DISC1 Regulates the Proliferation and Migration of Mouse Neural Stem/Progenitor Cells through Pax5, Sox2, Dll1 and Neurog2

Background: Disrupted-in-schizophrenia 1 (DISC1) regulates neurogenesis and is a genetic risk factor for major psychiatric disorders. However, how DISC1 dysfunction affects neurogenesis and cell cycle progression at the molecular level is still unknown. Here, we investigated the role of DISC1 in regulating proliferation, migration, cell cycle progression and apoptosis in mouse neural stem/progenitor cells (MNSPCs) in vitro. Methods: MNSPCs were isolated and cultured from mouse fetal hippocampi. Retroviral vectors or siRNAs were used to manipulate DISC1 expression in MNSPCs. Proliferation, migration, cell cycle progression and apoptosis of altered MNSPCs were analyzed in cell proliferation assays (MTS), transwell system and flow cytometry. A neurogenesis specific polymerase chain reaction (PCR) array was used to identify genes downstream of DISC1, and functional analysis was performed through transfection of expression plasmids and siRNAs. Results: Loss of DISC1 reduced proliferation and migration of MNSPCs, while an increase in DISC1 led to increased proliferation and migration. Meanwhile, an increase in the proportion of cells in G0/G1 phase was concomitant with reduced levels of DISC1, but significant changes were not observed in the number MNSPCs undergoing apoptosis. Paired box gene 5 (Pax5), sex determining region Y-box 2 (Sox2), delta-like1 (Dll1) and Neurogenin2 (Neurog2) emerged as candidate molecules downstream of DISC1, and rescue experiments demonstrated that increased or decreased expression of either molecule regulated proliferation and migration in DISC1-altered MNSPCs. Conclusion: These results suggest that Pax5, Sox2, Dll1 and Neurog2 mediate DISC1 activity in MNSPC proliferation and migration.

Despite our understanding of the DISC1 interactome (Camargo et al., 2007), few studies have focused on the role of DISC1 in the proliferation, migration, cell cycle progression and apoptosis of mouse neural stem/progenitor cells (MNSPCs). In this investigation, embryonic mouse hippocampal NSPCs were used as an in vitro cellular model to overexpress or silence DISC1 with retroviral expression vectors or siRNAs, and the biology of these altered MNSPCs was examined in a series of functional assays. To further understand the mechanisms of DISC1 regulated neurogenesis, a neurogenesis quantitative polymerase chain reaction (PCR) array was used to identify potential genes or signaling pathways associated with altered expression of DISC1. Paired box gene 5 (Pax5), sex determining region Y-box 2 (Sox2), Delta-like1 (Dll1) and Neurogenin2 (Neurog2) were among the candidate genes identified and were further examined in functional assays to determine whether they mediated DISC1 activity.

Ethics Statement
Animal protocols were approved by the ethics committee associated with Central South University (Changsha, Hunan, China). All experiments were performed in accordance with official recommendations by the Chinese animal community, and efforts were made to minimize the number of animals used and animal suffering.

Mice
C57/BL6 pregnant female mice (n = 4; P 14.5; Slac Laboratory Animal Co. Ltd, Shanghai, China) were used for the isolation and culture of MNSPCs.

Cell Culture
MNSPCs were isolated from the hippocampus of C57/BL6 fetal mouse brains as detailed in previous experiments (Yoneyama et al., 2011;Louis et al., 2013). Pregnant female mice (P 14.5) were anesthetized by i.p. injection with 10% chloral hydrate (0.1-0.2 mL/10 g), then the fetal mice were isolated from the uterus. The hippocampi of fetal mice were isolated using a dissecting microscope (Olympus). The hippocampi were washed in sterile cold phosphate buffered saline (PBS) in a 10 cm petri dish and cut into pieces using tissue scissors. The tissues were blown gently into smaller pieces using a 1 mL pipette, transferred into a 15 mL centrifuge tube, then D-HBSS was added, mixed into 10 mL suspension, and the tissues were incubated in room temperature (RT; 20-25 • C) for 20 min with 1 mL ACCUTASE (A6964, Sigma). The tissue chunks were then spun down in a low-speed centrifuge at 700 g for 1 min at RT, and the supernatant was removed and transferred to a new 15 mL centrifuge tube. We determined the number of cells using a blood cell counting plate, seeded the cells at a density of 1 × 10 5 /mL in a petri dish, and cultured the cells in NeuroCult TM Prol iferation Kit (Stem Cell; Vancouver, BC, Canada) to keep them in an undifferentiated, proliferative state.

