Etv5a Suppresses Neural Progenitor Cell Proliferation by Inhibiting sox2 Transcription

Abstract

Neural progenitor cells are self-renewable, proliferative, and multipotent cell populations that generate diverse types of neurons and glia to build the nervous system. Transcription factors play critical roles in regulating various cellular processes; however, the transcription factors that regulate the development of neural progenitors are yet to be identified. In the present study, we demonstrated that zebrafish etv5a is expressed in the neural progenitor cells of the neuroectoderm. Downregulation of endogenous Etv5a function by etv5a morpholino or an etv5a dominant-negative variant increased the proliferation of sox2-positive neural progenitor cells, accompanied by inhibition of neurogenesis and gliogenesis. These phenotypes in Etv5a-depleted embryos could be rescued by a co-injection with etv5a cRNA. Etv5a overexpression reduced sox2 expression. Direct binding of Etv5a to the regulatory elements of sox2 was affirmed by chromatin immunoprecipitation. These data revealed that Etv5a directly suppressed sox2 expression to reduce the proliferation of neural progenitor cells. In addition, the expression of foxm1, a putative target gene of Etv5a and a direct upstream transcription factor of sox2, was upregulated in Etv5a-deficient embryos. Moreover, the suppression of Foxm1 function by the foxm1 dominant-negative construct nullified the phenotype of upregulated sox2 expression caused by Etv5a deficiency. Overall, our results indicated that Etv5a regulates the expression of sox2 via direct binding to the sox2 promoter and indirect regulation by inhibiting foxm1 expression. Hence, we revealed the role of Etv5a in the transcriptional hierarchy that regulates the proliferation of neural progenitor cells.

Introduction

Neural progenitor cells are self-renewing, proliferative, and multipotent cells that give rise to neuronal and glial derivatives of the nervous system. These cells are generated from the neuroectoderm, which is derived from the ectoderm during early embryogenesis. The maintenance, proliferation, and differentiation of neural progenitor cells must be tightly controlled by environmental signals and intrinsic regulators.

The dysregulated function of these factors results in the defective development of neural progenitors, leading to neurological disorders and tumors of the nervous system. Signaling pathways, such as bone morphogenetic protein (BMP), Wnt, fibroblast growth factor (FGF), Sonic hedgehog (SHH), and Notch, are major players in regulating neural progenitor development. Further, their downstream transcription factors serve as convergence points of cell signaling by directly binding to the promoters or enhancers of target genes and activating or repressing DNA transcription.

SOX2 is a core transcription factor that is involved in the induction, maintenance, and differentiation of pluripotent embryonic stem cells [1 –3]. SOX2 also regulates neuroectoderm specification and governs the maintenance, self-renewal, and proliferation of neural progenitor cells [reviewed in Ref. (4)]. Neural progenitors differentiate into neuronal precursors that are regulated by transcription factors, such as hairy and enhancer of split-1 (HES1) and HES5.

These transcription factors regulate self-renewal and inhibit neuronal differentiation by repressing the expression of proneural genes such as achaete-scute family bHLH transcription factor 1 (ASCL1), neurogenin 1 (NEUROG1), and NEUROG2. In contrast, oligodendrocyte differentiation is regulated by oligodendrocyte transcription factor 1 (OLIG1) and OLIG2 [5,6].

The E26 transformation-specific (ETS) transcription factor family plays multiple roles in cellular processes, such as proliferation, differentiation, migration, and apoptosis, and is implicated in many disorders and cancers [7 –9]. The ETS family members contain a conserved winged helix–turn–helix ETS domain that commonly binds the GGA(A/T) sequence in the ETS-binding site (EBS).

Further, 28 human ETS proteins have been identified and are divided into 12 subfamilies according to their sequence similarity [10]. The ETS variant transcription factor 5 (ETV5; also known as ERM) belongs to the PEA3 subfamily, which includes ETV1, ETV4, and ETV5. ETV5 has attracted attention because of its role in various cancers, including thyroid cancer [11], colorectal cancer [12], synovial sarcoma [13], and neuroblastoma [14,15].

ETV5 regulates the proliferation of mouse embryonic stem cells [16], and it is required for the proliferation of neural stem cells in the ventricular zone of mouse embryos [17]. Further, ETV5 regulates the differentiation of Schwann cells from neural crest progenitor cells [18] and represses NEUROG2 expression to regulate the differentiation of glutamatergic neurons from neural progenitor cells [19].

Zebrafish contain two paralogs of mammalian ETV5, namely Etv5a and Etv5b (also known as Erm in zebrafish) [20]. Previous studies have demonstrated that Etv5a regulates hematopoiesis [20] and kidney development [21], whereas Etv5b mediates the development of hypothalamic serotoninergic neurons [22]. In the present study, we discovered that zebrafish Etv5a regulates the proliferation of neural precursor cells in the neuroectoderm by repressing sox2 and foxm1 expression. This study provides further insights into the role of transcription factors in neural progenitor processes.

Materials and Methods

Ethics statement and zebrafish maintenance

The experiments performed in this study followed standard guidelines for zebrafish research and were approved by the Institutional Animal Care and Use Committee of Chang Gung University (IACUC approval no.: CGU11-118).

Tü (wild-type) zebrafish were purchased from the Zebrafish International Resource Center (Oregon). Adult fish were raised, maintained, and paired, and embryos were collected according to standard guidelines. Embryos were staged according to the hours post-fertilization (hpf) and days post-fertilization [23].

cRNA and morpholino for injection

Zebrafish etv5a, etv5a^Δacidic , and etv5b constructs were used to create capped RNA for injection, and they were prepared as previously described [20]. etv5a morpholino (MO) (5′-TCACCTGGGTCTTCAAAGAGGCTCC-3′) that overlaps the ATG start codon (−26 to −2) was purchased from Gene Tools, LLC (Oregon), and the concentration for injection was determined as described by Chen et al. [20].

