The self-renewal activity of neural stem cells (NSCs) has been suggested to decrease with aging, resulting in age-dependent declines in brain function, such as presbyopia and memory loss. The molecular mechanisms underlying decreases in NSC proliferation with age need to be elucidated in more detail to develop treatments that promote brain function. We have previously reported that the expression of esophageal cancer-related gene 4 (Ecrg4) was upregulated in aged NSCs, whereas its overexpression decreased NSC proliferation, suggesting a functional relationship between Ecrg4 and NSC aging. Using Ecrg4-deficient mice in which the Ecrg4 locus was replaced with the lacZ gene, we here show that Ecrg4 deficiency recovered the age-dependent decline in NSC proliferation and enhanced spatial learning and memory in the Morris water-maze paradigm. We demonstrate that the proliferation of Ecrg4-deficient NSCs was partly maintained by the increased expression of Foxg1. Collectively, these results determine Ecrg4 as a NSC aging factor.
Tissue-specific stem cells have been suggested to self-renew throughout the life of an animal and give rise to the many differentiated cells required to maintain tissue homeostasis and repair damaged tissues (Goodell et al., 2015; Krieger and Simons, 2015). However, stem cell proliferation declines with aging, resulting in defects in the residing tissue function, decreased tissue-repairing abilities and increases in the incidences of degenerative diseases (Fuchs and Chen, 2013; Krieger and Simons, 2015; Liu and Rando, 2011). Therefore, the molecular mechanisms responsible for stem cell aging need to be elucidated in more detail.
Esophageal cancer-related gene 4 (Ecrg4) was originally identified as one of the genes that are expressed at lower levels in esophageal cancer cells than in the surrounding non-cancer tissues (Su et al., 1998). Decreases in Ecrg4 expression were subsequently detected in various types of cancer tissues (Götze et al., 2009; Sabatier et al., 2011; Yue et al., 2003). While overexpression of Ecrg4 was found to prevent tumorigenesis by inhibiting cancer cell proliferation, it was revealed that Ecrg4 fragments secreted from cancer cells induced tumor eradication by activating both innate and adaptive immune system in vivo (Li et al., 2009, 2010; Moriguchi et al., 2016; Xu et al., 2013). Taken together, these findings indicated that Ecrg4 acts as a novel type of tumor suppresser.
We have independently identified Ecrg4 as a new cell senescence-associating factor; the expression of Ecrg4 was induced in senescent oligodendrocyte precursor cells (OPCs) that were cultured in the presence of high concentrations of fetal bovine serum, whereas its overexpression induced OPCs to become senescent with cell-cycle arrest, a large flat shape and senescence-associated β-galactosidase (β-gal) activity (Kujuro et al., 2010). We further demonstrated that some neural cells, including OPCs and neural stem cells (NSCs), in aged mouse brains were Ecrg4 positive, whereas this immunoreactivity was weaker in young brains (Kujuro et al., 2010). These findings suggested that Ecrg4 is involved in NSC aging in vivo.
In order to address this possibility, we generated Ecrg4 knock-in (Ecrg4KI) mice by replacing the Ecrg4 locus with the lacZ gene, and analyzed NSC functions. We herein demonstrate that Ecrg4-deficient NSCs maintained their proliferation activity during cultivation, even in the presence of serum that normally prevents NSC proliferation, and with age. Furthermore, the Morris water maze test revealed that spatial learning and memory were markedly better in Ecrg4KI mice than in wild-type mice. Taken together, these results determine Ecrg4 as a novel NSC-aging factor.
Ecrg4 is predominantly expressed in ependymal cells and NSCs
We replaced the first exon, which contains the first ATG codon on the Ecrg4 locus, with a floxed-lacZ gene and generated Ecrg4floxed-lacZ/+ (Ecrg4hetero) mice that developed normally and were fertile. We then crossed heterozygote mice and obtained Ecrg4floxed-lacZ/floxed-lacZ (Ecrg4KI) mice that were also viable and developed normally, indicating that Ecrg4 is not essential for development (Fig. S1).
