SEARCH

SEARCH BY CITATION

Keywords:

  • Astrocytes;
  • Differentiation;
  • Neural stem cell;
  • Zac1;
  • Socs3;
  • Imprinting

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

Cell-fate decisions and differentiation of embryonic and adult neural stem cells (NSC) are tightly controlled by lineage-restricted and temporal factors that interact with cell-intrinsic programs and extracellular signals through multiple regulatory loops. Imprinted genes are important players in neurodevelopment and mental health although their molecular and cellular functions remain poorly understood. Here, we show that the paternally expressed transcriptional regulator Zac1 (zinc finger protein regulating apoptosis and cell cycle arrest) is transiently induced during astroglial and neuronal differentiation of embryonic and adult NSC lines. Thereby, Zac1 transactivates Socs3 (suppressor of cytokine signaling 3), a potent inhibitor of prodifferentiative Jak/Stat3 signaling, in a lineage-specific manner to prevent precocious astroglial differentiation. In vivo, Zac1 and Socs3 colocalize in the neocortical ventricular zone during incipient astrogliogenesis. Zac1 overexpression in primary NSCs delays astroglial differentiation whereas knockdown of Zac1 or Socs3 facilitates formation of astroglial cells. This negative feedback loop is unrelated to Zac1′s cell cycle arrest function and specific to the Jak/Stat3 pathway. Hence, reinstating Jak/Stat3 signaling in the presence of increased Zac1 expression allows for timely astroglial differentiation. Overall, we suggest that the imprinted gene Zac1 curtails astroglial differentiation of NSCs in the developing and adult brain. STEM Cells 2013;31:1621–1632


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

All neurons and macroglia (i.e., astrocytes and oligodendrocytes) in the developing central nervous system (CNS) are derived from neuroepithelial cells that give rise to radial glial cells [1]. These line the forebrain ventricles and spinal canal and represent precursors of neurons and glia at embryonic stages of development. They first produce neuronal cell types during the “neurogenic” phase followed by glial cell types during the later “gliogenic” phase. The transition from neuronal to glial subtype-specific precursors is tightly controlled, whereby cell-intrinsic programs and extracellular signals closely interact through multiple regulatory loops [2]. For example, the cytokine-induced Jak/Stat3 pathway is poorly active at early, neurogenic stages, when neurogenic factors are highly expressed. At later, gliogenic periods, decreased expression of neurogenic factors strengthened Jak/Stat3 signaling and promotes astrogliogenesis through a positive autoregulatory loop [3].

Adult mammalian neurogenesis occurs throughout life in restricted brain regions comprising the dentate gyrus of the hippocampus and from the subventricular zone (SVZ) of the lateral ventricle, the rostral migratory stream to the olfactory bulb and recapitulates many features of embryonic neurogenesis [1]. Astrocytes are important regulators of the local environment in adult neurogenic niches through the secretion of diffusible morphogenic factors, such as Wnts, which support adult neurogenesis [4]. Adding a new layer of epigenetic control, loss of imprinting of Dlk1 (δ-like homolog 1, an atypical NOTCH ligand) has been recently shown to restrain postnatal neural stem cell (NSC) proliferation from postnatal day 7 onward in NSCs and niche astrocytes [5].

Imprinted genes comprise a subset of mammalian genes that are subject to developmentally determined, parent-of-origin dependent, epigenetic modifications resulting in monoallelic expression. Although small in number, they are important to embryonic development, postnatal physiology and behavior [6]. Loss of their imprinting status frequently leads to severe metabolic and mental syndromes during prenatal and postnatal life [7].

The maternally imprinted Zac1 gene is transiently expressed in proliferating stem/progenitor cells of the telencephalic and cerebellar ventricular zones [8–11], the external granular cell layer [12], and the retina [13, 14], the function of which is, however, presently unknown.

Zac1 is a zinc finger protein conferring transcriptional activation and repression following monomer or dimer binding to GC-rich palindromic and repeat DNA elements [15–17]. Direct Zac1 target genes identified so far include the nuclear receptor PPARγ1 (peroxisome proliferator activated receptor) [18], the G-protein-coupled receptor PAC1 (pituitary adenylate activating receptor 1) [19–21], the exchange factor Rasgrf1 (RAS protein-specific guanine nucleotide-releasing factor 1) [22], and the cyclin-dependent kinase inhibitor p21Waf1/Cip1 [23]; all of which regulate apoptosis and cell cycle arrest. Here, we show that transient changes in Zac1 expression regulate astroglial differentiation by fine tuning Jak/Stat3 signaling in embryonic and adult NSCs through transcriptional regulation of Socs3.

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 Culture and Transfection Experiments

C17.2 [24], a mouse NSC line and SK-N-MC, a human neuroectodermal cell line (ATTC #HTB-10), were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum (FCS). Mouse embryonic NS-5 and adult O4ANS NSC lines were grown as reported [25, 26]. Tetracycline (Tc)-regulated Zac1 expression in C17.2 cells was established, and proliferation was measured as described [18, 22]. Primary cells from whole fetal brain (embryonic day 18; E18) of CD1 mice were dissected as described previously [27] and grown as suspension in DMEM/F12 and neurobasal medium (1:1) supplemented with N2 (1% v/v), B27 (2% v/v) (both Life Technologies GmbH, Darmstadt, Germany, http://www.lifetech.com), epidermal growth factor (EGF), and fibroblast growth factor (FGF) (both 10 ng/mL; PeproTech, Hamburg, Germany, Rocky Hill, NJ, http://www.peprotech.com). Neurospheres were dissociated with Accutase (Millipore, Schwalbach, Germany, http://www.millipore.com) and grown as monolayers on poly-d-lysine hydrobromide (Sigma, Munich, Germany, http://www.sigma-aldrich.com)-coated plates in the presence of EGF and FGF. For astroglial differentiation, cells were kept with 1% FCS or alternatively with 10 ng/mL bone morphogenetic protein 4 (BMP4) (R&D Systems, Wiesbaden, Germany, http://www.rndsystems.com) or 100 ng/mL ciliary neurotrophic factor (CNTF) (PeproTech) as described in the figure legends. Neuronal differentiation was initiated with 10 ng/mL FGF on Matrigel-coated dishes (0.4 μL/mL; BD Biosciences, Heidelberg, Germany, http://www.bdbiosciences.com). All media contained penicillin/streptomycin (Life Technologies GmbH).

Transient and stable transfections were performed using Lipofectamine (Life Technologies GmbH) or Turbofect transfection reagent (Fermentas, St. Leon-Roth, Germany) according to manufacturers' instructions using 1–5 × 105 cells per centimeter square. For stable Zac1 or Zac1 shRNA expression, O4ANS cells were transfected with the vectors pRK7.SV40-Pur-CMV.Flag-Zac1 or pLKO.1-Pur-Zac1.shRNA pool (Mission shRNA, clone NM_009538, Sigma), respectively, selected with puromycin (250 ng/mL, Merck KGaA, Darmstadt, Germany, http://www.merck.com), pooled, and amplified. Zac1 overexpressing O4ANS cells were additionally transfected with a Socs3 shRNA expression vector pLKO.1-CMV.Neo-Socs3 shRNA (Mission shRNA, clone NM_007707; Sigma), selected with gentamycin (300 μg/mL, Sigma) and processed as above. Scrambled RNA vector pLKO.1-Pur(Neo)-Non-Target shRNA (Mission shRNA, SHC016; Sigma) was used as a control in these experiments. Master stocks were prepared and cultivated on demand as described previously [22]. Expression of the transgenes was verified by immunoblotting and quantitative reverse transcriptase polymerase chain reaction (PCR). Luciferase reporter activities were normalized on β-galactosidase activity of a cotransfected expression vector [28]. Amounts of transfected plasmids are indicated in the corresponding figure legends.

