (Santa Cruz, CA), respectively. All other chemicals were bought from Sigma (Sydney, Australia) or BDH (Kilsyth, Melbourne, Australia); Cytosensor Microphysiometer and Cytosoft software (Molecular Devices, Menlo Park, CA); LaserSharp Acquisition software (Bio-Rad, Hercules, CA); Optimas Image Analysis (Media Cybernetics, Del Mar, CA) and Scion Image software (Scion, Frederick, MD).

Statistical Analysis. Statistics performed were the unpaired t test and one-way analysis of variance (Newman-Keuls multiple-comparison test) calculated by Prism data analysis software (v.2.0; GraphPad, San Diego, CA). Figures are presented as mean ± S.E.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Comparison of the AT_1A-EGFP and Wild-Type Receptors. To examine the internalization of AT_1A receptors, we generated a NHA-AT_1A-EGFP chimera suitable for following receptor trafficking by confocal microscopy. Figure 1 shows a comparison of receptor binding, signaling and internalization for the wild-type (NHA-AT_1A) and NHA-AT_1A-EGFP receptors. Competition binding studies, using the radio-labeled peptide antagonist [¹²⁵I]Sar¹Ile⁸AngII, demonstrated that both receptors expressed equally well in CHO-K1 cells (~1000 fmol of receptor/mg of protein) and displacement experiments with AngII, EXP3174 (an AT₁-selective nonpeptide antagonist), or PD123319 (an AT₂-selective nonpeptide antagonist) yielded identical binding profiles (Fig. 1A). AngII displaced the iodinated ligand with pK_I values of 8.13 ± 0.06 for the NHA-AT_1A receptor and 8.10 ± 0.08 for the NHA-AT_1A-EGFP receptor. pK_I values for the AT₁ receptor selective nonpeptide antagonist, EXP3174, were calculated as 7.60 ± 0.08 for NHA-AT_1A and 7.67 ± 0.07 for NHA-AT_1A-EGFP. PD123319, an AT₂ receptor selective nonpeptide antagonist, did not significantly displace the labeled peptide as indicated by pK_I values <6.0 for both receptors.

View larger version (23K): [in this window] [in a new window]   Fig. 1.   Comparison of ligand affinity, signaling capacity, and internalization of the NHA-AT1A and NHA-AT1A-EGFP receptors transiently expressed in CHO-K1 cells. A, competition binding assay for [125I]Sar1Ile8AngII displaced with AngII, EXP3174, and PD123319 for the NHA-AT1A (filled symbols, solid lines) or NHA-AT1A-EGFP (empty symbols, dashed lines) receptors. Binding assays (5 h, 4°C) were performed with ~30 pmol of [125I]Sar1Ile8AngII (60,000 cpm) in the presence of increasing concentrations of competing ligand. Data are mean ± S.E. from four experiments. B, signaling of the NHA-AT1A (--) and NHA-AT1A-EGFP (- -- -) receptors was measured by a Cytosensor Microphysiometer. Cells transiently expressing the NHA-AT1A or NHA-AT1A-EGFP receptors was perfused with various concentrations of AngII and the rate of pH change of the surrounding fluid was measured. Data are expressed as the percentage change from baseline (normalized to 100%) and represented as mean ± S.E. from three experiments. C, radioligand internalization comparing the NHA-AT1A (--) and NHA-AT1A-EGFP (- -- -) receptors. Cells transiently expressing the NHA-AT1A or NHA-AT1A-EGFP receptors were incubated with [125I]AngII at 37°C for indicated times. Bound ligand was stripped from the cell surface receptors and the amount of intracellular radioactivity was expressed as a percentage of the total specifically bound counts (intracellular + cell surface). Data are mean ± S.E. of four experiments.

11

6

Internalization of NHA-AT_1A-EGFP Receptor Measured by Confocal Microscopy. AngII-induced internalization of the NHA-AT_1A-EGFP receptor was examined by confocal microscopy in CHO-K1 cells that transiently expressed the receptor (see Fig. 2). Cells expressing a moderate level of receptor were chosen and optical sections were taken through cells at a point where the nucleus was largest. To differentiate between cell surface receptor and intracellular receptor, the plasma membrane of cells was defined by staining with a WGA-TR conjugate, which binds to carbohydrates on the cell surface of nonpermeabilized cells.

