Sox6 Up-Regulation by Macrophage Migration Inhibitory Factor Promotes Survival and Maintenance of Mouse Neural Stem/Progenitor Cells

Macrophage migration inhibitory factor (MIF) has important roles in supporting the proliferation and/or survival of murine neural stem/progenitor cells (NSPCs), but downstream effectors remain unknown. We show here that MIF robustly increases the expression of Sox6 in NSPCs in vitro. During neural development, Sox6 is expressed in the ventricular zone of the ganglionic eminence (GE) of mouse brains at embryonic day 14.5 (E14.5), cultured NSPCs from E14.5 GE, and NSPCs in the subventricular zone (SVZ) around the lateral ventricle (LV) of the adult mouse forebrain. Retroviral overexpression of Sox6 in NSPCs increases the number of primary and secondary neurospheres and inhibits cell differentiation. This effect is accompanied with increased expression of Hes1 and Bcl-2 and Akt phosphorylation, thus suggesting a role for Sox6 in promoting cell survival and/or self-renewal ability. Constitutive activation of the transcription factor Stat3 results in up-regulation of Sox6 expression and chromatin immunoprecipitation analysis showed that MIF increases Stat3 binding to the Sox6 promoter in NSPCs, indicating that Stat3 stimulates Sox6 expression downstream of MIF. Finally, the ability of MIF to increase the number of primary and secondary neurospheres is inhibited by Sox6 gene silencing. Collectively, our data identify Sox6 as an important downstream effector of MIF signaling in stemness maintenance of NSPCs.


Introduction
The Sox protein family of transcription factors has been identified as a major group of developmental regulators in vertebrates and invertebrates [1]. Sox transcription factors induce or suppress progenitor cell properties, such as proliferation and multipotentiality, or initiate differentiation programs by activating the expression of cell type-specific genes. The Sox family is comprised of 20 genes, classified into 8 groups (A to H), which encode transcription factors with a high-mobility-group (HMG) box DNA-binding domain highly similar to that of the sexdetermining region (Sry) protein [2]. Sox2, which is a SoxB protein, is a mandatory maintenance factor in neural/stem progenitor cells (NSPCs) in fetal and adult mouse brains [3], [4], [5]. However, the detailed function of most Sox genes in the developing nervous system and in NSPCs remains to be elucidated.
Sox6 belongs to the SoxD family, along with Sox5 and Sox13. SoxD proteins harbor two highly conserved functional domains: the family-specific HMG box DNA-binding domain and a group-specific coiled-coil domain that mediates homodimerization [6]. They have no known transactivation or transrepression domain, but participate in transcriptional activation and repression by utilizing various cofactors to modulate cell proliferation, survival, differentiation, and terminal maturation in a number of mesoderm-, ectoderm-, and endoderm-derived cell lineages [7]. Sox6 contributes to erythropoiesis and chondrogenesis and Sox6 knockout mice die soon after birth, presumably from cardiac malformation [8], [9]. In addition, Sox6 inhibits terminal differentiation of oligodendrocytes [10] and contributes to the specification of diverse types of neurons in vitro and in vivo [11], [12], [13] and favors the differentiation of rat NSPCs in vitro towards the astrocyte rather than neuronal lineage [14]. However, the mechanisms underlying Sox6 expression and the exact functions of Sox6 in NSPCs remain underexplored.
We previously reported that SOX6 is highly expressed in human gliomas [15] and that Macrophage migration inhibitory factor (MIF) supports the proliferation and/or survival of murine NSPCs in vitro [16]. Here we show that Sox6 is expressed in the ventricular zone of the ganglionic eminence (GE) of E14.5 mouse fetal brains, and in the subventricular zone (SVZ) around the lateral ventricle (LV) of mouse adult forebrains, where NSPCs are located. We also show that Sox6 expression is increased by MIF in NSPCs in vitro, and that this effect is mediated by the transcription factor Stat3. Further, we demonstrate an important role for Sox6 in supporting the viability and self-renewal ability of NSPCs under the control of MIF.