Primary NSCs of dissociated neurospheres from E18 brain were transfected with 1 μg of expression vectors pRK7-eGFP, pRK7-Zac1-eGFP, pLKO.1-Pur-Zac1.shRNA pool, pLKO.1-CMV.Neo-Socs3 shRNA, and pLKO.1-Pur(Neo)-Non-Target shRNA as described in the respective figure legend. Following transfection, cells were allowed to recover for 12 hours before astroglial differentiation was initiated for 1 day. Images of 100 green fluorescent protein (GFP) positive cells per transfection corresponding to a total number of 3,000–5,000 cells were scored for Gfap immunoreactivity by an independent investigator.

Fluorescence-Activated Cell Sorting Analysis

Cells (2 × 106) were stained with propidium iodide as described previously [21] and analyzed on a Beckman Coulter Epics XL using Expo32 ADC analysis.

Plasmids

The mouse Socs3 promoter spanning 3.8 kb of sequence upstream of the translational start site [29] was digested by SacI and StuI to generate the construct (−1,267/+929). A SacII digest, followed by blunt religation, was used to create the construct (−451/+929) and a BglI digest to obtain the construct (−141/+929). The constructs (+2/+929) and (+286/+929) were generated by PCR using primer pairs with internal MluI and XhoI sites. All promoter fragments were cloned in the pGL3 basic vector (Promega, Mannheim, Germany, http://www.promega.com) and verified by sequencing. The human SOCS3 promoter spanned 1.7 kb of sequence upstream of the translational start site in the vector pGL3 [30]. Zac1 expression constructs are described elsewhere [16].

To obtain pRK7-eGFP, the coding sequence of pEGFP-C1 (Clontech, Heidelberg, Germany, http://www.clontech.com) was amplified with PCR primers containing terminal BamHI and EcoRI restriction sites, cloned into the expression vector pRK7, and sequence verified. Thereafter, the Zac1 cDNA was inserted in-frame upstream of eGFP at the BamHI site to generate pRK7.Zac1-eGFP.

Chromatin Immunoprecipitation, Methylated DNA Immunoprecipitation, RNA Extraction, and PCR Experiments

Chromatin immunoprecipitation (ChIP) assays were performed as described [31] with Zac1-LPR antiserum. Sequential ChIP assays were done as described previously [22] by precipitation with the antibodies acH3 (06-599, Millipore) or H3K9me2 (pAb060050, Diagenode, Liege, Belgium, http://www.diagenode.com) followed by precipitation with Zac1-LPR antiserum. The ChIP against activated polymerase II was done with anti-c-terminal domain phospho serine-5 (ab5131, Abcam, Cambridge, U.K., http://www.abcam.com).

The methylated DNA immunoprecipitation (MeDIP) assay (55009, Active Motif, La Hulpe, Belgium, http://www.activemotif.com) was performed according to manufacturers' instructions. Data are diagrammed as percent of input normalized to control sera (IgG or preimmune-serum). RNA extraction and PCR experiments were performed as described [32]. Primers for PCR and ChIP experiments are listed in Supporting Information Table S1. The housekeeping genes Gapdh (glycerinaldehyde-3-phosphat-dehydrogenase), Atp5j (ATP synthase-coupling factor 6), or β-actin served for normalization.

Immunoblotting, Immunocytochemistry, and Immunohistochemistry

Whole cell extracts (20–70 μg) were fractionated by SDS-PAGE gel electrophoresis and tested with the following antibodies: Socs3 (ab16030, Abcam), Stat3 (sc-482, Santa Cruz, Heidelberg, Germany, http://www.scbt.com), phospho-Stat3-Y705 (sc-7993, Santa Cruz), Actin (sc-8432, Santa Cruz), Flag (F3165, Sigma), mouse and human Zac1 [18, 32].

For immunocytochemistry, cells were grown on 20 μg/mL poly-d-lysin hydrobromide-coated coverslips, fixed with 5% formaldehyde, permeabilized with 0.1% saponin (Sigma) in phosphate buffered saline (PBS), blocked with donkey normal serum (Sigma), and subjected to indirect immunofluorescence with antibodies to A2B5 (ab53521, Abcam), Gfap (Z0334, Dako, Hamburg, Germany, http://www.dako.com), Socs3 (ab16030, Abcam), and Zac1. Nuclei were stained with 2-(4-amidinophenyl)-6-indolecarbamidine dihydrochloride (DAPI).

Brains of CD1 mice at E16, E18, and postnatal day 1 (P1) were dissected and immediately cryofrozen by isopentan (Sigma) on dry-ice. Immunohistochemistry was done on 30-μm thick sections, postfixed by 4% paraformaldehyde, permeabilized with 0.1% saponin in PBS, blocked with donkey normal serum, and subjected to indirect immunofluorescence with antibodies to Gfap, Socs3, and Zac1. Nuclei were stained with DAPI.

Microscopic analysis was done with fluorescence microscope (BX61, Olympus, Hamburg, Germany, http://www.olympus.de) and CCD camera (E-620 SRL, Olympus) or confocal microscope (FluoView 1000, Olympus). All images were captured using identical laser power and gain settings.

Software and Statistical Analysis

Computational analysis of the Socs3 gene was done with Genomatix MatInspector. Results represent the means and SDs from at least five independent experiments. Numerical data were analyzed by unpaired Student's t test and ANOVA with post hoc Tukey test. The threshold for significance was set at *, p < .05 and **, p < .01.

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

Zac1 Induces Socs3 Expression

We previously generated a panel of inducible Zac1 clones using a Tet-off system [33] in hippocampal progenitor (HW3.5) and cerebellar stem (C17.2) cells. By cultivation in the absence or presence of tetracycline, corresponding to ectopic Zac1 expression being switched on and off, respectively, we screened 30 clones in either case for maximal Zac1 induction by immunoblot analysis (data not shown) [18]. Following on, we carried out a comparative, genome-wide expression analysis by cDNA microarrays for two representative Zac1 clones from each cell line cultivated for 3, 6, and 9 hours with or without tetracycline. Only genes exhibiting a >1.2-fold difference between tetracycline-positive and tetracycline-negative conditions in both clones were taken into account. Microarray results from C17.2 cells showed enhanced Socs3 expression at the time of early and intermediary (1.5-fold and 1.3-fold increases at 3 and 6 hours, respectively) but not of late (9 hours) Zac1 induction, whereas Socs3 expression was unaffected in HW3.5 cells at any time points (data not shown).

