G_βγ-independent constitutive association of G_αs with SHP-1 and angiotensin II receptor AT₂ is essential in AT₂-mediated ITIM-independent activation of SHP-1

Abstract

Conventional mode of activation of SH2 domain-containing phosphatase 1 (SHP-1) by a single transmembrane (TM) inhibitory receptor such as killer cell inhibitory receptor, Fcγ receptor type IIb1, and paired Ig-like receptors of inhibitory types requires tyrosine phosphorylation of immunoreceptor tyrosine-based inhibitory (ITIM) motifs in the cytoplasmic domains of the inhibitory receptors. Contrary to this paradigm, AT₂, a G protein-coupled 7TM receptor that does not undergo tyrosine phosphorylation in response to angiotensin II (Ang II) stimulation, also activates SHP-1. Here we show that SHP-1 constitutively and physically associates with AT₂ receptor in transfected COS-7 cells. On stimulation by Ang II, SHP-1 becomes activated and dissociated from AT₂ receptor, independent of pertussis toxin. Cotransfection of transducin G_βγ inhibits SHP-1/AT₂ association and the SHP-1 activation, whereas cotransfection of C-terminal of β-adrenergic receptor kinase, which abrogates G_βγ signaling, facilitates SHP-1 activation. Surprisingly, SHP-1/AT₂ association and the SHP-1 activation requires the presence of G_αs as shown by differential coimmunoprecipitation, dominant negative G_αs, constitutively active G_αs, and G_α peptides. A mutant AT₂ receptor D141A–R142L that is inactive in G_α protein activation constitutively associates with SHP-1 and activates it. Together, these results indicate that G_αs alone, rather than exclusively in the form of G_αβγ heterotrimer may facilitate signal transduction for G protein-coupled receptors, suggesting a novel mechanism distinct from the classic paradigm of heterotrimeric G proteins. The AT₂-mediated ITIM-independent activation of SHP-1 that is distinct from the conventional mode of activation, may represent a general paradigm for activation of SHP-1/2-class tyrosine phosphatases by G protein-coupled receptors.

The G protein-coupled receptors (GPCR) AT₁ and AT₂ are the two major subtypes that transduce signals for angiotensin II (Ang II), a potent vasoactive octapeptide (Asp¹–Arg²–Val³–Tyr⁴–Ile⁵–His⁶–Pro⁷–Phe⁸) and a growth factor. Activation of AT₁ and AT₂ receptors by Ang II results in opposing actions of stimulation and inhibition. The AT₁-mediated stimulatory actions as well as signal transductions are well documented, whereas the AT₂-mediated inhibitory actions and signal transductions remain poorly understood (1–4). Most recently, studies have shown that the AT₂ receptor not only couples to pertusis toxin (PTX)-sensitive G_iα2 and G_iα3 proteins (5), but also activates a SH2 domain-containing tyrosine phosphatase (SHP-1) (5–10). Deletion or alanine substitutions of residues 240–244 within the intermediate portion of the cytoloop 3 has been shown to result in a complete loss of AT₂-mediated apoptosis, inhibition of extracellular signal-regulated kinases, and SHP-1 activation (10).

SHP-1, an intracellular protein tyrosine phosphatase (PTPase) containing two SH2 domains at the N-terminal end, plays a central role in blocking cell activation signals and prevent target cell lysis by the natural killer cells. The SH2 domains of SHP-1 interact with the catalytic domain to self-constrain the SHP-1 activity in an inactive state. Removal of this self-constraint imposed by the SH2 domains thus activates the SHP-1. The inhibitory receptors of the immune system such as killer cell inhibitory receptor, Fcγ receptor type IIb1, and paired Ig-like receptors of inhibitory types that initiate an inhibitory signal of SHP-1 activation bear an immunoreceptor tyrosine-based inhibitory motif (ITIM) in the cytoplasmic tail (11–13). Stimulation of the inhibitory receptors result in tyrosine phosphorylation of the ITIM and subsequently activates SHP-1 because the SH2 domain of SHP-1 selectively binds to the phosphoylated ITIM. The activated form of SHP-1 can then associate through its catalytic active site, C(X)₅R motif, with diverse target molecules such as JAK2 that bear phosphorylated tyrosine sequence signature. The ITIM has become the hallmark of a growing family of receptors with inhibitory potential, which are expressed in various cell types such as monocytes, macrophages, dendritic cells, leukocytes, and mast cells. The presence of an ITIM in a cytoplasmic tail is of sufficient predictive value that it can be used as an initial criterion for identification of candidate inhibitory receptors (14–17). However, sporadic reports have shown that GPCRs bearing no ITIM also can activate SH2 domain containing PTPases such as SHP-1 and SHP-2 (18–21).

