Address correspondence and reprint requests to John F. MacDonald, Medical Sciences Building, 1 King's College Circle, University of Toronto, Toronto, ON, M5S 1A8, Canada. E-mail: email@example.com
NMDA receptor function is modulated by both G-protein-coupled receptors and receptor tyrosine kinases. In acutely isolated rat hippocampal neurons, direct activation of the platelet-derived growth factor (PDGF) receptor or transactivation of the PDGF receptor by D4 dopamine receptors inhibits NMDA-evoked currents in a phospholipase C (PLC)-dependent manner. We have investigated further the ability of D2-class dopamine receptors to modulate NMDA-evoked currents in isolated rat prefrontal cortex (PFC). We have demonstrated that, similar to isolated hippocampal neurons, the application of PDGF-BB or quinpirole to isolated PFC neurons induces a slow-onset and long-lasting inhibition of NMDA-evoked currents. However, in contrast to hippocampal neurons, the inhibition of NMDA-evoked currents by quinpirole in PFC neurons is dependent upon D2/3, rather than D4, dopamine receptors. In PFC slices, application of both PDGF-BB and quinpirole induced a phosphorylation of the PDGF receptor at the PLCγ binding and activation site, Tyr1021. The PDGF receptor kinase inhibitor, tyrphostin A9, and the D2/3 dopamine receptor antagonist, raclopride, inhibited quinpirole-induced Tyr1021 phosphorylation. These finding suggest that quinpirole treatment inhibits NMDAR signaling via PDGF receptor transactivation in both the hippocampus and the PFC, and that the effects of quinpirole in these regions are mediated by D4 and D2/3 dopamine receptors, respectively.
The activity of ligand-gated ion channels, such as the NMDA receptor, is highly regulated by intracellular signaling pathways initiated by the activation of both G-protein-coupled receptors (GPCRs) and receptor tyrosine kinases, such as the platelet-derived growth factor (PDGF) receptor. PDGF is a growth factor initially isolated from platelets, but is widely expressed in the CNS and has multiple effects on neuronal growth, development, and signaling (Valenzuela et al. 1996; Lei et al. 1999; Hoch and Soriano 2003). The four isoforms of PDGF (A–D) form homo- or hetero-dimers, and activate PDGF α and β receptors (for review see Heldin et al. 1998). Activation of the PDGFβ receptor by PDGF-BB results in the dimerization and autophosphorylation of several tyrosine residues, and the subsequent binding and activation of multiple signaling proteins including src, phospholipase C (PLC) γ, and PI3-kinase (Heldin et al. 1998). PLCγ is activated by binding to phosphorylated Tyr1021 of the PDGF receptor via its N-terminal SH2 domain (Sultzman et al. 1991; Ronnstrand et al. 1992; Valius et al. 1993).
In addition to PDGF binding directly, the PDGFβ receptor can also be transactivated by reactive oxygen species (Saito et al. 2002) and GPCRs such as the D2-class dopamine receptors (D2, D3, D4) (Oak et al. 2001). We have recently reported that D4 dopamine receptor transactivation of the PDGF receptor in the hippocampus inhibited NMDA-evoked currents (Kotecha et al. 2002). In isolated hippocampal neurons, either quinpirole or dopamine application resulted in a long-lasting depression of peak NMDA-evoked currents that was slowly reversible (approximately 25 min) (Kotecha et al. 2002). Inhibition or inactivation of Gαi/o, inhibition of PLC, chelation of intracellular calcium, or stabilization of actin filaments abolished the ability of quinpirole to depress NMDA-evoked currents in hippocampal neurons (Kotecha et al. 2002).
In the prefrontal cortex (PFC), dopamine receptor activation has multiple effects on NMDA receptor signaling as well as on other ion channels. In isolated pyramidal cortical neurons from the rat PFC, D4 dopamine receptor activation by the D4 receptor-specific agonist, PD168077, transiently inhibited NMDA-evoked currents (Wang et al. 2003). The D2-class dopamine receptor agonist, quinpirole, also decreases excitability in PFC slices by modulating both NMDA and AMPA receptors (Tseng and O'Donnell 2004).