Animals
All interventions and animal care procedures were performed in accordance with the Laboratory Animal Welfare Act, the Guide for the Care and Use of Laboratory Animals (National Institutes of Health, USA), and the Guidelines and Policies for Animal Surgery provided by the Animal Study Committee of the Keio University and were approved by the Animal Study Committee of Keio University (IRB approval number 12017-0). Pregnant C57BL/6J mice were purchased from Sankyo Labo Service (Tokyo, Japan). Sox6 knockout mice [9] were maintained on a 129 background.

Neurosphere Culture
NSPCs were isolated from mouse E14.5 GEs and the cells were cultured as neurospheres [17] at a cell density of 50 cells/ml in neurosphere culture medium (NSP medium) consisting of neurobasal medium (Invitrogen, Carlsbad, CA, www.invitrogen. com) supplemented with B27 (Invitrogen), human recombinant (hr) EGF (20 ng/ml; Peprotech, Rocky Hill, NJ, www.peprotech. com), and hrFGF2 (10 ng/ml; Peprotech). Neurosphere formation assays were performed as described previously [16]. In the experiments using NSPCs, mouse recombinant MIF (R&D systems, Mineapolis, MN, www.rndsystems.com) and MIF inhibitor ISO-1 (Calbiochem, La Jolla, CA, www.merckmillipore.com) were used. Human NSPCs were cultured as reported by Hattori Y et al., [18]. The study using human NSPCs was carried out in accordance with the principles of the Helsinki Declaration, and the Japan Society of Obstetrics and Gynecology. Approval to use human fetal neural tissues was obtained from the ethical committees of both Osaka National Hospital and Keio University. Written informed consent was obtained from all parents through routine legal terminations performed at Osaka National Hospital.

RNA Extraction and Quantitative (q) RT-PCR
Total RNA was isolated from tissues or cultured cells using TRIZOL (Invitrogen). Total RNA (0.5 m g) was subjected to the cDNA Synthesis using ReverTra Ace H qPCR RT Master Mix with gDNA Remover (Toyobo, Osaka, Japan, www.toyobo.co.jp). Quantitative RT-PCR analysis was performed with a FastStart Universal SYBR Green Master (Roche, Tokyo, www.app.rochebiochem.jp), using the ABI prism 7900 HT Sequence Detection System (Applied Biosystems, Life Technologies, Carlsbad, www. appliedbiosystems.com). The PCR conditions were as follows: 1 cycle of 5 min at 95uC, followed by 40 cycles of 95uC for 30 sec, 60uC for 60 sec, and 72uC for 30 sec. PCR reactions were performed in triplicate. Relative gene expression levels were determined using the DDCt-method. GAPDH mRNA levels were used as the internal normalization control. The primer sequences are listed in Table.S1 and the sequence for Hes1, Hes3, and GAPDH primers were described in the previous study [16].

Luciferase Assay
NSPCs were plated at a density of 5,000 cells/well on 96-well plates. 100 ng of a pGL3 reporter plasmid and 1 ng of a pRL-TK (Promega, Tokyo, Japan, www.promega.co.jp) internal control plasmid were co-transfected using the X-tremeGENE HP DNA Transfection Agent (Roche) according to the manufacturer's instructions. After 24 h of transfection, the cells were treated with MIF (400 ng/ml). Luciferase activity was measured after 24 h using the Dual-Glo Luciferase Assay System (Promega). Relative luciferase activity was calculated by dividing the firefly luciferase activity of the constructs by the Renilla luciferase activity of the tyrosine kinase promoter, pRL-TK (Promega).

Chromatin Immunoprecipitation
Cells were subjected to chromatin immuneprecipitation with the ChIP-IT kit (Active Motif, Carlsbad, CA, www.activemotif. com) using anti-Stat3 antibody (Cell Signaling technologies, Danvers, MA, www.cellsignal.com) or normal rabbit IgG following the manufacturer's protocol. After elution, samples were quantified using FastStart Universal SYBR Green Master and primers, which targets the upstream region of the starting point for Sox6 transcription (EpiTct ChIP qPCR Primers,

Cell Proliferation and Apoptosis Assay
Cell viability were assessed using Cell Titer-Glo Luminescent Cell Viability Assay kits (Promega) and a luminometer (En-Vision TM multilabel reader, Perkin Elmer, Waltham, MA, www. perkinelmer.com). Cell apoptosis was assessed using Caspase-Glo 3/7 Assay Kits (Promega). In both assays, single cells dissociated from neurospheres were seeded onto 96-well plates at a density of 5610 3 cells/well, and caspase activity was assayed 4 days postinfection.

