Activation of Hes1 and Msx1 in Transgenic Mouse Embryonic Stem Cells Increases Differentiation into Neural Crest Derivatives

The neural crest (NC) comprises a multipotent cell population that produces peripheral neurons, cartilage, and smooth muscle cells, among other phenotypes. The participation of Hes1 and Msx1 when expressed in mouse embryonic stem cells (mESCs) undergoing NC differentiation is unexplored. In this work, we generated stable mESCs transfected with constructs encoding chimeric proteins in which the ligand binding domain of glucocorticoid receptor (GR), which is translocated to the nucleus by dexamethasone addition, is fused to either Hes1 (HGR) or Msx1 (MGR), as well as double-transgenic cells (HGR+MGR). These lines continued to express pluripotency markers. Upon NC differentiation, all lines exhibited significantly decreased Sox2 expression and upregulated Sox9, Snai1, and Msx1 expression, indicating NC commitment. Dexamethasone was added to induce nuclear translocation of the chimeric proteins. We found that Collagen IIa transcripts were increased in MGR cells, whereas coactivation of HGR+MGR caused a significant increase in Smooth muscle actin (α-Sma) transcripts. Immunostaining showed that activation in HGR+MGR cells induced higher proportions of β-TUBULIN III+, α-SMA+ and COL2A1+ cells. These findings indicate that nuclear translocation of MSX-1, alone or in combination with HES-1, produce chondrocyte-like cells, and simultaneous activation of HES-1 and MSX-1 increases the generation of smooth muscle and neuronal cells.


Introduction
The neural crest (NC) comprises a transient multipotent cell population that emerges between the forming neural plate and nonneural ectoderm. The NC is a unique and highly specialized group of cells found in all vertebrates. The NC delaminates from the dorsal neural tube following its closure to

Expression of Hes1 and Msx1 in Wild-type ES Cells in Pluripotent Conditions and after NC Differentiation
To analyze the role of Hes1 and Msx1 in the differentiation of mESCs into neural crest cells, we used the stromal cell-inducing activity of Pre-adipose 6 (PA6) cells for 5 days [17], followed by the addition of BMP4, which commits cells to differentiate into NC derivatives [6]. Cultures were treated from day 5 to day 8 with 0.5 nM BMP4. At day 8, the cells were harvested by Papain treatment and plated on Fibronectin-coated dishes in neural differentiation medium with chick embryo extract without BMP4, a condition reported to favor differentiation into smooth muscle cells (α-SMA + ) and neural derivatives (β-TUBULIN III + ) [6] (Figure 1). Although we did not separate PA6 cells after replating, these cells did not proliferate, and their contribution to the measured parameters was constant for all conditions. At day 8, some cultures were dissociated and replated in micromass conditions to test their chondrogenic potential ( Figure 1). induction/differentiation of the NC using pluripotent mammalian ESCs as a model. To test this hypothesis, we overexpressed inducible forms of HES-1 and MSX-1 proteins in mouse ESCs and evaluated differentiation into NC derivatives, including neural, smooth muscle, and chondrocytelike cells, after activation of these transcription factors.

Expression of Hes1 and Msx1 in Wild-type ES Cells in Pluripotent Conditions and after NC Differentiation
To analyze the role of Hes1 and Msx1 in the differentiation of mESCs into neural crest cells, we used the stromal cell-inducing activity of Pre-adipose 6 (PA6) cells for 5 days [17], followed by the addition of BMP4, which commits cells to differentiate into NC derivatives [6]. Cultures were treated from day 5 to day 8 with 0.5 nM BMP4. At day 8, the cells were harvested by Papain treatment and plated on Fibronectin-coated dishes in neural differentiation medium with chick embryo extract without BMP4, a condition reported to favor differentiation into smooth muscle cells (α-SMA + ) and neural derivatives (β-TUBULIN III + ) [6] (Figure 1). Although we did not separate PA6 cells after replating, these cells did not proliferate, and their contribution to the measured parameters was constant for all conditions. At day 8, some cultures were dissociated and replated in micromass conditions to test their chondrogenic potential ( Figure 1). Figure 1. Differentiation of pluripotent cells into neural crest derivatives. The scheme shows the neural crest differentiation protocol. PA6 stromal cells were mitotically inactivated and plated in a monolayer to induce neural differentiation of mouse embryonic stem cells. The addition of BMP4 from day 5 to day 8 promoted the acquisition of a neural crest fate. At day 8, the cells were harvested with Papain and replated in Fibronectin-coated dishes or plated on micromass conditions. After 2 additional days, neurons and smooth muscle cells were detected. The micromasses were stained with Alcian blue after 10 days. Blue arrows indicate replating. The red asterisks indicate the addition of dexamethasone, starting at day one, to promote nuclear entry of the different constructs. The PA6 protocol was modified from [6] and the procedure for micromasses formation was modified from [34].
To assess the levels of expression of endogenous Hes1 and Msx1 in wild-type pluripotent cells, end-point RT-PCR was performed in RNA extracted from cultures grown in 2i/Leukemia inhibitory factor (LIF) without Dex. In W9.5 cells subjected to NC differentiation, we observed a non-significant Figure 1. Differentiation of pluripotent cells into neural crest derivatives. The scheme shows the neural crest differentiation protocol. PA6 stromal cells were mitotically inactivated and plated in a monolayer to induce neural differentiation of mouse embryonic stem cells. The addition of BMP4 from day 5 to day 8 promoted the acquisition of a neural crest fate. At day 8, the cells were harvested with Papain and replated in Fibronectin-coated dishes or plated on micromass conditions. After 2 additional days, neurons and smooth muscle cells were detected. The micromasses were stained with Alcian blue after 10 days. Blue arrows indicate replating. The red asterisks indicate the addition of dexamethasone, starting at day one, to promote nuclear entry of the different constructs. The PA6 protocol was modified from [6] and the procedure for micromasses formation was modified from [34].
To assess the levels of expression of endogenous Hes1 and Msx1 in wild-type pluripotent cells, end-point RT-PCR was performed in RNA extracted from cultures grown in 2i/Leukemia inhibitory factor (LIF) without Dex. In W9.5 cells subjected to NC differentiation, we observed a non-significant increase in Hes1 and a significant rise in Msx1 (Figure 2). These results indicates that upregulation of Msx1 is observed after NC induction of mESCs. However, it is not known if a controlled and sustained stimulation, causing nuclear translocation of either or both transcription factors, can influence NC commitemnt during differentiation of pluripotent cells. Therefore, an inducible system might be useful to test this idea. increase in Hes1 and a significant rise in Msx1 ( Figure 2). These results indicates that upregulation of Msx1 is observed after NC induction of mESCs. However, it is not known if a controlled and sustained stimulation, causing nuclear translocation of either or both transcription factors, can influence NC commitemnt during differentiation of pluripotent cells. Therefore, an inducible system might be useful to test this idea. Semiquantitative RT-PCR to measure the endogenous transcripts of these two genes in pluripotent W9.5 cells cultured in 2i/LIF, and after 10 days of differentiation in the PA6 protocol. The expression of Hes1 increased non-significantly. On the other hand, we detected a significant increase in Msx1 expression after NC differentiation. mean ± SEM is represented. Statistical analysis was conducted via one-way analysis of variance (ANOVA) followed by Tukey's multiple comparison test to establish differences. * p = 0.032.

