ABSTRACT

Net1 is a well-characterized oncoprotein with RhoA-specific GEF activity. Oncogenic Net1, which lacks the first 145 amino acids, is present in the cytosol and contributes to the efficient activation of RhoA and the formation of actin stress fibers in a number of tumor cell types. Meanwhile, wild-type Net1 is predominantly localized in the nucleus at steady state due to its N-terminal nuclear localization sequences, where the function of nuclear Net1 has not been fully determined. Here, we find that zebrafish net1 is expressed specifically in mesendoderm precursors during gastrulation. Endogenous Net1 is located in the nucleus during early embryonic development. Gain- and loss-of-function experiments in zebrafish embryos and mammalian cells demonstrate that, regardless of its GEF activity, nuclear Net1 is critical for zebrafish mesendoderm formation and Nodal signal transduction. Detailed analyses of protein interactions reveal that Net1 associates with Smad2 in the nucleus in a GEF-independent manner, and then promotes Smad2 activation by enhancing recruitment of p300 (also known as EP300) to the transcriptional complex. These findings describe a novel genetic mechanism by which nuclear Net1 facilitates Smad2 transcriptional activity to guide mesendoderm development.

INTRODUCTION

Members of the transforming growth factor β (TGF-β) superfamily, including Nodal and bone morphogenetic proteins (BMPs), are key players in embryonic development and homeostasis (Massagué, 2008; Wu and Hill, 2009). The Nodal subfamily of these growth factors [Squint (Sqt) and Cyclops (Cyc) in zebrafish (also known as Ndr1 and Ndr2, respectively); Nodal in mammals] are secreted morphogens that form concentration gradients required for mesendoderm induction and patterning (Feldman et al., 1998; Gritsman et al., 1999; Zhou et al., 1993). These ligands bind and activate heteromeric complexes of type I and II transmembrane receptors, which in turn phosphorylate intracellular mediators, such as Smad2 and Smad3. Phosphorylated Smad2 and Smad3 can then form complexes with Smad4 and translocate into the nucleus to regulate transcription of target genes (Schier and Shen, 2000; Schmierer and Hill, 2007). While this signal transduction pathway is well-characterized and seemingly straightforward, how it is spatio-temporally modulated during embryonic development remains unclear.

By using ChIP-chip assays, we previously identified several hundreds of genes directly targeted by Nodal signaling, such as the proto-oncogene neuroepithelioma transforming gene 1 (NET1) (Liu et al., 2011). Net1, the protein product of NET1, was originally isolated from neuroepithelioma cells and acts as a RhoA-specific guanine nucleotide exchange factor (GEF) to enhance cancer cell migration and invasion (Alberts and Treisman, 1998; Chan et al., 1996; Murray et al., 2008; Tu et al., 2010). In addition, this gene is expressed in a dorsoventral-gradient in the blastoderm margin at the onset of gastrulation and has important functions during early embryonic development (Liu et al., 2011). Net1-mediated RhoA activation is required for gastrulation movements during embryonic development of Xenopus and chicken (Miyakoshi et al., 2004; Nakaya et al., 2008), and, in zebrafish (Danio rerio), Net1 and its GEF activity are indispensable for dorsal cell fate specification through promotion of maternal β-catenin activation (Wei et al., 2017). TGF-β- and Nodal-regulated genes usually participate in feedback modulation of signal transduction (Liu et al., 2016; Nicklas and Saiz, 2013; Stroschein et al., 1999). The expression of zebrafish net1 in the blastoderm margin, where mesendoderm precursors are located, is regulated by Nodal signaling (Liu et al., 2011; Wei et al., 2017). However, whether Net1 is involved in Nodal signal regulation and mesendoderm induction remains unknown.

As a typical Dbl family GEF, Net1 has a catalytic Dbl homology (DH) domain and an adjacent pleckstrin homology (PH) domain flanked by N- and C-terminal extensions (Alberts and Treisman, 1998). Two nuclear localization sequences (NLS) in the N-terminus function to localize Net1 predominantly to the nucleus (Schmidt and Hall, 2002). Deletion of the N-terminal or mutating the NLS results in translocation of Net1 from the nucleus to the cytosol, which is critical for RhoA activation and filamentous actin formation (Qin et al., 2005; Schmidt and Hall, 2002). Recent data have demonstrated that enzymatically inactive Net1 can bind to components of the CARD11–BCL10–MALT1 (CBM) complex and modulate nuclear factor (NF)-κB transcriptional activity, suggesting a GEF activity-independent role for Net1 (Vessichelli et al., 2012). Nuclear Net1 also activates RhoA and RhoB in response to DNA damage-associated stimulation (Dubash et al., 2011; Srougi and Burridge, 2011). Furthermore, our previous study uncovered that both nuclear and cytoplasmic Net1 enhance β-catenin S675 phosphorylation in a GEF activity-dependent manner (Wei et al., 2017). These observations raise interesting questions concerning the developmental functions of nuclear Net1, and suggest the need for further investigation.

In this study, the functions of nuclear Net1 were characterized using various gain- and loss-of-function experiments in both zebrafish embryos and mammalian cells. These studies revealed that Net1 localizes to the nucleus during gastrulation, where it associates with Smad2 in a GEF-independent manner, and ultimately facilitates p300 (also known as EP300) recruitment to the transcriptional complex to promote Smad2 activation and mesendoderm induction.

RESULTS

Zebrafish net1 is essential for mesendoderm formation

In gastrulating embryos, zebrafish net1 transcripts were present in the dorsal organizer and lateral marginal region where mesendoderm progenitors originate, and then became enriched in the axial mesoderm at the mid-gastrulation stage (75% epiboly) (Fig. 1A) (Wei et al., 2017), suggesting that net1 might take part in mesendoderm formation. We previously generated two zebrafish null mutant lines named net1Δ20 and net1Δ53, in which an integrated compensatory network is activated to buffer against the loss of net1 (Wei et al., 2017). Therefore, we disturbed net1 expression in embryos by introducing a splice-blocking morpholino (net1 MO1) whose efficiency and lack of off-target effects had been validated (Wei et al., 2017). Knockdown of net1 using 4 ng MO1 resulted in a notable decrease in the expression of mesoderm marker genes goosecoid (gsc) and no taila (ntla) at the shield and 75% epiboly stages (Fig. 1B,C). Meanwhile, endoderm marker gene expression (sox32 and sox17) was also significantly reduced in net1 morphants (Fig. 1D). These results suggest that net1 is critical for mesendoderm formation in zebrafish embryos.

Fig. 1.

net1 is crucial for mesendoderm formation in zebrafish embryos. (A) Visualization of net1 expression in zebrafish embryos at shield and 75% epiboly (ep) stages using whole-mount in situ hybridization. Shield stage, lateral views with dorsal to the right; 75% epiboly stage, dorsal views with animal pole to the top. (B–F) The expression of mesendoderm marker genes and mesendoderm derivative marker genes in cMO- and net1 MO1-injected embryos at the indicated stages. (B–D) Shield stage, animal views with dorsal to the right; 75% epiboly stage, dorsal views with animal pole to the top. (E,F) Bud stage, and 24 and 36 hpf. All images are shown in dorsal views with anterior to the top. (G) Wild-type embryos were injected with 50 pg ΔN-tcf3 mRNA alone or co-injected with 3 pg tar* mRNA and 4 ng net1 MO1, and then collected at the shield and 75% epiboly stages for whole-mount in situ hybridization with a gsc or sox17 probe. (H) Overexpression of net1 rescues mesendoderm induction defects in net1 morphants. Wild-type embryos were injected with cMO or net1 3Mix (4 ng net1 MO1, 200 pg Flag-net1 mRNA and 1 ng AS-Flag-photo-MO), and then a subset of net1 3Mix-injected embryos were exposed to UV light at the sphere stage. The expression of gsc and sox17 was examined by in situ hybridization at shield and 75% epiboly stages, respectively. In B–H, the percentages (mean±s.d.) of the affected embryos are shown as calculated from three independent biological repeats with ∼15–30 embryos in each group. UIC, uninjected control.