Zebrafish foxM1 ^ΔC was amplified with TOOLS Ultra High Fidelity DNA Polymerase (BIOTOOLS, Taiwan). The sox2 promoter region (sox2^p , −2.2 to −1.2 kb) and constructs containing the different EBS sequences (EBS-i, EBS-ii, and both EBS-i and EBS-ii deletions) were amplified by polymerase chain reaction (PCR) and cloned into the pEGFP-N3 vector (see Supplementary Table S1 for a full list of primers). We injected 0.15 ng sox2^p:eGFP variants and 2.512 ng etv5a cRNA. Embryos were harvested at 8 hpf and subjected to quantitative real-time PCR (qRT-PCR) analysis.

Histological analysis

For in situ hybridization, digoxigenin-uridine triphosphate-labeled riboprobes were synthesized, and in situ hybridization was performed as previously described [24]. mRNA expression was observed using the nitro blue tetrazolium (NBT)/5-Bromo-4-chloro-3-indolyl phosphate (BCIP) substrate (Roche). For double in situ hybridization, digoxigenin- and fluorescein-labeled riboprobes were used and detected using anti-digoxigenin or anti-fluorescein antibodies. These antibodies were conjugated with alkaline phosphatase.

The color reaction protocol involved the following two steps: First, ImmPACT Vector Red (Vector Laboratories, Inc.) was used. Once the initial signals were developed, alkaline phosphatase was inactivated using 0.1 M glycine-HCl at a pH of 2.2. Second, an alkaline phosphatase-conjugated anti-digoxigenin or anti-fluorescein antibody was introduced, followed by the addition of BCIP/NBT for the second color reaction.

For immunohistochemistry, rabbit phospho-histone H3 antibody or rabbit monoclonal anti-active caspase-3 antibody (Abcam) was applied to the samples. The signal was detected using fluorochrome-conjugated Alexa Fluor 488 goat anti-rabbit IgG antibodies (Invitrogen). Goat anti-rabbit IgG horseradish peroxidase (HRP) (Invitrogen) was used to detect the primary antibodies, and 3,3′-diaminobenzidine was used as a substrate for secondary antibody-conjugated HRP (Amresco) (see Supplementary Table S2 for the full list of antibodies).

The embryos were mounted using Vectashield mounting medium and stained with 4′,6-diamidino-2-phenylindole (Vector Laboratories, Inc.) to visualize nuclear DNA and preserve fluorescence. Phospho-histone H3- and caspase-3-positive cells were manually counted.

Quantitative real-time polymerase chain reaction

Total RNA was extracted from the embryos and subjected to cDNA synthesis using the TOOLS Easy Fast RT Kit (BIOTOOLS, Taiwan). RT-qPCR was performed using the TOOLS 2X SYBR qPCR Mix (BIOTOOLS, Taiwan). Gene expression levels were normalized to that of gapdh and measured using the comparative Ct method (40 cycles) according to the manufacturer's instructions (Applied Biosystems).

Cell maintenance and transfection

The 293T (human embryonic kidney) and U87 (human gliobastoma) cell lines were obtained from American Type Culture Collection and cultured in Dulbecco's modified Eagle's medium (Gibco) supplemented with 10% fetal bovine serum (Avantor) and 1% Antibiotics-Antimycotics (GeneTeks). For plasmid transfection, the plasmid DNA was diluted and incubated with DreamFect Gold reagents (OZ biosciences) for 15 min. Subsequently, the mixture was added to the culture medium, which was used to incubate the cells.

Chromatin immunoprecipitation

Chromatin immunoprecipitation (ChIP) was performed as described by Lin et al. [25]. In brief, etv5a was cloned into a pCS2-MT vector containing a Myc tag, transcribed into cRNA in vitro, and injected into zebrafish embryos. The injected embryos were fixed with formaldehyde to crosslink the proteins and DNA. Chromatin was isolated using a Magna ChIP A/G ChIP Kit (EMD Millipore) and fragmented to an average size of ∼300 bp using sonication.

Protein A- and G-conjugated magnetic beads and anti-Myc antibodies were added to the chromatin solution. Protein A- and G-conjugated beads—antibody and chromatin complexes—were collected using a magnetic separator, and the protein/DNA complexes were isolated. The immunoprecipitated DNA was subjected to PCR using primers that frame the EBS-i and EBS-ii regions of the 5′-untranslated region of sox2.

Human ETV5 was cloned into a pCS2-MT vector containing a Myc tag and transfected into U87 cells. The transfected cells were subjected to a ChIP assay using a Magna ChIP A/G ChIP Kit (EMD Millipore). Thereafter, the immunoprecipitated DNA was amplified using primers designed to amplify the EBS sequence in the SOX2 promoter.

Statistical analysis

Quantitative data are presented as mean ± standard deviation. Statistical analyses were performed using Student's t test or one-way analysis of variance (ANOVA). All experiments were performed in triplicate for each sample, and P < 0.05 was considered significant.

Results

etv5a is expressed in the neuroectoderm and developing central nervous system

We found that zebrafish etv5a was expressed in the neuroectoderm, which mainly contains neural progenitors (Fig. 1A). Accordingly, we validated the characteristics of these etv5a-positive cells at 8 h post-fertilization (hpf) by counter-staining for the neural progenitor marker sox2 using double in situ hybridization. sox2 expression was displayed as a rhombus representing the location of neural progenitors (Fig. 1A), and etv5a expression was co-localized with sox2 in the areas shown in Fig. 1B and C.

FIG. 1.