We stained Ecrg4hetero mouse brains of 6 m and 15 m using X-gal and detected strong β-gal activity in the choroid plexus, which was consistent with previous findings (Gonzalez et al., 2011; Kujuro et al., 2010) and with data in the Allen Brain Atlas (www.brain-map.org) (Fig. 1A and Fig. S2A). We also observed β-gal+ cells in the NSC niches, ependymal layer, ventricular zone/subventricular zone (VZ/SVZ) and dentate gyrus (DG) (Fig. 1A and Fig. S2A, I and II). We confirmed the Ecrg4 expression in these area by immunolabeling brain sections for Ecrg4 (Fig. S3). These results apparently indicated that Ecrg4 is expressed in the NSC niches.
To identify which cells in the ependymal layer and VZ/SVZ express β-gal, we immunolabeled brain sections from Ecrg4KI mice for β-gal and neural markers, and found that the β-gal-expressing cells in the ependymal layer were positive for nestin, glial fibrillary acidic protein (GFAP) and S100β (Fig. 1B-D and Fig. S2B) (Doetsch et al., 1999; Steiner et al., 2007). A careful examination further revealed that Nestin+ activated NSCs and GFAP+ SVZ astrocytes, both of which are localized under the ependymal cell layer (Farkas and Huttner, 2008; Tramontin et al., 2003), were positive for β-gal (arrows in Fig. 1B,C). In contrast, migrating S100β+ cells and polysialylated neural cell adhesion molecule (PSA-NCAM)+ developing neuronal precursor cells were negative for β-gal (arrow and arrowheads in Fig. 1D and E, respectively). To further characterize β-gal+ cells in the ependymal layer and VZ/SVZ, we dissected these areas from Ecrg4hetero mouse brains, purified β-gal+ cells by flow cytometry (Fig. S4A), and then compared their gene expression profile with that of the cultured Ecrg4hetero NSCs. We found that β-gal+ cells from the ependymal layer and VZ/SVZ were enriched in ependymal cells expressing connective tissue growth factor, vascular endothelial growth factor A, aquaporins, vimentin and Foxj1, while the cells also expressed NSC markers, Foxg1, Notch and prominin 1, more than in the cultured NSCs (Fig. S4B). It should be noted that Ecrg4 expression was higher in cultured NSCs than that in β-gal+ cells from the ependymal layer and VZ/SVZ, suggesting that Ecrg4 expression was enhanced in the activated NSCs. Together, these results indicated that ependymal cells and activated NSCs express Ecrg4.
Ecrg4 deficiency prolongs NSC proliferation activity in the brain
There is much evidence that Ecrg4 prevents cell proliferation (Kujuro et al., 2010; Li et al., 2009, 2010; Xu et al., 2013). To examine the cell proliferation in the ependymal layer and VZ/SVZ cells, we immunolabeled sections of both wild-type and Ecrg4KI mouse brains at 6 months and 15 months for the proliferation marker Ki67 and nestin. The number of Ki67+/nestin+ cells in the ependymal layer and VZ/SVZ in wild-type mouse brains markedly decreased at 15 months (28 cells at 6 months versus one cell at 15 months), whereas a large number of nestin+ cells were positive for Ki67 in the same aged Ecrg4KI mouse brains (32 cells at 6 months versus 14 cells at 15 months) (Fig. 2A, 6 m and Fig. 2B, 15 m). We also found Ki67+/nestin+ NSCs in the DG of 6-month-old Ecrg4KI mice (9 cells), but not in the same aged wild-type mice (Fig. 2C) or in the 15-month-old Ecrg4KI and wild-type mice (Fig. 2D). We confirmed that significant number of the cells in Ecrg4KI mouse ependymal layer, VZ/SVZ and DG were positive for another proliferation marker phosphorylated histone H3 at serine 10 (pH3 S10), compared with those in the same areas of the wild-type mouse (10 cells versus 1 cell in the ependymal layer and VZ/SVZ at 15 months; eight cells versus two cells in the DG at 6 months) (Fig. S5). We further found the increased number of PSA-NCAM+ immature neurons in VZ/SVZ and DG of Ecrg4KI, compared with those in wild type (Fig. S6). Together, these findings indicated that Ecrg4 deficiency extends the proliferation activity of NSCs and keeps generating new neurons in the brain.