The AT₂ receptor belongs to the superfamily of GPCR. Interestingly, It exceptionally contains an ITIM-like motif (VIY¹⁴⁸PFL) in the cytoloop domain. Substitution of this cytoloop with corresponding region of the AT₁ receptor abolishes AT₂-mediated SHP-1 activation. Although results from our laboratory and Dzau's laboratory (10) have shown that the cytoloop 3 also plays an important role in AT₂-mediated SHP-1 activation, the underlying mechanism is not clear. This ambiguity is mainly because no direct evidence is available to show that the cytoloop 3 is involved in AT₂/SHP-1 association. Thus, it is logical to hypothesize that an inhibitory receptor-like mechanism of activation could be used by the AT₂ receptor in activating SHP-1. We show here that, contrary to our initial hypothesis, mutations of Y148F and Y148A (the tyrosine of the ITIM-like motif) for the AT₂ receptor, however, fail to inhibit the SHP-1 association and activation. We also show that G_βγ-independent constitutive association of G_αs with SHP-1 and AT₂ is essential in AT₂-mediated activation of SHP-1. Moreover, the AT₂-mediated SHP-1 activation is ITIM-independent. These data contribute to our understanding of how GPCR activates SHP-1 with a mechanism distinct from ITIM-bearing inhibitory receptors.

Methods and Materials

Materials.

The monoclonal anti-c-Myc antibody (Roche Molecular Biochemicals, clone 9E10), monoclonal anti-hemagglutinin (HA) antibody (Roche Molecular Biochemicals, clone 12CA5), monoclonal anti-1D4 antibody (Cell Culture Center, Endotronics, Minneapolis), oligonucleotides (Sigma-Genosys and MWG Biotech, Ebersburg, Germany), analogues of [Sar¹]Ang II (GeneMed Synthesis, San Francisco), and G_α protein peptides (Biosynthesis, Lewisville, TX, and GeneMed Synthesis) were synthesized. Ang II and [Sar¹]Ang II (Bachem), PD123319 and CGP42112A (Research Biochemicals, Natick, MA), Protein Tyrosine Phosphatase Assay kit (Upstate Biotechnology), DTT, and other chemicals (Sigma) were purchased. ¹²⁵I-[Sar¹,Ile⁸]Ang II (2,200 Ci/mmol; 1 Ci = 37 GBq) was supplied by Robert Speth (Univ. of Washington, Pulman). Genes of G_αs, dominant negative G_αs, and constitutively active G_αs were purchased from American Type Culture Collection. The gene of SHP-1 was a gift from Matthew L. Thomas, Washington University (St. Louis), and genes of Gα_i, Gβ_1γ2, and β-adrenergic receptor kinase (βARK) were gifts from Robert Lefkowitz, Duke University (Durham, NC). Rat AT₂ receptor gene and synthetic rat AT₁ receptor gene were gifts from Sadashiva Karnik, Cleveland Clinic Foundation (Cleveland).

Plasmid Construction, Mutagenesis, and Expression of the AT₂ Receptors.

The rat AT₂ and AT₁ receptor genes in vector pcDNA3.1 were constructed based on the gift genes from Sadashiva Karnik (22, 23). These receptor genes were tagged with c-Myc (EQKLISEEDL) and HA (YPYDVPDYA) epitope at the N terminus after the start coden to generate c-Myc-AT₂, HA-AT₂, C-Myc-AT₂, and HA-AT₂ receptor genes. The original 1D4 epitope tag (TETSQVAPA) attached to the C terminus was removed from the receptor genes. The SHP-1 gene received as gift from Matthew L. Thomas, was subcloned into pcDNA3.1 vector for expression in mammalian cells. All these genes in pcDNA3.1 vector contain a consensus Marilyn–Kozak sequence at the N terminus to facilitate the expression in mammalian cells. Mutant AT₂ and chimera AT₂/AT₁ receptors were prepared based on HA-AT₂ receptor gene unless otherwise indicated. All plasmid construction and mutagenesis were performed with the restriction fragment-replacement method and the PCR method (22, 23). All genes, mutations, and replaced regions were sequenced by Cleveland Genomics to confirm the sequences.

To express the genes, 3–8 μg of column (Qiagen) purified plasmid DNA per 10⁷ cells was used in transfection. N1E-115, Chinese hamster ovary (CHO)-K1, and COS-7 cells (American Type Culture Collection), cultured in DMEM supplemented with 10% FBS and 1% MEM nonessential amino acids (CHO-K1 only), were transfected by the GenePORTER transfection reagents (Gene Therapy Systems, San Diego). Transfected cells cultured for 48 h were used for either ligand binding assay, SHP-1 activity assay, immunoprecipitation, or cell membrane preparation. The receptor expression was assessed in each case by immunoblot analysis and by ¹²⁵I-[Sar¹,Ile⁸]Ang II saturation binding analysis for receptors. Low-level expression was used for all signal activation related experiments. For ligand binding and other experiments, over-expression was applied. The expression levels were controlled and optimized by altering the amounts of plasmid DNA in transfections with GenePORTER transfection reagents. Efficiency of peptide transfection was examined by fluorescent microscopy (24).

