SEARCH

SEARCH BY CITATION

Keywords:

  • Neural stem cell;
  • Fluorescent-activated cell sorting;
  • Microarray;
  • Gene expression;
  • Neural differentiation;
  • Cell migration

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. SUMMARY
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

During mammalian brain development, neural stem cells transform from neuroepithelial cells to radial glial cells and finally remain as astrocyte-like cells in the postnatal and adult brain. Neuroepithelial cells divide symmetrically and expand the neural stem cell pool; after the onset of neurogenesis, radial glial cells sequentially produce deep layer neurons and then superficial layer neurons by asymmetric, self-renewing divisions during cortical development. Thereafter, gliogenesis supersedes neurogenesis, while a subset of neural stem cells retain their stemness and lurk in the postnatal and adult brain. Thus, neural stem cells undergo alterations in morphology and the capacity to proliferate or give rise to various types of neural cells in a temporally regulated manner. To shed light on the temporal alterations of embryonic neural stem cells, we sorted the green fluorescent protein-positive cells from the dorsolateral telencephalon (neocortical region) of pHes1-d2EGFP transgenic mouse embryos at different developmental stages and performed gene expression profiling. Among dozens of transcription factors differentially expressed by cells in the ventricular zone during the course of development, several of them exhibited the activity to inhibit neuronal differentiation when overexpressed. Furthermore, knockdown of Tcf3 or Klf15 led to accelerated neuronal differentiation of neural stem cells in the developing cortex, and neurospheres originated from Klf15 knockdown cells mostly lacked neurogenic activities and only retained gliogenic activities. These results suggest that Tcf3 and Klf15 play critical roles in the maintenance of neural stem cells at early and late embryonic stages, respectively. STEM CELLS 2011;29:1817–1828


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. SUMMARY
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

The complex structure of the mammalian brain is an outgrowth of a homogeneous sheet of epithelial cells composing the neural tube. During the course of development, neural stem cells transform from neuroepithelial cells to radial glial cells and finally remain as astrocyte-like cells in the postnatal and adult brain [1–3]. Neuroepithelial cells initially divide symmetrically to multiply their copies and exponentially expand the neural stem cell pool. Around embryonic day 11 (E11), the division mode switches from symmetric to asymmetric, and neural stem cells commence to produce one differentiated neuronal daughter and another stem cell daughter (self-renewal) [4, 5]. After the onset of neurogenesis, neural stem cells adopt a radial glial morphology with radial processes extending apically to the ventricular surface and basally to the pial surface [2, 3]. During the period of cortical neurogenesis, radial glial cells sequentially produce deep layer neurons and then superficial layer neurons by asymmetric, self-renewing divisions. Gliogenesis then supersedes neurogenesis at the late embryonic or perinatal stage, at which time cortical neural stem cells generate mainly astrocytes and finally transform into astrocytes or ependymal cells, while a subset of neural stem cells retain their stemness and lurk in the postnatal and adult brain [6, 7].

Thus, neural stem cells undergo alterations in morphology and the capacity to proliferate or give rise to various types of neural cells in a temporally regulated manner [8]. If this controlled transition timing is disturbed, the size and shape of the brain and its cellular composition are severely affected. It has been well-established in Drosophila that neuroblasts sequentially alter their characteristics over time and express stage-specific transcription factors (Hunchback [RIGHTWARDS ARROW] Krüppel [RIGHTWARDS ARROW] Pdm [RIGHTWARDS ARROW] Castor), and their progeny are marked by the specific gene expression and differentiate into neurons with different functions [9]. However, such stage-specific gene expression in neural stem cells has not yet been elaborately investigated in the mammalian brain.

To shed light on the temporal alterations in characteristics of embryonic neural stem cells, we sorted the green fluorescent protein (GFP)-positive cells at different developmental stages from the dorsolateral telencephalon (neocortical region) of pHes1-d2EGFP transgenic mouse embryos [10] and performed gene expression profiling. Among dozens of transcription factors differentially expressed by cells in the ventricular zone (VZ) during the course of development, several of them exhibited the activity to inhibit neuronal differentiation when overexpressed. Furthermore, knockdown of Tcf3 (transcription factor 3) or Klf15 (Krüppel-like factor 15) resulted in accelerated neuronal differentiation, suggesting that these transcription factors play critical roles in the maintenance of neural stem cells at early and late embryonic stages, respectively.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. SUMMARY
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Cell Sorting

Dorsolateral parts of telencephalon (neocortical regions) were excised from pHes1-d2EGFP transgenic mouse embryos [10] at E11.5, 13.5, 15.5, and 17.5; digested with 0.25% trypsin-EDTA and 50 μg/ml DNaseI; and dissociated completely by pipetting. Trypsin inhibitor (0.25%) and Dulbecco's modified Eagle's medium (DMEM) were added, and then the cells were spun down, resuspended in serum-free culture medium (DMEM/F-12 [1:1] (Gibco, Carlsbad, CA, http://www.invitrogen.com/site/us/en/home/brands/Gibco.html) supplemented with B-27 (Gibco), N-2 (Gibco), 20 ng/ml epidermal growth factor (EGF) (Invitrogen, Carlsbad, CA, http://www.invitrogen.com/), and 20 ng/ml basic fibroblast growth factor (bFGF) (Invitrogen)), and filtered with a cell-strainer. GFP-positive cells were isolated using fluorescent-activated cell sorting (FACS) (FACSVantage and FACSAria cell sorter; BD Biosciences, San Jose, CA (for FACSVantage and FACSAria cell sorter) http://www.bdbiosciences.com/). Dead cells were excluded by gating on forward and side scatter properties and by eliminating cells stained with propidium iodide. Cells in the GFP-positive fraction were sorted and collected into culture medium; cell solutions were then spun down and cell pellets were dissolved in TRIzol reagent (Invitrogen). This procedure was completed within 2 hours to avoid cell differentiation following dissociation. More than 2 × 101 fluorescence intensity was defined as GFP-positive.

Microarray Analysis

Total RNA samples were extracted from the sorted cells using TRIzol reagent and the RNeasy Mini Kit (QIAGEN, Valencia, CA, http://www.qiagen.com/). Because of the limited number of sorted cells, Two-Cycle Target Labeling Assays were used to amplify cDNA, according to the manufacturer's protocol (Affymetrix, Santa Clara, CA, http://www.affymetrix.com/). Synthesized biotin-labeled cRNA was fragmented and hybridized to the GeneChip Mouse Genome 430 2.0 Array (Affymetrix). Data were analyzed using GCOS (Affymetrix) and were normalized to the signal intensity of the E11.5 sample using GeneSpring (Agilent Technologies, Santa Clara, CA, http://www.agilent.com/), as described elsewhere [11]. To explore the candidate genes differentially expressed at different embryonic stages, two criteria were set. First, the signal intensities should be flagged as “Present” and higher than 300. Second, the candidates should exhibit more than twofold changes in signal value between two given time points.

In Situ Hybridization

In situ hybridization was performed as described previously [12]. Digoxigenin-labeled antisense RNA probes were synthesized in vitro using the full- or partial-length cDNAs cloned into the pTA2 vectors (TOYOBO, Osaka, Japan, http://www.toyobo-global.com/) as templates, and hybridized to brain cryosections. Labeled preparations were imaged using a Zeiss Axiophot microscope (Carl Zeiss, Göttingen, Germany, http://www.zeiss.com/) equipped with an AxioCam color charge-coupled device (CCD) camera. Fluorescent in situ hybridization (FISH) was performed using fast red (Roche, Indianapolis, IN, http://www.roche.com/) as substrates instead of nitro blue tetrazolium (NBT)/5-bromo-4-chloro-3-indolyl-phosphate (BCIP).

