Scholarly article on topic 'Necdin controls EGFR signaling linked to astrocyte differentiation in primary cortical progenitor cells'

Necdin controls EGFR signaling linked to astrocyte differentiation in primary cortical progenitor cells Academic research paper on "Biological sciences"

CC BY-NC-ND
0
0
Share paper
Academic journal
Cellular Signalling
OECD Field of science
Keywords
{Necdin / EGFR / ERK / Gliogenesis / "MAGE family"}

Abstract of research paper on Biological sciences, author of scientific article — Izumi Fujimoto, Koichi Hasegawa, Kazushiro Fujiwara, Masashi Yamada, Kazuaki Yoshikawa

Abstract Cellular signaling mediated by the EGF receptor (EGFR) plays a key role in controlling proliferation and differentiation of cortical progenitor cells (CPCs). However, regulatory mechanisms of EGFR signaling in CPCs remain largely unknown. Here we demonstrate that necdin, a MAGE (melanoma antigen) family protein, interacts with EGFR in primary CPCs and represses its downstream signaling linked to astrocyte differentiation. EGFR was autophosphorylated and interacted with necdin in EGF-stimulated CPCs. Necdin bound to autophosphorylated EGFR via its tyrosine kinase domain. EGF-induced phosphorylation of ERK was enhanced in necdin-null CPCs, where the interaction between EGFR and the adaptor protein Grb2 was strengthened, suggesting that endogenous necdin suppresses the EGFR/ERK signaling pathway in CPCs. In necdin-null CPCs, astrocyte differentiation induced by the gliogenic cytokine cardiotrophin-1 was significantly accelerated in the presence of EGF, and inhibition of EGFR/ERK signaling abolished the acceleration. Furthermore, necdin strongly suppressed astrocyte differentiation induced by overexpression of EGFR or its ligand binding-defective mutant equivalent to a glioblastoma-associated EGFR variant. These results suggest that necdin acts as an intrinsic suppressor of the EGFR/ERK signaling pathway in EGF-responsive CPCs to restrain astroglial development in a cell-autonomous manner.

Academic research paper on topic "Necdin controls EGFR signaling linked to astrocyte differentiation in primary cortical progenitor cells"

ELSEVIER

Contents lists available at ScienceDirect

Cellular Signalling

journal homepage: www.elsevier.com/locate/cellsig

Cellular Signalling

Necdin controls EGFR signaling linked to astrocyte differentiation in primary cortical progenitor cells

Izumi Fujimoto a, Koichi Hasegawa a, Kazushiro Fujiwara a, Masashi Yamada b, Kazuaki Yoshikawa a'*

a Laboratory of Regulation of Neuronal Development, Institute for Protein Research, Osaka University, Osaka, Japan b Laboratory of Extracellular Matrix Biochemistry, Institute for Protein Research, Osaka University, Osaka, Japan

ARTICLE INFO ABSTRACT

Cellular signaling mediated by the EGF receptor (EGFR) plays a key role in controlling proliferation and differentiation of cortical progenitor cells (CPCs). However, regulatory mechanisms of EGFR signaling in CPCs remain largely unknown. Here we demonstrate that necdin, a MAGE (melanoma antigen) family protein, interacts with EGFR in primary CPCs and represses its downstream signaling linked to astrocyte differentiation. EGFR was autophosphorylated and interacted with necdin in EGF-stimulated CPCs. Necdin bound to autophosphorylated EGFR via its tyrosine kinase domain. EGF-induced phosphorylation of ERK was enhanced in necdin-null CPCs, where the interaction between EGFR and the adaptor protein Grb2 was strengthened, suggesting that endogenous necdin suppresses the EGFR/ERK signaling pathway in CPCs. In necdin-null CPCs, astrocyte differentiation induced by the gliogenic cytokine cardiotrophin-1 was significantly accelerated in the presence of EGF, and inhibition of EGFR/ERK signaling abolished the acceleration. Furthermore, necdin strongly suppressed astrocyte differentiation induced by overexpression of EGFR or its ligand binding-defective mutant equivalent to a glioblastoma-associated EGFR variant. These results suggest that necdin acts as an intrinsic suppressor of the EGFR/ERK signaling pathway in EGF-responsive CPCs to restrain astroglial development in a cell-autonomous manner.

© 2015 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license

(http://creativecommons.org/licenses/by-nc-nd/4.0/).

CrossMark

Article history:

Received 13 October 2015

Received in revised form 23 November 2015

Accepted 30 November 2015

Available online 2 December 2015

Keywords: Necdin EGFR ERK

Gliogenesis MAGE family

1. Introduction

The growth factors EGF and bFGF (also known as FGF-2) are indispensable for maintaining the self-renewal and multipotency of neural stem cells residing in developing mammalian telencephalon [1-4]. CPCs proliferate and differentiate into neuronal progenitors in response to bFGF at early stages of embryonic cortical development and into glial progenitors in response to EGF at the late stages [5-8]. These temporal changes of CPCs in the responsiveness to EGF and bFGF are involved in the generation of specific cell types during cortical development [9].

The EGF receptor (EGFR, also known as ErbB1 or HER1), a member of the receptor tyrosine kinase family, is expressed in the embryonic cortex, and its expression levels are low during the early period and high during the late period [8,10]. Mutant mice lacking the EGFR gene exhibit abnormal development and postnatal neurodegeneration in the

Abbreviations: EGFR, epidermal growth factor receptor; CPC, cortical progenitor cell; ERK, extracellular signal-regulated kinase; E, embryonic day; TKD, tyrosine kinase domain; DIV, day in vitro; CT-1, cardiotrophin-1; MEK, mitogen-activated protein kinase kinase; PI3K, phosphoinositide 3-kinase; MAGE, melanoma antigen; PWS, Prader-Willi syndrome.

* Corresponding author at: Laboratory of Regulation of Neuronal Development, Institute for Protein Research, Osaka University, Suita, Osaka 565-0871, Japan. E-mail address: yoshikaw@protein.osaka-u.ac.jp (K. Yoshikawa).

cerebral cortex [11,12], indicating that EGFR plays an important role in normal cortical development. EGFR is asymmetrically distributed in a subset of CPCs and differentially inherited by their daughter cells, which exhibit different responsiveness to EGF in a developmental stage-dependent manner [8]. Furthermore, EGFR promotes the differentiation and proliferation of astrocytes at late embryonic and neonatal stages of cortical development [5,7,12,13]. Thus, the temporal and spatial differences in EGFR expression contribute to the fate diversification of CPCs by changing their responsiveness to EGF during normal cortical development. However, there is limited information about regulatory mechanisms of EGFR signaling in CPCs.

Necdin is expressed abundantly in postmitotic neurons and moderately in neural precursor cells [14-17]. Necdin interacts with major nuclear proteins such as E2Fs, p53 and Sirt1 to suppress mitosis of proliferative cells and promote neuronal survival [18-21]. Necdin also interacts with the NGF receptor TrkA, another receptor tyrosine kinase, to promote differentiation and survival of sensory neurons [22]. Moreover, a study using yeast two-hybrid screening has identified necdin as one of the EGFR-interacting proteins [23]. These findings prompted us to investigate the physical and functional interactions between EGFR and necdin in CPCs.

In this study, we demonstrate that necdin interacts with EGFR in its active state to suppress the EGFR/ERK signaling pathway in primary

http://dx.doi.org/10.1016/j.cellsig.2015.11.016

0898-6568/© 2015 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/40/).

CPCs. We also show that necdin suppresses astrocyte differentiation induced by EGF-dependent activation of the EGFR/ERK signaling pathway in CPCs. Furthermore, necdin inhibits astroglial differentiation induced by overexpression of EGFR and its ligand-binding defective mutant. Because dysregulation of EGFR signaling has been proposed to contribute to the malignant transformation of astrocytes or astroglial progenitors, these observations provide insights into the molecular mechanisms underlying gliogenesis and gliomagenesis via EGFR signaling pathway under physiological and pathological conditions.

2. Materials and methods

2.1. Primary cortical progenitor cells

Neocortical tissues were dissected from mouse embryos at embryonic day (E) 14.5, incubated for 5 min at 37 °C in Ca2+/Mg2+-free glucose-supplemented HBSS with 0.05% trypsin, dissociated in DMEM supplemented with 10% fetal bovine serum, and centrifuged at 200 xg for 3 min to obtain cell pellets. Resuspended cells were incubated at 37 °C under humidified 5% CO2 conditions in CPC medium containing DMEM/F12 (Thermo Fisher Scientific), B-27 (1:50 dilution; Thermo Fisher Scientific), 14 mM sodium bicarbonate, 1 mM N-acetyl-L cysteine, 33 mM D ( + ) -glucose, 1 mg/ml bovine serum albumin, 2 mM GlutaMAX (Thermo Fisher Scientific), kanamycin/pen-icillin, and 20 ng/ml bFGF (PeproTech). CPCs were grown as floating spheres, dispersed with TrypLE (TrypLE Select; Thermo Fisher Scientific), and passaged every 48 h unless stated otherwise. Necdin gene (Ndn) mutant mice were generated and maintained as described previously [22]. Heterozygous male mice (Ndn+/-) (>25 generations in the ICR background) were crossed with wild-type female mice (Ndn+/+) (Japan SLC) to obtain wild-type (Ndn+m/+p) and paternal Ndn-deficient (Ndn+m/-p) littermates. All mice were housed in a 12 h light/dark cycle with room temperature at 23 ± 3 °C. Pregnant female mice at gestation day 14.5 were sacrificed by cervical dislocation, and embryos were collected. Genotypes of all mice were analyzed for mutated Ndn locus. The study was approved by the Animal Experiment Committee (Approval No. 24-04-0) and Recombinant DNA Committee (Approval No. 3642) of Institute for Protein Research, Osaka University, and were performed in accordance with national, institutional, and the ARRIVE guidelines.

