Agonist-specific Transactivation of Phosphoinositide 3-Kinase Signaling Pathway Mediated by the Dopamine D 2 Receptor*

Bromocriptine, acting through the dopamine D 2 re- ceptor, provides robust protection against apoptosis induced by oxidative stress in PC12-D 2 R and immortalized nigral dopamine cells. We now report the characteriza-tion of the D 2 receptor signaling pathways mediating the cytoprotection. Bromocriptine caused protein kinase B (Akt) activation in PC12-D 2 R cells and the inhi- bition of either phosphoinositide (PI) 3-kinase, epidermal growth factor receptor (EGFR), or c-Src eliminated the Akt activation and the cytoprotective effects of bromocriptine against oxidative stress. Co-immunoprecipi-tation studies showed that the D 2 receptor forms a com- plex with the EGFR and c-Src that was augmented by bromocriptine, suggesting a cross-talk between these proteins in mediating the activation of Akt. EGFR repression by inhibitor or by RNA interference eliminated the activation of Akt by bromocriptine. D 2 receptor stim- ulation by bromocriptine induced c-Src tyrosine 418 phosphorylation and EGFR phosphorylation specifically at tyrosine 845, a known substrate of Src kinase. Furthermore, Src tyrosine kinase inhibitor or dominant negative Src interfered with Akt translocation and phosphorylation. Thus, the predominant signaling cascade mediating cytoprotection by the D 2 receptor in- volves c-Src/EGFR transactivation by D 2 receptor,

The hallmark of Parkinson's disease (PD) 1 is the progressive loss of dopaminergic neurons in the substantia nigra pars com-pacta (1), causing a profound reduction in dopamine-mediated signaling (2). The prominent locomotor deficits that occur in this disease are in large part attributable to the loss of stimulation of the dopamine D 2 receptor (3), a member of the rhodopsin-like heptahelical receptor family (4). The D 2 receptor is an important target for anti-parkinsonian drugs that ameliorate the motor deficits associated with this disorder. In recent years, dopamine agonists have also been found to have neuroprotective activity in some experimental models, and the possibility that they may decrease the progression of PD has been proposed (5). However, the mechanisms underlying the agonist-mediated neuroprotection reported in experimental models are poorly understood, and the potential for dopamine agonists to alter the clinical course of this disease remains an area of controversy (6).
Many heptahelical receptors couple to multiple signal transduction pathways, including various heterotrimeric G-proteinsecond messenger pathways and growth factor receptor-protein kinase cascades (7). The signal for activation of the proximal mediators of signaling such as heterotrimeric G-proteins, receptor kinases, or other protein partners, is an alteration in the receptor's conformation that occurs following complexing with agonist. Studies in several heptahelical receptors suggest that these proteins exist in multiple, functionally significant conformations that may differ in their relative activation of different signaling pathways (8 -12). Studies with several receptors, including the dopamine D 2 receptor, suggest that agonists acting at the same receptor select among different active receptor conformations and determine the relative levels of activation of downstream signaling pathways, a hypothesis called agonistdirected signal trafficking (13)(14)(15)(16).
We had previously investigated the role of the D 2 receptor expressed in the PC12 cell line (PC12-D 2 R) in modulating the induction of apoptotic cell loss caused by hydrogen peroxideinduced (H 2 O 2 ) oxidative stress (17). Although the mechanism of neuronal loss in PD is not known, many studies have implicated oxidative stress (reviewed in Refs. 18 and 19). Oxidation of dopamine by auto-oxidation and monoamine oxidase produces reactive oxygen species, including H 2 O 2 . H 2 O 2 reacts with ferrous (Fe 2 ϩ ) iron to produce hydroxyl radicals, which can damage proteins, nucleic acids, and membrane phospholipids, and induce apoptosis (20). Some animal model and human PD postmortem studies provide evidence that the degeneration of DA neurons occurs via apoptosis (21,22). We found that activation of the D 2 receptor in the PC12-D 2 R line caused a robust, concentration-dependent increase in cell survival during oxidative stress that required activation of phosphoinositide 3-kinase (PI 3-kinase). Among the agonists studied, we found significant discrepancies in the capacity of individual agonists to mediate anti-apoptosis and to stimulate G-protein activation, assayed via [ 35 S]GTP␥S binding (17).