In Utero Electroporation

Polymerase chain reaction fragments of cDNAs of candidate genes with or without hemagglutinin (HA)-tag sequences were cloned into the pEF expression vector driven by the EF (elongation factor) promoter or into the pNestin expression vector containing the promoter and second intron of the nestin gene. In utero electroporation was performed using methods described previously [13]. Embryos or neonates were harvested 3 or 7 days after electroporation. Brains were excised, fixed in 4% paraformaldehyde, cryoprotected, embedded in OCT, and cryosectioned at 16 μm.

Knockdown Experiments

Small interfering RNA (siRNA) sequences, 19–21 nucleotides long, within the coding regions of the various target genes were designed using siRNA Wizard software (www.sirnawizard.com). Annealed siRNA inserts for various target genes were cloned into psiRNA-h7SKneo G1 plasmid vectors (InvivoGen, San Diego, CA, http://www.invivogen.com/), in which a small hairpin RNA is expressed under the control of an RNA Polymerase III (Pol III) promoter. The plasmid vectors were introduced into the VZ cells by in utero electroporation, and brain tissues were analyzed as described above.

Characterization of Transfected Cells

The dorsolateral part of the telencephalon, including the electroporated region (electroporated at E13.5), was excised from embryos at E16.5, and cells were dissociated with 0.25% trypsin-EDTA and 50 μg/ml DNaseI and were plated onto poly-L-lysine (PLL)-coated chamber slides. After cultivation in serum-free culture medium (DMEM/F-12 [1:1] supplemented with B-27, N-2, 20 ng/ml EGF, and 20 ng/ml bFGF) for 2 hours to allow cells to attach to the bottom, cells were fixed in 4% paraformaldehyde and immunocytochemically analyzed for several markers of neuronal differentiation. Experiments were repeated three times and the mean and standard deviation of cell counts were calculated. Unpaired Student's t test was used for the statistical analyses.

Immunohistochemistry/Immunocytochemistry

Fixed cryosections were washed with phosphate-buffered saline (PBS), preincubated in PBS containing 5% normal goat serum and 0.1% Triton X-100, then incubated in 1% normal goat serum and 0.1% Triton X-100 with the following primary antibodies: rabbit anti-GFP (1:500; Molecular Probes, Eugene, OR, http://www.invitrogen.com/site/us/en/home/brands/Molecular-Probes.html), chicken anti-GFP (1:500; Abcam, Cambridge, MA, http://www.abcam.com/), Alexa488-conjugated rabbit anti-GFP (1:500; Molecular Probes), mouse anti-TUJ1 (1:1,000; Covance, Berkeley, CA, http://www.covance.com/), goat anti-Brn-1 (1:100; Santa Cruz, Santa Cruz, CA, http://www.scbt.com/), goat anti-Foxp2 (1:100; Santa Cruz), mouse anti-Nestin (1:500; BD Pharmingen, San Diego, CA (for anti-Nestin, anti-Ki67, anti-BrdU antibody) http://www.bdbiosciences.com/), mouse anti-Pax6 (1:200; Developmental Studies Hybridoma Bank, University of Iowa, Iowa City, IA, http://dshb.biology.uiowa.edu/), rabbit anti-Tbr2 (1:500; Abcam), mouse anti-MAP2 (1:500; Sigma, St. Louis, MO, http://www.sigmaaldrich.com/), mouse anti-Ki67 (1:100; BD Pharmingen), mouse anti-glial fibrillary acidic protein (GFAP) (1:400; Sigma), rabbit anti-GFAP (1:500; DAKO, Carpinteria, CA, http://www.dako.com/), and mouse anti-galactocerebroside (GalC) (1:200; Millipore, Billerica, MA, http://www.millipore.com/). Sections were incubated with primary antibodies overnight at 4°C and then with secondary antibodies for 1–3 hours at room temperature. Primary antibodies were detected with Alexa488-conjugated goat anti-rabbit IgG (1:200; Molecular Probes), Alexa488-conjugated goat anti-chicken IgG (1:200; Molecular Probes), Alexa594-conjugated goat anti-mouse IgG (1:200; Molecular Probes), Alexa594-conjugated donkey anti-goat IgG (1:100; Molecular Probes), and Alexa594-conjugated goat anti-rabbit IgG (1:200; Molecular Probes) and DNA was stained with DAPI (4′,6-diamidino-2-phenylindole; Sigma). Fluorescent sections were imaged using a Zeiss LSM510 confocal microscope.

Neurosphere Assay

The dorsolateral part of the telencephalon, including the electroporated region (electroporated at E13.5), was excised from embryos at E14.5, and cells were dissociated with trypsin-EDTA and DNaseI as described above, and resuspended in neurosphere culture medium (DMEM/F-12 supplemented with 100 μg/ml transferrin, 25 μg/ml insulin, 20 nM progesterone, 30 nM sodium selenite, 60 μM putrescine, 20 ng/ml EGF, and 20 ng/ml bFGF). GFP-positive transfected cells were isolated using FACS and subjected to neurosphere assay. Neurosphere assays were performed as described previously [10]. For primary sphere formation assay, 100 μl of cell suspension (5 × 104 cells per milliliter) containing 5,000 cells was plated into each well of a 96-well ultra-low attachment plate (Corning, Corning, NY, http://www.corning.com/). The numbers of primary spheres larger than 50 μm in diameter were counted at day 7. Experiments were repeated more than three times and the mean and standard error of sphere numbers were calculated. Unpaired Student's t test was used for the statistical analyses. For the neurosphere differentiation assay, primary spheres were collected and plated onto PLL- and laminin-coated Lab-Tek chamber slides (Nalge Nunc, Rochester, NY, http://www.nalgenunc.com/) in the differentiation medium (DMEM/F-12 supplemented with B-27 and 2% fetal bovine serum (FBS)). After 5 days of culture, cells were fixed and examined immunocytochemically.

BrdU Labeling

Mice of 6 weeks old were given with drinking water containing 1 mg/ml bromodeoxyuridine (or 5-bromo-2′-deoxyuridine) (BrdU) for 7 consecutive days and sacrificed either 1, 2, or 4 weeks later for BrdU label retention. Brains were fixed by transcardial perfusion with 4% PFA, and the fixed cryosections were incubated in 2 N HCl solution for 30 minutes at 37°C, followed by neutralization in 0.1 M sodium tetraborate buffer. Sections were incubated with mouse anti-BrdU antibody (1:100; BD Biosciences) overnight at 4°C and then with Alexa488-conjugated goat anti-mouse IgG (1:200; Molecular Probes).