22. Immunoblot analysis

CPCs and brain tissues were homogenized with a lysis buffer containing 10 mM Tris-HCl (pH 8.0), 150 mM NaCl, 1 mM EDTA, 1% IGEP AL CA-630 (MP Biomedicals), 10% glycerol, and protease inhibitors (Complete, Roche Diagnostics). The protein concentration was determined by the Bradford method (Bio-Rad). Proteins (10 |ag per lane) were separated by 9% SDS-PAGE, and electroblotted to polyvinylidene difluoride membranes (Immobilon, Merck Millipore). Membranes were incubated with primary antibodies against EGFR (1005; 1:100; Santa Cruz Biotechnology), necdin (NC243; 1:3000) [24], nestin (ST-1; 1:1000) [25], Y-tubulin (GTU-88; 1:1000; Sigma-Aldrich), phospho-EGFR (Tyr1068) (D7A5; 1:1000; Cell Signaling Technology), PCNA (PC10; 1:1000; Santa Cruz Biotechnology), Myc (9E10; 1:10), phospho-ERK1/2 (E10; 1:1000; Cell Signaling Technology), ERK1/2 (K-23; 1:3000; Santa Cruz Biotechnology), phospho-Akt (193H12; 1:500; Cell Signaling Technology), Akt (1:500; Cell Signaling Technology), Grb2 (C-7; 1:300; Santa Cruz Biotechnology), Sos1 (C-23; 1:1000; Santa Cruz Biotechnology), glial fibrillary acidic protein (GFAP) (1:1000; gift from Dr. Seiichi Haga, Tokyo Metropolitan Institute of Medical Science), and FLAG (M2; 1:500; Sigma-Aldrich). After incubation with peroxidase-conjugated IgGs (Cappel), the proteins were visualized by chemiluminescence method (Chemiluminescence Reagent Plus, PerkinElmer). Signal intensities were measured by densitometry and quantified using NIH ImageJ 1.46 software.

2.3. Immunocytochemistry

Primary CPCs were dispersed with TrypLE Select, plated onto 24-well plates precoated with poly-L-ornithine (Sigma-Aldrich), and incubated for 3 h. Cells were fixed with 10% formalin solution at room temperature for 20 min and permeabilized with methanol at room temperature for 20 min. For EGFR staining, cells were fixed with 4% paraformaldehyde in phosphate buffer (pH 7.4) and permeabilized with 0.02% Tween 20 in PBS at room temperature for 10 min. Fixed cells were incubated with primary antibodies at 4 °C overnight, and with secondary antibodies at room temperature for 90 min. The primary antibodies used are against EGFR (1005; 1:50), necdin (GN1; 1:1000) [22], nestin (ST-1; 1:1000), phospho-EGFR (D7A5; 1:100), GFAP (1:1000), pIII-tubulin (5G8; 1:1000; Promega), and RFP (3G5, 1:100, MBL). The secondary antibodies Alexa 488-conjugated anti-rabbit IgG (1:500), Alexa 555-conjugated anti-guinea pig IgG (1:500), and Alexa 555-conjugated anti-rabbit IgG (1:500) were purchased from Molecular Probes. Nuclear DNA was counterstained with 3.3 |aM Hoechst 33342 (Sigma-Aldrich). Images were observed with a fluorescence microscope (BX51, Olympus) equipped with charge-coupled device camera system (DP73, Olympus) or by confo-cal laser scanning microscopy (FV1000 BX61, Olympus), and processed using Adobe Photoshop CS5 software.

2.4. Coimmunoprecipitation assay

For detection of endogenous binding between necdin and EGFR, ly-sates of CPCs (1 mg protein) cultured in the CPC medium for 4 days were incubated with guinea pig anti-necdin IgG (GN1; 1:10) [22] or preimmune IgG. Bound proteins were isolated with Dynabeads protein A (Thermo Fisher Scientific) and detected by immunoblotting with antibodies against phospho-EGFR (D7A5; 1:1000), EGFR (1005; 1:100) and necdin (NC243; 1:3000). For interactions between necdin and EGFR mutants in transfected cells, HEK293A cells were transfected with combinations of expression vectors by the calcium phosphate method and harvested after 24 h. Cell lysates (150 |ag protein) were incubated at 4 °C for 2 h with antibodies against Myc (9E10; 1:4) and necdin (NC243; 1:100), pelleted with Protein A-Sepharose (GE Healthcare), separated by 9% SDS-PAGE, and detected by immunoblot-ting. Full-length mouse EGFRcDNA (NCBI NM 207655.2) was synthesized by RT-PCR from mRNA expressed in CPCs. cDNAs encoding EGFR deletion mutants and its point mutants were generated using synthetic oligonucleotide primers based on their sequence information (NCBI NM 207655.2) and subcloned into 6xMyc-pcDNA3.1 +. For interactions of EGFR with Grb2 and Sos1, primary CPCs prepared from E14.5 mice were cultured for 4 days and treated with EGF (10 ng/ml) for 5 min. CPC lysates (500 |ag protein) were incubated with antibodies to Grb2 (C-7; 1:100) and Sos1 (C-23; 1:100). Bound proteins were pelleted with Protein A-Sepharose and detected by immunoblotting with the antibody to EGFR (1005; 1:100).

2.5. In vitro binding assay

The tyrosine kinase domain (TKD) of EGFR and its deletion mutants were generated using synthetic oligonucleotide primers based on their sequence information (NCBI NM 207655.2) and subcloned into pMAL-C2 vector to make maltose binding protein (MBP) fusion proteins. MBP-fused proteins were affinity-purified with amylose resin (New England Biolabs) and incubated with His-tagged necdin (200 ng) at 4 °C for 30 min in 0.5 ml of the binding buffer containing 20 mM Tris-HCl (pH 7.5), 200 mM NaCl, and 1 mM EDTA [21]. After washing, bound His-tagged necdin was eluted with 20 mM maltose and detected by immunoblotting with anti-necdin antibody. MBP fusion proteins were detected by Coomassie Brilliant Blue staining.

2.6. Cell proliferation assay

Primary CPCs were prepared from E14.5 mice and incubated for 48 h. Cells (2 x 105 cells) were replated in 35-mm dishes, incubated in the presence or absence of 20 ng/ml EGF for another 48 h, and harvested for manual cell counting. For EdU incorporation, CPCs at 4 day in vitro (DIV) were plated onto poly-L-ornithine-coated coverslips. Cells were cultured in the CPC medium supplemented with 10 ng/ml EGF for 24 h and fixed with 10% formalin solution at room temperature for 20 min, and permeabilized with methanol at room temperature for 20 min. EdU (Invitrogen, A10044; 10 pM) was added to the CPC medium 4 h before fixation. Fixed cells were incubated for 30 min with 3.3 pM Hoechst 33342 (Sigma-Aldrich) and then with EdU detection cocktail (A10044, Invitrogen) at room temperature for 15 min.

2.7. Cell differentiation assay

For astrocyte differentiation assay, CPCs cultured for 48 h in the CPC medium were incubated in the presence and absence of 20 ng/ml EGF for 48 h. CPCs were dissociated and plated onto poly-L-ornithine-coated coverslips in the CPC medium supplemented with 50 ng/ml cardiotrophin-1 (CT-1) (PeproTech) for 48 h. GFAP-immunopositive cells were detected by immunocytochemistry as above. For neuronal differentiation assay, CPCs cultured for 48 h in the CPC medium were incubated in the presence and absence of 20 ng/ml EGF for 48 h. CPCs were dissociated, plated onto poly-L-ornithine-coated coverslips and cultured in the CPC medium deprived of EGF and bFGF for 24 h. ßlll tubulin-expressing cells were detected by immunocytochemistry as above.

2.8. Kinase inhibitor assay

CPCs were cultured in the CPC medium for 48 h, and EGF (10 ng/ml) in the presence of the EGFR tyrosine kinase inhibitor gefitinib (5 |aM; Cayman Chemical), mitogen-activated protein kinase kinase (MEK) inhibitor U0126 (20 |aM; Merck Millipore) and phosphoinositide 3-kinase (P13K) inhibitor LY294002 (20 pM; Merck Millipore) were added to the CPC medium 30 min before treatment with EGF. CPCs were treated with EGF (10 ng/ml) for 5 min and harvested for immuno-blot analysis of phospho-EGFR, phospho-ERK, and phospho-Akt. For GFAP expression assay, CPCs were treated with EGF and the kinase inhibitors for 6 h, replated in 35-mm dish precoated with poly-L-ornithine, and treated with CT-1 (50 ng/ml) for 48 h. Expression of GFAP was analyzed by immunoblotting.

