Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2008 Nov 2.
Published in final edited form as: Dev Biol. 2006 Aug 10;299(2):398–410. doi: 10.1016/j.ydbio.2006.08.009

Smurf1 regulates neural patterning and folding in Xenopus embryos by antagonizing the BMP/Smad1 pathway

Evguenia M Alexandrova 1, Gerald H Thomsen 1,*
PMCID: PMC2577174  NIHMSID: NIHMS38773  PMID: 16973150

Abstract

The ubiquitin ligase Smurf1 can target a handful of signaling proteins for ubiquitin-mediated proteasomal destruction or functional modification, including TGF-β receptors, Smads, transcription factors, RhoA and MEKK2. Smurf1 was initially implicated in BMP pathway regulation in embryonic development, but its potential role in vertebrate embryogenesis has yet to be clarified. Here we demonstrate that inhibition of Smurf1 in Xenopus laevis embryos with an antisense morpholino oligonucleotide or a dominant-negative protein disrupts early development, with the nervous system being the principal target. Smurf1 is enriched on the dorsal side of gastrula stage embryos, and blocking Smurf1 disturbs neural folding and neural, but not mesoderm differentiation, enhances BMP/Smad1 signaling, and elevates phospho-Smad1 levels in the dorsal ectoderm. We conclude that in Xenopus embryos, the BMP pathway is a major physiological target of Smurf1, and we propose that in normal development Smurf1 cooperates with secreted BMP antagonists to limit BMP signaling in dorsal ectoderm. Our data also reveal a novel role for Smurf1 and Smad1 in neural plate morphogenesis.

Keywords: Smurf1, Smad1, BMP, Ubiquitin ligase, Neural tube, Neural folding, Neural patterning, Signal transduction, Embryo, Xenopus laevis

Introduction

Smurf1 is a member of the HECT class of E3 ubiquitin ligases, and it is evolutionarily conserved from Drosophila through man (Zhu et al., 1999; Podos et al., 2001; Ebisawa et al., 2001). Smurf1 and the related Smurf2 are characterized by an N-terminal phospholipid binding or C2 domain, two or three WW domains that bind PPXY consensus motifs in partner proteins and substrates, and a C-terminal catalytic HECT domain (Zhu et al., 1999; Pickart, 2001a). Ubiquitin ligases catalyze transfer of ubiquitin from an E2, ubiquitin-conjugating enzyme, onto target proteins that results in their proteasomal or lysosomal degradation, or regulates their subcellular localization, trafficking or protein–protein interactions (Pickart, 2001a, b). We originally isolated Smurf1 as a Smad1-interacting factor by a yeast two-hybrid screen (Zhu et al., 1999).

Smad1 is a signal transducer in the canonical bone morphogenetic protein (BMP) signal transduction pathway that plays an important role in several events during vertebrate embryonic development: (1) the patterning of the ventro-lateral mesoderm; (2) the decision between epidermal and neural cell fate, in which high activity of Smad1/5 specifies epidermis, intermediate activity specifies the “neural border” fates (e.g. neural crest and cement gland), and in the absence of BMP/Smad1 signaling, neural induction takes place; (3) dorsoventral patterning of the neural tube, wherein BMPs are responsible for differentiation of dorsal neuronal subtypes (Dale and Wardle, 1999; Harland, 2000; Hill, 2001; De Robertis and Kuroda, 2004; Chizhikov and Millen, 2005; Wilson and Maden, 2005).

BMP signaling commences when homo- or heterodimers bind a complex of type I and type II Ser/Thr kinase receptors, Smads 1, 5 or 8 (Smad1/5/8) get phosphorylated and activated, bind to the co-partner Smad4 and translocate as a complex to the nucleus where they regulate target gene transcription (Lutz and Knaus, 2002). The BMP/Smad1 pathway can be negatively regulated at several levels: by extracellular BMP antagonists such as Noggin and Chordin, pseudoreceptors (e.g. BAMBI), inhibitory Smads, MAP kinases and Smad ubiquitylation regulatory factors or Smurfs (reviewed by von Bubnoff and Cho, 2001; Lutz and Knaus, 2002; De Robertis and Kuroda, 2004).