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. SUMMARY
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

A Large Number of Genes Are Dynamically Expressed in Embryonic Neural Stem Cells

In the pHes1-d2EGFP transgenic mouse embryos, in which destabilized EGFP is driven by the Hes1 (hairy and enhancer of split 1) promoter (2.5 kb) (Fig. 1A), GFP expression marked the undifferentiated neural progenitor cells including neural stem cells within the VZ of the dorsal telencephalon, as we previously demonstrated [10] (Fig. 1B). Here, we further confirmed the cell types that express d2EGFP under the Hes1 promoter by both immunohistochemistry and immunocytochemistry. At all stages examined (E11.5, E13.5, E15.5, and E17.5), most GFP-positive cells in the dorsolateral telencephalon were positive for Pax6, a marker for neural stem cells (radial glial cells) during these developing stages (Supporting Information Figs. 1–4). Especially, almost all (>95%) GFP-positive cells were Pax6-positive at E11.5, E13.5, and E15.5. As development proceeds from E11.5 [RIGHTWARDS ARROW] E13.5 [RIGHTWARDS ARROW] E15.5 [RIGHTWARDS ARROW] E17.5, the telencephalic wall grows in thickness and the proportion of the VZ demarcated by GFP expression tapers off (Fig. 1C–1F; Supporting Information Figs. 1–4). Consistent with this observation, the GFP-positive fraction of cells gradually diminished over the same period, as shown in the FACS plots (Fig. 1G–1J).

thumbnail image

Figure 1. Sorting of green fluorescent protein (GFP)-positive cells from pHes1-d2EGFP transgenic mice. (A): Structure of pHes1-d2EGFP transgene [10]. (B): GFP expression (green) in pHes1-d2EGFP mice at embryonic day 11.5 (E11.5), E13.5, E15.5, and E17.5. Arrowheads indicate GFP expression in the dorsal telencephalon. (C–F): Anti-GFP (green) and anti-TUJ1 (red) immunostaining of pHes1-d2EGFP brain sections. Dorsolateral telencephalon (depicted by yellow rectangles) was excised and GFP-positive cells were sorted by fluorescent-activated cell sorting (FACS). (G–J): FACS plots of dissociated cells from dorsolateral telencephalon. Abbreviations: GFP, green fluorescent protein.

Download figure to PowerPoint

thumbnail image

Figure 2. Dynamic gene expression of transcription factors during cortical development. Representative transcription factors exhibiting temporal alterations in expression as development proceeds. Each gene showed significant upregulation (pink) or downregulation (blue) (more than double or less than half, respectively) between the two time points designated above. Klf15 was flagged “Absent” (A) at embryonic day 11.5 (E11.5) and E13.5.

Download figure to PowerPoint

thumbnail image

Figure 3. Expression patterns of candidate genes in the ventricular zone (VZ) of the telencephalon. In situ hybridization in coronal sections of mouse brains at different developmental stages (embryonic day 11.5 [E11.5]–E17.5), showing the expression patterns of mRNA for the indicated transcription factors. Although it is difficult to quantitatively compare the expression levels at different stages, many genes exhibited apparent changes in expression levels in the VZ of the dorsolateral telencephalon; rightmost panels are higher magnification images showing the dorsolateral part of the telencephalon. Although expression of some genes is restricted to the VZ, some are also expressed in the outer layers including the cortical plate besides the VZ.

Download figure to PowerPoint

thumbnail image

Figure 4. Summary of transcription factors dynamically expressed in neural stem cells. (A): Only genes that were confirmed to exhibit temporary alterations in expression by in situ hybridization are shown. Hmga2 showed consecutive downregulation between all three intervals. (B): Schematic diagram of dynamic gene expressions in each category on the basis of in situ hybridization. (C): Characterization of embryonic neural stem cells of each developmental stage, corresponding to embryonic day 11.5 (E11.5), E13.5, E15.5, and E17.5, by a combination of expression patterns for multiple transcription factor genes.

Download figure to PowerPoint

We sorted the GFP-positive cells at different developmental stages using FACS and then performed gene expression profiling to comprehensively analyze the temporal alterations in characteristics of embryonic neural stem cells (Supporting Information Fig. 5). Among 45,101 genes on the GeneChip array, 25,940 were flagged as Present in the GFP-positive cells at E11.5 (Supporting Information Table 1). To limit these results somewhat and to identify those genes differentially expressed at different embryonic stages, we focused first on transcription factors. A total of 92 transcription factor genes were differentially expressed among the various developmental stages; 53 genes were upregulated as development proceeds (E11.5<E13.5: 25 genes, E13.5<E15.5: 16 genes, and E15.5<E17.5: 12 genes) and 39 genes were downregulated (E11.5>E13.5: 12 genes, E13.5>E15.5: 9 genes, and E15.5>E17.5: 18 genes) between two given time points (Supporting Information Table 2). Temporal expression patterns of representative genes are shown in Figure 2. Hmga2 (high mobility group AT-hook 2) showed consecutive downregulation between all three intervals (E11.5–E13.5, E13.5–E15.5, and E15.5–E17.5), consistent with reports that its expression declines over the course of development [14].

thumbnail image

Figure 5. Inhibition of neuronal differentiation following overexpression of transcription factors. Coronal sections of dorsolateral telencephalon were immunostained to analyze the fates of cells transfected with expression vectors for the indicated genes 3 (A–M) or 7 (A–P′) days following in utero electroporation at embryonic day 13.5 (E13.5). Typical images that were reproducibly observed in multiple experiments using at least two independent animals are shown. Anti-green fluorescent protein (green; A–M, A–P′), anti-TUJ1 (red; A–M), anti-Brn-1 (red; A–D′, I–L′), and anti-Foxp2 (red; E–H′, M–P′). Abbreviations: CP, cortical plate; IZ, intermediate zone; SVZ, subventricular zone; VZ, ventricular zone; GFP, green fluorescent protein; II–IV, prospective cortical layers II–IV; V–VI, prospective cortical layers V and VI.

Download figure to PowerPoint

It has become evident that basic helix-loop-helix (bHLH) genes play critical roles in neuronal fate determination and differentiation during development of the central nervous system (CNS). We previously reported that repressor-type bHLH genes Hes1 and Hes5 are expressed in the VZ, function to inhibit neuronal differentiation as effectors of Notch signaling [15–19], and have the activity to maintain neural stem cells [13]. Therefore, we examined whether those bHLH genes display dynamic expression patterns during development (Supporting Information Fig. 6). Among the Hes genes (Hes1, Hes5, and Hes6) and Hes-related genes (Hey1 and Hey2), only Hey2 exhibited a remarkable increase between E11.5 and E13.5, although the raw signal was quite low. Other Hes or Hes-related genes showed rather stationary expression patterns throughout the periods examined. In contrast, some activator-type bHLH genes exhibited dynamic alterations in expression during this period. In particular, Neurogenin2 (Ngn2) and Neurod6 (Math2) were conspicuously upregulated between E11.5 and E13.5, in agreement with the onset of extensive neurogenesis [20, 21], maintained high expression during the neurogenic period through E15.5 and then declined between E15.5 and E17.5 in accordance with the transition from neurogenesis to gliogenesis.