2.9. In vitro electroporation

CPCs were prepared from E14.5 mice and cultured for 48 h. CPCs were centrifuged and resuspended at 5 x 106 cells in 100 |al of Opti-MEM containing expression vectors (total DNA, 20 |ag). Electric pulses (pore pulse; 125 V/10 msec x 1: transfer pulse; 20 V/50 msec/50 msec interval; 10 cycles) were applied to the cell suspension in 2-mm gap cuvette using a pulse generator (CUY21-ED1T II, BEX). cDNA encoding EGFR lacking residues 6-273 (eGFRANT) was generated using synthetic oligonucleotide primers based on their sequence information (NCBI NM 207655.2) and subcloned into 6xMyc-pcDNA3.1 +. cDNAs for EGFR, EGFR mutants (KR and ANT), FLAG-tagged mouse necdin, necdinAEB (residues 144-184 deletion mutant) [26], FLAG-tagged human MAGEA1 [27], and td-Tomato (ptdTomato-N1, Clontech) were transfected into primary CPCs. Transfection efficiency was ~70% as analyzed 12 h after transfection

Fig. 1. EGFR and necdin are coexpressed in primary CPCs. (A) CPCs were prepared from the neocortex at E14.5 and cultured in the presence of bFGF (bFGF+) for the indicated durations. Expression of EGFR, necdin, nestin, and Y-tubulin (Y-Tub) was analyzed by immunoblotting. (B) CPCs were immunostained for EGFR (green), necdin (red), and nestin (green) and observed by confocal microscopy. Nuclear DNA was counterstained with Hoechst33342 (blue) and merged with confocal images. (C) CPCs were treated with EGF for the indicated durations. Expression of phospho-EGFR (pEGFR) and total EGFR (EGFR) was analyzed by immunoblotting or (D) immunocytochemistry. Confocal images (green) and nuclear DNA stain (blue) are merged. Scale bars; 10 |jm (in B) 5 |jm (in D).

by fluorescence immunocytochemistry for td-Tomato expression. Expression of transfected cDNAs was analyzed by immunoblotting 12 h after electroporation. For GFAP expression, transfected CPCs were cultured for 24 h in the CPC medium and treated with CT-1 (50 ng/ml) for 48 h. GFAP expression of transfected CPCs was analyzed by immunoblotting and fluorescence immunocytochemistry using antibodies to GFAP and RFP for td-Tomato.

2.10. Statistics

Statistical significance was tested using an unpaired Student's t test or one-way ANOVA followed by Tukey-Kramer post hoc test. A significance of P <0.05 was required for rejection of the null hypothesis.

3. Results

3.1. EGFR expression is induced by bFGF in primary CPCs

We first analyzed the expression levels of EGFR and necdin in primary CPCs prepared from the mouse cortex at E14.5 by immunoblotting (Fig. 1A). When CPCs were treated with bFGF for up to 4 DIV, the EGFR levels increased in CPCs incubated in the presence of bFGF for 2 days or more, whereas necdin was expressed in a constitutive manner. The neural stem/progenitor cell marker nestin was also expressed in these CPCs. Immunocytochemical analysis showed that EGFR was hardly detectable in CPCs at 1 DIV but was clearly detected at the plasma membrane and in the cytoplasm at 4 DIV, whereas necdin was predominantly cytoplasmic at 1 and 4 DIV (Fig. 1B). Thus, we used CPCs cultured

EGF- EGF+ EGF- EGF+

Ndn WT WT КО WT КО

Fig. 2. Autophosphorylated EGFR and necdin are colocalized in primary CPCs. (A, B) Immunocytochemistry. CPCs at 4 DIV were treated with (EGF+) or without (EGF—) EGF for 5 min. EGFR (green) (A) or phospho-EGFR (pEGFR) (green) (B) and necdin (red) were immunostained and observed by confocal microscopy. Immunostained images and nuclear DNA stain (blue) are merged (Merge) for colocalization (yellow). (C) Three-dimensional analysis. CPCs were double immunostained for pEGFR (green) and necdin (red), and observed by confocal microscopy. Multiple z-stack images in XY,XZ and YZ axes are shown. Scale bars; 10 |jm (in A, B), 5 |jm (inC). (D) Coimmunoprecipitation assay. CPCs were prepared from wild-type (WT) and necdin-null (KO) mice and treated at 4 DIV with EGF (EGF+) or without (EGF—) EGF for 5 min. Cell lysates were immunoprecipitated with anti-necdin IgG (aNecdin IgG) or preimmune IgG (Pre IgG). pEGFR, EGFR, and necdin were analyzed by immunoblotting. Lysate, CPC lysate (10 |g).

in the bFGF-supplemented CPC medium for 4 days or more in the following experiments.

We then analyzed autophosphorylation of EGFR in EGF-stimulated CPCs. Autophosphorylation of EGFR (phospho-EGFR), which was detected with a site-specific anti-phosphotyrosine antibody against pY1068, increased markedly in CPCs 5 min after EGF treatment (Fig. 1C). The levels of the total EGFR and phospho-EGFR levels decreased at 120 min. Consistent with the immunoblot results, the phospho-EGFR immunoreactivity increased in the cytoplasm at 5 and 30 min, and both phospho-EGFR and total EGFR levels were reduced at 120 min (Fig. 1D). These results suggest that autophosphorylated EGFR undergoes degradation in primary CPCs.

3.2. Necdin and phosphorylated EGFR colocalize in the cytoplasm of CPCs

We investigated whether EGFR and necdin are colocalized in primary CPCs (Fig. 2A). Immunocytochemistry showed that EGFR was localized at the plasma membrane and in the cytoplasm of CPCs. When CPCs were treated with EGF, EGFR was accumulated in the cytoplasm, and the EGFR immunoreactivity overlapped partially with the necdin immunoreactivity. We also analyzed the colocalization of autophosphorylated EGFR and necdin in CPCs (Fig. 2B). The phospho-EGFR immunoreactivity was hardly detected in unstimulated CPCs but increased appreciably in the cytoplasm of EGF-treated CPCs, where it overlapped partially with that of necdin. Furthermore, three-

Fig.3. Necdin interacts with the tyrosine kinase domain of EGFR. (A) Diagrams ofEGFRandits C-terminal-truncated mutants. WT, wild-type; KR, K723R mutant; AC126, C-terminal 126 residue deletion; AC240, C-terminal 240 residue deletion; AC497, C-terminal 497 residue deletion; TM, transmembrane domain; TKD, tyrosine kinase domain; Y, autophosphorylated tyrosine residues. (B) HEK293A cells were co-transfected with cDNAs for Myc-tagged EGFR (WT, KR) and necdin. Cell lysates were immunoprecipitated (IP) with antibodies to Myc and necdin, and immunoblotted (IB) for necdin, Myc, and Y-tubulin (Y-Tub). Results are shown in A. (C) Diagrams of TKD deletion mutants. TKD-WT, wild-type TKD; TKD-KR K723R mutant TKD; TKDAC91, C-terminal 91 residue deletion; TKDAC30, C-terminal 30 residue deletion; TKD-C91, C-terminal 91 residues; TKD-C30, C-terminal 30 residues; PL, phosphate-binding loop; CH, C-helix; AL, activation loop. (D) Coimmunoprecipitation assay for TKD-deletion mutants. HEK293A cells were co-transfected with cDNAs for Myc-tagged TKD mutants and necdin. Cell lysates were immunoprecipitated (IP) with antibodies to Myc and necdin and immunoblotted (IB) for necdin and Myc. Results are shown in C. (E) In vitro binding assay. Purified maltose-binding protein (MBP)-fused EGFR deletion mutants immobilized on amylose resin were incubated with His-tagged necdin (His-Necdin). Bound His-necdin was detected by immunoblotting with anti-necdin antibody (upper panel). MBP-EGFR TKD mutants were separated by 7.5% SDS-PAGE and stained with Coomassie Brilliant Blue (lower panel).

dimensional analysis of confocal images demonstrated that phospho-EGFR and necdin were colocalized under the plasma membrane (Fig. 2C). We then analyzed the interaction between endogenous necdin and EGFR in CPCs by coimmunoprecipitation assay (Fig. 2D). Phospho-EGFR and total EGFR were coprecipitated with necdin in the lysate of wild-type CPCs treated with EGF. Although expression levels of total EGFR were similar between EGF-treated and untreated CPCs, necdin failed to interact with EGFR in wild-type CPCs without EGF

treatment. This indicates that necdin binds only to phospho-EGFR. Together, these results suggest that endogenous necdin and phospho-EGFR form a stable complex in EGF-stimulated CPCs.