View larger version (39K): [in this window] [in a new window]   Fig. 2.   Confocal imaging and time-dependence of internalization of the NHA-AT1A-EGFP receptor transiently expressed in CHO-K1 cells. A, single optical section of a transfected CHO-K1 cell taken with a confocal microscope. Left, NHA-AT1A-EGFP receptor fluorescence (green, labeled AT1A-EGFP) for an unstimulated cell. Scale bar, 2 µm. Middle, wheat-germ agglutinin-Texas Red (red, labeled WGA-TR) fluorescence bound to the cell membrane of the same cell. Right, an overlay of red and green images resulting in a yellow pseudocolor, where red and green colocalize. B, CHO-K1 cells expressing the AT1A-EGFP receptor were stimulated with AngII (107 M) for 0, 1, 2, 5, 10 and 20 min, fixed, and stained with WGA-TR. Representative overlay pictures are shown for each time-point. C, the proportion of intracellular receptor was quantified for nine cells at each time-point (as described under Materials and Methods). Data are mean ± S.E. of three experiments.

7

11

7

7

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Characterization of AT1A-EGFP receptor internalization. A, concentration-response curve for NHA-AT1A-EGFP receptor internalization. Cells expressing the receptor were stimulated with AngII at concentrations over the range 1011 to 107 M for 20 min. The proportion of intracellular receptor was measured by confocal microscopy and the concentration response-curve exhibited a pEC50 of 9.27 ± 0.27. Data are mean ± S.E. of three experiments, nine cells per concentration in each experiment. B, the nonpeptide AT1 receptor antagonist EXP3174 does not stimulate internalization of the NHA-AT1A-EGFP receptor and inhibits AngII-mediated internalization. Cells expressing the NHA-AT1A-EGFP receptor were stimulated with AngII (107M, 20 min), EXP3174 (105M, 20 min) or with AngII (107M, 20 min) after pretreatment with EXP3174 (105M, 20 min), and processed for confocal microscopy. (**, p < 0.01 of AngII; Con, unstimulated).

5

5

7

Internalization by Substituted Analogs of AngII. Cells expressing the NHA-AT_1A-EGFP receptor were stimulated for 20 min with analogs of AngII at a concentration equal to 100-fold the K_D of that analog at the AT_1A receptor (Table 1). Saturating concentrations and the 20-min time-point were chosen to allow maximal internalization of each analog. Because substitutions in AngII at positions 4 and 8 lead to profound decreases in AT_1A receptor activation, as measured by inositol phosphate generation [see Table 1 (Miura et al., 1999; Miura and Karnik, 1999)], we measured the internalization caused by AngII with analogs substituted at these positions. As shown in Fig. 4A, both mono-substituted (Sar¹Ala⁴AngII, Sar¹Ala⁸AngII) and di-substituted analogs (Sar¹Ile⁴Ile⁸AngII, Sar¹Gly⁴Gly⁸AngII) induced full internalization of the NHA-AT_1A-EGFP receptor.

View larger version (25K): [in this window] [in a new window]   Fig. 4.   Effect of substitution of AngII side chains on AT1A-EGFP receptor internalization. A, internalization of NHA-AT1A-EGFP receptor by analogs of AngII substituted at positions 4 and 8. Cells expressing the NHA-AT1A-EGFP receptor were stimulated for 20 min at a concentration of 100 × KD of the specified analog and intracellular receptor quantified by confocal microscopy. (A4, Sar1Ala4AngII; A8, Sar1Ala8AngII; I48, Sar1Ile4Ile8AngII; G48, Sar1Gly4Gly8AngII; Con: unstimulated). B, NHA-AT1A-EGFP receptor internalization by analogs of AngII substituted at positions 1, 2, 3, 6, 7, and the -carboxyl group of Phe8 (see Table 1). Cells expressing the receptor were stimulated for 20 min at a concentration of 100 × KD of specified analogs. (A1, Ala1AngII. G2, Sar1Gln2AngII; A3, Sar1Ala3AngII; A6, Sar1Ala6AngII; A7, Sar1Ala7AngII; Am, Sar1AngII-CONH2). Quantification was performed by confocal microscopy from nine cells/analog/experiment. Data are mean ± S.E. of four experiments. AngII analog codes are from Table 1.

