Matrin-3 is essential for fibroblast growth factor 2-dependent maintenance of neural stem cells

To investigate the mechanisms underlying the maintenance of neural stem cells, we performed two-dimensional fluorescence-difference gel electrophoresis (2D-DIGE) targeting the nuclear phosphorylated proteins. Nuclear phosphorylated protein Matrin-3 was identified in neural stem cells (NSCs) after stimulation using fibroblast growth factor 2 (FGF2). Matrin-3 was expressed in the mouse embryonic subventricular and ventricular zones. Small interfering RNA (siRNA)-mediated knockdown of Matrin-3 caused neuronal differentiation of NSCs in vitro, and altered the cerebral layer structure of foetal brain in vivo. Transfection of Matrin-3 plasmids in which the serine 208 residue was point-mutated to alanine (Ser208Ala mutant Matrin3) and inhibition of Ataxia telangiectasia mutated kinase (ATM kinase), which phosphorylates Matrin-3 Ser208 residue, caused neuronal differentiation and decreased the proliferation of neurosphere-forming stem cells. Thus, our proteomic approach revealed that Matrin-3 phosphorylation was essential for FGF2-dependent maintenance of NSCs in vitro and in vivo.

2 scanner. Fluorescent gel images obtained from ProQ Diamond and SYPRORuby were merged with Cy-Dye labelled gel images on DeCyder software. For instance, using DeCyder software, protein spots were detected from total protein profiles (Cy-Dye staining) and 2D-phospho protein profile (Pro Q Diamond staining), respectively. Among them, protein spots commonly and highly expressed in each profile were selected. These selected protein spots were served as the landmarks. Then, both protein profiles were matched and compared by calculating the distance between the landmarks by "co-detection function" using DeCyder software proprietary algorithm. After spot matching and profiling, statistical quantitative analysis of the protein patterns was performed.

In vivo analyses
To prepare samples for western blotting, the cerebral tissues at the embryonic, postnatal, and adult stages were obtained from ICR mice.
To perform immunohistochemical analyses, the mice were anaesthetised with a lethal dose of intraperitoneal pentobarbital (150 mg/kg) and were perfused with PBS followed by 4% paraformaldehyde. The brains were fixed for 2 h on ice. After fixation, the brains were dehydrated in a 30% sucrose solution and embedded in OCT compound. The frozen brains were sectioned into 5-µm coronal slices. These slices were treated in a citrate buffer (pH 6.0) for antigen-retrieval in a microwave. Immunoreactive signals were detected using a tyramide signal amplification procedure (Life Technologies).

Cell transfection
NSCs were transfected with plasmids using electroporation (MicroPorator-mini, AR 3 BROWN, Tokyo, Japan) and/or FuGENE 6 transfection reagent (Promega). Transfection experiments were performed with 2-3 µg plasmid DNA in a final volume of 100 µL of OPTI-MEM I medium (Life Technologies). After the transfection media were removed, the media were replaced by NSC culture medium, and the cells were incubated for 72 h. These samples were used for immunostaining.

Immunofluorescence staining
NSCs were fixed with PBS containing 4% paraformaldehyde for 10 min. After washing with PBS 3 times, the NSCs were treated with PBS + 0.1% Tween 20 and then with 10% serum in PBST at room temperature for 30 min. The cells were incubated overnight with antibodies diluted in 10% serum in PBST. The immunoreactive signals were observed using an epifluorescent microscope system (Olympus, Tokyo, Japan).

Tissue sections of mouse and human hippocampus
Tissue sections of normal human adult brain hippocampus from an 82-year-old woman were obtained from US Biomax, Inc. (Rockville, MD, USA). Normal adult mouse brain tissue sections were isolated at postnatal day 30.

Fig.S7. KU55933 reduced the number of neurosphere-forming stem cells and those sizes.
Neurospheres were cultured with NSC culture medium containing 50 nM KU55933 for 7 days. Bar, 100 μm.