miR-17 regulates the proliferation and differentiation of the neural precursor cells during mouse corticogenesis

Authors

  • Susu Mao,

    1. Jiangsu Engineering Research Center for microRNA Biology and Biotechnology, State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University School of Life Sciences, China
    Search for more papers by this author
  • Hanqin Li,

    1. Jiangsu Engineering Research Center for microRNA Biology and Biotechnology, State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University School of Life Sciences, China
    Search for more papers by this author
  • Qi Sun,

    1. Jiangsu Engineering Research Center for microRNA Biology and Biotechnology, State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University School of Life Sciences, China
    Search for more papers by this author
  • Ke Zen,

    Corresponding author
    1. Jiangsu Engineering Research Center for microRNA Biology and Biotechnology, State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University School of Life Sciences, China
    2. Department of Virology, University of California School of Public Health, Berkeley, CA, USA
    • Correspondence

      K. Zen, Jiangsu Engineering Research Center for microRNA Biology and Biotechnology, State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University School of Life Sciences, Nanjing 210093, China

      Fax: +86 25 84530990

      Tel: +86 25 84530990

      E-mail: kzen@nju.edu.cn

      C.-Y. Zhang, Jiangsu Engineering Research Center for microRNA Biology and Biotechnology, State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University School of Life Sciences, Nanjing 210093, China

      Fax: +86 25 84530990

      Tel: +86 25 84530990

      E-mail: cyzhang@nju.eud.cn

      L. Li, Jiangsu Engineering Research Center for microRNA Biology and Biotechnology, State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University School of Life Sciences, Nanjing 210093, China

      Fax: +86 25 84530990

      Tel: +86 25 84530990

      E-mail: lijing@sibs.ac.cn

    Search for more papers by this author
  • Chen-Yu Zhang,

    Corresponding author
    1. Jiangsu Engineering Research Center for microRNA Biology and Biotechnology, State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University School of Life Sciences, China
    • Correspondence

      K. Zen, Jiangsu Engineering Research Center for microRNA Biology and Biotechnology, State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University School of Life Sciences, Nanjing 210093, China

      Fax: +86 25 84530990

      Tel: +86 25 84530990

      E-mail: kzen@nju.edu.cn

      C.-Y. Zhang, Jiangsu Engineering Research Center for microRNA Biology and Biotechnology, State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University School of Life Sciences, Nanjing 210093, China

      Fax: +86 25 84530990

      Tel: +86 25 84530990

      E-mail: cyzhang@nju.eud.cn

      L. Li, Jiangsu Engineering Research Center for microRNA Biology and Biotechnology, State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University School of Life Sciences, Nanjing 210093, China

      Fax: +86 25 84530990

      Tel: +86 25 84530990

      E-mail: lijing@sibs.ac.cn

    Search for more papers by this author
  • Liang Li

    Corresponding author
    1. Jiangsu Engineering Research Center for microRNA Biology and Biotechnology, State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University School of Life Sciences, China
    • Correspondence

      K. Zen, Jiangsu Engineering Research Center for microRNA Biology and Biotechnology, State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University School of Life Sciences, Nanjing 210093, China

      Fax: +86 25 84530990

      Tel: +86 25 84530990

      E-mail: kzen@nju.edu.cn

      C.-Y. Zhang, Jiangsu Engineering Research Center for microRNA Biology and Biotechnology, State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University School of Life Sciences, Nanjing 210093, China

      Fax: +86 25 84530990

      Tel: +86 25 84530990

      E-mail: cyzhang@nju.eud.cn

      L. Li, Jiangsu Engineering Research Center for microRNA Biology and Biotechnology, State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University School of Life Sciences, Nanjing 210093, China

      Fax: +86 25 84530990

      Tel: +86 25 84530990

      E-mail: lijing@sibs.ac.cn

    Search for more papers by this author

Abstract

MicroRNAs (miRNAs) are endogenously expressed small, non-coding nucleotides that repress gene expression at the post-transcriptional level. In mammals, the developing brain contains a large, diverse group of miRNAs, which suggests that they play crucial roles in neural development. In the present study, we analyzed the miRNA expression patterns in the mouse cortex at various developmental stages. We found that miR-17 family miRNAs were highly expressed in the cortex during early developmental stages, and that their expression levels gradually decreased as the cortex developed. Further investigation revealed that the change in miR-17-5p expression occurred in the ventricular zone/sub-ventricular zone. In addition to promoting cell proliferation, miR-17-5p also influences the differentiation fate of neural precursor cells exposed to bone morphogenetic protein 2. Moreover, we show that these effects of miR-17-5p were mainly the result of regulating the bone morphogenetic protein signaling pathway by repressing expression of the bone morphogenetic protein type II receptor. Taken together, these findings suggest that miR-17 family members play a pivotal role in regulating cell activity during early development of the mouse cortex.

Abbreviations
bFGF

bovine fibroblast growth factor

BMP2

bone morphogenetic protein 2

BMPR2

bone morphogenetic protein type II receptor

DAPI

4′,6-diamidino-2-phenylindole

EdU

5-ethynyl-2′-deoxyuridine

EGFP

enhanced green fluorecence protein

GFAP

glial fibrillary acidic protein

hEGF

human epidermal growth factor

MAP2

microtubule-associated protein 2

miRNA

microRNA

mmu

Mus musculus

NPCs

neural precursor cells

PH3

phosphohistone H3

P-Smad1/5

phosphorylated Smad1/5

RT

room temperature

VZ/SVZ

ventricular zone/subventricular zone

Introduction

MicroRNAs (miRNAs) are approximately 22 nucleotide fragments derived from hairpin precursors that regulate gene expression by targeting the 3′ UTR of mRNAs [1, 2]. Large diverse groups of miRNAs have been identified in mammals, and have been found to influence almost every cellular and developmental process investigated [3, 4]. A substantial amount of evidence suggests that miRNAs are enriched in the central nervous system, and that they play significant roles in neural development and physiology [5-7]. For example, mutations in Drosophila Ago1 cause a severe loss in all types of neurons and glial cells, and Dicer-deficient mice display abnormal neurogenesis and gliogenesis in the developing central nervous system [8, 9]. These findings indicate that miRNAs are essential for the differentiation, maturation and survival of both neurons and glial cells.