AT_1A Receptor Phosphorylation by AngII Analogs. We examined the capacity of the various AngII analogs to cause phosphorylation of the AT_1A receptor, expressed in CHO-K1 cells, using the N-terminal HA tag to immunoprecipitate the receptor. AngII caused a robust phosphorylation of the AT_1A receptor (Fig. 5). Maximal receptor phosphorylation in response to AngII is observed at concentrations 10- to 100-fold higher than the K_D and at 10 min after ligand stimulation (Thomas et al., 2000). In comparison to AngII, substitutions at Arg² (G2), Val³ (A3), Tyr⁴ (A4), His⁶ (A6), and Pro⁷ (A7) or substitution of the -carboxyl group of Phe⁸ (Am) had no significant effect on receptor phosphorylation. In contrast, alanine substitution of Asp¹ (A1) led to a ~40% reduction in AT_1A receptor phosphorylation, which was mirrored by a similar reduction with AngIII (data not shown). The most important residue for phosphorylation seems to be Phe⁸. Most substitutions at this position caused significant decreases in receptor phosphorylation. Single substitutions of Phe⁸ to alanine (A8), -cyclohexylalanine (a nonaromatic, saturated ring side chain of equivalent size and hydrophobicity to phenylalanine/tyrosine) (C8), isoleucine (I8), or diphenylalanine (a strongly aromatic side chain of increased size compared with phenylalanine/tyrosine) (D8) caused significant (45 to 70%) decreases in AT_1A receptor phosphorylation compared with AngII. Double isoleucine substitution at positions 4 and 8 (I48) decreased receptor phosphorylation similar to the single isoleucine substitution at position 8 (I8), consistent with a lack of effect of position 4 substituted analogs on receptor phosphorylation. Remarkably, the double substitution to glycine at position 4 and 8 (G48) did not significantly reduce receptor phosphorylation compared with AngII. So, although single substitutions at Phe⁸ can inhibit receptor phosphorylation somewhat, there was no clear preference for size, hydrophobicity, or aromaticity at this position. Moreover, based on the strong phosphorylation induced by G48, it would seem that the presence of both "agonism" defining residues, Tyr⁴ and Phe⁸, is not essential for AT_1A receptor phosphorylation.

View larger version (60K): [in this window] [in a new window]   Fig. 5.   Phosphorylation of the NHA-AT1A receptor transiently expressed in CHO-K1 cells stimulated with AngII analogs. A, CHO-K1 cells expressing the receptor were preloaded with 0.2 mCi/ml [32P]Pi and stimulated for 20 min with AngII or substituted analogs of AngII at concentrations 100 times their respective KD (see Table 1). The cells were lysed and the receptor immunoprecipitated and resolved on a 10% SDS-PAGE gel. Radioactively labeled receptors were visualized and quantified with a filmless autoradiographic system (Fujix Bio-imaging Analyzer BAS 1000) and presented as a percentage of the phosphorylation caused by AngII. The phosphorylated receptor migrates as a broad band running at 60 to 110 kDa. B, gels were quantified according to the band intensity and expressed as a percentage of the phosphorylation induced by AngII (100%). Data are mean ± S.E. of three experiments (*, p < 0.05; **, p < 0.01 compared with AngII). AngII analog codes are from Table 1; Con, unstimulated.

Activation of MAP Kinase by AngII Analogs. Previous studies on the structure-activity profiles of AngII analogs with respect to their ability to activate AT₁ receptor signaling have focused principally on the G_q/11-PLC/IP₃ pathway. These studies have revealed strict requirements for position 4 and 8 in generating the active signaling form of the receptor (Miura et al., 1999; Miura and Karnik, 1999). To examine whether other signaling pathways activated by the AT_1A receptor are similarly affected by substitutions in AngII, the activation of MAPK (p42/44; extracellular signal-regulated kinase) after stimulation by the series of AngII analogs was determined (Fig. 6). AngII stimulation (4 min at 37°C) of the AT_1A receptor lead to a robust (~6-fold) increase in MAPK activation, primarily through activation of p42. Substitutions at position 1, 2, 3, 4, 6, and 7, the -carboxyl of Phe⁸, and the substitution of Phe⁸ with diphenylalanine produced levels of MAPK activation similar to that of AngII. In contrast, substitution of Phe⁸ with alanine completely abolished activation. Single substitutions of Phe⁸ with -cyclohexylalanine or isoleucine also significantly reduced MAPK activation compared with AngII. Similarly, double isoleucine and glycine substitutions of Tyr⁴ and Phe⁸ peptides weakly activated MAPK.