We investigated the possibility that transactivation of the PDGF receptor might be involved in D2-class dopamine receptor regulation of NMDA receptors in the PFC. Activation of D2-class dopamine receptors by quinpirole inhibited NMDA-evoked currents in isolated PFC neurons. The inhibition of NMDA-evoked currents was attenuated by incubation with the PDGF receptor kinase inhibitor, tyrphostin A9, and the D2/3 dopamine receptor antagonist, raclopride. Quinpirole also induced a tyrosine phosphorylation on the PDGFβ receptor at Tyr1021 in PFC. These findings suggest that quinpirole treatment of PFC neurons leads to a long-lasting inhibition of NMDA-evoked currents that is dependent on D2/3 dopamine receptors and occurs via transactivation of PDGF receptors.
Materials and methods
(–)-Quinpirole hydrochloride and PD168077 maleate were purchased from Tocris (Ellisville, MO, USA), and Tyrphostin A9 was purchased from Biomol (Plymouth, PA, USA). All other chemical reagents were purchased from Sigma (St. Louis, MO, USA). The anti-PDGFRβ and the anti-phospho-Tyr1021-PDGFRβ antibodies were purchased from Santa Cruz (Santa Cruz, CA, USA).
Cell isolation and whole-cell recording
CA1 neurons were isolated from hippocampal slices of post-natal day 14–21 Wistar rats as previously described (Wang and MacDonald 1995). PFC neurons were isolated from slices prepared as follows. A coronal cut was made approximately 1.5 mm posterior to the front end of the cerebral cortex, followed by a gentle cut along the rhinal fissure. Tissue attached ventrally to the rhinal fissure was discarded. Sagittal cuts were made to the remaining cortical tissues, giving rise to sagittal slices of the PFC with a thickness of approximately 300–400 µm. The extracellular solution (ECF) was composed of 140 mm NaCl, 1.3 mm CaCl2, 25 mm HEPES, 33 mm glucose, 5.4 mm KCl, 0.5 µm tetrodotoxin and 0.5 µm glycine, with a pH of 7.3–7.4 and osmolarity ranging from 320 to 330 mOsm. Recordings were carried out at room temperature (22°C). The membrane potential was held at − 60 mV throughout the recordings and a voltage step of 10 mV was applied prior to NMDA application to monitor series resistance. The intracellular solution consisted of 11 mm EGTA as intracellular calcium chelating buffer, 10 mm HEPES, 2 mm MgCl2, 2 mm tetraethyl ammonium chloride (TEA-Cl) to block the K+ channel, 1 mm CaCl2, 140 mm CsF and 4 mm K2ATP. NMDA currents were evoked by rapid application of NMDA solution delivered from a multi-barreled, fast perfusion system for 2 s in every minute. The solution was delivered at a rate of approximately 1 mL/min.
PFC and CA1 slices were dissected as described above. After recovery in oxygenated ECF, slices were incubated with drugs for 5 min. In experiments involving PDGFR inhibitors or dopamine receptor antagonists, slices were pre-incubated for 5 min prior to quinpirole treatment. After drug treatment, the ECF was decanted and tissues were dounce-homogenized in solubilization buffer [20 mm Tris, pH = 7.5, 150 mm NaCl, 1 mm EDTA, 1 mm EGTA, 1% triton, 30 mm Na pyrophosphate, 1 mm betaglycerophosphate, 1 mm Na3VO4, 40 mm NaF, 4 mm Pefabloc (Roche, Mannheim, Germany), MINI cocktail (Roche), 5 µg/mL aprotinin, 1 µg/mL pepstatin, 2.5 mm benzamidine]. Lysates were incubated on ice for 20–30 min and insoluble material was pelleted by centrifugation at 14 000 g for 10 min. Total protein concentration was determined by bicinchoninic acid (BCA) protein assay (Pierce, Rockford, IL, USA). Samples were separated by electrophoresis on 7.5% acrylamide gels, transferred to nitrocellulose membrane and probed with primary antibodies directed against PDGFRβ (1 : 300) or phospho-1021-PDGFRβ (1 : 1000). After washing and additional incubation with donkey anti-rabbit secondary antibody (1 : 4000) (Amersham, Little Chalfont, UK), immunoreactivity was determined by enhanced chemiluminescence (ECL) (Pierce), and imaging and analysis were performed using a Kodak Image Station 2000R (Kodak, Rochester, NY, USA).