C17.2 cells showed in the presence of tetracycline (+Tc) Zac1 mRNA expression levels similar to parental cells. In contrast, following 6 hours of tetracycline removal (−Tc), Zac1 mRNA expression rapidly increased (Fig. 1A). Time course analysis evidenced increased expression of Zac1 and Socs3 mRNAs and proteins at early time points which gradually declined thereafter (Fig. 1B, 1C). In accord with previous reports [15, 17, 18, 21, 34], Zac1 expression inhibited cell proliferation (Fig. 1D). Hereby, Zac1 led to G1 arrest due to upregulation of p21Waf1/Cip1 and concomitant apoptotic cell death (data not shown).

thumbnail image

Figure 1. Zac1 induces Socs3 expression. (A): C17.2 neural stem cells harboring regulated Zac1 expression were grown for 6 hours in the presence (+Tc) or absence (−Tc) of tetracycline. Expression of Zac1 and the house keeping gene Gapdh were detected by reverse transcriptase polymerase chain reaction (RT-PCR) and compared to parent C17.2 cells. (B): Time course of Zac1 and Socs3 expression following Tc removal for the indicated periods as measured by quantitative RT-PCR. (C): Immunoblot analysis (70 μg whole cell extract) for Zac1, Socs3, and β-actin expression at indicated periods of Tc removal. (D): Proliferation of C17.2 cells harboring regulated Zac1 in the presence or absence of Tc for the indicated number of days. Representative results from three (A, C) or means with SDs (±SDs) from six independent (B, D) experiments are shown. *, p < .05; **, p < .01. Abbreviations: Gapdh, glycerinaldehyde-3-phosphat-dehydrogenase; Socs3, suppressor of cytokine signaling 3; Tc, tetracycline; Zac1, zinc finger protein regulating apoptosis and cell cycle arrest.

Download figure to PowerPoint

Zac1 Transactivates the Socs3 Gene

We examined the Socs3 promoter for potential Zac1 binding sites by computational analysis. Zac1 recognizes GC-rich palindromic and repeat DNA elements [16] both of which are present in multiple copies at upstream, exonic, and intronic regions of the Socs3 gene schematically depicted in Figure 2A. To corroborate transcriptional activation of Socs3 by Zac1, we transfected a Socs3 promoter reporter plasmid - encompassing the region from −1,267 bp to +929 bp - and different Zac1 constructs into C17.2 cells. Reporter activities (Fig. 2B) were adjusted for expression differences between Zac1 constructs (Fig. 2D), whereby Zac1 expression was set to 1. Accordingly, wild-type Zac1 conferred a sixfold induction, whereas mutations in the DNA binding domain (broken zinc finger 6 or 7) reduced activation by more than half [16]. Mutations or deletions preferentially disruptive to direct repeat binding (ZF2mt and Δ1-5, devoid of ZF 1-5) reduced transactivation by one third compatible with the idea that Zac1 confers transactivation through either class of DNA element (Fig. 2B). Moreover, Zac1 devoid of the central transactivation domain (ΔLPR) did not stimulate Socs3 reporter activity indicating that regulation is unrelated to Zac1 coactivation [32]. Likewise, transfected Zac1, but not ΔLPR, induced Socs3 gene and protein expression in C17.2 cells (Fig. 2C, 2E, data not shown).

thumbnail image

Figure 2. Zac1 transactivates Socs3. (A): Scheme of the Socs3 5′-end including the promoter, exons (filled boxes), and the first intron. The signal flag symbolizes the transcriptional start site. Predicted Zac1 DNA-binding sites are depicted by filled bars and circles. Chromatin immunoprecipitation (ChIP) primer pairs flanking amplified regions are shown by filled triangles and roman numbers, respectively. (B): Reporter assays in C17.2 cells cotransfected with wild-type or mutated Zac1 constructs (100 ng each) and the 2 kb Socs3 promoter (200 ng). (C): Increasing doses of Zac1, but not of ΔLPR (10 ng, 50 ng and 100 ng each) enhanced Socs3 expression as measured by quantitative reverse transcriptase polymerase chain reaction. (D): Immunoblot (WCE 50 μg) of the different Zac1 constructs (100 ng each) visualized by an antibody against the Zac1 C-terminus. (E): Immunoblot (whole cell extract 50 μg) analysis of Socs3 and Zac1 expression following transfection of increasing amounts of Zac1 (50 ng, 100 ng, and 250 ng). (F): Reporter assays in C17.2 cells cotransfected with Zac1 (100 ng) and the indicated Socs3 promoter constructs (200 ng each). (G): ChIP assays in C17.2 cells with an antibody against Zac1 showed binding at the Socs3 promoter (II), exon 1 (III), and intron (IV). The upstream region (I) without binding sites served as negative control. (H): ChIP of transfected Flag-tagged Zac1 or ZF7mt (100 ng each) with a Flag antibody. Zac1, but barely ZF7mt, bound to the Socs3 gene as above. Representative results from three (D, E) or means ± SD from six to eight independent (B, C, F–H) experiments are shown. *, p < .05; **, p < .01. Abbreviations: Δ1-5 or ΔLPR, Zac1 without zinc fingers 1-5 or without the central transactivation domain; Socs3, suppressor of cytokine signaling 3; zac1, zinc finger protein regulating apoptosis and cell cycle arrest; ZF7mt, Zac1 containing broken zinc fingers 7; ZF6mt, Zac1 containing broken zinc fingers 6; ZF2mt, Zac1 containing broken zinc fingers 2.

Download figure to PowerPoint

In view of multiple potential Zac1 binding sites throughout the Socs3 upstream, exonic and intronic region, we used a series of progressive 5′-terminal truncations to identify actual regulatory sites. Absence of the distal (−451/+929 bp), proximal (−141/+929 bp), and core (+2/+929 bp) promoter regions stepwise decreased Zac1-dependent reporter activities (Fig. 2F). Zac1-dependent regulation was unaffected by deletion of exon 1 whereas deletion of the intronic region abrogated transactivation. Consistent with this finding, the remaining part of exon 2 and the parent vector are both devoid of Zac1 DNA elements (Fig. 2F). Together, these results suggest that Zac1 transactivates the Socs3 gene through the promoter and the first intron in an additive manner.

The SOCS3 gene is conserved between human and mice [35] including multiple potential ZAC1 DNA-binding sites (Supporting Information Fig. S1A). We tested regulation by ZAC1 in the human neuroectodermal cell line SK-N-MC, in which both ZAC1 and SOCS3 proteins are well expressed (Supporting Information Fig. S1D). In agreement with the results from mice, increasing doses of ZAC1 conferred transactivation to a SOCS3 promoter reporter plasmid and induced SOCS3 mRNA and protein levels (Supporting Information Fig. S1B–S1D).

Zac1 binding was corroborated by ChIP assays with a Zac1 antibody and primer pairs flanking each of the potential binding regions at the Socs3 gene labeled II–IV (Fig. 2A). Another primer pair (I) spanning an upstream region devoid of Zac1 binding motifs served as a negative control in these experiments. We detected Zac1 binding to slightly different degrees at regions II–IV (Fig. 2G), including exon 1. To assess Zac1 binding during transactivation, we transfected a Flag-tagged version of Zac1 which occupied indistinguishably to naive Zac1 the Socs3 gene. In contrast, a Zac1 protein defective in DNA binding (ZF7mt) was barely bound (Fig. 2H).

A similar pattern was observed for ZAC1 which preferentially occupied the human SOCS3 promoter and intron in SK-N-MC cells. Following ZAC1 overexpression, additional binding occurred at the exonic region (Supporting Information Fig. S1E).