thumbnail image

Figure 6. Premature neuronal differentiation of neural stem cells following knockdown of Tcf3 or Klf15. (A–J): Immunohistochemistry of anterior (A–E) and posterior (F–J) telencephalon following overexpression of Tcf3 (B, G) or Klf15 (D, I), and siRNA knockdown of Tcf3 (siTcf3) (C, H) or Klf15 (siKlf15) (E, J). Anti-green fluorescent protein (GFP) (green) and 4′,6-diamidino-2-phenylindole (DAPI) nuclear staining (blue). (K–P): Representative images of immunocytochemical observations of dissociated cells; following transfection with elongation factor promoter (pEF)-EGFP (K–M) and pEF-HA-Tcf3 (N–P). Anti-GFP (green), anti-Pax6 (red), and DAPI nuclear staining (blue). Arrows indicate double-labeled cells (yellow) with anti-GFP and anti-Pax6 antibodies (M, P). (Q, R): Immunocytochemical characterizations of cells transfected with overexpression vectors or small interfering RNA (siRNA) vectors for Tcf3 or Klf15 (electroporated at embryonic day 13.5 [E13.5]) after dissociation at E16.5. The graph shows averages of three independent experiments; error bars represent the standard deviation (Q) and the standard error (SE) (R), respectively. *, p < .05, **, p < .01. (S): Neurosphere forming assay. The graph shows average numbers of primary neurospheres originated from 5,000 cells; error bars represent the SE. *, p < .05, **, p < .01. (A′–K′): Neurosphere differentiation assay. Primary neurospheres were cultured in the differentiation conditions for 5 days and immunostained with anti-TUJ1, anti-GFAP, and anti-GalC antibodies (A′–J′), and the cellular compositions of the colonies derived from each sphere were evaluated (K′). A/N/O: colonies with astrocytes (A), neurons (N), and oligodendrocytes (O). A/O: colonies with only glial cells. Essentially all colonies contained a small number of oligodendrocytes. Abbreviations: DAPI, 4′,6-diamidino-2-phenylindole; GFP, green fluorescent protein; GFAP, glial fibrillary acidic protein; GalC, galactocerebroside.

Download figure to PowerPoint

To check whether the 92 transcription factor genes are indeed expressed by the VZ cells, we then performed in situ hybridization using probes for most of them (Supporting Information Table 3). Among these, 30 transcription factor genes were expressed in the VZ of the dorsolateral telencephalon during the periods examined (E11.5–E17.5) (Fig. 3). A total of 20 genes were expressed exclusively in the VZ and 10 were expressed in both the VZ and the cortical plate (CP). Some of them exhibited temporal alterations in expression levels during the course of development, showing similar patterns to those observed in the microarray data (Figs. 3 and 4). These results indicate that embryonic neural stem cells can be categorized into at least four stages by a combination of expression patterns for multiple transcription factors: Jarid2 (jumonji, AT rich interactive domain 2), Tcf3, Trp53 (transformation related protein 53), Tcf4 (transcription factor 4), Klf15, and Pippin (Csdc2: cold shock domain containing C2); (1) at ∼E11.5: Jarid2high/Tcf3high/Trp53high/Tcf4low/Klf15low/Pippinlow; (2) at ∼E13.5: Jarid2low/Tcf3high/Trp53high/Tcf4high/Klf15low/Pippinlow; (3) at ∼E15.5: Jarid2low/Tcf3low/Trp53high/Tcf4high/Klf15high/Pippinlow; and (4) at ∼E17.5: Jarid2low/Tcf3low/Trp53low/Tcf4high/Klf15high/Pippinhigh (Fig. 4C).

Several Transcription Factors Exhibited Inhibitory Activity on Neuronal Differentiation

We then analyzed the functional roles of transcription factors that were differentially expressed within the VZ during the course of development (Supporting Information Table 4). We first performed gain-of-function analyses by overexpressing the genes via in utero electroporation. To investigate the ratio of neural stem cells versus other progenitor cells (e.g., intermediate progenitor cells [IPCs]) or postmitotic neurons among cells transfected by in utero electroporation, we transfected the control vectors (pEF-EGFP) into the dorsolateral telencephalon of E13.5 mouse embryos, sacrificed them after a short period (6 hours) and examined immunohistochemically and immunocytochemically with several markers (Supporting Information Fig. 7). In the immunohistochemistry, most transfected cells were positive for Pax6, a marker for neural stem cells (radial glial cells) but negative for Tbr2, a marker for IPCs or a neuronal marker TUJ1, 6 hours after in utero electroporation. We further performed the immunocytochemistry to precisely characterize the transfected cells and found that more than 98% of transfected cells were Pax6-positive and nearly 90% were Ki67-positive, and Tbr2- or TUJ1-positive cells were less than 1% (Tbr2+: 0.25% and TUJ1+: 0.64%), indicating that almost all cells transfected by in utero electroporation are Pax6-positive, dividing neural stem cells in the VZ. Furthermore, we performed in situ hybridization for some selected genes and confirmed that the electroporation of expression vectors actually lead to overexpression of each gene (Supporting Information Fig. 8).

thumbnail image

Figure 7. Summary diagram of gene expression profiling during cortical development. The characteristics of neural stem cells are temporally altered. They undergo stepwise differentiation during development and give rise to various types of neural cells in a temporally regulated manner. In this study, green fluorescent protein (GFP)-positive cells of pHes1-d2EGFP mouse embryos were sorted from the dorsolateral telencephalon (neocortical region) at different developmental stages and were subjected to gene expression profiling to shed light on the temporal alterations of embryonic neural stem cells. Abbreviations: NSC, neural stem cell; NPC, neural progenitor cell; GRP, glial restricted precursor cell; OPC, oligodendrocyte precursor cell; APC, astrocyte precursor cell.

Download figure to PowerPoint

We used two kinds of expression vectors, pEF (elongation factor promoter; expressed more ubiquitously) and pNestin (nestin promoter and second intron; expressed specifically in neural progenitor cells) and introduced them into the VZ cells and examined the fates of the transfected cells 3 or 7 days later. When transfected with the control vectors (pEF-EGFP) at E13.5, the majority of transfected cells migrated out of the VZ and many of them migrated radially and reached the CP at E16.5 (Fig. 5A). Several of the overexpressed transcription factors exhibited the activity to inhibit neuronal differentiation or migration. For example, most of the cells transfected at E13.5 with expression vectors for Nfia (nuclear factor I/A), Tcf4, Fhl1 (four and a half LIM domains 1), Klf15, Etv5 (ets variant gene 5), Zbtb20 (zinc finger and BTB domain containing 20), Hmga2, Jarid2, Tcf3, Trp53, Emx1 (empty spiracles homolog 1), or Hbp1 (high mobility group box transcription factor 1) remained within the VZ and subventricular zone (SVZ) 3 days later at E16.5; moreover, the number of cells that migrated out of the VZ and reached the CP was considerably less than following transfection with the control vectors (Fig. 5A–5M). Given that most transfected cells within the VZ were negative for the neuronal marker TUJ1, it is unlikely that the accumulations of these cells were due to impaired migration of differentiated neurons or neuronal progenitors, but rather that neuronal differentiation itself was inhibited by the overexpression of these factors. In such cases, transfection with either the pEF or pNestin vectors produced similar results (data not shown). Interestingly, many cells transfected with pEF-Etv5 exhibited an abnormal shape (broad cell body with short neurites) within the VZ, suggesting that overexpression of Etv5 exerted a significant influence on cell morphology (Fig. 5G).

Seven days after in utero electroporation at postnatal day 1 (P1), most GFP-positive cells transfected with expression vectors for Klf15, Etv5, Zbtb20, Tcf3, or Trp53 were aberrantly located in superficial layers of the CP (prospective cortical layers II and III), in contrast to cells transfected with the control vectors that were localized in deeper layers (prospective cortical layers III and IV) (Fig. 5A′–5P′). Cortical layer structure was indicated by immunoreactivities for Brn-1, a marker of layer II–IV [22], and Foxp2, a marker of layer VI [23]. These results are also supportive of the previous finding that cell migration was not disturbed but neuronal differentiation was inhibited, delaying the onset of neuronal differentiation and migration for a few days. Overexpression of Hmga2 did not cause such a noticeable aberration (Fig. 5J′, 5N′). On the other hand, cells transfected with pEF-Nfia were distributed in a disorderly fashion throughout the cortex, and some of them displayed an abnormal, elongated shape with a thick process, indicating that overexpression of Nfia had an impact on cell morphology and motility (Fig. 5B′, 5F′). Furthermore, this defect led to poorly organized superficial layers with reduced numbers of Brn-1-positive neurons, as shown in Figure 5B′.