View larger version (61K): [in this window] [in a new window]   Fig. 6.   MAPK activation induced by AngII analogs. A, cells expressing the AT1A receptor were stimulated for 4 min (maximum activation) with AngII or substituted analogs of AngII at concentrations 100 times their respective KD. Cells were lysed and cellular proteins resolved on 10% SDS-PAGE followed by Western blotting. Blots were probed with a monoclonal antibody (E10) to phospho-p44/42 MAPK (T202/Y204) to identify the phosphorylated (active) forms of MAPK and reprobed with a rabbit polyclonal antibody (SC93) to detect total MAPK. Blots were developed using enhanced chemiluminescence and band intensity quantified (B) using Scion Image software. Active MAPK was normalized for loading differences using the intensity of the corresponding Total MAPK lane. The data are means ± S.E. from three to five experiments. ***, p < 0.001 compared with AngII; §§, p < 0.01 compared with control; §§§, p < 0.001 compared with control. AngII analog codes are from Table 1.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

A major initial aim of this study was to generate an AT_1A-EGFP chimeric receptor, which would allow visualization and quantification of AT_1A receptor endocytosis using confocal microscopy. Direct quantification was preferred because previous methods for examining AT₁ receptor internalization have been predominantly indirect (e.g., acid-insensitive sequestration of radioactive AngII). The unobtrusive nature of the carboxyl-terminal EGFP tag on AT_1A receptor binding affinity, selectivity, and receptor-mediated cellular activation was confirmed by competition binding experiments, Cytosensor measurements, and receptor phosphorylation assays. That these parameters were indistinguishable for EGFP-tagged and -untagged AT_1A receptors agrees with studies on other GFP-tagged GPCRs (Milligan, 1999) and with a recent report (Miserey-Lenkei et al., 2001), in which an AT_1A-EGFP receptor was used to investigate AT_1A receptor trafficking in human embryonic kidney 293 cells.

AT_1A-EGFP receptor internalization was quantified using confocal microscopy, using staining with WGA-TR to delineate the cell surface. Receptor endocytosis measured in this way was qualitatively equivalent to the traditional acid-insensitive [¹²⁵I]AngII binding method: internalization was rapid (t_1/2 = 2.1 min) and reached a steady state by 10 to 20 min, indicating that ligand sequestration correlates with receptor internalization. The pEC₅₀ of AT_1A-EGFP receptor internalization as determined by confocal microscopy (9.27 ± 0.27) approximated the K_D value of AngII at the AT_1A receptor (~ 1 nM), which highlights the close relationship between ligand binding and internalization. The graded response to AngII demonstrates that the assay can distinguish different levels of receptor internalization. The nonpeptide antagonist EXP3174 caused no internalization of the AT_1A-EGFP receptor, demonstrating that ligand binding to the receptor is not sufficient for internalization and that additional points of contact, or conformational changes in the receptor not provided by EXP3174, are required for this process. Importantly, the internalization in response to AngII was prevented by EXP3174, verifying that a specific interaction between AngII and the AT_1A receptor causes internalization.

The principal, yet unexpected, finding of our study was that AT_1A receptor internalization is unaffected by substitutions at any position within the AngII peptide sequence. This result contrasts with the accepted view that the two aromatic amino acid side-chains of AngII, Tyr⁴ and Phe⁸, are required for activating IP₃ signaling through the AT₁ receptor. Substitution of these residues, reducing their bulk and removing their aromaticity, has been shown in vivo, in isolated tissues, (Regoli et al., 1974; Samanen et al., 1989), and in cells expressing a recombinant AT₁ receptor (Noda et al., 1996; Miura et al., 1999), to significantly reduce responses compared with AngII. Substitution of Phe⁸ or Tyr⁴ with isoleucine yields peptides with reduced ability to generate IP₃ (~20% and 80% of AngII, respectively) (Miura et al., 1999), yet we observed no effect of these two individual substitutions on the level of internalization. Even the di-substituted analogs (Sar¹Ile⁴Ile⁸AngII and Sar¹Gly⁴Gly⁸AngII) that display no agonism with respect to inositol phosphate signaling (see Table 1) gave full AT_1A-EGFP receptor internalization. Consistent with our findings, previous studies have reported strong sequestration of [¹²⁵I]Sar¹Ile⁸AngII (Conchon et al., 1994; Thomas et al., 1996) and other Phe⁸ analogs of AngII also cause AT_1A-EGFP receptor internalization (Miserey-Lenkei et al., 2001). This evidence shows that the AngII residues involved in coupling the receptor to inositol phosphate signaling are not required for receptor internalization, indicating that distinct ligand-receptor interactions and receptor conformations subserve PLC signaling and internalization.

