The CRHR1/CREB/REST signaling cascade regulates mammalian embryonic neural stem cell properties

Abstract

Growing evidence suggests that the corticotropin‐releasing hormone (CRH) signaling pathway, mainly known as a critical initiator of humoral stress responses, has a role in normal neuronal physiology. However, despite the evidence of CRH receptor (CRHR) expression in the embryonic ventricular zone, the exact functions of CRH signaling in embryonic brain development have not yet been fully determined. In this study, we show that CRHR1 is required for the maintenance of neural stem cell properties, as assessed by in vitro neurosphere assays and cell distribution in the embryonic cortical layers following in utero electroporation. Identifying the underlying molecular mechanisms of CRHR1 action, we find that CRHR1 functions are accomplished through the increasing expression of the master transcription factor REST. Furthermore, luciferase reporter and chromatin immunoprecipitation assays reveal that CRHR1‐induced CREB activity is responsible for increased REST expression at the transcriptional level. Taken together, these findings indicate that the CRHR1/CREB/REST signaling cascade plays an important role downstream of CRH in the regulation of neural stem cells during embryonic brain development.

In addition to acting in stress‐related responses in the adult brain, the corticotropin‐releasing hormone receptor 1 (CRHR1) signaling pathway regulates brain development by modulating neural stem cell properties.

Introduction

Brain development is orchestrated by precise spatiotemporal regulation of neural stem cell proliferation and cell fate specification, for proper neural cell production, laminar organization, and size of a mature brain (Rakic, 1995; Breunig et al, 2011). This regulation is achieved by diverse cellular signaling pathways triggered by various cues such as secretory factors diffused from the microenvironment and their cell surface receptors (Yoon & Gaiano, 2005; Kohwi & Doe, 2013; Kim et al, 2018).

Corticotropin‐releasing hormone (CRH) is a neuropeptide produced in the paraventricular nucleus of the hypothalamus in response to stress or threat and stimulates the secretion of adrenocorticotropin (ACTH) from the anterior pituitary and cortisol from the adrenal gland (Vale et al, 1981). This activation of the hypothalamic–pituitary–adrenal (HPA) axis is the central physiological system that mediates stress‐induced behaviors in the adult stages (De Kloet et al, 2005). In humans, alterations in CRH expression and signaling are associated with major depression, psychiatric disorders, and compromised hippocampal function (Sautter et al, 2003; Laryea et al, 2012; Grimm et al, 2015; Jokinen et al, 2018), presumably resulting from a misregulated HPA axis. In addition to stress‐related responses at the organism level, CRH has been known to be involved in neuronal cellular events such as neurite elongation and synaptogenesis (Garcia et al, 2014; Inda et al, 2017). There are two CRH receptors, CRHR1 and CRHR2, and although CRH receptors are expressed in the dentate gyrus of the hippocampus (Brunson et al, 2002; Chen et al, 2004), which is one of the important neural stem niches in the adult brain, few research groups have been interested in the role of CRH signaling in terms of neural stem cell regulation. In this regard, reports showing that CRH is required for adult hippocampal neural stem cell homeostasis (Koutmani et al, 2019) and that CRH has a neuroprotective effect on embryonic neural stem cells (Koutmani et al, 2013) are noteworthy.

In this study, we show that CRHR1 enhances neural stem cell properties during brain development. Furthermore, among the diverse intracellular effectors downstream of CRHR1, we show that the cAMP response element (CRE)‐binding protein (CREB)/RE1 silencing transcription factor (REST) cascade is a direct target of CRHR1, providing a detailed molecular mechanism by which CRHR1 signaling regulates embryonic neural stem cells and thereby brain development.

Results and Discussion

CRH signaling enhances embryonic neural stem cell characteristics in vitro

Based on reports showing the embryonic expression of CRH and its receptors (Keegan et al, 1994; Chen et al, 2004; Koutmani et al, 2013), we attempted to elucidate the role of the CRH signaling pathway in embryonic neural stem cells. First, we examined the effects of the CRH ligand by the neurosphere assay. Neural progenitor cells isolated from the brain tissues of embryonic day 14.5 (E14.5) mouse embryos were cultured in defined neural stem cell media containing 1 μM CRH. CRH addition caused an increase in the number of neurospheres up to 2‐fold (Fig 1A) compared with the control, and antalarmin, a selective CRHR1 receptor antagonist (Habib et al, 2000), efficiently blocked CRH‐facilitated neurosphere formation, suggesting that CRH signaling enhances neural stemness. Consistently, expression of a neural stem cell marker, SOX2, was greatly increased in neural progenitor cells treated with CRH (Fig 1B).

Figure 1. CRH signaling enhances embryonic neural stem cell properties.

• A
  Neurosphere assays were performed using mouse E14.5 primary neural progenitors treated with CRH with or without antalarmin (Ant) (n = 3 biological replicates).
• B
  Effects of CRH treatment on the stemness of neural progenitor cells at 2 days post‐treatment were analyzed by Western blotting using an anti‐SOX2 antibody (n = 3 biological replicates).
• C, D
  Effects of CRHR1 overexpression on the (C) neurosphere formation and (D) SOX2 expression (n = 3 biological replicates).
• E
  CRHR1 knockdown efficiency of shRNAs used in this study (n = 4 biological replicates).
• F–I
  Effects of CRHR1 knockdown on the (F) neurosphere formation, (G) SOX2 expression, (H) neurosphere morphology, and (I) βIII‐tubulin expression. n = 3 biological replicates for each experiment except for (G) (n = 5 biological replicates).
• J, K
  Immunostaining using anti‐GFP (green) together with (J) anti‐βIII‐tubulin (clone TuJ1; red) (n = 5 biological replicates) or (K) anti‐GFAP (red) antibodies (n = 3 biological replicates) after in vitro differentiation of E14.5 neural progenitor cells transduced with a lentiviral vector expressing shCRHR1.