Ecrg4 deficiency maintains the proliferation activity of NSCs in a long-term culture
We investigated the molecular mechanism of how Ecrg4 deficiency keeps NSC proliferation. Ecrg4 was shown to induce the expression of pro-inflammatory cytokines, which prevent NSC proliferation, through the activation of a scavenger receptor/NF-κB signaling pathway in microglia (Lee et al., 2015; Kizil et al., 2015; Moriguchi et al., 2016, 2018). As cultured NSCs express both Ecrg4 and potential Ecrg4 receptors, including CD36 and Scarf1 (Fig. S4) (Moriguchi et al., 2018), secreted Ecrg4 may affect NSC proliferation in an autocrine/paracrine manner. To examine this possibility, we prepared three types of NSCs: Ecrg4+/+, Ecrg4+/− and Ecrg4−/− cells, from wild type, Ecrg4hetero and Ecrg4KI embryonic brains, respectively. During culture, we noted that the proliferation of Ecrg4+/− and Ecrg4−/− cells was markedly faster than that of Ecrg4+/+ cells. To quantify their proliferation activities, we performed a bromo-deoxyuridine (BrdU) incorporation assay. After 5 days in culture, 63% of Ecrg4+/− and 54% of Ecrg4−/− cells were positive for BrdU, whereas 29% of Ecrg4+/+ cells were positive (Fig. 3A left panel). After 60 days in culture, Ecrg4+/− and Ecrg4−/− cells maintained higher proliferation activities than that of Ecrg4+/+ cells (36% and 32% versus 8%, respectively) (Fig. 3A right panel), although the proliferation activities of all types of NSC appeared to decrease by 60 days. We also confirmed this tendency in the adult NSCs: adult Ecrg4+/− and Ecrg4−/− cells from 6-month-old VZ/SVZ and DG have proliferated faster than Ecrg4+/+ cells from these areas (33% and 35% versus 19% in VZ/SVZ, 33% and 35% versus 19% in DG, respectively) (Fig. S7A,B).
We then addressed whether the secreted Ecrg4 affects NSC proliferation in an autocrine/paracrine manner. We cultured Ecrg4−/− cells with or without Ecrg4+/+ cells for 7 days and then examined Ecrg4−/− cell proliferation using a BrdU incorporation assay. As shown in Fig. S7C, Ecrg4−/− cell proliferation significantly decreased when co-cultured with Ecrg4+/+ cells, indicating that the secreted Ecrg4 inhibited Ecrg4−/− cell proliferation through its receptor(s) on the cells.
We analyzed the expression of cell-cycle regulators in Ecrg4+/+, Ecrg4+/− and Ecrg4−/− cells, and their differentiation abilities. The expression of the cyclin-dependent kinase inhibitors (CDKIs) p16ink4a (Cdkn2a), p19arf and p21cip1 (Cdkn1a) was significantly low in 5-day-old cultured Ecrg4−/− cells, whereas the expression of p27kip1 decreased in 60-day-old cultured Ecrg4−/− cells, compared with that in Ecrg4+/+ cells (Jäkel et al., 2012; Nishino et al., 2008; Pagano et al., 1995; Rolfe et al., 1997) (Fig. 3B). No significant difference was observed in the expression of other CDKIs, p15ink4b (Cdkn2b), p18ink4c (Cdkn2c) and p57kip2 (Cdkn1c) between the three NSC lines (Fig. S8). Western blotting analysis confirmed the decreased expression of p16ink4a and p27kip1 in 5- and 60-day-old cultured Ecrg4−/− cells, respectively (Fig. 3C and Fig. S9A). We also analyzed the expression of cyclin D1-D3 (Ccnd1-Ccnd3) and found that they were similarly expressed in the three NSC lines and decreased during the culture (Fig. 3C and Fig. S9A)
All of the 60-day-old cultured NSCs were positive for the definitive NSC markers nestin and Sox2, in NSC medium with the mitogens bFGF and EGF (Fig. 3D, not shown about Ecrg4+/−). In differentiation medium, these NSCs became positive for the neuronal marker microtubule-associated protein 2 (MAP2) (a+b) (25 and 32% in Ecrg4+/+ and Ecrg4−/− cells, respectively) and astrocyte marker GFAP (32% and 24% in Ecrg4+/+ and Ecrg4−/− cells, respectively) (Fig. 3E). In DMEM+10% FCS, all cells were positive for GFAP. Unexpectedly, we found that Ecrg4−/− cells continued to proliferate in this culture condition, although Ecrg4+/+ cells stopped proliferating (Fig. 3F). Western blotting analysis revealed that p27kip1 remained low in Ecrg4−/− cells cultured in DMEM+10% FCS, whereas Ccnd3 levels slightly increased compared with those in Ecrg4+/+ cells (Fig. 3G and Fig. S9B). The expression of other cyclin D proteins was not detected in Ecrg4+/+ or Ecrg4−/− NSCs (data not shown). These data suggest that Ecrg4 deficiency drives NSC proliferation by regulating the expression of p27kip1, although it was not involved in either the expression of NSC markers or in differentiation.