These data suggest that Socs3 is a direct target of Zac1 in human and mice, whereby Zac1 binds at multiple palindromic and repeat DNA elements throughout the 5′-end of the Socs3 gene to confer transactivation.

Zac1 Induces Socs3 During Astroglial Differentiation

C17.2 cells, but not other NSC lines used in this study, allowed stable integration of tetracycline-regulated Zac1 expression; however, C17.2 cells poorly differentiate in vitro, possibly due to transformation by avian myc oncogene [24]. Embryonic and adult neurogliogenesis and astrogliogenesis vary in terms of spatio-temporal and quantitative criteria although many fundamental processes appear well conserved. To investigate whether Zac1-mediated Socs3 induction is shared at different developmental stages, we chose the nontransformed neural stem cell lines NS-5 and O4ANS for further experiments, which are derived from mouse embryonic and adult stem cells, respectively [25, 26]. These cells show in the undifferentiated state a radial glia-like phenotype and can be efficiently differentiated into either neurons or astrocytes in vitro >95% neuronal or astroglial cell types, data not shown) [36].

Zac1-dependent transactivation of the parent Socs3 promoter construct was preserved in either cell line (Fig. 3A; Supporting Information Fig. S2A). With the onset of lineage-directed differentiation, expression of the NSC-related marker Nestin rapidly declined in parallel to increased expression of the neuronal and astroglial markers Tuj1 and Gfap, respectively (Fig. 3B; Supporting Information Fig. S2B). Interestingly, upregulation of Socs3 mRNA and protein was confined to the astroglial lineage where Zac1 mRNA and protein expression peaked at day 4 and diminished thereafter (Fig. 3C, 3D; Supporting Information Fig. S2C, S2D). On the other hand, sustained Zac1 upregulation during neuronal differentiation did not induce Socs3 expression.

thumbnail image

Figure 3. Zac1 induces Socs3 during astroglial differentiation. (A): Promoter reporter assays. Cotransfection of increasing amounts of Zac1 (25 ng, 50 ng, 100 ng, and 250 ng) enhanced Socs3 promoter (200 ng) activity in O4ANS cells. (B): Astroglial and neuronal differentiation of the adult NSC line O4ANS. Expression of Nestin, Gfap, and Tuj1 was monitored by reverse transcriptase polymerase chain reaction (RT-PCR). (C): Time course analysis of Zac1 and Socs3 mRNA expression during astroglial and neuronal cell differentiation as measured by quantitative RT-PCR. (D): Immunoblot analysis (whole cell extract 70 μg) revealed that Socs3 induction time shifted to Zac1 during astroglial differentiation, while Zac1 expression during neuronal differentiation did not enhance Socs3 expression. (E): Chromatin immunoprecipitation (ChIP) analysis during astroglial or neuronal differentiation evidenced Zac1 occupancy at the Socs3 promoter, exon 1, and intron solely during the former. (F, G): ChIP assays during astroglial differentiation at day 4 with antibodies against acH3 and H3K9me2. (H, I): Sequential ChIP analysis. During astroglial differentiation, Zac1 associated with active chromatin marks (acH3/Zac1) at the promoter and intron of Socs3, whereas association with repressive chromatin marks (H3K9me2/Zac1) decreased in parallel. Representative results from three (B, D) or means ± SD from six to eight independent (A, C, E–I) experiments are shown. *, p < .05; **, p < .01. Abbreviations: NSC, neural stem cell; Socs3, suppressor of cytokine signaling 3; Zac1, zinc finger protein regulating apoptosis and cell cycle arrest.

Download figure to PowerPoint

Socs3's role as a lineage-specific target gene was corroborated by ChIP assays which evidenced that Zac1 occupies promoter, exonic, and intronic regions to varying degrees during astroglial differentiation (Fig. 3E; Supporting Information Fig. S2E). In contrast, Zac1 poorly bound the Socs3 gene during neuronal differentiation although it localized similarly to the nuclei of astroglial and neuronal cells (data not shown).

To elucidate whether exonic Zac1 binding contributes to transactivation under astroglial differentiation, we carried out sequential ChIP experiments with antiserum against active (pan-acetyl histone 3, acH3) or repressive (dimethyl lysine 9 of histone 3; H3K9me2) histone marks followed by immunoprecipitation with Zac1 antiserum. Astroglial differentiation led to an increase in acH3 across the 5′-end of the Socs3 gene concomitant to a decrease in H3K9me2 (Fig. 3F, 3G). Zac1 and the active marks preferentially associated in the same chromatin fraction at the promoter and intronic, but not at the exonic region, suggesting that transactivation occurs at the former (Fig. 3H, 3I). In contrast, active and repressive histone marks were unchanged following neuronal differentiation (Supporting Information Fig. S3A–S3D). Consistent with these results, a ChIP assay against the active (serine 5-phosphorylated) form of polymerase II evidenced high binding following astroglial but not neuronal differentiation at the Socs3 upstream region (Supporting Information Fig. S4). Moreover, a MeDIP scan of the overlapping CpG island evidenced reduced DNA methylation at different Zac1 binding sites preferentially during astroglial differentiation (Supporting Information Fig. S5B). Accordingly, overall methylation was decreased by 64% and 23% under astroglial and neuronal differentiation (Supporting Information Fig. S5C). Together, these results suggest that sustained DNA methylation confers Socs3 silencing during neuronal differentiation and prevents Zac1 binding and associated histone marks. Conversely, Socs3 demethylation during astroglial differentiation allows for Zac1 DNA binding and subsequent transactivation.

Zac1 and Socs3 Colocalize in the Neocortical Ventricular Zone

Gliogenesis generally follows neurogenesis in the developing mammalian brain, with the same progenitor domains switching developmental programs from neuron production mainly to oligodendrocyte or astrocyte production. Progenitors in the subventricular zone from late embryonic stages to early postnatal stages (E17–P14) generate predominantly glial lineages [1]. Therefore, we investigated Zac1 and Socs3 expression in the neocortical ventricular zone at E16 (prior to astrogliogenesis), E18 (concurrent to the onset of astrogliogenesis), and at P1 (advanced astrogliogenesis). Confocal microscopy visualized Zac1 expression in the subventricular and ventricular zone at E16 in the absence of Socs3. In contrast, nuclear Zac1 and cytoplasmatic Socs3 immunoreactivities strongly colocalized at E18 in the ventricular zone (Fig. 4A, 4B). Thereafter, both Zac1 and Socs3 expression declined and poorly colocalized in the neocortical ventricular zone. Moreover, consistent with a role of Zac1 in early astrogliogenesis, only few cells were Gfap positive (a marker of astrocytes) and did not colocalize with Zac1 (Fig. 4C). Together, these results evidence Zac1 and Socs3 coexpression in the neocortical ventricular zone during incipient astrogliogenesis.

thumbnail image

Figure 4. Zac1 and Socs3 colocalize in the neocortical ventricular zone. (A): Mouse brains from E16, E18, and P1 were stained with antibodies for Zac1 (red) and Socs3 (green). Representative images of the subventricular and ventricular zones are shown. Scale bar = 100 μm. (B): Within the neocortical ventricular zone Zac1 and Socs3 double positive cells appeared at E18. Insets show boxed regions at higher magnification and confirm colocalization of Zac1 and Socs3 (C): Gfap positive cells in the neocortical ventricular zone at E18 did not colocalize with Zac1. Scale bar = 20 μm. Nuclei were stained with 2-(4-amidinophenyl)-6-indolecarbamidine dihydrochloride (blue). Representative results from three independent experiments (A–C) are shown. Abbreviations: Socs3, suppressor of cytokine signaling 3; Zac1, zinc finger protein regulating apoptosis and cell cycle arrest.