Ligand Binding Study.

¹²⁵I-[Sar¹,Ile⁸]Ang II binding experiments were carried out under equilibrium conditions as described earlier (22, 23). For competition binding studies, membranes expressing the wild-type or mutant receptors were incubated for 1 h with 0.3 nM ¹²⁵I-[Sar¹,Ile⁸]Ang II and various concentrations of the ligands in assay buffer. The K_d value of ¹²⁵I-[Sar¹,Ile⁸]Ang II was determined by competition binding study in the presence of various concentrations (0.03 to 3 nM) of ¹²⁷I-[Sar¹,Ile⁸]Ang II. For the saturation binding study, a 10-fold higher increasing concentrations of ¹²⁵I-[Sar¹,Ile⁸]Ang II than the K_d value of the receptor was used to get >90% bound form of the receptor. Nonspecific binding (which was <5% of total binding in our experiments) of the radioligand to the membrane was measured in the presence of 50 μM ¹²⁷I-[Sar¹,Ile⁸]Ang II. All binding experiments were carried out at 22°C for 1 h in a 250-μl volume. After equilibrium was reached, the binding reactions were stopped by filtering the binding mixture under vacuum (Brandel Type M-24R) through FP-200 GF/C glass fiber filters (Whatman), which were extensively washed further with binding buffer to wash the free radioligand. The bound ligand fraction was determined from the cpm remaining on the membrane. Equilibrium binding kinetics were determined by using the computer program ligand. The K_d values represent the mean ± SEM of three to five independent determinations.

Immunoprecipitation and Western Blot.

At 24 h after cotransfection of AT₂ receptors and SHP-1, CHO-K1, or COS-7 cells were starved overnight and then treated with Ang II. The stimulation of cells was stopped by putting the cells on ice and rinsing briefly with ice-cold PBS. The cells were then solubilized for 30 min at 4°C in lysis buffer (50 mM Tris⋅HCl, pH 7.5/1.5% CHAPS/150 mM NaCl/5 mM EDTA), containing 0.5 mM sodium orthovanadate, 10 μg/ml of benzamidine, 10 μg/ml of bacitracine, 10 μg/ml of leupeptin, and 2 μg/ml of aprotinine, and 50 μg/ml PMSF. After centrifugation at 15,000 × g for 15 min, the resulting supernatant (1–1.5 mg of proteins) was incubated for 3 h at 4°C with either 10 μl of monoclonal anti-SHP-1 antibody prebound to Sepharose Protein A (Santa Cruz Biotechnology) or 10 μl of anti-HA antibody. The immunocomplexes were washed five times with washing buffer (identical to the lysis buffer). At the end of washing, the immunocomplexes were used for SDS/PAGE and Western blotting analysis. Proteins were visualized by using their specific antibodies and corresponding secondary antibodies coupled to horseradish peroxidase and enhanced chemiluminescence system (Amersham Pharmacia).

Immunoprecipitation and SHP-1 Activity Assay.

After transfection and treatment, cells were solubilized for 30 min at 4°C in lysis buffer containing 0.5 mM sodium orthovanadate and protease inhibitors. After centrifugation at 15,000 × g for 15 min, the resulting supernatant (1–1.5 mg of proteins) was incubated for 3 h at 4°C with 10 μl of monoclonal anti-SHP-1 antibody prebound to Sepharose Protein A. The immunocomplexes were washed five times with washing buffer (identical to the lysis buffer but without 1.5% CHAPS and 0.5 mM sodium orthovanadate). At the end of washing, the immunocomplexes were resuspended in PTPase buffer (50 mM Tris⋅HCl, pH 7.0/1 mg/ml BSA/5 mM DTT). The PTPase activity was assayed by measuring release of inorganic phosphate from phosphopeptides based on a malachite green detection system (11) using PTPase assay kit (Upstate Biotechnology) following the company's protocol.

Production of Total Inositol Phosphates (IP).

COS-7 Cells cultured in 60 mm Petri dishes, 24 h after transfection, were labeled for 24 h with [³H]myo-inositol (1 μCi/ml) at 37°C in inositol-free DMEM containing 10% bovine calf serum. On the day of IP assay (i.e., 48 h after transfection), the labeled cells were washed three times with serum-free medium and incubated with the media containing 10 mM LiCl for 20 min. Then medium alone or ligands were added to the cells. After incubation for 45 min at 37°C, the medium was removed, and total soluble IP was extracted from the cells by perchloric acid extraction method as described (23).