Tcf3 Is Essential for Maintenance of Neural Stem Cells at Early Embryonic Stages

To further analyze the molecular functions of those factors encoded by the differentially expressed genes, we carried out knockdown experiments, by examining the fates of cells transfected with various siRNAs following in utero electroporation. Intriguingly, cells transfected with siRNA against Tcf3 (siTcf3) exhibited conspicuously accelerated neuronal differentiation and migration; cells transfected at E13.5 swiftly migrated out of the VZ and reached the CP without lingering in the SVZ. Thus, the majority of transfected cells were located in the CP at E16.5 (Fig. 6C, 6H; Supporting Information Fig. 9).

To further characterize the overexpressing and knocked down cells in more detail, we immunocytochemically analyzed several markers of neuronal differentiation in dissociated telencephalic cells 3 days following in utero electroporation (Fig. 6K–6Q). Whereas only 9.7% of the VZ cells transfected with the control vector (pEF-EGFP) remained positive for Pax6, a marker for radial glial cells (neural stem cells) (Fig. 6K–6M) [24], most transfected cells differentiated into neurons (TUJ1+: 65.4%, MAP2+: 62.9%) and 15.2% of them were positive for Tbr2, a marker for IPCs in the SVZ [25]. In this experiment, we observed most of the Tbr2-positive cells were also positive for Nestin and a few of them coexpressed TUJ-1. Approximately 25% of the control-transfected cells were Nestin-positive (Nestin+: 25.4%), which were comprised of neural stem cells and IPCs (i.e., they were also positive for Pax6 and Tbr2, respectively).

In agreement with the immunohistochemical observations in brain sections, when Tcf3 was overexpressed with pEF-Tcf3 vectors, the numbers of cells positive for Nestin and Ki67, a marker for proliferating cells, significantly increased (Nestin+: 47.3% and Ki67+: 20.2%) at the expense of differentiated neurons (TUJ1+: 41.7% and MAP2+: 40.6%) (Fig. 6Q, see also Fig. 6B, 6G; Supporting Information Fig. 9). The increase in Pax6-positive cells was most prominent, more than double compared with the control (Pax6+: 21.5%) (Fig. 6N–6P). Tbr2-positive cells also increased (Tbr2+: 23.4%), probably resulting from the expansion of the neural stem cell population. By contrast, among the cells transfected with siRNA against Tcf3, the Tbr2-positive population dramatically increased (Tbr2+: 29.9%) and the sum of Tbr2- and TUJ1-positive cells also increased at the expense of Pax6-positive cells (Pax6+: 4.8%) (Fig. 6R), even though the number of differentiated neurons was unexpectedly less than in the control, and despite the apparent increase in transfected cells localized in the CP observed in the immunohistochemistry (Fig. 6C, 6H; Supporting Information Fig. 9).

Together, these findings show that overexpression of Tcf3 leads to retention and expansion of neural stem cells by inhibiting neuronal differentiation, and reduction in Tcf3 induces premature neuronal differentiation of neural stem cells into IPCs and neurons, indicating that Tcf3 is critical for maintaining the population of embryonic neural stem cells.

Klf15 Is Involved in Maintenance of Neural Stem Cells at Late Embryonic Stages

Results of the knockdown experiments with siRNA against Klf15 (siKlf15) were even more noticeable. Surprisingly, most of these cells transfected at E13.5 had accumulated 3 days later at E16.5 in the SVZ, where Tbr2-positive IPCs reside, and in the intermediate zone (IZ) (Fig. 6E, 6J; Supporting Information Fig. 9). The number of transfected cells within the VZ was noticeably less than in the control, while the number of transfected cells that migrated radially and reached the CP dramatically decreased, indicating that these cells prematurely exited the neural stem cell population and migrated out of the VZ but were inhibited from radial migration en route to the CP, thereby accumulating in the SVZ and IZ.

Immunocytochemistry of dissociated cells after in utero electroporation demonstrated that overexpression of Klf15 led to increases in the percentages of Nestin-, Pax6-, and Tbr2-positive cells (Nestin+: 32.8%, Pax6+: 17.2%, and Tbr2+: 22.2%), with a decrease in the number of differentiated neurons (TUJ1+: 47.0% and MAP2+: 47.6%) (Fig. 6Q, see also Fig. 6D, 6I; Supporting Information Fig. 9). Following the knockdown of Klf15 with siRNA, the Tbr2-positive IPCs remarkably increased (Tbr2+: 31.2%) and the sum of Tbr2- and TUJ1-positive cells was higher than the control at the expense of Pax6-positive cells (Pax6+: 5.5%) (Fig. 6R, see also Fig. 6E, 6J; Supporting Information Fig. 9). As the percentages of differentiated neurons only slightly decreased compared with the control (TUJ1+: 53.5% and MAP2+: 44.9%), it is likely that the transfected cells that accumulated in the SVZ and the IZ consist mostly of Tbr2-positive IPCs and differentiated neurons. These results indicate that Klf15 is essential for the maintenance of neural stem cells in the VZ at later embryonic stages. It remains to be addressed whether the migration defects seen following the knockdown of Klf15 are due to its primary or secondary effect on neuronal differentiation or specification.

To further confirm the effects of overexpression and knockdown of Tcf3 and Klf15 on the maintenance of neural stem cells, we then performed neurosphere forming assays (Fig. 6S). One day after in utero electroporation at E13.5, dorsolateral part of the telencephalon, including the electroporated region, was excised from embryos at E14.5. Cells were dissociated and GFP-positive transfected cells were isolated using FACS and subjected to neurosphere assay. Cells transfected with pEF-Tcf3 or pEF-Klf15 generated apparently more primary neurospheres (pEF-Tcf3: 129.67 ± 17.50 and pEF-Klf15: 134.00 ± 34.52 from 5,000 cells) compared with the control (pEF-EGFP: 87.00 ± 6.22). On the other hand, the numbers of neurospheres generated from the cells transfected with siRNA against Tcf3 or Klf15 were less than the control (siTcf3: 54.67 ± 16.29 and siKlf15: 55.50 ± 22.49). Next, we cultured those primary neurospheres in the differentiation conditions for 5 days and examined the cell lineages immunocytochemically (Fig. 6A′–6J′). Surprisingly, whereas nearly all spheres generated astrocytes (A), neurons (N), and oligodendrocytes (O) in other cases (pEF-EGFP: A/N/O = 95.57 ± 3.91%, pEF-Tcf3: A/N/O = 100%, siTcf3: A/N/O = 100%, and pEF-Klf15: A/N/O = 96.33 ± 3.39%), most spheres originated from the cells transfected with siRNA against Klf15 generated only glial cells with a lack of differentiated neurons (siKlf15: A/O = 92.79 ± 1.24% and A/N/O = 7.21 ± 1.24%), indicating that knockdown of Klf15 lead to the loss of multipotential sphere-forming cells (Fig. 6K′). These results reinforce the above notion that Tcf3 and Klf15 have critical functions to maintain neural stem cells by inhibiting neuronal differentiation, and reduction in either Tcf3 or Klf15 induces premature neuronal differentiation of neural stem cells.