Data information: Scale bars, 25 μm for (H), 50 μm for (A, C, F), and 100 μm for (J, K). Error bars represent SEM. Statistical significance was determined by the unpaired Student's t‐test (B–D, J, K) and ordinary one‐way ANOVA (A, E–I). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns, not significant.

Source data are available online for this figure.

To examine CRH signaling at the cell membrane, among the two CRH receptors, we selected CRHR1 because it is expressed in both the embryonic and adult neural stem cell niches (Brunson et al, 2002; Koutmani et al, 2013) and treatment of antalarmin alone sufficiently reduced CRH‐enhanced neurosphere formation. For CRHR1 overexpression, we used a lentiviral vector pWPI that contains an internal ribosome entry sequence (IRES), enabling dual expression of the CRHR1 and GFP reporter genes. Lentiviral CRHR1 transduction of neural progenitor cells phenocopied the effects of CRH treatment in vitro; CRHR1 expression increased neurosphere formation rate (Fig 1C) and SOX2 expression (Fig 1D).

In parallel to the gain‐of‐function experiments, we used a loss‐of‐function approach employing RNAi‐mediated knockdown of CRHR1 expression. For this, a lentiviral vector pLKO.3G was used to drive the expression of short hairpin RNA (shRNA) against CRHR1 (shCRHR1) and GFP. Two shCRHR1 vectors were constructed with different target sequences of CRHR1 to rule out the possibility of off‐target effects associated with shRNAs, and both shRNA vectors showed approximately 50% of CRHR1 knockdown efficiency at the mRNA level (Fig 1E). As shown in Fig 1F and G, downregulation of CRHR1 expression reduced the number of neurospheres and SOX2 expression in vitro. Because it was previously reported that the CRH signaling pathway affects apoptosis (Koutmani et al, 2013), we tested whether CRHR1‐regulated neurosphere generation and SOX2 expression were due to the modulation of cell survival/apoptosis pathways. We found that activation of apoptosis, as assessed by staurosporine‐induced caspase‐3 cleavage, was not significantly different between CRHR1 and control groups (Fig EV1A and B).

Figure EV1. CRHR1 does not alter the cell survival of neural progenitor cells.

• A, B
  E14.5 neural progenitor cells were transduced with a lentiviral vector expressing (A) CRHR1 or (C) shCRHR1. At 2 days post‐transduction, cells were treated with 2 μM staurosporine for 8 h to induce cell death and harvested for Western blotting for cleaved caspase‐3.

Data information: n = 3 biological replicates. Error bars represent the SEM. The unpaired Student's t‐test was used to determine statistical significance. ns, not significant.

Source data are available online for this figure.

Of note, during long‐term culture of more than 7 days, most neurospheres transduced with the shCRHR1 vectors started to establish contact with the plate and spread neurite‐like extensions (Fig 1H), implying accelerated differentiation even under nonadherent and proliferation conditions. Indeed, Western blot analysis revealed that expression of the neuronal marker βIII‐tubulin was greatly increased in the CRHR1‐knockdown neurospheres (Fig 1I). A subsequent in vitro differentiation assay also substantiated that βIII‐tubulin⁺ cell generation was enhanced in the shCRHR1‐transduced neural progenitor cells (Fig 1J), whereas GFAP⁺ cell production was comparable between the two groups (Fig 1K). Taken together, these results suggest that CRH/CRHR1 signaling enhances neural stemness and inhibits neurogenesis in vitro.

Our data showing that inhibition of CRHR1 expression by shRNA‐mediated gene silencing impaired neural stem cell properties was consistent with the results of CRHR1‐specific pharmacological inhibition by antalarmin. These results also suggested that CRHR2 has a negligible role in the embryonic neural stem cells. Similar to our findings, CRHR1, but not CRHR2, was reported to play a crucial role in the activation of the HPA axis (Müller et al, 2001), the most well‐known effect of CRH in the adult brain. In addition, the knockout of CRHR1 alone was sufficient to yield modulated dendritic differentiation (Chen et al, 2004). These observations may be due to different tissue distribution patterns and binding ability to CRH; unlike CRHR1, CRHR2 expression dominates in the periphery, and CRH has a 4‐ to 20‐fold greater binding affinity for CRHR1 than for CRHR2 (Hauger et al, 2006). Furthermore, the counteraction of CRHR2 against CRHR1 signaling has also been hypothesized (Hauger et al, 2006).

CRHR1 expression positively regulates neural stem cell properties in vivo

Next, we examined the effects of CRHR1 expression in the developing brain. Providing important confirmation of the in vitro results, a greatly higher fraction of cells overexpressing CRHR1 was detected in the ventricular/subventricular zone (VZ/SVZ), the embryonic neural stem cell region, at 2 days post‐in utero electroporation (Fig 2A and B), suggesting attenuated differentiation by CRHR1. Indeed, a higher number of CRHR1‐overexpressing cells were co‐labeled with SOX2 in vivo (Fig 2A and C). Consistently, downregulation of CRHR1 expression by shRNA vectors resulted in increased cell exit from the VZ/SVZ, and co‐expression of the shRNA‐resistant human CRHR1 gene restored cell distribution patterns to control levels, ruling out the possibility of off‐target effects of CRHR1 RNAi (Fig 2D).

Figure 2. CRHR1 expression enhances neural stem cell properties in the developing brain.