To elucidate the mechanisms underlying agonist-specific modulation of cell survival, we have now investigated the antiapoptotic signaling pathway activated by the D 2 receptor. We find that D 2 receptor-mediated protection against oxidative stress involves a novel c-Src-dependent transactivation of the epidermal growth factor receptor (EGFR) that activates PI 3-kinase/protein kinase B (Akt) and that agonists differ in their capacity to activate this pathway.
Cell Culture and Viability Analysis-The development and characterization of the PC12-D 2 R cell line, which is stably transfected with the human D 2L receptors, were previously described (17). The cells were maintained in DMEM supplemented with 10% horse serum, 5% fetal bovine serum, and 500 g/ml G418 in a humidified atmosphere containing 5% CO 2 at 37°C. For differentiation, PC12-D 2 R cells were plated onto collagen-coated plates in DMEM containing 10% horse serum and 5% fetal bovine serum and allowed to attach overnight. The cells were then induced to differentiate by growing in DMEM supplemented with 1% fetal bovine serum and 100 ng/ml NGF for 10 -14 days. Nigral dopamine cell line SN4741 (generous gift from Dr. J. H. Son, Columbia University, New York, NY) was cultured as described previously (23). Cell viability was measured by the MTT method 24 h after various treatments as described (17).
Transfections and DNA Constructs-For live cell fluorescence microscopy, PC12-D 2 R cells (1 ϫ 10 5 ) were plated into 60-mm culture dishes and incubated in the media for 24 h. The media was replaced with serum-free DMEM, and a mixture containing 5 g of the plasmid DNA encoding the pleckstrin homology domain of Akt protein kinase (1-167) tagged with green fluorescent protein (PH-Akt-GFP) (24) (kindly provided by Dr. T. Balla, National Institutes of Health, Bethesda, MD) and 30 l of LipofectAMINE reagent were gently added to each plate and incubated for 3 h at 37°C at 5% CO 2 . The DNA-containing medium was replaced with fresh DMEM-containing serum. When cotransfection of c-Src (wild type) or dominant negative c-Src (K295R/ Y527F) (both were generous gifts from Dr. J. Burgge, Harvard Medical School, Boston, MA) with PH-Akt-GFP was carried out, the DNA concentration used was 1:1. Green fluorescent protein (GFP) plasmid was from Clontech. SN4741 cells were co-transfected with D 2L and PH-Akt-GFP or GFP by calcium phosphate method (25).
Epifluorescence Imaging-Microscopy of live cells transfected to express PH-Akt-GFP was performed on the Olympus (BX65) upright fluorescent microscope using a water immersion objective lens (ϫ40) fitted with a heated stage and an objective lens heater. Images were collected at 2-min intervals after the addition of the drugs and processed using Adobe Photoshop (5.5).
Immunoblotting and Immunoprecipitation-PC12-D 2 R cells (1 ϫ 10 6 cells/100-mm plate) were grown for 24 h, and following respective treatments, the cells were washed twice with ice-cold phosphate-buffered saline and lysed in buffer containing 20 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1% Igepal C630, 1 mM phenylmethylsulfonyl fluoride, 1 mM sodium orthovanadate, 5 g/ml aprotinin, and mixture of protease inhibitors (Roche Applied Science, GmbH) at 4°C for 20 min. After centrifugation at 14,000 ϫ g for 20 min at 4°C, equal amounts of proteins were resolved by SDS-polyacrylamide gel electrophoresis. The resolved proteins were electrotransferred to nitrocellulose membranes. Detection of proteins by immunoblotting was conducted using ECL system according to the manufacturer's recommendations. The blots were then stripped in buffer containing 62.5 mM Tris-HCl (pH 6.8), 2% SDS, and 100 mM ␤-mercaptoethanol for 30 min at 50°C and re-probed with respective antibodies.
For immunoprecipitation, the protein extract was incubated sequentially (2 h for each incubation at 4°C) with anti-D 2 receptor antibody and protein A/G-agarose with gentle agitation. Immunoprecipitates were washed three times with lysis buffer, boiled for 5 min in 3ϫ Laemmli sample buffer, and processed for Western blotting using EGFR or c-Src antibody. The EGFR blots were stripped and reprobed with anti-phospho tyrosine (PY20) antibody.