Acutely dissected tissue slices were treated with vehicle or PDGF-BB for 10 min. After drug treatment, the slices were incubated with 0.5–1.0 mg/mL sulfo-NHS-LC-biotin (Pierce) at 4°C for 30 min. Slices were washed three times with Tris-buffered saline (TBS) and homogenized in solubilization buffer + 1% sodium dodecyl sulfate (SDS). After a 30 min incubation on ice, insoluble material was removed by centrifugation. Total protein concentrations were determined and sample protein concentrations were normalized. Equal volumes of samples were incubated overnight with 50 µL avidin-cojugated beads (Sigma). Beads were subsequently washed in phosphate-buffered saline (PBS) with 0.5% SDS/0.5% triton-X and boiled with loading buffer for 5 min before SDS/polyacrylamide gel electrophoresis (PAGE) gel electrophoresis.
All animal experimentation was conducted in accordance with the Policies on the Use of Animals at the University of Toronto.
Statistical analysis of the data was completed using the Prism graphpad program. Graphs and sample tracings were made from the Origin program. All data are reported as mean ± SEM. Significance level was set at α = 0.05. Data were analyzed by two-way anova, with time and drug effects being the variable factors, or Student's t-test where appropriate.
Application of NMDA (50 µm) to PFC pyramidal neurons evoked a large inward current characterized by a rapid onset and decay to a steady-state value (steady-state current, Iss). In acutely isolated PFC neurons, peak NMDA-evoked currents were stable over a 25 min time course (peak at t = 25 min: 92.0 ± 5.4%, p = 0.085 n = 6, Student's t-test). However, there was a slight run-down of steady-state current (73.3 ± 3.5%, p < 0.001). Treatment of isolated PFC neurons with PDGF-BB (10 ng/mL) inhibited NMDA-evoked currents by > 10%, and this inhibition was not completely reversible after 25 min of recording (data not shown). We have previously demonstrated that PDGF-induced depression of NMDA-evoked currents is PLC-dependent in CA1 hippocampal neurons (Kotecha et al. 2002). To investigate the role of PLC in PDGF receptor-induced inhibition of NMDA currents, we pre-incubated the cells with the PLC inhibitor, U73122, or the inactive analog, U73343, for 10 min. PDGF-BB-induced inhibition of NMDA-evoked currents was absent in cells pre-treated with U73122, but not in cells pre-treated with U73343, suggesting that PDGF-induced inhibition of NMDA-evoked currents is PLC-dependent in PFC neurons (Fig. 1).
We have previously demonstrated that application of a selective D2-class dopamine receptor agonist, quinpirole, reduced NMDA-evoked currents by about 30% in hippocampal CA1 neurons (Kotecha et al. 2002). In isolated PFC neurons, quinpirole application depressed the peak current up to 21.1 ± 4.0% (n = 7; p < 0.05, two-way anova for time and drug effects) (Figs 2a and b) compared with control recordings (data not shown). The quinpirole-induced depression of NMDA-evoked currents was attenuated in the presence of the PDGF receptor kinase inhibitor, tyrphostin A9 (Fig. 2a). This result suggests that, similar to hippocampal neurons, activation of D2-class dopamine receptors in the PFC inhibits NMDA-evoked currents by transactivating PDGF receptors. We then investigated which D2-class dopamine receptor subtype was responsible for inhibiting NMDA-evoked currents after quinpirole treatment. Pre-incubation of PFC neurons with raclopride [500-fold more selective for D2/3 dopamine receptors compared with D4 dopamine receptors (Asghari et al. 1995) also attenuated quinpirole-induced inhibition of NMDA-evoked currents (Fig. 2a).