33. Necdin interacts with the tyrosine kinase domain of EGFR

To characterize the interaction between necdin and EGFR, we performed the coimmunoprecipitation assay using transfected HEK293A

Fig. 4. Necdin suppresses the EGFR/ERK signaling pathway in CPCs. (A, B) EGF-induced phosphorylation of EGFR, ERK, and Akt. CPCs prepared from wild-type (Ndn WT) and necdin-null (Ndn KO) mice at E14.5 were cultured for 4 days and treated with EGF (10 ng/ml) for the indicated durations. CPC lysates were immunoblotted for phospho-EGFR (pEGFR), EGFR, phospho-ERK (pERK), ERK, phospho-Akt (pAkt), Akt, necdin, and Y-tubulin. pEGFR, pERK and pAkt levels were quantified by densitometry and normalized to those of EGFR, ERK and Akt, respectively, and then to Y-tubulin (B). Each value represents the mean ± SEM, n = 3 (pEGFR), n = 4 (pERK, pAkt). P-values were calculated using one-way ANOVA followed by Tukey-Kramer post hoc test. (C, D) Interactions of EGFR with Grb2 and Sosl in EGF-stimulated CPCs. CPCs at 4 DIV were treated with (EGF+) or without (EGF-) EGF for 5 min. CPC lysates were immunoprecipitated (IP) with antibodies to Grb2 and Sosl and immunoblotted (IB) for EGFR, Grb2, Sosl, pEGFR, pERK, ERK, necdin, and Y-tubulin (C). Signal densities of EGFR coprecipitated with Grb2 and Sosl were normalized to those of Grb2 and Sosl, respectively, and then to Y-tubulin (D). Each value represents the mean ± SEM (Grb2, n = 6; Sosl, n = 4). P-values were calculated using Student's t-test

cells. In mouse EGFR cDNA-transfected HEK293A cells, overexpressed EGFR underwent ligand-independent autophosphorylation (Fig. S1A). To validate the assay system, we tested the interactions of necdin with wild-type EGFR and its kinase dead (K723R) mutant using the adaptor protein Grb2 as a positive control (Fig. S1B). Both necdin and Grb2 bound preferentially to wild-type EGFR, suggesting that necdin binds to EGFR only in its active state.

We then examined the interactions of necdin with EGFR deletion mutants that lack 126 and 240 amino acid residues of the C-tail containing autophosphorylated tyrosine residues (Fig. 3A). Necdin was expressed in transfected HEK293A cells as major 42 kDa and minor 37 kDa proteins (the minor protein is presumably a degraded product of the 42 kDa protein). In this assay, necdin bound to wild-type EGFR (WT) and an EGFR mutant lacking C-terminal 126 residues (WTAC126) (Fig. 3B). Notably, necdin strongly bound to the deletion mutants WTAC240 and KRAC240 lacking the entire C-terminal 240 residues. These results suggest that the C-tail of EGFR interferes with the interaction between necdin and EGFR. Necdin failed to interact with an EGFR mutant (AC497) lacking both the C-tail and the tyrosine kinase domain (TKD) (residues 690-946), indicating that necdin binds to the TKD.

To determine the necdin-binding region of the TKD, we constructed Myc-tagged TKD deletion mutants (Fig. 3C). Necdin bound to wild-type TKD and the TKD KR mutant, but not to wild-type TKDAC91 lacking the C-terminal 91 residues in the C-lobe (Fig. 3D). In contrast, necdin

strongly interacted with the C-terminal 91 residues (C91). The necdin-binding region was narrowed down to the C-terminal 30 residues of the TKD (TKD-C30). To test whether necdin directly binds to the TKD-C30 region, we performed in vitro pull-down assay using MBP-fused wild-type EGFR TKD (WT), TKD mutants (TKDAC30 and TKD-C30), and His-tagged necdin (Fig. 3E). Consistent with the results of coimmunoprecipitation assay, His-tagged necdin bound to wild-type TKD and TKD-C30, but not to TKDAC30, indicating that necdin binds directly to the C-terminal region (residues 917-946) of the TKD.

3.4. Necdin suppresses phosphorylation of ERK via inhibition of the interaction between EGFR and Grb2

To examine whether endogenous necdin affects the tyrosine kinase activity, we analyzed EGF-induced autophosphorylation of EGFR in necdin-null CPCs (Fig. 4A, B). Immunoblot analysis showed that phospho-EGFR levels were not significantly changed in necdin-null CPCs (Fig. 4A, top panel; Fig. 4B, top graph). We also examined the effects of necdin on the EGFR signaling pathways by analyzing the phosphorylation levels of ERK and Akt. In this analysis, minor and major ERK bands at 44 kDa (ERK1) and 42 kDa (ERK2) were detected. Notably, phospho-ERK levels in necdin-null CPCs increased ~ 2-fold at 5 min in response to EGF (Fig. 4A, 3rd panel; Fig. 4B, middle graph). In contrast, phospho-Akt levels were not significantly changed in necdin-null CPCs (Fig. 4A, 5th panel, Fig. 4B, bottom graph).

Fig. 5. Necdin suppresses EGF-promoted proliferation of CPCs. (A, B) Total CPC count. CPCs prepared from wild-type (Ndn WT) and necdin-null (Ndn KO) E14.5 mice were cultured for 48 h, replated at 2 x 105 cells per 35-mm dish, and cultured in the presence (EGF+) or absence (EGF-) of EGF for another 48 h (A). Cells were harvested for manual cell counting (n = 4) and presented as fold change (B). (C-E) EdU incorporation assay. CPCs were cultured for 96 h, replated onto poly-L-ornithine-precoated 24-well plates, cultured in the presence (EGF+) or absence (EGF-) of EGF for 24 h, and fixed for fluorescence microscopy (D). EdU was added to the medium 4 h before fixation, and EdU incorporated into nuclear DNA was chemically stained (green, arrowheads). EdU+ cells were counted (each > 190 DNA-stained cells examined, n = 3), and theEdU+ population is presented as percentage (E). P-values were calculated using one-way ANOVA followed by Tukey-Kramer post hoc test. NS, not significant at P >0.05.

To investigate whether necdin affects the EGFR/RAS/ERK signaling pathway, we examined the interactions of EGFR with the adaptor protein Grb2 and the Grb2-binding RAS activator Sos1 in wild-type and necdin-null CPCs (Fig. 4C). The amount of EGFR coprecipitated with Grb2 or Sos1 increased markedly (~2.5-fold) in necdin-null CPCs treated with EGF (Fig. 4C, top and 3rd panels, Fig. 4D). Grb2 failed to interact with necdin as analyzed by coimmunoprecipitation using transfected HEK293A (Fig. S2). These results suggest that necdin suppresses the interaction between EGFR and Grb2 to downregulate the RAS/ERK signaling in CPCs.

3.5. Necdin suppresses proliferation of EGF-responsive CPCs

Because the EGFR/ERK pathway is involved in the control of cellular proliferation, we next examined whether necdin affects the proliferation rate of primary CPCs (Fig. 5A, B). When CPCs were treated with EGF for 48 h, the total cell number of CPCs increased significantly (1.37 times the wild-type control level) in necdin-null CPCs, whereas there was no significant difference in the absence of EGF. We also used EdU incorporation assay to determine the proliferation rate of CPCs prepared from wild-type and necdin-null mice (Fig. 5C-E). The

Fig. 6. Necdin suppresses EGF-promoted astrocyte differentiation of CPCs. (A-C) Astrocyte differentiation assay. CPCs prepared from wild-type (Ndn WT) and necdin-null (Ndn KO) mice at E14.5 were cultured for 48 h, treated with (EGF+) or without (EGF-) EGF for 48 h, and then with CT-1 for 48 h (A). CPCs were immunostained for GFAP (green) (B), andGFAP+ cells were counted (each > 170 cells examined, n = 3) (C). (D-F) Neuronal differentiation assay. CPCs were cultured for 48 h, treated with (EGF+) or without (EGF-) EGF for 48 h and incubated in the medium deprived of EGF and bFGF for 24 h (D). CPCs were immunostained for JIII-tubulin (red) (E), and (JIII-tubulin+ cells were counted (each >300 cells examined, n = 3) (F). Immunostained images are merged with nuclear DNA stain (B, E). Scale bars, 50 ^m (in B, E). Each value represents the mean ± SEM (in C, F). P-values were calculated using one-way ANOVA followed by Tukey-Kramer post hoc test. NS, not significant at P >0.05.

population of EdU+ S-phase cells in necdin-null CPCs increased significantly (1.35 times the wild-type control level). These results suggest that endogenous necdin suppresses EGF-stimulated proliferation of primary CPCs through inhibition of the EGFR/ERK pathway.

3.6. Necdin suppresses astrocyte differentiation in EGF-responsive CPCs

CPCs differentiate into astrocytes in response to the interleukin-6 (IL-6) family of cytokines such as CNTF and LIF [28-30]. EGFR regulates the competence of CPCs to interpret LIF as an astrocyte-inducing signal [13]. To investigate whether necdin affects astrocyte differentiation of CPCs by suppressing EGFR signaling, we used cardiotrophin-1 (CT-1), one of the IL-6 family cytokines [31,32], to induce astrocyte differentiation of CPCs (Fig. 6A-C). The number of GFAP+ astrocytes differentiated from necdin-null CPCs increased by 39% in the presence of EGF and CT-1, whereas the GFAP+ cell population was not significantly changed in necdin-null CPCs in the absence of EGF. These results indicate that endogenous necdin suppresses EGF-promoted glial differentiation via inhibition of the EGFR/ERK signaling.

In this analysis, we found that the GFAP immunoreactivity increased appreciably in each GFAP-expressing cell differentiated from necdin-null CPCs (Fig. S3A). Thus, we analyzed the expression levels of the GFAP protein by immunoblotting (Fig. S3B). EGF markedly increased

expression levels of the GFAP protein in necdin-null CPCs treated with CT-1, indicating that the GFAP protein level correlates well with the extent of astrocyte differentiation.

We also examined neuronal differentiation of EGF-treated CPCs by withdrawing both bFGF and EGF (Fig. 6D-F). EGF markedly reduced the population of (JIII-tubulin+ neurons, suggesting that EGF prevents neuronal differentiation of CPCs. However, there was no significant difference in the number of differentiated (JIII-tubulin+ neurons between wild-type and necdin-null CPCs. These observations suggest that necdin specifically suppresses EGF-promoted as-trocyte differentiation of CPCs.