Download figure to PowerPoint

Zac1 Induces Socs3 in Differentiating Astrocytes During Transition to Maturation

Stat3-dependent activation of target genes plays an important role in the formation and maturation of astroglial cells [37]. This process can be counterbalanced by Socs3 which potently inhibits Jak/Stat3 signaling by binding to transmembrane tyrosine kinase receptors following dimerization [38]. Consistent with this concept, tyrosine phosphorylation (Y705) of Stat3 was strongly reduced following Zac1 transfection of C17.2 cells or during astroglial differentiation of O4ANS stem cells (Supporting Information Fig. S6A, S6B). This led us to investigate whether Zac1-dependent Socs3 expression constitutes a negative feedback loop to restrain astroglial differentiation.

The A2B5 antibody recognizes the c-series gangliosides [39] which are expressed during early stages of astroglial differentiation as a cell adhesion molecule [40]. Following 24 hours of differentiation, all O4ANS cells were A2B5 positive and mostly Gfap, Zac1, or Socs3 negative (Fig. 5A, a, f, a', f'). The subsequent course of differentiation was characterized by a steady decrease in A2B5 immunoreactivity and an inverted increase in Gfap immunoreactivity (Fig. 5A, a–j), implying the transition of astroglial precursors into a more mature state.

thumbnail image

Figure 5. Zac1 induces Socs3 in differentiating astroglial cells during the transition to maturation. (A): Astroglial differentiation of O4ANS cells. At different days cells were subjected to indirect immunofluorescence with antibodies against A2B5 (a–e) and Gfap (f–j), merge (k–o) or Zac1 (a'–e') and Socs3 (f'–j'), merge (k'–o'). Nuclei were stained with DAPI (blue). Scale bar = 40 μm. (B): Quantitative analysis of glial marker A2B5 and Gfap positive cells are shown as percentage of the total number of DAPI positive cells. Mean of five analyzed images per day of the type shown in (A, a–o). (C): Quantitative analysis of Zac1 and Socs3 positive cells are shown as percentage of the total number of DAPI positive cells. Mean of five analyzed images per day of the type shown in (A, a'–o'). Abbreviations: DAPI, 2-(4-amidinophenyl)-6-indolecarbamidine dihydrochloride; Socs3, suppressor of cytokine signaling 3; Zac1, zinc finger protein regulating apoptosis and cell cycle arrest.

Download figure to PowerPoint

In transition to maturation, A2B5/Gfap double positive cells appeared (Fig. 5A, m) and quantitative analysis revealed 69% colocalization around day 4 (Fig. 5B). Interestingly, Zac1 and Socs3 expression peaked at day 4 suggesting a function in early maturation of astroglial precursors (Fig. 5A, m') in agreement with the findings from the necortical ventricular zone at E18 (Fig. 4A, 4B). While Zac1 staining in Zac1/Socs3 double positive cells was confined to the nucleus, we detected appreciable Socs3 staining in both the cytoplasm and nucleus as described previously in cells undergoing acute or chronic stimulation by cytokines in vivo [41, 42]. Zac1/Socs3 double positive cells peaked at day 4 (85%) and declined thereafter progressively (day 8, double positive 46%; day 12, double positive 17%) (Fig. 5C). These results conform to gene expression data (Fig. 3C) showing transient changes in Zac1 and Socs3 mRNA levels. We performed additional colocalization studies to elucidate whether early transient Zac1-mediated Socs3 expression associated preferentially with one of the astroglial markers A2B5 and Gfap (Supporting Information Fig. S7). A2B5 and Zac1 or Socs3 showed intense costaining at day 4, where also Gfap and Zac1 or Socs3 were for the most part colocalized. As a result, Zac1 and Socs3 were mainly coexpressed in A2B5/Gfap double positive cells. Similar results were measured in NS-5 cells (Supporting Information Fig. S8) and demonstrated that Zac1-mediated expression of Socs3 is temporarily confined to A2B5/Gfap double positive cells appearing during early maturation of astroglial precursors.

Zac1 Fine Tunes Differentiation of Astroglial Cells

To address Zac1′s function during astroglial differentiation, we stably transfected O4ANS cells with either Flag-tagged Zac1 or Zac1 shRNA expression vectors. Recombinant pools showed a doubling or a reduction by 80% of Zac1 mRNA levels with corresponding changes in protein expression (Fig. 6A, 6B). Expression of exogenous Zac1 was further assessed by immunoblot with the Flag antibody or PCR analysis for the expression of the selection marker puromycin-N-acetyltransferase present in the backbone of either vector (Fig. 6B).

thumbnail image

Figure 6. Zac1 fine-tunes differentiation of astroglial cells. (A): Zac1 expression following stable Zac1 (Flag-tagged) or Zac1 shRNA overexpression in O4ANS cells as evidenced by quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) whereby parental expression was set to 100%. (B): Immunoblot (whole cell extract 70 μg) with Zac1 and Flag antibodies shows expression of Flag-Zac1. Integration of Zac1 shRNA plasmids was evidenced by RT-PCR for the selection marker puromycin. (C): Analysis of Zac1 and Socs3 expression during astroglial differentiation of parental, Zac1, or Zac1 shRNA overexpressing cells as measured by qRT-PCR. (D): Distribution of cell cycle phases of parental, Zac1 or Zac1 shRNA overexpressing cells during astroglial differentiation as measured by FACS analysis. (E): Immunocytochemistry of parental, Zac1, or Zac1 shRNA overexpressing O4ANS cells during astroglial differentiation. Cells were subjected to indirect immunofluorescence with antibodies against A2B5 (red) and Gfap (green). Nuclei were stained with DAPI (blue). Scale bar = 40 μm. (F): Astroglial differentiation of O4ANS cells by serum (FCS) or following treatment with CNTF or BMP4 in serum free medium. Zac1 and Socs3 expression were measured by qRT-PCR at day 4. (G): Light microscopy of O4ANS cells differentiated with CNTF or BMP4. (H): Double staining of astroglial markers A2B5 (red) and Gfap (green) in parent, Zac1, or Zac1 shRNA overexpressing O4ANS cells differentiated with CNTF or BMP4 for 4 days. Scale bar = 40 μm. Representative results from 3 to 5 (B, E, G, H) or means ± SD from 6 to 10 independent (A, C, D, F) experiments are shown. *, p < .05; **, p < .01. Abbreviations: BMP4, bone morphogenetic protein 4; CNTF, ciliary neurotrophic factor; FCS, fetal calf serum; FACS, fluorescence-activated cell sorting; shRNA, small hairpin RNA; Socs3, suppressor of cytokine signaling 3; Zac1, zinc finger protein regulating apoptosis and cell cycle arrest.