• A
  Embryonic brains that were intraventricularly electroporated CRHR1‐expressing plasmids at E13.5 were harvested at E15.5 and immunostained using anti‐GFP (green) and SOX2 (red) antibodies.
• B
  Quantification of the cell fraction in each laminar layer in (A) (n = 3 biological replicates).
• C
  Percentage of SOX2⁺ cells among GFP⁺ cells in (A) (n = 4 biological replicates).
• D
  Embryonic brains were electroporated with shCRHR1 plasmids with or without shCRHR1‐resistant CRHR1 expression vectors at E14.5, and samples were harvested at E17.5 for anti‐GFP immunolabeling. GFP⁺ cell localization was quantified (bottom panel) (n = 5 biological replicates).
• E, F
  Brains electroporated with CRHR1 expression vectors at E13.5 were harvested at E15.5 and co‐immunostained using anti‐GFP (green) together with (E) anti‐TBR2 (red) or (F) anti‐CTIP2 (red) antibodies (n = 3 biological replicates).
• G
  E14.5 Embryonic brains that were electroporated with CRHR1‐expression vectors at E13.5 were pulse‐labeled with EdU and harvested at E15.5. Samples were then visualized for GFP (green), EdU (blue), and Ki67 (red) (n = 4 biological replicates).

Data information: Error bars represent SEM. Two‐way ANOVA with the Sidak's multiple comparisons test (B), the unpaired Student's t‐test (C, E–G), and two‐way ANOVA with the Tukey's multiple comparisons test (D) were used to determine statistical significance. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.001, ns, not significant. Scale bars, 20 μm. VZ, ventricular zone; SVZ, subventricular zone; IZ, intermediate zone; CP, cortical plate. The upper and lower dashed lines in (A, D–G) indicate the boundaries between the CP and IZ, and IZ and VZ/SVZ, respectively.

To further investigate the molecular nature of the CRHR1‐overexpressing cells in vivo, we performed immunofluorescence analysis for other cell‐type‐specific markers, TBR2 and CTIP2, which label intermediate progenitor cells and early cortical neurons, respectively (Englund et al, 2005; Edri et al, 2015). We found that TBR2⁺ cells were increased by up to 2‐fold (Fig 2E), while CTIP2⁺ cells were greatly reduced in the CRHR1‐electroporated samples (Fig 2F), confirming the hindrance of neuronal differentiation by CRHR1 overexpression in vivo.

Based on our in vitro results showing CRHR1‐promoted stemness of neural progenitor cells and the in vivo cell‐type analysis data, we postulated that the relatively higher number of CRHR1‐electroporated cells in the VZ/SVZ is due to delayed cell cycle exit of neural progenitor cells. To examine this possibility, we performed in utero electroporation at E13.5 and labeled proliferating cells by a pulse of 5‐ethynyl‐2′‐deoxyuridine (EdU) at 24 h postelectroporation. Brains were collected at E15.5 and analyzed for EdU and Ki67 (markers of cell proliferation). EdU⁺ and Ki67⁺ double‐positive cells are progenitors that were actively undergoing the cell cycle at the time of sample collection, and cells labeled by EdU, but not Ki67, are the cells that have exited the cell cycle. As shown in Fig 2G, CRHR1‐overexpressing cells revealed a decrease in cell cycle exit, which can account for the increased cell localization of CRHR1‐overexpressing cells in the VZ/SVZ. Taken together, these results indicate that CRHR1 signaling enhances neural stem cell characteristics in the in vivo context.

REST, a master regulator of neurogenesis, is the transcriptional target of CRHR1 signaling

To identify intracellular signaling components mediating CRHR1 function, we performed real‐time reverse transcription quantitative PCR (qPCR) analysis using E14.5 neural progenitor cells transduced with CRHR1 shRNA and searched for diverse neural stemness/differentiation‐related genes whose altered expression levels were consistent with the effects of CRHR1 knockdown. The most likely candidate was a transcription factor since cell fate‐determining signals eventually lead to changes in the levels of a specific lineage‐instructive transcription factor in the nucleus. Fig 3A displays an example of qPCR analysis, showing that expression levels of most transcription factors tested were not significantly changed (NEUROG2, NEUROD2, and NEX1) or could not explain the effect of CRHR1 knockdown (NEUROG1). One transcription factor, RE1 silencing transcription factor (REST), stood out since it had been reported to be a positive regulator for the quiescence of neural stem cells and a negative factor for neuronal differentiation (Su et al, 2004; Urbán et al, 2019), and reduced REST expression levels were consistent with the effects of CRHR1 knockdown. Reduced REST expression by the CRHR1 shRNA vector was confirmed by Western blot analysis (Fig 3B), and consistently, increased REST expression by CRHR1 overexpression was also verified by qPCR and Western blot analyses (Fig 3C and D), indicating that the expression levels of REST correlate with CRHR1 expression. Then, we performed in utero electroporation to address whether REST is the key determinant for CRHR1‐enhanced neural stem cell characteristics in vivo. Co‐expression of REST shRNA efficiently disrupted CRHR1‐increased cell localization in the VZ/SVZ (Fig 3E and F) and SOX2⁺ cell number (Fig 3E and G). Furthermore, the cell positional changes upon CRHR1 knockdown were reverted toward the VZ/SVZ upon co‐expression of REST (Fig 3H and I). Taken together, these data strongly indicate that CRHR1 actions in neural stem cell regulation are accomplished through increasing REST expression.

Figure 3. REST is a key transcriptional mediator of CRHR1 action in the embryonic neural stem cells.

• A, B
  Analysis of the effects of CRHR1 knockdown on neural stemness‐ and cell fate‐regulating transcription factors in neural progenitor cells by (A) qPCR (n = 4 biological replicates) and (B) Western blotting (n = 3 biological replicates) at 2 days post‐transduction.
• C, D
  The effects of CRHR1 overexpression on REST expression in neural progenitor cells analyzed by (A) qPCR (n = 4 biological replicates) and (B) Western blotting (n = 6 biological replicates).
• E
  Double‐immunolabeling of E15.5 brain sections electroporated in utero with CRHR1‐expressing plasmid with or without shREST vectors at E13.5 using anti‐GFP (green) and SOX2 (red) antibodies.
• F, G
  Quantification of (F) GFP⁺ cell distribution (n ≥ 4 biological replicates) and (G) percentage of SOX2⁺ cells (n ≥ 3 biological replicates) in (E).
• H
  Embryonic brains were electroporated with shCRHR1 plasmids with or without REST expression vectors at E14.5, and brain samples were harvested at E17.5 for anti‐GFP immunolabeling.
• I
  Quantification of GFP⁺ cell localization in (H) (n ≥ 4 biological replicates).

