Volume 93, Issue 4
Free Access

Dopamine D2 receptor stimulation of mitogen‐activated protein kinases mediated by cell type‐dependent transactivation of receptor tyrosine kinases

Chunhe Wang

Departments of Behavioral Neuroscience

Neurology, Oregon Health & Science University

Search for more papers by this author
David C. Buck

Departments of Behavioral Neuroscience

VA Medical Center, Portland, Oregon, USA

Search for more papers by this author
Rui Yang

Departments of Behavioral Neuroscience

VA Medical Center, Portland, Oregon, USA

Search for more papers by this author
Tara A. Macey

Departments of Behavioral Neuroscience

Search for more papers by this author
Kim A. Neve

Departments of Behavioral Neuroscience

VA Medical Center, Portland, Oregon, USA

Search for more papers by this author
First published: 22 April 2005
Citations: 46
Address correspondence and reprint requests to: Kim A. Neve, Research Service (R&D‐30), VA Medical Center, 3710 SW US Veterans Hospital Road, Portland, OR 97239–2999, USA.
E‐mail: [email protected]

Abstract

Dopamine D2 receptor activation of extracellular signal‐regulated kinases (ERKs) in non‐neuronal human embryonic kidney 293 cells was dependent on transactivation of the platelet‐derived growth factor (PDGF) receptor, as demonstrated by the effect of the PDGF receptor inhibitors tyrphostin A9 and AG 370 on quinpirole‐induced phosphorylation of ERKs and by quinpirole‐induced tyrosine phosphorylation of the PDGF receptor. In contrast, ectopically expressed D2 receptor or endogenous D2‐like receptor activation of ERKs in NS20Y neuroblastoma cells, which express little or no PDGF receptor, or in rat neostriatal neurons was largely dependent on transactivation of the epidermal growth factor (EGF) receptor, as demonstrated using the EGF receptor inhibitor AG 1478 and by quinpirole‐induced phosphorylation of the EGF receptor. The D2 receptor agonist quinpirole enhanced the coprecipitation of D2 and EGF receptors in NS20Y cells, suggesting that D2 receptor activation induced the formation of a macromolecular signaling complex that includes both receptors. Transactivation of the EGF receptor also involved the activity of a matrix metalloproteinase. Thus, although D2 receptor stimulation of ERKs in both cell lines was decreased by inhibitors of ERK kinase, Src‐family protein tyrosine kinases, and serine/threonine protein kinases, D2‐like receptors activated ERKs via transactivation of the EGF receptor in NS20Y neuroblastoma cells and rat embryonic neostriatal neurons, but via transactivation of the PDGF receptor in 293 cells.

Abbreviations used

  • EGF
  • epidermal growth factor
  • EGFP
  • enhanced green fluorescent protein
  • ERK
  • extracellular signal‐regulated kinase
  • GPCR
  • G protein‐coupled receptor
  • HEK
  • human embryonic kidney 293 cells
  • HSV
  • herpes simplex virus
  • MEK
  • mitogen‐activated protein kinase/ERK kinase
  • PDGF
  • platelet‐derived growth factor
  • RTK
  • receptor tyrosine kinase
  • G protein‐coupled receptors (GPCRs) use multiple overlapping pathways to stimulate the mitogen‐activated protein kinases that include extracellular signal‐regulated kinases (ERKs) and stress‐activated protein kinase/Jun‐amino‐terminal kinase (van Biesen et al. 1996; Gutkind 1998). The activation of ERKs by GPCRs contributes to the regulation of DNA synthesis and mitogenesis in many different cell types (Dhanasekaran et al. 1998). Among the many tracks by which GPCRs stimulate ERKs is transactivation of receptor tyrosine kinases (RTKs) (Daub et al. 1996; Wetzker and Bohmer 2003). RTK transactivation refers to a process by which activation of a GPCR in turn activates an RTK, thus recruiting the ERK cascade of the classical RTK signaling pathway (Schlessinger 2000), and represents a subset of the potential mechanisms or tracks by which GPCRs activate the ERK cascade. The most extensively studied track within this subset is transactivation of the epidermal growth factor (EGF) receptor by a triple membrane‐passing signal mechanism in which stimulation of a GPCR activates a membrane‐bound metalloproteinase that stimulates the extracellular release of an endogenous ligand for the EGF receptor (Prenzel et al. 1999; Kalmes et al. 2000; Gschwind et al. 2001). It is not clear to what extent this model is applicable to the transactivation of other RTKs such as the platelet‐derived growth factor (PDGF) receptor, the insulin growth factor‐1 receptor, and the fibroblast growth factor receptor (Wetzker and Bohmer 2003). There are also ligand‐independent mechanisms for RTK transactivation that involve RTK phosphorylation by Src‐family protein tyrosine kinases and/or the participation of the GPCR and the RTK in a multi‐receptor signaling complex (Maudsley et al. 2000; Wetzker and Bohmer 2003).

    A single GPCR can engage different RTKs to activate ERKs in different types of cells (Wetzker and Bohmer 2003). For example, the angiotensin II receptor transactivates the EGF, PDGF, and fibroblast growth factor receptors in cardiomyocytes (Asakura et al. 2002), vascular smooth‐muscle cells (Linseman et al. 1995), and bovine adrenal medulla cells (Peng et al. 2002), respectively, and the lysophosphatidic acid receptor transactivates the PDGF receptor in L cells that lack the EGF receptor, the EGF receptor in COS‐7 cells expressing the EGF receptor but not the PDGF receptor, and predominantly the EGF receptor in Rat‐1 cells that express both EGF and PDGF receptors (Herrlich et al. 1998). Similar cell type‐specific transactivation of multiple RTKs by one GPCR is observed for the bradykinin B2 receptor (Zwick et al. 1997; Adomeit et al. 1999).