RNA Interference-Custom SMARTpool plus small interfering RNA (siRNA) to target rat EGFR (catalog no. M-004710-00) was designed and synthesized by Dharmacon (Lafayette, CO). siRNA (50 pmol) was co-transfected with PH-Akt-GFP (2 g) into PC12-D 2 R cells using transit-TKO and -neural transfection reagents (Mirus, Madison, WI) according to manufacturer's protocol. For immunofluorescence, 24 h after transfection, cells were serum-starved for 1 h and were treated with bromocriptine for 10 min. The cells were fixed with ice-cold methanol and immunostained for EGFR and visualized using CY3-conjugated secondary antibody. A nonspecific RNA duplex (Dharmacon, catalog no. D-001206-09-05) was used in control experiments.

Neuroprotection by D 2 Receptor Activation Involves PI 3-Kinase/Akt
Signaling Cascade-We have previously reported that the increased cell survival in PC12-D 2 R cells mediated by D 2 receptor activation was completely abolished by inhibitors of PI 3-kinase, suggesting that the D 2 receptor may be altering cell survival by activating PI 3-kinase (17). We therefore studied whether PI 3-kinase/Akt signaling was modulated by the D 2 receptor when complexed with an agonist that prevents apoptosis in these cells. Activation of PI 3-kinase generates phosphatidylinositol 3,4,5-triphosphate (PIP 3 ) and thereby stimulates anti-apoptotic proteins (26). The downstream PI 3-kinase target, protein kinase B (Akt), has been reported to be important in mediating survival in many cell types (27). Akt is activated by phosphorylation at Thr 308 in the catalytic loop and Ser 473 in the C-terminal domain (28,29).
We first determined whether the anti-apoptotic dopamine agonist bromocriptine induced phosphorylation of Ser 473 of endogenous Akt in PC12-D 2 R cells. As shown in Fig. 1A, Akt phosphorylation was increased 15 min after exposure to bromocriptine. In some cell lines, H 2 O 2 has been reported to activate Akt (30,31). However, we found that in PC12-D 2 R cells H 2 O 2 alone had no effect on the phosphorylation of Akt (Fig.  1A). Akt phosphorylation occurs after it is recruited to the plasma membrane through an interaction of its N-terminal pleckstrin homology (PH) domain with PIP 3 (32), thereby bringing the enzyme into the proximity of additional PIP 3 -dependent and -independent protein kinases (33). We studied the redistribution of Akt by D 2 receptor signaling using a PH-Akt-GFP fusion protein (24). The localization of PH-Akt-GFP in quiescent PC12-D 2 R cells was indistinguishable from that of transfected GFP alone. Receptor activation by bromocriptine, however, caused a rapid (Ͻ5 min) translocation of the PH-Akt-GFP to ruffled membrane regions (Fig. 1B, top panels). No response to bromocriptine was observed in control PC12-D 2 R cells expressing GFP alone (Fig. 1B, middle panels) or in the parent PC12 cells, which lack the D 2 receptor, expressing PH-Akt-GFP (Fig. 1B, bottom panels). We also tested whether this pathway was active in cells exposed to oxidative stress. As shown in Fig. 1C, the bromocriptine-induced phosphorylation of Akt and translocation of PH-Akt-GFP (Fig. 1D) were unaffected in the presence of H 2 O 2 . Thus, D 2 receptor stimulation by bromocriptine caused translocation and phosphorylation of Akt in PC12-D 2 R cells during oxidative stress. D 2 Receptor Activation of Akt Is PTX-insensitive and PI 3-Kinase-dependent-The D 2 receptor is a member of the rhodop-sin-like heptahelical receptor family, whose classic signaling pathway involves the activation of the G i /G o subtype heterotrimeric G-proteins, which can be inactivated by PTX (34). To examine the role of G i /G o coupling in activation of PI 3-kinase/ Akt, we determined the effects of PTX on these responses. Pretreatment of PC12-D 2 R cells with 100 ng/ml PTX (16 h) eliminated bromocriptine-stimulated [ 35 S]GTP␥S binding (data not shown). In contrast, the D 2 receptor-mediated phosphorylation of Akt was unaffected by PTX ( Fig. 2A, compare lanes 3 and 4). To assess the role of PI 3-kinase in the activation of Akt induced by D 2 receptors, PC12-D 2 R cells were pretreated with the inhibitor wortmannin at 100 nM, a concentration that selectively blocks PI 3-kinase (35). The cultures were then exposed to bromocriptine. Wortmannin