Data information: Error bars represent SEM. Two‐way ANOVA with the Sidak's multiple comparisons test (A), the unpaired Student's t‐test (B–D), two‐way ANOVA with the Tukey's multiple comparisons test (F, I), and ordinary one‐way ANOVA (G) were used to determine statistical significance. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns, not significant. Scale bars, 20 μm. The upper and lower dashed lines in (E, H) indicate the boundaries between the CP and IZ, and IZ and VZ/SVZ, respectively.

Source data are available online for this figure.

CRHR1‐increased CREB activity is responsible for the enhancement of neural stem cell characteristics

Our next question was to identify the factor acting between CRHR1 and REST. CRHR1 is known to be involved in the activation of several intracellular signaling networks including PKA‐CREB, MEK‐ERK, or PLC‐CaMKII (Bayatti et al, 2005; Mo et al, 2015; Inda et al, 2017). After an extensive literature search, we obtained an important clue from Kreisler et al (2010) showing that a partial CRE site matching 5 out of 8 consensus CRE nucleotide sequences (TGACGTCA) exists in the human REST promoter. In addition, simultaneous upregulation of CREB phosphorylation and REST expression in kainic acid‐treated astrocytes was observed by Zhang et al (2022), but the determination of the causal relationship between CREB and REST was not their subject of interest. Based on these previous studies, we hypothesized that CREB could be the functional mediator that links CRHR1 and REST. First, we examined whether CREB activity was modulated by CRHR1 in our experimental settings and found that the level of CREB phosphorylation positively correlated with that of CRHR1 expression: CRHR1 overexpression in E14.5 neural progenitor cells increased CREB phosphorylation (Fig 4A), and CRHR1 knockdown produced the opposite results (Fig 4B). Consistent with the previous studies identifying PKA as a downstream regulator of CRHR1 (Inda et al, 2017; Hu et al, 2020) and an upstream kinase phosphorylating CREB (Blanchet et al, 2015), treatment of the PKA inhibitor H89 abolished CRHR1‐induced CREB phosphorylation whereas the CaM kinase inhibitor KN‐93 did not, indicating that the CRHR1‐induced CREB phosphorylation is PKA‐dependent (Fig 4C). In vivo analysis also revealed that phosphorylated CREB was immunostained in CRHR1‐overexpressing cells in the VZ/SVZ (Fig 4D). The importance of CREB activity in CRHR1‐induced neural stemness was further verified in vivo. After translocation into the nucleus, CREB cooperatively assembles on DNA to form homodimers (Wu et al, 1998), and phosphorylation of serine 133 of CREB is required for the recruitment of the transcriptional co‐activator paralogues CBP and p300 (Mayr & Montminy, 2001) and thus for target gene expression. However, the nonphosphorylatable CREB S133A, a widely used dominant‐negative form of CREB (Gonzalez et al, 1989; Rammes et al, 2000), still retains the ability to form dimers with wild‐type CREB (Huggins et al, 2007), which is presumed to be how CREB S133A can act as an inhibitor of CREB function. As shown in Fig 4E–G, CRHR1‐increased cell localization in the VZ/SVZ and SOX2⁺ cell numbers were completely blocked by the co‐expression of CREB S133A. Also, CREB overexpression effectively prevented CRHR1 shRNA‐expressing cells from exiting the VZ/SVZ (Fig 4H and I), confirming the importance of CREB as a downstream effector of CRHR1 signaling. Due to the nature of rescue experiments using plasmid overexpression, we cannot rule out the possibility that levels of pCREB in (shCRHR1 + CREB) samples would be higher than those in the control due to CREB overexpression, and this might have made cells stay in the VZ/SVZ independent of CRHR1 signaling. However, the complementary findings in Fig 4E–G address this concern considerably.

Figure 4. CREB activity is upregulated by CRHR1 expression.

• A, B
  Western blot analysis of E14.5 neural progenitor cells transduced with a lentiviral vector expressing (A) CRHR1 or (B) shCRHR1 using an anti‐CREB and anti‐phospho‐CREB (pCREB) (S133) antibodies (n = 5 biological replicates).
• C
  E14.5 neural progenitor cells were transduced with CRHR1‐containing lentiviral vectors. At 2 days post‐transduction, cells were treated with 10 μM H89 (PKA inhibitor) or 1 μM KN93 (CaM kinase inhibitor) for 8 h and harvested for Western blot analysis for CREB (top panel). Band intensities were quantified (bottom panel) (n = 3 biological replicates).
• D
  Double‐immunolabeling of E15.5 brain sections electroporated in utero with CRHR1‐expressing plasmid at E13.5 using anti‐GFP (green) and pCREB (S133) (red) as primary antibodies. Immunostaining was quantified (right panel) (n = 5 biological replicates). White inset boxes were magnified and shown in the panels D′ and D″.
• E
  Double‐immunolabeling of E15.5 brain sections electroporated in utero with CRHR1‐expressing plasmid with or without a vector expressing CREB S133A, a dominant‐negative mutant form of CREB, at E13.5 using anti‐GFP (green) and SOX2 (red) antibodies.
• F, G
  Quantification of (F) GFP⁺ cell distribution (n = 4 biological replicates) and (G) percentage of SOX2⁺ cells (n ≥ 3 biological replicates) in (E).
• H
  Embryonic brains were electroporated with shCRHR1 plasmids with or without CREB expression vectors at E14.5, and then, brain samples were harvested at E17.5 for anti‐GFP immunolabeling.
• I
  Quantification of GFP⁺ cell localization in (L) (n ≥ 4 biological replicates).