Fig. 1.

net1 is crucial for mesendoderm formation in zebrafish embryos. (A) Visualization of net1 expression in zebrafish embryos at shield and 75% epiboly (ep) stages using whole-mount in situ hybridization. Shield stage, lateral views with dorsal to the right; 75% epiboly stage, dorsal views with animal pole to the top. (B–F) The expression of mesendoderm marker genes and mesendoderm derivative marker genes in cMO- and net1 MO1-injected embryos at the indicated stages. (B–D) Shield stage, animal views with dorsal to the right; 75% epiboly stage, dorsal views with animal pole to the top. (E,F) Bud stage, and 24 and 36 hpf. All images are shown in dorsal views with anterior to the top. (G) Wild-type embryos were injected with 50 pg ΔN-tcf3 mRNA alone or co-injected with 3 pg tar* mRNA and 4 ng net1 MO1, and then collected at the shield and 75% epiboly stages for whole-mount in situ hybridization with a gsc or sox17 probe. (H) Overexpression of net1 rescues mesendoderm induction defects in net1 morphants. Wild-type embryos were injected with cMO or net1 3Mix (4 ng net1 MO1, 200 pg Flag-net1 mRNA and 1 ng AS-Flag-photo-MO), and then a subset of net1 3Mix-injected embryos were exposed to UV light at the sphere stage. The expression of gsc and sox17 was examined by in situ hybridization at shield and 75% epiboly stages, respectively. In B–H, the percentages (mean±s.d.) of the affected embryos are shown as calculated from three independent biological repeats with ∼15–30 embryos in each group. UIC, uninjected control.

Importantly, the formation of the notochord [as indicated by the expression of ntl and sonic hedgehog a (shha)], a derivative of the axial mesoderm, was severely reduced and ruptured in net1 morphants (Fig. 1E). Likewise, the expression of marker genes in endodermal derivatives, including liver bud, pancreas rudiment and pharyngeal pouches, was almost abolished in net1-depleted embryos (Fig. 1F). These data indicate that the mesendoderm formation defects in net1 morphants were not caused by a developmental delay.

Shortly after the mid-blastula transition, maternal β-catenin activates the transcription of the Nodal family member sqt to induce mesendodermal fate at the margin of embryos (Bellipanni et al., 2006; Zorn and Wells, 2009). We previously published that zebrafish Net1 is essential for dorsal axis specification through promotion of maternal β-catenin activation (Wei et al., 2017). In addition, net1 morphants exhibit a much smaller organizer at shield stage and ventralized phenotypes at later stages, including a variably reduced head size and shortened body axis, which partially resemble Nodal mutants (Wei et al., 2017). To confirm that mesendoderm defects in net1 morphants are not secondary effects of Wnt/β-catenin signaling inhibition, we examined the role of net1 in Wnt signaling-deficient embryos generated by injecting one-cell-stage embryos with 50 pg ΔN-tcf3 mRNA, which encodes a dominant negative form of Tcf3 (Molenaar et al., 1996). As expected, introducing the ΔN-tcf3 mRNA significantly decreased mesoderm and endoderm formation, which was restored upon reintroduction of the Nodal signal by co-injecting 3 pg tar* mRNA that encodes a constitutively active form of the Nodal type I receptor (Fig. 1G). Meanwhile, embryonic knockdown of net1 decreased Tar*-induced expression of mesendoderm marker genes (Fig. 1G). Therefore, net1 is required for mesendodermal cell fate specification in a Wnt/β-catenin signaling-independent manner.

Next, we sought to exclude possible off-target effects of net1 MO injection during mesendoderm formation. Because injection of net1 mRNA leads to very early embryonic lethality (Wei et al., 2017), an antisense photo-cleavable morpholino targeting the N-terminal Flag sequence of Flag-net1 mRNA (AS-Flag-photo-MO) was used to inhibit translation in the very early embryo. To examine the expression of mesendoderm marker genes, one-cell-stage embryos were co-injected with a mixture of net1 MO1, AS-Flag-photo-MO and Flag-net1 mRNA (denoted net1 3Mix) and exposed to UV light to turn on net1 expression at the sphere stage. As shown in Fig. 1H, reintroduction of net1 expression in morphants was sufficient to reverse the mesendoderm defects, suggesting the specificity of net1 MO1. Taken together, these results demonstrate an indispensable role for net1 in mesendoderm development.

Net1 promotes Nodal signaling during mesendoderm formation

Because depletion of Net1 suppressed Nodal receptor-induced mesendoderm induction (Fig. 1G), we expanded our analysis to embryos overexpressing Nodal ligand. As expected, injection of sqt mRNA significantly enlarged the regions of mesoderm and endoderm marker gene expression. However, upon co-injection of the net1 MO1, this Nodal ligand-induced mesendoderm expansion was notably reduced (Fig. 2A,B). These results suggest that net1 is essential for mesendoderm induction through promoting Nodal signal activity. Furthermore, embryos overexpressing net1 exhibited ectopic expression of mesoderm and endoderm markers compared to what was seen in the wild-type control, which was eliminated by treating with the Nodal signal-specific inhibitor SB431542 (Fig. 2C). In addition, the expression of fscn1a and cyclops (cyc), the genes directly targeted by the Nodal signal (Liu et al., 2016), was reduced in shield-stage net1 morphants (Fig. 2D). Therefore, Net1 is potentially a positive regulator of Nodal signal transduction during mesendoderm formation.

Fig. 2.

Net1 upregulates Nodal signal transduction. (A,B) Knockdown of net1 decreases Nodal ligand-induced mesendoderm expansion. Wild-type embryos were injected with 1 pg sqt mRNA alone or concurrently with 4 ng net1 MO1 at the one-cell stage, harvested at the shield and 75% epiboly stages, and analyzed for the expression of indicated marker genes by using whole-mount in situ hybridization. (C) Zebrafish embryos were injected with net1 2Mix (300 pg Flag-net1 mRNA and 1 ng AS-Flag-photo-MO) at the one-cell stage. A subset of these embryos was treated with 25 μM SB431542 (SB) from the 64-cell stage, of which a portion was exposed to UV light at the sphere stage. The expression of gsc and sox17 was examined by in situ hybridization at the shield and 75% epiboly stages, respectively. (D) The expression of Nodal target genes cyc and fscn1a in cMO- and net1 MO1-injected embryos at shield stage. Lateral view with animal pole to the top. In A–D, the percentages (mean±s.d.) of the affected embryos are shown as calculated from three independent biological repeats with ∼15–25 embryos in each group. UIC, uninjected control. (E,F) Overexpression of Net1 enhanced ARE-luciferase expression. (E) HeLa cells were transfected with the ARE-luciferase reporter concurrently with increasing amounts of Net1. At 36 h post transfection, the cells were treated with TGF-β1 or left untreated overnight, and then harvested for luciferase assays. *P<0.05; NS, not significant (Student's t-test). (F) At the one-cell stage, zebrafish embryos were injected with the ARE-luciferase reporter alone or together with net1 2Mix. At the sphere stage, half of the net1 2Mix-injected embryos were exposed to UV light, and then all embryos were harvested at the shield stage for luciferase assays. *P<0.05 (Student's t-test). (G) Knockdown of net1 inhibits ARE-luciferase expression. At the one-cell stage, zebrafish embryos were co-injected with the ARE-luciferase reporter and cMO or net1 MO1, and then harvested at the shield stage for luciferase assays. **P<0.01 (Student's t-test). Results in E–G are mean±s.d. (n=3).

Fig. 2.