Download figure to PowerPoint

As expected, Zac1 overexpression or knockdown led to concordant changes in Socs3 mRNA expression in proliferating and differentiating O4ANS cells consistent with the role of Socs3 as lineage-specific Zac1 target gene (Fig. 6C).

In accord with previous reports [15, 17, 21, 34] Zac1 overexpression caused an increase in the number of cells in G1-phase with concomitant decreases in S- and G2/M-phase during early time points (Fig. 6D). In contrast, Zac1 knockdown reduced the number of cells in G1-phase and led to concomitant increases in S- and G2/M-phase preferentially during late time points. Because cell-cycle arrest in G1-phase can potentiate differentiation [43], we assessed next the effects of altered Zac1 expression on astroglial differentiation.

Interestingly, Zac1 overexpression postponed the appearance of A2B5 and Gfap (Fig. 6E and data not shown). In opposite, Zac1 knockdown cells adopted an astroglial phenotype clearly ahead of time at day 4. These findings suggest that Zac1′s effects on astroglial differentiation are unrelated to the regulation of G1-phase arrest but reflect the induction of Socs3.

To test this concept, we derived stable Socs3 knockdown populations from Zac1 overexpressing cells. These cells showed similar levels of Socs3 mRNA expression during proliferation and astroglial differentiation when compared to parent cells (Supporting Information Fig. S9A, S9C). Thereby, reduced Socs3 expression reinstated timely astroglial differentiation in the presence of sustained G1 arrest (Supporting Information Fig. S9D, S9E).

To assess whether Zac1 controls additional pathways in astroglial differentiation, we studied differentiation in serum free medium following treatment with CNTF or BMP4 to activate either the Jak/Stat3 or Smad pathways [44]. CNTF treatment doubled Zac1 and Socs3 expression compared to serum conditions, whereas BMP4 showed the opposite effect (Fig. 6F). Zac1 overexpressing or knockdown cells maintained with CNTF differentiated into elongated glial cells showing sparse Gfap expression (Fig. 6G, 6H), which probably reflects the absence of extracellular matrix components in the differentiation protocol [45]. In contrast, following treatment with BMP4, astroglial cells displayed a star-like morphology with robust Gfap expression (Fig. 6G, 6H) [46]. Interestingly, astroglial differentiation was postponed under CNTF, but not under BMP4 treatment, on day 4 in Zac1 overexpressing cells. In contrast, knockdown of Zac1 affected neither CNTF nor BMP4 dependent astroglial differentiation when compared to the parental population (Fig. 6H).

These results support that Zac1 controls Jak/Stat3 signaling through Socs3 induction to fine tune astroglial differentiation, whereas unrelated pathways, such as BMP4, were unaffected by differences in Zac1 expression.

Zac1 Induces Socs3 During Astroglial Differentiation of Primary NSC

Long-term maintenance of NSCs can alter aspects of cellular identity and potential [25, 26]. Therefore, we investigated the lineage-specific role of Zac1-dependent Socs3 induction additionally in primary cells of embryonic mice brain from day 18 (E18). Cells dissected at different developmental periods differentiate into progeny that reflect the developmental capacity occurring at the time of dissection, that is, cells from early development differentiate into neurons, whereas those from late prenatal and early postnatal stages differentiate into glia [47].

The Socs3 promoter construct was robustly transactivated by Zac1 in E18 cells (data not shown); moreover, differentiation of neurospheres into astroglial and neuronal cells (Fig. 7A) induced Zac1 expression whereby upregulation of Socs3 was confined to the astroglial lineage (Fig. 7B). Zac1 and Socs3 mRNA levels peaked around day 3 to decline thereafter and matched changes in protein expression during differentiation (Fig. 7B, 7C). As expected, ChIP analysis confirmed high Zac1 occupancy at the Socs3 gene solely during early astroglial differentiation (data not shown).

thumbnail image

Figure 7. Zac1 induces Socs3 in primary differentiating astrocytes. (A): Nestin, Gfap, and Tuj1 expression during astroglial and neuronal differentiation of E18 primary neural stem cells (NSCs) as evidenced by reverse transcriptase polymerase chain reaction (RT-PCR). (B, C): Zac1 and Socs3 expression during astroglial and neuronal differentiation as evidenced by quantitative RT-PCR (B) and immunoblot (C). (D): Following 3 days of astroglial differentiation, primary E18 cells were subjected to staining with antibodies A2B5, Gfap, Zac1, and Socs3. Nuclei were stained with DAPI (blue). Scale bar = 40 μm (rows 1–5) or 100 μm (row 6). (E): Green fluorescent protein (GFP) imaging and Gfap immunocytochemistry of cotransfected primary E18 NSCs following 1 day of astroglial differentiation. Expression vectors harbored (a) Zac1-eGFP fusion protein, (b) Zac1-eGFP and shRNA Socs3 (ratio 1:2), (c) shRNA Zac1 and eGFP (ratio 2:1), (d) shRNA Socs3 and eGFP (ratio 2:1). (F): Quantitative analysis of Gfap positive cells is shown as percentage of the total number of GFP positive cells. Representative results from three to five (A, C–E), means (F) or means ± SD (B) from five independent experiments are shown. *, p < .05; **, p < .01. Abbreviations: DAPI, 2-(4-amidinophenyl)-6-indolecarbamidine dihydrochloride; GFP, green fluorescent protein; Socs3, suppressor of cytokine signaling 3; Zac1, zinc finger protein regulating apoptosis and cell cycle arrest.

Download figure to PowerPoint

Double immunostaining revealed that Zac1 and Socs3 colocalized in A2B5 positive astroglial cells following 3 days of differentiation (Fig. 7D) in agreement with the results from above stem cell lines. In contrast, Zac1, Socs3, and A2B5 weakly localized with Gfap immunoreactivity and were undetectable on day 6, when 80% of the cells had become Gfap positive (Fig. 7D and data not shown).

To corroborate Zac1′s role in Socs3 induction and astroglial differentiation, we carried out cotransfection experiments in primary E18 NSCs. Following 1 day of astroglial differentiation, one third of the cells showed moderate to high Gfap immunoreactivity as was the case for GFP positive cells following transfection of an eGFP expression vector (Fig. 7F and data not shown). In contrast, transfection of a Zac1-eGFP expression vector reduced the number of GFP/Gfap double-positive cells (Fig. 7E, 7F) and led to an increase in Socs3 expression (data not shown). This effect was reversed following simultaneous knockdown of Socs3. Moreover, knockdown of either Zac1 or Socs3 advanced astroglial differentiation whereas scrambled shRNA was without effect (Fig. 7E, 7F). Together, these results show that Zac1 restrains early astroglial differentiation of E18 NSCs via Socs3 induction.

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

In this study, we have shown that Zac1 is transiently induced in differentiating NSCs and transactivates in a lineage-specific manner Socs3—a known potent negative regulator of the Jak/Stat3 signaling pathway [48]. As a result, Zac1 curtails prodifferentiating signaling and astroglial differentiation.

Genome-wide expression profiling in the C17.2 neural stem cell line identified Socs3 as potential Zac1 target gene. A series of complementary experiments including transfection, mutational, and immunoprecipitation studies corroborated that Zac1 binds to the promoter and intronic regions of Socs3 to confer transactivation in human and mice.