completely prevented the phosphorylation of Akt that is inducible by D 2 receptor activation ( Fig. 2A, compare lanes 4 and 5). Similar results were also obtained with LY294002 (20 M), another commonly used but less potent inhibitor of PI 3-kinase (35) (Fig. 2A, compare lanes 4 and 6). Translocation of PH-Akt-GFP by bromocriptine was also inhibited by pretreatment with the PI 3-kinase inhibitor wortmannin, whereas it was unaffected by PTX pretreatment (Fig. 2B). Similar results were obtained with LY294002 (data not shown). These results indicate that the D 2 R-mediated activation of Akt occurs through PI 3-kinase by a mechanism independent of G i /G o class heterotrimeric G-proteins. The arrows indicate localized areas of PH-Akt-GFP translocation following addition of bromocriptine. Bromocriptine (100 nM) had no effect on the translocation of or EGFP expressed in PC12-D 2 R cells (middle panels) or PH-Akt-GFP expressed in native PC12 cells, which lack D 2 receptors (bottom panels). Panels shown are from one of eight independent experiments. C, oxidative stress had no effect on the phosphorylation or translocation of Akt in the presence of bromocriptine. Western blot of phospho-Akt. to PI 3-kinase/Akt observed in PC12-D 2 R cells was present when the D 2 receptor was expressed in a different cellular context, we studied this signaling pathway in the mouse immortalized nigral dopamine cell line SN4741, which expresses tyrosine hydroxylase, the dopamine transporter, and D 2 autoreceptors (23). Activation of D 2 receptors by bromocriptine in these cells was found to induce phosphorylation of endogenous Akt (Fig. 3A). When the cells were transfected with the PH-Akt-GFP construct, bromocriptine induced redistribution of this reporter (data not shown). The capacity of the activated D 2 receptor to induce Akt redistribution in this model was enhanced in cells co-transfected with D 2 receptor and PH-Akt-GFP. Bromocriptine caused a translocation of the PH-Akt-GFP protein into discrete regions of the SN4741 cells (Fig. 3B) that was similar to the response observed in differentiated PC12-D 2 R cells (Fig. 3C). The bromocriptine-stimulated translocation of PH-Akt-GFP in SN4741 cells was eliminated by pretreatment with the PI 3-kinase inhibitor LY290042 (data not shown). Control experiments in which cells were transfected with the D 2 receptor and GFP showed no change in the distribution of fluorescence in response to bromocriptine. These results suggest that the D 2 receptor can couple to the PI 3-kinase/ Akt signaling pathway in dopaminergic neurons. D 2 Receptor-mediated Neuroprotection and Activation of PI3-Kinase/Akt Involves EGFR Transactivation-We previously reported that bromocriptine showed significant PI 3-kinase-dependent anti-apoptotic activity in PC12-D 2 R cells and have demonstrated, as described above, that bromocriptine also induced Akt phosphorylation and translocation. We next attempted to delineate the signal mediators connecting the D 2 receptor to PI 3-kinase. It has been reported that the PI 3-kinase/Akt pathway in PC12 cells can be activated by receptor tyrosine kinases (36). The effectiveness of several receptor tyrosine kinase inhibitors on bromocriptine-mediated neuroprotection was evaluated. H 2 O 2 exposure caused significant loss of PC12-D 2 R cell viability at 24 h, as determined using the MTT metabolism assay, and this cell loss was nearly completely reversed by the D 2 receptor agonist bromocriptine (Fig. 4A), consistent with our previous results (17). The effects of various growth factor receptor inhibitors on the capacity of bromocriptine to protect cells against cell death due to H 2 O 2 exposure were studied. AG1296 (200 nM), AG1478 (200 nM), and k252a (50 nM) in the presence of H 2 O 2 and the presence or absence of bromocriptine (100 nM) for 24 h were evaluated. As shown in Fig. 4A, AG1478, a specific inhibitor of EGFR intrinsic tyrosine kinase activity (37), completely abolished the neuroprotection provided by bromocriptine exposure, an effect similar to that observed with inhibition of PI 3-kinase (17). In contrast, inhibition of platelet-derived growth factor (PDGF) receptors by AG1296 or NGF receptor by k252a had no effect on D 2 receptormediated cell survival.
