Growth factor-mediated Fyn signaling regulates alpha-amino-3- hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor expression in rodent neocortical neurons

Abstract

Src-family protein tyrosine kinases (PTKs) transduce signals to regulate neuronal development and synaptic plasticity. However, the nature of their activators and molecular mechanisms underlying these neural processes are unknown. Here, we show that brain-derived neurotrophic factor (BDNF) and platelet-derived growth factor enhance expression of alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA)-type glutamate receptor 1 and 2/3 proteins in rodent neocortical neurons via the Src-family PTK(s). The increase in AMPA receptor levels was blocked in cultured neocortical neurons by addition of a Src-family-selective PTK inhibitor. Accordingly, neocortical cultures from Fyn-knockout mice failed to respond to BDNF whereas those from wild-type mice responded. Moreover, the neocortex of young Fyn mutants exhibited a significant in vivo reduction in these AMPA receptor proteins but not in their mRNA levels. In vitro kinase assay revealed that BDNF can indeed activate the Fyn kinase: It enhanced tyrosine phosphorylation of Fyn as well as that of enolase supplemented exogenously. All of these results suggest that the Src-family kinase Fyn, activated by the growth factors, plays a crucial role in modulating AMPA receptor expression during brain development.

Figure 1

Effects of growth factors on AMPA receptor expression in cultured rat neocortical neurons. Neocortical neurons were grown at a lower density (<800 cells/mm²) for 5 days with or without (−) daily applications of BDNF (50 ng/ml), EGF (20 ng/ml), PDGF (30 ng/ml), and bFGF (20 ng/ml). Protein was extracted from two sister cultures (for each factor), was separated by SDS/PAGE, was transferred to a membrane, and probed with antibodies directed against GluR1, GluR2/3, and α-tubulin. Note that the treatment of EGF suppressed GluR1 levels (P < 0.001) but not GluR2/3 levels. In contrast, bFGF effects on the AMPA receptors displayed an increasing tendency (57), but its effect failed to reach statistical significance. Control protein levels were 100 ± 4.9% for GluR1 and 100 ± 12.7% for GluR2/3, EGF-treated levels were 63 ± 3.7% for GluR1 and 111 ± 4.5% for GluR2/3, and bFGF-treated levels were 130 ± 13% for GluR1 and 140 ± 10% for GluR2/3 as determined by four independent blots (n = 4). Numbers on the right indicate the positions of protein size markers (kDa). See Fig. 2 for the effects of the other factors.

Figure 2

Effects of a Src-family-specific inhibitor on the growth factor-mediated increases in AMPA receptors. Rat neocortical cultures were prepared as shown in Fig. 1 and were treated with BDNF (A) or PDGF (B) in the presence or absence of a Src-family PTK inhibitor, PP1 (n = 3–4 for each condition). The effects of these factors and PP1 on the level of GluR1, GluR2/3, and NMDAR2A/B proteins were determined by immunoblotting. For statistical analyses, independent blots were made, and each immunoreactivity on the blots was measured by densitometry. The average control receptor level was set as 100% for each immunoblot. Statistical analysis was performed by using one-way ANOVA followed by the Bonferroni test. Protein levels of NMDAR2A/B and α-tubulin (data not shown) were not significantly altered by the treatments (F [2,6] = 1.1 in A and F [2,9] = 3.7 in B for NMDAR2A/B, and F [2,6] = 1.2 in A and F [2,9] = 0.7 in B for α-tubulin). The action of the inhibitor was controlled by monitoring the basal tyrosine phosphorylation level with an antiphosphotyrosine antibody (PY20); tyrosine phosphorylation of p60 was reduced to 58.1 ± 0.4% (P <0.001; n = 3). Protein recovery was not affected significantly by the growth factors or by PP1; control for BDNF; 0.550 ± 0.017 mg/dish, BDNF; 0.578 ± 0.022 mg/dish and PP1 plus BDNF; 0.480 ± 0.031 mg/dish (F [2,6] = 4.43), and control for PDGF; 0.616 ± 0.022 mg/dish, PDGF; 0.600 ± 0.025 mg/dish; PP1 plus PDGF; 0.578 ± 0.006 mg/dish (F [2,6] = 0.43).

