This study was conducted in accordance with the animal care guidelines issued by the National Institutes of Health and by the San Francisco Veterans Affairs Medical Center Animal Care and Use Committee. Adult mice heterozygous for Emx1 gene derived from cryopreserved embryos at Jackson Laboratory were bred to generate heterozygous and homozygous offspring that were maintained in house in the institutional standard cages (4 mice per cage) on a 12-h light/12-h dark cycle, with ad libitum access to water and food before and during experimental procedures. All in-vitro and in-vivo procedures were conducted by examiners blinded to experimental conditions (for the detailed methodology of Western blot analysis and antibodies used, please refer to Data Sheet 1 in the Supplementary Material).

Forebrains were dissected from Emx1 WT or KO embryos at embryonic day 14 (E14) harvested from time-pregnant mice. Neurosphere assay for the adult mice (2.5 months of age) were conducted from the dissected subventricular zone (SVZ) or the hippocampus. After enzymatic digestion with Papain (10 U/ml) and DNase (0.5 mg/ml) (Sigma Aldrich, St. Louis, MO, USA) solution at 37°C for 20 min, brain tissue was mechanically triturated with a fire-polished Pasteur pipette into single cell suspension and resuspended in DMEM-F12 (1:1) medium in the presence of hEGF (20 ng/ml) or hEGF (20 ng/ml) + FGF (10 ng/ml) (Invitrogen, Eugene, Oregon, USA) and heparin (2 μg/ml) (StemCell Technologies, Inc., Vancouver, BC, Canada). Neural colony forming cell (NCFC) assay kit (StemCell Technologies, Inc., Vancouver, BC, Canada) was used for neurosphere assay. Cells were suspended in complete NeuroCult NCFC medium, and mixed with collagen solution, 7500 cells were dispensed into 35 mm culture dishes and cultured for 21 days. hEGF and FGF-2 (20 ng/ml each) was added into the semi-solid culture every 7 days. The diameters of spheres were measured using a 2 mm-grid dish on a Zeiss invert microscope. Remaining cells were suspended in proliferation medium composed of DMEM/F-12 (1:1) containing B27 supplement and 20 ng/ml hEGF, and were cultured in a T-75 tissue culture flask to form primary floating neurospheres. Floating neurospheres were passaged every 7 days.

Primary neurospheres were mechanically dissociated into single cells suspension and ~400 viable cells were plated into each well of a 6-well cell culture plate containing proliferation medium. The number of secondary neurospheres generated in each well was counted 10 days after plating and averaged (n = 3–4 replicates/neurosphere samples/genotype).

Neurospheres were harvested after 2nd passage by centrifugation at 500 rpm for 5 min and resuspended in growth factor-free DMEM/F-12 (1:1) medium. Cells in neurospheres were mechanically dissociated by trituration and filtered through 0.45 μm cell strainer. Viable single cells (10⁵ cells/well) were plated on poly-D-lysine/laminin coated 6-well culture plate in DMEM-F12 (1:1) plus NeuroCult NSFC differentiation supplements for 10 days. The differentiated cells were fixed with 4% paraformaldehyde in 0.1 M PBS for 15 min, and then washed with PBS twice. Differentiated cells were immunostained with following reagents: mouse anti-GFAP; mouse anti-Tuj1; rat anti-nestin; Cy3 goat anti-mouse; biotinylated sheep anti-rat; Cy3 streptavidine. Cells were count stained with DNA dye Hoechst. Stained cells were photographed with a Zeiss invert fluorescence microscope. Differentiation ratio was calculated as immuno-positive cells divided by total cells stained with Hoechst.