etv5a expression in neural progenitors and neural derivatives. In situ hybridization was performed to determine the type of etv5a-expressing cells in the zebrafish embryos. (A) The left-most panels are schematic illustrations showing the location of the neuroectoderm, se, sox2 expression area, and etv5a expression area. The other panels are the in situ hybridization results showing that etv5a was ubiquitously expressed in the neuroectoderm and mesoderm, whereas sox2 was only expressed in the neuroectoderm (the expression pattern was rhombus shaped when observed dorsally). i, ii, iii, and iv are enlargements of the areas indicated by dashed rectangles. (B) Lateral view with dorsal to the right and animal pole to the top. The whole-mount double in situ hybridization with etv5a (blue) and the neural progenitor marker, sox2 (red), with cryosection (bottom rows), demonstrating partial co-localization of etv5a and sox2. The different color channels were separated after in situ hybridization. (C) After double in situ hybridization, the embryos were dissected and flat-mounted to present their dorsal neuroectoderm (anterior to the top). sox2 was strongly expressed in this area, whereas etv5a displayed different expression levels, co-localized with sox2 expression. Purple: sox2 co-localized with strong etv5a expression; merlot/magenta: sox2 co-localized with weak etv5a expression. Cell nuclei were detected using DAPI (blue-fluorescent). The dashed areas indicate representative cells co-expressing etv5a and sox2. (D) At 10 hpf, etv5a and sox2 expression did not overlap. Left panels, dorsal view; right panels, lateral view. (E, F) Lateral view with anterior to the left and posterior to the right. (E) At 24 hpf, etv5a is expressed in the developing brain and sensory cells of the spinal cord located in the dorsal region of the spinal cord. The enlarged view of the spinal cord in “D” highlights etv5a expression in the sensory neurons. (F) At 48 hpf, etv5a is expressed in the forebrain, cerebellum, hindbrain, and pharyngeal arches. ce, cerebellum; DAPI, 4′,6-diamidino-2-phenylindole; e, eye; fb, forebrain; hb, hindbrain; int, intestine; ne, neuroectoderm; pa, pharyngeal arches; se, surface ectoderm; sn, sensory neurons.

Outside of the sox2-positive area, ubiquitous weak expression of etv5a was observed in the entire neural ectoderm and mesoderm, whereas relatively strong expression was observed in the dorsal neuroectoderm and ventral mesoderm, as previously described [20] (Fig. 1A, B). This result confirmed that etv5a is expressed in neural progenitors and may regulate the development of these cells. Further, etv5a was expressed in the brain and spinal cord during the segmentation and organogenesis stages (Fig. 1D–F), indicating that Etv5a may also regulate the formation and differentiation of neural derivatives.

Etv5a deficiency induces the proliferation of neural progenitor cells

To study the role of Etv5a in zebrafish neural development, we used etv5a MO and an etv5a dominant-negative variant to perform loss-of-function analysis. The specificity and effectiveness of this etv5a MO knockdown have been demonstrated in hematopoiesis, and it does not cause off-target p53 activation [20].

Moreover, the effectiveness of etv5 MO was confirmed by a deletion variant that lacks the acidic transcription regulatory domain (etv5a^Δacidic ) and causes a dominant-negative effect [20]. Injection of etv5 MO or etv5a^Δacidic cRNA induced expression of the neural progenitor marker sox2 (Fig. 2A, C). The phenotypes caused by etv5a MO were rescued by a co-injection with etv5a cRNA (Fig. 2A, C).

FIG. 2.

Etv5a deficiency induces the proliferation of neural progenitor cells. (A) Dorsal view with the animal pole to the top. Neural progenitors were identified by the expression of neural progenitor marker sox2 using in situ hybridization at 8 hpf. Injection of etv5 MO or etv5a^Δacidic upregulated sox2 expression. This result was quantified using qRT-PCR shown in (C) The percentages (%) indicate the proportion of embryos that displayed the same phenotype as that shown on each photo. (B) Embryos in (A) were flat-mounted for a better presentation of the neuroectoderm, and the pH3-positive cell fraction of all the sox2-positive areas was counted (upper panels). The bottom panels are enlargements of the areas indicated in the upper panels that show counter-stained cells for sox2 (fluorescent red), phospho-histone 3 (pH3, fluorescent green), and nuclei marker DAPI (fluorescent blue). The sox2 and pH3 double-positive cells are encircled in the red squares. The sox2-negative and pH3-positive cells are encircled in black squares. Injection of etv5a MO or etv5a^Δacidic increased the number of sox2 and pH3 double-positive cells as well as sox2-negative pH3-positive cells. The number of cells were counted using positive DAPI staining and quantified in (D). Values are presented as the mean ± SD. *P < 0.05; **P < 0.01. MO, morpholino; n.s., not significant; pH3, phosphorylated histone-3; qRT-PCR, quantitative real-time polymerase chain reaction; SD, standard deviation.

These results demonstrate the specificity of etv5a MO. Next, we analyzed neural progenitor proliferation by counter-staining for sox2 with the cell proliferation marker phosphorylated histone-3 (pH3). Our results revealed that the number of sox2 and pH3 double-positive cells had increased by etv5 MO or etv5a^Δacidic injection (Fig. 2B, D), indicating that Etv5a deficiency is sufficient to induce the proliferation of neural progenitors.

Notably, Etv5a deficiency also induced the proliferation of sox2-negative neuroectodermal cells (Fig. 2B, D). This result suggests that Etv5a also regulates the proliferation of neural progenitor cells that give rise to the posterior central nervous system, because sox2 is expressed strongly in the presumptive forebrain and correlates with future subdivisions of the brain [26].

To examine the effect of Etv5a on the apoptosis of neural progenitor cells, we counter-stained for sox2 and apoptotic marker activated caspase-3. Our results revealed that injection of etv5 MO or etv5a^Δacidic did not affect the number of sox2 and caspase-3 double-positive cells (Supplementary Fig. S1), indicating that Etv5a deficiency does not regulate the apoptosis of neural progenitors. To further confirm this, we knocked down p53 expression in Etv5a-deficient embryos by co-injecting etv5a MO with tp53 MO and found that tp53 MO could not restore the phenotype of increased sox2 expression caused by etv5a MO (Fig. 2B, D, and Supplementary Fig. S1).

Etv5a deficiency reduces the formation of neural derivatives

Next, we evaluated the effect of Etv5a deficiency on neural derivatives. Injection of etv5a MO or etv5a ^Δacidic downregulated the expression of the neuronal precursor marker neurog1 at 10 hpf (Fig. 3). In contrast, Etv5a deficiency did not alter the number of caspase-3 positive apoptotic cells in the neurog1-positive areas (Supplementary Fig. S1).

FIG. 3.

Etv5a deficiency inhibits the development of neuronal derivatives. (A) Dorsal view with anterior to the top. Neuronal precursors were analyzed with neurog1 riboprobe using in situ hybridization at 10 hpf. Injection of etv5a MO or etv5a ^Δacidic cRNA downregulated neurog1 expression. This result was quantified using qRT-PCR (B) Values are presented as the mean ± SD. *P < 0.05.