To investigate whether necdin-null mice express high GFAP levels in the cortex in vivo during the neonatal period, we analyzed cortical GFAP levels at postnatal day 4 by immunoblotting. We found no significant difference in the GFAP level between wildtype and necdin-null mice (Fig. S4A, B). Intriguingly, EGF mRNA levels in necdin-null cortex in vivo were reduced by 54% at E18.5, a late stage of cortical development when astrocyte differentiation occurs (Fig. S4C). Expression of EGF mRNA in primary CPCs was also reduced by 50% (Fig. S4D). These results suggest that necdin-null CPCs express low levels of EGF mRNA to prevent astrocyte hyperproliferation due to enhanced EGFR signaling via a negative feedback regulation.

Fig. 7. Necdin suppresses EGF-promoted GFAP expression in CPCs. (A) Effects of kinase inhibitors on EGFR signaling pathways. CPCs were incubated for 48 h and treated with EGF for 5 min. EGFR tyrosine kinase inhibitor (ETKI, gefitinib; 5 |jM), MEK inhibitor (MEKI, U0126; 20 |jM) and PI3K inhibitor (PIKI, LY294002; 20 |jM) were added to the medium 30 min before EGF treatment. Phospho-EGFR (pEGFR), phospho-ERK (pERK), and phospho-Akt (pAkt) levels in CPCs were analyzed by immunoblotting. (B, C) GFAP expression in CT-1-treated CPCs. CPCs were cultured for 48 h, treated with EGF (EGF+) or without (EGF-) EGF and kinase inhibitors for 6 h, and incubated with CT-1 for 48 h (B). GFAP expression in CPCs was analyzed by immunoblotting (C). (D, E) Effects of kinase inhibitors on GFAP levels in necdin-null CPCs. CPCs prepared from wild-type (Ndn WT) and necdin-null (Ndn KO) E14.5 mice were treated with EGF and kinase inhibitors as in B. Expression of GFAP, necdin, and Y-tubulin (Y-Tub) was analyzed by immunoblotting (D) and quantified by densitometry (E). GFAP levels were normalized to Y-tubulin levels. Each value represents the mean ± SEM, n =4. P-values were calculated using one-way ANOVA followed by Tukey-Kramer post hoc test. NS, not significant at P >0.05.

3.7. Inhibition of EGFR/ERK signaling blocks astrocyte differentiation enhanced in necdin-null CPCs

To determine whether EGF-promoted astrocyte differentiation is mediated by the EGFR signaling pathway in CPCs, we used the EGFR tyrosine kinase inhibitor (gefitinib), MEK1/2 inhibitor (U0126), and PI3K inhibitor (LY294002). We first examined the effects of these inhibitors on EGF-induced phosphorylation levels of EGFR, ERK and Akt by immu-noblot analysis (Fig. 7A). Autophosphorylation of EGFR was detected 5 min after EGF stimulation in CPCs treated with the MEK inhibitor and the PI3K inhibitor but not in those treated with the EGFR tyrosine kinase inhibitor. As expected, phosphorylation of ERK and Akt was hardly detected in EGF-stimulated CPCs treated with the inhibitors of MEK and PI3K, respectively.

We then analyzed the effects of these kinase inhibitors on astrocyte differentiation by immunoblot assay. CPCs were treated with EGF in the presence or absence of these inhibitors for 6 h and then with CT-1 for 48 h (Fig. 7B). The GFAP levels in EGF-stimulated CPCs increased appreciably when incubated in the absence of the inhibitors or in the presence of the PI3K inhibitor (Fig. 7C). These observations indicate that EGF-promoted astrocyte differentiation is mediated by the EGFR/ERK signaling pathway in CPCs.

To test whether the EGFR/RAS/MEK/ERK pathway mediates EGF-promoted enhancement of the GFAP levels in necdin-null CPCs, we treated CPCs with EGF and these kinase inhibitors (Fig. 7D, E). The EGF-dependent increase in the GFAP levels was significantly enhanced (1.54 times the wild-type control level) in necdin-null CPCs, and this enhancement was completely blocked by EGFR tyrosine kinase inhibitor (gefitinib) or MEK1/2 inhibitor (U0126), but not by PI3K inhibitor (LY294002). These results suggest that necdin suppresses astrocyte differentiation through attenuation of the EGFR/RAS/MEK/ERK signal transduction.

3.8. Necdin antagonizes astrocyte differentiation promoted by transient overexpression of EGFR in CPCs

Sustained activation of the EGFR/RAS/ERK pathway is involved in astrocyte proliferation and pathogenesis of gliomas [33,34]. Thus, we investigated whether astrocyte differentiation of primary CPCs is promoted by overexpression of EGFR and an EGFR mutant (EGFRANT) lacking the N-terminal ligand-binding domain (L1), which is equivalent to human glioblastoma-associated EGFR variant III (EGFRvIII) (Fig. 8A). CPCs were transiently transfected with cDNAs encoding wild-type EGFR, kinase-dead K723R mutant, and EGFRANT by electroporation. Autophosphorylation of wild-type EGFR and EGFRANT mutant, but not that of the KR mutant, was detected 12 h after transfection by immunoblot analysis of phospho-EGFR (Fig. 8B).

We analyzed whether forced expression of EGFR and EGFR mutants enhances astrocyte differentiation. Transiently transfected CPCs were incubated for 24 h and then treated with CT-1 for 48 h (Fig. 8C). Immu-nocytochemistry showed that overexpression of wild-type EGFR or EGFRANT increased the GFAP immunoreactivity in transfected td-Tomato+ CPCs (Fig. 8D). These cells exhibited morphological characteristics of differentiated astrocytes such as extended processes and larger cell bodies. Consistent with these morphological changes, GFAP expression levels increased markedly in CPCs transfected with wild-type EGFR or EGFRANT as analyzed by immunoblotting (Fig. 8E).

We then examined whether necdin prevents EGFR- or EGFRANT-promoted astrocyte differentiation of CPCs (Fig. 8F, G). We found that a necdin mutant lacking residues 144-184 located in the MAGE homology domain (necdinAEB) [26] failed to interact with EGFR (Fig. S5). Thus, we used this EGFR binding-defective mutant as a negative control. Necdin strongly decreased the GFAP level to near basal level of td-Tomato+ control in EGFR-overexpressing CPCs. Similarly, necdin reduced the GFAP level in EGFRANT-overexpressing CPCs. In contrast,

necdinAEB had no antagonizing effects. In this analysis, neither necdin nor necdinAEB affected the basal GFAP level in transfected CPCs.

We analyzed whether human MAGEA1, a MAGE family member expressed in many cancers, affects EGFR-mediated astrocyte differentiation induced by EGFR overexpression. MAGEA1 bound to both EGFR and kinase-dead mutant (Fig. S6A). In contrast to necdin, MAGEA1 enhanced EGFR-induced GFAP expression, and coexpression of necdin strongly antagonized the effect of MAGEA1 (Fig. S6B, C). These results suggest that necdin and MAGEA1 exert opposite effects on the EGFR signaling.

We also examined the GFAP expression levels in wild-type and necdin-null CPCs overexpressing EGFR (Fig. 8H, I). The GFAP levels increased by 57% in necdin-null CPCs overexpressing EGFR. The increase in the GFAP level was completely suppressed by coexpression of necdin, but not of necdinAEB. These results suggest that endogenous necdin in CPCs acts as an intrinsic suppressor of astrocyte differentiation induced by EGFR overexpression,

4. Discussion

EGFR signaling in CPCs plays an important role in astrocyte differentiation during the late embryonic period when the transition from neurogenesis to gliogenesis occurs in developing cortex [9]. Furthermore, the RAS/ERK signaling pathway is involved in astrocyte differentiation during brain development [35-38]. These findings indicate that the RAS/MEK/ERK signaling pathway is involved in the control of astrocyte differentiation during cortical development. Necdin suppresses the EGFR/ERK signaling pathway by interacting with autophosphorylated EGFR in EGF-responsive CPCs. Thus, we propose that necdin is an intrinsic suppressor of the EGF/EGFR signaling in CPCs to restrict astrocyte differentiation during the late period of cortical development.

Stimulation of EGFR by EGF induces autophosphorylation of the C-tail followed by activation of downstream signaling pathways including the RAS/MEK/ERK and PI3K/Akt cascades [39]. Necdin suppresses the EGFR/ERK signal transduction by interacting with the C-terminal region of the TKD. Necdin interacts with the TKD of EGFR only in its activated state, and the interaction is enhanced in the absence of the C-tail. This suggests that the unphosphorylated C-tail interferes with the interaction between necdin and the TKD. This idea is supported by the previous observations on the conformational changes of the C-tail and TKD in inactive and active states of EGFR [40]. We speculate that necdin binds to the TKD when autophosphorylation of the C-tail uncovers the necdin-binding site on the TKD. Thus, the C-tail may serve as a molecular switch to control the interaction between necdin and the TKD in CPCs (Fig. 9).