siRNA Interference
Five siRNAs for each gene were designed against the functional sequences of DISC1, Pax5, Sox2, Dll1 and Neurog2 and sense and antisense sequences were chemically synthesized (Sigma). MNSPCs were seeded at a density of 2 × 10 5 /mL in 24-well plates, and after 24 h were transfected with siRNAs at a concentration of 100 nM in Lipofectamine 2000 Reagent (Invitrogen, New York, NY, USA). The medium was replaced 48 h later, and the efficiency of gene knockdown was evaluated using real-time PCR (RT-PCR; Supplementary Figure S1).

Packaging of Retroviral Vectors and Determination of Titers
A full-length mouse DISC1 sequence was ligated into the retroviral vector pCAG-EGFP (a generous gift from Dr. Chunmei Zhao, Laboratory of Genetics, The Salk Institute for Biological Studies, La Jolla, CA, USA; also available for purchase from Addgene; Cambridge, MA, USA), linearized with Pmel and Sfil restriction sites to generate the DISC1 expression construct CAG-DISC1-EGFP. The DISC1 retroviral construct and pCAG-EGFP were packaged into 293T cells with packaging plasmids pCMV-vsv-g and pCMV-gp (Addgene) and titered (as high as 10 8 c.f.u./mL) as previously described (Tashiro et al., 2006;Zhao et al., 2006;Jessberger et al., 2007).

Multiplicity of Infection (MOI)
Viral titers were serially diluted (10 4 , 10 5 and 10 6 ) and applied to MNSPCs to determine the multiplicity of infection (MOI) prior to experimentation. MOI was assessed under fluorescence microscopy with expression of EGFP encoded by the retroviral vector. More than 60% of the MNSPCs were EGFP-positive after incubation with virus for 4 days at 10 4 c.f.u./mL and an MOI of 100.

Infection of Cells
MNSPCs (2 × 10 5 /mL) were plated, and the medium was replaced 24 h later with 0.5 mL of 10 4 c.f.u./mL diluted, filtered virus containing polybrene at a concentration of 6 µg/mL (Sigma; St. Louis, MO, USA). The medium was changed after 2 days, and EGFP-positive colonies were collected for experiments.

Transfection of Cells by Electroporation
Expression plasmids for Pax5 (pPax5), Sox2 (pSox2), Dll1 (pDll1) and Neurog2 (pNeurog2) were constructed. One hundred microliter of cell suspension (10 7 cells/mL) was pipetted into a 4-mm electroporation cuvette and plasmids were added to a final concentration of 100-200 nM (unless otherwise specified) immediately before electroporation. Electroporation was performed with a Fischer electroporator (Fischer; Heidelberg, Germany) with a rectangle pulse of 330 V for 10 min. Cells were incubated for 15 min at RT, diluted 20-fold with culture medium, and incubated at 37 • C in 5% CO 2 .