Foxg1 is essential for Ecrg4−/− proliferation
To further analyze how Ecrg4 deficiency enhances NSC proliferation, we compared gene expression profiles between Ecrg4+/+, Ecrg4+/− and Ecrg4−/− cells and selected the increased 21 genes and decreased 19 genes (log2 ratio >2 times and <−2 times, respectively) (Fig. S10). Using quantitative RT-PCR, we confirmed decreases in the expression levels of iroquois homeobox 1 (Irx1) as well as Ecrg4, and increases in the expression levels of forkhead box g1 (Foxg1), six homeobox 1 (Six1) and DNA damage inducible transcript 4 like (Ddit4l) in Ecrg4−/− cells (Fig. 4A, Fig. S10B,C). Of these, we focused on Foxg1 for the following reasons: Allen Brain Atlas analysis has shown that Foxg1 is expressed in the VZ/SVZ and DG. Furthermore, the knockout of Foxg1 reduces NSC proliferation and induces premature neuronal differentiation (Genin et al., 2014; Tian et al., 2012; Xuan et al., 1995). Foxg1 has also been shown to prevent p21cip1-dependent cell-cycle arrest and to target the p27kip1 gene (Kumamoto et al., 2013; Siegenthaler et al., 2008; Yip et al., 2012). Immunolabeling confirmed the increased expression of Foxg1 in Ecrg4−/− cells in the culture (Fig. 4B). Moreover, immunohistochemical analysis revealed that the Foxg1+ cell layer in the VZ/SVZ of 15 m Ecrg4KI mice was thicker than that of wild-type mice (Fig. 4C). These indicated that Ecrg4 negatively regulates Foxg1 expression in NSCs in vitro and in vivo.
We next investigated whether Foxg1 regulates NSC proliferation using Foxg1 and its specific shRNA, foxg1sh, expression vectors. As shown in Fig. 4D, the overexpression of Foxg1 increased the proliferation of Ecrg4+/+ cells in culture. In contrast, the knockdown of Foxg1 prevented the proliferation of Ecrg4−/− cells, whereas foxg1sh did not influence Ecrg4+/+ cell proliferation (Fig. 4E). Taken together, these results demonstrated that Ecrg4 governs NSC proliferation by regulating Foxg1 expression.
Ecrg4 deficiency enhanced spatial learning and memory in mice
Age-dependent memory decline is thought to be caused by the proliferation defect and malfunction of NSCs in hippocampus (Ren et al., 2009). As Ecrg4−/− cells extended proliferation in the Ecrg4KI hippocampus (Fig. 2C), we investigated whether spatial learning and memory were better in Ecrg4KI mice than in wild-type mice using the Morris water maze system. As shown in Fig. 5A, wild-type and Ecrg4KI mice both swam at similar speeds (18.07 and 19.69 cm/s, respectively) in all experiments. On day 1, wild-type and Ecrg4KI mice arrived at the platform in the pool after similar times (49.6 s and 52.5 s, respectively; Movies 1 and 2). However, Ecrg4KI mice arrived faster than wild-type mice on day 2 (wild type versus KO, 45.5 s versus 29.3 s), and this improved further on day 3 (wild type versus KO, 41.7 versus 18.9 s; Movies 3 and 4), whereas the arrival time of wild-type mice became gradually shorter during the session although the difference was not drastic (Fig. 5B). Thus, these findings revealed that Ecrg4 deficiency enhances spatial learning and memory in mice by maintaining NSC proliferation.