The TKD of EGFR consists of N-terminal (N-lobe) and C-terminal (C-lobe) regions, and the N-lobe contains the specific sites for the tyrosine kinase activity such as ATP-binding site, C-helix, and phosphate-binding loop (P-loop) [41,42]. Necdin binds to the C-terminal end of the C-lobe, and this location may be critical for the reduced interaction between autophosphorylated C-tail and Grb2 without affecting the tyrosine ki-nase activity. Intriguingly, the EGFR feedback inhibitor MIG6 (also known as RALT or ERRFI1), like necdin, binds to the TKD when EGFR is activated [43]. However, MIG6, unlike necdin, inhibits the tyrosine kinase activity of EGFR to suppress both RAS/ERK and PI3K/Akt pathways [44]. This may be attributable to the structural characteristics of MIG6 associated with the allosteric control of the EGFR kinase activity [43, 45]. The differences between necdin and MIG6 imply that necdin is a unique inhibitor of the EGFR/RAS/ERK signaling pathway linked to cellular proliferation but not of the PI3K/Akt signaling pathway linked to cell survival. These findings are consistent with the notion that necdin suppresses cell proliferation and maintains cell survival.

Neural precursor cells in the embryonic forebrain of necdin-null mice are highly proliferative in vivo during the early period of development when neurogenesis occurs actively [16,17]. Necdin suppresses the proliferation of CPCs by increasing p16 expression through interaction with the p16 transcriptional repressor Bmi1 and by reducing Cdk1

expression through interaction with the Cdk1 transcriptional activator E2F1 [17], suggesting that necdin prevents hyperproliferation of early CPCs, which are competent to differentiate into neurons, by modulating the activities of these cell cycle regulatory proteins. The present study provides evidence that necdin suppresses astrocyte generation from EGF-responsive CPCs through attenuation of the EGFR/ERK pathway.

Thus, necdin may control the EGFR/ERK signaling pathway in late CPCs, which are competent to differentiate into astrocytes. We propose that necdin fine-tunes the regulatory systems involved in neurogenesis and gliogenesis at different stages of normal cortical development.

Numerous studies have indicated that EGFR signaling is central to the pathogenesis of many cancers in which genetic alterations of EGFR

Fig. 9. A schematic model for suppression of EGFR signaling by necdin. Necdin binds to EGF-activated EGFR via the TKD C-lobe and blocks the interaction between EGFR and Grb2, resulting in the suppression of the RAS/ERK signaling pathway. For details, see Discussion. ICD, intracellular domain; TKD, tyrosine kinase domain; P, phospho-tyrosine; Ndn +, wild-type necdin; Ndn —, necdin deficiency.

such as gene amplification and mutations are involved [46]. Furthermore, the importance of EGFR signaling in cancer progression is supported by the fact that several anticancer drugs such as gefitinib and erlotinib target the EGFR TKD [47]. The present study has clarified that necdin acts as an endogenous suppressor of EGFR signaling. Furthermore, necdin is not expressed in many cancer cells, where the necdin gene is hypermethylated [48]. These findings raise the possibility that necdin potentially suppresses malignant transformation induced by enhanced EGFR signaling in various types of cancers. As shown in this study, necdin suppresses EGFR overexpression-induced astrocyte differentiation, whereas MAGEA1, another MAGE family member expressed in many kinds of malignant tumors [49], promotes it, suggesting that necdin and MAGEA1 exert opposite effects on EGFR-mediated signaling events. Thus, it is tempting to speculate that these different types of MAGE family members antagonistically control malignant transformation induced by EGFR or its mutants.

Dysregulation of EGFR signal transduction has been suggested to contribute to the etiology of brain tumors including gliomas [33]. Aberrant activation of the RAS/MEK/ERK pathway is also involved in the pathogenesis of astrocytoma or glioma [34,38]. On the other hand, necdin is not expressed in many cell lines derived from neural cell-derived tumors such as glioma, neuroblastoma, pheochromocytoma, and ependymoma [25,50,51]. A network modeling study of DNA copy

number aberrations implicates necdin in suppressing cell growth of glioblastomas [52]. Furthermore, reduced NDN expression in low grade gliomas strongly correlates with reduced overall survival of patients with gliomas [53]. The gene overexpression experiments demonstrated that necdin strongly prevents astrocyte differentiation induced by overexpression of wild-type EGFR or its mutant equivalent to human EGFR variant III (EGFRvIII) in CPCs. Overexpression of EGFR and its deletion mutants such as EGFRvIII is commonly found in glioblastomas [33]. Gliomas are proposed to arise from neural stem cells that are generated from glial progenitor cells or astrocytes via reprogrammed processes [34,54]. Thus, we infer that necdin serves as a tumor suppressor that controls the EGFR/ERK pathway in astrocyte progenitors to prevent their malignant transformation.

The human necdin gene (NDN) is located in chromosome 15q11.2-q12 [55], a region deleted in Prader-Willi syndrome (PWS), a typical genomic imprinting-associated neurodevelopmental disorder. NDN is maternally imprinted, expressed only from the paternal allele, and not expressed in individuals with PWS [56,57]. Although patients with PWS exhibit symptoms due to hypothalamic abnormalities such as hy-perphagia and hypogonadism, there is limited information on neuro-pathological lesions in PWS. In contrast, the mouse necdin gene (Ndn) located in chromosome 7C is also imprinted, and paternal Ndn-mutant mice display failure-to-thrive and early neonatal lethality [58,59].

Fig. 8. Necdin antagonizes astrocyte differentiation induced by EGFR overexpression in CPCs. (A) Diagrams of EGFR and its mutants. WT, wild-type EGFR; KR EGFR K723R mutant; ANT, EGFRN-terminal (residues 6-273) deletion mutant; L1/2, ligand-binding domain 1/2 (LigandBD); TM, transmembrane domain; Necdin BD, necdin-binding domain; TKD, tyrosine kinase domain; CT, C-terminal domain. (B) Autophosphorylation of EGFR and ANT mutant. CPCs were transiently transfected with wild-type EGFR, KR mutant and ANT mutant by electropora-tion, and expression of phospho-EGFR (pEGFR) and EGFR in transfected CPCs was analyzed 12 h after transfection by immunoblotting. Tmt, td-Tomato vector control. (C-E) GFAP expression in EGFR-overexpressing CPCs. CPCs were transiently transfected with cDNAs for wild-type EGFR and its mutants, cultured for 24 h, and treated with CT-1 for 48 h (C). GFAP expression in transfected CPCs was analyzed by triple staining of GFAP (green), td-Tomato (red), and nuclear DNA (blue) (D) and immunoblotting (E). Triple-stained images are merged for GFAP expression in td-Tomato+ cells (yellow in D). Arrowheads point to td-Tomato+ CPCs. Scale bar, 20 |jm. (F, G) Effect of necdin on EGFR-induced GFAP expression. CPCs were transiently transfected with combinations of cDNAs for wild-type EGFR, EGFRANT, FLAG-necdin (Ndn) and NdnAEB (EGFR binding-defective necdin mutant). GFAP expression in transfected CPCs was analyzed immunoblotting (F). Necdin-immunoreactive bands at46,42,34 kDa are of FLAG-necdin, endogenous necdin, and necdinAEB, respectively. GFAP expression was quantified by densitometry (G). The broken line indicates the basal GFAP expression in CPCs expressing td-Tomato alone ( = 1). (H, I) Effect of endogenous necdin on GFAP expression induced by EGFR overexpression. CPCs prepared from wild-type (Ndn WT) and necdin-null (Ndn KO) mice at E14.5 were transfected with EGFR, FLAG-necdin (Ndn) and NdnAEB, and GFAP expression levels were analyzed. Each value represents the mean ± SEM, n = 4(G, I). P-values were calculated using one-way ANOVA followed by Tukey-Kramerpost hoc test. NS, not significant at P >0.05.

Moreover, PWS model mice created by transgene insertion or PWS imprinting-center mutation fail to express necdin and display early postnatal lethality [60,61]. The Ndn-mutant mice, mostly examined at their embryonic and neonatal stages, exhibit various morphological and functional abnormalities, in which several are reminiscent of PWS [22,59,62-72]. Thus, primary cells such as CPCs prepared from necdin-null mice are indispensable for dissecting molecular mechanisms underlying these abnormalities. We assume that other stem/progenitor cells expressing both necdin and EGFR also exhibit enhanced EGFR signaling in the absence of endogenous necdin. The present findings warrant close examination of the pathologies associated with dysregu-lation of EGFR signaling in PWS.

5. Conclusions

EGFR and necdin are coexpressed in primary CPCs. Necdin binds to autophosphorylated EGFR via its tyrosine kinase domain. In necdin-null CPCs, the EGFR/ERK pathway is activated through increased interaction between EGFR and Grb2. Furthermore, EGF-promoted astrocyte differentiation is accelerated via the EGFR/ERK pathway in necdin-null CPCs. Necdin restrains astrocyte differentiation induced by overexpression of EGFR or its EGF binding-defective mutant in CPCs. These results suggest that necdin is an intrinsic suppressor of EGFR/ERK signaling in CPCs under physiological and pathological conditions.

Funding

This work was supported by a Grant-in-Aid for Scientific Research B2 (24300134; to K.Y.) from the Japan Society for the Promotion of Science.

Author contributions

I.F., K.H., K.F., M.Y., and K.Y. conceived and designed the research. I.F. performed the experiments. I.F., K.H. and K.F. analyzed the data. I.F. and K.Y. wrote the manuscript.