Real-Time PCR (RT-PCR)
RT-PCR was used to assess interference efficiency 48 h after cells were transfected with retrovirus vector, plasmids, or siRNAs. Total RNA was isolated with Trizol Reagent per the manufacturer's protocols (Life Technologies; Grand Island, NY, USA). The samples were incubated with 1 µL DNase I (Life Technologies) for 30 min at 37 • C to remove contaminating genomic DNA. cDNA was synthesized from RNA (200 ng) in a 20 µL reverse transcription reaction (Promega; Madison, WI, USA). RT-PCR was performed with SYBR Green qPCRSuperMix (Life Technologies) with the recommended program of one cycle at 95 • C for 15 min, followed by 40 cycles at 95 • C for 30 s and 60 • C for 30 s on the ABI PRISM 7500 Sequence Detection System. Primers used for RT-PCR of DISC1, Pax5, Sox2, Dll1 and Neurog2 are listed in Table 1.
Relative expression levels were calculated as ratios normalized with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) expression. The RT-PCR data were analyzed using the ∆Ct method (Wang et al., 2010), and results were determined as the mean ± SD of three independent experiments. Danvers, MA, USA), rabbit anti-Dll1 (1:500, Abcam), and rabbit anti-Neurog2 (1:10,000, Abcam)), followed by incubation with a conjugated secondary antibody (goat anti-rabbit IgG-HRP (1:10,000; Thermo Scientific; Pittsburgh, PA, USA)). Membranes were rinsed three times for 10 min after each incubation with Tris-buffered saline (TBS) containing 0.1% Tween-20. Protein bands were visualized using ECL reagent (Thermo Scientific), and densitometric analysis of protein bands was performed with Image-Pro Plus 6.0 software. All western blotting data are representative of at least three independent experiments.

Flow Cytometric Detection of Apoptosis
Trypsinized cells (without EDTA) were rinsed twice in PBS and re-suspended in 1x binding buffer at a concentration of 1 × 10 6 cells/mL. Cells (100 µL of solution containing ∼1 × 10 5 cells) were transferred to a 5 mL culture tube, and Annexin V-FITC antibody (5 µL) and propidium iodide (5 µL; BD Biosciences; San Jose, CA, USA) were added and incubated for 15 min at RT in the dark. Binding buffer (1×; 400 µL) was added, and the samples were analyzed by flow cytometry within 1 h on the FACSCalibur (Becton Dicknson; Franklin Lakes, NJ, USA).

Cell Cycle Analysis
Cells were harvested, rinsed twice with PBS, and re-suspended in the residual PBS. Fixation was performed with cold 75% ethanol (10 mL) added dropwise while vortexing, and cells were incubated at −20 • C for a minimum of 2 h. Cells were pelleted by centrifugation, rinsed twice with PBS, and adjusted to a final concentration of 1 × 10 7 cells/mL. Aliquots (100 µL; 1 × 10 6 cells) of the cell suspension were then transferred into 12 × 75 mm tubes and incubated with PI-PBS (500 µL; 50 µg/mL PI, 100 µg/mL RNase A, 0.2% Triton X-100) after gentle vortexing for 30 min at 4 • C. Samples were analyzed by flow cytometry on the FACSCalibur (Becton Dicknson, Franklin Lakes, NJ, USA), and data were analyzed with ModFit LT software (Becton Dickinson, Franklin Lakes, NJ, USA).

Transwell Migration Assay
Trypsinized cells suspended in serum-free DMEM medium were seeded onto untreated BD Falcon Cell Culture Inserts (8.0 µm Polyethylene terephthalate (PET) membranes; 0.8 × 10 6 pores 1 cm 2 ) at a density of approximately 2.5 × 10 4 cells/cm 2 . The inserts were placed in the wells of BD Falcon multi-well plates containing 600 µL 10% FBS DMEM and incubated at 37 • C in 5% CO 2 . After 24 h and 48 h, the cells were wiped off of the inserts with a cotton swab. The remaining cells migrating to the other side of the membrane were fixed with 4% paraformaldehyde (PFA) for 15 min and stained with Crystal Violet for 10 min, and then the average number of those cells was calculated to evaluate their migrated ability. Images were captured with the Olympus microscope.