    Dopamine receptors are GPCRs that are classified as D1‐like (D1 and D5 receptors), or D2‐like (D2, D3 and D4 receptors) (Neve and Neve 1997). D2‐like receptors signal primarily through the pertussis toxin‐sensitive Gαi/o class of heterotrimeric G proteins to elicit a number of cellular responses, including inhibition of adenylate cyclase and Ca2+ channels, increased K+ conductance and Na+/H+ exchange, and activation of phospholipase C (Huff 1996). The D2 receptor also activates ERKs (Luo et al. 1998; Welsh et al. 1998; Choi et al. 1999; Ghahremani et al. 2000; Oak et al. 2001; Kim et al. 2004) and Jun‐amino‐terminal kinase (Luo et al. 1998). Endogenous D2‐like receptors activate ERKs in brain slices (Otani et al. 1999; Yan et al. 1999) and in vivo (Cai et al. 2000). In postmitotic neurons, activation of ERKs might be involved not only in cell survival and in synaptic plasticity (Fukunaga and Miyamoto 1998; Impey et al. 1999; Otani et al. 1999), but also in acute behavioral responses to stimulation of dopamine receptors (Cai et al. 2000).

    The mechanisms by which the D2 receptor activates ERKs have not been fully elucidated. In CHO cells, D2 receptor activation of ERKs is mediated by transactivation of the PDGF receptor (Oak et al. 2001). D2‐like receptors also transactivate the PDGF receptor in hippocampal neurons, an interaction implicated in D2‐like receptor inhibition of glutamate receptor‐mediated synaptic neurotransmission (Kotecha et al. 2002). In contrast, the D2 receptor transactivates the EGF receptor in PC12 cells, a pathway implicated in D2 receptor‐mediated cytoprotection against oxidative stress (Nair and Sealfon 2003). Importantly, the role of RTK transactivation in D2 receptor activation of ERKs has not been studied in neurons or neuronal cell lines.

    In this report, we investigated the role of differential RTK transactivation in D2 receptor‐mediated activation of ERKs in NS20Y neuroblastoma cells, primary neostriatal neuronal culture, and non‐neuronal HEK293 cells transfected with the rat D2L receptor. We now report that although PDGF receptor transactivation was necessary for D2 receptor activation of ERKs in HEK293 cells, transactivation of the EGF receptor contributed substantially to D2 and D2‐like receptor‐stimulated ERK activation in NS20Y cells and neostriatal neurons.

    Experimental procedures

    Materials

    [3H]Spiperone (95 Ci/mmol) was purchased from Amersham Biosciences (Piscataway, NJ, USA). Quinpirole, spiperone, (+)‐butaclamol, and inhibitors for the EGF and PDGF receptors were purchased from Sigma (St. Louis, MO, USA). GM6001, PD98059, H‐89, PP2, bisindolylmalemide I, wortmannin, staurosporine, and ophiobolin A were from Calbiochem (San Diego, CA, USA). Cell transfection kit was purchased from Ambion (Austin, TX, USA). The polyclonal antibody for dually phosphorylated (i.e. activated) ERKs was from Promega (Madison, WI, USA), EGF and PDGFβ receptor antibodies were from Upstate (Charlottesville, VA, USA), and phospho (p‐Tyr1045)‐EGF receptor and phospho (p‐Tyr751)‐PDGFβ receptor antibodies were purchased from New England Biolabs (Beverly, MA, USA). Antiserum for D2 receptor was obtained from Chemicon (Temecula, CA, USA). The SuperSignal™ West Pico chemiluminescent kit was from Pierce Biotechnology (Rockford, IL, USA). Alexa 568‐labeled goat anti‐rabbit IgG antibody and Prolong™ anti‐fade kit were obtained from Molecular Probes (Eugene, OR, USA). Modified L‐15 medium and Dulbecco's modified Eagle's medium were purchased from Sigma, and Neurobasal medium and B‐27 supplements were from Invitrogen (Carlsbad, CA, USA). Fetal bovine and calf bovine sera for cell culture were purchased from HyClone (Logan, UT, USA). HEK293 (HEK) cells, a transformed cell line from human embryonic kidney, and NS20Y cells, a mouse neuroblastoma cell line established from an A/Jax mouse strain with neuroblastoma, were purchased from ATCC (Manassas, VA, USA).

    Production and maintenance of D2‐transfected cell lines

    The HEK‐D2 cell line was created by transfection of a rat D2L cDNA construct in pcDNA1 together with the plasmid pBabe Puro into HEK cells and selection with puromycin (2 µg/mL) as described (Watts and Neve 1996). The NS20Y‐D2 clone was created by transfection of a rat D2L cDNA in pcDNA3 into NS20Y cells and selection with G418 (600 µg/mL). Puromycin‐ or G418‐resistant colonies were isolated and screened for expression of [3H]spiperone binding sites. Cells were maintained in Dulbecco's modified Eagle's medium supplemented with 5% fetal bovine serum and 5% calf bovine serum, penicillin–streptomycin, and puromycin (2 µg/mL) or G418 (600 µg/mL), and grown in a humidified incubator at 37°C in the presence of 10% CO2. D2 receptors are not detectable in untransfected HEK293 or NS20Y cells, as determined by the lack of specific binding of [3H]spiperone.