In the developing cortex, two germinal zones, the pseudostratified epithelium of the ventricular zone (VZ) and the underlying sub-ventricular zone (SVZ), have been implicated as major sites of neurogenesis and gliogenesis [10]. Furthermore, it has been shown that neurons and glial cells may be generated from a common neural precursor in the fetal telencephalic neuroepithelium [11]. However, the mechanisms underlying regulation of the neurogenesis/gliogenesis transition remain to be elucidated. Currently, it is believed that commitment of neural precursors is controlled by both intrinsic signals and external cues [12, 13]. It has been reported that bone morphogenetic protein 2 (BMP2) signaling plays a crucial role in determining cell fate by promoting astrocytogenesis during neural development [14, 15]. BMP2 activation is mediated by heterotetrameric serine/threonine kinase BMP receptors (BMPR1/BMPR2) that phosphorylate the receptor-regulated Smads proteins (Smad1, 5 and 8), which then initiates transcription of the downstream negative helix-loop-helix (HLH) factors Id1, Id3 and Hes-5 that inhibit expression of neurogenic genes [16, 17].

Numerous studies have indicated that several miRNAs are involved in determining neural cell fate and proliferation during neural development [18, 19]. To further understand the role that miRNAs play during cortex development, we screened the miRNA expression levels in the mouse cortex at various developmental stages. We found that miR-17 family miRNAs are robustly expressed during early development of the cortex (embryonic day E12.5), and that their expression levels gradually decrease as the cortex develops. In addition, we also provide evidence suggesting that miR-17-5p represses the expression of BMPR2 in cortical cells by targeting the 3′ UTR of its mRNA. A functional analysis revealed that over-expressing miR-17-5p in neural precursor cells (NPCs) increases cell proliferation under maintenance conditions but promotes neurogenesis when cells are treated with BMP2 under differentiation conditions. Furthermore, down-regulating miR-17-5p expression enhanced the cellular response to BMP2, which facilitates astrocytogenesis during differentiation. We also provide evidence indicating that the effects of miR-17-5p are predominantly the result of its ability to regulate the BMP signaling pathway by targeting BMPR2. Taken together, these observations suggest that miR-17 family miRNAs play an important role during cortex development.

Results

Expression of miR-17 family members in the mouse cortex at various developmental stages

In the present study, we first analyzed the miRNA expression levels in the mouse cortex at various developmental stages (embryonic day E12.5, E15.5, E18.5 and postnatal day P60) using a miRNA array (Table 1). We selected candidates based on the following criteria: signal intensity of each miRNA at E12.5 > 1000, and fold change of each miRNA (E12.5/P60) > 5. Interestingly, we found that the expression levels of miR-17 family members were higher at the early developmental stage (E12.5) compared with the postnatal stage (P60). Additionally, we also observed that the expression levels of miR-17 family members gradually decreased over the late embryonic stages (E15.5 and E18.5). The miR-17 family consists of six members [20] located in three separate genome clusters (Fig. 1A). Our data show that not only are the miR-17 family members highly expressed in the cortex at E12.5, but that the other miRNAs (miR-18a, miR-19b, miR-25 and miR-92a) located in these three clusters are also highly expressed in the cortex at E12.5. This finding indicates that transcription of these three clusters is increased during the early development of the cortex (Table 1). We further assessed the expression levels of the miR-17 family members by quantitative real-time PCR (Fig. 1B). The results are consistent with those from the miRNA array (Table 1). Because these miRNAs have the same 5′ seed region, which is important for mRNA target recognition [21], and because it is the most abundantly expressed miRNA belonging to the miR-17 family, we focused on miR-17-5p for the remainder of this study (Table 1).

Table 1. Selected miRNAs whose levels increased as the mouse cortex developed. Asterisks indicate members of the miR-17 family.
Gene nameDevelopmental stagesFold change
E12.5E15.5E18.5P60E12.5/P60
mmu-miR-15b2184.81434.31045.5154.714.1
mmu-miR-17-5p*21162.217140.915685.21048.620.2
mmu-miR-18a5679.74839.02212.099.357.2
mmu-miR-19b3540.02360.62192.2220.616.0
mmu-miR-20a*13760.611676.29009.6461.529.8
mmu-miR-20b*2045.41109.6678.974.627.4
mmu-miR-252456.21464.51092.9106.923.0
mmu-miR-92a5586.94507.42185.8330.816.9
mmu-miR-93*15034.916196.713492.22204.16.8
mmu-miR-106a*5960.33554.62533.4178.633.4
mmu-miR-106b*8152.27546.97619.71133.87.2
mmu-miR-130a4293.22459.82039.0327.813.1
mmu-miR-130b7357.66027.93468.5236.531.1
mmu-miR-181d4288.13281.33072.2616.87.0
mmu-miR-296-3p1898.93321.42380.2121.815.6
mmu-miR-4941276.0419.6276.9168.07.6
mmu-miR-21326441.16510.25153.21158.55.6
mmu-miR-21373832.24151.72341.2400.19.6
mmu-miR-21401421.91452.11808.2236.36.0
Figure 1.