Mouse Neurogenesis PCR Array
The Mouse Neurogenesis RT 2 Profiler TM PCR Array (PAHM 404Z, Qiagen, Shanghai, China) was used to profile 84 genes related to the process of neurogenesis (Supplementary Table S1). RNA was isolated with the RNasy Mini Kit and RNase-Free DNase Set (Qiagen; Valencia, CA, USA). The RT 2 First Strand Kit (Qiagen) was used to synthesize cDNA, and RT-PCR was performed using SYBR Green Master mixes with the arrays on the Applied Bio-systems 7500HT system per the manufacturer's instructions. Expression of each gene was normalized against the following endogenous control genes: β-actin (Actb), β2 microglobulin (B2m), GAPDH, β-Glucuronidase (Gusb), and heat shock protein 90 α, cytosolic, class B (Hsp90ab1). The Ct values of all the wells were exported to a blank Excel spread sheet and analyzed with the SABiosciences (Qiagen) PCR Array Data Analysis Web-based software (RT 2 Profiler PCR Array Data Analysis version 3.5).

Statistical Analysis
All statistical analyses were performed with Statistical Package for the Social Sciences (SPSS Inc. 2007, version 16.0; Chicago, IL, USA). Results are expressed as the mean ± SD. Data were analyzed with Student's t-test, and p-value < 0.05 were considered statistically significant.
To determine whether the siRNAs (negative control siRNA (NCsiRNA) and DISC1siRNA) and expression constructs (CAG-EGFP and CAG-DISC1) affected protein levels, DISC1 protein was examined by Western blotting. RT-PCR and Western blotting were performed to detect the expression of DISC1 48 h after transfection with siRNAs or the expression constructs. siRNA led to decreases in DISC1 mRNA (Figure 2A, Left) and protein (Figure 2A, Right) levels by 53.2-fold and 2.9-fold, respectively, compared to controls. In contrast, transfection of the retroviral vector CAG-DISC1 increased mRNA (Figure 2B, Left) and protein (Figure 2B, Right) expression by 32.5-and 1.7-fold respectively. Changes in protein levels demonstrated that protein function may be altered as a result of transfection with siRNA or the expression construct.

DISC1 Promotes Proliferation of MNSPCs
To examine the role of DISC1 on MNSPC proliferation, MTS assays were performed in cells transfected with siRNAs or expression construct. Proliferation of MNSPCs significantly diminished after siRNA knock-down of DISC1, whereas proliferation increased with increased DISC1 expression. We observed that siRNA inhibited proliferation of MNSPCs over time by 5.7%, 22% and 23.4% after 24 h, 48 h and 72 h respectively. In contrast, increased expression of DISC1 promoted cell growth by 10.6%, 15.2% and 19.6% after 24 h, 48 h, and 72 h, respectively (Figures 3A-C).
These results indicate that the DISC1 gene had a role in promoting growth of MNSPCs in a time-dependent manner.

DISC1 Promotes MNSPC Migration
To determine how DISC1 influences migration, the ability of transfected MNSPCs to cross the Transwell membranes was examined. The average number of MNSPCs that migrated across the membrane after siRNA administration was fewer relative to controls (53.5 vs. 90.7 cells; p < 0.05; Figures 4A,B). DISC1 overexpression, however, led to increased MSPNC migration relative to the control group (53.7 vs. 34.8 cells; p < 0.01; Figures 4C,D). These results show that DISC1 mediates MNSPC.