Statistical Analysis
All values are expressed as mean6S.D. Student's t tests were used to determine the statistical significance of differences between groups (*P,0.05, **P,0.01).
(Sox6). The cells were cultured in the presence of both EGF and FGF2. In the secondary neurosphere assay, neurospheres infected with a Sox6expressing retrovirus or control for 5DIV were sorted as GFP-positive cells onto a 96-well plate and cultured in the presence of both EGF and FGF2. (B) Sox6 targeting using retroviral shRNA significantly attenuated the efficiency of both primary and secondary neurosphere formation. In the primary neurosphere assay, single dissociated neurospheres were infected with a retrovirus expressing either Luc-shRNA (shLuc) or Sox6-shRNA (shSox6). In the secondary neurosphere assay, neurospheres infected with a retrovirus expressing either Luc-shRNA or Sox6-shRNA for 5 DIV were dissociated into single cells and plated onto a 96-well plate for the culture in the presence of EGF and FGF2. (C) Schematic representation of the self-inactivating lentivirus vector expressing the Venus reporter gene under the control of the SOX6 promoter. Flow cytometric analysis of neurospheres infected with lentivirus expressing the Venus reporter regulated by a SOX6 promoter for 5 days. (D) Venus-positive cells (II) and Venus-negative cells (I) were sorted onto a 96-well plate and cultured in the presence of EGF and FGF2, and the number of secondary neurospheres was quantified. All neurosphere formation assays were performed at a low cell density (

Sox6 is a Downstream Target of MIF Signaling in NSPCs
Changes in the expression levels of the Sox genes upon treatment of NSPCs with MIF were examined by qRT-PCR. MIF treatment increased the RNA level of Sox6 in NSPCs (Fig. 1A). In addition, cell treatment with ISO-1, a MIF inhibitor, led to a decrease in Sox6 RNA level in a dose-dependent manner (Fig. 1B). Interestingly, MIF treatment increased the RNA level of Sox6, but not that of Sox1 and Sox2 (Fig. 1C). Together, these results suggest that the Sox6 gene is a downstream target of MIF signaling in NSPCs.

Sox6 is Expressed in Mouse Embryonic Brain NSPCs
The expression of the Sox6 protein in the mouse embryonic brain at E14.5 was examined by immunofluorescence. Sox6 was expressed in the ventricular zone of the cortex and the dorsal part of the lateral ganglionic eminence (LGE), as previously shown [12] ( Fig. 2A and B). The Sox6-positive cells also expressed Sox2, a   Sox6 Promotes Characteristics of NSPCs PLOS ONE | www.plosone.org known marker of NPSCs (Fig. 2C), and these cells were located more towards the ventricle compared to cells expressing Ascl1, a marker of neural progenitor cells, in the E14.5 GE (Fig. 2D). In the adult mouse forebrain, Sox6 was expressed in the subventricular zone (SVZ) (Fig. 2E). Some Sox6-expressing cells were positive for BrdU after short-term labeling, indicating that they were actively proliferating, and they were also positive for Sox2, proving that they corresponded to NSPCs (Fig. 2, E-G). In the sub-granular zone (SGZ) of the adult mouse hippocampus, Sox6 expression coincided with BrdU labeling expression (data not shown), indicating that Sox6 was expressed in NSPCs in the adult as it is in the embryo. Next, we generated neurospheres from E14.5 GEs and observed co-expression of Sox6 with Nestin, a marker of NSPCs (Fig. 2H). To examine the expression of Sox6 in differentiated cells, neurospheres were cultured without growth factors for 5 days in vitro (DIV). Besides a few exceptions, Sox6 expression was detected only in cells negative for differentiation markers, namely TuJ1 (neurons), GFAP (glia), and CNPase (oligodendrocytes) (Fig. 2, I-K). This result is consistent with a report showing that Sox6 is highly expressed in oligodendrocyte progenitor cells (OPCs) but not in post-mitotic differentiating oligodendrocytes [10]. Sox6 protein expression in NSPCs from 14.5 GE was also confirmed by Western blot (Fig. 2L). Finally, we compared the expression level of Sox6 to that of other Sox genes by qRT-PCR in NSPCs derived from E14.5 GE. The data clearly indicated that NSPCs contained higher levels of RNA for Sox2 and Sox6 than for their respective close relatives, Sox1 and Sox5 (Fig. 2M).