The migration of the NSCs in response to fetal calf serum (FCS) or recombinant human VEGF₁₆₅ (R&D Systems, Minneapolis, USA) was assessed using a modified Boyden chamber assay as previously described (Lamszus et al., 1998; Schmidt et al., 2009). Briefly, quadruplicates of media containing 10% FCS or VEGF were added to the lower wells of a 96-well modified Boyden chamber (Neuro Probe, Gaithersburg, MD, USA), and wells were covered with an 8-μ m pore size Nucleopore filter that had previously been coated with 10 μg/ml laminin (Invitrogen, Eugene, Oregon, USA). NSCs from dissociated neurospheres passages 3–4 were then suspended at 2.5 × 10⁴ cells in 50 μl of either serum-free DMEM/F12 medium containing 0.1% bovine serum albumin (BSA) or in growth medium containing EGF alone or EGF+FGF and seeded into the upper wells. After incubation for 15 h at 37°C, non-migrated cells were scraped off the upper side of the filter and filters were stained with Diff-Quik (VWR, Brisbane, CA, USA). Nuclei of migrated cells were counted in an average of 180 high power fields (field size 20 × 20 to 100 × 100 μm) using a 40x objective by using the Stereo Investigator software (Stereo Investigator, MicroBrightField, VA). Values were assessed in quadruplicate from at least three independent experiments and expressed as mean ± standard deviation in percentage of the control migration (= 100%). The control migration rate for each assay was assessed in response to either serum-free medium containing 0.1% BSA or growth medium containing EGF or EGF+FGF.

Forty milliliters of NSCs (passage 2), seeded with a density of 100,000 cells/ml, were harvested after 5 days of growing in defined media by centrifugation at 400 × g. Cell pellets were lysed in 1 ml of lysis-buffer containing 9 M urea (Sigma), 4% (w/v) CHAPS (Calbiochem, San Diego, CA), 0.5% (v/v) pharmalytes pH 3–10 (Amersham Bioscience), 1 mM EDTA (American Bioanalytical, Natick, MA), Complete protease inhibitor tablets were from Roche (Indianapolis, IN). Homogenates were triturated using a syringe with 20G-needle and incubated for 1 h at RT. After samples were centrifuged at 4000 rpm for 10 min, supernatants were transferred into an empty 1.5 ml tube, while pellets were discarded. To determine protein concentrations the 2-D-Quant-protein concentration kit (Amersham Bioscience) was used following the instructor's guideline. The first dimension was performed using an Ettan IPGphor II isoelectric focusing unit (Amersham Bioscience). Five-hundred micrograms of protein sample were mixed with 0.2% (v/v) ampholytes (Amersham Bioscience) and 170 μl of 2 × rehydration-buffer containing 9 M urea, 2% (w/v) CHAPS, 0.002% (w/v) bromphenol blue, 18 mm of dithiothreitol (DTT (ThermoFisher), and 0.5% IPG-buffer 3–10 (Amersham Bioscience). To reach the final volume of 340 μl, lysis-buffer (see above) was added. Each sample was pipetted into a slot of the dehydration tray in which an 18 cm-Immobilene™ Dry Strip pH 3–10 (Amersham Bioscience) was placed upside down. In order to inhibit drying-out of the solution overnight each slot was covered with mineral oil (Amersham Bioscience). The strips passively rehydrated at constant 25°C for 18 h. Afterwards the strips were transferred to the focusing tray where the proteins were separated according to their isoelectric points using the following protocol: 1 h at 150 V, 3 h at 300 V, 6 h at 1000 V using the linear ramp, 3 h at 6000 V using the linear ramp, and 3 h at 6000 V. To inhibit diffusion of the proteins after focusing the machine held 50 V until manually stopped. After the focusing process strips were equilibrated two times for 10 min in a buffer containing 50 mM of Tris-HCl (pH 8.8), 6 M Urea, 30% (v/v) glycerol, 2% (w/v) SDS, 0.002% (w/v) bromphenol blue and either 65 mM DTT. Equilibrated strips were put on top of a 12% polyacrylamide gel, made by using standard laboratory procedures, and covered with 0.5% (w/v) melted agarose also containing 0.002% (w/v) bromphenol blue. Ten microliters of a molecular weight standard (BioRad, Hercules, CA) was dropped on a piece of blotting paper and was put next to the strip into the agarose solution before polymerization. Gels were run over night at 10 μA per gel at 4°C using an Ettan DALTsix electrophoreses unit (Amersham Bioscience, Piscataway, NJ; for the detailed methodology of Gel staining with silver or Coomassie Blue, please refer to Data Sheet 1 in the Supplementary Material).