Statistical Analysis.

The results are expressed as the mean ± SEM of 2–5 independent determinations. The significance of measured values was evaluated with an unpaired Student's t test.

Results and Discussion

Inactive SHP-1 Constitutively and Physically Associates with AT₂ Receptor, Whereas Active SHP-1 Dissociates from AT₂ Receptor.

Consistent to the previously reported observation (8), stimulation of AT₂ receptor by 0.1 μM of Ang II in COS-7 cells cotransfected with HA-AT₂ and SHP-1 induced rapid SHP-1 activation as assayed in immunocomplexes prepared with anti-SHP-1 antibodies (Fig. 1A). The activation reached the maximum of 180% at about 5 min and returned to the basal level at 30 min. Similar activation profile was observed with N1E-115 cells expressing endogenous AT₂ and SHP-1. With respect to the basal phosphatase activity in the absence of Ang II, COS-7 cells cotransfected with genes of HA-AT₂ receptor and SHP-1 displayed no significant difference from COS-7 cells transfected with SHP-1 alone. However, mock-transfected or AT₂-transfected COS-7 cells show significantly less basal phosphatase activity (Fig. 1A). This suggests that the AT₂ receptor is not constitutively active in SHP-1 activation or the assay itself is not sensitive enough to demonstrate the constitutive activity of AT₂ receptor as reported in an AT₂-mediated apoptosis study (25).

Fig 1.

AT₂ receptor mediates SHP-1 activation. (A) N1E-115, and COS-7 cells were incubated with 0.1 μM Ang II for various periods of time before immunoprecipitation with polyclonal anti-SHP-1 antibodies. The phosphatase activities measured as described in the experimental section were expressed as percentage of basal activity of mock-transfected COS-7 cells. Results are means ± SEM of at least three independent experiments performed in duplicate. (B) Immunodetection of physical interaction between AT₂ and SHP-1. Cell lysates (500 μg of protein) or immunocomplexes using anti-SHP-1 and anti-HA antibodies were prepared from COS-7 cells cotransfected with HA–AT₂ and SHP-1. After separation on 10% SDS/PAGE, immunoblotting were performed with antibodies as indicated in the figure. (C) Quantitation of the relative extent of SHP-1/AT₂ association. The amount of SHP-1 immunoprecipitated by anti-HA antibodies and the amount of AT₂ immunoprecipitated by anti-SHP-1 antibodies, as assessed by densitometric scanning (arbitrary units) were normalized to the similarly quantitated amount of AT₂ and SHP-1, respectively, recovered in each immunoprecipitate. The mock background were substracted.

To determine whether SHP-1 physically associates with AT₂ receptor, we carried out differential immunoprecipiation and immunoblotting experiments. Genes of HA-AT₂ and SHP-1 were cotransfected in COS-7 cells and the immunocomplexes of the cell lysates precipitated with anti-HA or anti-SHP-1 antibodies were detected by immunoblot with anti-SHP-1 or anti-HA antibodies. Fig. 1B shows that SHP-1 constitutively and physically associates with AT₂ receptor in the absence of Ang II. On stimulation by 0.1 μM of Ang II, the association decreases. The decrease in association reached the maximum at about 5 min and returned to the basal level at 30 min. This pattern is very similar to the pattern of increase in SHP-1 activity as shown in Fig. 1A. It indicates that the activated SHP-1 dissociates from activated AT₂ receptors. To further confirm this conclusion, phosphatase activity was assayed in immunocomplexes precipitated with anti-HA rather than anti-SHP-1 antibodies. As expected, the phosphatase activity of immunocomplexes precipitated with anti-HA antibodies remains unchanged regardless of the time course.

In N1E-115 cells and transfected COS-7 cells, 0.1 μM of Ang II induces maximal SHP-1 activity. No significant difference in SHP-1 activity was seen between 0.1 μM and 50 μM of Ang II concentrations.

The maximal decrease in physical association (about 3-fold) is much greater than the maximal increase in SHP-1 activity (180%) at 5 min after Ang II stimulation. This difference possibly reflects the relatively low sensitivity of phosphatase assay.

AT₂ Receptor Is Not Constitutively Active in SHP-1 Activation.