Conflict of interest

The authors declare no competing financial interests.

Acknowledgments

We are grateful to all the past and present members of the K.Y. group for discussions, research information and experimental materials.

Appendix A. Supplementarydata

Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.cellsig.2015.11.016.

References

[1 ] BA Reynolds, W. Tetzlaff, S. Weiss, A multipotent EGF-responsive striatal embryonic progenitor cell produces neurons and astrocytes, J. Neurosci. 12 (1992) 4565-4574.

[2] V. Tropepe, M. Sibilia, B.G. Ciruna, J. Rossant, E.F. Wagner, D. van der Kooy, Distinct neural stem cells proliferate in response to EGF and FGF in the developing mouse telencephalon, Dev. Biol. 208 (1999) 166-188.

[3] D. Maric, I. Maric, Y.H. Chang, J.L. Barker, Prospective cell sorting of embryonic rat neural stem cells and neuronal and glial progenitors reveals selective effects of basic fibroblast growth factor and epidermal growth factor on self-renewal and differentiation, J. Neurosci. 23 (2003) 240-251.

[4] T.J. Kilpatrick, P.F. Bartlett, Cloned multipotential precursors from the mouse cerebrum require FGF-2, whereas glial restricted precursors are stimulated with either FGF-2 or EGF, J. Neurosci. 15 (1995) 3653-3661.

[5] R.C. Burrows, D. Wancio, P. Levitt, L. Lillien, Response diversity and the timing of progenitor cell maturation are regulated by developmental changes in EGFR expression in the cortex, Neuron 19 (1997) 251-267.

[6] X. Qian, Q. Shen, S.K. Goderie, W. He, A. Capela, AA Davis, S. Temple, Timing of CNS cell generation: a programmed sequence of neuron and glial cell production from isolated murine cortical stem cells, Neuron 28 (2000) 69-80.

L. Lillien, H. Raphael, BMP and FGF regulate the development of EGF-responsive neural progenitor cells, Development 127 (2000) 4993-5005. Y. Sun, S.K Goderie, S. Temple, Asymmetric distribution of EGFR receptor during mitosis generates diverse CNS progenitor cells, Neuron 45 (2005) 873-886. S. Temple, The development of neural stem cells, Nature 414 (2001) 112-117. H.I. Kornblum, R.J. Hussain, J.M. Bronstein, C.M. Gall, D.C. Lee, K.B. Seroogy, Prenatal ontogeny of the epidermal growth factor receptor and its ligand, transforming growth factor alpha, in the rat brain, J. Comp. Neurol. 380 (1997) 243-261. D.W. Threadgill, A.A. Dlugosz, L.A. Hansen, T. Tennenbaum, U. Lichti, D. Yee, C. LaMantia, T. Mourton, K. Herrup, R.C. Harris, et al., Targeted disruption of mouse EGF receptor: effect of genetic background on mutant phenotype, Science 269 (1995) 230- 234.

M. Sibilia, J.P. Steinbach, L. Stingl, A. Aguzzi, E.F. Wagner, A strain-independent postnatal neurodegeneration in mice lacking the EGF receptor, EMBOJ. 17 (1998) 719-731. J. Viti, A. Feathers, J. Phillips, L. Lillien, Epidermal growth factor receptors control competence to interpret leukemia inhibitory factor as an astrocyte inducer in developing cortex, J. Neurosci. 23 (2003) 3385-3393.

K. Maruyama, M. Usami, T. Aizawa, K. Yoshikawa, A novel brain-specific mRNA encoding nuclear protein (necdin) expressed in neurally differentiated embryonal carcinoma cells, Biochem. Biophys. Res. Commun. 178 (1991) 291-296. T. Uetsuki, K. Takagi, H. Sugiura, K. Yoshikawa, Structure and expression of the mouse necdin gene. Identification of a postmitotic neuron-restrictive core promoter, J. Biol. Chem. 271 (1996) 918-924.

Z. Huang, K. Fujiwara, R Minamide, K. Hasegawa, K. Yoshikawa, Necdin controls proliferation and apoptosis of embryonic neural stem cells in an oxygen tension-dependent manner, J. Neurosci. 33 (2013) 10362-10373.

R Minamide, K. Fujiwara, K. Hasegawa, K. Yoshikawa, Antagonistic interplay between necdin and Bmi1 controls proliferation of neural precursor cells in the embryonic mouse neocortex, PLoS One 9 (2014) e84460.

H. Taniura, N. Taniguchi, M. Hara, K. Yoshikawa, Necdin, a postmitotic neuron-specific growth suppressor, interacts with viral transforming proteins and cellular transcription factor E2F1, J. Biol. Chem. 273 (1998) 720-728. H. Taniura, K. Matsumoto, K. Yoshikawa, Physical and functional interactions of neuronal growth suppressor necdin with p53, J. Biol. Chem. 274 (1999) 16242-16248. M. Kurita, T. Kuwajima, I. Nishimura, K. Yoshikawa, Necdin downregulates cdc2 expression to attenuate neuronal apoptosis, J. Neurosci. 26 (2006) 12003-12013. K. Hasegawa, K. Yoshikawa, Necdin regulates p53 acetylation via Sirtuin1 to modulate DNA damage response in cortical neurons, J. Neurosci. 28 (2008) 8772-8784. K. Kuwako, A. Hosokawa, I. Nishimura, T. Uetsuki, M. Yamada, S. Nada, M. Okada, K. Yoshikawa, Disruption of the paternal necdin gene diminishes TrkA signaling for sensory neuron survival, J. Neurosci. 25 (2005) 7090-7099. Y.L. Deribe, P. Wild, A. Chandrashaker, J. Curak, M.H. Schmidt, Y. Kalaidzidis, N. Milutinovic, I. Kratchmarova, L. Buerkle, M.J. Fetchko, P. Schmidt, S. Kittanakom, K.R. Brown, I. Jurisica, B. Blagoev, M. Zerial, I. Stagljar, I. Dikic, Regulation of epidermal growth factor receptor trafficking by lysine deacetylase HDAC6, Sci. Signal. 2 (2009) ra84.

M. Niinobe, K. Koyama, K. Yoshikawa, Cellular and subcellular localization of necdin in fetal and adultmouse brain, Dev. Neurosci. 22 (2000) 310-319. T. Aizawa, K. Hasegawa, T. Ohkumo, S. Haga, K. Ikeda, K. Yoshikawa, Neural stem cell-like gene expression in a mouse ependymoma cell line transformed by human BK polyomavirus, Cancer Sci. 102 (2011) 122-129.

H. Taniura, M. Kobayashi, K. Yoshikawa, Functional domains of necdin for proteinprotein interaction, nuclear matrix targeting, and cell growth suppression, J. Cell. Biochem. 94 (2005) 804-815.

I. Gur, K. Fujiwara, K. Hasegawa, K. Yoshikawa, Necdin promotes ubiquitin-dependent degradation of PIAS1 SUMO E3 ligase, PLoS One 9 (2014) e99503. K.K. Johe, T.G. Hazel, T. Muller, M.M. Dugich-Djordjevic, R.D. McKay, Single factors direct the differentiation of stem cells from the fetal and adult central nervous system, Genes Dev. 10 (1996) 3129-3140.

A. Bonni, Y. Sun, M. Nadal-Vicens, A. Bhatt, D.A. Frank I. Rozovsky, N. Stahl, G.D. Yancopoulos, M.E. Greenberg, Regulation of gliogenesis in the central nervous system by the JAK-STAT signaling pathway, Science 278 (1997) 477-483.

K. Nakashima, S. Wiese, M. Yanagisawa, H. Arakawa, N. Kimura, T. Hisatsune, K. Yoshida, T. Kishimoto, M. Sendtner, T. Taga, Developmental requirement of gp130 signaling in neuronal survival and astrocyte differentiation, J. Neurosci. 19 (1999) 5429-5434.

W. Ochiai, M. Yanagisawa, T. Takizawa, K. Nakashima, T. Taga, Astrocyte differentiation of fetal neuroepithelial cells involving cardiotrophin-1-induced activation of STAT3, Cytokine 14 (2001) 264-271.

F. Barnabe-Heider, J.A. Wasylnka, K.J. Fernandes, C. Porsche, M. Sendtner, D.R. Kaplan, F.D. Miller, Evidence that embryonic neurons regulate the onset of cortical gliogenesis via cardiotrophin-1, Neuron 48 (2005) 253-265. M.K Nicholas, R.V. Lukas, N.F. Jafri, L. Faoro, R Salgia, Epidermal growth factor receptor-mediated signal transduction in the development and therapy of gliomas, Clin. Cancer Res. 12 (2006) 7261-7270.

Y. Zhu, L.F. Parada, The molecular and genetic basis of neurological tumours, Nat. Rev. Cancer 2 (2002) 616-626.

B. Dasgupta, D.H. Gutmann, Neurofibromin regulates neural stem cell proliferation, survival, and astroglial differentiation in vitro and in vivo, J. Neurosci. 25 (2005) 5584-5594.

X. Li, J.M. Newbern, Y. Wu, M. Morgan-Smith, J. Zhong, J. Charron, W.D. Snider, MEK is a key regulator of gliogenesis in the developing brain, Neuron 75 (2012) 1035-1050.