This result suggests that Etv5a deficiency causes the proliferation of neural progenitors and maintains progenitor cells, consequently inhibiting neuronal differentiation, as demonstrated by the decreased number of neuronal precursors.

To address the role of Etv5a in gliogenesis, we first analyzed the radial glial marker glial fibrillary acidic protein (Gfap) in Etv5a-depleted embryos at 24 hpf. Injection of etv5a MO or etv5a ^Δacidic induced gfap expression (Fig. 4). Next, we examined the effect of Etv5a deficiency in glial derivatives using the oligodendrocyte marker olig1 and mag (myelin-associated glycoprotein).

FIG. 4.

Etv5a deficiency induces radial cell formation and inhibits oligodendrocyte formation. (A) Dorsal view with anterior to the top. Radial glial cells were examined by gfap expression, and oligodendrocytes were detected by olig1 and mag expression using in situ hybridization. Injection of etv5a MO or etv5a ^Δacidic cRNA induces gfap expression and inhibits olig1 and mag expression. (B) The result in (A) was confirmed and quantified using qRT-PCR. (C) The brain sizes were measured using the width of the distal margins of two eyes, indicated in yellow dashed lines in (A), which demonstrates that Etv5a-deficiency reduced brain sizes. Values are presented as the mean ± SD. *P < 0.05; **P < 0.01; ****P < 0.00001.

The injection of etv5a MO or etv5a ^Δacidic inhibited the expression of olig1 and mag at 48 hpf (Fig. 4), indicating that Etv5a is essential for oligodendrocyte formation. Co-injection with etv5a cRNA rescued these glial phenotypes in Etv5a-depleted embryos.

Etv5a directly binds to the upstream region of sox2 and inhibits its transcription

Etv5a deficiency induced sox2 expression (Fig. 2), whereas overexpression of the full coding sequence of etv5a cRNA reduced sox2 expression, suggesting that Etv5a may bind to sox2 (Fig. 5A, B). Mammalian ETV5 recognizes the consensus sequence ACCGGAAGT(G/A) of the 5′ untranslated region of target genes [27]. We identified two putative Etv5a binding sites (namely EBS-i and EBS-ii) at −2 kb upstream from the start codon of sox2 (Fig. 5C) and discovered the direct binding of Etv5a to the sox2 promoter using ChIP analysis.

FIG. 5.

Etv5a represses sox2 expression by directly binding to the promoter region. (A, B) Overexpression of full coding sequence of etv5a cRNA reduced the sox2 expression at 8 hpf analyzed using in situ hybridization (A) and quantified using qRT-PCR (B, C) Schematic illustration of the promoter region of sox2. Two fragments containing putative EBSs, named EBS-i and EBS-ii, are indicated. A fragment, located at an anti-sense strand with an EBS sequence, was named pseudo-EBS (−1), and a fragment with no predicted EBS binding sequence, adjacent to EBS-ii, the juxtapose control, is indicated. These two fragments were used as non-binding controls. (D) The full coding sequence of etv5a was cloned into a pCS2 vector containing a Myc-tag (etv5a-MT). The expression level of etv5a-MT was affirmed using qRT-PCR. (E, F) ChIP–qPCR was performed using etv5a-myc injected embryos, harvested at 8 hpf. EBS-i and EBS-ii fragments were amplified using qPCR, demonstrating the direct binding of Etv5a to these regions. No signal was detected on juxtaposed and pseudo-EBS controls, demonstrating the specific binding of Etv5a on EBS-i and EBS-ii. One percent input of ChIP samples was used as a positive control. (G) Schematic illustration of Etv5a-sox2^p promoter-reporter assay. This sox2^p:eGFP construct was assembled in a pEGFP-N3 backbone with the zebrafish sox2 promoter region (−2.2 to −1.2 kb), containing EBS-i and EBS-ii fragments. (H) qRT-PCR analysis of eGFP expression level in sox2^p:eGFP injected embryos at 8 hpf. Injection of etv5a cRNA significantly reduced eGFP expression in sox2^p:eGFP injected samples, whereas injection of sox2^p:eGFP alone displayed higher eGFP expression. (I) Left, schematic illustration showing the deletion constructs derived from sox2^p:eGFP. Either one or both EBS-i and EBS-ii were deleted [sox2^pΔEBS-i:eGFP (EBS-i deletion), sox2^pΔEBS-ii:eGFP (EBS-ii deletion), and sox2^pΔEBS-(i+ii):eGFP (EBS-i and EBS-ii deletion)]. Co-injection of etv5a with sox2^p:eGFP inhibited GFP expression, whereas etv5a did not repress GFP expression when co-injected with sox2^pΔEBS-(i+ii):eGFP. In contrast, Etv5a inhibited the expression of GFP when co-injected with sox2^pΔEBS-i:eGFP (EBS-i deletion) or sox2^pΔEBS-ii:eGFP (EBS-ii deletion). However, the inhibitory level was less effective than that of Etv5a co-injected with sox2^p:eGFP (with intact EBS-i and EBS-ii). Values are presented as the mean ± SD. *P < 0.05; **P < 0.01; ***P < 0.001. ChIP, chromatin immunoprecipitation; EBS, ETS-binding site.

Myc-tagged etv5a cRNA was injected into fertilized eggs (Fig. 5D), and chromatin fragments isolated from these embryos were immunoprecipitated with an antibody against the Myc tag. In addition to the two putative binding fragments, we designed two non-binding sequence controls, namely, a fragment adjacent to the putative binding sites and a fragment at the anti-sense strand.

qPCR analysis was performed, using primers designed for the putative binding or non-binding sequences. The putative binding fragments were selectively amplified (Fig. 5E, F). To confirm the results of the ChIP-qPCR assay, we performed an Etv5a-sox2 promoter-reporter assay. The zebrafish sox2 promoter region (−2.2 kb to −1.2 kb), which contains the Etv5a-binding fragments, was cloned into a pEGFP-N3 vector, named sox2^p:eGFP.

This sox2^p:eGFP construct was injected with or without etv5a cRNA into zebrafish embryos. A concomitant injection of etv5a and sox2^p:eGFP inhibited eGFP expression (Fig. 5G, H), demonstrating that Etv5a binds to the sox2 promoter to inhibit eGFP expression. To confirm whether EBS-i and/or EBS-ii are essential for Etv5a to inhibit sox2 expression, we generated GFP reporter constructs driven by the sox2 promoter and containing different EBS-i and EBS-ii sequences.