Three lines of evidence suggest that wild type AT₂ receptor is constitutively active. First, in an AT₂/AT₁ chimera receptor with the cytoloop 3 domain of AT₂ replaced by the corresponding region of AT₁ receptor, Ca²⁺ movement was observed in the absence of Ang II stimulation (26). Second, Ang II binding profile has shown that in agreement with constitutively active AT₁ mutant N111G (27–29), AT₂ is insensitive to modifications of Ang II side chains at any position (22, 30), suggesting that AT₂ receptor is already in a relaxed active state (R*). Third, Karnik's group has shown that AT₂ induces apoptosis in the absence of Ang II (25). However, stimulation by Ang II failed to facilitate the AT₂-mediated constitutive apoptosis (25).

[Sar¹,Ile⁴,Ile⁸]Ang II that is completely incapable of activating wild-type AT₁ receptor has been shown to fully activate constitutively active AT₁ receptor mutant N111G (27–29). [Sar¹,Ile⁴,Ile⁸]Ang II binds wild type AT₂ receptor with >100-fold greater affinity with comparison to the wild-type AT₁ receptor. However, 5 or 50 μM of [Sar¹,Ile⁴,Ile⁸]Ang II fails to activate AT₂-mediated SHP-1 activity (Fig. 2A). Consistently, this inactive Ang II analog also fails to induce any significant change in AT₂/SHP-1 association (data not shown). These observations are consistent to the findings regarding basal phosphotase activities as described above.

Fig 2.

(A) Ang II analogs induce SHP-1 activity. COS-7 cells transfected with HA-AT₂ and SHP-1, CHO-K1 cells transfected with HA-AT₂, and N1E-115 cells were incubated with Ang II (0.1 μM) and [Sar¹]Ang II (0.1 μM), and [Sar¹,Ile⁴,Ile⁸]Ang II (50 μM) for 5 min before immunoprecipitation with anti-SHP-1 antibodies. (B) Ang II (0.1 μM) induces SHP-1 activity in N1E-115 cells transfected with or without G_β1γ2 and βARK. Each data point represents the mean ± SE of two duplicate determinations. *, P < 0.01 with comparison to G_β1γ2 and βARK.

Taken together, these findings indicate that the wild type AT₂ receptor is not constitutively active in activation of SHP-1. However, these findings do not exclude the possibility that AT₂ receptor may be constitutively active in activation of other signal pathways as exemplified by Perez et al. (31) with α1-adrenergic receptor.

AT₂-Mediated SHP-1 Activation Is Independent of G Protein Activation.

It is known that AT₂ receptor couples to the G_αi proteins only (5). However, the AT₂-mediated SHP-1 activation is PTX-insensitive (8). These contradictory findings do not support an involvement of G_αi proteins in AT₂-mediated SHP-1 activation. To examine whether AT₂-mediated SHP-1 activation could occur independently of activation of any Gα protein, we need an AT₂ mutant receptor that does not activate any Gα protein but is still capable of activating SHP-1. Initially, we constructed a mutant AT₂ receptor (D141A–R142L) in which the highly conserved Asp-141–Arg-142 residues were substituted with Ala¹⁴¹Leu¹⁴². In structurally related GPCRs, this sequence is the switch for release of GDP from the receptor-coupled G_α protein. Mutations of the sequence destabilize receptor–G protein complex and prevent the mutant receptor from inducing GDP–GTP exchange (32). The D141A–R142L mutant receptor expressed in COS-7 cells binds Ang II and AT₂ specific antagonist PD123319 with similar affinities to wild-type AT₂ receptor. Interestingly, this mutant receptor fully activates SHP-1 (Fig. 3A) and physically associates with SHP-1 (Fig. 3B). Consistently, the AT₂-mediated SHP-1 activation was PTX-insensitive. These findings indicate that AT₂-mediated SHP-1 activation can occur in the absence of Gα protein activation.

Fig 3.

AT₂ receptor mutants mediate SHP-1 activation. (A) COS-7 cells cotransfected with SHP-1 and HA-tagged AT₂ wild type (WT), or D141A–R142L or D90A mutants were incubated with Ang II (0.1 μM) for 5 min before immunoprecipitation with anti-SHP-1 antibodies. The SHP-1 activities were measured in duplicate. *, P < 0.01 between wild type and D141A–R142L in the presence of Ang II. (B) Immunodetection of receptor-SHP-1 association in the presence or absence of 0.1 μM of Ang II. Each data point represents the mean ± SE of at least two duplicate determinations. (C) Quantitation of the relative extent of SHP-1/AT₂ association. The amount of SHP-1 immunoprecipitated by anti-HA antibodies as assessed by densitometric scanning (arbitrary units) were normalized to the similarly quantitated amount of AT₂ receptors recovered in each immunoprecipitate. The mock background were substracted.

Significantly less SHP-1 activity of D141A–R142L mutant may reflect reduced binding affinity between mutant receptor and SHP-1 (Fig. 3A). This speculation is consistent to the immunoblot result as shown in Fig. 3B.