HES-1-GR and MSX-1-GR Translocate to the Nucleus
The use of GR as an inducible system, activated by Dex, allows the timely stimulation of nuclear translocation of transcriptional regulators, to study the effects on cell differentiation. To test if the inducible system was activated by Dex to promote nuclear translocation, we cloned GR or Hes1-GR (HGR) cDNA sequences into a pcDNA3.1/myc-His vector, which has a 6X-His tag that can be recognized with a specific antibody. We also cloned the Msx1-GR (MGR) cDNA sequence into pcDNA3.1-Hygromycin B (Figures 3 and S1). The resulting plasmids were named GR, HGR, and MGR, respectively. These vectors were used to transiently transfect mouse embryo fibloblast (NIH-3T3) cells to evaluate nuclear translocation after Dex addition more easily than in mESCs, which have a very reduced cytoplasm. We assessed different concentrations of Dex (10, 50 and 100 μM) to obtain maximal nuclear translocation of each construct, and we decided to stimulate with the lowest tested concentration that caused a significant translocation effect. We measured the nuclear translocation using ImageJ software. In HGR cells, we quantified the fluorescence using an equation to obtain the Corrected Cell Fluorescence (CCF) in the nucleus ( Figure S2A). We observed an increase in the nuclear CCF after inducing nuclear translocation with Dex. In the case of MGR, we counted cells that had an exclusive nuclear signal after immunostaining and quantified the percentage of MSX-1 cells with positive labeling ( Figure S2B). After confirming that this system worked in mouse cells, we generated stable transgenic mESC lines. Semiquantitative RT-PCR to measure the endogenous transcripts of these two genes in pluripotent W9.5 cells cultured in 2i/LIF, and after 10 days of differentiation in the PA6 protocol. The expression of Hes1 increased non-significantly. On the other hand, we detected a significant increase in Msx1 expression after NC differentiation. mean ± SEM is represented. Statistical analysis was conducted via one-way analysis of variance (ANOVA) followed by Tukey's multiple comparison test to establish differences. * p = 0.032.

HES-1-GR and MSX-1-GR Translocate to the Nucleus
The use of GR as an inducible system, activated by Dex, allows the timely stimulation of nuclear translocation of transcriptional regulators, to study the effects on cell differentiation. To test if the inducible system was activated by Dex to promote nuclear translocation, we cloned GR or Hes1-GR (HGR) cDNA sequences into a pcDNA3.1/myc-His vector, which has a 6X-His tag that can be recognized with a specific antibody. We also cloned the Msx1-GR (MGR) cDNA sequence into pcDNA3.1-Hygromycin B (Figure 3 and Figure S1). The resulting plasmids were named GR, HGR, and MGR, respectively. These vectors were used to transiently transfect mouse embryo fibloblast (NIH-3T3) cells to evaluate nuclear translocation after Dex addition more easily than in mESCs, which have a very reduced cytoplasm. We assessed different concentrations of Dex (10, 50 and 100 µM) to obtain maximal nuclear translocation of each construct, and we decided to stimulate with the lowest tested concentration that caused a significant translocation effect. We measured the nuclear translocation using ImageJ software. In HGR cells, we quantified the fluorescence using an equation to obtain the Corrected Cell Fluorescence (CCF) in the nucleus ( Figure S2A). We observed an increase in the nuclear CCF after inducing nuclear translocation with Dex. In the case of MGR, we counted cells that had an exclusive nuclear signal after immunostaining and quantified the percentage of MSX-1 cells with positive labeling ( Figure S2B). After confirming that this system worked in mouse cells, we generated stable transgenic mESC lines.

Generation of Transgenic Stable mESC Lines
We transfected W9.5 mESCs with the GR, HGR, MGR, or both the HGR and MGR plasmids by electroporation, followed by antibiotic selection with 300 μg/mL G418 (GR and HGR), 200 μg/mL Hygromycin B (MGR) or both antibiotics for HGR+MGR for 10 days. The cells were maintained with 100 μg/mL G418 or 110 μg/mL Hygromycin B for 11 more days to obtain several colonies, which were picked and expanded in Knock Out-Dubelco´s Modified Eagle´s Medium (KO-DMEM) with LIF and 2i. We obtained genomic DNA from these clones to identify the transgenes using specific primers for GR/myc-His (208 bp), GR (577 bp), MGR (482 bp) or HGR (900 bp), as presented in the scheme shown in Figures 3B and 3C. Gapdh was also detected (576 bp). After genotyping, we identified several clones with the GR (four clones), HGR (six clones), MGR (five clones) or HGR+MGR (two clones) constructs ( Figures 3B and 3C). The identified positive clones were selected for differentiation based on the best