We have shown that Smurf1 can ubiquitylate and down-regulate Smad1/5 (Zhu et al., 1999; see below), but it also has a number of other potential targets that depend on the cell. For example, in C2C12 and 2T3 cells, Smurf1 can suppress BMP/Smad5 signaling and osteoblast differentiation by ubiquitylating Smad5 (Ying et al., 2003) or the osteoblast-specific transcription factor Cbfα1/Runx2 (Zhao et al., 2003, 2004; Kaneki et al., 2006). In overexpression assays, Smurf1 can target the TGF-β type I receptor (TBRI), BMP type I receptor (ALK6), Smad4 and inhibitory Smad7 for proteasomal degradation (Moren et al., 2005; Ebisawa et al., 2001; Suzuki et al., 2002; Murakami et al., 2003; Zhu et al., 1999 supplementary data). Furthermore, endogenous Smurf1-dependent ubiquitylation can trigger degradation of the small GTPase RhoA to affect cell protrusive activity and polarity (Wang et al., 2003), neurite outgrowth (Bryan et al., 2005) or epithelial cell tight junction dissolution in TGF-β-induced epithelial–mesenchymal transition (Ozdamar et al., 2005).

By misexpressing Smurf1 in Xenopus embryos, we previously found that Smurf1 can cause incomplete secondary axis formation by dorsalizing ventral marginal zone tissue, and Smurf1 can neuralize embryonic ectodermal explants (Zhu et al., 1999). However, a loss-of-function analysis of Smurf1 in Xenopus embryos is needed to reveal which, if any, of these phenomena are relevant in vivo. Smurf1 loss-of-function studies have been accomplished in Drosophila and mouse, with somewhat different results. Drosophila maternalzygotic dSmurf mutants display enhanced and prolonged DPP/BMP signaling (Podos et al., 2001) as a consequence of stabilized phospho-MAD, the activated Drosophila homolog of vertebrate Smad1/5 (Liang et al., 2003). In contrast, Smurf1 knockout (KO) mice do not have developmental defects, but are characterized by an age-dependent increase in bone mass through enhanced osteoblast activity (Yamashita et al., 2005). Although osteoblasts from these mice are sensitized to BMP signaling, Smurf1 does not directly affect the levels of Smad1 or BMP receptors. Instead, MEKK2 is stabilized and activates JNK. The mouse results in particular raise the question of whether or not Smurf1 targets the BMP/ Smad1 pathway under physiological situations in developing vertebrate embryos.

Here we report that blocking endogenous Smurf1 in Xenopus embryos with an antisense morpholino oligonucleotide (MO), or a dominant-negative mutant protein, disrupts neural folding and patterning, greatly affecting head development. We show that up-regulation of the BMP/Smad1 signaling pathway is the underlying cause of the knockdown phenotypes.

Materials and methods

Embryo manipulations, in situ hybridization, morpholino oligos and synthetic mRNAs

Xenopus laevis embryos were obtained by standard in vitro fertilization, de-jellied in 2% cysteine pH8.0, microinjected and incubated for several hours in 3% ficoll+0.5×MMR+10 μg/ml gentamycin and grown in 0.1×MMR+10 μg/ml gentamycin thereafter. 30–40 ng Smurf1 MO or 0.10–0.20 ng Smurf1CA mRNA was injected into the dorsal-animal region of the 4–8 cell stage embryos, unless indicated otherwise. All stages are according to Nieuwkoop and Faber (1967). Animal caps were excised at stage 8; prospective neural ectoderm (NE) and ventral ectoderm (VE) were excised at stage 10.25 as 60°-wide sectors on either the dorsal or the ventral side, from the animal pole to the bottom of the pigmented zone. The explants were cut and cultured in 0.5×MMR+10 μg/ml gentamycin until the sibling embryos reached the appropriate stage. In situ hybridization was as previously described (Harland, 1991) using BM purple as a chromogenic substrate (Roche). Template plasmids for making probes were pCS2-Smurf1, pGEM-Pax6, pBS-SK-Otx2, pBS-KS+En2, pGEM-Krox20, pBS-NCAM, pCS2-reverse-Msx1, pBS-SK-XNkx2.2, pBS-Sox2 and pBS-Xep. Morpholino oligo sequences are: standard control MO-5′cctcttacctcagttacaatttata3′, Smurf1 MO-5′attcgacatccctccaaacgccg3′ (Gene Tools, LLC, Philomath, OR). Full-length X. laevis Smad6 and zebrafish Danio rerio zSmurf1 were obtained as EST clones (I. M.A.G.E. consortium, clone ID numbers 6317366 and 5915182, respectively) and verified by sequencing. zSmurf1 was subcloned into pCS2 using EcoRI and XbaI sites. Capped synthetic mRNAs were in vitro transcribed using mMESSAGE mMACHINE kits (Ambion) from the following linearized plasmids: pCS2-Smurf1, pCS2-Smurf1CA, pCS2-zSmurf1, pCS2-Smad1, pSPYS-Chordin, pCS2-BMP4, pCMV-SPORT6-Smad6.