In these constructs, either one of or both EBS-i and EBS-ii were deleted [sox2^pΔEBS-i:eGFP (EBS-i deletion), sox2^pΔEBS-ii:eGFP (EBS-ii deletion), and sox2^pΔEBS-(i+ii):eGFP (EBS-i and EBS-ii deletion)]. Co-injection of etv5a with sox2^p:eGFP inhibited GFP expression, whereas etv5a did not repress GFP expression when co-injected with sox2^pΔEBS-(i+ii):eGFP (Fig. 5I).

In contrast, Etv5a inhibited GFP expression when co-injected with sox2^pΔEBS-i:eGFP or sox2^pΔEBS-ii:eGFP, demonstrating that Etv5a binding to either EBS-i or EBS-ii inhibited GFP expression (Fig. 5I). However, the inhibitory level was less effective than that of Etv5a co-injected with sox2^p:eGFP (with intact EBS-i and EBS-ii) (Fig. 5I).

These results affirmed that both EBS-i and EBS-ii are essential for Etv5a binding and that interaction with either inhibits sox2 expression. Taken together, these results indicate that Etv5a inhibits sox2 transcription by directly binding to the EBSs in the promoter region.

Etv5a suppresses neural progenitor proliferation via inhibiting Foxm1 function

In Xenopus embryos, Foxm1 is expressed in the neuroectoderm and is required for the proliferation of neural plate cells [28]. In primary glioblastoma cells, FOXM1 directly binds to the SOX2 promoter, and the overexpression of FOXM1 upregulates SOX2 expression and induces the growth of glioblastoma cells [29]. In zebrafish gastrula, foxm1 was expressed in the neuroectoderm and mesoderm, a pattern very similar to that of etv5a expression (Fig. 6A).

FIG. 6.

Etv5a deficiency induces sox2 expression via upregulating foxm1 expression. (A) In situ hybridization results showing that both etv5a and foxm1 are expressed in the neuroectoderm and mesoderm. (B) Results from double in situ hybridization of etv5a and foxm1. Embryos were flat-mounted to show the neuroectoderm. The dashed areas are representative cells co-expressing etv5a and sox2. (C) etv5 MO injection upregulates foxm1 expression examined using in situ hybridization (left) and qRT-PCR (right). (D) In situ hybridization demonstrated that overexpressed foxm1 ^ΔC cRNA inhibits sox2 expression. Injection of etv5a MO induces sox2 expression, and this phenotype could be restored by concomitant injection of foxm1 ^ΔC (left). The result in (B) was confirmed and quantified using qRT-PCR (right). (E) The proliferation of neural progenitors was examined by counter-staining for sox2 with pH3. The pH3-positive cell fraction of all the sox2-positive areas was counted. The bottom panels are enlargements of the areas indicated in the upper panels that show counter-stained cells for sox2 (fluorescent red and pH3 (fluorescent green). Injection of foxm1 ^ΔC reduced the number of sox2 and pH3 double-positive cells, whereas injection of foxm1 ^ΔC in Etv5a-deficient embryos restored the number of sox2 and pH3 double-positive cells. The number of cells were counted and quantified (right). Values are presented as the mean ± SD. *P < 0.05; **P < 0.01.

Double in situ hybridization results demonstrated that the expression of etv5a and foxm1 was co-localized in the neuroectoderm (Fig. 6B), suggesting that Etv5a also regulates Foxm1 in the zebrafish neuroectoderm. To gain insights into the regulatory mechanism of Etv5a in controlling neural progenitor proliferation, we examined the role of Foxm1 in Etv5a-Sox2 mediated neural progenitor proliferation. Injection of etv5a MO significantly upregulated foxm1 expression, as analyzed using in situ hybridization and qRT-PCR (Fig. 6C), suggesting negative regulation of Etv5a on Foxm1.

To confirm that foxm1 upregulation is critical for neural progenitor development in Etv5a-deficient embryos, we used a dominant-negative Foxm1 variant to reduce the effect of foxm1 upregulation and to inhibit Foxm1 function. The design of this variant was based on a previous study demonstrating that deletion of the C-terminal domain of FOXM1 (FOXM1 ^ΔC) abolishes the transcriptional activity of endogenous FOXM1 in cancer cell lines [30].

In the present study, the overexpression of zebrafish foxm1 ^ΔC cRNA reduced sox2 expression and restored the upregulated sox2 expression caused by Etv5a deficiency (Fig. 6D). To verify the role of FoxM1 in neural progenitor proliferation, we overexpressed foxm1 ^ΔC and examined the proliferation of neural progenitors by counter-staining for sox2 with the proliferation marker, pH3.

The foxm1 ^ΔC injection reduced the proliferation of neural progenitors, whereas foxm1 ^ΔC overexpression in Etv5a-deficient embryos restored the number of proliferating neural progenitors, caused by Etv5a deficiency (Fig. 6E). This demonstrated that foxm1 ^ΔC disrupted the function of Foxm1 and thus nullified the function of the upregulated foxm1 caused by Etv5a deficiency, thereby confirming that Foxm1 is a downstream factor of Etv5a. Thus, Etv5a indirectly regulates sox2 expression and neural progenitor proliferation by suppressing Foxm1 function.

Etv5a, Etv5b, and human ETV5 are functionally diverse in regulating neural progenitors

Zebrafish Etv5a and Etv5b are paralogues to the mammalian ETV5. To examine whether Etv5b is functionally conserved with or diverted from Etv5a, we first examined etv5b expression. In the late gastrula stage, etv5b expression was very similar to that of etv5a, which was expressed in the entire neuroectoderm. At later stages, etv5b expression resembled that of etv5a at 8, 24, and 48 hpf in the developing nervous system (Supplementary Fig. S2).

However, etv5b was not as highly expressed as etv5a in the tail region at 10 hpf. These observations are similar to those previously published [31]. etv5b overexpression did not alter the expression of sox2 (neural progenitor marker) or neurog1 (neural precursor marker) (Fig. 7A), which suggests that etv5b alone did not affect the formation of neural progenitor cells or neural precursors.