Generation of Transgenic Stable mESC Lines
We transfected W9.5 mESCs with the GR, HGR, MGR, or both the HGR and MGR plasmids by electroporation, followed by antibiotic selection with 300 µg/mL G418 (GR and HGR), 200 µg/mL Hygromycin B (MGR) or both antibiotics for HGR+MGR for 10 days. The cells were maintained with 100 µg/mL G418 or 110 µg/mL Hygromycin B for 11 more days to obtain several colonies, which were picked and expanded in Knock Out-Dubelco's Modified Eagle's Medium (KO-DMEM) with LIF and 2i. We obtained genomic DNA from these clones to identify the transgenes using specific primers for GR/myc-His (208 bp), GR (577 bp), MGR (482 bp) or HGR (900 bp), as presented in the scheme shown in Figure 3B,C. Gapdh was also detected (576 bp). After genotyping, we identified several clones with the GR (four clones), HGR (six clones), MGR (five clones) or HGR+MGR (two clones) constructs ( Figure 3B,C). The identified positive clones were selected for differentiation based on the best morphology of the pluripotent colonies.
To further assess whether these transgenes affected pluripotency markers, we performed immunostaining for OCT4, SOX2, and NANOG in the parental cell line W9.5 and the selected clonal lines ( Figure 4A), and we observed that the transgenic lines were positive for these markers. Consistently, RT-qPCR detection of Oct4, Sox2 and Nanog showed similar levels in all transgenic cell lines, except for a significant increase in Sox2 in HGR cells ( Figure 4B). This increase does not affect the pluripotent state, since mESCs can express high levels of Nanog [35] and HGR cells continue to express Oct4 and Sox2 at the mRNA and protein levels; these 3 transcription factors conform the core positive autoregulatory network of pluripotency [36]. the pluripotent state, since mESCs can express high levels of Nanog [35] and HGR cells continue to express Oct4 and Sox2 at the mRNA and protein levels; these 3 transcription factors conform the core positive autoregulatory network of pluripotency [36].

Expression of the Human Glucocorticoid Receptor, Hes1 and Msx1 are Similar in the Transgenic Cell
Lines before Differentiation.
Since the integration of plasmids in mESCs is randomized, we tested the expression level of the human GR, which is included in all the transgenic constructs, in cDNA prepared from cultures growing with 2i/LIF, using wild-type W9.5 mESCs as a negative control. Quantitative PCR in W9.5 cells did not result in amplification as expected. All transgenic lines express the transgene. We did not find significant differences, suggesting that all lines express comparable levels of transgenes ( Figure 5A). Additionally, to compare the endogenous levels of expression of Hes1 and Msx1 in wild- Since the integration of plasmids in mESCs is randomized, we tested the expression level of the human GR, which is included in all the transgenic constructs, in cDNA prepared from cultures growing with 2i/LIF, using wild-type W9.5 mESCs as a negative control. Quantitative PCR in W9.5 cells did not result in amplification as expected. All transgenic lines express the transgene. We did not find significant differences, suggesting that all lines express comparable levels of transgenes ( Figure 5A). Additionally, to compare the endogenous levels of expression of Hes1 and Msx1 in wild-type and transgenic pluripotent cells, end-point PCR was performed in cDNA prepared from cultures grown in 2i/LIF without Dex. No significant differences were detected in the expression of these two genes in undifferentiated conditions ( Figure 5B). grown in 2i/LIF without Dex. No significant differences were detected in the expression of these two genes in undifferentiated conditions ( Figure 5B).

Figure 5.
RT-PCR to detect the transgenic human glucocorticoid receptor and the endogenous Hes1 and Msx1 in pluripotent wild-type and transgenic lines. RNA was obtained from W9.5, GR, HGR, MGR, and HGR+MGR cells growing in 2i/LIF. Retrotranscription was made to prepare cDNA, which was amplified by qPCR (A) with the primers listed in Methods. Wild-type W9.5 mESCs did not express the transgene. The expression of the transgenic sequence present in all plasmids was normalized to β-actin. (B) Semiquantitative RT-PCR was used to measure transcripts of these two genes in pluripotent W9.5 (non-transgenic) and transgenic (GR, HGR, MGR, and HGR+MGR) cells, cultured in 2i/LIF. mean ± SEM is represented. Statistical analysis was conducted via one-way ANOVA followed by Tukey's multiple comparison test to establish differences. No significant differences were found between the four transgenic cell lines in 3-4 independent experiments.

Neural and NC Marker Expression in Transgenic Pluripotent and Differentiated Cells
After 10 days of differentiation under control conditions, we extracted RNA and performed RT-PCR to analyze the expression of different NC markers ( Figure 6). The expression of each gene was normalized to β-actin levels and is presented as fold change with respect to expression in control pluripotent cells (GR). In this assay, we analyzed the expression of endogenous Msx1 and Hes1, as well as Sox2 and NC markers (Sox9, Sox10, and Snai1) after NC differentiation. We observed significant differences when mESCs were subjected to differentiation into NC derivatives. Sox2 expression decreased, whereas Sox9, Snai1, and Msx1 expression was augmented in all cell lines relative to pluripotent GR cells. No significant differences were found in Sox10 and Hes1 between the four cell lines under pluripotent or differentiated conditions. These molecular changes are consistent with the transcriptional response observed after NC differentiation of ESCs [19,22]. RT-PCR to detect the transgenic human glucocorticoid receptor and the endogenous Hes1 and Msx1 in pluripotent wild-type and transgenic lines. RNA was obtained from W9.5, GR, HGR, MGR, and HGR+MGR cells growing in 2i/LIF. Retrotranscription was made to prepare cDNA, which was amplified by qPCR (A) with the primers listed in Methods. Wild-type W9.5 mESCs did not express the transgene. The expression of the transgenic sequence present in all plasmids was normalized to β-actin. (B) Semiquantitative RT-PCR was used to measure transcripts of these two genes in pluripotent W9.5 (non-transgenic) and transgenic (GR, HGR, MGR, and HGR+MGR) cells, cultured in 2i/LIF. mean ± SEM is represented. Statistical analysis was conducted via one-way ANOVA followed by Tukey's multiple comparison test to establish differences. No significant differences were found between the four transgenic cell lines in 3-4 independent experiments.