Net1 upregulates Nodal signal transduction. (A,B) Knockdown of net1 decreases Nodal ligand-induced mesendoderm expansion. Wild-type embryos were injected with 1 pg sqt mRNA alone or concurrently with 4 ng net1 MO1 at the one-cell stage, harvested at the shield and 75% epiboly stages, and analyzed for the expression of indicated marker genes by using whole-mount in situ hybridization. (C) Zebrafish embryos were injected with net1 2Mix (300 pg Flag-net1 mRNA and 1 ng AS-Flag-photo-MO) at the one-cell stage. A subset of these embryos was treated with 25 μM SB431542 (SB) from the 64-cell stage, of which a portion was exposed to UV light at the sphere stage. The expression of gsc and sox17 was examined by in situ hybridization at the shield and 75% epiboly stages, respectively. (D) The expression of Nodal target genes cyc and fscn1a in cMO- and net1 MO1-injected embryos at shield stage. Lateral view with animal pole to the top. In A–D, the percentages (mean±s.d.) of the affected embryos are shown as calculated from three independent biological repeats with ∼15–25 embryos in each group. UIC, uninjected control. (E,F) Overexpression of Net1 enhanced ARE-luciferase expression. (E) HeLa cells were transfected with the ARE-luciferase reporter concurrently with increasing amounts of Net1. At 36 h post transfection, the cells were treated with TGF-β1 or left untreated overnight, and then harvested for luciferase assays. *P<0.05; NS, not significant (Student's t-test). (F) At the one-cell stage, zebrafish embryos were injected with the ARE-luciferase reporter alone or together with net1 2Mix. At the sphere stage, half of the net1 2Mix-injected embryos were exposed to UV light, and then all embryos were harvested at the shield stage for luciferase assays. *P<0.05 (Student's t-test). (G) Knockdown of net1 inhibits ARE-luciferase expression. At the one-cell stage, zebrafish embryos were co-injected with the ARE-luciferase reporter and cMO or net1 MO1, and then harvested at the shield stage for luciferase assays. **P<0.01 (Student's t-test). Results in E–G are mean±s.d. (n=3).

To confirm the role of Net1 in Nodal signal transduction, the effects of Net1 on the expression of ARE-luciferase, a TGF-β/Nodal-responsive reporter, were investigated in mammalian cells and zebrafish embryos. In HeLa cells, Net1 overexpression upregulated the luciferase activity of this reporter in a dose-dependent manner (Fig. 2E). In addition, the luciferase activity of the reporter was also significantly increased in zebrafish embryos overexpressing Net1 (Fig. 2F). Conversely, Net1-deficient embryos displayed a greater reduction in reporter gene expression (Fig. 2G). Taken together, these results suggest that Net1 promotes mesendoderm formation through upregulating the Nodal signal.

Nuclear Net1 potentiates the Nodal signal independently of its GEF activity

Previous work has revealed that epitope-tagged Net1 localizes mainly in the nucleus of Xenopus embryonic cells (Miyakoshi et al., 2004). To examine the subcellular localization of endogenous Net1 during embryonic development, nuclear and cytoplasmic proteins were extracted from zebrafish embryos and examined by western blotting. It was observed that endogenous Net1 was primarily in the nucleus during different developmental stages of zebrafish embryogenesis (Fig. 3A). It has been shown that TGF-β stimulates the transfer of Net1 from the nucleus to the cytoplasm in human retinal pigment epithelial cells (Lee et al., 2010). Therefore, we determined whether Nodal signaling similarly influences Net1 translocation in zebrafish embryos. Surprisingly, despite a moderate increase in Net1 expression in sqt mRNA-injected embryos, the subcellular localization of Net1 did not change upon Nodal signal activation (Fig. 3B). The Nodal signal is specifically activated in the blastoderm margin (Bennett et al., 2007), suggesting that the other cells in zebrafish gastrulas are deficient in this signal. Since the cytoplasmic Net1 was almost undetectable in wild-type embryos (Fig. 3A,B), Net1 is rarely present in the cytoplasm in the absence of Nodal signal. Thus, the increased nuclear Net1 proteins in sqt mRNA-injected embryos might be the result of the upregulated expression of net1, as it is a direct Nodal-targeted gene (Liu et al., 2011; Wei et al., 2017), and not the nuclear translocation of cytoplasmic Net1 induced by Nodal signal. Taken together, these data reveal a predominant nuclear localization of Net1 during early embryonic development, which is not changed in response to Nodal signal stimulation.

Fig. 3.

Nuclear Net1 enhances Nodal signaling and mesendoderm formation in a GEF-independent manner. (A,B) Endogenous Net1 primarily localizes in the zebrafish embryonic cell nucleus. Nuclear and cytosolic fractions were extracted from wild-type embryos at indicated stages (A) or from sqt mRNA-injected embryos at the shield stage (B). Then these resulting samples were assessed by western blotting with the indicated antibodies. (C) Net1 promotes Nodal signaling in the nucleus. HeLa cells were transfected with the ARE-luciferase reporter together with Net1, Net1-NLS or Net1-ΔN-NES. After 36 h of transfection, the cells were treated with or without TGF-β1 overnight, then harvested for luciferase assays. **P<0.01; ***P<0.001; NS, not significant (Student's t-test). (D) Nuclear Net1 facilitates mesendoderm induction. At the sphere stage, zebrafish embryos injected with net1 2Mix or net1-NLS 2Mix or net1-ΔN NES 2Mix were exposed to UV light, and then harvested for whole-mount in situ hybridization to detect the expression of gsc and sox17 at the shield and 75% epiboly stages. (E,F) Overexpression of the nuclear, but not cytosolic, Net1 rescues the mesendoderm defects in net1 morphants. Embryos were injected with cMO, net1 MO1, net1-NLS 3Mix or net1-ΔN NES 3Mix. Embryos injected with net1-NLS 3Mix or net1-ΔN NES 3Mix were exposed to UV light at the sphere stage. The expression of gsc (E) and sox17 (F) was examined by in situ hybridization at the shield and 75% epiboly stages, respectively. (G,H) Net1 promotes Nodal signaling independently of its GEF activity. HeLa cells were co-transfected with the ARE-luciferase reporter and wild-type Net1 and its GEF-deficient mutants (G) or nuclear Net1 and its GEF-deficient mutants (H). After 36 h of transfection, the cells were treated with or without TGF-β1 overnight, then harvested for luciferase assays. *P<0.05, **P<0.01, ***P<0.001 (Student's t-test). (I) The influence of nuclear Net1 and its GEF-deficient mutants in mesendoderm formation. Zebrafish embryos were injected with net1-NLS 2Mix, NLS-L266E 2Mix and NLS-W437L 2Mix, respectively. These embryos were exposed to UV at the sphere stage, then harvested for whole-mount in situ hybridization. In D,E,F and I, the percentages (mean±s.d.) of the affected embryos are shown as calculated from three independent biological repeats with ∼20–30 embryos in each group. Results in C,G and H are mean±s.d. (n=3). UIC, uninjected control.

Fig. 3.

Nuclear Net1 enhances Nodal signaling and mesendoderm formation in a GEF-independent manner. (A,B) Endogenous Net1 primarily localizes in the zebrafish embryonic cell nucleus. Nuclear and cytosolic fractions were extracted from wild-type embryos at indicated stages (A) or from sqt mRNA-injected embryos at the shield stage (B). Then these resulting samples were assessed by western blotting with the indicated antibodies. (C) Net1 promotes Nodal signaling in the nucleus. HeLa cells were transfected with the ARE-luciferase reporter together with Net1, Net1-NLS or Net1-ΔN-NES. After 36 h of transfection, the cells were treated with or without TGF-β1 overnight, then harvested for luciferase assays. **P<0.01; ***P<0.001; NS, not significant (Student's t-test). (D) Nuclear Net1 facilitates mesendoderm induction. At the sphere stage, zebrafish embryos injected with net1 2Mix or net1-NLS 2Mix or net1-ΔN NES 2Mix were exposed to UV light, and then harvested for whole-mount in situ hybridization to detect the expression of gsc and sox17 at the shield and 75% epiboly stages. (E,F) Overexpression of the nuclear, but not cytosolic, Net1 rescues the mesendoderm defects in net1 morphants. Embryos were injected with cMO, net1 MO1, net1-NLS 3Mix or net1-ΔN NES 3Mix. Embryos injected with net1-NLS 3Mix or net1-ΔN NES 3Mix were exposed to UV light at the sphere stage. The expression of gsc (E) and sox17 (F) was examined by in situ hybridization at the shield and 75% epiboly stages, respectively. (G,H) Net1 promotes Nodal signaling independently of its GEF activity. HeLa cells were co-transfected with the ARE-luciferase reporter and wild-type Net1 and its GEF-deficient mutants (G) or nuclear Net1 and its GEF-deficient mutants (H). After 36 h of transfection, the cells were treated with or without TGF-β1 overnight, then harvested for luciferase assays. *P<0.05, **P<0.01, ***P<0.001 (Student's t-test). (I) The influence of nuclear Net1 and its GEF-deficient mutants in mesendoderm formation. Zebrafish embryos were injected with net1-NLS 2Mix, NLS-L266E 2Mix and NLS-W437L 2Mix, respectively. These embryos were exposed to UV at the sphere stage, then harvested for whole-mount in situ hybridization. In D,E,F and I, the percentages (mean±s.d.) of the affected embryos are shown as calculated from three independent biological repeats with ∼20–30 embryos in each group. Results in C,G and H are mean±s.d. (n=3). UIC, uninjected control.