With respect to the Zac1 expression pattern in the developing mouse nervous system [8, 11, 17], we focused on the role of Zac1-mediated Socs3 regulation in the context of NSC differentiation. We chose the NS-5 and O4ANS cell lines (derived from fetal mice brain (E15) and from the adult mouse SVZ, respectively) as cellular model, which show a radial glia-like phenotype [25, 26]. Astroglial as well as neuronal differentiation resulted in transient upregulation of Zac1 mRNA and protein whereby transactivation of Socs3 was confined to the astroglial lineage in either model consistent with a role as a lineage-specific target gene. Hereby, sustained DNA methylation of the Socs3 gene during neuronal differentiation maintained gene silencing and prevented Zac1 binding and transactivation.

Upon astroglial differentiation, Zac1 and Socs3 expression colocalized exclusively at an early stage concurrent with a strong decrease in receptor activation-dependent tyrosine phosphorylation of Stat3 (Y705). In accord with these findings, we detected Zac1 and Socs3 colocalization in the neocortical ventricular zone at E18 with the onset of astrogliogenesis. This led us to propose that Zac1-dependent Socs3 induction during early astroglial differentiation constitutes a negative feedback loop restraining Jak/Stat3 signaling.

The onset of astroglial differentiation depends on initial activation of Jak/Stat3 signaling by cardiotrophin-1, CNTF, or leukemia inhibitory factor (LIF) [3, 37, 49, 50]. In accord with these findings, genetic deletions in the Jak/Stat3 signaling pathway (i.e., gp130, LIF and Stat3) impair astroglial differentiation [3, 51, 52].

Likewise, Socs3 overexpression inhibits Stat3-mediated astrogliogenesis [53]. Socs3 null mutant mice die between days 11 and 13 of gestation due to defects in placental development [54]. NSC/neural precursor cell cultures generated from E10.5 embryos show sustained Stat3 phosphorylation and induction of Gfap in response to LIF treatment [55]. Moreover, Socs3 has been conditionally deleted by Cre recombinase expressed under the control of the Nestin promoter [50]. This leads to enhanced astrogliogenesis in Socs3-deficient neonatal mouse brain and primary neuroepithelial cells.

Consistent with an inhibitory role in Jak/Stat3 signaling, doubling of Zac1 expression postponed astroglial differentiation despite enhanced G1-phase arrest. By contrast, silencing of Zac1 advanced early astroglial differentiation in the presence of delayed cell cycle arrest. In many somatic cells, the length of the G1-phase of the cell cycle is tightly coupled to differentiation, and its elongation can trigger differentiation [56]. Although it has been proposed that a rapid cell cycle and a short G1-phase maintain embryonic stem cells in the pluripotent state, lengthening of the cell cycle seems insufficient to induce differentiation on its own or to facilitate differentiation [57]. Our results in NSCs conform to this idea and suggest that cell cycle arrest acts in concert with extrinsic signaling pathways to initiate differentiation. In fact, delayed astroglial differentiation in Zac1 overexpressing cells was rescued by additional Socs3 knockdown reinstating Jak/Stat3 signaling.

While O4ANS cells did not reassume cell division following astroglial differentiation, most likely due to the absence of appropriate signals, cortical astroglia have been shown to undergo symmetric cell division during early postnatal life raising the prospect that the downregulation of Zac1 after the initial peak correlates with a release from cell cycle arrest.

BMPs participate in astroglial maturation via Smad signaling to activate differentiation determinant genes and cell cycle exit by the cyclin-dependent kinase inhibitors p21Waf1/Cip1 and p27Kip1 [44, 50]. In contrast to impaired astroglial differentiation under CNTF treatment, BMP4 signaling was unaffected by Zac1 expression supporting that its regulatory function is limited to the prodifferentiating Jak/Stat3 signaling pathway.

Our results from embryonic and adult NSC lines were in agreement with the findings from primary E18 NSCs, in particular Zac1 and Socs3 colocalization was confined to early astroglial differentiation, Zac1 overexpression delayed and knockdown of either Zac1 or Socs3 advanced differentiation.

Zac1 null mutant mice (Zac1+/−pat) suffer from intrauterine growth retardation, an overall reduction in brain size and die from embryonic cardiovascular and postnatal breathing defects [58, 59]. Accordingly, conditional deletion of Zac1 in NSCs appears necessary to dissect system-wide from cell autonomous effects.

Repression of Zac1 in epithelial and mesenchymal cells by mitogenic pathways [60–62] indicates that withdrawal of EGF/FGF may also underlie Zac1 induction in NSCs. Consistent with a role in differentiation, the proneural transcription factors neurogenin 1 and 2 [63], the morphogen retinoid acid [64, 65], and the neuropeptide pituitary adenylate cyclase-activating polypeptide (PACAP) [66, 67] have been reported to enhance Zac1 expression. These findings suggest that Zac1 is downstream to mitogenic and prodifferentiative pathways in developing and adult mice and operates during the transition from proliferation to early differentiation in different types of progenitor/stem cells.

There is increasing evidence that imprinted genes influence brain function, behavior, and mental diseases by affecting neurodevelopmental processes [6, 68] and for brain's particular sensitivity to imprinting compared to other tissues [69]. Work in humans on the patterns of inheritance associated with a number of common mental and neurological disorders indicate “parent of origin” effects consistent with the action of imprinted genes [70]. Therefore, it appears conceivable that changes in ZAC1's imprinting status cause delayed or precocious astroglial differentiation leading to inappropriate wiring, disorganization, and eventually dysfunction of the CNS.

Despite improved insight into the mechanisms of genomic imprinting, the role of imprinted genes in brain function remains incompletely understood [6]. Together with other recent work [5] our results suggest that imprinted genes can influence brain function by controlling cell-fate decisions and differentiation in NSCs. These activities extend the critical functions of imprinted genes and provide new inroads for their role in the developing and adult brain and related disorders.

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

We conclude that Zac1 is transiently induced in differentiating NSCs and transactivates in a lineage-specific manner Socs3 - a known potent negative regulator of the Jak/Stat3 signaling pathway. As a result Zac1 curtails prodifferentiating signaling and prevents precocious astroglial differentiation. Overall, our findings provide new insight into the mechanisms by which imprinted genes can affect brain development and disease.

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 Albin Varga and Laurent Journot for help in animal experiments and discussion. The NS-5 and O4ANS mouse NSC lines were generously gifted by Austin Smith (Cambridge University, U.K.). Mice and human SOCS3 plasmids were kindly provided by Christoph Auernhammer (University of Munich, Germany) and Jason Matthews (University of Toronto, Canada). This research was supported by the funding provided by the Deutsche Forschungsgemeinschaft (SP 386/5-1 to D.S.)