Neural and NC Marker Expression in Transgenic Pluripotent and Differentiated Cells
After 10 days of differentiation under control conditions, we extracted RNA and performed RT-PCR to analyze the expression of different NC markers ( Figure 6). The expression of each gene was normalized to β-actin levels and is presented as fold change with respect to expression in control pluripotent cells (GR). In this assay, we analyzed the expression of endogenous Msx1 and Hes1, as well as Sox2 and NC markers (Sox9, Sox10, and Snai1) after NC differentiation. We observed significant differences when mESCs were subjected to differentiation into NC derivatives. Sox2 expression decreased, whereas Sox9, Snai1, and Msx1 expression was augmented in all cell lines relative to pluripotent GR cells. No significant differences were found in Sox10 and Hes1 between the four cell lines under pluripotent or differentiated conditions. These molecular changes are consistent with the transcriptional response observed after NC differentiation of ESCs [19,22]. The transgenic cells were differentiated into neural crest derivatives over 10 days. The expression of several markers was measured using end point RT-PCR. The expression of each gene was normalized to β-actin expression, and the fold change was determined relative to the levels found in pluripotent GR cells. mean ± SEM. Statistical analysis was conducted via two-way ANOVA followed by Tukey's multiple comparison test to establish differences between pluripotent and differentiated cells from 3 independent experiments. * p < 0.01; ** p < 0.001. No differences were found between the four cell lines when compared at in the pluripotent state or after differentiated conditions, indicating that the significant differences are caused by the differentiation protocol.
To evaluate the role of HES-1 and MSX-1 activation, Dex was added from day 1 of differentiation. RNA was obtained at day 10 of differentiation to evaluate the same genes using semiquantitative RT-PCR. It is noteworthy that the primers for Hes1 and Msx1 detect the endogenous forms and not the transgenes. We only observed significant increases in Sox10 expression in MGR and HGR+MGR cells relative to the control GR and HGR cells after stimulation of nuclear translocation (Figure7A). Evaluation of NC markers via RT-qPCR revealed that in Dex-treated double-transgenic cells, the activation of HES-1 and MSX-1 nuclear translocation increased α-Sma transcripts, suggesting increased differentiation to smooth muscle cells, one type of NC derivative ( Figure 7B). In contrast, the levels of other genes related to neural differentiation (Pax6, Sox1, and β-Tubulin III) did not change. Analysis of additional NC markers such as Type II Collagen gene (Col2a1) expression, characteristic of chondrocytes, was increased in MGR cells, but other chondrocytic markers (Aggrecan and Col10a1) were unaffected ( Figure 7B). The transgenic cells were differentiated into neural crest derivatives over 10 days. The expression of several markers was measured using end point RT-PCR. The expression of each gene was normalized to β-actin expression, and the fold change was determined relative to the levels found in pluripotent GR cells. mean ± SEM. Statistical analysis was conducted via two-way ANOVA followed by Tukey's multiple comparison test to establish differences between pluripotent and differentiated cells from 3 independent experiments. * p < 0.01; ** p < 0.001. No differences were found between the four cell lines when compared at in the pluripotent state or after differentiated conditions, indicating that the significant differences are caused by the differentiation protocol.
To evaluate the role of HES-1 and MSX-1 activation, Dex was added from day 1 of differentiation. RNA was obtained at day 10 of differentiation to evaluate the same genes using semiquantitative RT-PCR. It is noteworthy that the primers for Hes1 and Msx1 detect the endogenous forms and not the transgenes. We only observed significant increases in Sox10 expression in MGR and HGR+MGR cells relative to the control GR and HGR cells after stimulation of nuclear translocation ( Figure 7A). Evaluation of NC markers via RT-qPCR revealed that in Dex-treated double-transgenic cells, the activation of HES-1 and MSX-1 nuclear translocation increased α-Sma transcripts, suggesting increased differentiation to smooth muscle cells, one type of NC derivative ( Figure 7B). In contrast, the levels of other genes related to neural differentiation (Pax6, Sox1, and β-Tubulin III) did not change. Analysis of additional NC markers such as Type II Collagen gene (Col2a1) expression, characteristic of chondrocytes, was increased in MGR cells, but other chondrocytic markers (Aggrecan and Col10a1) were unaffected ( Figure 7B).  The transgenic cells were differentiated into neural crest derivatives over 10 days. The expression of several markers was measured using end point RT-PCR. The expression of each gene was normalized to β-actin expression, and the fold change was determined relative to the levels found in pluripotent GR cells. mean ± SEM. Statistical analysis was conducted via two-way ANOVA followed by Tukey's multiple comparison test to establish differences between pluripotent and differentiated cells from 3 independent experiments. * p < 0.01; ** p < 0.001. No differences were found between the four cell lines when compared at in the pluripotent state or after differentiated conditions, indicating that the significant differences are caused by the differentiation protocol.
To evaluate the role of HES-1 and MSX-1 activation, Dex was added from day 1 of differentiation. RNA was obtained at day 10 of differentiation to evaluate the same genes using semiquantitative RT-PCR. It is noteworthy that the primers for Hes1 and Msx1 detect the endogenous forms and not the transgenes. We only observed significant increases in Sox10 expression in MGR and HGR+MGR cells relative to the control GR and HGR cells after stimulation of nuclear translocation (Figure7A). Evaluation of NC markers via RT-qPCR revealed that in Dex-treated double-transgenic cells, the activation of HES-1 and MSX-1 nuclear translocation increased α-Sma transcripts, suggesting increased differentiation to smooth muscle cells, one type of NC derivative ( Figure 7B). In contrast, the levels of other genes related to neural differentiation (Pax6, Sox1, and β-Tubulin III) did not change. Analysis of additional NC markers such as Type II Collagen gene (Col2a1) expression, characteristic of chondrocytes, was increased in MGR cells, but other chondrocytic markers (Aggrecan and Col10a1) were unaffected ( Figure 7B). RT followed by qPCR. The fold induction was relative to the levels in GR cells. The data for each transcript were normalized to β-actin levels. mean ± SEM. Statistical analysis of all conditions was conducted via ANOVA followed by Tukey's multiple comparison test from 3-4 independent experiments. The marked significant differences are relative to GR, unless indicated otherwise. * p < 0.05.