Data information: Error bars represent SEM. The unpaired Student's t‐test (A, B, D), two‐way ANOVA with the Tukey's multiple comparisons test (F, I), and ordinary one‐way ANOVA (C, G) were used to determine statistical significance. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns, not significant. Scale bars, 20 μm for (D, E, H), 10 μm for (D′ and D″). The upper and lower dashed lines in (D, E, H) indicate the boundaries between the CP and IZ, and IZ and VZ/SVZ, respectively.

Source data are available online for this figure.

CREB is the direct upstream factor of REST expression in the embryonic neural stem cells

Finally, we tested CREB as an upstream factor of REST expression. CREB overexpression in the E14.5 neural progenitor cells caused a large increase in the amount of REST at the levels of both mRNA (Fig 5A) and protein (Fig 5B). To confirm these effects, we employed a pharmacological approach using a selective CREB inhibitor, 666‐15 (Xie et al, 2015). We found that 666‐15 efficiently reduced endogenous REST mRNA and proteins in a dose‐dependent manner (Fig 5C and D) and chose 20 nM 666‐15 for further experiments. As expected, CRHR1‐induced REST expression was also effectively blocked by 666‐15 (Fig 5E and F). Considering that 666‐15 is used at the micromolar scale even for primary cells including mouse embryonic fibroblasts or microglial cells (Guan et al, 2019; Wan et al, 2021), this result shows that 666‐15 has an extremely potent effect on primary mouse neural progenitor cells.

Figure 5. REST is a direct transcriptional target of CREB in embryonic neural stem cells.

• A, B
  (A) qPCR (n = 5 biological replicates) and (B) Western blot analyses (n = 3 biological replicates) of E14.5 neural progenitor cells transduced with CREB‐expressing lentiviral vectors, measuring REST mRNA and protein levels, respectively.
• C, D
  Dose‐dependent inhibition of endogenous REST expression by a CREB inhibitor, 666‐15. E14.5 neural progenitor cells were treated with the indicated concentrations of 666‐15 for 48 h and subjected to (C) qPCR (n = 3 biological replicates) and (D) Western blot analyses (n = 4 biological replicates).
• E, F
  Inhibition of CRHR1‐induced REST expression by 666‐15. E14.5 neural progenitor cells transduced with a lentiviral vector expressing CRHR1 were treated with 20 nM of 666‐15, and after 48 h, cells were lysed and subjected to (E) qPCR (n = 3 biological replicates) and (F) Western blot analyses (n = 4 biological replicates).
• G
  Identification of a CRE sequence within the mouse REST promoter region. Consensus CRE sequences, the heptamer GACGTCA, were identified within the mouse REST promoter at position −971, and a region from −1,500 to +172 of the REST genomic sequence was cloned into the luciferase plasmid pGL3‐Basic. The function of this site was destroyed by mutating CG to AT.
• H
  Luciferase activity was measured using E14.5 neural progenitor cells transfected with CREB expression vectors and pGL3‐Basic luciferase plasmids harboring the wild‐type or mutated REST promoter at 2 days post‐transfection (n = 3 biological replicates).
• I
  A representative gel image of ChIP analysis showing the association of CREB proteins with regions of the REST promoter by PCR amplification of putative CRE sequences in the REST promoter (n = 3 biological replicates).
• J
  Immunolabeling of E15.5 brain sections electroporated in utero with CREB‐expressing plasmids with or without shREST vectors at E13.5 using anti‐GFP antibody. GFP⁺ cell localization was quantified (right panel) (n ≥ 5 biological replicates). The upper and lower dashed lines indicate the boundaries between the CP and IZ, and IZ and VZ/SVZ, respectively. Scale bar, 20 μm.

Data information: Error bars represent SEM. The unpaired Student's t‐test (A, B, I), ordinary one‐way ANOVA (C–F), two‐way ANOVA with the Sidak's multiple comparisons test (H), and two‐way ANOVA with the Tukey's multiple comparisons test (J) were used to determine statistical significance. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns, not significant.

Source data are available online for this figure.

Since a previous study reporting transactivation of the REST promoter by CREB was performed using human cell lines such as HEK293 cells (Kreisler et al, 2010), we needed to test the responsiveness of the mouse REST promoter to CREB due to species‐specificity. We obtained the genomic sequence of the mouse REST gene (Ensembl gene ID, ENSMUSG00000029249) and searched for possible CREB binding sites within 2 kb upstream of the transcriptional start site. One candidate sequence matching 7 out of 8 consensus CRE nucleotide sequences centered at position −971 was identified, where +1 represents the transcription initiation site (Fig 5G). For functional analysis of the promoter sequences containing the putative CRE, we amplified and cloned a region positioned from −1,500 to +172 of the REST gene into the luciferase plasmid pGL3‐Basic. We found that mouse REST promoter activity was significantly increased in CREB‐transduced neural progenitor cells and that mutations of CG to AT in the center of the putative CRE completely disrupted REST promoter activation by CREB (Fig 5H). Consistent with the luciferase promoter assays, chromatin immunoprecipitation (ChIP) experiments using chromatin extracts derived from mouse neuroblastoma cells (Neuro‐2a) transfected with Myc‐tagged CREB or Myc‐tagged GFP expression plasmids revealed that CREB is recruited to the mouse genomic REST promoter region (Fig 5I). Taken together, these results show that REST expression is directly regulated by CREB at the promoter level. Furthermore, the importance of CREB‐induced REST expression was confirmed in vivo. Co‐electroporation of shRNA constructs against REST returned CREB‐induced changes in laminal cell distribution to control levels in the developing brain (Fig 5J). These data indicate that REST is a direct transcriptional target of CREB and that CREB‐induced REST expression is the important regulator of CRHR1‐induced neural stemness.