    Radioligand binding

    The binding of [3H]spiperone was assessed as described previously (Watts et al. 1998). Confluent cells in 10‐cm plates were harvested by lysis with ice‐cold hypotonic buffer (1 mm Na+‐HEPES, pH 7.4, 2 mm EDTA). After swelling for 10–15 min, the cells were scraped from the plate and spun at 24 000 g for 20 min. The resulting crude membrane fraction was resuspended in Tris‐buffered saline (50 mm Tris‐HCl, pH 7.4, with 155 mm NaCl) with a Brinkmann Polytron homogenizer (Westbury, NY, USA) at setting 6 for 10 s. Aliquots of the membrane preparation (5–15 µg of protein) were added to duplicate assay tubes containing the following: Tris‐buffered saline, 0.001% bovine serum albumin, and radioligand in a final volume of 1 mL. (+)‐Butaclamol (2 µm) was used to define non‐specific binding. Incubations were performed at 37°C for 45 min and terminated by filtration using a 96‐well Tomtec cell harvester (Orange, CT, USA). Fifty microliters of BetaPlate scintillation fluid was added to each sample after filters were dried. Radioactivity on the filters was determined using a Wallac 1205 BetaPlate scintillation counter. The Bmax and Kd values were determined by non‐linear regression using prism 3.0 (GraphPad, San Diego, CA, USA).

    Cell stimulation and immunoblotting

    NS20Y‐D2 and HEK‐D2 cells were grown in 6‐well or 12‐well plates to 80–90% confluence. The cells were starved in serum‐free Dulbecco's modified Eagle's medium overnight, then incubated with receptor agonists at 37°C for the indicated durations. In some experiments, RTK inhibitors were added 45 min before agonist. Incubation was terminated by placing the tissue culture cluster on ice and rapidly aspirating the medium, followed by the addition of ice‐cold solubilization buffer (50 mm Tris‐HCl, pH 7.5, 150 mm NaCl, 1% NP‐40, 0.5% deoxycholate, 0.1% sodium dodecyl sulfate, 1 mm NaVO3, and protease inhibitors) and incubation for 15 min with shaking. After centrifugation (14 000 g at 4°C for 15 min), the supernatant was collected and the protein concentration was measured and adjusted using solubilization buffer. Samples (30 µL) with equal amounts of protein mixed with Laemmli loading buffer were denatured at 70°C for 10 min and separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis. Alternatively, for coprecipitation experiments, solubilized cell lysates were centrifuged at 20 000 g for 15 min at 4°C, and the supernatants were collected and pre‐cleared with protein G‐Sepharose beads. Equal amounts of cellular proteins were incubated with anti‐EGF receptor monoclonal antibody for 2 h at 4°C, followed by incubation with protein G‐Sepharose for 1 h. The resulting immune complexes were washed three times with lysis buffer and then resuspended in 40 µL of Laemmli loading buffer for separation by sodium dodecyl sulfate–polyacrylamide gel electrophoresis. The proteins were transferred to polyvinylidene difluoride membrane and visualized with the SuperSignal™ West Pico chemiluminescent kit. The intensity of bands was quantified using ImageQuant™ 5.2 from Amersham Biosciences. In addition to measuring protein concentrations to ensure equal loading, in many experiments in which phosphorylated forms of ERKs, EGF receptor, and PDGF receptor were detected, the blots were stripped and re‐probed for total ERKs, EGF receptor, and PDGF receptor.

    Primary neuronal cultures

    Striata were dissected from Sprague‐Dawley rat embryos (gestation day 21) and mechanically dissociated by gentle pipetting in modified L‐15 medium. After allowing 10 min for clumped tissue to settle, the suspended cells were pelleted by spinning at 4°C for 10 min at 200 g. Cell pellets were re‐suspended in minimal essential medium supplemented with 10% fetal bovine serum, 0.45% glucose, 5 pg/mL insulin, 0.5 mm glutamine, 100 U/mL penicillin, and 100 µg/mL streptomycin and plated on 18‐mm diameter poly d‐lysine‐coated glass coverslips (Fisher 12‐545‐84 18cir‐1D) at a density of 75 000 cells per coverslip. Neurobasal™ medium with B27 supplement and 0.5 mm l‐glutamine was added 1 h after initial plating. To obtain a near pure neuronal culture, 10 µm of cytosine arabinoside was added after 1 day of culturing and continued for 2 days to stop the proliferation of the microglials and astrocytes. The cultures were maintained in a humidified atmosphere with 5% CO2 for at least 10 days before experiments.

    Herpes simplex virus‐mediated expression of the D2 receptor in neuronal cultures

    A herpes simplex virus (HSV) vector was used to drive the expression of the D2 receptor with enhanced green fluorescent protein (EGFP) fused to the C teminus (D2‐EGFP) (Macey et al. 2004). Replication‐defective HSV vectors were packaged at a titer of approximately 2 × 105 infectious units/µL as described (Neve et al. 1997). Cells were infected with HSV‐D2‐EGFP (approximately 1 infectious unit/cell) 18 h prior to detection of the D2 receptor by fluorescence microscopy or stimulation of ERKs.

    Confocal immunofluorescence imaging

    Neostriatal neurons were treated as described in Results, then fixed with 10% paraformaldehyde at 25°C for 30 min and permeabilized with 100% methanol at −20°C for 15 min. After permeabilization, neurons were immersed in blocking buffer (1% bovine serum albumin 3% goat serum in Dulbecco's phosphate‐buffered saline) with gentle agitation for 3 h before incubation with phospho‐ERK antibody (1 : 1000) overnight at 4°C. After five washes with Dulbecco's phosphate‐buffered saline (15 min each), coverslips were incubated with Alexa 568‐labeled goat anti‐rabbit IgG (1 : 500; 90 min at room temperature), followed by five additional washes. The coverslips were then mounted with the ProLong™ anti‐fade kit, dried in the dark, and scanned at 568 nm with a Leica TCS SP confocal laser scanning microscope. Images were digitally captured and deconvolved using Power HazeBuster™ imaging program (VayTek Inc., Fairfield, VA, USA).