To gain insight into the molecular functions of these genes, we examined the expression patterns of Tcf3 and Klf15 in the postnatal and adult brain (Supporting Information Fig. 10). As shown in Figure 3, Tcf3 is downregulated as development proceeds and its expression became faint in the VZ of E17.5 brain. Agreeing with these observation, Tcf3 mRNA could be detected in the anterior periventricular region very faintly at P0 (Supporting Information Fig. 10A), but could not be detected in P7, P14, or adult (6 weeks) brains (data not shown and Supporting Information Fig. 10J). In contrast, Klf15 mRNA was clearly detected in the anterior (Supporting Information Fig. 10C) and posterior (Supporting Information Fig. 10D) periventricular region at P0, and the expression was sustained in the anterior SVZ, where neurogenesis continuously occurs, in the adult (6 weeks) brain (Supporting Information Fig. 10E–10G). Klf15 was expressed in the Purkinje cell layer of the cerebellum (Supporting Information Fig. 10I), but could not be detected in the dentate gyrus of the hippocampus (Supporting Information Fig. 10H). Furthermore, we administered BrdU in drinking water to 6-week-old mice and sacrificed 1, 2, or 4 weeks later, and performed FISH using probes for Klf15, combined with immunohistochemistry using anti-BrdU or anti-GFAP antibodies (Supporting Information Fig. 11). BrdU-retaining cells gradually decreased in number as weeks passed and only a small fraction of cells retained BrdU at 9 or 11 weeks. Expression of Klf15 was detected in the periventricular region, especially in the lateral SVZ. Although not all BrdU-retaining cells expressed Klf15, it was detected by some of BrdU-retaining cells. Furthermore, Klf15 expression domain overlapped with GFAP expression in the SVZ, suggesting that Klf15 is mainly expressed adjacent to the stem cell niche in the adult brain, including quiescent (BrdU-retaining) and active neural stem cells in the SVZ. These results suggest that Klf15 plays a role in the maintenance of neural stem cells in the postnatal and adult brain as well.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. SUMMARY
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

During cortical development, neural stem cells initially proliferate symmetrically and thereafter sequentially produce deep layer neurons, superficial layer neurons, and glial cells with a strictly regulated timing. In this way, the characteristics of neural stem cells are temporally altered. They undergo stepwise differentiation and produce a variety of cell types during the course of development, thereby providing the basis for the complex structure and higher functions of the mammalian neocortex. We performed gene expression profiling to reveal temporal alterations in gene expression in embryonic neural stem cells (Fig. 7). Supporting Information Table 4 summarizes the process of gene expression profiling and functional analyses of genes differentially expressed during embryonic development performed in this study. A clear understanding of the profile of neural stem cells and the mechanisms regulating their stepwise differentiation will contribute to the advancement of neuroregenerative therapy. Although enormous efforts have been made to establish cell replacement therapies for damaged brains or neurodegenerative disorders using neural stem/precursor cells, it remains a difficult challenge to achieve structurally and functionally intact integrations of manipulated neurons into the CNS [26–31]. Furthermore, embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs) have recently been regarded as attractive new donor sources for cell replacement, and several lines of evidence indicate that specific types of neural cells that can be used for cell replacement therapies can be selectively induced from ESCs/iPSCs [32, 33]. Thus far, directed differentiation of ESCs into dorsal or ventral telencephalic precursor cells has been achieved using optimized, serum-free suspension cultures combined with some factors [34]. Furthermore, iPSCs were also able to give rise to neurons and glial cells with therapeutic potential to improve the symptoms of Parkinson's disease [35] or spinal cord injury [36] after transplantation. However, it is still difficult to regulate the temporal or regional properties of neural stem cells and to expand a desired population that is suitable for the environment of transplantation sites. Dedifferentiating neural stem cells or initializing them into naïve neural stem cells with broader potential to generate various cell types, or expanding stage-specific neural stem cells via controlled gene expression, and thereby voluntarily generating desired types of neurons or glial cells, would be powerful tools for cell replacement therapies for damaged brains (Fig. 7). Our results demonstrated that numerous genes exhibit dynamic expression patterns in neural stem cells during cortical development, and some genes have critical functions to maintain neural stem cells.

Dynamic Gene Expression in Neural Stem Cells During Cortical Development

Gene expression profiling revealed that neural stem cells are dynamically altering their characteristics during cortical development, as demonstrated by the dynamic gene expression. More than half of the genes present on the GeneChip array (25,940 of 45,101) were flagged as Present and thousands of them exhibited more than twofold changes between two given time points between E11.5, E13.5, E15.5, or E17.5. Genes with peak expressions at each stage are expected to have functions in the “maintenance of neural stem cells,” “specification of deep layer neurons,” “specification of superficial layer neurons,” and “glial differentiation,” respectively. Ratio of Pax6-positive cells in d2EGFP-positive cells was slightly less than 90% at E17.5, although d2EGFP-positive cells mostly expressed Pax6 at other stages. As only a few cells were neuronal progenitor cells (Tbr2+: 1.17%) or neurons (TUJ1+: 0.37%) at E17.5, remaining population might be composed of glial cells or glial progenitors. Thus, we should be careful to analyze the GeneChip data when comparing that of E17.5 and others. Although we focused on transcription factors here, there are still many valuable genes in other categories that remain to be explored. We also tried to find stage-specific markers that sharply demarcate each step, and identified transcription factors that exhibited peak expressions at either E13.5 or E15.5. We performed in situ hybridization to examine the expression patterns of individual genes; unfortunately we did not detect any prominent expression in the VZ at the specific time point. Therefore, it appears difficult to clearly distinguish the neural stem cells of each developmental stage by single transcription factors, unlike the stepwise expression of transcription factors in Drosophila neuroblasts. However, we could possibly characterize the neural stem cells of each stage by a combination of multiple transcription factors that demonstrated temporal alterations in expression, as revealed in this study.

Transcription Factors with Activities to Inhibit Neuronal Differentiation

As functional analyses of candidate genes involved in neural stem cell differentiation, we adopted overexpression and knockdown experiments using in utero electroporation. Although several transcription factors exhibited the activity to affect neuronal differentiation, intriguingly, most of them exerted inhibitory effects. Neurogenic bHLH transcription factors such as Ngn1, Ngn2, and Mash1 (mammalian achaete-scute complex homolog 1) are well known to promote neuronal differentiation and induce premature differentiation of neural stem cells when overexpressed. Unfortunately, we have so far failed to dig out novel transcription factors that positively regulate neuronal differentiation. We found that overexpression of the inhibitory factors mostly inhibited neuronal differentiation, leading to delayed onset of migration without hindering neuronal migration itself; the exception being the impaired migration that followed overexpression of Nfia. Among these transcription factors, Hmga2 was reported to be a developmental regulator of stem cell self-renewal and tumor suppressor expression, promoting fetal and young-adult neural stem cell self-renewal by decreasing p16Ink4a/p19Arf expression [14]. In our study, the expression of Hmga2 steadily declined from E11.5 through E17.5 (Fig. 2) and Hmga2 exerted an inhibitory effect on neuronal differentiation when overexpressed by neural stem cells (Fig. 5H). Trp53, a well-known tumor suppressor, is reported to be involved in maintaining the self-renewal of ESCs [37]. The expression of this gene declined between E15.5 and E17.5 in our GeneChip study (Fig. 2), and Trp53 inhibited neuronal differentiation when excessively expressed by neural stem cells (Fig. 5K). These results indicate that there are many transcription factors that are capable of inhibiting neuronal differentiation and maintaining neural stem cells. It is possible that the maintenance of undifferentiated state is very critical for normal development, and many factors that inhibit neuronal differentiation can be present redundantly, while not so many factors are required to keenly trigger the chain reaction of neuronal differentiation that may proceed automatically and stepwise thereafter. However, it remains to be addressed whether they are essentially or redundantly involved in the maintenance of neural stem cells at a physiological level or whether such activities are exerted only when excessively expressed. To address this question, we then proceeded to conduct a loss-of-function study, in which knockdown vectors were introduced into neural stem cells by in utero electroporation. Among several genes examined with knockdown vectors expressing specific siRNAs, reductions in Tcf3 and Klf15 resulted in striking consequences.