In this study, we focused on the role of the CRH signaling pathway during embryonic brain development by modulating CRHR1 expression levels. Particularly, in terms of neurogenesis, we observed that CRHR1 signaling inhibits neuronal differentiation and this warrants discussion due to apparent discrepancies with a previous report. In vitro analyses of Fig 1I and J showed increased βIII‐tubulin⁺ cell production by CRHR1 downregulation, and in vivo results revealed changes in SOX2⁺ cell number and cell localization toward VZ/SVZ by CRHR1 overexpression, clearly demonstrating that the CRH pathway negatively regulates neurogenesis. However, a previous study reported reduced DCX⁺ (an immature neuronal marker) cells in the adult hippocampus of CRH knockout mice (Koutmani et al, 2019). Different experimental conditions may account for the different conclusions: First of all, it is highly likely that CRH signaling has distinct roles in the regulation of neural stem cells depending on age (embryonic vs. adult stages) and brain region (cortex vs. hippocampus). Consistently, many cellular regulators are known to produce spatiotemporally different outcomes. In addition, the experimental techniques used were very different. The previous study used a conventional CRH knockout mouse lacking the CRH gene in all cells and tissues, not only in the brain (Muglia et al, 1995). They also targeted the soluble and diffusible CRH ligand. Thus, our approach employing CRHR1 gene delivery into the wild‐type background by in vitro lentiviral transduction and in utero electroporation has great advantages for determining the direct and cell‐autonomous role of the CRH signaling pathway. These techniques also allowed us to perform gain‐of‐function and loss‐of‐function experiments. Regarding the stemness of neural stem cells, both studies obtained similar results showing that CRH signaling positively regulates SOX2⁺ neural stem cells.

Our study raises the question of where embryonic cortical CRH comes from. The source of CRH can be either outside of or within the embryos. We first would like to emphasize that CRH is an endocrine hormone. It is known that the placenta synthesizes CRH and releases it into the bloodstream (Thomson, 2013), and also that placental CRH acts on the fetal brain (McLean & Smith, 2001; Vrachnis et al, 2012). These reports indicate that the embryonic cortex does not necessarily have to express CRH to stimulate cortical cells since extrinsic CRH proteins can freely diffuse within the embryonic brain in an endocrine fashion. In addition to the placenta, the embryonic forebrain and cerebellum also express CRH (Keegan et al, 1994; Bishop & King, 1999), and this intrinsic CRH could also act on many regions of the embryonic body including the embryonic cortex.

Here, we attempted to understand why CRHR1, the key receptor of one of the well‐known adult endocrine systems, exists in the embryonic stages, and provide evidence of a new function and novel intracellular circuit during embryonic brain development. Similar to the case of CRH, thyroid hormone, which is most commonly known to control body metabolism in the adult, is reported to be involved in the process of neuroectoderm maturation (Fernández et al, 2020). Likewise, erythropoietin (EPO) and EPO receptors, which are critical for erythroid cell production, were found to promote cell survival in the embryonic brain (Chen et al, 2007). Therefore, the identification of an unexpected role of CRH signaling at the embryonic stages provides yet another interesting example of the diverse role of adult hormone systems at different developmental stages and different tissues.

Materials and Methods

Plasmid construction

To construct expression vectors for CRHR1, CREB and REST, total RNA was isolated from primary mouse neural progenitor cells or human embryonic kidney 293T (HEK293T) cells using TRIzol reagent (Invitrogen). The isolated RNA (2 μg) and oligo (dT) primers (Promega) were utilized to synthesize single‐stranded cDNA using a reverse transcription kit (Promega) according to the manufacturer's instructions. One microliter of the cDNA product was amplified by PCR using each gene‐specific primer pair: human CRHR1‐F (5′‐GAATTCACGCGTGCCACCATGGGAGGGCACCCGCAGCTCCGTCTCGTC‐3′), human CRHR1‐R (5′‐GCGGCCGCGGATCCTCAAGCGTAATCTGGAACATCGTATGGGTAGACTGCTGTGGACTGCTTGATGCTGTG‐3′), mouse CREB1‐F (5′‐TCTGTATTTGTCTGAAAATAATGCCAGCAGCTCATGCAAC‐3′), mouse CREB1‐R (5′‐GAATTGGATCCGCGGGCCCATCACAGATCCTCTTCTGAGATGAGTTTTTGTTCATCTGATTTGTGGCAGTAAA‐3′), human REST‐F (5′‐AAGCAGCTCAAGGGCAGGAGGGTAAGCCTATCCCTAACCCTCTCCTCGGTCTCGATTCTACGTGAGGATCCAATTCCGCCCC‐3′), and human REST‐R (5′‐GGGGCGGATTGGATCCTCACGTAGAATCGAGACCGAGGAGAGGGTTAGGGATAGGCTTACCCTCCTGCCCTTGAGCTGCTT‐3′). Amplified sequences were cloned into the lentiviral vector pWPI (Addgene, #12254). For CRHR1 RNAi studies, two CRHR1 target sequences, #1 (5′‐GCAAAGTGCACTACCACAT‐3′) and #2 (5′‐GGGCCATTGGGAAACTTTACT‐3′), were cloned into the pLKO.3G vector (Addgene, #14748). A short hairpin RNA (shRNA) vector targeting mouse REST was purchased from Sigma‐Aldrich (TRCN0000321488). The CREB S133A mutant was generated by site‐directed mutagenesis using PrimeSTAR polymerase (Takara) and synthetic oligonucleotides: CREB S133A up (5′‐AAATCCTTTCAAGGAGGCCTGCCTACAGGAAAATTTTGAATGA‐3′) and CREB S133A down (5′‐TCATTCAAAATTTTCCTGTAGGCAGGCCTCCTTGAAAGGATTT‐3′). The REST promoter‐luciferase plasmid was generated by PCR amplification of the −1,500 to +172 region (relative to the transcription start site) of mouse REST genome sequence (Ensembl gene ID, ENSMUSG00000029249) using genomic DNA isolated from mouse neural progenitor cells. The amplified sequences were then cloned into the pGL3‐Basic vector (Promega). The putative CRE site mutation from GACGTCA to GAATTCA was created by site‐directed mutagenesis using synthetic oligonucleotides: CRE SA upper (5′‐CTGTCCCCATTATAAGAATTCAGGTACTGAACTGGAC‐3′), and CRE SA lower (5′‐GTCCAGTTCAGTACCTGAATTCTTATAATGGGGACAG‐3′).