    Results

    Stable expression of D2 receptors in NS20Y‐D2 and HEK‐D2 cells

    Clonal NS20Y‐D2 and HEK‐D2 cell lines were chosen for our study based on their comparable and constant receptor expression levels. The density of receptors was determined by saturation analysis of the binding of [3H]spiperone to membranes prepared from the cells. Bmax values determined by non‐linear regression were 538 ± 20 and 496 ± 124 fmol/mg of protein for NS20Y‐D2 and HEK‐D2 cells, respectively, with Kd values of 44 ± 4 and 40 ± 14 pm.

    Activation of extracellular signal‐regulated kinases by the D2 receptor in NS20Y‐D2 and HEK‐D2 cells

    Receptor‐stimulated activation of ERKs (ERK1, 44 kDa and ERK2, 42 kDa) was measured using an antibody for phospho‐ERKs to quantify the abundance of dually phosphorylated ERKs. Treatment with the D2 receptor agonist quinpirole induced rapid and robust activation of ERKs in both NS20Y‐D2 and HEK‐D2 cells (Fig. 1). The two cell lines were indistinguishable in terms of the time‐ and concentration‐dependence of stimulation, with maximal responses for both observed within 5 min of treatment with 1 µm quinpirole (Figs 1a and b). In both cell lines the responses were mediated by the D2 receptor, as demonstrated by blockade with the D2 receptor antagonist spiperone (Fig. 1c).

    image

    Quinpirole‐induced activation of extracellular signal‐regulated kinases (ERKs) in NS20Y and HEK293 cells transfected with the D2L receptor. NS20Y‐D2 or HEK‐D2 cells were incubated overnight in serum‐free medium before treatment with quinpirole (Quin). Cells were lysed on ice and the levels of ERK activation were quantified by immunoblotting for the dually phosphorylated ERKs. In each panel, an immunoblot from one experiment for each cell line representative of three independent experiments is shown above the averaged data, which are expressed as a percentage of the maximum response observed in that cell line ± SE. (a) ERK activation in response to treatment with the indicated concentrations of quinpirole for 5 min. (b) Time course of the activation of ERKs by 1 µm quinpirole. (c) Quinpirole‐induced activation of ERKs (1 µm for 5 min), compared to untreated cells (NT) was prevented by the D2 receptor antagonist spiperone (Spip, 10 µm).

    Cell type‐specific transactivation of epidermal growth factor and platelet‐derived growth factor receptors by the D2 receptor

    We used selective RTK inhibitors to determine if EGF or PDGF receptors mediate D2 receptor activation of ERKs. In HEK‐D2 cells, the PDGF receptor inhibitor tyrphostin A9 abolished quinpirole‐ or PDGF‐induced accumulation of phospho‐ERKs (Figs 2a and c), but had no effect on stimulation by EGF (Fig. 2b), whereas the EGF receptor inhibitor AG 1478 prevented stimulation by EGF (Fig. 2b), but had no effect on ERK activation by quinpirole or PDGF (Figs 2a and c). In contrast, in NS20Y‐D2 cells the EGF receptor inhibitor AG 1478 substantially decreased quinpirole‐induced activation of ERKs (Fig. 2e), and prevented ERK activation by EGF (Fig. 2f), but had no effect on the modest activation of ERK induced by PDGF (Fig. 2g). Interestingly, although tyrphostin A9 inhibited PDGF‐induced activation of ERKs in HEK‐D2 cells, it could not be used to assess the activity of the PDGF receptor in NS20Y cells because it caused robust activation of ERKs on its own (Figs 2e–g). To confirm the respective roles of the EGF and PDGF receptors in NS20Y and HEK293 cells, we used additional EGF receptor and PDGF receptor inhibitors. The PDGF receptor inhibitor AG 370 had no effect on activation by quinpirole in NS20Y‐D2 cells (Fig. 2h) although it inhibited activation of ERK by PDGF in the cells (data not shown), but AG 370 substantially decreased stimulation by quinpirole in HEK‐D2 cells (Fig. 2d), whereas the EGF receptor inhibitor AG 112 (0.1 µm) prevented quinpirole‐induced ERK activation in NS20Y‐D2 cells (data not shown).

    image

    Transactivation of the epidermal growth factor (EGF) or platelet‐derived growth factor (PDGF) receptors was required for D2 receptor‐mediated activation of extracellular signal‐regulated kinases (ERKs) in NS20Y or HEK293 cells, respectively. Cells were pre‐treated with vehicle (no inhibitor), the PDGF receptor inhibitors tyrphostin A9 (1 µm) or AG 370 (20 µm), or the EGF receptor inhibitor AG 1478 (2.5 µm) for 45 min before analysis of ERK activation in response to treatment with vehicle (NT, not treated) or receptor agonist for 5 min. The panels depict results using HEK‐D2 cells (a–d) or NS20Y‐D2 cells (e–h), with each panel including an immunoblot from a representative experiment above the averaged results (mean ± SE) from three independent experiments, which are expressed as a percentage of the maximum response in that set of experiments and normalized to the amount of total ERK immunoreactivity. (a and e) Effect of receptor tyrosine kinase (RTK) inhibitors on ERK activation induced by quinpirole (Quin, 1 µm). (b and f) Effect of RTK inhibitors on ERK activation induced by EGF (10 ng/mL). (c and g) Effect of RTK inhibitors on PDGF (100 ng/mL) induced activation of ERKs. (d and h) Effect of AG 370 on ERK activation induced by quinpirole (Quin, 1 µm). The apparently smaller stimulation by tyrphostin A9 in (f) compared to (e) and (g) is because of the relatively strong response to EGF; blot exposure times for EGF were reduced so as not to overexpose the image, and results are expressed as a percentage of the EGF response. *p < 0.05, **p < 0.01 compared to untreated cells (NT).