FIG. 7.

Etv5b has no effect on sox2 or neurog1 expression and does not bind to the sox2 promoter. (A) Overexpression of etv5b cRNA did not alter the expression of sox2 or neuog1, as demonstrated through in situ hybridization (left panels) and qRT-PCR (right panels). (B) ChIP-PCR results showing that no fragment was amplified using primer sets specific for EBS-I, EBSii, juxtaposed control, and pseudo-EBS controls, demonstrating that Etv5b did not bind to EBS-i or EBS-ii in the sox2 promoter.

In addition, we performed a ChIP-PCR analysis to determine whether Etv5b binds to the sox2 promoter and found that Etv5b did not bind to EBS-i or EBS-ii, the region where Etv5a binds (Fig. 7B), demonstrating that Etv5b could not bind to the sox2 promoter. Overall, these results demonstrated that Etv5a and Etv5b are functionally different in regulating neural progenitors.

Next, we performed a ChIP analysis to investigate the potential interaction between human ETV5 and the SOX2 promoter using the U87 cell line. Myc-tagged ETV5 was transfected into U87 cells, and ETV5-bound chromatin fragments were immunoprecipitated using anti-Myc antibodies. The EBS in the SOX2 promoter was described by Cooper et al. [27], and we designed the primers accordingly to amplify the EBS sequences.

ChIP-PCR and ChIP-qPCR results demonstrated that the EBS was selectively amplified, indicating that ETV5 binds to the EBS in the SOX2 promoter (Fig. 8A–C). In addition, we examined whether zebrafish Etv5a is functionally conserved with human ETV5 by transfecting zebrafish etv5a cRNA into human 293T embryonic kidney cells and U87 glioblastoma cells to examine its effect on human SOX2 expression. qRT-PCR results demonstrated that the transfection of zebrafish etv5a did not alter the expression of human SOX2 (Fig. 8D).

FIG. 8.

Human ETV5 binds to the EBS in the SOX2 promoter and regulates SOX2 expression. (A) Schematic illustration of the promoter region of human SOX2. The EBS and a fragment with no predicted EBS (the juxtapose control) are indicated. (B, C) After ChIP, the EBS fragment was amplified using PCR (B) and qPCR (C) No signal was detected on the juxtaposed control, demonstrating the specific binding of ETV5 to the EBS in the SOX2 promoter. (D) U87 and 293T cells were transfected with zebrafish etv5a or human ETV5 to examine their effects on SOX2 expression. Then, they were analyzed using qPCR, whose results showed that only human ETV5 induced SOX2 expression in U87 cells. Values are presented as the mean ± SD. *P < 0.05; ***P < 0.001.

In contrast, the transfection of human ETV5 induced SOX2 expression in U87 cells but not in 293T cells (Fig. 8D). These findings demonstrated that although both ETV5 and Etv5a are capable of binding to EBS in the SOX2 and sox2 promoters, respectively, they have distinct functions within their respective species. Further, human ETV5 induced SOX2 expression in U87 cells but not in 293T cells, which also indicated that the function of ETV5 varies depending on the cell type.

Discussion

The self-renewal and multipotency characteristics of neural progenitor cells make them attractive for studies on developmental biology, tissue regeneration, and stem cell therapy. However, the molecular regulation of the development of neural progenitor cells is unclear, and data obtained from in vivo and in vitro analyses are inconsistent.

In this study, we discovered that zebrafish Etv5a inhibited the proliferation of neural progenitor cells in the embryonic neuroectoderm by directly binding to sox2 regulatory elements and inhibiting sox2 transcription. Further, Etv5a indirectly inhibits sox2 expression by inhibiting the expression of foxm1, a sox2 upstream gene. Our results reveal a hierarchy of transcriptional regulation in the proliferation of neural progenitor cells.

Etv5a and Etv5b exhibit high similarity in terms of their amino acid sequences; however, their redundant roles have not been fully determined. Simultaneous knock-down of Etv5a, Etv5b, and Etv4 (previously named Pea3) causes defects in cardiac progenitors and isthmic organizers, demonstrating the redundant role in regulating cell development [32].

However, they also perform non-redundant roles in several tissues. Etv5a regulates hemato-vascular derivatives [20] and nephron epithelial cells [21], whereas Etv5b governs hypothalamic-serotoninergic neurons [22], retina [33], and rhombomere segmentation [34,35]. Our findings demonstrated that only Etv5a (and not Etv5b) regulates sox2 expression and the formation of neural progenitors.

In our previous study, we found that Etv5a downregulation increases the proliferation of ventral mesoderm cells and causes defects in hemato-vascular derivatives [20]. Therefore, Etv5a may regulate the proliferation of many cell types with stemness or progenitor characteristics. Cancer stem cells share many characteristics with normal ones, such as high proliferation, self-renewal, and multipotency.

Therefore, Etv5a deficiency induced the proliferation of neural progenitor cells, suggesting that ETV5 may act in the same manner in cancer cells as in the nervous system. However, studies have demonstrated an oncogenic role of ETV5 in tumors of the nervous system. In neuroblastoma cell lines, anaplastic lymphoma kinase activation induces ETV5 expression, and ETV5 binds to the RET promoter to drive neuroblastoma oncogenesis [14].

Hence, ETV5 is required for proliferation, migration, and colony formation in neuroblastoma cells [15]. Moreover, ETV5 is essential for glioma growth in a mouse model with a constitutively active Ras pathway [36]. FOXM1 is also oncogenic in nervous system tumors. FOXM1 binds to SOX2 to promote stemness and radiation resistance in glioblastoma cells [29].

Further, FOXM1 induces SOX2 expression and inhibits neuroblastoma cell differentiation [37]. SOX2 is considered to be an oncogene in many tumors. In tumors of the nervous system, SOX2 is highly expressed in glioblastomas [38] and maintains the stemness of glioma stem cells [39] and medulloblastoma cells [40]. According to these studies, ETV5, SOX2, and FOXM1 are oncogenic in nervous system tumors. This conclusion is inconsistent with our observation that Etv5a inhibits the proliferation of neural progenitors by inhibiting sox2 expression and foxm1-sox2 signaling.