One of the unanswered crucial questions for humans is how we age. Tissue-specific stem cells are undoubtedly involved in the aging process; tissue-specific stem cells continue to self-renew and supply functionally differentiated cells in their residing tissues; however, the cells gradually lose their abilities with age and eventually become senescent (Fuchs and Chen, 2013; Krieger and Simons, 2015; Liu and Rando, 2011). It is therefore crucial to find the factors that induce stem cell senescence and to characterize in order to elucidate the molecular mechanism of the aging process. Ecrg4 may fulfill the requirements of a stem cell aging factor; its expression increases in NSCs and ependymal cells with age (Kujuro et al., 2010; this study). Its overexpression induces cell senescence in NSCs and OPCs, whereas its deficiency maintains NSC proliferation in vitro and in vivo (Kujuro et al., 2010; this study). Proliferation of Ecrg4−/− cells significantly decreased when cultured with Ecrg4+/+ cells (this study). Ecrg4 has been shown to induce the expression of pro-inflammatory factors, which are involved in the initiation and maintenance of cell senescence in microglia/macrophages through the activation of scavenger receptor/NF-κB signaling pathway (Kizil et al., 2015; Moriguchi et al., 2016, 2018). Furthermore, we have demonstrated here that Ecrg4 deficiency increased spatial learning and memory in mice. Collectively, these findings suggest that Ecrg4 induces NSCs to be senescent in an autocrine/paracrine manner or through the activation of their surrounding cells, such as microglia/macrophage. Therefore, the inhibition of Ecrg4 may maintain brain function at high levels, even in the aged brain.
We detected Ecrg4 promoter activity in the ependymal layer and VZ/SVZ using Ecrg4KI mice, suggesting that NSC- and ependymal cell-specific transcription factors/regulators induce the expression of Ecrg4. This hypothesis is supported by the Champion ChiP Transcription Factor Search Portal (Qiagen) showing that the Ecrg4 promoter contains potential binding sites for Sox, Myb and Foxj1, which are essential for NSC maintenance and ependymal cell differentiation (Johe et al., 1996; Kyrousi et al., 2015; Stubbs et al., 2008; Zappone et al., 2000). Let7b and its targets, HMGA2 and Igf2bp1, are also candidates for Ecrg4 expression, because they are not only expressed in VZ/SVZ and DG but have also been shown to play roles in NSC aging (Nishino et al., 2008, 2013; Yang et al., 2015). Elucidation of the molecular mechanisms by which the expression of Ecrg4 is induced in NSCs and ependymal cells would provide an important insight involved in the initiation of NSC aging.
We found the differences in the expression of p16ink4a and p27kip1 between Ecrg4−/− and Ecrg4+/+ cells. p27kip1 was shown to be negatively regulated by Src-, JAK2- and Lyn-dependent phosphorylation and the proteasome pathway (Grimmler et al., 2007). As our microarray analysis revealed an increase in the expression of Lyn in Ecrg4−/− cells, Lyn may induce p27kip1 degradation in Ecrg4−/− cells. Foxg1 is another candidate factor that decreases p27kip1 expression, as it was shown that Foxg1 directly targeted p27kip1 gene (Kumamoto et al., 2013; this study). Much evidence has demonstrated the transcriptional and post-transcriptional regulation of p16ink4a by inhibitor of differentiation 1, polycomb group complexes, microRNA-24, etc. (LaPak and Burd, 2014). As the expression of these candidate factors did not change between Ecrg4−/− and Ecrg4+/+ cells in our microarray data, it remains unknown which factor induces the accumulation of p16ink4a in NSCs.