Activation of HES-1 and MSX-1 Increased the Number of β-TUBULIN III-and α-SMA-positive Cells
Transgenic mESC lines differentiated using the NC protocol with PA6 stromal cells were fixed at day 10 to perform immunofluorescence staining with specific antibodies. We detected the expression of β-Tubulin III in cells that were small and exhibited a multipolar morphology under all conditions ( Figure 8A); quantification of immuno-positive cells showed that the number of β-TUBULIN III-positive cells significantly increased when the nuclear translocation of HES-1 and MSX-1 was induced with Dex ( Figure 8B). Since real-time PCR indicated that nuclear translocation in double-transgenic cells upregulated Smooth muscle actin expression, antibodies for α-SMA were used to examine differentiated cells. The control and treatment conditions produced large, filamentous and polyhedral α-SMA + cells ( Figure 8C); however, quantification of the number of cells positive for this marker showed a significant increase when HGR+MGR cells were activated ( Figure  8D). Please note that the primers for Hes1 and Msx1 detect the endogenous forms and not the transgenes (B) Quantification of the neural and neural crest differentiation markers via RT followed by qPCR. The fold induction was relative to the levels in GR cells. The data for each transcript were normalized to β-actin levels. mean ± SEM. Statistical analysis of all conditions was conducted via ANOVA followed by Tukey's multiple comparison test from 3-4 independent experiments. The marked significant differences are relative to GR, unless indicated otherwise. * p < 0.05.

Activation of HES-1 and MSX-1 Increased the Number of β-TUBULIN III-and α-SMA-positive Cells
Transgenic mESC lines differentiated using the NC protocol with PA6 stromal cells were fixed at day 10 to perform immunofluorescence staining with specific antibodies. We detected the expression of β-Tubulin III in cells that were small and exhibited a multipolar morphology under all conditions ( Figure 8A); quantification of immuno-positive cells showed that the number of β-TUBULIN III-positive cells significantly increased when the nuclear translocation of HES-1 and MSX-1 was induced with Dex ( Figure 8B). Since real-time PCR indicated that nuclear translocation in double-transgenic cells upregulated Smooth muscle actin expression, antibodies for α-SMA were used to examine differentiated cells. The control and treatment conditions produced large, filamentous and polyhedral α-SMA + cells ( Figure 8C); however, quantification of the number of cells positive for this marker showed a significant increase when HGR+MGR cells were activated ( Figure 8D). Immunostaining with anti-α-SMA antibody after differentiation under control and dexamethasoneadded conditions. (D) Quantification of large α-SMA-positive cells with a filamentous aspect. Doubletransgenic cells produced a higher proportion of smooth muscle cells compared with GR and MGR cells. Experiments were performed in triplicate; pictures were taken in at least ten fields from three independent experiments. The mean ± SEM is presented. Statistical analysis of all conditions was conducted with ANOVA followed by Tukey's multiple comparison test and only the significant differences are indicated in the graph. * p < 0.05, ** p < 0.01, and *** p < 0.001.

Activation of MGR and Double-transgenic Cells Increases Chondrocyte Differentiation.
Immunostaining of cells stimulated to differentiate to NC for 10 days, with or without Dex, revealed a significant increase in COL2A1-positive cells in HGR+MGR transgenic cells, after activation of both transcription factors (Figure 9). To analyze if chondrogenic differentiation can proceed further, cells were differentiated on PA6 cells for 8 days and replated as micromass culures in differentiation medium without chick embryo extract (Figure 1) [34]. These cultures were subjected to Alcian blue staining to reveal proteoglycan characteristic of extracellular matrix produced by chondrocytes. After 3 days, no positive cells were detected, regardless of Dex addition. After 10 days all cell lines showed a few blue cells in the absence of Dex, and activation of MSX-1 clearly increased the proportion of Alcian blue-positive cells. Furthermore, blue aggregates of cobblestone-like cells were more abundant in MGR cells stimulated with Dex ( Figure 10). (D) Quantification of large α-SMA-positive cells with a filamentous aspect. Double-transgenic cells produced a higher proportion of smooth muscle cells compared with GR and MGR cells. Experiments were performed in triplicate; pictures were taken in at least ten fields from three independent experiments. The mean ± SEM is presented. Statistical analysis of all conditions was conducted with ANOVA followed by Tukey's multiple comparison test and only the significant differences are indicated in the graph. * p < 0.05, ** p < 0.01, and *** p < 0.001.

Activation of MGR and Double-transgenic Cells Increases Chondrocyte Differentiation.
Immunostaining of cells stimulated to differentiate to NC for 10 days, with or without Dex, revealed a significant increase in COL2A1-positive cells in HGR+MGR transgenic cells, after activation of both transcription factors (Figure 9). To analyze if chondrogenic differentiation can proceed further, cells were differentiated on PA6 cells for 8 days and replated as micromass culures in differentiation medium without chick embryo extract (Figure 1) [34]. These cultures were subjected to Alcian blue staining to reveal proteoglycan characteristic of extracellular matrix produced by chondrocytes.

Discussion
In this study, we used human GR as an inducible system. We verified that Msx1 and Hes1 fused to GR responded to Dex addition by nuclear translocation. Stable transgenic mESC lines containing GR, HGR, MGR, or HGR+MGR were generated, maintained the expression of pluripotency-associated markers, and responded to a differentiation protocol for NC induction. After differentiation of the transgenic cell lines to obtain neural crest derivatives and treatment with Dex to induce nuclear translocation of the transcription factors during the differentiation progress, we found the following:

Discussion
In this study, we used human GR as an inducible system. We verified that Msx1 and Hes1 fused to GR responded to Dex addition by nuclear translocation. Stable transgenic mESC lines containing GR, HGR, MGR, or HGR+MGR were generated, maintained the expression of pluripotency-associated markers, and responded to a differentiation protocol for NC induction. After differentiation of the transgenic cell lines to obtain neural crest derivatives and treatment with Dex to induce nuclear translocation of the transcription factors during the differentiation progress, we found the following: (1) Nuclear translocation of HES-1 and MSX-1 in double-transgenic cells significantly increased expression of Sox10, as well as differentiation to neurons and smooth muscle cells by augmenting the number of cells positive for β-TUBULIN III and Smooth Muscle Actin, respectively.
Hes basic Helix-Loop-Helix (bHLH) repressors encode nuclear proteins that repress transcription, and they have three conserved regions that confer transcriptional functions: the bHLH (DNA binding and region for dimerization), Orange (selection of bHLH heterodimer partners) and WRPW (represses transcription) domains [37][38][39]. Hes genes play an essential role in the development of many organs by maintaining progenitor cells and regulating binary cell fate decisions. During late neural development, HES-1 regulates an astrocyte versus oligodendrocyte fate choice in glial restricted precursors [40]. On the other hand, Hes genes have a role in cell cycle [28]. HES-1 has an important function in self-renewal and cell survival in multipotent neural cells in the central nervous system [41]. This transcription factor has an important role in neural development, but the role in NC induction in mammals remains unclear. We did not observe induction of further NC differentiation in HGR cells after Hes1 nuclear translocation.
It has been reported that HES-1 increases the promoter activity of α-Sma in hepatic stellate cells [42], although previous studies reported that activation of Hes1 by juxtacrine Notch signaling did not increase α-SMA protein levels in the same cell type [43]. In contrast, high levels of HES-1 inhibit the ability of MyoD to convert 10T1/2 mouse embryonic fibroblasts to muscle cells [44]. Our results show that neither activated HES-1 nor MSX-1 increased α-Sma transcript or protein levels. However, the coactivation in HGR+MGR cells showed a significant effect, increasing the number of α-SMA-positive cells and α-Sma mRNA levels. Interestingly, a previous paper showed that upon differentiation in the presence of TGF-β human iPSC-NC stem cells produced a Smooth Muscle lineage that expressed α-Sma, [45]. Thus, the regulation of expression of α-Sma seems to be dependent on the partners of Hes1 and/or the cell type.
On the other hand, there is extensive evidence that MSX-1 transcription factors have an important role in proliferation, differentiation, and NC induction during embryonic development. Msx1 and Msx2 are specifically required for osteogenesis in the cranial NC lineage. Msx1 is strongly expressed in these cranial cells and plays a critical role in regulating epithelial to mesenchymal transition (EMT) during organogenesis [46,47]. Msx1 and Msx2 function redundantly in multiple tissues and organs during vertebrate development, including the heart. Msx1-/-Msx2-/double-knockout mice have profound defects in the cardiac outflow tract. These defects were associated with excessive proliferation of cardiac NC cells [31]. A study with Msx1/Msx2 double-null mice investigated the effect of eliminating these genes on atrioventricular cushion and valve development. The researchers demonstrated that mutant mice had impaired EMT, resulting in hypoplastic cushions and malformed valves. Interestingly, α-Sma is expressed prior to EMT in smooth muscle cell precursors; in double-knockout animals, decreased expression of α-Sma is reported, which can contribute to defective atrioventricular cushion formation [48]. The NC is one of the major sources of smooth cell muscle cells in blood vessels (i.e., the aorta, ductus arteriosus, common carotid, innominate and right subclavian arteries) [49][50][51][52]. Taken together, these results suggest that simultaneous activation of MSX-1 and HES-1 participate in the regulation of α-Sma expression.
SOX proteins are candidates for participation in NC specification. They are involved in several processes during embryogenesis. Based on the amino acid sequences of the HMG domain, SOX proteins can be divided into ten subgroups. Subgroup E consists of three members (SOX8, SOX9 and SOX10) that are expressed in several developing tissues, including the NC [53]. Sox9 is an early marker of prospective NC cells, preceding markers of migratory cells. Sox9 is the first to be expressed, overlapping with the expression of BMP4 and Snai2 [48,54]. Sox10 is expressed shortly after SNAI2 and is required for the early survival and migration of NC cells, and later on for differentiation of Schwann cells and melanocytes. Sox10 expression continues in glia development but is downregulated in other NC lineages [55].
In vivo, neural crest cells constitute a transient population that arise from the ectoderm and migrates away and differentiate into many derivatives. Early work described that multipotent NC stem cells can be isolated from the rat sciatic nerve [56] or from postmigratory cells of the mouse first branchial arch [57]; such cells can differentiate in vitro to neurons, osteoblasts, myofibroblast (positive for α-SMA), and Schwann cells, even at clonal assays. This suggests that a rare niche of NC stem cells might be present in vivo, but cultured cells can present a different behavior in terms of progenitor gene expression and terminal differentiation. Supporting this notion, multipotent "crestospheres" from premigratory chick embryos express the NC genes Sox9, Ets1, Bmp4, and Snai2, as well as the neural marker Pax6 [58]. These NC cells differentiated to neurons, melanocytes, smooth muscle cells, and chondrocytes, both in vitro and after transplantation in the head mesenchyme of chicks. Induction of the premigratory NC population from neural rosettes derived from human ESCs produced crestospheres that differentiated to neurons, melanoblasts, smooth muscle cells, and chondrocytes, but in conjunction with neurospheres positive for Sox2. These results highlight the differences observed when studying NC progenitors in vivo and those differentiated in vitro from ESCs. We observed significant increases in Sox9 and Snai1 expression after NC induction under control conditions, indicating proper NC induction of mESCs. Furthermore, in MGR and HGR+MGR cells stimulated with Dex, significant upregulation of Sox10 was observed. This suggests that the NC progenitor state is enriched under these conditions. However, enhanced differentiation to neurons, smooth muscle, and chondrocyte-like cells was also observed, which might reflect cell populations at different stages of differentiation to NC derivatives in these cultures. To our knowledge, this is the first report that indicates that activation of HES-1 and MSX-1 in mESCs produce higher proportions of neurons, smooth muscle cells, and chondrocytes, derivatives obtained also from the differentiation of NC stem/progenitor cells, although the molecular mechanisms remain to be identified.
Previously reported evidence demonstrated that morphogens, such as BMPs, together with the transcription factor SOX9, are crucial for chondrogenesis. Wnt1-driven Sox9 conditional knockout present defects in cartilage production in cranial structures [32]. SOX9 is the transcription factor that activates Collagen IIa (Col2a1) [59]. On the other hand, the Col2a1 gene can also be controlled by Barx2, a homeodomain transcription factor [60]. Chondrocytes can be articular or growth plate (hypertrophic): the former give rise to cartilage and the latter provide a template on which new bone is formed, substituting hypertrophic cells. Hypertrophic chondrocytes express Col10a1 [61].
Activation of Msx genes by BMP signaling can promote cartilage development. For example, exogenous BMP-7, applied to lateral mandibular region at an early stage of chick development, causes ectopic expression of Msx1 and inhibited the formation of Meckel´s cartilage. In contrast, BMP-7 addition at later stages up-regulated Msx1, but the final result was increased chondrogenesis [62]. Another paper where chick embryos were transduced with BMP receptors in craniofacial structures, showed that the activation of these receptors upregulated Msx1 and Msx2 transcripts in regions that normally do not express these genes, causing overgrowth of cartilage elements in the head; the authors conclude that the activation of BMP receptors promotes chondrogenesis [63].
We found that MSX-1 activation in MGR cells induces upregulation of Col2a1 transcription, as well as an increase in Alcian blue staining of micromass cultures, an effect not previously reported, but that might have interesting implications for cartilage development. In fact, the expression of Col2a1 is a sign of chondrocyte maturation for cells differentiated from mESCs, and was observed when using a different protocol, involving embryoid bodies and cell aggregation, to produce chondrocytes [33]. Craft and coworkers used different induction protocols for the production of articular o hypertrophic chondrocytes. They described differences in morphology and Collagen2a1 protein levels: hypertrophic cells are larger and had higher levels of Collagen 2, compared to articular cells [33]. Our results indicated that in MGR cells stimulated with Dex chondrogenesis occurred by the expression of Sox9 and Col2a1. Furthermore, stimulated MGR cells replated in micromass conditions showed stronger staining for the extracellular matrix components detected by Alcian blue. Interestingly, HGR+MGR cells stimulated with Dex showed increased proportions of COLLAGEN 2A1-positive cells, larger chondrocyte-like cells and an increased, albeit non-significant, expression of the hypertrophic marker Col10a1, suggesting that simultaneous activation of HES-1 and MSX-1 promotes a growth plate chondrocyte-like phenotype. To avoid nutrient depletion of the culture medium, PA6 cells were inactivated mitotically with 10 µg/mL Mitomycin C for 2 h at 37 • C in 5% CO 2 . After 2 h, the cells were washed with PBS, harvested and maintained in culture with α-MEM until used for differentiation of mESCs.