Expression of miR-17 family members in the mouse cortex at various developmental stages. (A) Schematic representation of the murine miR-17-92 cluster and its paralogs, the miR-106a-363 and miR-106b-25 clusters. The miRNAs encoded in the three clusters are categorized into four separate miRNA families as indicated, including the miR-17 family. (B) Quantitative real-time PCR verifying miR-17 family miRNA expression levels at various developmental stages. The expression levels of the miRNAs are normalized to the U6 snRNA expression level, and the expression rates between the various samples are presented as fold changes relative to the lowest normalized expression value. Values are means ± SEM (= 3). (C) Expression patterns of miR-17-5p and miR-124 (a neuron-enriched miRNA used as a control) in the mouse cortex at various developmental stages were analyzed using in situ hybridization with digoxigenin-labeled locked nucleic acid probes. The miR-17-5p level was found to decrease significantly in the ventricular zone and the underlying sub-ventricular zone (VZ/SVZ) as the cortex develops. cp, cortical plate. Scale bar (bold) = 100 μm.

To further understand the role that miR-17 family members play during cortex development, we investigated the expression pattern of miR-17-5p in the mouse cortex at various developmental stages (E15.5, E18.5 and P1) during which NPCs are proliferating and differentiating into neurons and glial cells [22]. The in situ hybridization results showed that miR-17-5p is mainly expressed in the cortical plate and the VZ/SVZ region, and that another neuron-enriched miRNA, miR-124, is exclusively expressed in the cortical plate (Fig. 1C). These results indicate that in the VZ/SVZ region, miR-17-5p is expressed in the neural precursors. In addition, we also analyzed the expression pattern of miR-106a (another member of miR-17 family), which is similar to that of miR-17-5p in the developing cortex except that the signal intensity is lower (Fig. S1). Furthermore, in situ hybridization staining demonstrated that miR-17-5p expression gradually decreased in those two areas as the cortex developed (Fig. 1C). Because most of the neurons in the cortical plate are derived from the VZ/SVZ region, which is the predominant source of neural precursors for neurogenesis and gliogenesis, we focused our remaining experiments on evaluating the effects of miR-17-5p in the VZ/SVZ region.

miR-17-5p promotes proliferation of neural precursor cells derived from the embryonic cortex

The VZ/SVZ region is the main area in the cortex where neural precursors proliferate and give rise to neurons and glial cells at various developmental stages [23]. Because our results showed that the miR-17-5p expression pattern undergoes a remarkable change in the VZ/SVZ region during corticogenesis, we investigated the role that miR-17-5p plays in neural precursors derived from the VZ/SVZ region. First, we made use of a primary mouse NPC in vitro culture system as described previously [24]. We further characterized our culture using markers of various types of cells, and showed that most of our cultured cells are nestin-positive (> 95%) under maintenance conditions, and there are few β-tubulin III-positive or glial fibrillary acidic protein (GFAP)-positive cells (< 5%) (Fig. S2). To analyze the effects of miR-17-5p on NPC proliferation, we over-expressed miR-17-5p in NPCs using a lentivirus that increased the miR-17-5p expression level twofold (Fig. 2A). Using a 5-ethynyl-2′-deoxyuridine (EdU) incorporation assay and phosphohistone H3 (PH3) staining, we show that up-regulation of miR-17-5p expression results in a significant increase in proliferation of NPCs under maintenance conditions (Fig. 2C–F). When we knocked down the endogenous miR-17-5p expression level using antagomiR oligonucleotides specifically directed against miR-17-5p (antago-17-5p) (Fig. 2B), we found that both EdU and PH3 signals in the NPCs are reduced, indicating a decrease in cell proliferation (Fig. 2G–J). Furthermore, we performed this analysis in NPCs under maintenance conditions at multiple time points after transfection and found similar results (Fig. 2E,F,I,J).These findings are consistent with previous studies showing that miR-17 family members promote cell proliferation in other tissues as well as neural stem cells [25-27].

Figure 2.

miR-17-5p increases the proliferation of NPCs. (A,B) Quantitative real-time PCR was used to determine the efficiency of over-expressing and inhibiting miR-17-5p using lentivirus (lenti-17-5p) and antagomiR oligonucleotides (antago-17-5p), respectively. Values are means ± SEM (= 3). Asterisks indicate a statistically significant difference compared with the control (*< 0.05). (C,D) NPCs were infected with lentivirus over-expressing miR-17-5p (lenti-17-5p) or scrambled control lentivirus (control). After 12 h, the cells were placed in fresh medium and cultured for another 2 days, followed by EdU assay or immunofluorescent staining for PH3. Nuclear staining was performed using DAPI. Scale bars = 50 μm. (E,F) After infection with lenti-17-5p or control lentivirus for 12 h, NPCs were cultured in fresh medium for another 2, 3 or 4 days. Quantitative analysis was performed to determine the percentage of cells positively labeled for proliferation markers EdU and PH3. Values are means ± SEM (= 3). Asterisks indicate a statistically significant difference compared with the control (*< 0.05). (G,H) NPCs were transfected with antago-17-5p or scrambled control antagomiR for 2 days, and then the cells were subjected to EdU assay or immunofluorescent staining for PH3. Nuclear staining was performed using DAPI. Scale bars = 50 μm. (I,J) NPCs were transfected with antago-17-5p or control antagomiR for 2, 3 or 4 days. Quantitative analysis was performed to determine the the percentage of cells positively labeled for the proliferation markers EdU and PH3. Values are means ± SEM (= 3). Asterisks indicate a statistically significant difference compared with the control (*< 0.05).

miR-17-5p regulates the neurogenesis/astrocytogenesis transition during NPC differentiation by repressing BMP signaling via targeting BMPR2

It has been reported that miR-17 family members repress BMPR2 expression in human pulmonary artery endothelial cells [28], and thus regulate the BMP signaling pathway. Additionally, findings from previous studies suggest that the BMP signaling pathway is involved in the neurogenesis/astrocytogenesis transition during corticogenesis [11, 14]. Taken together, these findings suggest that miR-17 family members may play a role in fate determination during cell differentiation in cortex development.