Neural progenitor culture and in utero electroporation

All experimental procedures using animals were supervised by the Institutional Animal Care and Use Committee at Sungkyunkwan University (SKKUIACUC2021‐11‐01‐1). Timed pregnant CD1 mice (Koatech) were used for neural progenitor cell preparation and electroporation, and embryos were considered 0.5‐day‐old when a vaginal plug was detected in the morning. Primary neural progenitor cells were prepared from embryonic day 14.5 (E14.5) embryonic brains. Dissected brain tissue was minced, washed three times with PBS, and incubated in 0.25% trypsin (Gibco) at 37°C for 5 min. DNaseI and ovomucoid trypsin inhibitor (both from Worthington) were added, and samples were triturated using a fire‐polished Pasteur pipette. Cells were washed twice with DMEM/F12 media (Gibco), resuspended in PBS, and run through a 40 μm cell strainer (Corning). Isolated neural progenitors were cultured in DMEM/F12 media (Gibco) supplemented with 20 ng/ml FGF, B27 (Gibco), GlutaMAX (Gibco), 2 μg/ml heparin (Sigma‐Aldrich), and 1% penicillin and streptomycin. To modulate CRHR1 activity in vitro, 1 μM CRH (Sigma‐Aldrich, C3042) or 100 nM antalarmin (selective CRHR1 antagonist) (Sigma‐Aldrich, SKU A8727) was added to the stem cell medium at the beginning of the culture. The number of neurospheres was quantified after 7 days. For in vitro inhibition of CREB activity followed by qPCR of the REST gene, 20 nM 666‐15 (Sigma‐Aldrich, 5.38341) was incubated for 48 h, and for identification of the CREB upstream kinase, 10 μM H89 (a selective PKA inhibitor; Sigma‐Aldrich, B1427) and 1 μM KN93 (a CaMKII specific inhibitor; Sigma‐Aldrich, K1385) were treated for 8 h on neural progenitor cells. For the cell survival assay, neural progenitor cells were treated with 2 μM staurosporine (Sigma‐Aldrich, S5921) for 8 h to induce apoptosis in vitro.

Prior to injection, pregnant mice were anesthetized with pentobarbital sodium (Hanlim Pharm). For in utero electroporation, 1–2 μl of DNA solution (2 mg/ml) in PBS with 0.01% fast green dye (Merck) was injected into the lateral ventricle of E13.5 or E14.5 embryos. By using forceps‐type electrodes (Nepagene), DNA was electroporated to the embryos five times for 50 ms followed by a 950 ms gap. For the in vivo cell cycle exit assay, EdU (100 mg/kg body weight, Sigma‐Aldrich, 900584) was injected intraperitoneally into pregnant females at 24 h after in utero electroporation. Incorporation of EdU was visualized at 24 h postinjection using the Click‐iT™ EdU Imaging kit (Invitrogen, C10086) according to the manufacturer's instructions. Sections were further immunostained for GFP and Ki67 and analyzed using a confocal microscope (Zeiss, LSM 700).

Western blot

Cells were lysed using RIPA buffer (Merck) with a protease and phosphatase inhibitor cocktail (Pierce Biotechnology). Equal amounts of protein (20~40 μg, depending on target protein) were resolved in 8~10% (w/v) sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) and transferred to polyvinylidene fluoride (PVDF) membranes (Millipore). The membranes were blocked with TBST (150 mM NaCl, 10 mM Tris–HCl, 0.1% (v/v) Tween 20, pH 8.0) containing 5% (w/v) skim milk and analyzed with the following primary antibodies: anti‐SOX2 (Cell Signaling Technology, #23064S; 1:500), anti‐CREB (Cell Signaling Technology, #9197S; 1:500), anti‐phospho‐CREB (S133) (Cell Signaling Technology, #9198S; 1:500), anti‐REST (Bioss, bs‐2590R; 1:500), anti‐βIII‐tubulin (Covance, MMS‐435P; 1:1,000), anti‐Myc tag (Cell Signaling Technology, #2272S; 1:500), anti‐HA tag (Roche, 11867423001; 1:500), anti‐cleaved caspase‐3 (Cell Signaling Technology, #9661S; 1:500), and anti‐beta‐actin (Santa Cruz, sc‐47778; 1:5,000). All blots were incubated overnight with a primary antibody at 4°C, and then with horseradish peroxidase‐conjugated anti‐mouse (Santa Cruz, sc‐516102), anti‐rabbit (Santa Cruz, sc‐2357) or anti‐rat (Cell Signaling Technology, #7077S) secondary antibodies (1:10,000) for 2 h at room temperature. The protein bands were visualized with an enhanced chemiluminescence system (Cytiva) and X‐ray film (Agfa).