DISC1 Influences Apoptosis and Cell Cycle Progression of MNSPCs
Apoptosis is an important cell function inhibiting cell growth. Flow cytometry was used to determine how DISC1 expression influences apoptosis in MNSPCs. Based on the total number of apoptotic events calculated in the four MNSPC groups, only DISC1siRNA altered apoptosis in MNSPCs. The percentage of MNSPCs undergoing apoptosis was greater with transfection of DISC1siRNA (7.5%) than NCsiRNA, CAG-EGFP and CAG-DISC1 (5%-6%; Figures 5A-C). These results suggest that loss of DISC1 only slightly increases apoptosis of MNSPCs.
Interestingly, both decreased and increased DISC1 levels altered cell cycle progression. Decreased expression of DISC1 inhibited cell growth as the average percentage of cells in G0/G1 phase increased from 44.2% to 64.2% (p < 0.01) and those in the S phase decreased from 39.7% to 22.3% at 48 h (p < 0.01). Cells in the G2/M phase had a decrease from 16.2% to 13.5% (p < 0.01; Figures 5D,E,G). In contrast, increased expression of DISC1 reduced the average percentage of cells in the G0/G1 phase (43.5% to 41.5%; p < 0.01) as well as in the S phase (39.6% to 32.3%; p < 0.01) relative to controls. The proportion of cells in the G2/M fraction, however, increased from 16.9% to 26.2% (p < 0.01; Figures 5D,F,G). These results show that alterations in DISC1 levels affect nearly all phases of the cell cycle in MNSPCs, with DISC1 reduction resulting in some form of G1 arrest.

MNSPC Migration after Pax5, Sox2, Dll1 or Neurog2 Rescue Experiment
Rescue experiments were performed to investigate whether DISC1 regulates MNSPC migration via Pax5, Sox2, Dll1 or Neurog2. In DISC1siRNA MNSPCs, down-regulation of Pax5 by siRNA decreased the average number of migratory MNSPCs compared to the control (15.5 vs. 24.7; p < 0.01), contrary to our initial expectations. In CAG-DISC1 MNSPCs, up-regulation of Pax5 increased the average number of migratory MNSPCs compared to the control (30.2 vs. 21; p < 0.01). These results suggest that Pax5 regulates the migration of MNSPCs independent of DISC1 (Figures 4E-H).