Sox6 Supports Cell Survival and the Self-renewal Ability of NSPCs
To identify the function of Sox6 in NSPCs, we first performed neurosphere-forming assays, a commonly used method to measure NSPC self-renewal. Retrovirus-mediated Sox6 overexpression led to a 1.760.6 fold increase in the number of primary neurospheres, and 3.261.1 fold increase in the number of secondary neurospheres (Fig. 3A). In contrast, Sox6 silencing by retroviral expression of shRNA-Sox6 in NSPCs attenuated the formation of primary and secondary neurospheres by 0.2360.25 fold and 0.6260.22 fold, respectively (Fig. 3B). This result was also supported in experiments using Sox6-null mice (Fig. S1). Changes in Sox6 protein levels in NSPCs by Sox6 overexpression and gene silencing were assessed by Western blot analysis (Fig. S2). Moreover, we performed the neurosphere-forming assay using a lentivirus-based Sox6 reporter system. A reporter plasmid was constructed using a self-inactivating lentivirus vector harboring the Venus reporter gene under the control of the human Sox6 promoter region, which contains elements highly conserved in humans and mice in terms of tissue-specific expression [21] (Fig. 3C). Venus-positive cells (II) and Venus-negative cells (I) were sorted onto a 96-well plate and then cultured in the presence of EGF and FGF2, and the number of secondary neurospheres was assessed (Fig. 3D). The specificity of the reporter assay was confirmed in human cells and mouse NSPCs by showing a strong correlation between Venus activity and Sox6 protein level (Fig.  S3). In this system, Venus-positive NSPCs formed 2.861.0 fold more secondary neurospheres than Venus-negative cells (Fig. 3D). Collectively, these data indicated that Sox6 supports the selfrenewal ability of NSPCs in vitro. We then asked whether the Sox6 promoter featured MIF responsive regions. The promoter contains two regions, called A-box (79 bp) and B-box (48 bp), which are 85 and 96% identical, respectively, in the rat, mouse, and human genomes [21]. NSPCs were transfected with luciferase reporters containing either a promoter fragment (517 bp) or tandem repeats of the A-box (4xA-box) or B-box (4xB-box) and were then treated with or without MIF (Fig. S4). The construct containing the A-box tandem repeats showed strong transcriptional activity, and this activity was increased 3.460.8 fold by MIF treatment, suggesting that the A-box contains MIF-responsive elements (Fig. S4). We also examined changes in NSPC viability upon Sox6 overexpression and silencing. Retrovirus-mediated overexpression of Sox6 in NSPCs increased cell viability by 2.160.12 fold 5 days after infection (Fig. 3E), whereas retrovirus-mediated silencing of Sox6 decreased cell viability by half (0.5060.02 fold) 4 days after infection (Fig. 3F). Sox6 knockdown also led to increased caspase3/7 activity 4 days after infection, confirming that Sox6 acts as a survival factor in NSPCs (Fig. 3G, Fig. S5).

Sox6 is a Maintenance Factor in NSPCs
Changes in the multi-lineage differentiation potential of NSPCs were examined upon Sox6 overexpression. NSPCs infected with a Sox6-expressing retrovirus were cultured for 5 days, and then replated and cultured for 5 additional days in the absence of growth factors. Then, the number of cells differentiated into each neuronal cell type was assessed by immunocytochemistry. In this assay, only GFP-positive cells, i.e., infected by the retrovirus, were examined. Sox6 overexpression led to a decrease in the number of cells differentiated into neurons (TuJ1), astrocytes (GFAP), and oligodendrocytes (CNPase) (Fig. 4, A and B). In addition, the expression of Nestin and Musashi-1, which are known as NSPC markers [22], [23], was also upregulated by Sox6 overexpression (Fig. 4, C and D). Thus, Sox6 functions as a maintenance factor for NSPCs in vitro. Consistent with this experiment, the numbers of neurons, astrocytes and oligodendrocytes were increased in NSPC cultures from Sox6-null mice (Fig. S6).