In CA1 hippocampal slices, quinpirole treatment induces a tyrosine phosphorylation on the PDGF receptor (Kotecha et al. 2002). Treatment of PFC slices with both PDGF-BB or quinpirole induced a phosphorylation of the PDGF receptor at Tyr1021 (Fig. 2c). Pre-incubation (5 min) of PFC slices with either tyrphostin A9 (Fig. 2d) or raclopride (Fig. 2e) prevented quinpirole treatment from inducing Tyr1021 phosphorylation. These findings suggest that quinpirole transactivates the PDGFβ receptor in the PFC via D2/3 dopamine receptor activation.
The ability of raclopride to completely attenuate both the PDGF receptor phosphorylation and the inhibition of NMDA-evoked currents by quinpirole suggests that these signaling events are mediated by D2/3 rather than D4 dopamine receptors. However, activation of D4 dopamine receptors by the D4 dopamine receptor-selective agonist, PD168077 (Glase et al. 1997), is reported to inhibit NMDA-evoked currents in PFC neurons (Wang et al. 2003). In addition, the D4 dopamine receptor subtype is responsible for the quinpirole depression of NMDA-evoked currents in CA1 hippocampal neurons (Kotecha et al. 2002). To examine specifically the role of D4 dopamine receptors in modulating NMDA-evoked currents in PFC neurons, we treated the neurons with PD186077. PD168077 did not depress the NMDA-evoked current (PD168077: 100.9 ± 1.9% vs. control, n = 6) (Fig. 3a). As a positive control, we applied PD168077 under the same experimental condition to neurons isolated from the hippocampal CA1 region of the rat brain. In CA1 hippocampal neurons, PD168077 application resulted in a robust depression of NMDA-evoked current up to 49.8 ± 5.7% (n = 6, p < 0.01, two-way anova) (Fig. 3b). The highly selective D4 dopamine receptor antagonist, L745870 (Bristow et al. 1997), blocked the NMDA-evoked current depression induced by PD168077 (Fig. 3c).
The results of the electrophysiological studies suggest that D4 dopamine receptor activation does not inhibit NMDA-evoked currents in PFC neurons. We then compared the ability of D4 dopamine receptors to transactivate the PDGF receptor in PFC slices. In CA1 hippocampal slices, both quinpirole and PD168077 induced Tyr1021 phosphorylation on the PDGF receptor. However, in PFC slices, only quinpirole treatment resulted in a robust phosphorylation at Tyr1021 (Fig. 4). This suggests that D4 dopamine receptors do not transactivate PDGF receptors in the PFC, and may explain the lack of PD168077-induced inhibition of NMDA-evoked currents. D4 dopamine receptor activation has been shown to modulate NMDA-evoked currents in PFC slices in 3 to 5-week-old rats (Wang et al. 2003). To examine the possibility that there may be age-related changes in the ability of D4 dopamine receptors to transactivate PDGF receptors, we incubated PFC slices from 4 to 5-week-old rats with PD168077. However, PD168077 treatment did not increase PDGF receptor phosphorylation at Tyr1021 in these older animals (105 ± 14% vs. vehicle, n = 5). In addition, Wang et al. suggested that D4 dopamine receptor inhibition of NMDA-evoked currents involves a decrease in the surface expression of the NMDA receptor subunit, NR1 (Wang et al. 2003). To examine the mechanism of NMDA current inhibition downstream of PDGF receptors, we applied either vehicle or PDGF-BB (10 mg/mL) to PFC slices and examined NR1 surface expression. NR1 surface expression was not significantly different in PFC or CA1 hippocampal slices after PDGF-BB treatment (PFC: 125 ± 19% of vehicle, n = 5; CA1: 87 ± 10% of vehicle, n = 6).