Next, it was determined whether nuclear Net1 is responsible for Nodal signal transduction and mesendoderm induction. First, the effects of nuclear and cytoplasmic Net1 on the activity of ARE-luciferase reporter were examined in HeLa cells upon expression of Net1-NLS and Net1-ΔN NES constructs, which are restricted to the nucleus and cytoplasm, respectively (Wei et al., 2017). Interestingly, Net1-NLS potentiated luciferase activity equally to or better than the wild-type protein, while Net1-ΔN NES did not affect reporter gene expression (Fig. 3C), suggesting nuclear Net1, rather than cytosolic Net1, enhances the TGF-β signal in mammalian cells. Next, we tested the influence of Net1 subcellular localization on mesendoderm development in zebrafish embryos. As shown in Fig. 3D, nuclear, but not cytosolic, Net1 clearly promoted mesendoderm formation. Furthermore, the mesendoderm defects in net1 morphants were almost totally eliminated by overexpression of the nuclear but not cytosolic Net1 (Fig. 3E,F), indicating a critical requirement of the nuclear Net1 during zebrafish mesendoderm formation.

To investigate whether Net1 GEF activity is necessary for Nodal signal transduction, we examined two GEF-deficient zebrafish Net1 mutants named as Net1-L266E and Net1-W437L, which contain point mutations in the DH and PH domains, respectively (Wei et al., 2017). HeLa cells were co-transfected with the ARE-luciferase reporter and either wild-type Net1 or one of the Net1 mutants. Upon examination of the lysates from these transfected cells, the cells expressing Net1-L266E or Net1-W437L displayed similar TGF-β1-induced levels of ARE-luciferase expression to the wild-type Net1 (Fig. 3G). Moreover, nuclear Net1 (Net1-NLS) and its GEF mutants (Net1-NLS L266E and Net1-NLS W437L) contributed equally to the expression of the reporter gene in HeLa cells and mesendoderm formation in zebrafish embryos (Fig. 3H,I). This clearly indicates that nuclear Net1 enhances Nodal signaling and mesendoderm formation independently of its GEF activity.

Net1 upregulates the Nodal signal in parallel with or downstream of Smad2

Next, we determined which step of Nodal signal transduction is regulated by Net1. Because Net1 overexpression increased TGF-β1-induced ARE-luciferase expression and net1 knockdown in embryos decreased sqt-activated mesendoderm formation (Fig. 2A,B,E), the regulation of Nodal signal by Net1 is unlikely to occur at the ligand level. Next, it was tested whether Net1 affects Nodal signaling at the receptor level. A constitutively active form of TGF-β type I receptor ALK5 (CA-ALK5; ALK5 is also known as TGFBR1) was used to activate ARE-luciferase expression in HeLa cells. Net1 overexpression augmented this activation (Fig. 4A). Meanwhile, the enhanced expression of gsc and sox17 in zebrafish embryos mediated by constitutively activated Nodal type I receptor tar* was eliminated by the introduction of net1 MO1 (Fig. 4B). These results rule out the possibility that Net1-mediated regulation of TGF-β/Nodal signal acts at the receptor level.

Fig. 4.

Net1 promotes Nodal signal downstream of or parallel to Smad2. (A) Overexpression of Net1 promotes CA-ALK5-induced ARE-luciferase expression. At 36 h after transfecting with ARE-luciferase reporter together with increasing amounts of Net1 expression plasmids, HeLa cells were re-transfected with or without CA-ALK5 for 24 h, and then harvested for luciferase assays. *P<0.05; **P<0.01 (Student's t-test). (B) One-cell-stage zebrafish embryos were injected with 3 pg tar* mRNA alone or together with 4 ng net1 MO1, and then harvested for whole-mount in situ hybridization. (C) HeLa cells were co-transfected with the ARE-luciferase reporter and indicated plasmids. At 36 h after transfection, the cells were harvested for luciferase assays. **P<0.01 (Student's t-test). (D) One-cell-stage zebrafish embryos were injected with 25 pg ca-smad2 mRNA alone or together with 4 ng net1 MO1, and then harvested for whole-mount in situ hybridization. (E) Overexpression of Net1 does not influence the subcellular localization of Smad2. HeLa cells transfected with Flag–Net1 were treated with or without TGF-β1 for 1 h, and then immunostained with anti-Flag (green) and anti-Smad2 (red) antibodies. Nuclei were counterstained with DAPI and are shown in blue. Scale bar: 10 μm. (F) Overexpression of Net1 does not affect the phosphorylation of Smad2. HeLa cells were transfected with Flag-tagged wild-type (WT) Net1 or the Net1-L266E mutant. After 36 h of transfection, these cells were treated with TGF-β1 or left untreated for 4 h, and then harvested for western blotting using the indicated antibodies. Results in A and C are mean±s.d. (n=3). In B and D, the percentages (mean±s.d.) of the affected embryos are shown as calculated from three independent biological repeats with ∼20–25 embryos in each group. UIC, uninjected control.

Fig. 4.

Net1 promotes Nodal signal downstream of or parallel to Smad2. (A) Overexpression of Net1 promotes CA-ALK5-induced ARE-luciferase expression. At 36 h after transfecting with ARE-luciferase reporter together with increasing amounts of Net1 expression plasmids, HeLa cells were re-transfected with or without CA-ALK5 for 24 h, and then harvested for luciferase assays. *P<0.05; **P<0.01 (Student's t-test). (B) One-cell-stage zebrafish embryos were injected with 3 pg tar* mRNA alone or together with 4 ng net1 MO1, and then harvested for whole-mount in situ hybridization. (C) HeLa cells were co-transfected with the ARE-luciferase reporter and indicated plasmids. At 36 h after transfection, the cells were harvested for luciferase assays. **P<0.01 (Student's t-test). (D) One-cell-stage zebrafish embryos were injected with 25 pg ca-smad2 mRNA alone or together with 4 ng net1 MO1, and then harvested for whole-mount in situ hybridization. (E) Overexpression of Net1 does not influence the subcellular localization of Smad2. HeLa cells transfected with Flag–Net1 were treated with or without TGF-β1 for 1 h, and then immunostained with anti-Flag (green) and anti-Smad2 (red) antibodies. Nuclei were counterstained with DAPI and are shown in blue. Scale bar: 10 μm. (F) Overexpression of Net1 does not affect the phosphorylation of Smad2. HeLa cells were transfected with Flag-tagged wild-type (WT) Net1 or the Net1-L266E mutant. After 36 h of transfection, these cells were treated with TGF-β1 or left untreated for 4 h, and then harvested for western blotting using the indicated antibodies. Results in A and C are mean±s.d. (n=3). In B and D, the percentages (mean±s.d.) of the affected embryos are shown as calculated from three independent biological repeats with ∼20–25 embryos in each group. UIC, uninjected control.

By using the same strategy as above, we found that constitutively active Smad2 (CA-Smad2)-induced expression of the reporter gene and mesoderm and endoderm markers were notably changed by Net1 overexpression or depletion (Fig. 4C,D). In addition, neither ectopic wild-type Net1 nor its GEF-deficient mutants affected Smad2 nuclear translocation and phosphorylation in response to TGF-β1 stimulation in HeLa cells (Fig. 4E,F). Taken together, these results suggest that Net1 acts in parallel with or downstream of Smad2.