Angonist-induced conformational change of a receptor from inactive state (R) to active state (R*) is an absolute prerequisite for homologous activation of an effector pathway. AT₂ activation of SHP-1 must involve similar conformational change of the receptor triggered by the binding of agonist Ang II. Interruption of the conformational change should paralyze the AT₂ receptor in SHP-1 activation but probably not in SHP-1 association. To determine whether AT₂-mediated SHP-1 activation requires an active conformation (R*) of the AT₂ receptor, we constructed a mutant AT₂ receptor (D90A) in which the highly conserved Asp⁹⁰ residue in the TM3 domain was replaced with Ala⁹⁰. In structurally highly conserved GPCRs, mutation of the Asp residue has been shown in many receptors, including AT₁ receptor, to disable the receptors in signal activation (2, 33). As expected, the D90A mutant receptor recognizes Ang II and AT₂-specific antagonist PD123319 with similar affinities of the wild type. Interestingly, the D90A mutant associates with SHP-1 like the wild type but fails to activate SHP-1 (Fig. 3).

These findings indicate that activation of SHP-1 by AT₂ receptor requires conformational transition of the receptor from inactive state to active state, in a way similar to Gα protein activation but distinct from classical SHP-1 activation by immunoreceptors (11–17). In these non-GPCR receptors, phosphorylation of a tyrosine located in the ITIM is required for SHP-1 binding and activation (11–17).

Synergism of Intracellular Domains Rather than the ITIM-Like Motif (VIYPFL) of the AT₂ Receptor Mediates SHP-1 Activation.

An ITIM-like motif (VIYPFL) located in the cytoloop 2 of AT₂ receptor was identified originally by sequence analysis using PHI-blast program (34). However, substitutions of the tyrosine residue at this motif (Y148F and Y148A) with Phe and Ala failed to block SHP-1 association and activation in the presence of Ang II (data not shown). This result should be conclusive because Tyr¹⁴⁸ is the only tyrosine residue available in the cytoloop 2 of the AT₂ receptor (Tyr¹⁴³ is not included because Tyr¹⁴³ is part of the DRY motif that is highly conserved in most GPCRs, including AT₁ receptor). In addition, AT₂ receptor activation by Ang II resulted in dissociation from SHP-1, rather than association with the SHP-1. This dissociation is accompanied with increase in SHP-1 activity (Fig. 1). This finding does not support any significant role of this ITIM-like motif in AT₂-medaited SHP-1 activation.

To identify structural determinants of AT₂ receptor for SHP-1 association and activation, we took an AT₂/AT₁ chimera receptor approach. Because intracellular domains of GPCR often serve as the interface for coupling to signaling molecules, four chimera receptors that contain only minimal exchange of intracellular segments of AT₁ and AT₂ receptor in various combinations were used (Table 1). These receptors bind Ang II with high affinity (0.46–1.56 nM). They also bind subtype-specific antagonist losartan or PD123319 with affinity similar to their structurally dominant wild types, respectively. Among these chimera receptors, CR14, which contains the cytoloop 3 of AT₁ receptor, showed no SHP-1 association and activation (Table 1). This result indicates that the cytoloop 3 of AT₂ receptor is involved in AT₂-mediated SHP-1 activation. This finding is in agreement with Lehtonen's report (10) in which the key residues of Lys²⁴⁰, Asp²⁴², and Ser²⁴³ located in the third cytoloop of AT₂ receptor have been found important for AT₂-induced apoptosis, extracellular signal-regulated kinase inhibition and SHP-1 activation.

Table 1.

Functions of Ang II receptors

SHP-1 activity and coupling to the receptor were determined at 5 min of incubation with 0.1 μM Ang II. The maximal SHP-1 activity of wild-type AT₂ was referred as 100%. The maximal inositol phosphate (IP) production of wild type AT₁ in the presence of 0.1 μM Ang II was referred as 100%. CR13, the C-terminal region (309–363) of AT₂ replaced by the corresponding region (293–359) of AT₁; CR14, the cytoloop 3 (230–254) of AT2 replaced by the corresponding region (214–239) of AT₁; CR25, a combination of CR13 and CR14 replacements; CR26, the cytoloop 2 (134–162) of AT₂ replaced by the corresponding region (118–146) of AT₁. ND, not detectable.

However, CR26 that has the cytoloop 2 replaced also show no SHP-1 association and activation, suggesting an important role for the second cytoloop. Interestingly, CR25 that contains both cytoloop 3 and C terminus of AT₁ receptor associates with SHP-1 and activates SHP-1 (Table 1). CR13 with replacement of the C-terminal alone preserved the capability of wild-type AT2 receptor to associate with and activates SHP-1 (Table 1).