Peptide extracts were separated on-line by nanoLC utilizing a 2D LC NanoLC System (Eksigent/AB Sciex). Liquid chromatographic separations were performed using a PepMap (Dionex/LC Packings) trap column and a reversed phase nano-column (75 μm in diameter × 150 mm in height packed in-house with Jupiter Proteo C12 endcapped resin, 90 Å pore size, 4 μm particle size). An aliquot of peptide extract (3–4 μL) was loaded onto the trap column with loading solvent (0.1% formic acid) at a flow rate of 20 μL/min. The trap column was washed with the loading solvent for 3 min prior to switching it in line with the reversed phase nano-column. The nano-column and elution buffers were maintained at ambient temperature and mobile phase flow rate was 250 nL/min. The nano-column was equilibrated with 2% Solvent B in Solvent A for 20 min prior to sample injection (Solvent A: 2% acetonitrile/0.1% formic acid; Solvent B: 80% acetonitrile/0.1% formic acid). Peptide separation was accomplished using a binary gradient which consisted of a 5 min isocratic wash at 2% B followed by a linear gradient of 2–50% B over 45 min, and concluded with a column cleanup step of 95% B for 7 min. The nano-column was interfaced directly to a nanoelectrospray ion source (Protana) mounted on a QSTAR Elite quadrupole/quadrupole/time-of-flight (QqTOF) mass spectrometer (AB Sciex). Protein identification was accomplished by isolating sequentially eluting peptide populations with a single mass-to-charge ratio (m/z), within the mass spectrometer, fragmenting this population, and measuring the masses of the peptide fragment ions. Peak lists were generated with the Mascot Daemon (Matrix Sciences) and the Mascot v2.2 search engine (Matrix Science) was used to search the experimentally determined peptide fragment ion masses against a theoretical fragment ion mass database generated by in silico digestion and fragmentation of all proteins in the UniProtKB/Swiss-Prot Protein Knowledgebase (downloaded 03-09-2011; 525,997 protein sequences; 185,874,894 residues), taxonomy: rodentia (25,596 protein sequences). The following search parameters were utilized: precursor ion mass tolerance ± 150 ppm; fragment ion mass tolerance 0.15 Da; tryptic digestion; 2 missed cleavages; p < 0.05; ion score or expect cut-off 0.05; fixed modification: Cys-carbamidomethyl; variable modifications: deamidation (Asn and Gln), Met-sulfoxide, and Pyro-glu (N-term Gln; for the detailed methodology of In-gel digestion procedure, please refer to Data Sheet 1 in the Supplementary Material).

The Elsevier's Pathway Studio v. 10.0 (Ariadne Genomics/Elsevier) was used to deduce relationships among differentially expressed proteomics protein candidates using the Ariadne ResNet database (Bonnet et al., 2009; Yuryev et al., 2009). “Subnetwork Enrichment Analysis” (SNEA) algorithm was selected to extract statistically significant altered biological and functional pathways pertaining to each identified set of protein hits SNEA utilizes Fisher's statistical test used to determine if there are nonrandom associations between two categorical variables organized by specific relationship. SNEA starts by creating a central “seed” from all relevant entities in the database, and retrieving associated entities based on their relationship with the seed (that is, binding partners, expression targets, protein modification targets, regulation). The algorithm compares the sub-network distribution to the background distribution using one-sided Mann–Whitney U-Test, and calculates a p-value indicating the statistical significance of difference between two distributions. In our analysis, “GenBank ID symbols” of the 38 altered proteins from 2D-PAGE-MS/MS spectrometry data were imported to the software to form an experimental data set. For the reconstruction of networks of pathways, biological processes, and Molecular function were evaluated for each single protein hit and its associated targets (networks and pathways; Daraselia et al., 2012; Pyatnitskiy et al., 2014). Integrated Venn diagram analysis was performed using “the InteractiVenn”: a web-based tool for the analysis of comlex data sets. Green rectangles, violet rectangles, and blue hexagons reflect biological processes, disease processes, and functional classes, respectively. The different colors of proteins reflect their degree of expression.

Differentially expressed proteins were GO-classified based on the PANTHER (Protein ANalysis THrough Evolutionary Relationships) system (http://www.pantherdb.org; Mi et al., 2016). PANTHER software classifies genes and proteins by their functions, using published scientific experimental evidence and evolutionary relationships abstracted by curators with the goal of predicting function even in the absence of direct experimental evidence (Mi et al., 2010). Proteins are classified into families and subfamilies of shared function, which are then categorized using a highly gene controlled vocabulary (Gene Ontology terms) by biological process, molecular function and cellular localization (for the detailed methodology of Gene Ontology analysis, please refer to Data Sheet 1 in the Supplementary Material, a detailed description of these data is presented in the Supplemental Figures 1–3).