To determine whether miR-17-5p regulates the BMP signaling pathway in the nervous system, we first performed a luciferase assay showing that miR-17-5p significantly inhibits the translational efficiency of a reporter gene fused with the 3′ UTR of the BMPR2 mRNA. This inhibition was drastically reduced when we mutated the miR-17-5p binding site (Fig. 3A,B). Furthermore, we found that the over-expression of miR-17-5p results in a decrease in the BMPR2 expression level, and that inhibition of miR-17-5p significantly increases the cellular BMPR2 level in HEK 293T cells (Fig. 3C) and NPCs (Figs 4C and 5A). We also analyzed the BMPR2 expression level in the mouse cortex at various developmental stages. Our western blot analysis results show that there is a time-dependent increase in the BMPR2 protein expression level in the mouse cortex (Fig. 3D) that is negatively correlated with changes in the expression levels of miR-17 family members during corticogenesis (Fig. 1B). Moreover, we obtained similar results for the BMPR2 expression pattern in the VZ/SVZ region (Fig. 3E). Together with the results for the miR-17-5p expression pattern in the cortical VZ/SVZ region at the various developmental stages (Fig. 1C), these results strongly indicate that miR-17 family members regulate the BMPR2 expression level in the VZ/SVZ region during corticogenesis.

Figure 3.

miR-17-5p inhibits BMPR2 expression. (A) Schematic drawing of the luciferase reporter. The mutated sequence in the 3′ UTR of the BMPR2 mRNA is indicated. (B) Luciferase activity was measured 24 h after transfecting HEK 293T cells. Reporter plasmids with the wild-type (Wt) or mutated (Mut) 3′ UTR of BMPR2 were transfected either alone (vector) or with miR-17-5p (mimics) or scrambled control mimics (control). Values are means ± SEM (= 3). Asterisks indicate a statistically significant difference compared with the control (*< 0.05). (C) Protein analysis of the BMPR2 expression level in HEK 293T cells transfected with miR-17-5p mimics (miR-17-5p) or miR-17-5p inhibitors (inh-17-5p). A representative western blot image (upper panel) and quantitative analysis of the BMPR2 expression level (bottom panel) are shown. Values are means ± SEM (= 3). Asterisks indicate a statistically significant difference compared with the control (*< 0.05). (D) BMPR2 expression levels in the mouse cortex at various developmental stages were determined using western blot analysis. Three independent experiments were performed. (E) Representative photographs from the immunohistochemistry analysis of BMPR2 expression in the mouse cortex at various developmental stages. Boxes in the upper images indicate the area shown at magnification in the lower images; these results show that the BMPR2 expression level increased in the VZ/SVZ region as the cortex developed. Samples treated without primary antibody are shown as a control. Scale bars = 100 μm (upper panels) and 50 μm (bottom panels).

Figure 4.

Over-expression of miR-17-5p increased neurogenesis in NPCs. (A) Detached NPCs were cultured for 48 h, and then treated with or without 10 ng/mL BMP2 for 4 h. The cells were then cultured for another 48 h in the absence of bFGF and hEGF, and subjected to immunofluorescent staining. Inset: DAPI staining of the cell nuclei in the field of view. Scale bar = 50 μm. (B) The ratio of β-tubulin III-positive cells to GFAP-positive cells in the NPC population treated with or without BMP2. Values are means ± SEM (= 3). Asterisks indicate a statistically significant difference compared with the control (*< 0.05). (C) Expression of BMPR2 and p-Smad1/5 in the NPCs decreased 72 h after infecting them with lentivirus expressing miR-17-5p (lenti-17-5p) or control lentivirus (control). GAPDH and Smad1/5/8 were used as internal controls. (D) Quantification of the protein expression levels from the experiment shown in (C). Values are means ± SEM (= 3). Asterisks indicate a statistically significant difference compared with the control (*< 0.05). (E) NPCs were infected with lenti-17-5p or control lentivirus and cultured for 72 h. After treating the cells with BMP2 for 4 h, they were cultured for another 48 h in the absence of bFGF and hEGF, and then subjected to immunofluorescent staining. Inset: DAPI staining of cell nuclei in the field of view. Scale bar = 50 μm. (F) The ratio of β-tubulin III-positive cells to GFAP-positive cells was determined. Values are means ± SEM (= 3). Asterisks indicate a statistically significant difference compared with the control (**< 0.01). (G) NPCs were infected with lentivirus engineered to express miR-17-5p and EGFP (lenti-EGFP-17-5p) or EGFP only (control). Next, the cells were treated as described in (E), and then subjected to immunofluorescent staining. Inset: DAPI staining of cell nuclei in the field of view. Scale bar = 50 μm. (H) The percentage of marker-positive cells in each EGFP-positive cell population was determined. Values are means ± SEM (= 3). Asterisks indicate a statistically significant difference compared with the control (**< 0.01).

Figure 5.

Inhibition of miR-17-5p increased astrocytogenesis in NPCs. (A) The expression levels of BMPR2 and p-Smad1/5 in NPCs increased 48 h after transfection with antagomiR oligonucleotides specifically directed against miR-17-5p (antago-17-5p) or scrambled control antagomiR (control). GAPDH and Smad1/5/8 were used as internal controls. (B) Quantification of the protein expression levels from the experiment shown in (A). Values are means ± SEM (= 3). Asterisks indicate a statistically significant difference compared with the control (*< 0.05). (C,E) NPCs were incubated with antago-17-5p or control antagomiR for 48 h. After treatment with BMP2 for 4 h, the cells were cultured for another 48 h in the absence of bFGF and hEGF, and then subjected to immunofluorescent staining. Inset: DAPI staining of cell nuclei in the field of view. Scale bar = 50 μm. (D,F) The ratio of MAP2/β-tubulin III-positive cells to GFAP-positive cells was determined. Values are means ± SEM (= 3). Asterisks indicate a statistically significant difference compared with the control (**< 0.01).