Most studies involving mammalian ETV5 have been performed in immortalized tumor cell cultures whose tissue complexity, heterogeneity, and molecular regulation could be very different from those of animal models, namely the zebrafish model used in the present study. Overall, these results indicate that the role of Etv5a in embryonic neural progenitors is different from that of ETV5 in tumor cells of the nervous system.

Although we discovered that Etv5a is upstream of the transcriptional hierarchy that regulates the proliferation of neural progenitor cells, the upstream regulator of Etv5a is unknown in this cell population. The major signaling pathways that regulate neural progenitor development are BMP, Wnt, FGF, SHH, and Notch. By searching the published literature listed in PubMed, we did not find evidence describing the regulation of BMP, Wnt, SHH, and Notch on ETV5/Etv5a.

In contrast, FGF signaling induces the expression of ETV5/etv5. In zebrafish embryos, Fgf3 and Fgf8 induce etv5b expression in various regions of the developing nervous system [41,42]. During neural induction, FGF inhibits BMP signaling in neuroectoderm specification. As Etv5a is expressed in the neuroectoderm and inhibits sox2 expression, it is required for inhibiting neural progenitor cells.

One possible explanation is that FGF-Etv5a signaling induces neural induction, temporarily inhibits sox2 expression, and downregulation of Etv5a expression relieves this inhibition and favors Sox2 function in the formation of neural progenitor cells. However, this hypothesis and the upstream signaling that regulates Etv5a in the neuroectoderm remain to be confirmed.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This work was supported by grants from Chang Gung Medical Foundation and Chang Gung Memorial Hospital (CMRPD1I0093, CMRPD1M0091, and BMRP857 for Y.-C.C.; CORPG3J0261, CORPG 3J0262, and CMRPG3H0322 for Y.-C.H.), the Ministry of Science and Technology, Taiwan (MOST 106-2311-B-182-001-MY3 and 110-2320-B-182-008-MY3 for Y.-C.C.; NMRPG3J6141 and NMRPG3J6142 for Y.-C.H.), and Taipei Medical University (DP2-107-21121-01-N-10, DP2-108-21121-01-N-10-01, TMU106-AE1-B20 for T.-H.Y.).

Supplementary Material

Supplementary Figure S1

Supplementary Figure S2

Supplementary Table S1

Supplementary Table S2

References

Takahashi

and Yamanaka

. (2006). Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell, 126:663–676.

Wang

, Oron

, Nelson

, Razis

and Ivanova

. (2012). Distinct lineage specification roles for NANOG, OCT4, and SOX2 in human embryonic stem cells. Cell Stem Cell, 10:440–454.

Boyer

, Lee

, Cole

, Johnstone

, Levine

, Zucker

, Guenther

, Kumar

, Murray

, et al. (2005). Core transcriptional regulatory circuitry in human embryonic stem cells. Cell, 122:947–956.

Zhang

and Cui

. (2014). Sox2, a key factor in the regulation of pluripotency and neural differentiation. World J Stem Cells, 6:305–311.

Imayoshi

and Kageyama

. (2014). bHLH factors in self-renewal, multipotency, and fate choice of neural progenitor cells. Neuron, 82:9–23.

Kageyama

, Shimojo

and Imayoshi

. (2015). Dynamic expression and roles of Hes factors in neural development. Cell Tissue Res, 359:125–133.

Kar

and Gutierrez-Hartmann

. (2013). Molecular mechanisms of ETS transcription factor-mediated tumorigenesis. Crit Rev Biochem Mol Biol, 48:522–543.

Sizemore

, Pitarresi

, Balakrishnan

and Ostrowski

. (2017). The ETS family of oncogenic transcription factors in solid tumours. Nat Rev Cancer, 17:337–351.

Hsing

, Wang

, Rennie

, Cox

and Cherkasov

. (2020). ETS transcription factors as emerging drug targets in cancer. Med Res Rev, 40:413–430.

10.

, Shin

and Janknecht

. (2012). ETV1, 4 and 5: an oncogenic subfamily of ETS transcription factors. Biochim Biophys Acta, 1826:1–12.

11.

Meng

, Li

, Ma

, Wu

, Fu

and Qin

. (2020). ETV5 overexpression contributes to tumor growth and progression of thyroid cancer through PIK3CA. Life Sci, 253:117693.

12.

Feng

, Liu

, Shen

, Liang

, Wang

, Qiu

, Cheng

and Zhao

. (2020). Targeting tumor cell-derived CCL2 as a strategy to overcome Bevacizumab resistance in ETV5(+) colorectal cancer. Cell Death Dis, 11:916.

13.

DeSalvo

, Ban

, Li

, Sun

, Jiang

, Kerr

, Khanlari

, Boulina

, Capecchi

, et al. (2021). ETV4 and ETV5 drive synovial sarcoma through cell cycle and DUX4 embryonic pathway control. J Clin Invest, 131:e141908.

14.

Lopez-Delisle

, Pierre-Eugene

, Louis-Brennetot

, Surdez

, Raynal

, Baulande

, Boeva

, Grossetete-Lalami

, Combaret

, et al. (2018). Activated ALK signals through the ERK-ETV5-RET pathway to drive neuroblastoma oncogenesis. Oncogene, 37:1417–1429.

15.

Mus

, Lambertz

, Claeys

, Kumps

, Van Loocke

, Van Neste

, Umapathy

, Vaapil

, Bartenhagen

, et al. (2020). The ETS transcription factor ETV5 is a target of activated ALK in neuroblastoma contributing to increased tumour aggressiveness. Sci Rep, 10:218.

16.

Akagi

, Kuure

, Uranishi

, Koide

, Costantini

and Yokota

. (2015). ETS-related transcription factors ETV4 and ETV5 are involved in proliferation and induction of differentiation-associated genes in embryonic stem (ES) cells. J Biol Chem, 290:22460–22473.

17.

Ahmad

, Rogers

, Chen

, Dixit

, Adnani

, Frankiw

, Lawn

, Blough

, Alshehri

, et al. (2019). Capicua regulates neural stem cell proliferation and lineage specification through control of Ets factors. Nat Commun, 10:2000.