We next investigated the role of the EGFR in mediating the signaling from the D 2 receptor to Akt. Activation of the EGFR by EGF caused a rapid phosphorylation of Akt and a translocation of PH-Akt-GFP in PC12-D 2 R cells, similar to the response observed with bromocriptine (Fig. 4B). The involvement of EGFR transactivation in D 2 receptor stimulation of Akt phosphorylation and translocation was supported by finding a complete inhibition of these responses after pretreatment with AG1478 (Fig. 4C). These results suggest that PI 3-kinase/Akt is one of the downstream effectors of the EGFR and that the D 2 receptor activates PI 3-kinase/Akt via transactivation of the EGFR in PC12-D 2 R cells.
To confirm the role of the EGFR in the activation of PI-3 kinase/Akt by D 2 receptors, we reduced the levels of EGFR expression in PC12-D 2 R cells using RNA interference. After transfection with EGFR-specific or control small interfering RNA (siRNA), cultures were assessed for EGFR protein expression by immunofluorescence. As shown in Fig. 5, EGFR was substantially repressed by 24 h post-transfection in ϳ60 -70% of the cells. The involvement of EGFR transactivation in D 2 receptor-stimulation of Akt was studied in PC12-D 2 R cells co- transfected with PH-Akt-GFP and EGFR siRNA or control siRNA. 24 h after transfection the cells were serum-starved for 1 h, stimulated with bromocriptine, and assessed for the expression of EGFR and the translocation of PH-Akt-GFP. In control siRNA-transfected cells, PH-Akt-GFP translocation was similar to that observed in cells not transfected with siRNA ( Fig. 6A and data not shown). However, EGFR repression by siRNA completely blocked the translocation of PH-Akt-GFP by bromocriptine (Fig. 6B). We conclude that EGFR is essential for the translocation of PH-Akt-GFP following D 2 receptor stimulation.
To clarify the mechanisms through which the D 2 receptor transactivates the EGFR, we examined the association between these two membrane proteins. After exposure of cells to vehicle or bromocriptine and preparation of cell extracts, we used a specific anti-D 2 receptor monoclonal antibody (Fig. 7A) to perform immunoprecipitations followed by immunoblotting for the EGFR (Fig. 7B). These studies showed that the EGFR co-immunoprecipitated with the D 2 receptor and the association between these two proteins was augmented in the presence of bromocriptine. Furthermore, the EGFR that complexed with the D 2 receptor in the presence of bromocriptine showed an increase in Tyr phosphorylation. These data indicate that the D 2 receptor and EGFR form a complex and that their association is augmented by bromocriptine. We also examined this complex for the presence of c-Src, which was detected in extracts immunoprecipitated by the D 2 receptor antibody. The presence of c-Src in this complex was also enhanced by exposure of the cells to bromocriptine (Fig. 7C).
The sites of EGFR Tyr phosphorylation induced by bromocriptine were studied using site-specific anti-phosphotyrosine antibodies. We analyzed tyrosine phosphorylation of the EGFR at residues 992 and 1068, which are EGFR autophosphorylation sites (38) and at residue 845 (Tyr 845 ), a known Src phosphorylation site (39). As shown in Fig. 8, incubation of PC12-D 2 R cells with EGF increased the phosphorylation of tyrosine residues 845, 992, and 1068, whereas bromocriptine only enhanced phosphorylation of Tyr 845 . The bromocriptinemediated phosphorylation of Tyr 845 was inhibited by pretreatment with EGFR inhibitor (Fig. 8C).
EGFR Transactivation Is c-Src-dependent-The finding that bromocriptine enhanced the association of the D 2 receptor with c-Src and induced phosphorylation of Tyr 845 , a Src-dependent phosphorylation site of the EGFR (39), led us to study further the role of c-Src in this signaling. Src family kinases have been implicated in the phosphorylation of the EGFR and of PI 3-kinase (39 -41). We examined the phosphorylation of Tyr 418 in c-Src, which is an autophosphorylation site required for kinase activity of c-Src (42). Cells were exposed to bromocriptine (100 nM) for periods up to 30 min. To determine whether c-Src was activated by D 2 receptor stimulation, we performed immunoblotting using an antibody specific for c-Src phospho-Tyr 418 . Stimulation of the D 2 receptor by bromocriptine caused c-Src to be phosphorylated at Tyr 418 (Fig. 9A).  We then examined the effects of the Src family tyrosine kinase inhibitor PP2 on D 2 receptor-mediated Akt phosphorylation and cell survival. PP2 completely inhibited both the capacity of bromocriptine to induce phosphorylation of Akt (Fig.  9B) and to mediate cell survival in the presence of oxidative stress (data not shown). PP2 also prevented the ability of bromocriptine treatment to induce the phosphorylation of c-Src-Tyr 418 and EGFR-Tyr 845 (Fig. 8C). However, inhibition of the EGFR by AG1478 did not affect the capacity of bromocriptine to induce phosphorylation of c-Src (Fig. 9C), suggesting that the EGFR is downstream of c-Src in D 2 receptor signaling. The role of c-Src in the D 2 receptor signaling was further evaluated using a dominant negative c-Src construct. When co-expressed with PH-Akt-GFP, the dominant negative c-Src kinase (k295R/Y527F) completely abolished translocation of PH-Akt-GFP in response to bromocriptine (Fig. 9D). Thus both pharmacological inhibition and dominant negative studies indicate that c-Src activation is required for signaling from the D 2 receptor through the EGFR to the neuroprotective PI 3-kinase/Akt pathway.