Y. Wang, E. Kim, X. Wang, B.G. Novitch, K. Yoshikawa, L.S. Chang, Y. Zhu, ERK inhibition rescues defects in fate specification of Nf1-deficient neural progenitors and brain abnormalities, Cell 150 (2012) 816-830.

[38] S. Li, P. Mattar, R Dixit, S.O. Lawn, G. Wilkinson, C. Kinch, D. Eisenstat, D.M. Kurrasch, J.A. Chan, C. Schuurmans, RAS/ERK signaling controls proneural genetic programs in cortical development and gliomagenesis, J. Neurosci. 34 (2014) 2169-2190.

[39] R.N. Jorissen, F. Walker, N. Pouliot, T.P. Garrett, C.W. Ward, A.W. Burgess, Epidermal growth factor receptor: mechanisms of activation and signalling, Exp. Cell Res. 284 (2003) 31-53.

[40] N. Jura, N.F. Endres, K. Engel, S. Deindl, R Das, M.H. Lamers, D.E. Wemmer, X. Zhang, J. Kuriyan, Mechanism for activation of the EGF receptor catalytic domain by the juxtamembrane segment, Cell 137 (2009) 1293-1307.

[41] X. Zhang, J. Gureasko, K. Shen, P.A. Cole, J. Kuriyan, An allosteric mechanism for activation of the kinase domain of epidermal growth factor receptor, Cell 125 (2006) 1137-1149.

[42] M.A. Lemmon, J. Schlessinger, K.M. Ferguson, The EGFR family: not so prototypical receptor tyrosine kinases, Cold Spring Harb. Perspect Biol. 6 (2014) a020768.

[43] X. Zhang, KA Pickin, R Bose, N. Jura, P.A. Cole, J. Kuriyan, Inhibition of the EGF receptor by binding of MIG6 to an activating kinase domain interface, Nature 450 (2007) 741-744.

[44] O. Segatto, S. Anastasi, S. Alema, Regulation of epidermal growth factor receptor signalling by inducible feedback inhibitors, J. Cell Sci. 124 ( 2011) 1785-1793.

[45] S. Anastasi, M.F. Baietti, Y. Frosi, S. Alema, O. Segatto, The evolutionarily conserved EBR module of RALT/MIG6 mediates suppression of the EGFR catalytic activity, Oncogene 26 (2007) 7833-7846.

[46] N. Normanno, A. De Luca, C. Bianco, L. Strizzi, M. Mancino, M.R. Maiello, A. Carotenuto, G. De Feo, F. Caponigro, D.S. Salomon, Epidermal growth factor receptor (EGFR) signaling in cancer, Gene 366 (2006) 2-16.

[47] R.S. Herbst, M. Fukuoka, J. Baselga, Gefitinib—a novel targeted approach to treating cancer, Nat. Rev. Cancer 4 (2004) 956-965.

[48] L.E. De Faveri, C.D. Hurst, F.M. Platt, C.F. Taylor, JA. Roulson, M. Sanchez-Carbayo, M.A Knowles, E.J. Chapman, Putative tumour suppressor gene necdin is hypermethylated and mutated in human cancer, Br. J. Cancer 108 (2013) 1368-1377.

[49] P. Chomez, O. De Backer, M. Bertrand, E. De Plaen, T. Boon, S. Lucas, An overview of the MAGE gene family with the identification of all human members of the family, Cancer Res. 61 (2001) 5544-5551.

[50] T. Aizawa, K. Maruyama, H. Kondo, K. Yoshikawa, Expression of necdin, an embryonal carcinoma-derived nuclear protein, in developing mouse brain, Brain Res. Dev. Brain Res. 68 (1992) 265-274.

[51] M. Kobayashi, H. Taniura, K. Yoshikawa, Ectopic expression of necdin induces differentiation of mouse neuroblastoma cells, J. Biol. Chem. 277 (2002) 42128-42135.

[52] R Jornsten, T. Abenius, T. Kling, L. Schmidt, E. Johansson, T.E. Nordling, B. Nordlander, C. Sander, P. Gennemark, K. Funa, B. Nilsson, L. Lindahl, S. Nelander, Network modeling of the transcriptional effects of copy number aberrations in glioblastoma, Mol. Syst. Biol. 7 (2011) 486.

[53] X. Li, S.C. Hughes, R Wevrick Evaluation of melanoma antigen (MAGE) gene expression in human cancers using The Cancer Genome Atlas, Cancer Genet. 208 (2015) 25-34.

[54] N. Sanai, A. Alvarez-Buylla, M.S. Berger, Neural stem cells and the origin of gliomas, N. Engl. J. Med. 353 (2005) 811-822.

[55] Y. Nakada, H. Taniura, T. Uetsuki, J. Inazawa, K. Yoshikawa, The human chromosomal gene for necdin, a neuronal growth suppressor, in the Prader-Willi syndrome deletion region, Gene 213 (1998) 65-72.

[56] P. Jay, C. Rougeulle, A. Massacrier, A. Moncla, M.G. Mattei, P. Malzac, N. Roeckel, S. Taviaux, J.L. Lefranc, P. Cau, P. Berta, M. Lalande, F. Muscatelli, The human necdin gene, NDN, is maternally imprinted and located in the Prader-Willi syndrome chromosomal region, Nat. Genet. 17 (1997) 357-361.

[57] H.R. MacDonald, R. Wevrick The necdin gene is deleted in Prader-Willi syndrome and is imprinted in human and mouse, Hum. Mol. Genet. 6 (1997) 1873-1878.

[58] M. Gerard, L. Hernandez, R Wevrick, C.L. Stewart, Disruption of the mouse necdin gene results in early post-natal lethality, Nat. Genet. 23 (1999) 199-202.

[59] F. Muscatelli, D.N. Abrous, A. Massacrier, I. Boccaccio, M. Le Moal, P. Cau, H. Cremer, Disruption of the mouse necdin gene results in hypothalamic and behavioral alterations reminiscentof the human Prader-Willi syndrome, Hum. Mol. Genet. 9 (2000) 3101-3110.

[60] T. Yang, T.E. Adamson, J.L. Resnick S. Leff, R Wevrick U. Francke, N.A. Jenkins, N.G. Copeland, C.I. Brannan, A mouse model for Prader-Willi syndrome imprinting-centre mutations, Nat. Genet. 19 (1998) 25-31.

[61] J.M. Gabriel, M. Merchant, T. Ohta, Y. Ji, R.G. Caldwell, M.J. Ramsey, J.D. Tucker, R Longnecker, R.D. Nicholls, A transgene insertion creating a heritable chromosome deletion mouse model of Prader-Willi and Angelman syndromes, Proc. Natl. Acad. Sci. U. S. A. 96 (1999) 9258-9263.

[62] J. Ren, S. Lee, S. Pagliardini, M. Gerard, C.L. Stewart, J.J. Greer, R Wevrick Absence of Ndn, encoding the Prader-Willi syndrome-deleted gene necdin, results in congenital deficiency of central respiratory drive in neonatal mice, J. Neurosci. 23 (2003) 1569-1573.

[63] T. Kuwajima, I. Nishimura, K. Yoshikawa, Necdin promotes GABAergic neuron differentiation in cooperation with Dlx homeodomain proteins, J. Neurosci. 26 (2006) 5383-5392.

[64] T. Kuwajima, K. Hasegawa, K. Yoshikawa, Necdin promotes tangential migration of neocortical interneurons from basal forebrain, J. Neurosci. 30 (2010 ) 3709-3714.

[65] S. Zanella, F. Watrin, S. Mebarek F. Marly, M. Roussel, C. Gire, G. Diene, M. Tauber, F. Muscatelli, G. Hilaire, Necdin plays a role in the serotonergic modulation of the mouse respiratory network: implication for Prader-Willi syndrome, J. Neurosci. 28 (2008) 1745-1755.

[66] A.A. Tennese, C.B. Gee, R Wevrick Loss of the Prader-Willi syndrome protein necdin causes defective migration, axonal outgrowth, and survival of embryonic sympathetic neurons, Dev. Dyn. 237 (2008) 1935-1943.

[67] N.L. Miller, R Wevrick P.L. Mellon, Necdin, a Prader-Willi syndrome candidate gene, regulates gonadotropin-releasing hormone neurons during development, Hum. Mol. Genet. 18 (2009) 248-260.

[68] J.R. Bush, R Wevrick Loss of necdin impairs myosin activation and delays cell polarization, Genesis 48 (2010) 540-553.

[69] J. Aebischer, R. Sturny, D. Andrieu, A. Rieusset, F. Schaller, S. Geib, C. Raoul, F. Muscatelli, Necdin protects embryonic motoneurons from programmed cell death, PLoS One 6 (2011) e23764.

[70] K. Fujiwara, K. Hasegawa, T. Ohkumo, H. Miyoshi, Y.H. Tseng, K. Yoshikawa, Necdin controls proliferation of white adipocyte progenitor cells, PLoS One 7 (2012) e30948.

[71 ] J.R. Bush, R Wevrick Loss of the Prader-Willi obesity syndrome protein necdin promotes adipogenesis, Gene 497 (2012) 45-51.

[72] K. Hasegawa, T. Kawahara, K. Fujiwara, M. Shimpuku, T. Sasaki, T. Kitamura, K. Yoshikawa, Necdin controls FoxO1 acetylation in hypothalamic arcuate neurons to modulate the thyroid axis, J. Neurosci. 32 (2012) 5562-5572.