DISCUSSION
We used an in vitro model to characterize the function of DISC1 during neural development. We show that DISC1 promotes the proliferation and migration of MNSPCs, and loss of this protein profoundly alters cell cycle progression does not significantly alter apoptosis. Furthermore, experiments with knockdown or overexpression of Pax5, Sox2, Dll1 and Neurog2 suggest that DISC1 regulates the MNSPC proliferation and migration through one or more of these transcription factors.
Our observation that DISC1 positively regulates the proliferation and migration of MNSPCs corroborates previous studies performed in vivo on the embryonic and developing brain (Brandon and Sawa, 2011;Ishizuka et al., 2011). Neuronal proliferation and migration are essential to the process of neurogenesis and neural development (Tomita et al., 2011). DISC1 mediates the proliferation of neuronal progenitors in the developing cortex through the Wnt/β-catenin pathway (S) crystal violet, CAG-DISC1 (control); CAG-DISC1 + Neurog2siRNA; (T) bar graph, CAG-DISC1 (control) vs. CAG-DISC1 + Neurog2siRNA. * p < 0.05, * * p < 0.01 (controls-NCsiRNA, CAG-EGFP, DISC1siRNA + Con, CAG-DISC1 + Con). All results are from three independent experiments. (Mao et al., 2009), thus the phosphorylation of DISC1 may be a key molecular event from maintaining the proliferation of mitotic progenitor cells to activating the migration of post-mitotic neurons in the developing cortex (Ishizuka et al., 2011). However, the function of DISC1 on neuronal migration is complicated. DISC1 deficiency generally results in migration defects. For example, DISC1 knockdown results in migration delay in the developing neocortex (Duan et al., 2007;Kubo et al., 2010), but, DISC1 has opposite effects on neuronal migration in the developing vs. adult hippocampal neurons. While DISC1 deficiency delays migration in the developing hippocampal neurons, loss of DISC1 induces inappropriate migration in newly generated neurons in the adult hippocampus (Duan et al., 2007;Enomoto et al., 2009;Kim et al., 2009). DISC1 function also affects migration in a temporal manner. DISC1 knockdown does not alter migration of pyramidal cells in the developing CA1 at postnatal day 2 (P2; Meyer and Morris, 2009), whereas migration appears to be especially sensitive to DISC1 suppression from P2 to P3 (Meyer and Morris, 2009;Tomita et al., 2011). Moreover, DISC1 is known to regulate migration in a cell type specific manner. Loss of function of DISC1 hinders granule cell migration in the embryonic hippocampus, but fails to influence neuronal migration of pyramidal neurons in CA1 (Meyer and Morris, 2009), and DISC1 seems to have no impact on postnatal neuroblast migration within the rostral migratory stream (RMS; Wang et al., 2011). Further complicating the issue is an underlying difference in mouse strains due to a spontaneous 25-base pair (bp) deletion of DISC1. Swiss Webster, ICR, and 129S6/SvEv mice, for example, lose production of some specific isoforms of DISC1, which leads to a reduced or no apparent effect on neuronal migration, compared to C57BL/6 mice without the deletion (Ishizuka et al., 2006;Koike et al., 2006;Meyer and Morris, 2009). Taken together, the influence of DISC1 on neuronal migration seems to be affected by temporal and spatial differences, cell type specificity, and animal model. Although DISC1 is known to interact with several protein partners and extracellular/intracellular signaling pathways to co-regulate mainly neuronal proliferation and migration ( Figure 1B; Kamiya et al., 2005;Duan et al., 2007;Namba et al., 2011), the exact molecular mechanisms remain to be elucidated.
In the present study, dysregulation of DISC1 strongly affects cell cycle progression of MNSPCs but only moderately affects apoptosis (Figure 5). Expression of mutated DISC1 has been previously associated with abnormalities in the expression of cell cycle genes and oligodendroglial progenitor markers (Katsel et al., 2011). DISC1 suppression reduces MNSPC proliferation and leads to premature exit from the cell cycle and differentiation (Mao et al., 2009). In addition, a partner of DISC1, the FEZ1, has been associated with the cell cycle and apoptosis in neuronal development (Assmann et al., 2006). A role for DISC1 in the cell cycle and apoptosis is observed in diseases of the central nervous system (CNS). For example, mitochondria involved in apoptosis are the predominant site of DISC1 expression, suggesting that the effect of DISC1 on apoptosis is related to neuronal disturbances in the pathogenesis of SZ . Furthermore, DISC1 is shown to be a target of miR-181b, a miRNA which may have a role in the pathogenesis of SZ (Carroll et al., 2013). Therefore, understanding the role of DISC1 in the cell cycle and apoptosis in different cell types requires further investigation.
Expression array analysis reveals Sox2, Pax5, Dll1 and Neurog2 as potential downstream or partner mediators of DISC1 (Figures 6A,D,H,I,L,O,S,T). Overexpression of the gene leads to increased Sox2 and Neurog2 but decreased Pax5 and Dll1 levels while suppression of DISC1 induces the opposite expression profile (Figures 2C-F). Therefore, we propose that DISC1 regulates proliferation and migration through one or more of these transcription factors. The results of Sox2, Dll1 and Neurog2 interference in functional assays parallel DISC1 expression, hence Sox2, Dll1 and Neurog2 appear to be necessary for proliferation and migration induced by DISC1 (Figures 3G-O, 4I-T). However, alterations in Pax5 expression yielded unexpected results in the functional assays, namely that increased Pax5 also promoted proliferation and migration rather than inhibiting these functions in response to DISC1 activity (Figures 3D-F, 4E-H). These results suggest that: (1) DISC1 regulates the proliferation and migration of MNSPCs through Sox2, Dll1 and Neurog2 and that exogenous intervention of Sox2, Dll1 and Neurog2 may help maintain the ability of MNSPCs to proliferate and migrate; and (2) Pax5 regulates the proliferation and migration of MNSPCs independent of DISC1, and endogenous changes in Pax5 levels may be a protective factor against DISC1 dysfunction, and thus Pax5 may act as a partner or co-regulator of DISC1.