Translocation and Phosphorylation of Akt by D 2 Receptor
Stimulation Are Agonist-specific-We had previously found that D 2 receptor agonists varied greatly in their capacity to mediate increased survival of PC12-D 2 R cells and that their protective efficacy showed no correlation with G-protein activation, as assayed by GTP␥S binding. In particular, the efficacy of the agonists bromocriptine and pramipexole for GTP␥S binding were indistinguishable, whereas pramipexole was essentially devoid of neuroprotective activity in the PC12-D 2 R model. Having implicated the PI 3-kinase/Akt signaling pathway in the neuroprotection mediated by the D 2 receptor when complexed with bromocriptine, we were interested in determining the effects of pramipexole on this pathway. As shown in Fig.  10, pramipexole failed both to induce translocation of PH-Akt-GFP and phosphorylation of Akt in PC12-D 2 R cells. These results suggest that specific agonists that interact with the dopamine D 2 receptor differ markedly in their relative activation of classic and growth factor signaling pathways when complexed with the D 2 receptor (see "Discussion"). DISCUSSION We have delineated a D 2 receptor-activated signaling pathway that mediates neuroprotection by specific D 2 agonists in dopaminergic cell lines. Bromocriptine stimulates the PI 3-kinase/Akt pathway through a PTX-insensitive mechanism involving c-Src and transactivation of the EGFR. Our results suggest that the relative activation of classic G-protein and growth factor signaling pathways by the D 2 receptor is agonist-specific.
Because bromocriptine can induce the activation of the PI 3-kinase/Akt pathway and in many circumstances the modulation of Akt signaling normally occurs via growth factor stimulation, we sought to determine if the effects of bromocriptine on PC12-D 2 R cells were mediated through a growth factor receptor. Here, we report that the bromocriptine induced the activation of Akt within minutes and this activation required the EGFR. We show that EGFR-specific tyrosine kinase inhibitor completely block bromocriptine-induced activation of Akt. Furthermore, EGFR repression by siRNA also inhibited the translocation of PH-Akt-GFP by bromocriptine. Inhibitor and dominant negative Src studies show that the activation of the EGFR by the D 2 receptor involves Src. Co-immunoprecipitation studies show that the D 2 receptor complexes with the EGFR and with c-Src and that this association is enhanced by D 2 FIG. 6. Gene silencing of EGFR inhibits D 2 receptor signaling to Akt. EGFR siRNA inhibited bromocriptine-induced PH-Akt-GFP translocation. PC12-D 2 R cells were transfected with EGFR siRNA or control siRNA together with PH-Akt-GFP plasmid DNA. The cells were treated with bromocriptine (100 nM) for 10 min, and PH-Akt-GFP translocation and EGFR expression were examined in these cells. Each set of three vertical panels represents the same field. EGFR immunofluorescence is indicated in red (top panels). PH-Akt-GFP signal is indicated in green (middle panels). The bottom panels are overlays of both EGFR and PH-Akt-GFP signals. A, bromocriptine induced a characteristic ring-like margination of PH-Akt-GFP and concentration of PH-Akt-GFP into membrane processes in cells transfected with control siRNA. B, suppression of EGFR expression by EGFR siRNA eliminated the redistribution of PH-Akt-GFP by bromocriptine. Note that these images are from fixed cells and are not identical in appearance to the live cell images shown in Fig. 1. Data shown are representative of three independent experiments. FIG. 7. Bromocriptine induced the association of D 2 receptor with EGFR and c-Src. PC12-D 2 R cells were treated with either vehicle or 1 M bromocriptine for 10 min, and whole cell extracts were prepared. A, equal amount of cell extract, prior to immunoprecipitation, probed with the anti-D 2 receptor monoclonal antibody to establish specificity of the antibody. B, cell extracts were immunoprecipitated with monoclonal anti-D 2 receptor antibody and immunoblotted with antibodies to EGFR (upper panel). Stripping and reprobing with monoclonal anti-phosphotyrosine antibody revealed EGFR tyrosine phosphorylation associated with bromocriptine stimulated D 2 receptor (lower panel). C, immunoblot of D 2 receptor immunoprecipitate stained with antic-Src antibody. Note that activation of the D 2 receptor with bromocriptine increased the association of the receptor with both the EGFR and c-Src. All blots shown are representative of two to three independent experiments. receptor activation. The neuroprotective D 2 receptor signaling pathway we have characterized is summarized in Fig. 11.