Next, we further investigated the effects of miR-17-5p on the neurogenesis/astrocytogenesis transition during NPC differentiation. We demonstrate that BMP2 stimulation significantly decreases the ratio of β tubulin III-positive to GFAP-positive cells in the NPC population under differentiation conditions (Fig. 4A,B), and this finding is consistent with a previous report [11]. Additionally, we also found that over-expressing miR-17-5p reduces the BMPR2 protein level in cells and decreases the level of phosphorylated Smad1/5, which suggests that BMP signaling is repressed (Fig. 4C,D). Moreover, the ratio of β-tubulin III-positive/GFAP-positive cells was significantly higher in cells over-expressing miR-17-5p compared with that in control cells that were treated with BMP2 under differentiation conditions (Fig. 4E,F). To evaluate the direct impact of miR-17-5p on fate determination during NPC differentiation, we removed exogenous BMP2 and used noggin to inhibit potential functional BMP signaling remaining in our system, as it has been reported that endogenously expressed BMPs are present in SVZ cells [11]. We found that miR-17-5p on its own has little impact on the neurogenesis/astrocytogenesis transition during NPC differentiation, as the β-tubulin III-positive/GFAP-positive cell ratio shows no significant changes (Fig. S3). These observations indicate that the effect of miR-17-5p on cell fate determination is predominantly the result of its ability to regulate the BMP signaling pathway by targeting BMPR2.

To further elucidate the precise effects of miR-17-5p on neurogenesis and astrocytogenesis, we measured the ratio of microtubule-associated protein 2 (MAP2)-positive cells and GFAP-positive cells to DAPI-positive cells separately. We found that over-expressing miR-17-5p increases the MAP2-positive/enhanced green fluorescence protein (EGFP) -positive cell ratio and decreases the GFAP-positive/EGFP-positive cell ratio (Fig. 4G,H). Furthermore, inhibiting miR-17-5p resulted in an increase in Smad phosphorylation through up-regulation in BMPR2 expression upon BMP2 treatment (Fig. 5A,B). When we blocked miR-17-5p function using antago-17-5p, the neuron/astrocyte ratio derived from NPCs (MAP2-positive/GFAP-positive or β-tubulin III-positive/GFAP-positive) was markedly reduced (Fig. 5C–F). These results suggest that miR-17-5p promotes neurogenesis at the expense of astrocytogenesis in NPCs treated with BMP2.

Taken together, these findings indicate that miR-17 family members play a crucial role in regulating the neurogenesis/astrocytogenesis transition, and this regulatory effect is the result of repressing BMP signaling via targeting BMPR2.

BMPR2 siRNA increases the proliferation and neurogenesis of NPCs

It has been reported that miR-17 family members affect multiple signaling pathways that regulate cellular activity [29, 30]. To evaluate the effects of miR-17-5p mediated by only the BMP signaling pathway in neural precursors, we specifically knocked down BMPR2 expression in NPCs using a modified siRNA. As a result, the BMPR2 protein level was significantly reduced compared with the control. Additionally, the phosphorylation of Smad1/5 was also decreased, indicating that BMP signaling was inhibited (Fig. 6A,B). Next, we examined the cell proliferation and found that down-regulation of BMPR2 increased cell proliferation of NPCs under maintenance conditions (Fig. 6C,D). This result suggests that BMPR2 is one of the major miR-17-5p targets used to regulate NPC proliferation. Furthermore, we also demonstrate that the BMPR2 siRNA mimics the effects of miR-17-5p on cell fate determination during NPC differentiation. The depletion of BMPR2 significantly increased the MAP2-positive/GFAP-positive cell ratio (Fig. 6E,F). These findings support the proposal that the BMP signaling pathway is a major pathway through which miR-17-5p regulates NPC proliferation and fate determination during corticogenesis.

Figure 6.

BMPR2 siRNA mimics the effects of miR-17-5p in NPCs. (A) The expression levels of BMPR2 and p-Smad1/5 in NPCs decreased 48 h after incubating them in 200 nm cholesterol, 2′-O-methyl and phosphorothioate (Chol-OMe-PS) modified BMPR2 siRNA (si-BMPR2). GAPDH and Smad1/5/8 were used as internal controls. (B) Quantifications for the protein expression levels from the experiment shown in (A). Values are means ± SEM (= 3) Asterisks indicate a statistically significant difference compared with the control (*< 0.05). (C) NPCs were incubated with 200 nm Chol-OMe-PS-modified si-BMPR2 or control siRNA (control) for 48 h before performing an EdU incorporation assay. Representative images of individual groups are shown. Scale bar = 50 μm. (D) Quantitative analysis of the EdU assay presented as the ratio of EdU-positive cells to total DAPI-positive cells. Values are means ± SEM (= 3). Asterisks indicate a statistically significant difference compared with the control (*< 0.05). (E) NPCs were incubated in 200 nm si-BMPR2 or control siRNA for 48 h. After treatment with BMP2 for 4 h, the cells were cultured for another 48 h in the absence of bFGF and hEGF, and then subjected to immunofluorescent staining. Inset: DAPI staining of cell nuclei in the field of view. Scale bar = 50 μm. (F) The ratio of MAP2-positive cells to GFAP-positive cells was determined. Values are means ± SEM (= 3). Asterisks indicate a statistically significant difference compared with the control (**< 0.01).

Taken together, our results suggest that miR-17-5p inhibits BMP signaling by targeting BMPR2 expression, through which it further regulates NPC activity during corticogenesis.

Discussion

The accumulating evidence suggests that miRNAs play important roles in the developing central nervous system, including roles in regulating cell proliferation, differentiation, maturation and even survival [31, 32]. However, to obtain a complete understanding of this regulatory network, further investigations are required to identify the functional consequences of specific miRNAs. Here we investigated the role that miR-17-5p plays in NPC proliferation and differentiation during corticogenesis.