NSCs were plated onto poly-L-ornithine coated plates at 10⁵/well (24-well plate). Twenty-four hours post plating, recombinant mouse VEGF-A (100 ng/ml, R&D system) was added to culture medium and BrdU was added into culture 24 h later at a concentration of 10 μM prior to fixation with 4% PFA. SU1498 was added into culture at 5 or 10 μM prior to the addition of VEGF. The number of BrdU positive cells was quantified only in live cells.

Data were expressed as mean ± s.e.m. Statistical tests were carried out with one-way or two-way analyses of variance (ANOVAs) followed by post-hoc paired comparisons using the Fisher's PLSD test when appropriate (Statview 5.0.1, SAS Institute Inc., Cary, NC). Values of p < 0.05 were considered as significant.

Emx1 deletion is associated with the reduction of volume and immature neurons in the dentate gyrus (DG) of adult mice (Hong et al., 2007). To determine whether this abnormality originated from the early stage of brain development, neurosphere assay was performed with embryonic or adult brains from Emx1 knockout (KO) and wild type (WT) mice. Although embryonic brains from both genotypes formed neurospheres, the number of neurospheres or neuroprogenitor cells (0.5 mm < diameter < 1 mm) in E14 Emx1^−/− forebrains was significantly less than those of Emx1^+/+ brains. In addition, the size of neurospheres was smaller, indicating that Emx1^−/− embryonic NSCs had a limited proliferative potential (Figures 1A,B). The frequency of neurospheres formed during 21 days of culture also differed depending on the presence of growth factors. There was a significant effect of FGF on the frequency of neurospheres greater than 0.5 mm in diameter in both genotypes (Figure 1). In contrast to the embryonic brains, the adult brains harbor very few true NSCs in either genotype. There was also no significant difference in the overall frequency of NSCs within the progenitor regions of subventricular zone or the hippocampus between genotypes (Figure 1C).

Neurospheres contain multipotent NSCs with heterogeneous capacity in clonal expansion. To determine whether the observed difference in the number or size of neurospheres is attributed to differences in self-renewal capacity between genotypes, we quantified the number of secondary neurospheres generated by single primary neurospheres following 10 days of culture in a self-renewal assay. We found that primary neurospheres from forebrains of both Emx1 WT and KO embryos had the ability of self-renewal (Figure 2A), but differed in the extent of producing secondary neurospheres (Figure 2B). The numbers of primary and secondary neurospheres from Emx1 KO brains were significantly reduced to ~30 and 50% of those from the Emx1 WT ones, respectively (Figure 2).

To compare the extent of multipotency of NSCs derived from Emx1 WT and KO embryonic forebrains, cells from neurospheres of both genotypes were exposed to differentiation medium for 10 days. The majority of dissociated NSCs cultured as monolayer cells displayed immunoreactivity for nestin, a marker of neuroprogenitor cells, prior to the induction of differentiation (Figures 3A–D). Multipotency was demonstrated by differentiation into TuJ1-positive neurons, and GFAP-positive astrocytes. However, no difference was found in the differentiation ratio of neurons or astrocytes between Emx1 WT and Emx1 KO NSCs (Figure 3E).

Neuroblasts migrate in close association with blood vessels and long-distance migration of neuroblasts was found in peri-infarct tissue in human stroke, implicating a role for VEGF or other angiogenic factors in neurogenesis and neuroblast migration. To quantify the chemotactic response of NSCs, a modified Boyden Chamber assay was used in the presence of FCS or chemokine VEGF₁₆₅ as chemoattractants. In response to 10% FCS, Emx1 WT NSCs exhibited a 2.1-fold increase in chemotaxis activity relative to Emx1 KO NSCs. When cultured in the presence of EGF, NSCs of either Emx1 genotype failed to migrate in response to VEGF₁₆₅ at 10 or 100 ng/ml (Figure 4A). In contrast, when cultured in the presence of EGF and FGF, VEGF₁₆₅ at 50 ng/ml was able to induce a robust migration of Emx1 WT NSCs compared to Emx1 KO NSCs (Figure 4B). Out findings indicate that Emx1 gene deletion negatively impacted the chemotactic response of NSCs toward serum or VEGF, and the migration Emx1 WT NSCs induced by VEGF requires the presence of FGF in addition to EGF (Zhang et al., 2014).