Given that phosphorylation within the central region of the AT_1A carboxyl terminus is associated with receptor internalization (Smith et al., 1998; Thomas et al., 1998), we investigated whether substitutions in AngII would affect receptor phosphorylation. Only changes to Asp¹ (to alanine) and Phe⁸ were found to inhibit phosphorylation significantly, although none completely. A comparison of Phe⁸-substituted AngII analogs revealed similar reduced phosphorylation for alanine, isoleucine, diphenylalanine, and -cyclohexylalanine replacement, suggesting that neither position 8 side chain size, hydrophobicity, nor aromaticity is predominant in determining AT_1A receptor phosphorylation. Moreover, double substitution on Tyr⁴ and Phe⁸ with isoleucine produced a level of phosphorylation similar to the single isoleucine replacement for Phe⁸, indicating a minimal contribution of Tyr⁴ to AT_1A phosphorylation. This result was surprising because mutation of Asn¹¹¹ in the AT_1A receptor, the residue presumed to make productive contact with Tyr⁴ of AngII, produces a receptor incapable of AngII-induced phosphorylation (Thomas et al., 2000). Remarkably, double-glycine substitution of Tyr⁴ and Phe⁸ produced a level of receptor phosphorylation equivalent to that generated by AngII, which would seem difficult to reconcile given the contribution of position 8 revealed by the other analogs. We can only speculate that removal of the Phe⁸ (and Tyr⁴) side chains favors a phosphorylatable conformation.

An additional insight into the differing requirements for receptor function is provided by the activation of MAPK by the AngII analogs. Alanine was the least tolerated residue at position 8, resulting in little MAPK activation. In contrast, -cyclohexylalanine (nonaromatic) or isoleucine (nonpolar) substitution was moderately tolerated, whereas diphenylalanine (strongly aromatic) was fully tolerated, indicating a key contribution of side chain aromaticity and perhaps hydrophobicity to MAPK activation. Single substitutions of Tyr⁴ to alanine (A4) and -cyclohexylalanine (C4) failed to reduce MAPK activation compared with AngII and Sar¹Ile⁴Ile⁸AngII (I48) produced a similar level of MAPK activation to Sar¹Ile⁸AngII (I8). Thus, in contrast to inositol phosphate signaling, where substitution with -cyclohexylalanine at Tyr⁴ (C4) leads to significant inhibition of signaling, Tyr⁴ seems to contribute little to MAPK signaling. Moreover, double substitution at Tyr⁴ and Phe⁸, with isoleucine or glycine, which abrogates inositol phosphate signaling, reduces MAPK signaling by only half.