PDGF receptor activation results in a decrease in NMDA-evoked currents in hippocampal CA1 slices, hippocampal cultures, and isolated hippocampal neurons (Valenzuela et al. 1996; Lei et al. 1999). Recombinant PDGFRβ also inhibits NMDA-evoked currents in oocytes (Valenzuela et al. 1996). In oocytes expressing a PDGF receptor mutant lacking tyrosine residues important for signaling to phosphatidylinositol 3-kinase (PI3K), ras-GAP, tyrosine phosphatases and PLCγ, no inhibition of NMDA-evoked currents was observed (Valenzuela et al. 1996). Restoration of the PLCγ-binding tyrosine 1021 was sufficient to restore the inhibition of NMDA-evoked currents (Valenzuela et al. 1996). The PLC inhibitor, U73122, also abolished the depression of NMDA-evoked currents in isolated CA1 hippocampal neurons after treatment with PDGF-BB or quinpirole (Lei et al. 1999; Kotecha et al. 2002).
The signaling pathway described herein is similar to our previous work in isolated CA1 hippocampal neurons. However, in the CA1, it was the D4 dopamine receptor subtype that was responsible for the depression of NMDA-evoked currents after quinpirole treatment (Kotecha et al. 2002). In the present study in PFC slices, the D4 dopamine receptor agonist, PD168077, had no effect on NMDA-evoked currents or on PDGF receptor Tyr1021 phosphorylation. There are several possible reasons why D4 dopamine receptor activation failed to transactivate PDGF receptors in the prefrontal cortex. There may be differences in the co-localization of PDGF receptors and D2/3 or D4 receptors in the PFC. In addition, although both D2/3 or D4 receptor activation transactivate PDGF receptors and lead to a subsequent phosphorylation of extracellular signal-regulated kinase (ERK)1/2 in Chinese hamster ovary (CHO) cells (Oak et al. 2001), the sequestration of Gβγ subunits prevented D4, but not D2, dopamine receptor signaling to ERK1/2 in CHO cells (Oak et al. 2001). This result highlights the possibility that the mechanism(s) of transactivation may not be identical amongst all dopamine receptor subtypes. Furthermore, D2 and D4 receptors have also been shown to couple to different Gα subunits, and can differentially signal in recombinant cells lines (O'Hara et al. 1996a,b; Obadiah et al. 1999).
These results appear to conflict with a recent report demonstrating that direct D4 dopamine receptor activation by the D4 dopamine receptor-selective agonist, PD168077, inhibited NMDA-evoked currents in a pathway involving protein kinase A (PKA), protein phosphatase 1 (PP1) and calcium-calmodulin kinase II (CaMKII) (Wang et al. 2003). However, in that study, the effects of PD168077 were fast-onset and disappeared after removal of the agonist (Wang et al. 2003), whereas the effect of quinpirole in the present study developed more slowly and was much longer lasting. In addition, Wang et al. used isolated cells from older rats (3–5 weeks) than those used in our studies (2–3 weeks) (Wang et al. 2003). Another report from the same group demonstrated that the activation of the neuregulin receptor tyrosine kinase, inhibited NMDA currents by activating a PLC-dependent pathway (Gu et al. 2005). In addition, D2 dopamine receptor activation inhibits inhibitory synaptic currents in the rat PFC by transactivating PDGF receptors (Trantham-Davidson et al. 2004). Further experiments should be directed at elucidating the differences in the kinetics and signaling pathways modulating NMDA-evoked currents after D2-class dopamine receptor and receptor tyrosine kinase activation in PFC neurons.
PDGF receptors undergo both an N- and an O-glycosylation before expression at the membrane (Claesson-Welsh et al. 1987; Daniel et al. 1987). Removal of the N- and O-linked oligosaccharide chains reduce the apparent molecular mass by 40 and 7 kDa, respectively (Daniel et al. 1987). The antibodies used in this study detected two major bands, the glycosylated 180 kDa band that is presumably located on the cell surface, and an unglycosylated band of approximately 130 kDa. Interestingly, both PDGF-BB and quinpirole induced a much more robust Tyr1021 phosphorylation on the unglycosylated form of the PDGF receptors. Much of the biochemical and pharmacological detail regarding PDGF receptor localization and signaling is from non-neuronal cells, therefore, further investigation into the role of the different post-translational forms of the PDGF receptor may help in the understanding of how PDGF receptor signaling inhibits NMDA-evoked currents.
This work was supported by a grant to JFM from the Canadian Institutes of Health Research.