    Involvement of a metalloproteinase in D2 receptor‐stimulated epidermal growth factor receptor transactivation

    According to the triple membrane‐passing signal model, a soluble EGF‐like ligand, heparin‐binding‐EGF (HB‐EGF), is generated by proteolytic processing of proHB‐EGF by metalloproteinases in response to GPCR stimulation (Prenzel et al. 1999, 2000; Kalmes et al. 2000; Gschwind et al. 2001). The involvement of metalloproteinases in EGF receptor transactivation by D2 receptor was studied using the selective inhibitor, GM6001. As shown in Fig. 3, although GM6001 had no effect on the ERK activation induced by EGF, the metalloproteinase inhibitor significantly and dose‐dependently decreased quinpirole‐induced activation of ERKs in NS20Y‐D2 cells. Substantial quinpirole‐induced ERK activation was observed even in the presence of 10 µm GM6001.

    image

    Metalloproteinase involvement in quinpirole‐stimulated, but not epidermal growth factor (EGF)‐stimulated, extracellular signal‐regulated kinase (ERK) activation in NS20Y cells. Cells were pre‐treated with various concentrations of the matrix metalloproteinase inhibitor GM6001 for 15 min before stimulation with quinpirole (Quin, 1 µm) or EGF (10 ng/mL) for 5 min. The ERK activation levels were measured by immunoblotting with an antibody specific for phosphorylated ERKs. Immunoblots from experiments representative of three independent experiments for each agonist are shown above the averaged data (mean ± SE) for quinpirole and EGF, which are expressed as a percentage of the maximal response for that agonist and, for quinpirole, normalized to the amount of total ERK immunoreactivity. *p < 0.05, **p < 0.01 compared to cells treated with quinpirole but not GM6001.

    Signaling pathways regulating D2 receptor activation of extracellular signal‐regulated kinases in HEK and NS20Y cells

    To investigate further the pathways by which the D2 receptor activates ERKs in HEK and NS20Y cells, we used a variety of protein kinase inhibitors (Fig. 4). Pre‐treatment of cells with PD98059, a selective mitogen‐activated protein kinase/ERK kinase (MEK) inhibitor, abolished the quinpirole‐induced activation of ERKs in both HEK and NS20Y cells. Substantial inhibition was also observed after treatment with PP2, an inhibitor of the Src family of tyrosine kinases, and staurosporine, a broad spectrum serine/threonine protein kinase inhibitor. Wortmannin, an inhibitor of phosphatidylinositol 3‐kinase, partially inhibited quinpirole‐stimulated phosphorylation of ERKs in HEK‐D2 cells, but not in NS20Y‐D2 cells. Selective inhibitors of protein kinase C (bisindolylmaleimide I), protein kinase A (H‐89), or calcium/calmodulin‐dependent protein kinase (ophiobolin A) had little or no effect on quinpirole‐induced activation of ERKs.

    image

    Effect of protein kinase inhibitors on D2 receptor‐stimulated activation of extracellular signal‐regulated kinases (ERKs). Cells were serum‐starved overnight and treated with PD98059 (250 µm, 40 min), H‐89 (2 µm, 40 min), PP2 (10 µm, 15 min), bisindolylmaleimide I (BisI, 5 µm, 40 min), wortmannin (Wort, 2.34 µm, 10 min), staurosporine (Stau, 40 µm, 40 min), or ophiobolin A (25 µm, 3 min) prior to treatment with vehicle (NT) or quinpirole (Quin, 1 µm, 5 min). Cells were lysed on ice and the levels of ERK activation were quantified by immunoblotting for the dually phosphorylated ERKs. An immunoblot from one experiment for each cell line representative of three independent experiments is shown above the averaged data (mean ± SE), which are expressed as a percentage of the maximal response in that cell line.

    Phosphorylation of the epidermal growth factor and platelet‐derived growth factor receptors and inducible association of the D2 and epidermal growth factor receptors

    To extend the observation that D2 receptor‐stimulated activation of ERKs was decreased by EGF receptor inhibitors in NS20Y‐D2 cells, but by PDGF receptor inhibitors in HEK‐D2 cells, we determined the levels of phosphorylation of the EGF and PDGF receptors in these cells upon D2 receptor stimulation. Application of either quinpirole or PDGF increased the phosphorylation of the PDGF receptor in HEK‐D2 cells, whereas PDGF receptor or phospho‐PDGF receptor immunoreactivity was not detected in NS20Y cells after either treatment (Fig. 5a). In NS20Y‐D2 cells, on the other hand, phosphorylation of the EGF receptor was markedly increased by treatment with either quinpirole or EGF for 2 min (Fig. 5b). Treatment with EGF, but not treatment with quinpirole, also increased the phosphorylation of the EGF receptor in HEK‐D2 cells. Moreover, treatment with quinpirole for 2 min dramatically increased the coprecipitation of D2 and EGF receptors in NS20Y‐D2 cells (Fig. 5c). Treatment with EGF, on the other hand, caused little or no increase in the coprecipitation of D2 and EGF receptors. No treatment changed the abundance of total EGF or PDGF receptor immunoreactivity. EGF receptor immunoreactivity was not detected in HEK cells, and PDGF receptor immunoreactivity could not be detected in NS20Y cells. The D2 receptor did not precipitate with the PDGF receptor in HEK‐D2 cells when stimulated with quinpirole (data not shown), even though transactivation of the PDGF receptor was detected in this cell line (Fig. 5).