Figure 3

Altered BDNF responses in neocortical culture prepared from Fyn knockout mice. (A) Neocortical cultures were prepared separately from neonates of three homozygous Fyn-knockout mice (lanes 1–3) and three wild-type mice (lanes 4–6). Each number represents a pair of cultures prepared from an individual animal. One set of sister cultures from each animal was treated for 4 days with 50 ng/ml BDNF administered daily, and the other set was used as a control. Cultures were processed individually for immunoblotting and were probed with antibodies against GluR1, GluR2/3, NMDAR1, NMDAR2A/B, and α-tubulin (an internal standard). (B) Protein levels of the glutamate receptors were measured and normalized with those of α-tubulin. Statistical analysis was performed on averaged values obtained from two independent blots using one-way ANOVA followed by the Bonferroni test. Filled bars represent BDNF-dependent protein levels (%) in neocortical cultures prepared from wild-type (+/+) mice, and clear bars represent those from homozygous Fyn-knockout (−/−) mice. Results are expressed as means ± SEM (n = 3).

Figure 4

Impaired expression of the AMPA receptors in the neocortex of young Fyn knockout mice. (A) Protein from the neocortex of Fyn knockout mice (postnatal 13–14 days old) was processed for immunoblotting with anti-Fyn, anti-Src, anti-GluR1, anti-GluR2/3, and anti-NF antibodies. Five wild-type mice (+/+), eight heterozygotes (+/−), and 10 homozygotes (−/−) were examined. Four examples representative for each genotype are shown. Note that some Fyn mutants exhibited rather normal GluR2/3 levels but expressed higher levels of Src (indicated with asterisks). (B) Fyn, Src, GluR1, GluR2/3, and NF immunoreactivity was quantified by densitometry. Protein levels for Fyn, Src, GluR1, and GluR2/3 were normalized by those of NF in each animal, and the average values for wild-type mice were set as 100%. Statistical analysis was performed with one-way ANOVA followed by Bonferroni test. Results are expressed as means ± SEM. GluR1 levels also were altered by Fyn mutation (F [2,20] = 3.52): 100 ± 16.9% (+/+), 59.5 ± 15.4% (+/−), and 50.3 ± 7.5% (−/−). In contrast, NF levels were not changed significantly among genotypes (F [2,20] = 0.80). (C) mRNA levels for GluR2, GluR3, and cyclophilin (an internal control) were measured by RNA blotting in four wild-type mice as well as four heterozygous and five homozygous Fyn mutants. Signal intensities for mRNA in each animal were measured and normalized with those for cyclophilin mRNA levels. Control values of wild-type mice were set as 100%. Statistical analysis was performed with one-way ANOVA (F [2,13] = 0.064 for GluR2 mRNA and F [2,13] = 0.31 for GluR3 mRNA).

Figure 5

Effects of BDNF on tyrosine phosphorylation of Fyn. Fyn activation by BDNF was examined by an in vitro kinase assay of the immunoprecipitates obtained with an anti-Fyn antibody. Tissue homogenate was prepared from the hippocampus of young rats (postnatal 4–6 days), was treated with (+) or without (−) BDNF, and was subjected to immunoprecipitation with an anti-Fyn polyclonal antibody. Half of the anti-Fyn immunoprecipitate from the hippocampus was used for immunoblotting to control for the quality of the immunoprecipitation (Left). Half of the anti-Fyn immunoprecipitate was incubated with [³²P]-γ-ATP, and SDS/PAGE then was performed. Phosphorylation of Fyn was visualized by film autoradiography (Center). Fyn immunoprecipitates also were taken from the neocortex, and the kinase assay was performed similarly (Right). The tyrosine kinase activity of Fyn was monitored by measuring phosphorylation levels of enolase that was supplemented exogenously into other sets of the reaction mixture (data not shown).