Replacement of the second or third cytoloop alone results in loss of SHP-1 activation, whereas replacement of C-terminal in CR25 rescued the CR14 in SHP-1 activation. These findings suggest that all three intracellular domains are important in AT₂-mediated SHP-1 activation. It is their collective actions that dictate the SHP-1 activation. This is reminiscent to what we learned from receptor–G protein interactions. Interestingly, somatostatin receptor sst2-mediated SHP-1 activation also requires participation of G_αi protein (35).

Cotransfection of G_βγ Inhibits AT₂-Mediated SHP-1 Activation, Whereas Cotransfection of βARK Potentiates the Function.

Because AT₂-mediated SHP-1 activation involves multiple intracellular domains, it becomes logical to hypothesize that ancillary molecules are required for AT₂-SHP-1 interaction. To test whether G_βγ proteins contribute to the interactions, N1E-115 cells were cotransfected with G_β1 and G_γ2 proteins. It has been shown that overexpression of G_b1g2 proteins in rabbit tubule epithelial cells augments Ang II-induced arachidonic acid release, whereas overexpression of the C terminus of βARK abrogates Ang II-induced arachidonic acid release under conditions with and without overexpression of Gβ_1γ2 proteins (36). Surprisingly, overexpression of G_β1γ2 proteins inhibited AT₂-mediated SHP-1 activation by 92%, whereas overexpression of βARK potentiated the action by 95% (Fig. 2B). This result supports a role of G_α rather than G_β1γ2 in AT₂-mediated SHP-1 activation.

G_αs Constitutively and Physically Associates with SHP-1 and AT₂ and Is Essential in AT₂-Mediated SHP-1 Activation.

It is well established that both α and βγ subunits of signal-transducing G proteins are maintained in inactive states by their mutual association in a heterotrimeric complex, which is coupled to 7TM GPCRs. As a result, neither GDP-bound α nor βγ can interact with or activate downstream effectors. Conformational change of the receptor induced by specific agonist binding can result in conformational changes in GDP-bound α to favor GTP binding. The GTP-bound α dissociates from the receptor and releases βγ. These two activated signal generating molecules, the GTP-bound α and the free βγ, will activate their own effectors, initiating signal transductions (37). However, whether α protein or βγ dimer alone can serve as an independent adapter protein or ancillary protein for other signal pathways is unknown. The fact that AT₂-mediated SHP-1 activation can be inhibited by overexpression of G_β1γ2 and potentiated by overexpression of βARK strongly suggests that a G_α protein is involved.

To test this possibility, the immunocomplexes precipitated with anti-HA antibody in cell lysates of COS-7 cells expressing HA-AT₂ and SHP-1 were immunoblotted with various anti-G_α antibodies. We identified a band with anti-G_αs antibodies (Fig. 4). The AT₂-G_αs association decreases in the presence of Ang II. Similar bands were also demonstrated from immunocomplexes precipitated with anti-SHP-1 antibodies in cell lysates of COS-7 cells expressing HA-AT₂ and SHP-1. However, the G_αs-SHP-1 association shows no significant decrease in the presence of Ang II. This observation suggests that G_αs protein remains associated with activated SHP-1 molecule and the sustained association may serve as a mechanism to maintain the activated SHP-1 in its active conformation until it binds to its target molecules of dephosphorylation.

Fig 4.

Role of G_α proteins in AT₂-mediated SHP-1 activity. (A) Immunodetection of physical association of Gas with AT₂ or SHP-1 in COS-7 cells cotransfected with HA-AT₂, SHP-1 and G_αs genes. (B) Immunodetection of AT₂-SHP-1 and G_αs-SHP-1 associations in the presence or absence of G_αs peptides. (C) SHP-1 activities induced by 0.1 μM Ang II in N1E-115 cells in the presence of G_αs peptides and G_αs mutants. G_αs peptides are: GiP, IKNNLKDCGLF; rGiP, NGIKCLFNDKL (Randomized peptide of Gi); GqP, LQLNLKEYNAV; rGqP: LAYQVNKLNLE (Randomized peptide of Gq); GsP, QRMHLRQYELL; rGsP: QQRLEMYLRHL (Randomized peptide of Gs); G13P, LHDNLKQLMLQ; rG13P, MQLDNKLQLHL (Randomized peptide of G13). Data for rGiP, rGqP, and rG13P peptides are not included in C. Here caG_αs and dnG_αs represent constitutively active G_αs mutant R201C and dominant negative mutant A366S-G226A-E268A, respectively. Each data point represents the mean ± SE of at least two duplicate determinations. *, P < 0.01 compared with mock-transfected N1E-115 cells; ¶, P < 0.01 compared with mock-transfected N1E-115 cells.