    image

    Cell type‐dependent transactivation of the epidermal growth factor (EGF) and platelet‐derived growth factor (PDGF) receptors and D2 receptor‐stimulated coprecipitation of the D2 and EGF receptors. (a) PDGF receptor tyrosine phosphorylation was increased by quinpirole and PDGF in HEK‐D2 cells, but not in NS20Y‐D2 cells. Cells were treated for 2 min with vehicle (NT), quinpirole (Quin, 1 µm) or PDGF‐ββ (PDGF, 100 ng/mL), after which the level of PDGF receptor tyrosine phosphorylation was measured by immunoblotting with antibodies for Tyr751‐phosphorylated PDGF receptor (p‐PDGFR). Blots were subsequently stripped and probed with an antibody for total PDGF‐β receptor. (b) Phosphorylation of the EGF receptor was increased by quinpirole treatment in NS20Y‐D2 cells, but not in HEK‐D2 cells, and increased by EGF treatment in both cell lines. Cells were treated for 2 min with vehicle (NT), quinpirole (Quin, 1 µm), or EGF (10 ng/mL) after which the level of EGF receptor tyrosine phosphorylation was measured by immunoblotting with antibodies for Tyr1045‐phosphorylated EGF receptor (p‐EGFR). Blots were subsequently stripped and probed with an antibody for total EGF receptor. (c) D2 receptor‐stimulated coprecipitation of the D2 and EGF receptors. NS20Y‐D2 cells were treated for 2 min with vehicle (NT), quinpirole (Quin, 1 µm),or EGF (10 ng/mL). Cells were lysed, and the cell lysates were collected and precipitated with an EGF receptor antibody. The presence of D2 and EGF receptors in the precipitate was measured by immunoblotting with a D2 or EGF receptor antibody. *p < 0.05, **p < 0.01 compared to untreated cells (NT).

    Transactivation of the epidermal growth factor receptor was required for D2‐like receptor‐stimulated extracellular signal‐regulated kinase activation in primary neostriatal neurons

    We next examined whether transactivation of the EGF receptor is required for the activation of ERK by the D2 receptor in primary neurons from rat neostriatum, which is the richest source of dopamine receptors (Bouthenet et al. 1987). Quinpirole stimulation of endogenous D2‐like receptors in neostriatal neurons increased the phosphorylation of the EGF receptor, but not of the PDGF receptor (Fig. 6a). The presence of functional EGF and PDGF receptors in the neurons was demonstrated by the ability of EGF and PDGF to enhance phosphorylation of their respective receptors (Fig. 6a), and by the robust activation of ERK that was induced by EGF and PDGF (Fig. 6b). Treatment with quinpirole also increased the abundance of phospho‐ERK in a subset of neostriatal neurons (Figs 6b and c), an effect inhibited by the D2‐like receptor antagonist (+)‐butaclamol (data not shown). Interestingly, ERK phosphorylation induced by PDGF appeared to be concentrated in cell bodies, whereas that induced by quinpirole or EGF was in the cell bodies and also distributed widely throughout the neuronal processes. The level of ERK activation induced by stimulation of endogenous D2‐like receptors was substantially decreased by the EGF receptor inhibitor AG 1478 (Fig. 6c), an effect confirmed with another EGF receptor inhibitor AG 112 (0.1 µm, data not shown). In contrast, the PDGF receptor inhibitors AG 1295 (Fig. 6c) and AG 370 (data not shown) had little effect on the response to quinpirole. Tyrphostin A9 caused marked activation of ERKs on its own, similar to what was observed in NS20Y‐D2 cells (data not shown).

    image

    Transactivation of the epidermal growth factor (EGF) receptor, but not the platelet‐derived growth factor (PDGF) receptor, was required for ERK activation by the endogenous D2 receptor in neostriatal neurons. Primary cultures of rat neostriatal neurons were prepared from Sprague‐Dawley rat embryos and grown on poly d‐lysine treated glass coverslips. (a) Treatment with either quinpirole or EGF enhanced the phosphorylation of the EGF receptor in rat neostriatal neurons, whereas only PDGF treatment increased the phosphorylation of the PDGF receptor. (b) Activation of extracellular signal‐regulated kinase (ERK) in neostriatal neurons by quinpirole, EGF and PDGF. Neuron cultures in (a) and (b) were treated with vehicle (NT), quinpirole (Quin, 1 µm), EGF (10 ng/mL), or PDGF‐ββ (100 ng/mL) for 5 min before the phosphorylation level of the EGF and PDGF receptors (a) and ERKs (b) was measured by immunoblotting (a) or confocal immunofluorescence imaging (b) with an antibody against phospho‐ERK. (c) Role of receptor tyrosine kinase (RTK) transactivation in D2‐like receptor activation of ERKs in neostriatal neurons. Neuron cultures were pre‐treated for 45 min with vehicle (no inhibitor), AG 1478 (2.5 µm), or AG 1295 (1 µm) before incubation with vehicle (NT) or quinpirole (Quin, 1 µm) for 5 min to stimulate ERKs. The lower panel shows the intensity of immunofluorescence in the cell bodies (∼6 per experiment) quantified using ImageQuant™ 5.2 (mean of three experiments ± SE). Images shown in (b) and (c) are representative of 20–30 images obtained from three independent experiments in which the abundance of immunoreactivity for phospho‐ERK was examined in the most intensely stained cells from each treatment condition. The presence of neurons in each frame was verified by inspection of the transmitted light image (not shown). *p < 0.05, **p < 0.01 compared to untreated cells (NT).