Generation of cDNA Constructs and the Inducible System
The development of hormone-inducible proteins allows accurate control of the time at which they act [65,66]. To temporally control overexpression of Hes1 and Msx1 and coexpression of Hes1-Msx1, the ligand binding domain of the human GR was amplified with the GR-Fw (SacI) 5 -GAGCTCcacaagaaacctctgaaaatcctggtaac-3 and GR-Rev (XhoI) 5 -CTCGAGctgattggtgatgatttcagc-3 [15] and cloned upstream in frame (into SacI and XhoI sites) with the coding sequence of a myc-His construct. This control plasmid only expresses encoded GR and the myc-His tag. The Mus musculus Hes1 ORF and Msx1 ORF sequences were high-fidelity PCR amplified from the full-length cDNA clone ID 6478994 (cat. # MMM1013-9200434, Open BioSystems Inc., Dharmacon, Lafayette, CO, USA) using the following primers: mHes1-Fw (HindIII) 5 -AAGCTTatgccagctgatataatggagaaa-3 ; mHes1-Rev (BamHI) 5 -GGATCCagttccgccacggtctcca ; mMsx1-Fw (HindIII) 5 -AAGCTT atgacttctttgccactcggtg-3 ; mMsx1-Rev (BamHI) 5 -GGATCC aagtcaggtggtacatgctgtagcc-3 ) and ligated into a pCR-Topo-TA vector. The amplified fragment encoded 284 amino acid residues of Hes1, without the stop codon, and included its three structural domains, bHLH, Orange and WRPW. Msx1 (cDNA clone ID 4923403, cat. # MMM1013-202766935) contained the repressor domain, homeodomain (DNA binding) and nuclear localization signal (NLS). The pTopo-Hes1 vector and pTopo-Msx1 vectors were cut and then directionally subcloned into HindIII and BamHI sites in the GR vector to produce the Hes1-GR-pcDNA3.1-myc-His construct and Msx1-GR was subcloned into pcDNA3.1-Hygro, as shown in Table 1. DH5α cells were used to expand the plasmid, and the colonies were selected with 100 µg/mL ampicillin. Plasmid DNA was obtained and sequenced to verify each construct.

Nuclear Translocation after Dexamethasone Stimulation
NIH-3T3 cells were used to express the hGR-pcDNA3.1-myc-His (GR), Hes1-hGR-pcDNA3.1myc-His (HGR) or Msx1-hGR-pcDNA3.1-Hygro (MGR) constructs. Cells were transfected using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's instructions. Twenty-four hours before transfection, NIH-3T3 cells were cultured in DMEM without phenol red supplemented with 10% charcoal-stripped FBS to avoid activation of GR before Dex addition. Forty-eight hours after transfection, the cells were treated with 10 µM dexamethasone (Dex; D2915, Sigm -Aldrich, Milwaukee, WI, USA) to evaluate nuclear translocation of the constructs. We stimulated the cells with Dex for 24 h, assessed nuclear translocation using specific antibodies and analyzed the staining using ImageJ program (1.51 p 22). In brief, we selected, using the freeform drawing tool, the nucleus of each cell stained with the specific antibody recognizing the 6X-His tag to identify GR and HGR translocation or anti-MSX-1/2 antibodies for MGR. After selection of the area of interest, we chose a region next to the cell that had no fluorescence, defined this as the background level, and subtracted the contribution of the background in the selected area. We then quantified the integrated density and mean gray value and used these data to calculate the corrected cell fluorescence (CCF) in the nucleus using the equation CCF = Integrated density − (area of selected cell × mean fluorescence of background readings).

Generation of Stable mESC Lines
Two-and-a-half million mESCs were electroporated separately with 3 µg of the GR, HGR, or MGR plasmids, or with both the HGR and MGR plasmids using the Nucleofector system (Lonza, Basilea, Suiza). After electroporation, cells were cultured in KO-DMEM with LIF and 2i. Two days after electroporation, mESCs were treated with 200 µg/mL G418 (Sigma-Aldrich, Milwaukee, WI, USA) or 300 µg/mL Hygromycin B (Sigma-Aldrich, Milwaukee, WI, USA) for 10 days to select resistant clones. Clones were expanded and maintained with 100 µg/mL G418 or 110 µg/mL Hygromycin B in the presence of LIF and 2i before differentiation.