Net1 interacts with nuclear Smad2 in a GEF-activity-independent manner

Since Net1 acts in parallel with or downstream of Smad2, it was hypothesized that Net1 interacts with Smad2. To test this, epitope-tagged Net1 and Smad2 were overexpressed in HEK293T cells and co-immunoprecipitations were performed, revealing a physical interaction between Net1 and Smad2 (Fig. 5A,B). As HeLa cells are more responsive to TGF-β stimulation than HEK293T cells, we examined the association between Net1 and Smad2 in HeLa cells in the absence or presence of TGF-β1. We found that overexpressed Net1 was able to interact with endogenous Smad2 (Fig. 5C). Interestingly, this Net1–Smad2 association was enhanced by TGF-β1 treatment (Fig. 5D). Consistent with this, endogenous Net1 in zebrafish gastrula stage embryos could be co-immunoprecipitated with an anti-Smad2 antibody, indicating that Net1 and Smad2 form a complex in vivo, and the endogenous Net1–Smad2 interaction was obviously strengthened in sqt mRNA-injected embryos (Fig. 5E). Since Net1 is mainly located in the nucleus (Figs 4E and 5F) and has a higher affinity for activated Smad2 (Fig. 5C,E), we speculate that the Net1–Smad2 binding occurs in the nucleus. To address this hypothesis, bimolecular fluorescence complementation (BiFC) assay was performed in HeLa cells, and the reconstituted fluorescent YFP from YC-Smad2 (fusion of Smad2 to C-terminal half of YFP) and YN-Net1 (fusion of N-terminal half of YFP to Net1) was observed in the nuclei but not in the cytosol (Fig. 5G). These observations strongly support the idea that Net1 associates with nuclear Smad2.

Fig. 5.

Net1 interacts with Smad2. (A,B) Flag-tagged Net1 interacts with overexpressed Smad2. HEK293T cells were transfected with Flag–Net1 and Myc–Smad2 expression plasmids, then harvested for immunoprecipitation with anti-Flag M2 agarose beads (A) or anti-Myc agarose beads (B). Total lysates and immunoprecipitation (IP) samples were analyzed by western blotting with the indicated antibodies. (C,D) Overexpressed Net1 interacts with endogenous Smad2. HeLa cells transfected with Flag-tagged Net1 were treated with or without TGF-β1 for 4 h, and then harvested for immunoprecipitation with anti-Flag M2 agarose beads. Note that TGF-β1 treatment enhanced the interaction between Flag–Net1 and endogenous Smad2 (D). (E) The physical interaction of endogenous Net1 and Smad2. Whole wild-type embryos were lysed with TNE at the shield stage for immunoprecipitation with anti-Net1 antibody. sqt mRNA was injected when indicated. (F,G) Detection of the Net1–Smad2 interaction in living cells by means of a BiFC assays. YC-Smad2 and YN-Net1 were transfected alone or simultaneously into HeLa cells. The expression of YN-Net1 or YC-Smad2 was detected by immunostaining with an anti-GFP antibody and DyLight 549-conjugated secondary antibody (F). The reconstituted YFP fluorescence was detected by confocal microscopy at 488 nm (G). Scale bars: 10 μm. (H) Smad2 binds to Net1 through its L-MH2 domain. HEK293T cells were co-transfected with Flag–Net1 and Myc-tagged Smad2 deletion mutants, and then harvested for immunoprecipitation. (I) Net1 interacts with Smad2 independently of its GEF activity. HEK293T cells were co-transfected with Myc–Smad2 and Flag-tagged wild-type (WT) Net1 or Net1 GEF-deficient mutants, and then harvested for immunoprecipitation.

Fig. 5.

Net1 interacts with Smad2. (A,B) Flag-tagged Net1 interacts with overexpressed Smad2. HEK293T cells were transfected with Flag–Net1 and Myc–Smad2 expression plasmids, then harvested for immunoprecipitation with anti-Flag M2 agarose beads (A) or anti-Myc agarose beads (B). Total lysates and immunoprecipitation (IP) samples were analyzed by western blotting with the indicated antibodies. (C,D) Overexpressed Net1 interacts with endogenous Smad2. HeLa cells transfected with Flag-tagged Net1 were treated with or without TGF-β1 for 4 h, and then harvested for immunoprecipitation with anti-Flag M2 agarose beads. Note that TGF-β1 treatment enhanced the interaction between Flag–Net1 and endogenous Smad2 (D). (E) The physical interaction of endogenous Net1 and Smad2. Whole wild-type embryos were lysed with TNE at the shield stage for immunoprecipitation with anti-Net1 antibody. sqt mRNA was injected when indicated. (F,G) Detection of the Net1–Smad2 interaction in living cells by means of a BiFC assays. YC-Smad2 and YN-Net1 were transfected alone or simultaneously into HeLa cells. The expression of YN-Net1 or YC-Smad2 was detected by immunostaining with an anti-GFP antibody and DyLight 549-conjugated secondary antibody (F). The reconstituted YFP fluorescence was detected by confocal microscopy at 488 nm (G). Scale bars: 10 μm. (H) Smad2 binds to Net1 through its L-MH2 domain. HEK293T cells were co-transfected with Flag–Net1 and Myc-tagged Smad2 deletion mutants, and then harvested for immunoprecipitation. (I) Net1 interacts with Smad2 independently of its GEF activity. HEK293T cells were co-transfected with Myc–Smad2 and Flag-tagged wild-type (WT) Net1 or Net1 GEF-deficient mutants, and then harvested for immunoprecipitation.

To determine which domain of Smad2 is responsible for binding with Net1, truncated Smad2 mutants expressing the MH1 domain, MH2 domain, MH1 domain and linker region (MH1-L), and linker region and MH2 domain (L-MH2), respectively, were constructed. Domain mapping revealed that the linker region and MH2 domain of Smad2 were both required for the interaction between Smad2 and Net1 (Fig. 5H). In addition, in agreement with the GEF-independent function of Net1 in Nodal signaling, wild-type Net1 and its GEF-deficient mutants bound to Smad2 with a similar affinity (Fig. 5I). Thus, Net1 interacts with Smad2 in a GEF-activity-independent manner.

The interaction of Net1 and Smad2 facilitates p300 recruitment

Net1 promotes Nodal signaling without affecting the nuclear translocation or phosphorylation of Smad2 (Fig. 4E,F). Therefore, it was hypothesized that Net1 enhances Smad2 transcriptional activity. Histone acetylation is essential for the regulation of transcription (Struhl, 1998). The histone acetyl transferase p300 is a general transcriptional coactivator, which functions by loosening the chromatin in an acetylation-dependent manner, while histone deacetylases (HDACs) act as transcriptional repressors (Ogryzko et al., 1996; Taunton et al., 1996; Wolffe, 1996). Interactions of the Smad complex with p300 or HDACs represent transcriptional activation or repression, respectively (Feng et al., 1998; Janknecht et al., 1998; Pouponnot et al., 1998; Wotton et al., 1999). Therefore, it was examined whether Net1 affects the association of Smad2 with p300 or HDAC1 by performing co-immunoprecipitation experiments. As shown in Fig. 6A, Smad2 could bind to HDAC1, and this binding was dramatically decreased upon Net1 overexpression. By contrast, overexpression of Net1 resulted in a clear promotion of Smad2–p300 interactions (Fig. 6B), indicating that Net1 promotes Smad2 transcriptional activity by enhancing p300 recruitment. Interestingly, compared to wild-type Net1, Net1 GEF mutants had similar effects on the associations of Smad2 with HDAC1 or p300 (Fig. 6A,B), suggesting that the contribution of the GEF activity of Net1 was negligible in the regulation of Smad2 transcriptional activity. This may explain why Net1 promotes Nodal signaling and mesendoderm induction independently of its GEF activity. Taken together, these results suggest that nuclear Net1 interacts with Smad2 in a GEF-activity-independent manner. Net1–Smad2 interaction induces the dissociation of the corepressor HDAC1 from and recruitment of the coactivator p300 to Smad transcriptional complexes, thereby activating Nodal target gene transcription during mesendoderm development (Fig. 6C).

Fig. 6.

Net1 increases the affinity of Smad2 for p300. (A,B) Net1 inhibits the interactions between Smad2 and HDAC1 (A), but enhances the association of Smad2 with p300 (B). HEK293T cells were transfected with indicated plasmids, and then harvested for immunoprecipitation with anti-Flag M2 agarose beads. Total lysates and immunoprecipitation (IP) samples were analyzed by western blotting using the indicated antibodies. (C) Proposed model of the activation of the Nodal signal by Net1. When Net1 is absent, Smad2 interacts with the corepressor HDAC1, which results in transcriptional inhibition. When Net1 is present, Net1–Smad2 interaction induces the recruitment of the coactivator p300, which ultimately leads to the expression of Nodal target genes during mesendoderm formation.

Fig. 6.