To test whether specific blockade of AT₂–G_αs interaction would inhibit AT₂-mediated SHP-1 activation, 0.5 μM of peptides (Fig. 4) derived from the last 11 residues of G_αq, G_αs, G₁₃, and G_αi proteins are introduced into N1E-115 cells by transfection with Lipofectamine reagent (Life Science). These peptides are able to block G protein coupling to GPCR (37). Fig. 4 shows that only Gs peptide (GsP) inhibits the AT₂-mediated SHP-1 activation by about 88%. The rGsP, the control peptide of scrambled sequence of Gs peptide and other G_α peptides failed to inhibit the SHP-1 activation. In COS-7 cells cotransfected with HA-AT₂ and SHP-1, Gs peptide also blocked AT₂-SHP-1 association.

Because AT₂-mediated SHP-1 activation requires only conformational change of AT₂ receptor to its active state but not G protein activation as described earlier with D141A–R142L and D90A mutants, we speculated that constitutively active G_αs would not constitutively activate SHP-1, whereas dominant negative G_αs would facilitate the SHP-1 activation. As expected, overexpression of constitutively active G_αs (R201C mutant) failed to show any effect on SHP-1 activation, whereas overexpression of dominant negative G_αs (A366S-G226A-E268A mutant) augmented SHP-1 activation up to 78%. Constitutively active G_αs is monomer in GTP-bound form and no longer binds to the receptor. However, the dominant negative G_αs forms heterotrimer with G_βγ subunits with reduced binding affinity for GDP and GTP. It consumes endogenous G_βγ to facilitate AT₂–SHP-1 interaction. The 78% increase in SHP-1 activation reflects a mixed effect of sequestration of G_βγ and modified binding to the receptor (38).

SHP-1 Activation by G_αs-Coupled AT₂ Receptor: ITIM vs. Non-ITIM.

SHP-1 binding to tyrosine-phosphorylated ITIM (I/VxYxxL) of inhibitory receptors represents a general paradigm for SHP-1 activation (11–17). This finding is further supported by evidence that has identified sequence YxxL as a previously uncharacterized ITIM in the death domain of death receptor of the tumor necrosis factor and nerve growth factor superfamily (39). The resultant free catalytic domain becomes active and ready to interact with its substrate. However, similar ITIMs are not present in the GPCRs that activate SHP-1 (18–21, 35). Moreover, no similar ITIMs are present in any of G protein subunits as indicated by our sequence analysis. The ITIM-like motif identified in the AT₂ receptor is not required for the SHP-1 activation. Thus, it is conceivable that GPCR-mediated SHP-1 activation is ITIM-independent and certainly distinct from the general paradigm of ITIM-mediated SHP-1 activation.

The fact that G_αs involves in AT₂-mediated SHP-1 activation has illustrated a previously undescribed mode of SHP-1 activation by GPCRs. Besides physical interaction with AT₂ receptor, SHP-1 may interact with G_αs protein at the interface to which G_βγ proteins bind. Alternatively, the SHP-1 binding site may overlap with the G_βγ binding site in G _αs protein. AT₂, G _αs, and SHP-1 are capable of forming complexes at basal state.

SHP-1 (11, 14, 15) consists of (i) two SH2 domains, designated SH2N and SH2C, (ii) a catalytic domain, (iii) a 55-residue fragment connecting the SH2C and CA, and (iv) an 80-residue C-terminal tail (Fig. 5). It is well known that SH2 domains are functional protein motifs that bind tyrosine-phosphorylated targets by recognizing phosphotyrosine and specific adjacent residues. It is thought that the free cytosolic SHP-1 maintains a self-constrained inactive state by interaction of the catalytic domain with the two SH2 domains. Preferential interaction of the SH2 domain (mainly SH2C) with a tyrosine phosphorylated ITIM automatically removes the constraints imposed on the catalytic domain, leading to SHP-1 activation (14, 40, 41). This activation mechanism is based on evidence that removal of the SH2 domains from SHP-2, belonging to SHP-1 PTPase family, results in a 12- to 45-fold increase in phosphatase activity and exogenous SH2 domains reverse this effect (14, 40, 41). Because the functions of SH2N, the 55 fragment and C tail of SHP-1 are not yet defined, it is possible that binding of SHP-1 to AT₂ and G_αs through these functional domains may also lead to SHP-1 activation.

Fig 5.

Functional domains of SHP-1 and the ITIMs.

Abbreviations

• GPCR, G protein-coupled receptor

• TM, transmembrane helix

• SHP-1, SH2 domain containing phosphatase 1

• ITIM, immunoreceptor tyrosine-based inhibitory motif

• βARK, β-adrenergic receptor kinase

• PTX, pertusis toxin

• GDT, guanosine diphosphate

• HA, hemagglutinin