    Quantitative immunoblot analysis of ERK activation by endogenous D2‐like receptors in the neuronal cultures was not possible due to low and variable quinpirole‐stimulated activity, presumably because only a subset of cells expressed D2‐like receptors. To quantify the effect of RTK inhibitors on activation of ERKs by the D2 receptor in neurons, we used an HSV vector to express recombinant D2‐EGFP in the neuronal cultures (Neve et al. 1997). Whereas neurons infected with HSV‐LacZ showed little fluorescence, cells infected with HSV‐D2‐EGFP displayed robust EGFP autofluorescence (Fig. 7a). Treating HSV‐D2‐EGFP‐infected neuronal cultures with quinpirole caused robust activation of ERKs that was prevented by the EGF receptor inhibitors AG 1478 and AG 112, but not by the PDGF receptor inhibitor AG 1295 (Fig. 7b).

    image

    Transactivation of the epidermal growth factor (EGF) receptor was required for extracellular signal‐regulated kinase (ERK) activation by the D2 receptor heterologously overexpressed in rat neostriatal neurons. (a) Herpes Simplex virus (HSV) mediated expression of a D2 receptor‐enhanced green fluorescent protein fusion protein (D2‐GFP) in neostriatal neuronal cultures resulted in fluorescence that was not detectable in cells infected with HSV‐LacZ. Eighteen hours after infection with HSV‐LacZ or HSV‐D2‐GFP, the neurons were imaged with an Olympus CX640 fluorescence microscope. Transmitted light images in the lower row demonstrate the presence of cells in both preparations. (b) Eighteen hours after infection with HSV‐D2‐GFP, neurons were treated for 45 min with vehicle (no inhibitor), the EGF receptor inhibitors AG 1478 (2.5 µm) and AG 112 (0.1 µm), or the PDGF receptor inhibitor AG 1295 (1 µm), followed by incubation with vehicle (NT, no treatment) or quinpirole (Quin, 1 µm for 5 min). Neurons were lysed on ice and the levels of ERK activation were quantified by immunoblotting for the dually phosphorylated ERKs. An immunoblot from one experiment for each cell line representative of three independent experiments is shown above the averaged data (mean ± SE), which are expressed as a percentage of the maximum response to quinpirole. **p < 0.01 compared to untreated cells (NT).

    Discussion

    G protein βγ subunits, phosphotidylinositol‐3‐kinase, Src, RTKs, and the small molecular weight G protein Ras have all been implicated in D2 receptor activation of ERKs (Huff 1996; Luo et al. 1998; Welsh et al. 1998; Choi et al. 1999; Yan et al. 1999; Cai et al. 2000; Ghahremani et al. 2000; Oak et al. 2001). In CHO cells, the D2L receptor activates ERKs via transactivation of the PDGF receptor (Oak et al. 2001; Kim et al. 2004). Transactivation of the PDGF receptor by D2‐like receptors is also observed in hippocampal neurons (Kotecha et al. 2002), whereas a recombinant D2 receptor transactivates the EGF receptor in neuron‐like PC12 cells (Nair and Sealfon 2003). Because of the relatively low abundance of D2 dopamine receptors in the hippocampus (Missale et al. 1998), and the lack of evidence in neuronal cells concerning mediation of ERK activation by RTK transactivation, the aim of this study was to evaluate the role of RTK transactivation in D2 receptor stimulation of ERKs in D2 receptor‐expressing cell lines and in neurons of the neostriatum, where the D2 receptor is present at higher densities.

    We used non‐neuronal HEK293 cells and NS20Y neuroblastoma cells, both stably transfected with a rat D2L receptor cDNA, and primary cultures of neurons obtained from embryonic rat neostriatum. In both HEK‐D2 and NS20Y‐D2 cells, the D2 receptor agonist quinpirole caused a rapid and robust increase in the abundance of activated ERKs. Quinpirole‐induced activation of ERKs was decreased by the D2‐like receptor antagonist (+)‐butaclamol and by inhibitors of the ERK kinase MEK, of serine‐threonine protein kinases, and of Src‐family protein tyrosine kinases. The protein tyrosine kinase inhibitor PP2 has also been reported to be an inhibitor of the PDGF receptor (Oak et al. 2001), an effect that might have contributed to its ability to prevent stimulation of ERK in HEK293 cells. The two cell lines differed in sensitivity to the phosphatidylinositol 3‐kinase inhibitor wortmannin, with partial inhibition observed in HEK‐D2 cells but not in NS20Y‐D2 cells.

    In HEK‐D2 cells, two types of evidence support the conclusion that selective transactivation of the PDGF receptor mediates activation of ERKs. First, the PDGF receptor inhibitors tyrphostin A9 and AG 370 prevented the quinpirole‐induced ERK activation, whereas inhibitors of the EGF receptor had no effect. Second, quinpirole treatment enhanced the tyrosine phosphorylation of the PDGF receptor, but not the EGF receptor. These results are consistent with the D2L and D4 receptor‐stimulated transactivation of the PDGF receptor observed in CHO cells (Oak et al. 2001; Kim et al. 2004); interestingly, Kim et al. (2004) observed that D2S receptor‐stimulated activation of ERKs occurs via an arrestin‐mediated pathway instead of transactivation the PDGF receptor.

    In NS20Y‐D2 cells, on the other hand, the EGF receptor inhibitor AG 1478, but not the PDGF receptor inhibitor AG 370, greatly decreased quinpirole‐induced activation of ERKs, and treatment with quinpirole enhanced the tyrosine phosphorylation of the EGF receptor, but not the PDGF receptor, suggesting that selective transactivation of the EGF receptor mediates the response to stimulation of the D2 receptor. We also obtained evidence for the mechanism of D2 receptor transactivation of the EGF receptor; quinpirole‐stimulated ERK activation was decreased by pre‐treatment with a metalloproteinase inhibitor, and quinpirole treatment induced an interaction between the D2 and EGF receptors that increased their coprecipitation. Quinpirole treatment also activated ERKs in neostriatal neurons, an effect that was diminished by EGF receptor inhibitors but not by PDGF receptor inhibitors, and enhanced the tyrosine phosphorylation of the EGF receptor but not the PDGF receptor. These results suggest a model in which D2 receptor activation of ERKs in neostriatal neurons or neuroblastoma cells involves the formation of a macromolecular signaling complex that includes both D2 and EGF receptors, and also the release of HB‐EGF that activates EGF receptors directly. It is noteworthy that transactivation of EGF receptor might not be the only mechanism by which the D2 receptor activates ERKs, as quinpirole stimulated ERK activation was not completely blocked by AG 1478 or by GM6001.