Net1 increases the affinity of Smad2 for p300. (A,B) Net1 inhibits the interactions between Smad2 and HDAC1 (A), but enhances the association of Smad2 with p300 (B). HEK293T cells were transfected with indicated plasmids, and then harvested for immunoprecipitation with anti-Flag M2 agarose beads. Total lysates and immunoprecipitation (IP) samples were analyzed by western blotting using the indicated antibodies. (C) Proposed model of the activation of the Nodal signal by Net1. When Net1 is absent, Smad2 interacts with the corepressor HDAC1, which results in transcriptional inhibition. When Net1 is present, Net1–Smad2 interaction induces the recruitment of the coactivator p300, which ultimately leads to the expression of Nodal target genes during mesendoderm formation.

DISCUSSION

Nodal ligands emerged as endogenous mesendoderm inducers in different vertebrate embryos (Feldman et al., 1998; Gritsman et al., 1999; Zhou et al., 1993). Interestingly, loss of Smad2 disrupts the development of the mesoderm and endoderm, while Smad3 inactivation does not result in early embryonic defects. This suggests that Smad2 is the major downstream mediator of Nodal signaling during mesendoderm formation (Nomura and Li, 1998; Weinstein et al., 1998; Yang et al., 1999; Zhu et al., 1998). We previously demonstrated that zebrafish net1 is a direct target of the Nodal pathway during gastrulation (Liu et al., 2011). Many Nodal target genes have been identified as important feedback regulatory factors involved in the control of Nodal-mediated embryonic development (Liu et al., 2016; Shen, 2007). In light of these reports, we set out to determine whether Net1 functions in the formation of mesoderm and endoderm by regulating Nodal signal transduction. Our data suggest that Net1 is expressed in mesendoderm progenitors and required for the specification of mesendodermal cell fates through promoting Smad2 transcriptional activation. However, zebrafish net1 mutants exhibit no detectable developmental defects, and a set of genes encoding GEFs, GTPases and GTPase effectors are upregulated and suspected to compensate for the loss of net1 (Wei et al., 2017). It will be interesting to further investigate the compensatory mechanisms activated in the mutants.

Full-length mammalian Net1, a RhoA guanine nucleotide exchange factor, mainly localizes in the nucleus in quiescent cells. Meanwhile, the oncogenic form of this protein is found in the cytoplasm and enhances cellular proliferation and tumorigenesis by activating RhoA (Alberts and Treisman, 1998; Qin et al., 2005; Schmidt and Hall, 2002). Likewise, Xenopus Net1 is in the nucleus and the activation of the non-canonical Wnt signaling pathway does not alter its localization in animal cap cells (Miyakoshi et al., 2004). Interestingly, our study shows that zebrafish Net1 also predominantly localizes to the nucleus, and its location does not change in response to Nodal signal stimulation. However, TGF-β signaling has been reported to induce redistribution of Net1 to the cytoplasm in human retinal pigment epithelial cells, which is required for TGF-β-induced epithelial–mesenchymal transition (EMT) (Lee et al., 2010). Therefore, it is possible that the different effects of TGF-β signaling on Net1 subcellular translocation may be dependent on the context.

Thus far, the prevailing model is that Net1 export from the nucleus to the cytoplasm leads to RhoA activation through its GEF activity. However, recent studies have suggested that Net1 is required for the activation of RhoA and RhoB in the nucleus in response to DNA damage (Dubash et al., 2011; Srougi and Burridge, 2011). In zebrafish, based on the nuclear localization of Net1 and the severe defects in mesendoderm formation noted in net1 morphants, we hypothesized that Net1 might have a previously unidentified function in the nucleus. Indeed, we found that nuclear Net1 plays a role in Nodal signal transduction and mesendoderm development. We further demonstrated that Net1 interacts with Smad2 in the nucleus and promotes Smad2 activation through p300 recruitment in a GEF-independent manner. Importantly, to further support our hypothesis, the overexpression of net1 had did not promote mesendoderm formation in embryos treated with SB431542, as Smad2 could not be phosphorylated and translocated into the nucleus when Nodal signal was inhibited. Thus, nuclear Net1 has at least two functions: (1) activating nuclear Rho GTPases as a GEF in response to DNA damage, and (2) regulating gene expression as a scaffold protein in Smad2-containing transcriptional complexes that direct cell fate during early embryogenesis. Identification of other nuclear binding partners of Net1 will be important for further understanding of its functions within the nucleus.

Several previous studies have revealed a link between the maternal Wnt/β-catenin pathway and early mesendoderm induction. It has been well established that maternal β-catenin is required for the expression of dorsal genes, such as Nodal ligands, and for the activation of MAPK and the mesodermal markers Xbra and eomesodermin (Bellipanni et al., 2006; Schohl and Fagotto, 2003; Zorn and Wells, 2009). We previously reported that Net1, as well as its GEF activity, is critical for maternal β-catenin activation (Wei et al., 2017). In this study, extensive experimental evidence and analysis established that Net1 promotes mesendoderm formation via activation of Smad2 transcriptional activity. Several lines of evidence suggest that the mesendoderm defects in net1-depleted embryos, rather than the secondary effects of the suppression of Wnt/β-catenin signal, are the direct results of the reduction of Nodal activity. First, when Wnt/β-catenin activity is blocked by ΔN-tcf3 expression, Net1 depletion significantly decreases Nodal receptor-induced mesendoderm formation. Second, both cytosolic and nuclear Net1 promote β-catenin activation (Wei et al., 2017), but only nuclear Net1 is responsible for promoting Smad2 transcriptional activity. Third, Net1 GEF activity is indispensable for β-catenin activation, but is unnecessary for Nodal signal transduction. Therefore, Net1 integrates Wnt/β-catenin and Nodal signals by shifting from its role as a GEF to that of a scaffold protein during embryonic development.

MATERIALS AND METHODS

Zebrafish maintenance

Wild-type embryos were obtained from zebrafish (Tuebingen strain) matings. Adult zebrafish and embryos were raised and maintained under standard laboratory conditions. Embryo stages were determined by morphology as previously described (Kimmel et al., 1995). All zebrafish experiments were in strict accordance with the Regulations for the Care and Use of Laboratory Animals as published by the Ministry of Science and Technology of China, and the Institute of Zoology's Guidelines for the Care and Use of Laboratory Animals.

mRNAs, morpholinos and microinjection

Capped mRNAs of zebrafish net1, tar*, sqt, ca-smad2 and ΔN-tcf3 were synthesized in vitro from the corresponding linearized plasmids by using mMessage mMachine kits (Ambion). The morpholinos were designed and synthesized by Gene Tools. net1 MO1 was 5′-CTTGCTCCGGCTGTACTCACCTCTT-3′, control MO (mis-MO1) was 5′-CTTCCTGCGCCTGTAGTCACGTCTT-3′, and AS-Flag-photo-MO was 5′-TCATCGTCGTpCTTGTAGTCCAT-3′. To overexpress net1 in zebrafish, Flag-net1 mRNA and AS-Flag-photo-MO were premixed then co-injected into one-cell-stage embryos. Once at the sphere stage, these embryos were exposed to 365 nm UV light for 10 min using a Lightbox (Gene Tools, LLC) and then harvested at the shield or 75% epiboly stages.

Whole-mount in situ hybridization

Digoxigenin-UTP-labeled antisense RNA probes were synthesized in vitro by using a MEGAscript® Kit (Ambion) according to the manufacturer's instructions. Whole-mount in situ hybridizations were performed as previously reported (Liu et al., 2011).

Cell lines and transfections

HEK293T (CRL-3216, ATCC) and HeLa cell lines (CCL-2, ATCC) were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin at 37°C in a humidified incubator with 5% CO2. Cell transfections were performed by using Lipofectamine 2000 (11668019, Invitrogen) according to the manufacturer's instructions.

Dual-luciferase reporter assays

For the luciferase reporter assays in mammalian cells, HeLa cells were transfected with ARE-luciferase (provided by Prof. Yeguang Chen at Tsinghua University, China) and the indicated plasmids together with a Renilla luciferase reporter, which serves as an internal control. At 36 h post transfection, cells were incubated in DMEM supplemented with 2% FBS and 5 ng/ml TGF-β1 (240-B, R&D Systems) overnight, and then harvested in order to determine luciferase activity assays.