Downstream Signal of Sox6 in NSPCs
Hes1 is a transcription factor acting downstream of Notch signaling and well known for its ability to maintain the stemness of NSPCs [24]. Interestingly, Hes1 expression was increased upon treatment of NSPCs with MIF (Fig. 5A) as well as upon overexpression of Sox6 (Fig. 5B). Furthermore, expression of the Bcl-2 gene, which is important for cell survival and is regulated by MIF in NSPCs [16], also increased with Sox6 overexpression (Fig. 5C). This increase was confirmed at the protein level for Bcl-2 ( Fig. 5D) and found to occur concomitantly with an increase in Akt phosphorylation (1.560.11 fold) (Fig. 5E).

MIF-regulated Stat3 Controls Sox6 Gene Expression in NSPCs
We previously demonstrated that MIF treatment results in activation of the transcription factor Stat3 [16]. Hes3 expression, which is activated by Stat3-pSer727 in NSPCs [25], was also shown to be upregulated by MIF [16]. Thus, we examined the changes in gene expression of Sox6, Hes1, and Hes3 in NSPCs upon overexpression of constitutively active Stat3 in NSPCs, and found that the expression level of all genes was increased (Fig. 6A). Moreover, ChIP analysis showed an increase in Stat3 binding to the Sox6 promoter following MIF treatment of NSPCs (Fig. 6B). These data thus suggest that Sox6 is upregulated by MIF signaling in NSPCs through binding of Stat3 to its promoter.

MIF-mediated Up-regulation of Sox6 Supports the Selfrenewal Ability of NSPCs
To examine whether Sox6 plays a role in neurosphere formation in response to MIF, we performed neurosphere formation assays using cells treated with a retrovirus expressing Sox6-shRNA. While MIF treatment of control cells resulted in a larger number of primary (Fig. 7A) and secondary neurospheres (Fig. 7B), this effect was blunted by knockdown of Sox6, indicating that Sox6 is a downstream effector of MIF signaling in NSPCs. The same result was observed in a cell growth assay using lentivirally expressed Sox6 shRNA in NSPCs (Fig. S7).