Robust expression of miR-17 family members in the early developing mouse cortex

The cell activity that takes place during cortical development may be divided into the following three phases: the expansion phase, the neurogenic phase and the gliogenic phase. During these phases, multipotent cells mainly undergo proliferation in the first phase and differentiation in the following two phases, respectively [22]. In the present study, we provide strong evidence that miR-17 family miRNAs remain up-regulated during early development of the mouse cortex, which mainly corresponds to the first two phases. Our miRNA array and quantitative analysis showed that six members of the miR-17 family are highly expressed in the E12.5 cortex, and that other miRNAs located in the same clusters as the miR-17 family members also have similar expression patterns during cortex development. As the embryo develops, the expression levels of the miR-17 family members decreased in the developing cortex, especially within the VZ/SVZ region. The VZ/SVZ region is the dorsal germinal zone of the telencephalon, which is the main region of cell proliferation during corticogenesis [33]. Additionally, it has also been suggested that the VZ/SVZ region is a major source of neurogenesis and gliogenesis during cortex development [34]. The progenitors in the VZ/SVZ region generate many types of neural cells, and this differentiation is regulated by a combination of extrinsic and intrinsic factors [34]. Taken together, these results suggest that these miR-17 family members may play a role in NPC proliferation and differentiation during corticogenesis.

miR-17-5p increases NPC proliferation under maintenance conditions

miR-17 family miRNAs have been extensively studied, and their ability to regulate cell proliferation in many tissues has been reported [30]. However, it remains unknown whether miR-17 family members promote or inhibit cell growth. For example, it has been reported that up-regulation of miR-17 promotes cell proliferation and tumor growth by targeting the Rho family GTPase 3 (RND3) tumor suppressor gene in colorectal carcinoma, but there is also evidence suggesting that mir-17-5p represses breast cancer cell proliferation by targeting amplified in breast cancer 1 (AIB1) [35, 36]. In the adult mouse brain, the miR17-92 cluster facilitates cell proliferation in the SVZ after ischemic stress by repressing the transforming growth factor-β signaling pathway [37]. In the present study, we examined the effects of miR-17-5p on NPC proliferation during the corticogenic stage. and demonstrate that miR-17-5p over-expression increases the proliferation of NPCs derived from the VZ/SVZ region under maintenance conditions. This observation appears to be contradictory to the result of a previous report, which found that over-expression of the miR-17-92 cluster reduced neural stem cell proliferation by targeting E2F transcription factor 1(E2F1) [38]. However, another study has also suggested that the miR-17-92 cluster promotes NPC expansion by targeting phosphatase and tensin homolog deleted on chromosome ten (PTEN) [27]. There are several possible reasons for the discrepancy between these observations. First, the experimental models used in these studies are different. Palm et al. used a primary mouse NSC in vitro culture from a niche-independent system [38], while our cultured NPCs were directly separated from the E14.5 mouse telencephalons as previously described [24]. Bian et al. performed their study using genetic mouse models [27]. Furthermore, the effects of the miR-17 family on cell proliferation are complicated as miR-17 family members regulate many proteins involved in cell proliferation. Among those targeted genes, some (such as E2F1 or AIB1) are reported to promote proliferation while others (such as PTEN and RND3) are suggested to decrease cell expansion [27, 35, 36, 38]. Therefore, the effects of miR-17 family members may be different in various models, and the difference between these in vitro and in vivo models may result in different responses under similar conditions. Second, there is also difference between miR-17-5p and the miR-17-92 cluster in regulation of expression of downstream proteins. The miR-17-92 cluster contains miRNAs belonging to other families, and thus may have more target genes than miR-17-5p alone. As a result, the physiological effects of miR-17-5p and the miR-17-92 cluster may be different.

Interestingly, when we specifically blocked BMPR2 expression using siRNA, we also observed an increase in NPC proliferation in the absence of exogenous BMP2 stimulation. This finding may be the result of endogenously expressed BMPs in the SVZ cells [11]. As there is evidence suggesting that a high concentration of BMP inhibits neural cell proliferation [39], it is thought that inhibition of BMP signaling may also contribute to the miR-17-5p-mediated increase in NPC proliferation.

miR-17-5p regulates the neurogenesis/astrocytogenesis transition by repressing BMP signaling via targeting BMPR2

It has been suggested that miR-17-5p affects BMP signaling by negatively regulating the BMPR2 expression level in other tissues [28]. Therefore, we investigated whether this type of regulation occurs in the neural progenitors during corticogenesis. Using a luciferase assay and western blot analysis, we provide strong evidence that miR-17-5p directly represses the BMPR2 translational efficiency in NPCs. Furthermore, we demonstrate that the BMPR2 expression level in the VZ/SVZ region increases as the cortex develops, and that this expression pattern has a negative correlation with the miR-17-5p expression pattern in the VZ/SVZ region. Based on these results, we propose that the miR-17 family members regulate the BMPR2 expression level in the VZ/SVZ region during cortex development.

The regulation of the BMP signaling pathway in the VZ/SVZ region is thought to influence the neurogenesis/astrocytogenesis transition during corticogenesis [11, 14]. Furthermore, the BMP antagonist noggin inhibits the gliogenic activity of the BMPs, and creates a neurogenic niche within the SVZ [40]. Therefore, we wished to determine whether miR-17-5p affects neurogenesis and astrocytogenesis during NPC differentiation. We found that miR-17-5p drastically reduced the phosphorylation of Smad1/5 by down-regulating the BMPR2 level in NPCs, which indicates that BMP signaling is suppressed. In addition, we show that miR-17-5p significantly increases neurogenesis at the expense of astrocytogenesis during NPC differentiation via regulating BMP signaling, which is similar to the effect of BMP inhibition [41]. We also demonstrate that miR-17-5p on its own has little impact on the neurogenesis/astrocytogenesis transition during NPC differentiation when functional BMP signaling is inhibited by noggin. Furthermore, down-regulation of miR-17-5p in the VZ/SVZ region is accompanied by the developmental transition from neurogenesis to astrocytogenesis in vivo during corticogenesis [42]. These findings provide strong evidence that miR-17-5p inhibits BMP signaling by repressing BMPR2, which further regulates the neurogenesis/astrocytogenesis transition in NPCs during corticogenesis.