Sox2 is a transcription factor that is expressed in MNSPCs and mature neurons in the mouse CNS and is necessary to maintain self-renewal or pluripotency of undifferentiated embryonic stem cells (Fong et al., 2008). Activated Sox2 promotes cell proliferation and differentiation in diverse cell types (Ferri et al., 2004;Que et al., 2009;Tompkins et al., 2011), and changes in levels of Sox2 transcription leads to abnormalities in the growth and development of stem cells (Bani-Yaghoub et al., 2006;Suh et al., 2007). In this study, Sox2 was also found to promote proliferation (Figures 3G-I) and migration (Figures 4I-L) of MNSPCs. Taken together, the data suggest that DISC1 regulates the proliferation and migration of MNSPCs directly or indirectly through Sox2.
Pax5 was originally identified as the B-cell-specific activator protein (BSAP; Adams et al., 1992), a transcription factor expressed at early but not late stages of B lymphocyte differentiation (Urbánek et al., 1994;Eberhard and Busslinger, 1999). Although expression of this gene was originally associated with proliferation of immune cell types, the phenotype of knock-out mice reveals a role for Pax5 in neural development. Transient expression of Pax5 occurs during embryogenesis in the mesencephalon and spinal cord in a spatial and temporal manner (Adams et al., 1992). Functionally, Pax5 over-expression reduces migration and motility of invasive MDA-MB231 cells, whereas the depletion of endogenous Pax5 enhances the migration of MCF-7 cells (Vidal et al., 2010). Moreover, Pax5 modulates the transcription of genes involved  in the migration and adhesion of B cells, and promotes the adhesion of intercellular junctions in tumor cells (Schebesta et al., 2007). Our study found that Pax5 regulates the proliferation (Figures 3D-F) and migration (Figures 4E-H) of MNSPCs with DISC1. These findings suggest that Pax5 may stimulate proliferation and migration in a variety of cell types.
Notch ligand Dll1 may also be associated with DISC1regulated neurogenesis. In the developing brain, steady Dll1 expression inhibits the proliferation of neural progenitors and accelerates neurogenesis (Shimojo et al., 2016), whereas oscillatory expression of Dll1 during embryonic development of the mammalian telencephalon is critical in neural fate decision (Barton and Fendrik, 2013) and cell differentiation in neural progenitor cells (Formosa-Jordan et al., 2012, Barton andFendrik, 2013) through Notch signaling. In this study, Dll1 was found to regulate proliferation (Figures 3J-L) and migration (Figures 4M-P) of MNSPCs. Taken together, the data suggest that DISC1 regulates the proliferation and migration of MNSPCs directly or indirectly through Dll1.
Neurog2 is a member of the neurogenin subfamily of basic helix-loop-helix (bHLH) transcription factor genes that plays an important role in neurogenesis (Heng et al., 2008). Neurog2 is involved in the regulation of proliferation (Vied et al., 2014), migration, and apical progression (Kawaue et al., 2014) of the ventricular zone during neocortical development, as well as the control of successive steps of neurogenesis in the embryonic cerebral cortex (Azzarelli et al., 2014;Li et al., 2014). In this study, we found that Neurog2 also regulates proliferation (Figures 3M-O) and migration (Figures 4Q-T) of MNSPCs with DISC1, but how they work together requires further investigation.
The results from the expression array further suggest Pax3, Sox3, Dvl3 and Neurog1 as potential partners/mediators of DISC1 in regulating the proliferation or migration of MNSPCs. These observations provide new molecular clues for future studies of the ''DISC1 interactome''. Pax3 mediates the migration of myoblasts in the developing limb bud through a signaling pathway which includes the activation of the c-Met (MET or Methylnitronitrosoguanidine HOS Transforming gene) receptor (Bottaro et al., 1991;Galland et al., 1992;Bober et al., 1994;Goulding et al., 1994;Bladt et al., 1995;Epstein, 1996). In addition, Pax3 is required for fine-tuning migration behavior of cardiac neural crest cells, but the protein is not essential for regulating their migration (Epstein et al., 2000). Pax3 was also found to be necessary for migration rather than differentiation in limb muscle precursors of mice (Daston et al., 1996). Although the exact role of Pax3 in MNSPC migration is unclear, Pax3 is considered as a candidate gene associated with DISC1 function based on our study.
Sox3 belongs to the high mobility group (HMG) family of transcription factors and is expressed in neural progenitor cells throughout the developing CNS. The protein is essential for the proliferation of neuroepithelial precursors (Rizzoti et al., 2004) and existing progenitors (Rogers et al., 2009). Furthermore, Xnr5, an early zygotic gene regulated by β-catenin/VegT, is a direct target of Sox3, suggesting that Sox3 mediates cell proliferation through Wnt signaling pathways (Zhang et al., 2003). Taken together, DISC1 appears to shape an intricate network associated with Sox2, Sox3 and Wnt signaling pathways which regulate cell proliferation.
Dvl3 is a Wnt signaling related gene whose mRNA levels changed in response to DISC1 in our study. A previous study reports that disheveled (Dvl; three isoforms) not only promotes β-catenin signaling in the Wnt pathway by stimulating GSK-3β phosphorylation but also promotes AKT signaling pathways by facilitating AKT phosphorylation. DISC1 participates in the WNT signaling pathway and in WNT-mediated cell proliferation, either in primary neural stem/progenitor in vitro or in embryonic brains (Mao et al., 2009). In addition to Dvl3, Pafah1b1 (also named lis1) has been shown to participate in neurodevelopmental processes such as neuronal precursor proliferation and differentiation, neuronal migration and neurite outgrowth, possibly in together with DISC1 through direct or indirect interactions. However, the exact mechanism remains to be defined (Lipska et al., 2006;Bradshaw et al., 2008Bradshaw et al., , 2011Rastogi et al., 2009;Bradshaw and Porteous, 2012). Additionally, Neurog1 is shown to influence the differentiation and migration of neural crest stem/progenitor cells and fate specification of embryonic stem cells (Velkey and O'Shea, 2013). Whether Neurog1 regulates neurogenesis via DISC1 requires further investigation.