Shown in Fig. 7 are separate comparisons of receptor internalization, inositol phosphate production, phosphorylation and MAPK activation, with respect to each other, after stimulation by AngII and the substituted analogs. These comparisons reveal differences in the tolerance of the various receptor processes to substitutions in AngII. For example, when internalization is compared separately with inositol phosphate production (Fig. 7A), phosphorylation (Fig. 7B), and MAPK activation (Fig. 7C), it is clear that the di-substituted analogs (Sar¹Ile⁴Ile⁸AngII and Sar¹Gly⁴Gly⁸AngII; I48 and G48) promote full receptor internalization, slightly reduced (80 to 90% of AngII) phosphorylation (Fig. 7B), a reduced (40% of AngII) MAPK activation (Fig. 7C), and completely fail to elicit inositol phosphate production (Fig. 7A). Ala¹AngII (A1) fully activates IP and MAPK, but shows a slight reduction (60% of AngII) in receptor phosphorylation (Fig. 7B). Sar¹Ala⁸AngII (A8) elicits greatly reduced IP and MAPK responses (20% of AngII) and reduced capacity for phosphorylation (50% of AngII, Fig. 7B). Figure 7D highlights the structural requirements for position 8 with respect to receptor phosphorylation and MAPK activation. Although substitution Phe⁸ with isoleucine (I8) reduced phosphorylation (50% of AngII) and MAPK activation (40% of AngII), substitution with the strongly aromatic diphenylalanine (D8) causes reduced (50%) receptor phosphorylation but full MAPK activation. An AngII peptide carrying the nonaromatic ring side chain, -cyclohexylalanine, at position 8 (C8) reduces receptor phosphorylation (30% of AngII) more than reducing MAPK activation (60% of AngII). These results suggest that aromaticity and bulk at position 8, although important for MAPK activation, are not crucial for phosphorylation. The bulky, but not aromatic, substitution at position 4 (Sar¹Cha⁴AngII, C4) produces a ligand that promotes full phosphorylation (Fig. 7E) and MAPK (Fig. 7F) but only an intermediate inositol phosphate response. These data demonstrate differing structural and size requirements in the AngII peptide for AT_1A receptor activation, signaling, phosphorylation and internalization.

View larger version (24K): [in this window] [in a new window]   Fig. 7.   Comparison of individual receptor activities in response to stimulation by AngII analogs. Data for receptor internalization (Fig. 4), phosphorylation (Fig. 5), MAPK activation (Fig. 6), and inositol phosphate production (Table 1) were expressed as a percentage of AngII (100%). Values are corrected for basal values of the unstimulated control samples. This normalized data for the individual receptor processes was plotted to compare all receptor activities. A theoretical line (dashed), representing the position at which a particular substitution would affect two receptor functions to the same degree, is shown. For clarity, only analogs whose responses are shifted from full activity are labeled using the code described in Table 1.

Multiple functional ligand-receptor contacts, variously referred to as agonist-receptor trafficking (Kenakin, 1995) or biased agonism (Jarpe et al., 1998), support the concept of separate receptor conformational states. Based on contemporary two- or three-state models for GPCR activation (Lefkowitz et al., 1993; Kenakin, 1995; Leff, 1995; Leff et al., 1997; Leurs et al., 1998; Scaramellini and Leff, 1998), receptor conformations that activate signaling (e.g., R*) are also targeted for phosphorylation by GPCR kinases, leading to arrestin binding, which inhibits further signaling and targets receptors for clathrin-mediated internalization. Thus, substitutions in the AngII peptide (specifically at Tyr⁴ and Phe⁸) should equally affect signals emanating from the receptor, phosphorylation and internalization. Instead, we observed differences in the capacity of AngII analogs to promote receptor signaling, phosphorylation, and internalization. Most compelling was the observation that all substituted analogs stimulated robust AT_1A receptor internalization, indicating that internalization is more tolerant than signaling of modifications in the AngII peptide. These data argue strongly against a linear transition from an inactive to a signaling receptor, which is then phosphorylated, desensitized, and internalized. Instead, they support the idea of ligand-specific receptor states, each selected by unique contacts between peptide and receptor and potentially capable of coupling to one or more distinct receptor activities. A caveat to this interpretation is that we used saturating concentrations of ligand with single, maximal time-points of stimulation in one cell type (CHO-K1). Clearly, cellular environment (i.e., the complement and/or concentration of G proteins and other signaling and regulatory molecules) may affect the capacity of specific analogs to promote signaling, phosphorylation, and internalization. Also, experiments measuring initial rates of reactions may reveal subtle differences in the kinetics and/or efficacy of substituted AngII analogs not revealed by the current experimental approach.

Nevertheless, the concept of distinct AT_1A receptor states is consistent with AT_1A receptor mutants that display proclivity for specific receptor functions. For example, an Asp⁷⁴Asn AT_1A mutant undergoes full internalization yet does not signal (Conchon et al., 1994; Miserey-Lenkei et al., 2001); an AT_1A receptor carrying five tyrosine mutations is deficient in inositol phosphate and calcium signaling (Doan et al., 2001) but retains robust AngII-induced intracellular tyrosine kinase signaling and cell proliferation; and AngII causes internalization, but not phosphorylation, of constitutively active AT_1A receptor mutants (Thomas et al., 2000).