A potential signaling pathway involved in the neurogenesis/astrocytogenesis transition is regulated by miR-17

It has been reported that activation of the transcription factor signal transducer and activator of transcription 3 (STAT3) is involved in astrocytogenesis as it directly increases the transcription of GFAP [43]. Smad1, activated by BMP2, cooperates with STAT3 by recruiting the transcriptional co-activator cAMP response element-binding protein (CREB)-binding protein (CBP) to the GFAP promoter [24]. Furthermore, miR-17 family members are believed to repress STAT3 expression during stem cell differentiation and in acute myeloid leukemia cells [44, 45]. Taken together, these findings suggest that miR-17 family members may regulate the neurogenesis/astrocytogenesis transition through multiple signaling pathways. To further evaluate the precise effects of the BMP signaling pathway on neural differentiation regulated by miR-17-5p, we specifically knocked down the BMPR2 expression level using siRNA and examined the differentiation of NPCs stimulated by BMP2. We found that BMPR2 inhibition reduces the phosphorylation of Smad1/5, which significantly increased the neuron/astrocyte ratio. These results suggest that inhibition of the BMP signaling pathway significantly contributes to the miR-17-mediated increase in neurogenesis.

In summary, our results demonstrate that miR-17 family members remain up-regulated during early development of the mouse cortex, especially in the VZ/SVZ region. The high expression levels of the miR-17 family members contribute to proliferation and neurogenesis of the NPCs during the early stages of corticogenesis. These regulatory effects are the result of regulating the BMP signaling pathway by targeting BMPR2.

Experimental procedures

Animals and cell culture

All of the animal care and experimental procedures were performed in accordance with the Laboratory Animal Care Guidelines approved by the Model Animal Research Center of Nanjing University. Neural precursor cells were prepared from E14.5 mouse telencephalons as previously described [24]. Freshly isolated cells were plated onto 10 cm dishes pre-coated with poly-l-ornithine (Sigma-Aldrich, St Louis, MO) and human Fibronectin (R&D Systems, Minneapolis, MN, USA), and cultured in N2 plus media supplement (R&D Systems) Dulbecco's modified Eagle medium: Nutrient mixture F-12 (Life Technologies, Grand Island, NY, USA) containing 10 ng/mL recombinant bovine fibroblast growth factor (bFGF) (R&D Systems) and recombinant human epidermal growth factor (hEGF) (R&D Systems) for 4 days (maintenance conditions). Cells were then dissociated and re-plated on 48-well plates (1 × 104 cells per well) or six-well plates (5 × 104 cells per well) pre-coated with poly-l-ornithine and human Fibronectin. The supplements bFGF and hEGF were included unless otherwise indicated. When the supplements bFGF and hEGF are removed, the cells are being cultured under differentiation conditions. BMP2 (10 ng/mL, ProsPec, East Brunswick, NJ, USA) and noggin (500 ng/mL, Prospec) are used as indicated. HEK 293T cells were maintained in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum (Invitrogen).

miR-17-5p mimics (50 nm, Genepharma, Shanghai, China) and miR-17-5p inhibitors (50 nm, Genepharma) were transfected into HEK 293T cells using Lipofectamine 2000 (Invitrogen). AntagomiR oligonucleotides (100 nm, RiboBio Inc., Guangzhou, China) directed against miR-17-5p and Chol-OMe-PS-modified BMPR2 siRNA duplexes (GGGAGCACGUGUUAUGGUCtt [46], 200 nm, RiboBio Inc.) were transfected into NPCs 48 h after re-plating without any transfection reagents. Lentiviruses over-expressing only miR-17-5p (lenti-17-5p) or EGFP and miR-17-5p (lenti-EGFP-17-5p) were generated by Genechem Inc. (Shanghai, China) and Genomeditech Inc. (Shanghai, China), respectively. NPCs were exposed to lentivirus (at a multiplicity of infection of 3) 48 h after re-seeding, and incubated overnight at 37 °C. The medium was completely refreshed 12 h later. Cells were harvested at the indicated times for subsequent assays.

RNA isolation, microRNA array and quantitative real-time PCR

Total RNA was isolated using Trizol reagent (Invitrogen) according to the manufacturer's instructions. Embryo and pups cortices at various developmental stages (E12.5, E15.5, E18.5 and P60) were dissected from C57BL/6J mice. Samples of 50 μg total RNA were extracted and sent to Capitalbio Corporation (Beijing, China) for miRNAs Affymetrix GeneChip® miRNA 2.0 array analysis (Santa Clara, CA, USA). Quantitative real-time PCR of mature miRNAs was performed using TaqMan microRNA probes (Applied Biosystems, Foster City, CA, USA) according to the manufacturer's instructions. Briefly, single-stranded cDNA was reverse transcribed using stem-loop reverse transcription primers (Ambion, Foster City, CA, USA) and reverse transcriptase XL (AMV) (Takara, Dalian, China). PCR was performed using a TaqMan PCR kit and an Applied Biosystems 7300 sequence detection system. All reactions were run in triplicate. The threshold cycle for each sample was chosen within the linear range and converted into a starting quantity by normalizing to internal standard U6 snRNA run on the same plate.