Discussion
In the present study, we newly identified Sox6, which is expressed in the GE of the fetal mouse brain, as a maintenance factor for NSPCs. Moreover, we showed that Sox6 gene expression is upregulated by Stat3 downstream of MIF, a signaling molecule that supports the proliferation and/or survival of murine NSPCs.
To determine whether MIF can regulate Sox genes to maintain NSPC stemness, we analyzed change in gene expression levels of Sox1, Sox2 and Sox6 in response to MIF treatment in NSPCs. Sox2 is known as a maintenance factor for NSPC stemness in fetal and adult mouse brains [3]. Interestingly, in this assay system, MIF treatment increased Sox6, but not Sox1 or Sox2 gene expression. Consistent with this finding, treatment of NSPCs with ISO-1, a MIF inhibitor, decreased Sox6 gene expression, suggesting that Sox6 is a downstream molecule of MIF in NSPCs. However, it is possible that MIF may modulate Sox1 or Sox2 gene expression in other assay conditions and in NSPCs derived from different tissue types, including the cortex and spinal cord. Analysis by luciferase assay using a Sox6 promoter also showed that Sox6 gene transcription is upregulated by MIF in NSPCs, suggesting that Sox6 is a direct target of MIF signaling in this system.
We demonstrated that Sox6 overexpression can increase the survival and self-renewal ability of NSPC, and that Sox6 gene silencing in NSPCs has opposite effects. We observed an increase in Bcl2 expression and Akt phosphorylation in response to MIF treatment, both of which play an important role in NSPC cell survival [16], further suggesting that Sox6 is a survival factor acting downstream of MIF in NSPCs. In contrast, a previous report showed that SOX6 overexpression decreased cell proliferation in INS-1E insulinoma cells and NIH3T3 cells [26], indicating the possibility that Sox6 may regulate cell proliferation in a cell type-dependent manner. Furthermore, it is known that Sox proteins function in concert with partner proteins in a cell type-specific manners [27]. Thus, the identification of Sox6 partner proteins in NSPCs is a topic for future studies that would allow for an understanding of Sox6 protein mode of action.
We showed here that Sox6 is a maintenance factor for NSPCs in neurosphere-forming and differentiation assays in vitro, and observed up-regulation of Hes1 gene expression upon Sox6 overexpression, as also seen upon MIF treatment. Hes1 is known as an important factor for NSPC maintenance [24]. Although Hes3 was also found to be up-regulated by MIF [16], Hes1 has Sox6 binding elements in its promoter, while Hes3 does not. Thus, Sox6 may possibly activate Hes1 with other transcription factors, as seen in chondrocytes [28], while Hes3 may be regulated directly by Stat3 without the intervention of Sox6. In previous reports, Sox6 was shown to support neurogenesis and gliogenesis and to inhibit terminal differentiation of oligodendrocytes [10], [11], [14]. Scheel et al have reported that Sox6 overexpression induces astrocytic differentiation from rat hippocampal NSPCs [14]. Thus, it may be important to examine Sox6 function in NSPCs derived from different species and tissue types (e.g. adult SVZ of forebrain, adult hippocampus, fetal cortex), especially if different NSPC culture methods are used (neurosphere floating culture VS adherent culture). Additionally, it would also be interesting to examine the regulation of Sox6 function by MIF in different types of tissue stem cells, including hair follicle stem cells [29], [30]. Moreover, Sox6 was reported to regulate the differentiation of neural progenitor cells into different neural cell types in the mouse embryonic brain in vivo [12], [13]. In those in vivo studies, the expression patterns of many transcription factors changed in the mouse embryonic GE in response to the loss of Sox6, showing that Sox6 may play a role in controlling the maintenance of stemness of NSPCs via or together with several transcription factors.
In our previous report [16] we found that although MIF induced the self-renewal ability of NSPCs, it did not change the cell fate of NSPCs in vitro, as there was no significant difference in the cell differentiation potential of NSPCs upon MIF treatment. However, in the present study, Sox6 overexpression in NSPCs in vitro resulted in fewer differentiated cells belonging to three lineages. As Sox6 is just one of multiple genes activated by MIF, changes in the cell fate of NSPCs by Sox6 overexpression do not necessarily mirror changes induced by MIF treatment. Strong Sox6 activity may have been responsible for the decrease in differentiation markers in this system, which was not achieved in a previous study which examined the effects of MIF, a factor that lies upstream of Sox6.
The expression of SOX genes has been reported in many tumors [31]. We reported high expression of SOX genes in human gliomas in a previous report [15]. To date, high expression of SOX6 in gliomas has been confirmed by in silico gene expression databases, including Oncomine (www.oncomine.org). We have observed higher levels of SOX6 expression in glioma-initiating cells generated from glioma specimens compared to neural stem cells (Ohta et al., unpublished data). Thus, it may be important to analyze the detailed function of Sox6 in gliomas and gliomainitiating cells in the future.
In this study, we focused on Sox6 function in NSPCs derived from the GE of mouse embryonic brains in vitro, showing a new function of Sox6 as a maintenance factor of NSPCs stemness. Functional analyses of Sox6 in adult mouse NSPCs, and especially in human gliomas and glioma-initiating cells, will pave the way for evaluating Sox6 as a therapeutic target for many brain diseases. Figure S1 Self-renewal ability of Sox6 mutant NSPCs. In the primary neurosphere formation assay, single dissociated cells from E14.5 GEs of mouse fetal brains taken from of Sox6 mutant and littermate controls were seeded onto a 96-well plate at a cell density of 10 cells/ml in the presence of EGF and FGF2. WT (n = 9), KO (n = 6). In the secondary neurosphere assay, primary neurospheres were dissociated into single cells and seeded onto a 96-well plate at a cell density of 10