Tcf3 and Klf15 Are Involved in Maintenance of Neural Stem Cells at Different Embryonic Stages

Overexpression of Tcf3 and Klf15 led to similar inhibitions of neuronal differentiation, and the knockdown of each of these genes resulted in the opposite effect on neural stem cells. The VZ cells transfected with siRNA vectors against Tcf3 or Klf15 failed to retain their neural stem cell identity and morphology as radial glial cells, and instead, exited the neural stem cell pool and differentiated directly or via IPCs into neurons. When in utero electroporation was performed at E13.5 and analyzed at E16.5, Tcf3 exhibited a stronger effect in the gain-of-function study than did Klf15, as evidenced by the significant increase in Pax6-positive cells along with a decrease of differentiated neurons. Tcf3 is one of the downstream effectors of the canonical Wnt signaling, in which Tcf/Lef interacts with β-catenin in the nucleus and regulates transcription of target genes [38]. When stabilized β-catenin is misexpressed, radial glial cells are inhibited from generating IPCs, and thus, neural stem cells continue to proliferate with a consequent enlarged brain [39–41]. Our results support these findings and demonstrate that Tcf3 is an indispensable effector in the function of Wnt signaling to inhibit the transition from neural stem cells to IPCs.

Overexpression of Klf15 exhibited a weaker inhibitory effect on neuronal differentiation than that of Tcf3, probably because the amounts of Klf15 start to rise after E13.5 and might already have reached sufficient levels by around E15. On the other hand, siRNAs against either Tcf3 or Klf15 resulted in increased populations of Tbr2-positive cells accompanied by decreases in Pax6-positive cells, indicating that the knockdown of either Tcf3 or Klf15 promoted neuronal differentiation into IPCs at the expense of neural stem cells. In the case of siTcf3, the slight decrease in differentiated neurons might have been due to the expansion of IPCs that were actively proliferating [42], as suggested by the increase in Ki67-positive cells. The knockdown of Klf15 brought about more noteworthy results. Most transfected cells accumulated in the SVZ and just outside the SVZ, where migrating neurons usually pass through with a short period of stagnation. Most of these cells adopted a morphology resembling multipolar migrating cells, which are generally oriented horizontally in the SVZ and IZ with multiple processes instead of a radially oriented leading process [43, 44]. One possible cause of this migration defect is that the downregulation of Klf15 affected the specification or maturation of postmitotic neurons, thereby impairing the acquisition of migration ability. Another possibility is that an abrupt loss of neural stem cells (i.e., radial glial cells), due to premature neuronal differentiation caused by the transfection of siKlf15, led to a devastating loss of radial fibers that serve as a scaffold for neuronal migration, although such a drastic depletion of radial fibers was not apparent in immunostaining with anti-Nestin antibodies (data not shown). As Klf15 expression is restricted to the VZ, as shown by in situ hybridization, it is unlikely that the knockdown of Klf15 affects the specification or maturation of postmitotic neurons such as multipolar migrating cells, although it is still possible to affect neuronal fate specification or acquisition of essential items for radial migration before exit from the VZ. Given that siRNAs should be expressed and active not only in the VZ cells but also in migrating cells in the SVZ and IZ, it remains to be elucidated whether such a migration defect is a manifestation of the intrinsic activity of Klf15 or a secondary effect caused ectopically. Intriguingly, it was reported that KLF15 and Sp1 activate the transcription of human LRP5 (low-density lipoprotein receptor-related protein 5), a coreceptor of Wnt signaling [45], suggesting that Klf15 is involved in the modulation of the Wnt signaling activity as well. Furthermore, neurosphere differentiation assay revealed that primary spheres originated from the cells transfected with siRNA against Klf15 mostly lacked neurogenic activities and only retained gliogenic activities, suggesting that Klf15 is essential for the maintenance of neural stem cells with neurogenic activities during late developmental stages.

Together, these results demonstrate that both Tcf3 and Klf15 are expressed in the VZ and are involved in the maintenance of neural stem cells by inhibiting premature neuronal differentiation. Taking the temporal alterations in expression of each gene into consideration, whereas Tcf3 plays a key role from the beginning of the early neurogenic period until the late stage when its expression declines, Klf15 functions mainly during a later neurogenic period, beginning around E15, when superficial layer neurons are actively generated from neural stem cells, suggesting that these two transcription factors work in a relay during embryonic cortical development (Fig. 7).

SUMMARY

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. SUMMARY
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Gene expression profiling of GFP-positive cells from the dorsolateral telencephalon of pHes1-d2EGFP transgenic mouse embryos at different developmental stages revealed dynamic gene expression patterns in neural stem cells during cortical development. These results suggest the possibility of characterizing the embryonic neural stem cells of each developmental stage by a combination of multiple transcription factors. Among dozens of transcription factors differentially expressed within the VZ during the course of development, several of them exhibited the activity to inhibit neuronal differentiation when overexpressed, and it was revealed that Tcf3 and Klf15 play critical roles in the maintenance of neural stem cells by inhibiting premature neuronal differentiation at early and late embryonic stages, respectively.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. SUMMARY
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