Using tandem mass spectrometry, we analyzed a number of most distinctly different spots on the 2-D PAGE comparing the patterns of protein expression from the lysates of Emx1 WT and KO embryonic NSCs (Figure 5). Our neuroproteomics analysis detected 38 differentially expressed proteins between Emx1^−/− and Emx1^+/+ (see Table 1 for protein Ids and their respective biological functions). Detailed characteristics of each individual protein with its corresponding mass spectrometry data are included in Supplemental Tables 1, 2 and Data Sheet 1 in the Supplementary Material.

Based on the above neuroproteomics findings, we utilized a neurosystem biology platform for protein network and pathway analysis relevant to the altered proteins in the Emx1^+/+ vs. Emx1^−/− NSCs. For this purpose, we utilized two bioinformatics analysis approaches: a global protein interaction and unsupervised pathway assessment on one side and a cluster targeted approach for selective implicated functional biological pathways on the other hand.

For the targeted cluster analysis, selected Emx1-implicated pathways were assessed against Emx1-identified differential proteomic data. Using the SNEA algorithm, targeted enriched pathways revealed that several altered proteins were associated with Emx1 differential proteome-implicated neural pathways including: cell proliferation, cell migration, cell differentiation, cytoskeletal actin organization, and neuronal migration as illustrated in Figures 6A–E. For detailed cluster analysis in silico validation including statistical analysis, protein entity, biological process along with the interaction type and the PubMed reference utilized see Supplemental Tables 3A–C, 4A–C, 5A–C, 6A–C and Data Sheet 1 in the Supplementary Material.

For the global interaction and unsupervised pathway assessment, differential pathways were generated using the “direct interaction” algorithm to map the relationship among the identified proteins. We found that among the 38 altered proteins, 17 had direct regulatory function, including binding, post-translational modifications and transcriptional regulation. Unsupervised pathway assessment, showed that these proteins are implicated in several biological pathways involving brain development, axonal guidance, synaptic transmission, neurogenesis along with the hippocampal morphology biological process (highlighted in blue rectangles). Of interest, these altered proteins showed a centrality relation with VEGF protein being an upstream regulator for several of the Emx1-identified proteome involving Cofilin1, enolase, phosphatidyl-ethanolamine-binding protein, and neurite growth-promoting factor 2 (Figure 7). Statistical significance of the interaction was performed in silico for the validation process. A detailed depiction of these data is presented in the (Supplemental Tables 7A–C and Data Sheet 1 in the Supplementary Material). These include the protein entity and biological process involved along with the interaction type and directionality and the PubMed reference utilized to derive these interaction type. Due to the well-recognized central role of Cofilin1 and its diverse interaction with several of the identified biological pathways, we performed a functional validation analysis to determine the expression of Cofilin1/p-Cofilin1in the NSCs of both genotypes in response to VEGF signaling.

To further determine whether Cofilin1 plays a role in VEGF signaling, lysates of NSCs from both genotypes stimulated with VEGF or media were analyzed by western blotting for the expression of p-Cofilin1, total Cofilin1, and VEGF RECEPTOR 2 FLK. We found that Emx1 KO NSCs had significantly reduced p-Cofilin1 compared to those from Emx1 WT, when cultured in medium containing EGF only or EGF+FGF (Figures 8A,B). VEGF further increased the level of p-Cofilin only in the WT NSCs (Figures 8A,B). VEGF also modestly induced the expression of FLK only in the WT NSCs (Figures 8A,C). Our results suggest that Emx1 gene deletion reduces phosphorylated Cofilin1, likely contributes to the difference in neurogenesis and chemotactic response to VEGF.

Because VEGF receptors are expressed in NSCs (Maurer et al., 2003) and VEGF stimulates the expansion of NSCs in vitro (Schänzer et al., 2004), we determine the effect of VEGF on the proliferation and survival of NSCs. We found that VEGF-A (100 ng/ml) significantly increased the percentage of BrdU immunoreactive cells already after 4 h (p < 0.05) of treatment. The effect was even more pronounced after 17 h (p < 0.001; Supplemental Figure 1A and Data Sheet 1 in the Supplementary Material). In addition, VEGF-induced NSC expansion was dose dependently reduced by the specific VEGF RECEPTOR 2 antagonist SU1498 (Supplemental Figure 1B and Data Sheet 1 in the Supplementary Material), suggesting that VEGF increased the number of NSCs possibly by stimulating proliferation or enhancing their survival.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.