In situ hybridization

This protocol is a modified version of a previously described protocol [47]. Brains were fixed in 4% paraformaldehyde overnight at 4 °C, and cryoprotected in 15% sucrose overnight followed by 30% sucrose overnight. Cryostat sections (30 μm thick) were cut and mounted onto Superfrost Gold Plus slides (Fisher, Pittsburgh, PA). The sections were then post-fixed using 4% paraformaldehyde for 20 minutes at room temperature (RT) followed by washing in NaCl/Pi (3 × 10 min, RT) and once in 75% alcohol at RT for 10 min. Next, the sections were dried at RT and then incubated in hybridization buffer (diethypyrocarbonate-treated H2O with 50% v/v de-ionized formamide, 5 × SSC, 0.2 mg/mL yeast tRNA, 0.5 mg/mL herring sperm DNA, 10% w/v dextran sulfate, 0.1% Tween-20) containing 10 nm digoxigenin-labeled locked nucleic acid probe (Exiqon, Woburn, MA) overnight at 50 °C in a humidified chamber. After hybridization, the sections were washed in 0.1 × SSC at 55 °C (3 × 30 min), and then blocked in 10% fetal bovine serum in Tris-buffered saline (pH 7.5) for 1 h at RT. The sections were then incubated in alkaline phosphatase-conjugated antibody against digoxigenin (1:1000, Roche Diagnostics, Indianapolis, IN, USA) overnight. A 5-Bromo-4-chloro-3-indolyl phosphate/Nitro blue tetrazolium(BCIP/NBT) kit (Invitrogen) was used for the color reaction.

Luciferase assay

The BMPR2 3′ UTR containing the predicted target sequence was cloned and inserted into the pMIR-REPORT™ luciferase vector (Ambion, Austin, TX). The forward primer 5′-CTCGAGATGTTTTCAAGGTTATGGAG-3′ and reverse primer 5′-AAGCTTTCTGAATTTGCTTCTGTTTT-3′ were used. A mutated vector was generated by Invitrogen by replacing the predicted target region with its reverse sequence (from GCACT to CGTGA). HEK 293T cells were seeded onto 24-well plates for 12 h. Afterwards, 0.2 μg of firefly luciferase reporter plasmid, 0.2 μg of β-galactosidase (β-gal) expression vector (Ambion) and 50 nm of miR-17-5p mimics or scrambled control were transfected into the cells. Cells were harvested for the luciferase assay (Promega, Madison, WI, USA) 24 h later, and the luciferase activity was normalized to the β-gal activity.

EdU incorporation assay

NPCs were incubated in 50 μM EdU (RiboBio Inc.) for 16 h, and fixed with 4% paraformaldehyde for 30 min at RT. Next, the cells were washed in NaCl/Pi (2 × 5 min, RT) and then permeabilized using NaCl/Pi containing 0.3% Triton X-100 for 10 min. After extensive washes in NaCl/Pi, the cells were incubated in Apollo® staining solution (RiboBio Inc.) for 20 min, washed with NaCl/Pi (3 × 10 min, RT), and then incubated in 4′,6-diamidino-2-phenylindole (DAPI, 1:2500; Roche Diagnostics, Mannheim, Germany) for 10 min at RT.

Immunofluorescence

NPCs were briefly washed twice with cold NaCl/Pi, and then fixed in 4% paraformaldehyde for 10 min at RT. After fixation, the cells were washed with NaCl/Pi (3 × 5 min, RT), and then permeabilized and blocked using 2% BSA (Sigma-Aldrich) and 0.05% Triton X-100 in NaCl/Pi for 1 h at RT. Next, the cells were incubated with primary antibody [p-Histone H3 (1:200, Santa Cruz Biotechnology, Santa Cruz, CA), nestin (1:1000, Abcam, Hong Kong, China), GFAP (1:500, Abcam), MAP2 (1:2000, Abcam) or β-tubulin III (1:500, Abcam)] in 2% BSA/NaCl/Pi in a humidified chamber overnight at 4 °C, and then rinsed in NaCl/Pi (3 × 5 min, RT). The cells were then incubated in secondary fluorescent antibodies (Invitrogen) in 2% BSA/NaCl/Pi in a light-proof container for 1 h at RT. Finally, the cells were stained with DAPI and visualized using a fluorescent microscope. For quantitative analysis, the numbers of each type of cell scored in four random fields were averaged.

Immunohistochemistry

Sections were washed and blocked in NaCl/Pi containing 0.3% Triton X-100 and 5% BSA, and then incubated in primary antibody (BMPR2, 1:200; Abgent, San Diego, CA, USA) overnight at 4 °C. Next, the sections were washed in NaCl/Pi (3 × 15 min, RT), and then incubated in biotinylated secondary antibody (1:500, Vector Laboratories, Burlingame, CA) overnight at 4 °C. The signal was further amplified using the VECTASTAIN Elite ABC kit (Vector Laboratories). 3,3′-diaminobenzidine (Sigma-Aldrich) was used for the color reaction.

Western blot analysis

The tissues and cultured cells were lysed using RIPA buffer (Thermo Scientific, Rockford, IL, USA) and the protein was extracted according to the manufacturer's instructions. Protein samples were quantified using a BCA kit (Thermo Scientific, Rockford, IL). Samples of 50 μg protein were separated using 10% SDS/PAGE, and transferred onto poly(vinylidene difluoride) western blotting membranes (Roche Diagnostics). The membranes were incubated in primary antibody overnight at 4 °C. The following primary antibodies were used: anti-BMPR-II (1:400, BD Transduction Laboratories, San Jose, CA), anti-BMPR-II (1:1000, Cell Signaling Technology, Beverly, MA, USA), anti-P-Smad1/5 (1:2000, Cell Signaling Technology), anti-Smad1/5/8 (1:2000, Santa Cruz Biotechnology) and anti-GAPDH (1:2000, Abcam). Horseradish peroxidase-conjugated anti-mouse and anti-rabbit secondary antibodies (Santa Cruz Biotechnology) were used. The signal was detected using the SuperSignal enhanced chemiluminescence system (Pierce, Rockford, IL, USA).

Statistical analysis

All data are presented as means ± SEM. Differences were assessed for statistical significance at < 0.05 using Student's t test. Prism 5.0 software (GraphPad Inc., La Jolla, CA) was used for data analysis.

Acknowledgements

This work was supported by grants from the National Natural Science Foundation of China (31000478, 31100777 and 31271378).

Ancillary