18.

Hagedorn

, Paratore

, Brugnoli

, Baert

, Mercader

, Suter

and Sommer

. (2000). The Ets domain transcription factor Erm distinguishes rat satellite glia from Schwann cells and is regulated in satellite cells by neuregulin signaling. Dev Biol, 219:44–58.

19.

Liu

and Zhang

. (2019). ETV5 is essential for neuronal differentiation of human neural progenitor cells by repressing NEUROG2 expression. Stem Cell Rev Rep, 15:703–716.

20.

Chen

, Shih

, Lin

, Hsiao

, Li

and Cheng

. (2013). Etv5a regulates the proliferation of ventral mesoderm cells and the formation of hemato-vascular derivatives. J Cell Sci, 126:5626–5634.

21.

Marra

and Wingert

. (2016). Epithelial cell fate in the nephron tubule is mediated by the ETS transcription factors etv5a and etv4 during zebrafish kidney development. Dev Biol, 411:231–245.

22.

Bosco

, Bureau

, Affaticati

, Gaspar

, Bally-Cuif

and Lillesaar

. (2013). Development of hypothalamic serotoninergic neurons requires Fgf signalling via the ETS-domain transcription factor Etv5b. Development, 140:372–384.

23.

Kimmel

, Ballard

, Kimmel

, Ullmann

and Schilling

. (1995). Stages of embryonic development of the zebrafish. Dev Dyn, 203:253–310.

24.

Cheng

, Scotting

, Hsu

, Lin

, Shih

, Hsieh

, Wu

, Tsao

and Shen

. (2013). Zebrafish rgs4 is essential for motility and axonogenesis mediated by Akt signaling. Cell Mol Life Sci, 70:935–950.

25.

Lin

, Chiang

, Shih

, Hsu

, Yeh

, Huang

, Lin

and Cheng

. (2017). Regulator of G protein signaling 2 (Rgs2) regulates neural crest development through Ppardelta-Sox10 cascade. Biochim Biophys Acta Mol Cell Res, 1864:463–474.

26.

Okuda

, Yoda

, Uchikawa

, Furutani-Seiki

, Takeda

, Kondoh

and Kamachi

. (2006). Comparative genomic and expression analysis of group B1 sox genes in zebrafish indicates their diversification during vertebrate evolution. Dev Dyn, 235:811–825.

27.

Cooper

, Newman

, Aitkenhead

, Allerston

and Gileadi

. (2015). Structures of the Ets protein DNA-binding domains of transcription factors Etv1, Etv4, Etv5, and Fev: determinants of DNA binding and redox regulation by disulfide bond formation. J Biol Chem, 290:13692–13709.

28.

Ueno

, Nakajo

, Watanabe

, Isoda

and Sagata

. (2008). FoxM1-driven cell division is required for neuronal differentiation in early Xenopus embryos. Development, 135:2023–2030.

29.

Lee

, Kim

, Cho

, Kim

, Rheey

, Shin

, Seo

, Choi

, et al. (2015). FoxM1 promotes stemness and radio-resistance of glioblastoma by regulating the master stem cell regulator Sox2. PLoS One, 10:e0137703.

30.

Kim

, Choi

, Kim

, Lim

and Park

. (2013). C-terminus-deleted FoxM1 is expressed in cancer cell lines and induces chromosome instability. Carcinogenesis, 34:1907–1917.

31.

Munchberg

, Ober

and Steinbeisser

. (1999). Expression of the Ets transcription factors erm and pea3 in early zebrafish development. Mech Dev, 88:233–236.

32.

Znosko

, Yu

, Thomas

, Molina

, Li

, Tsang

, Dawid

, Moon

and Tsang

. (2010). Overlapping functions of Pea3 ETS transcription factors in FGF signaling during zebrafish development. Dev Biol, 342:11–25.

33.

Vinothkumar

, Rastegar

, Takamiya

, Ertzer

and Strahle

. (2008). Sequential and cooperative action of Fgfs and Shh in the zebrafish retina. Dev Biol, 314:200–214.

34.

Tambalo

, Mitter

and Wilkinson

. (2020). A single cell transcriptome atlas of the developing zebrafish hindbrain. Development, 147:dev184143.

35.

Walshe

, Maroon

, McGonnell

, Dickson

and Mason

. (2002). Establishment of hindbrain segmental identity requires signaling by FGF3 and FGF8. Curr Biol, 12:1117–1123.

36.

Breunig

, Levy

, Antonuk

, Molina

, Dutra-Clarke

, Park

, Akhtar

, Kim

, Town

, et al. (2016). Ets factors regulate neural stem cell depletion and gliogenesis in Ras pathway glioma. Cell Rep, 17:3407.

37.

Wang

, Park

, Carr

, Chen

, Zheng

, Li

, Tyner

, Costa

, Bagchi

, et al. (2011). FoxM1 in tumorigenicity of the neuroblastoma cells and renewal of the neural progenitors. Cancer Res, 71:4292–4302.

38.

de la Rocha

, Sampron

, Alonso

and Matheu

. (2014). Role of SOX family of transcription factors in central nervous system tumors. Am J Cancer Res, 4:312–324.

39.

Gangemi

, Griffero

, Marubbi

, Perera

, Capra

, Malatesta

, Ravetti

, Zona

, Daga

, et al. (2009). SOX2 silencing in glioblastoma tumor-initiating cells causes stop of proliferation and loss of tumorigenicity. Stem Cells, 27:40–48.

40.

Vanner

, Remke

, Gallo

, Selvadurai

, Coutinho

, Lee

, Kushida

, Head

, Morrissy

, et al. (2014). Quiescent sox2(+) cells drive hierarchical growth and relapse in sonic hedgehog subgroup medulloblastoma. Cancer Cell, 26:33–47.

41.

Raible

and Brand

. (2001). Tight transcriptional control of the ETS domain factors Erm and Pea3 by Fgf signaling during early zebrafish development. Mech Dev, 107:105–117.

42.

Roehl

and Nusslein-Volhard

. (2001). Zebrafish pea3 and erm are general targets of FGF8 signaling. Curr Biol, 11:503–507.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.19 MB

0.22 MB

0.25 MB

0.11 MB