Genotyping of the Resistant Clones
To detect integration of specific cDNA into the mouse genome, we obtained genomic DNA and evaluated the presence of human GR, HGR or MGR via end-point PCR using specific primers (all listed 5 -3 ). For GR: Forward (Fw)-gttggaggttattgaacctgaag and Reverse (Rev)-ttcatgcatagaatccaagag; GR/myc-His: Fw-ctgacaaaactcttggattctatgc and Rev-tgaatatgcataccggtcatcatc; mMsx1-GR (MGR): Fw-cctacgcaagcacaagacca and Rev-ggtaggggtgagttgtggta; mHes1-GR (HGR): Fw-tctacaccagcaacagtgg and Rev-tgaatatgcataccggtcatcatc. The amount of DNA was assessed with the following specific primers for Gapdh (NM_001289726.1): Fw-atcaccatcttccaggagcg and Rev-cctgcttcaccaccttcttg. The confirmed clones for each construct were chosen based on the best morphology of the pluripotent colonies. One clone per construction was analyzed: GR (G2), HGR (H5), MGR (M5) and HGR+MGR (H+M 20).

Neural Crest Differentiation
mESCs were harvested with Triple Xpress (Gibco), resuspended in fresh DMEM without phenol red LIF or 2i, counted, plated at 2.5 × 10 3 cells/mL in 24 well-plates with mitotically inactivated PA6 cells for 8 days in differentiation medium (DMEM without phenol red, 10% charcoal-stripped FBS, 2 mM Glutamax, 0.1 mM nonessential amino acids, 1 mM sodium pyruvate and 0.1 mM 2-mercaptoethanol) [6]. According to this protocol, 0.5 nM BMP4 was added from day 5 to day 8 of differentiation and cells were replated on fibronectin-coated dishes in medium containing chick embryo extract to promote neural and smooth muscle differentiation [6,56]. We evaluated the expression of NC, neuronal and smooth muscle cell markers via RT-PCR with specific primers and by immunofluorescence with specific antibodies.

Immunofluorescence Staining
NIH-3T3 fibroblasts were plated on gelatin-coated coverslips, washed three times in PBS, fixed with 4% paraformaldehyde solution for 10 min, washed again three times with PBS and permeabilized with 0.5% Triton X-100 for 1 h. After blocking for 1. Fw-agctcaccagaccctgagaa and Rev-ccagcaatcgttaccttc; Sox10 (NM_011437): Fw-gcaggaagggttagggtagg and Rev-gcggagaaaggatcagagtg. PCR products were analyzed via electrophoresis to determine their molecular weights after ethidium bromide staining. Reactions with RNA in the absence of retro-transcription were included as negative controls. The housekeeping genes β-Actin (NM_007393) (Primers: Fw-gggtcagaaggattcctatg and Rev-ggtctcaaacatgatctggg) and Gapdh (NM_001289726.1) (Primers: Fw-atcaccatcttccaggagcg and Rev-cctgcttcaccaccttcttg) were used to normalize the gene expression levels. The expression of neural and NC markers was evaluated in GR, HGR, MGR, and HGR+MGR cell lines after induction of nuclear translocation by Dex, and the results were compared to the condition without Dex.

qPCR
Real-time polymerase chain reaction (PCR) was carried out using iTaq Universal SYBR Green Supermix (Bio-Rad, Hercules, CA, USA) and a CFX96 Real-Time System (Bio-Rad, Hercules, CA, USA). The primer concentration was 0.3 µM. Forty reaction cycles were performed followed by melt curves to confirm the specificity of the primer sets.
Three biological replicates were measured. We used the following specific primers.

Micromass Culture
After 8 days of differentiation in stromal PA6 cells, cells were harvested and counted. Micromass cultures were obtained by pipetting a 10 µL droplet (2 × 10 5 cells) of cell suspension in 24-well plates. Following a 3 h attachment period without medium, differentiation medium was gently added, and changed every other day. The day of plating in micromass culture was designated as day 0. Starting on day 1, the culture medium was changed, and dexamethasone or vehicle was added, as indicated in Figure 1. After 3 or 10 days, micromass cultures were rinsed twice with PBS and fixed with Khale solution for 20 min at room temperature. This solution was substituted by 0.5% Alcian blue solution in 0.1 N HCl. After two days of staining, cultures were washed two times with 0.1 N HCl for 5 min. The cultures were stored in PBS until microphotography analysis [67]. The images of each well were digitalized by a color scanner (Hp Deskjet Ink Advantage 2135, Palo Alto, CA, USA), and are shown as insets for each condition. The microscopic pictures were acquired with an inverted microscope (Nikon DIAPHOT, Minato, Tokio, Japan) and digitalized to 8 bits. These images were further processed to subtract background and to add the "Cyan hot" filter, using the open access program Fiji (Figure 10 A-H). Higher magnifications images were acquired with a Nikon eclipse TE2000-U inverted microscope and minimally adjusted for brightness and contrast ( Figure 10 I-X).

Statistical Analysis
Experiments were performed in triplicate with at least 3 independent replicates. The results are expressed as the mean ± SEM. One-way or two-way ANOVA was applied followed by Tukey's multiple comparison test. p values ≤0.05 were considered statistically significant.

Conclusions
Altogether, our results show that the activation of two key genes that have been demonstrated to be pivotal in the development of multiple lineages in chick, mouse, and frog embryos has an effect on differentiation of the generated mESC lines into NC cells. Most protocols for NC differentiation of ESCs require several exogenous cell signaling molecules to facilitate differentiation into NC-derived cells [68]. Remarkably, in this work, the lineages obtained were differentiated through activation of the HES-1 and MSX-1 transcription factors in addition to the extracellular stimulation. It is tempting to speculate that these differentiated α-SMA-expressing cells will provide a valuable in vitro model to gain insights into the genes and pathways involved in the formation of smooth muscle cells for research in vascular development. Vascular smooth cells are neural crest derivatives, a cell population which is important in physiological and pathological processes that are not yet completely understood (i.e., the neurocristopathies CADASIL and Alagille syndromes) [1]. On the other hand, MGR cells might be used to produce chondrocytes at higher efficiencies, another NC-derived cell type which is highly relevant for craniofacial development