It was surprising that Ecrg4+/− cells showed an increase in proliferation in the same way that Ecrg4−/− cells did, although Foxg1 levels did not increase in Ecrg4+/− cells, suggesting that an unknown Ecrg4-downstream factor/mechanism, other than Foxg1, is also involved in cell proliferation. One possible mechanism is that the combination of the increased level of Ccnd2 with the decreased expression of p15ink4b and p57kip2, compared with Ecrg4−/− cells, might enhance Ecrg4+/− cell proliferation (Fig. 3C and Fig. S8). The expression of these cell-cycle regulators may be controlled by the crosstalk of Ecrg4 receptor signaling pathways, as Ecrg4 binds multiple scavenger receptors (Moriguchi et al., 2018). It is, therefore, essential in future to understand whole signaling pathways that Ecrg4 regulates.
Microarray analyses identified Foxg1 that is negatively regulated by Ecrg4 signaling. Because Ecrg4 has been shown to activate the NF-κB signaling pathway by binding with its receptors (Moriguchi et al., 2018), we searched the putative NF-κB binding sites in mouse Foxg1 promoter region using the transcription factor binding sites search engine PROMO (alggen.lsi.upc.es/cgi-bin/promo_v3/promo/promoinit.cgi?dirDB=TF_8.3) and found one NF-κB target site. Although there is much evidence that NF-κB acts as a transcription activator, it may repress Foxg1 expression with epigenetic modifiers, including enhancer of zeste homolog 2 (EZH2), binding sites for which also exist in the Foxg1 promoter.
Water maze analysis revealed that Ecrg4 deficiency increased the spatial learning and memory in mice. As Ecrg4 induces the expression of pro-inflammatory factors in microglia/macrophage, inhibition of Ecrg4-dependent inflammatory pathway may make it possible to maintain brain function better. Similarly, it is of great interest to study whether Ecrg4-dependent chronic inflammation causes a functional decline in various types of tissues.
MATERIALS AND METHODS
Animals and chemicals
Mice were obtained from Charles River Japan. Ecrg4KI mice (accession number CDB0712K; www2.clst.riken.jp/arg/mutant%20mice%20list.html) were generated using a knock-in method described previously (www2.clst.riken.jp/arg/methods.html) (Fig. S1A). Ecrg4KI mice were crossed with ACTB-FLPe mice (Rodríguez et al., 2000) to delete the neomycin selection cassette, generating Ecrg4floxed-lacZ/+ mice, and were backcrossed and maintained on a C57BL/6 genetic background. The genotype of each mouse was confirmed by a Southern blot analysis and PCR (Fig. S1B and S1C). All mouse experimental protocols were approved by the Animal Care and Use Committees of RIKEN Kobe Branch and Hokkaido University. Chemicals and growth factors were purchased from Sigma-Aldrich and PeproTech, respectively, except where otherwise indicated.
Preparation of the VZ/SVZ and hippocampus cells
The VZ/SVZ and hippocampus were dissected from 6-month-old Ecrg4hetero mice under a Stemi 2000 microscope (Carl Zeiss). The collected tissues were cut as small as possible and the cells were dispersed using the Papain dissociation system (Worthington).
Mouse primary NSCs were cultured as floating spheres in DMEM/F12 (Gibco) supplemented with the chemicals bFGF (10 ng/ml) and EGF (10 ng/ml) (NSC medium) in a standard culture dish (Falcon) for up to 60 days, as described previously (Kondo and Raff, 2000, 2004). For immunolabeling, cells were cultured as a monolayer on poly-D-lysine (PDL, 15 μg/ml) and fibronectin (1 μg/ml, Invitrogen)-coated dishes in NSC medium. In order to induce neural differentiation, NSCs were cultured on PDL- and fibronectin-coated eight-well chamber slides (Nunc) in mitogen-free NSC medium with 0.5% FCS (differentiation medium) or DMEM+10% FCS for 1 week. A BrdU incorporation assay was performed to quantify cell proliferation, as described previously (Kondo and Raff, 2004).