For the luciferase reporter assays in zebrafish embryos, ARE and Renilla luciferase plasmids and the indicated mRNAs or morpholinos were co-injected into one-cell-stage embryos, and then the embryos were harvested at the shield stage for the luciferase activity assays. Each luciferase reporter assay was performed in triplicate and the data represent the mean±s.d. of three independent biological repeats after normalization to Renilla activity.

Immunoprecipitation and western blot analysis

HEK293T or HeLa cells were transfected with the indicated plasmids. At 48 h post transfection, the cells were harvested and lysed with TNE lysis buffer (10 mM Tris-HCl, pH 7.5, 150 mM NaCl, 2 mM EDTA, and 0.5% Nonidet P-40) containing protease inhibitors. Immunoprecipitation and western blots were then performed on the resulting cell lysates as previously described (Liu et al., 2013).

For the immunoprecipitations, anti-Flag M2 (1:100, A2220, Sigma) or anti-Myc agarose beads (1:100, A7470, Sigma) were used. For western blots, affinity-purified anti-Flag (1:5000, F2555, Sigma), anti-HA (1:3000, CW0092A, CW), anti-Myc (1:3000, M047-3, MBL), anti-Smad2 (1:1000, 3122, Cell Signaling Technology) and anti-phospho-Smad2 (Ser245/250/255) (1:500; 3104, Cell Signaling Technology) antibodies were used. The rabbit polyclonal anti-Net1 antibody was generated by our laboratory. An epitope corresponding to residues SRGEQDLIEDLKLARKAC of zebrafish Net1 was chosen for immunization. This polyclonal antibody was affinity purified and validated for specificity by peptide competition assays. The purified antibody (concentration, 200 μg/ml) was used at a dilution of 1:1000 for western blots and 1:100 for protein coimmunoprecipitations in our previous study and current work (Wei et al., 2017).

Immunofluorescence analysis

HeLa cells were cultured on coverslips, fixed with 4% paraformaldehyde in PBS for 15 min at room temperature, and then permeabilized with 0.2% Triton X-100 in PBS for 10 min. After blocking for 1 h in 5% FBS in PBS, cells were incubated with anti-Flag M2 (1:1000, A2220, Sigma), anti-Smad2 (1:100, 5339, Cell Signaling Technology) or anti-GFP (1:1000, A11122, Invitrogen) antibodies for 4 h, followed by fluorescently conjugated Alexa secondary antibodies for 2 h. Cells were counterstained with DAPI (10236276001, Sigma) to visualize nuclei. All immunofluorescent images were captured with a Zeiss LSM780 inverted confocal microscope using the same settings for all experiments.

BiFC assay

For BiFC assay, Net1 was fused to the N-terminal half of YFP (YN-Net1) and Smad2 was fused to the C-terminal half of YFP (YC-Smad2). YN-Net1 and YC-Smad2 were individually or together transfected into HeLa cells. YFP fluorescence was detected 48 h after transfection with a Zeiss LSM780 inverted confocal microscope.

Acknowledgements

We are grateful to members of the Qiang Wang laboratory for assistance and discussion.

Footnotes

Author contributions

Conceptualization: Q.W.; Validation: S.W.; Formal analysis: Q.W.; Investigation: S.W., G.N., L.L., Y.Y., S.Y., Y.C.; Writing - original draft: S.W.; Writing - review & editing: Q.W.; Supervision: Q.W.; Project administration: Q.W.; Funding acquisition: Q.W.

Funding

This work was supported by grants from the Ministry of Science and Technology of the People's Republic of China (National Key Research and Development Program of China) (2016YFA0100503) and the National Natural Science Foundation of China (31271532, 31571501).