We thank Kayo Nishida for technical assistance. This work was supported by research grants from the Ministry of Education, Culture, Sports, Science and Technology of Japan (to T.O. and R.K.). K.K. is currently affiliated with the Laboratory for Lymphocyte Differentiation, RIKEN Research Center for Allergy and Immunology, Yokohama, Japan.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. SUMMARY
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. SUMMARY
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
STEM_731_sm_SuppFig1.pdf118KSupplementary Figure 1. Characterization of d2EGFP-positive cells in the dorsolateral telencephalon at E11.5. (A-L): Coronal sections of dorsolateral telencephalon of pHes1-d2EGFP transgenic mouse embryos at E11.5 were immunostained with several markers. (M-X): Dissociated cells from the same region were plated, fixed two hours later, and immunostained. Cells expressing each marker among GFP-positive cells were counted and average ratios of three independent experiments were calculated.
STEM_731_sm_SuppFig2.pdf138KSupplementary Figure 2. Characterization of d2EGFP-positive cells in the dorsolateral telencephalon at E13.5. (A-L): Coronal sections of dorsolateral telencephalon of pHes1-d2EGFP transgenic mouse embryos at E13.5 were immunostained with several markers. (M-X): Dissociated cells from the same region were plated, fixed two hours later, and immunostained. Cells expressing each marker among GFP-positive cells were counted and average ratios of three independent experiments were calculated.
STEM_731_sm_SuppFig3.pdf133KSupplementary Figure 3. Characterization of d2EGFP-positive cells in the dorsolateral telencephalon at E15.5. (A-L): Coronal sections of dorsolateral telencephalon of pHes1-d2EGFP transgenic mouse embryos at E15.5 were immunostained with several markers. (M-X): Dissociated cells from the same region were plated, fixed two hours later, and immunostained. Cells expressing each marker among GFP-positive cells were counted and average ratios of three independent experiments were calculated.
STEM_731_sm_SuppFig4.pdf87KSupplementary Figure 4. Characterization of d2EGFP-positive cells in the dorsolateral telencephalon at E17.5. (A-L): Coronal sections of dorsolateral telencephalon of pHes1-d2EGFP transgenic mouse embryos at E17.5 were immunostained with several markers. (M-X): Dissociated cells from the same region were plated, fixed two hours later, and immunostained. Cells expressing each marker among GFP-positive cells were counted and average ratios of three independent experiments were calculated. (Y): Summary of immunocytochemical characterizations at each stage. The graph shows averages of three independent experiments; error bars represent the SD.
STEM_731_sm_SuppFig5.pdf82KSupplementary Figure 5. Comparison of gene expressions between two given embryonic stages using the GeneChip array. Scatter plots of all genes comparing their expression levels between the first (x) and second (y) experiments at E11.5 (A), E11.5 (x) and E13.5 (y) (B), E13.5 (x) and E15.5 (y) (C), and E15.5 (x) and E17.5 (y) (D), where ‘x’ and ‘y’ represent the x- and y-axes, respectively. Note that the dispersion became wider as development proceeded. Green lines indicate the range of two-fold differences.
STEM_731_sm_SuppFig6.pdf68KSupplementary Figure 6. Temporal expression patterns of bHLH transcription factor genes from the GeneChip array. Alterations in gene expression of Hes and Hes-related genes (A, B), and neurogenic bHLH genes (C, D) during development. Note that Ngn2 and Neurod6 (Math2) were highly expressed during the neurogenic period from E13.5 through E15.5, and then declined between E15.5 and E17.5 in accordance with the transition of neural stem cells from neurogenesis to gliogenesis. (A, C) Relative expression values normalized to those at E11.5. (B, D) Raw signal intensities.
STEM_731_sm_SuppFig7.pdf105KSupplementary Figure 7. Characterization of transfected cells by in utero electroporation. (A-O): Coronal sections of dorsolateral telencephalon were immunostained to characterise the cells transfected with pEF-EGFP vectors six hours following in utero electroporation at E13.5. (A'-O'): Dissociated cells from the same region were plated, fixed two hours later, and immunostained. Cells expressing each marker among GFPpositive cells were counted and average ratios of three independent experiments were calculated. (P'): The graph shows averages of three independent experiments; error bars represent the SD.
STEM_731_sm_SuppFig8.pdf60KSupplementary Figure 8. Overexpression of genes transduced by in utero electroporation. (A,B): Immunostaining with anti-GFP antibodies, indicating the transfected regions of dorsolateral telencephalon 24 hours following in utero electroporation at E13.5. (C,D): In situ hybridization with probes for Tcf3 (C) or Klf15 (D) in the adjacent sections.
STEM_731_sm_SuppFig9.pdf188KSupplementary Figure 9. Impacts of overexpression or knock-down of Tcf3 or Klf15 on neuronal differentiation and migration. Coronal sections of dorsolateral telencephalon were immunostained to analyze the fates of transfected cells three days following in utero electroporation at E13.5; siRNA knock-down of Tcf3 (siTcf3) or Klf15 (siKlf15) along with scrambled siRNA for Tcf3 (siTcf3scr) or Klf15 (siKlf15scr) as a control, and overexpression of Tcf3 (pEF-HATcf3) or Klf15 (pEF-HA-Klf15). Anti-GFP (green) and antibodies for various markers (red).
STEM_731_sm_SuppFig10.pdf61KSupplementary Figure 10. Expression of Tcf3 and Klf15 in postnatal and adult brain. In situ hybridization with probes for Tcf3 (A,B,J) or Klf15 (C-I) on coronal sections of postnatal day 0 (P0) (A-D) and adult (six weeks) (E-J) brains. Anterior (A,C) or posterior (B,D) periventricular region of P0 brain, anterior periventricular region (E-G, J), dentate gyrus of hippocampus (H), and cerebellum (I) of adult brain. (F): Higher magnification image of the rectangular area in (E). (G): Higher magnification image of the rectangular area in (F).
STEM_731_sm_SuppFig11.pdf95KSupplementary Figure 11. Expression of Klf15 in the SVZ of adult brain. BrdU was administered in drinking water to 6 week old mice for one week, and BrdU retaining cells were examined with anti-BrdU antibodies (green; A-C,E-G,I) in the anterior periventricular region following 1 week (A), 2 weeks (B-E), or 4 weeks (F-I). Klf15 mRNA was detected by fluorescence in situ hybridization (FISH) (red; A,B,DF, H,I,K,L,N,O). (J-O): Double staining of FISH for Klf15 and GFAP (green; K,L,N,O).
STEM_731_sm_SuppTable1.pdf51KSupplementary Table 1. Summary of the numbers of genes that exhibited more than two-fold changes between the different embryonic stages. Among 45,101 genes on the GeneChip array, 25,940 were flagged as ‘Present’ in the GFP-positive cells at E11.5. The numbers of genes that exhibited more than two-fold changes between the two different embryonic stages, consecutive down- or upregulations from E11.5 through E15.5, or a peak expressions either at E13.5 or 15.5, are listed along with the expected functions.
STEM_731_sm_SuppTable2.pdf186KSupplementary Table 2. Summary of the transcription factors that exhibited more than two-fold changes between the two time points. All candidate transcription factors that showed more than two-fold changes between the two different embryonic stages are listed. The numbers of transcription factors that were up-regulated between E11.5 and E13.5 (E11.5<E13.5), E13.5 and E15.5 (E13.5<E15.5), or E15.5 and E17.5 (E15.5<E17.5), and that were down-regulated between E11.5 and E13.5 (E11.5>E13.5), E13.5 and E15.5 (E13.5>E15.5), or E15.5 and E17.5 (E15.5>E17.5), are summarized in the inset table.
STEM_731_sm_SuppTable3.pdf210KSupplementary Table 3. Expression patterns of candidate genes confirmed by in situ hybridization. The RNA probes for in situ hybridization were synthesized in vitro using the full- or partial-length cDNAs as templates. Genes shown in red were exclusively expressed in the VZ; those shown in blue were detected in both the VZ and outer layers including the CP. circles: obvious signal, triangles: vague signal, -: no signal, n.d.: not yet determined, *medial: expressed in dorsomedial telencephalon, *ventral: expressed in ventral telencephalon.
STEM_731_sm_SuppTable4.pdf56KSupplementary Table 4. Flowchart of gene expression profiling and functional analyses in this study. Among 45,101 genes on the GeneChip array, 25,940 were flagged as ‘Present’ in GFPpositive cells at E11.5. There were 8,431 genes that exhibited more than two-fold changes in signal value between two given time points, and 92 genes encoding transcription factors were subjected to further analyses. We performed in situ hybridization to confirm whether these genes are actually expressed by the VZ cells, and identified 30 genes (among approximately 80 genes examined) that were expressed in the VZ. The molecular functions of these genes were evaluated by overexpression and knock-down experiments.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.