Our data indicate that stimulation of c-Src/EGFR family kinases are required for Akt activation in response to bromocriptine in PC12-D 2 R cells. Src family kinases have been implicated in GPCR-induced EGFR tyrosine phosphorylation, and GPCRs can induce association of Src with the EGFR (43)(44)(45). In other studies, GPCR-induced EGFR tyrosine phosphorylation was found to be Src-independent (46,47). We found that bromocriptine caused activation of the EGFR, and inhibition of Src kinases had a significant effect on bromocriptine-induced EGFR tyrosine phosphorylation, implicating Src family kinases in bromocriptine-induced EGFR tyrosine phosphorylation. Src kinases can be activated by several heptahelical receptors (43,48,49) as well as by growth factor receptor stimulation (50), including the EGFR (39,43,51). c-Src has been reported to influence EGFR activity by mediating phosphorylation of Tyr 845 , a consensus Src phosphorylation site in the EGFR (39). Inhibition of either Src or the EGFR impaired the ability of bromocriptine to cause activation of Akt in PC12-D 2 R cells, indicating that activation of both proteins are required for this signaling. Inhibition of Src kinase also inhibited EGFR phosphorylation at Tyr 845 , whereas inhibition of the EGFR did not prevent phosphorylation of c-Src. These results suggest that c-Src is upstream of both the EGFR and Akt. Moreover, inhibiting either c-Src or the EGFR completely abolished the capacity of bromocriptine to increase cell survival during oxidative stress. Although Akt or upstream kinases have been reported to be a substrate for c-Src phosphorylation (52,53), in PC12-D 2 R cells both c-Src and EGFR phosphorylation were required for Akt activation. Therefore we propose that, in dopaminergic neurons, the D 2 receptor transactivates the EGFR through c-Src, which in turn activates the cytoprotective PI 3-kinase/Akt pathway.
Several mechanisms have been reported for heptahelical receptor activation of Src kinase. The ␤3 adrenergic receptor interacts with c-Src directly via Pro-rich domains in the receptor (49). Src activation by the ␤2 adrenergic receptor requires arrestin (48). The D 3 receptor has been found to contain noncanonical SH3 ligands (54). Putative SH3 domains are also present in the D 2 receptor, which might potentially mediate an interaction with Src. Using co-immunoprecipitation studies, we have demonstrated that the D 2 receptor and the EGFR form a complex that includes c-Src. Our inhibitor and dominant negative Src data further support the involvement of c-Src in activation of the EGFR. Whether the activation of c-Src by the D 2 receptor occurs directly or requires additional adaptor proteins remains to be determined.
Heptahelical receptors, including D 2 -class receptors, have been reported to induce activation of growth factor receptorcoupled pathways or PI 3-kinase (49,(55)(56)(57)(58)(59)(60)(61)(62)(63)(64). The cellular background in which a receptor is expressed may be important in determining its signaling potential. Platelet-derived growth factor receptor (PDGF) transactivation by the D 2 and D 4 receptors expressed in CHO cells has been reported (60). However, in contrast to our results for EGFR phosphorylation, the transactivation of the PDGF receptor in CHO cells showed sensitivity to PTX (60). The D 3 receptor expressed in CHO cells mediates activation of PI 3-kinase via atypical protein kinase C in a manner also sensitive to PTX (63). In striatal neurons, the D 2 agonist has been reported to activate Akt independently of PI 3-kinase activation (65). However, we find in both PC12-D 2 R and the dopaminergic SN4741 cells that Akt activation requires PI 3-kinase activity. Our results in the two dopamine cell lines studied are consistent with the observations of Kihara et al. (62) in cortical neurons, who also found that bromocriptine activated the PI 3-kinase/Akt pathway.