References

Alberts
,
A. S.
and
Treisman
,
R.
(
1998
).
Activation of RhoA and SAPK/JNK signalling pathways by the RhoA-specific exchange factor mNET1
.
EMBO J.
17
,
4075
-
4085
.
Bellipanni
,
G.
,
Varga
,
M.
,
Maegawa
,
S.
,
Imai
,
Y.
,
Kelly
,
C.
,
Myers
,
A. P.
,
Chu
,
F.
,
Talbot
,
W. S.
and
Weinberg
,
E. S.
(
2006
).
Essential and opposing roles of zebrafish beta-catenins in the formation of dorsal axial structures and neurectoderm
.
Development
133
,
1299
-
1309
.
Bennett
,
J. T.
,
Joubin
,
K.
,
Cheng
,
S.
,
Aanstad
,
P.
,
Herwig
,
R.
,
Clark
,
M.
,
Lehrach
,
H.
and
Schier
,
A. F.
(
2007
).
Nodal signaling activates differentiation genes during zebrafish gastrulation
.
Dev. Biol.
304
,
525
-
540
.
Chan
,
A. M.
,
Takai
,
S.
,
Yamada
,
K.
and
Miki
,
T.
(
1996
).
Isolation of a novel oncogene, NET1, from neuroepithelioma cells by expression cDNA cloning
.
Oncogene
12
,
1259
-
1266
.
Dubash
,
A. D.
,
Guilluy
,
C.
,
Srougi
,
M. C.
,
Boulter
,
E.
,
Burridge
,
K.
and
García-Mata
,
R.
(
2011
).
The small GTPase RhoA localizes to the nucleus and is activated by Net1 and DNA damage signals
.
PLoS ONE
6
,
e17380
.
Feldman
,
B.
,
Gates
,
M. A.
,
Egan
,
E. S.
,
Dougan
,
S. T.
,
Rennebeck
,
G.
,
Sirotkin
,
H. I.
,
Schier
,
A. F.
and
Talbot
,
W. S.
(
1998
).
Zebrafish organizer development and germ-layer formation require nodal-related signals
.
Nature
395
,
181
-
185
.
Feng
,
X.-H.
,
Zhang
,
Y.
,
Wu
,
R.-Y.
and
Derynck
,
R.
(
1998
).
The tumor suppressor Smad4/DPC4 and transcriptional adaptor CBP/p300 are coactivators for smad3 in TGF-beta-induced transcriptional activation
.
Genes Dev.
12
,
2153
-
2163
.
Gritsman
,
K.
,
Zhang
,
J.
,
Cheng
,
S.
,
Heckscher
,
E.
,
Talbot
,
W. S.
and
Schier
,
A. F.
(
1999
).
The EGF-CFC protein one-eyed pinhead is essential for nodal signaling
.
Cell
97
,
121
-
132
.
Janknecht
,
R.
,
Wells
,
N. J.
and
Hunter
,
T.
(
1998
).
TGF-beta-stimulated cooperation of smad proteins with the coactivators CBP/p300
.
Genes Dev.
12
,
2114
-
2119
.
Kimmel
,
C. B.
,
Ballard
,
W. W.
,
Kimmel
,
S. R.
,
Ullmann
,
B.
and
Schilling
,
T. F.
(
1995
).
Stages of embryonic development of the zebrafish
.
Dev. Dyn.
203
,
253
-
310
.
Lee
,
J.
,
Moon
,
H.-J.
,
Lee
,
J.-M.
and
Joo
,
C.-K.
(
2010
).
Smad3 regulates Rho signaling via NET1 in the transforming growth factor-beta-induced epithelial-mesenchymal transition of human retinal pigment epithelial cells
.
J. Biol. Chem.
285
,
26618
-
26627
.
Liu
,
Z.
,
Lin
,
X.
,
Cai
,
Z.
,
Zhang
,
Z.
,
Han
,
C.
,
Jia
,
S.
,
Meng
,
A.
and
Wang
,
Q.
(
2011
).
Global identification of SMAD2 target genes reveals a role for multiple co-regulatory factors in zebrafish early gastrulas
.
J. Biol. Chem.
286
,
28520
-
28532
.
Liu
,
X.
,
Xiong
,
C.
,
Jia
,
S.
,
Zhang
,
Y.
,
Chen
,
Y.-G.
,
Wang
,
Q.
and
Meng
,
A.
(
2013
).
Araf kinase antagonizes Nodal-Smad2 activity in mesendoderm development by directly phosphorylating the Smad2 linker region
.
Nat. Commun.
4
,
1728
.
Liu
,
Z.
,
Ning
,
G.
,
Xu
,
R.
,
Cao
,
Y.
,
Meng
,
A.
and
Wang
,
Q.
(
2016
).
Fscn1 is required for the trafficking of TGF-beta family type I receptors during endoderm formation
.
Nat. Commun.
7
,
12603
.
Massagué
,
J.
(
2008
).
TGFbeta in Cancer
.
Cell
134
,
215
-
230
.
Miyakoshi
,
A.
,
Ueno
,
N.
and
Kinoshita
,
N.
(
2004
).
Rho guanine nucleotide exchange factor xNET1 implicated in gastrulation movements during Xenopus development
.
Differentiation
72
,
48
-
55
.
Molenaar
,
M.
,
van de Wetering
,
M.
,
Oosterwegel
,
M.
,
Peterson-Maduro
,
J.
,
Godsave
,
S.
,
Korinek
,
V.
,
Roose
,
J.
,
Destrée
,
O.
and
Clevers
,
H.
(
1996
).
XTcf-3 transcription factor mediates beta-catenin-induced axis formation in Xenopus embryos
.
Cell
86
,
391
-
399
.
Murray
,
D.
,
Horgan
,
G.
,
MacMathuna
,
P.
and
Doran
,
P.
(
2008
).
NET1-mediated RhoA activation facilitates lysophosphatidic acid-induced cell migration and invasion in gastric cancer
.
Br. J. Cancer
99
,
1322
-
1329
.
Nakaya
,
Y.
,
Sukowati
,
E. W.
,
Wu
,
Y.
and
Sheng
,
G.
(
2008
).
RhoA and microtubule dynamics control cell-basement membrane interaction in EMT during gastrulation
.
Nat. Cell Biol.
10
,
765
-
775
.
Nicklas
,
D.
and
Saiz
,
L.
(
2013
).
Characterization of negative feedback network motifs in the TGF-beta signaling pathway
.
PLoS ONE
8
,
e83531
.
Nomura
,
M.
and
Li
,
E.
(
1998
).
Smad2 role in mesoderm formation, left-right patterning and craniofacial development
.
Nature
393
,
786
-
790
.
Ogryzko
,
V. V.
,
Schiltz
,
R. L.
,
Russanova
,
V.
,
Howard
,
B. H.
and
Nakatani
,
Y.
(
1996
).
The transcriptional coactivators p300 and CBP are histone acetyltransferases
.
Cell
87
,
953
-
959
.
Pouponnot
,
C.
,
Jayaraman
,
L.
and
Massagué
,
J.
(
1998
).
Physical and functional interaction of SMADs and p300/CBP
.
J. Biol. Chem.
273
,
22865
-
22868
.
Qin
,
H.
,
Carr
,
H. S.
,
Wu
,
X.
,
Muallem
,
D.
,
Tran
,
N. H.
and
Frost
,
J. A.
(
2005
).
Characterization of the biochemical and transforming properties of the neuroepithelial transforming protein 1
.
J. Biol. Chem.
280
,
7603
-
7613
.
Schier
,
A. F.
and
Shen
,
M. M.
(
2000
).
Nodal signalling in vertebrate development
.
Nature
403
,
385
-
389
.
Schmidt
,
A.
and
Hall
,
A.
(
2002
).
The Rho exchange factor Net1 is regulated by nuclear sequestration
.
J. Biol. Chem.
277
,
14581
-
14588
.
Schmierer
,
B.
and
Hill
,
C. S.
(
2007
).
TGFbeta-SMAD signal transduction: molecular specificity and functional flexibility
.
Nat. Rev. Mol. Cell Biol.
8
,
970
-
982
.
Schohl
,
A.
and
Fagotto
,
F.
(
2003
).
A role for maternal beta-catenin in early mesoderm induction in Xenopus
.
EMBO J.
22
,
3303
-
3313
.
Shen
,
M. M.
(
2007
).
Nodal signaling: developmental roles and regulation
.
Development
134
,
1023
-
1034
.
Srougi
,
M. C.
and
Burridge
,
K.
(
2011
).
The nuclear guanine nucleotide exchange factors Ect2 and Net1 regulate RhoB-mediated cell death after DNA damage
.
PLoS ONE
6
,
e17108
.
Stroschein
,
S. L.
,
Wang
,
W.
,
Zhou
,
S.
,
Zhou
,
Q.
and
Luo
,
K.
(
1999
).
Negative feedback regulation of TGF-beta signaling by the SnoN oncoprotein
.
Science
286
,
771
-
774
.
Struhl
,
K.
(
1998
).
Histone acetylation and transcriptional regulatory mechanisms
.
Genes Dev.
12
,
599
-
606
.
Taunton
,
J.
,
Hassig
,
C. A.
and
Schreiber
,
S. L.
(
1996
).
A mammalian histone deacetylase related to the yeast transcriptional regulator Rpd3p
.
Science
272
,
408
-
411
.
Tu
,
Y.
,
Lu
,
J.
,
Fu
,
J.
,
Cao
,
Y.
,
Fu
,
G.
,
Kang
,
R.
,
Tian
,
X.
and
Wang
,
B.
(
2010
).
Over-expression of neuroepithelial-transforming protein 1 confers poor prognosis of patients with gliomas
.
Jpn. J. Clin. Oncol.
40
,
388
-
394
.
Vessichelli
,
M.
,
Ferravante
,
A.
,
Zotti
,
T.
,
Reale
,
C.
,
Scudiero
,
I.
,
Picariello
,
G.
,
Vito
,
P.
and
Stilo
,
R.
(
2012
).
Neuroepithelial transforming gene 1 (Net1) binds to caspase activation and recruitment domain (CARD)- and membrane-associated guanylate kinase-like domain-containing (CARMA) proteins and regulates nuclear factor kappaB activation
.
J. Biol. Chem.
287
,
13722
-
13730
.
Wei
,
S.
,
Dai
,
M.
,
Liu
,
Z.
,
Ma
,
Y.
,
Shang
,
H.
,
Cao
,
Y.
and
Wang
,
Q.
(
2017
).
The guanine nucleotide exchange factor Net1 facilitates the specification of dorsal cell fates in zebrafish embryos by promoting maternal beta-catenin activation
.
Cell Res.
27
,
202
-
225
.
Weinstein
,
M.
,
Yang
,
X.
,
Li
,
C.
,
Xu
,
X.
,
Gotay
,
J.
and
Deng
,
C.-X.
(
1998
).
Failure of egg cylinder elongation and mesoderm induction in mouse embryos lacking the tumor suppressor smad2
.
Proc. Natl. Acad. Sci. USA
95
,
9378
-
9383
.
Wolffe
,
A. P.
(
1996
).
Histone deacetylase: a regulator of transcription
.
Science
272
,
371
-
372
.
Wotton
,
D.
,
Lo
,
R. S.
,
Lee
,
S.
and
Massagué
,
J.
(
1999
).
A Smad transcriptional corepressor
.
Cell
97
,
29
-
39
.
Wu
,
M. Y.
and
Hill
,
C. S.
(
2009
).
Tgf-beta superfamily signaling in embryonic development and homeostasis
.
Dev. Cell
16
,
329
-
343
.
Yang
,
X.
,
Letterio
,
J. J.
,
Lechleider
,
R. J.
,
Chen
,
L.
,
Hayman
,
R.
,
Gu
,
H.
,
Roberts
,
A. B.
and
Deng
,
C.
(
1999
).
Targeted disruption of SMAD3 results in impaired mucosal immunity and diminished T cell responsiveness to TGF-beta
.
EMBO J.
18
,
1280
-
1291
.
Zhou
,
X.
,
Sasaki
,
H.
,
Lowe
,
L.
,
Hogan
,
B. L. M.
and
Kuehn
,
M. R.
(
1993
).
Nodal is a novel TGF-beta-like gene expressed in the mouse node during gastrulation
.
Nature
361
,
543
-
547
.
Zhu
,
Y.
,
Richardson
,
J. A.
,
Parada
,
L. F.
and
Graff
,
J. M.
(
1998
).
Smad3 mutant mice develop metastatic colorectal cancer
.
Cell
94
,
703
-
714
.
Zorn
,
A. M.
and
Wells
,
J. M.
(
2009
).
Vertebrate endoderm development and organ formation
.
Annu. Rev. Cell Dev. Biol.
25
,
221
-
251
.

Competing interests

The authors declare no competing or financial interests.