Immunostaining of brain sections (6-10 μm) was performed as described previously (Kujuro et al., 2010; Nishide et al., 2009). Sections were permeabilized with 0.3% Triton X-100 in PBS for penetration, treated with a blocking solution (2% skim milk, 0.3% Triton X-100 and PBS) for 1 h, and incubated with primary antibodies for 16 h at 4°C. Cells were fixed and immunostained as described previously (Hide et al., 2009; Kondo and Raff, 2000, 2004; Nishide et al., 2009). The following antibodies were used to detect antigens: mouse monoclonal anti-nestin (1:200, BD, Rat401), rabbit polyclonal anti-GFAP (1:500, DakoCytomation, Z0334), mouse monoclonal anti-GFAP (Sigma; 1:400, G3893), chicken anti-β-gal (1:500, Abcam, ab9361), rabbit polyclonal anti-Sox2 (1:500, StemCell Technology, 01438), mouse monoclonal anti-MAP2(a+b) (1:200, Abcam, ab5392), rabbit monoclonal anti-Ki67 (1:100, Thermo Fisher Scientific, Clone SP6), rabbit polyclonal anti-Foxg1 (1:100, Novus Biologicals, NBP1-56594), mouse monoclonal anti-PSA-NCAM (1:500, 12E3, a kind gift from Dr Tatsunori Seki, Tokyo Medical University, Japan), rabbit polyclonal anti-Ecrg4 (1:100, Sigma, HPA008546), rabbit anti-Histone H3 phospho serine10 (1:200, Abcam, ab47297) and anti-BrdU (1:100, DSHB, G3G4). Antibodies were detected using Alexa568-conjugated goat anti-rabbit (1:500, Thermo Fisher Scientific, A-11036), Alexa568-conjugated goat anti-mouse (1:500, Thermo Fisher Scientific, A-11004), Alexa488-conjugated goat anti-rabbit (1:500, Thermo Fisher Scientific, A-11008), Alexa488-conjugated goat anti-mouse (1:500, Thermo Fisher Scientific, R37120) and Cy3-conjugated donkey anti-chicken (1:500, Jackson ImmunoResearch, 703-165-155). Cells were counterstained with DAPI (1 µg/ml, DOJINDO, D523) to visualize nuclei. Fluorescence images were obtained using an AxioImager M1 microscope (Carl Zeiss).
Flow cytometry analysis
The VZ/SVZ cells were treated with the SPiDER-βGal reagent for β-gal activity, according to the supplier's instructions (Dojin), and then analyzed in an Aria II (Becton Dickinson) using a dual-wavelength analysis (488 nm solid-state laser and 638 nm semiconductor laser). Propidium iodide (PI)-positive (i.e. dead) cells were excluded from the analysis.
Brain fixation and histochemistry
Dissected mouse brains were fixed in 4% paraformaldehyde at 4°C overnight. After fixation, the brains were cryoprotected with 12-18% sucrose in PBS and embedded in Tissue-Tek OCT compound (Miles). In X-gal staining, frozen 10 μm sagittal sections were fixed using 0.2% glutaraldehyde in PBS for 5 min, washed several times and stained using an X-gal staining solution (2 mg/ml X-gal) for 24 h, as described previously (Nishide et al., 2009). For immunohistochemical analyses, the antigens were retrieved by HistoVT One, according to the supplier's instructions (Nacalai Tesque). X-gal staining images were obtained using a BZ-X800 microscope (Keyence).
Real-time PCR was performed on the StepOnePlus (Thermo Fisher Scientific) using THUNDERBIRD SYBR qPCR Mix (TOYOBO), according to the manufacturer's protocol. Cycle parameters were 15 s at 95°C and 30 s at 60°C for 40 cycles. All gene expression data were normalized to the housekeeping gene 18S ribosomal RNA, according to the ΔΔCt method. The sequences of the oligonucleotide primers are shown in Table S1.
Complementary DNAs (cDNAs) were cloned as described previously (Kondo and Raff, 2000, 2004). Mouse Foxg1 cDNA was inserted into the pMX-EGFP vector to produce pMX-EGFP-mfoxg1. The following oligonucleotide DNA primers were synthesized to amplify full mouse Foxg1 cDNA: 5′ primer, 5′-AGAATTCGCCACCATGCTGGACATGGGAGATAGG-3′; 3′ primer, 5′-ACTCGAGTTAATGTATTAAAGGGTTGGAAGAAGA-3′.