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_1405_sm_SupplFigure1.tif132KFigure S1. ZAC1 transactivates the human SOCS3 gene. (A): Scheme of the SOCS3 5′end including the promoter, exons (filled boxes) and first intron. Signal flag symbolizes transcriptional start site. Predicted ZAC1 DNA-binding sites are depicted by filled bars and circles. ChIP primer pairs flanking amplified regions are shown by filled triangles and Roman numbers, respectively. (B): Reporter assays in SK-N-MC cells cotransfected with increasing amounts of human ZAC1 (50 ng, 100 ng, 250 ng, 500 ng) and a 1.7 kb human SOCS3 promoter construct (200 ng). (C): Transfection of ZAC1 (200 ng) increased SOCS3 expression compared to mock transfection as determined by qRT-PCR. (D): Immunoblot (WCE 50 μg) analysis of ZAC1 and SOCS3 expression following mock and ZAC1 (200 ng) transfection (E): ChIP assay with a ZAC1 antibody showed binding at the SOCS3 promoter (II), exon 1 (III) and intron (IV) in mock and ZAC1 (200 ng) transfected SK-N-MC cells. No binding could be detected at the upstream region (I) devoid of ZAC1 DNA binding motifs. Representative results from 3 (D) or means with standard deviations (± SD) from 6 independent (B, C, E) experiments are shown. * P < 0.05 and ** P < 0.01.
STEM_1405_sm_SupplFigure2.tif223KFigure S2. Zac1 induces Socs3 expression during astroglial differentiation of NS-5 cells. (A): Promoter reporter assays. Cotransfection of increasing amounts of Zac1 (5 ng, 25 ng, 50 ng, 100 ng) enhanced Socs3 promoter (200 ng) activity. (B): Astroglial and neuronal differentiation of the embryonic NSC line NS-5. Expression of Nestin, Gfap and Tuj1 was monitored by RT-PCR. (C): Time course analysis of Zac1 and Socs3 mRNA expression during astroglial and neuronal cell differentiation as measured by qRT-PCR. Zac1 was expressed in both lineages, whereas Socs3 induction was confined to the astroglial lineage. (D): Immunoblot analysis (WCE 70 μg) revealed concomitant increases in Zac1 and Socs3 expression during astroglial differentiation, while Zac1 expression during neuronal differentiation did not enhance Socs3 expression. (E): ChIP analysis during astroglial or neuronal differentiation evidenced Zac1 occupancy at the Socs3 promoter, exon 1 and intron solely during the former. Representative results from 3 (B, D) or means ± SD from 6-7 independent (A, C, E) experiments are shown. * P < 0.05 and ** P < 0.01.
STEM_1405_sm_SupplFigure3.tif170KFigure S3. Histone marks at the Socs3 gene during neuronal differentiation of O4ANS cells. (A and B): Chromatin marks during neuronal differentiation at day 4 were detected with antibodies against acH3 (A) and H3K9me2 (B). (C and D): Sequential ChIP analysis during neuronal differentiation showed unchanged association of Zac1 with active (acH3/Zac1) (C) or repressive (H3K9me2/Zac1) (D) chromatin marks. Means ± SD from 4 independent experiments are shown.
STEM_1405_sm_SupplFigure4.tif37KFigure S4. Activated polymerase II is recruited during astroglial differentiation. ChIP assay against activated polymerase II (Ser5-Pol) were performed on day 4 in differentiating O4ANS cells. Activated polymerase II is recruited during astroglial, but not neuronal, differentiation. Means ± SD from 4 independent experiments are shown. * P < 0.05 and ** P < 0.01.
STEM_1405_sm_SupplFigure5.tif115KFigure S5. Demethylation of Socs3 during astroglial differentiation (A): Scheme of Socs3 upstream region with overlaying CpG island shown beneath. Primer pairs flanking amplified regions are indicated. (B): MeDIP assay of the upstream region evidenced low methylation at the promoter and transcriptional start site, whereas higher levels of methylation were detected at the exonic and intronic regions. In either case reduced DNA methylation at different Zac1 binding sites preferentially occurred during astroglial differentiation. (C): Overall methylation at the Socs3 upstream region following astroglial and neuronal differentiation. Means ± SD from 4 independent (B, C) experiments are shown. * P < 0.05 and ** P < 0.01.
STEM_1405_sm_SupplFigure6.tif429KFigure S6. Zac1 expression inhibits Stat3 (Y705) phosphorylation (A): Immunoblot (WCE 70 μg) of mock or Zac1 (250 ng) transfected C17.2 cells. Zac1 transfection increased Socs3 expression and decreased Stat3 (Y705) phosphorylation while expression of Stat3 and β-actin proteins was unaffected. (B): Following astroglial differentiation of O4ANS, Stat3 (Y705) phosphoimmunoreactivity decreased as evidenced by immunoblot (WCE 70 μg). In contrast, Stat3 protein expression was unaltered under either astroglial or neuronal differentiation. Representative results from 3 independent experiments are shown.
STEM_1405_sm_SupplFigure7.tif2724KFigure S7. Glial markers colocalize with Zac1 or Socs3 during the transition to differentiation. (A): O4ANS undergoing astroglial differentiation were subjected to indirect immunofluorescence with antibodies A2B5 (red) (a-j), Gfap (green) (k-t), Zac1 (green) (a-e) and (red) (k-o), Socs3 (green) (f-j) and (red) (p-t). Nuclei were stained with DAPI (blue). Scale bar 40 μm. Representative results from 5 independent experiments are shown.
STEM_1405_sm_SupplFigure8.tif2630KFigure S8. Zac1 induces Socs3 in differentiating astroglial cells during the transition to maturation. (A): Astroglial differentiation of NS-5 cells. At different days cells were subjected to indirect immunofluorescence with antibodies against A2B5 (a-d) and Gfap (e-h), merge (i-l) or Zac1 (a′-d′) and Socs3 (e′-h′), merge; (i′-l′). Nuclei were stained with DAPI (blue). Scale bar 40 μm. (B): Quantitative analysis of A2B5 and Gfap positive cells are shown as percentage of the total number of DAPI positive cells. Mean of five analyzed images per day of the type shown in (A, a-l). (C): Quantitative analysis of Zac1 and Socs3 positive cells are shown as percentage of the total number of DAPI positive cells. Mean of five analyzed images per day of the type shown in (A, a′-l′). (D): Colocalization of Zac1 (green) (a-d) and (red) (i-l), and Socs3 (green) (e-h) and (red) (m-p) with glial markers A2B5 (red) (a-h) or Gfap (green) (i-p). Scale bar 40 μm. Representative results from 5 independent experiments are shown.
STEM_1405_sm_SupplFigure9.tif1145KFigure S9. Socs3 knock-down reinstates astroglial differentiation in Zac1 overexpressing O4ANS cells. (A): Zac1 and Socs3 expression in parental, Flag-Zac1 (FZ) overexpressing, and Flag-Zac1 overexpressing Socs-3 knock-down (FZ-shRNA-Socs3) cells were measured by qRT-PCR whereby parental expression was set to 100%. (B): Integration of Socs3 shRNA plasmids was evidenced by the selection marker neomycin. (C): Zac1 and Socs3 expression in FZ-shRNA-Socs3 cells during astroglial differentiation as measured by qRT-PCR. (D): Distribution of cell cycle phases in FZ and FZ-shRNA-Socs3 cells during astroglial differentiation. (E): Immunocytochemistry of FZ and FZ-shRNA-Socs3 cells during astroglial differentiation. Cells were subjected to indirect immunofluorescence with antibodies against A2B5 (red) and Gfap (green). Nuclei were stained with DAPI (blue). Scale bar 40 μm. Representative results from 2-4 (B, E) or means ± SD from 6-8 independent (A, C, D) experiments are shown. * P < 0.05 and ** P < 0.01.
STEM_1405_sm_SupplTable1.pdf22KTable S1. Primers for PCR and ChIP 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.