Real‐time reverse transcription quantitative PCR (qPCR)

Total RNA was prepared from mouse neural progenitor cells at 2 days post‐transduction using RNAiso Plus reagent (Takara) and cDNAs were synthesized from 500 ng of each RNA sample by using an oligo (dT) primer and M‐MLV reverse transcription enzyme (Toyobo). To search for CRHR1‐regulated genes, qPCR was performed according to the Smart Cycler System (Takara) protocol using the following primers: CRHR1 forward (5′‐GGGCAGCCCGTGTGAATTATT‐3′), CRHR1 reverse (5′‐ATGACGGCAATGTGGTAGTGC‐3′), NEUROG1 forward (5′‐CCAGCGACACTGAGTCCTG‐3′), NEUROG1 reverse (5′‐CGGGCCATAGGTGAAGTCTT‐3′), NEUROG2 forward (5′‐AACTCCACGTCCCCATACAG‐3′), NEUROG2 reverse (5′‐GAGGCGCATAACGATGCTTCT‐3′), NEUROD2 forward (5′‐AAGCCAGTGTCTCTTCGTGG‐3′), NEUROD2 reverse (5′‐GCCTTGGTCATCTTGCGTTT‐3′), NEX1 forward (5′‐TTAACACTACCGTTTGACGAGTC‐3′), NEX1 reverse (5′‐TGTTTTGGAAAGCTCTCTGGTT‐3′), REST forward (5′‐CATGGCCTTAACCAACGACAT‐3′), REST reverse (5′‐CGACCAGGTAATCGCAGCAG‐3′), β‐actin forward (5′‐CAAAAGCCACCCCCACTCCTAAGA‐3′), and β‐actin reverse (5′‐GCCCTGGCTGCCTCAACACCTC‐3′).

Immunofluorescence

Standard immunofluorescence procedures were conducted to detect target gene expression in electroporated brains. In brief, the gene‐transferred embryonic brains were fixed with 4% paraformaldehyde, dehydrated with 30% sucrose, and cryosectioned using a microtome (Leica, Germany, CM1850). Sections were washed in PBS, then blocked for 1 h with PBS containing 2% fetal bovine serum and 0.2% Triton X‐100. Samples were then incubated overnight at 4°C with the primary antibodies rabbit anti‐GFP (Invitrogen, A11122; 1:1,000), chicken anti‐GFP (Abcam, ab13970; 1:1,000), anti‐SOX2 (Cell Signaling Technology, #23064S; 1:500), and anti‐phospho‐CREB (S133) (Cell Signaling Technology, #9198S; 1:500), anti‐Ki67 (Abcam, ab15580; 1:1,000), anti‐TBR2 (Abcam, ab23345; 1:500), anti‐CTIP2 (Abcam, ab18465; 1:1,000), washed three times in PBS and incubated for 3 h at room temperature with Alexa Fluor 488‐ or 555‐conjugated secondary antibodies (Invitrogen) diluted in blocking solution. Images were analyzed using a confocal microscope (Zeiss, LSM 700).

Chromatin immunoprecipitation

Transfected Neuro‐2a cells were cross‐linked using 1% formaldehyde for 15 min. The samples were then washed three times in cold PBS and homogenized in ChIP cell lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris–HCl [pH 8.0], and protease inhibitor cocktail [Sigma]). After 10‐min incubation on ice, the lysates were sonicated (Cole‐Parmer) on ice using 14 pulses of 10 s each at 5 W and centrifuged for 10 min at 20,000 rcf at 4°C, and then, the supernatant was collected. An antibody against the Myc tag (Santa Cruz Biotechnology, sc‐40 X) was added, and the samples were rotated overnight at 4°C. Immunoprecipitates were isolated by incubating with protein G‐agarose (Invitrogen), and the beads were washed consecutively with low‐salt buffer (0.1% SDS, 1% Triton X‐100, 2 mM EDTA, 20 mM Tris–HCl [pH 8.0], 150 mM NaCl), high‐salt buffer (0.1% SDS, 1% Triton X‐100, 2 mM EDTA, 20 mM Tris–HCl [pH 8.0], 500 mM NaCl), LiCl buffer (0.25 M LiCl, 1% NP‐40, 1% deoxycholic acid, 1 mM EDTA, 20 mM Tris–HCl [pH 8.0]), and TE buffer (1 mM EDTA, 10 mM Tris–HCl [pH 8.0]). Chromatin was eluted and cross‐linking was reversed with 0.2 M NaCl. Samples were then digested with 10 μg of proteinase K (Sigma), and DNA was isolated via a DNA purification kit (GeneAll). Enrichment of the genomic REST promoter regions was determined by normalizing PCR product intensities to a negative control. PCR of the genomic REST promoter used the following primers: REST forward (5′‐AAGGCTACAAGGCCCACTTC‐3′) and REST reverse (5′‐GTCCTTGTGGGGGTTGAAC‐3′).

Statistical analysis

Statistical tests were performed using the Prism 8 software (GraphPad). Statistical analyses were performed using the unpaired two‐tailed Student's t‐test, one‐way ANOVA, and two‐way ANOVA. All data represent three or more independent experiments.

Author contributions

Mookwang Kwon: Conceptualization; data curation; formal analysis; validation; investigation; visualization; methodology; writing – original draft; writing – review and editing. Ju Hyun Lee: Formal analysis; validation; investigation; visualization. Youngik Yoon: Formal analysis; validation; investigation. Samuel J Pleasure: Writing – review and editing. Keejung Yoon: Conceptualization; resources; supervision; funding acquisition; methodology; writing – original draft; project administration; writing – review and editing.

Disclosure and competing interests statement

The authors declare that they have no conflict of interest.

Supporting information

Expanded View Figures PDF

Source Data for Expanded View

PDF+

Source Data for Figure 1

Source Data for Figure 3

Source Data for Figure 4

Source Data for Figure 5

EMBO reports (2023) 24: e55313

Data availability

No data were deposited in a public database.