Our investigations further suggest that the D 2 receptor, when activated by bromocriptine, can couple both to heterotrimeric G i /G o family G-proteins and, simultaneously, to the PI 3-kinase/Akt signaling pathway (Fig. 11). These data led us to propose that the coupling to heterotrimeric G-protein and the coupling to PI 3-kinase/Akt may be independent. First, the G-protein coupling, but not the Akt activation showed PTX sensitivity. Second, the agonists studied differed in their capacity to activate each pathway. Bromocriptine activated both signaling pathways, whereas pramipexole, although quite efficient at simulating GTP␥S binding in these cells (17), failed to activate the PI 3-kinase/Akt signaling pathway (see Fig. 10). Our data suggest that the anti-apoptotic activity induced by dopamine agonists in these cells resulted from transactivation of the PI 3-kinase/Akt pathway. In a previous study, we found little correlation between the capacity of agonists to confer protection against oxidative stress and their capacity to activate classic G-protein signaling (17). Based on these results, we FIG. 8. Tyrosine phosphorylation of residue 845 of the EGFR by bromocriptine. PC12-D 2 R cells were incubated with 1 M bromocriptine for the indicated time points and EGF for 10 min. The cells were then lysed, and the lysates were resolved by SDS-gel electrophoresis on 7.5% gel and immunoblotted. Tyrosine phosphorylation was detected using site-specific tyrosine antibodies to EGFR from total cell lysates using the indicated anti-phosphotyrosine antibodies. EGFR levels were detected using anti-EGFR antibody. A, bromocriptine had no effect on the phosphorylation of EGFR Tyr 992 and Tyr 1068 . B, bromocriptine-induced phosphorylation of EGFR Tyr 845 . C, phosphorylation of EGFR Tyr 845 by bromocriptine was inhibited by AG1478 pre-treatment (200 nM, 30 min).
propose a model where the specificity of the agonist complexed with the D 2 -receptor determines the switching of signaling between G-protein and growth factor signaling pathways.
Kenakin (13) originally proposed that specific agonists acting at the same receptor might differentially activate downstream signaling pathways, a phenomenon he called agonist-mediated signal trafficking. Signal trafficking could arise as a result of receptors having multiple active conformational states that differ in their activation of specific signaling pathways. Agonists could cause different patterns of signaling by each inducing a different relative distribution of the accessible active states. Many studies suggest that heptahelical receptors exhibit properties consistent with the existence of multiple conformational states. In rhodopsin, for example, the existence of multiple conformers is evident from absorbance changes (9).  Multiple receptor conformational states are also evident in single molecule spectroscopy studies of the ␤2-adrenergic receptor (11) and are supported by the presence of phenotypically different serotonin 5HT2C receptor activation mutants (12). Pharmacological evidence for signal trafficking has been reported in several heptahelical receptors (14,15,66,67). Evidence for signal trafficking at the D 2 receptor based on the G-protein sensitivity of binding affinity has been previously reported for D 2 receptor expressing Sf21 insect cell lines. Notably, the agonist bromocriptine was found to induce a distinct pattern of coupling (15). Our results are consistent with the signal trafficking hypothesis and suggest that agonists acting at the D 2 receptor may differ markedly in their capacity to stabilize conformations leading to classic and growth factor signaling.
We implicate the capacity of D 2 agonists in transactivating the PI 3-kinase/Akt pathway and in mediating anti-apoptosis in PC12-D 2 R cells. Furthermore, we find evidence that the effectiveness of an agonist to protect against oxidative stress by activating PI 3-kinase/Akt may differ greatly for specific agonists. Our results support the hypothesis that agonists have a conformationally specific effect at the D 2 receptor. Among the agonists studied to date, we find agonists that preferentially activate GTP␥S binding and agonists that activate both GTP␥S binding and anti-apoptotic signaling. Our results suggest that it may be possible to identify agonists that specifically traffic signaling to the EGFR-PI 3-kinase/Akt pathway in dopamine neurons. Given the central role of the dopamine D 2 receptor in brain function, the refined model of conformationally dependent D 2 receptor signaling has important implications for the pathophysiology and treatment of brain diseases involving altered dopamine neuronal survival or neurotransmission, such as Parkinson's disease and schizophrenia.