In order to knockdown mouse Foxg1, short-hairpin (sh) sequences were generated using InvivoGen's siRNA Wizard (www.sirnawizard.com/). These sh sequences were inserted into a psiRNA-h7SKhygro G1 expression vector (InvivoGen) to produce psiRNA-h7SKhygro-mfoxg1sh. The knockdown efficiency of the vector was analyzed by western blotting (Fig. S11). The sh target sequence for mouse foxg1 was 5′-GTCTTCTTCCAACCCTTTAAT-3′. The control sh target (egfp) sequence was 5′-GCAAGCTGACCCTGAAGTTCA-3′. The nucleotide sequences of cloned cDNA and inserted shRNA were verified using the BigDye Terminator Kit version 3.1 (Applied Biosystems) and ABI sequencer model 3130xl (Applied Biosystems). We transfected the cells with vectors using either the Nucleofector device according to the supplier's instructions (Lonza) or Polyethylenimine (PEI), as previously described (Kondo and Raff, 2004; Nishide et al., 2009).
Gene microarray analysis
Total RNA was prepared from the cultured NSCs and the dissociated VZ/SVZ cells, and then used for microarray analysis with the 3D-Gene Mouse Oligo chip 24k (TORAY) and the mouse GE 8×60K microarray (Agilent), as previously described (Kaneko et al., 2015; Ohtsu et al., 2016). Microarray data have been deposited in GEO under accession numbers GSE120626 and GSE120749.
Western blotting was performed as previously described (Takanaga et al., 2009). The blotted membranes were probed for the following antigens: p16Ink4a (Santa Cruz, 1:200), p21Cip1 (Santa Cruz, 1:100), p27Kip1 (Cell Signaling Technology, 1:500), Ccnd1 (Cell Signaling Technology, 1:500), Ccnd2 (Abcam, 1:100), Ccnd3 (Cell Signaling Technology, 1:2000) and GAPDH (Chemicon, 1:1000). An ECL system (Amersham) was used for detection. Protein quantification was performed using ImageJ software (NIH).
Morris water maze test
The Morris water maze test for the cognitive functional examination was performed in the New Drug Research Center (Sapporo, Japan) as previously described (Taniguchi et al., 2005). In brief, mice were tested in a maze that was 90 cm in diameter (LE82090, Nihon Bioresearch) containing water at 22°C with a round platform (diameter of 11 cm), which was submerged 0.5 cm beneath the water surface. The test was performed after 3 days of acquisition training with three trials per day. Mice (approximately 6 months old) were released at one of three randomly varied points and were given 60 s to find the platform. If the mice failed to find the platform within the time allotted, they were manually placed on the platform for 10 s and latency was recorded as 60 s. Mice were videoed and tracked using a LOGICOOL HD Webcam C615. Swimming speed was analyzed by Smart V3.0 software.
All studies were analyzed using a two-tailed Student's t-test, with a significant difference being defined as P<0.05.
We thank Shuji Takeda and Uichi Koshimizu for their helpful comments and critical reading of the manuscript; Tatsunori Seki for mouse anti-PSA-NCAM monoclonal antibody, Toshihiro Hata, Sachiyo Kawami, Kanako Ohyama and Naoki Ikeda for their technical assistance; and Satoshi Kondou for performing the DNA microarray analysis.
Conceptualization: T.K.; Methodology: Y.N., H.K., T.K.; Validation: Y.N., T.K.; Formal analysis: Y.N., H.K., T.K.; Investigation: Y.N., H.K., T.K.; Resources: Y.N., H.K., T.K.; Data curation: Y.N., H.K., T.K.; Writing - original draft: T.K.; Visualization: Y.N., T.K.; Supervision: T.K.; Project administration: T.K.; Funding acquisition: T.K.
This work was partly supported by a RIKEN internal grant (to T.K.), a research grant from Asubio Pharma Corporation (to T.K.) and was carried out under the Joint Research Program of the Institute for Genetic Medicine, Hokkaido University (to T.K.).
The authors declare no competing or financial interests.