    Yan et al. (1999) reported that D2 receptor activation of ERKs in neostriatal slices is mediated by activation of phospholipase C, because it is prevented by combined inhibition of protein kinase C (with Go6976) and chelation of intracellular calcium. These results are not necessarily in conflict with ours, as RTK inhibitors were not tested in the prior study and we did not evaluate the effect of protein kinase C inhibitors combined with chelation of calcium in neurons, but our finding that the protein kinase C inhibitor bisindoylmaleimide I had no effect on ERK activation in NS20Y‐D2 cells indicates that this branch of the phospholipase C pathway was not involved in the response. It is also possible that the process of adapting to culturing changes signaling mechanisms compared to native neurons.

    NS20Y‐D2 cells had little PDGF receptor expression, as demonstrated by the lack of PDGF receptor or phospho‐PDGF receptor immunoreactivity and by the modest PDGF‐stimulated activation of ERKs in the cells, and based on the present results we cannot rule out the possibility that cotransfection with the PDGF receptor would reveal D2 receptor transactivation of that RTK. Phospho‐PDGF receptor immunoreactivity and PDGF receptor‐mediated activation of ERKs were present in neostriatal neurons, but the pattern of activation (concentrated in the cell bodies as opposed to the D2 receptor‐mediated activation that was observed throughout the processes) suggests that the two receptors are not colocalized to the same subcellular compartments. Thus, the lack of D2 receptor‐mediated transactivation of the PDGF receptor in neurons and NS20Y cells could be due to little colocalization in the former and lack of expression in the latter cells. The lack of transactivation of the PDGF receptor in neurons observed in our study compared to D2‐like receptor transactivation of the PDGF receptor in hippocampal neurons (Kotecha et al. 2002) could be explained by tissue‐specific factors that differ between neostriatum and hippocampus such as the relative abundance of D2‐like receptor subtypes, by the age of the rats from which cell cultures were established, or by differences in culture conditions. Interestingly, we found that the PDGF receptor inhibitor tyrphostin A9 could not be used to evaluate PDGF receptor involvement in activation of ERKs in neurons or the neuroblastoma cells, because tyrphostin A9 has an additional site of action that caused robust accumulation of phospho‐ERKs.

    Although the EGF receptor was present at lower levels in HEK293 cells than in NS20Y cells, as shown in Fig. 5(b), this seems less likely to explain the lack of transactivation of the EGF receptor in HEK‐D2 cells, because treating the cells with EGF caused both enhanced tyrosine phosphorylation of the EGF receptor and robust activation of ERKs that was prevented by the EGF receptor inhibitor AG 1478. Moreover, transactivation of the EGF receptor by other GPCRs has been observed in HEK293 cells (Belcheva et al. 2001; Grewal et al. 2001; Shah et al. 2003, 2004b). D2 receptor transactivation of the PDGF receptor rather than the EGF receptor in HEK293 cells may reflect a preference for the PDGF receptor if it is present and appropriately localized in the cell, or it may reflect the absence of a neuronal protein that is required for formation of a macromolecular signaling complex that includes the D2 and EGF receptors.

    This study also revealed an important physiological role played by the EGF receptor in neostriatal neurons. Both EGF and PDGF have been demonstrated to exert neuromodulatory and neurotrophic effects (Morrison et al. 1987; Terlau and Seifert 1989; Abe et al. 1992; Valenzuela et al. 1996; Lei et al. 1999; Abe and Saito 2000). In hippocampal tissue, the PDGF receptor, whether activated by its cognate ligands (Lei et al. 1999; Valenzuela et al. 1996) or by GPCR transactivation (Kotecha et al. 2002), regulates excitatory synaptic neurotransmission via Ca2+‐dependent inactivation of N‐methyl‐d‐aspartate receptors. Although EGF causes a long‐lasting enhancement of excitatory synaptic transmission (Terlau and Seifert 1989; Abe et al. 1992), GPCR transactivation of the EGF receptor has been demonstrated in immortalized hypothalamic neurons (Shah et al. 2004a), and D2 receptor cytoprotection in PC12 cells is mediated by transactivation of the EGF receptor (Nair and Sealfon 2003), to our knowledge transactivation of the EGF receptor by GPCRs in neurons in primary culture has not been previously reported. Since the localization and ontogenetic time of appearance of EGF receptor immunoreactivity are not consistent with a direct mitogenic role of the receptor in astroglia proliferation during development, it has been proposed that the EGF receptor in brain might be involved in neuronal survival or neuron‐glia signaling (Gómez‐Pinilla et al. 1988), a hypothesis that is supported and extended by our finding that the EGF receptor mediated D2‐like receptor signaling.

    Taken together, our data suggest that although the D2 receptor activated ERKs via transactivation of the PDGF receptor in non‐neuronal HEK293 cells, D2 and D2‐like receptors in neuronal cells recruited the EGF receptor via agonist‐induced formation of a multi‐receptor signaling complex and the participation of matrix metalloproteinases. The activity of MEK, a Src family kinase, and one or more serine/threonine protein kinases was required in both types of cells for full activation of ERKs by D2 or D2‐like receptors, whereas evidence for involvement of phosphatidylinositol 3‐kinase was observed in HEK‐D2 cells but not in NS20Y‐D2 cells. Furthermore, these results suggest that the EGF receptor is involved not only in neuromodulatory and neurotrophic processes, but also in the acute responses to dopamine D2 receptor stimulation that are mediated by ERKs.

    Acknowledgements

    This work was supported by United States Public Health Service grants MH045372 and DA07262, and by the VA Merit Review and Career Scientist programs.

      The full text of this article hosted at iucr.org is unavailable due to technical difficulties.