CONCLUSION
Our findings reveal an important role for DISC1 in neurogenesis. First, DISC1 is required for the proliferation and migration of MNSPCs: DISC1 deficiency results in proliferation and migration defects, while DISC1 overexpression promotes both activities. Second, DISC1 regulates the proliferation and migration of MNSPCs in vitro with Pax5, Sox2, Dll1, and Neurog2, which may act as downstream mediators of DISC1 activity. Third, Pax3, Sox3, Dvl3, Neurog1 are potential partners of DISC1 in the regulation of neurogenesis. DISC1 appears to be a crucial regulator not only of neural development but also of specific neurogenesis-related genes/molecules and signaling pathways. Further investigation of the molecular components of the DISC1 regulatory complex, as well as the nature of more detailed molecular interactions of DISC1 with Pax5, Sox2, Dll1 and Neurog2 in regulating migration, proliferation and the processes of neurogenesis such as differentiation is needed.

AUTHOR CONTRIBUTIONS
QW: conception and design, collection and assembly of data, data analysis and interpretation, manuscript writing, final approval of manuscript, provision of study material. WT: conception and design, collection and assembly of data, data analysis and interpretation, final approval of manuscript. ZL: conception and design, data analysis and interpretation, final approval of manuscript. YL, YS and ZY: conception and design, data analysis and interpretation. BX and LF: conception and design, financial support, final approval of manuscript.

SUPPLEMENTARY MATERIAL
The Supplementary Material for this article can be found online at: http://journal.frontiersin.org/article/10.3389/fncel.2017.002 61/full#supplementary-material FIGURE S1 | mRNA levels of Disrupted-in-schizophrenia 1 (DISC1), Paired box gene 5 (Pax5), and sex determining region Y-box 2 (Sox2) knocked down with siRNA. (A-E) siRNAs designed against each gene were transfected into cells and assessed for efficiency of knockdown by Real-time-polymerase chain reaction (RT-PCR); (F) Sense and anti-sense for each gene. All results are from three independent experiments. * p < 0.05, * * p < 0.01.