In vitro translation, Western blot analysis, antibodies and phalloidin staining

Smurf1 protein was in vitro translated using TnT T7/SP6 Coupled Reticulocyte Lysate system (Promega) in the presence of control or Smurf1 MO and [35S] methionine, according to the manufacturer’s protocol. Half of each reaction was resolved by 8% SDS-PAGE, visualized by autoradiography and quantified using NIH Image software. For Western blot analysis, three total embryos (Fig. 2C) or 15 explants (Fig. 11) were lysed and the proteins were resolved by 8% SDS-PAGE. A Smurf1 monoclonal antibody (Wang et al., 2003) was used at 1:4 dilution; β-tubulin antibody was used at 1:20,000 dilution (Accurate Chemical and Scientific Corporation); P-Smad1/5 antibody was used at 1:200 dilution (Cell Signaling Technology, Inc.); AF680 goat anti-rabbit and AF800 goat anti-mouse secondary antibodies were used at 1:2000 dilution (Molecular Probes). Resolved proteins were visualized and quantified using the Odyssey Infrared Imager (LI-COR, Inc.). F-actin staining was done using 1:100 AF488-phalloidin (Molecular Probes) on stage 18 embryos fixed 30 min in MEMPFA.

Quantitative RT-PCR

Total RNA was extracted from 10–15 animal caps or one wt embryo in the presence of 0.25 μg/ml proteinase K, phenol/chloroform extracted and ethanol precipitated, treated with DNaseI and phenol/chloroform extracted/ethanol precipitated again. 1 μg of total RNA was used to synthesize cDNA and 1/30 of the resulting cDNA was used in each RT-PCR reaction. The primers to Xagr2 (Novoselov et al., 2003), 5′gaaccagctgatattgatcatttg3′ (upstream) and 5′aatggtctccttcatacaccac3′ (downstream), were used at the following conditions: 95°C/10 s, 55°C/5 s, 72°C/12 s, acquisition temperature 79°C. The primers to XAG1, 5′ctgactgtccgatcagac3′ (upstream) and 5′gagttgcttctctggcat3′ (downstream), were used at the following conditions: 95°C/10 s, 55°C/6 s, 72°C/12 s, acquisition temperature 85°C. Primers and cycling conditions for ODC, N-CAM, Wnt8, Vent1, GATA6 and αT4-globin were previously described (Kofron et al., 1999; Xanthos et al., 2001; Tao et al., 2005). Quantitative RT-PCR using LightCycler System (Roche Applied Science) was previously described (Kofron et al., 1999). We used 1:1, 1:10, 1:100 and 1:1000 dilutions of the cDNA from a wt embryo to generate standard curves. Expression levels of all genes were normalized to the expression level of ODC. That value was further normalized to the corresponding endogenous gene expression level in a whole embryo, and plotted as percentage thereof.

Results

Smurf1 knockdown disrupts anterior neural development

To analyze the in vivo function of Smurf1 in Xenopus embryos by a loss-of-function approach, we first re-examined the distribution of Smurf1 transcripts in more detail than previously reported (Zhu et al., 1999). Embryos were bisected at gastrula and neurula stages and subjected to in situ hybridization. This revealed an enrichment of Smurf1 at the onset of gastrulation in dorsal compared to ventral animal pole and marginal zone (Figs. 1A, A′). In early (stage 14) and late (stage 20) neurula embryos, Smurf1 transcripts are present in the neural plate, somitogenic mesoderm, somites, prechordal plate and notochord (Figs. 1B–C′). These patterns suggested to us that Smurf1 contributes to neural and/or dorsal mesodermal development. To address the endogenous function of Smurf1, we used gene knockdown and dominant negative approaches.

Fig. 1.

Fig. 1

Open in a new tab

Smurf1 transcripts are enriched in the dorsal tissues (A, A′) Lateral view and sagittal section of early gastrula, stage10 (animal pole up, dorsal side at the right, asterisks mark the dorsal blastopore lip). (B, B′) Early neurula, stage 14. (C, C′) Late neurula, stage 20. In panels B, B′ and C, anterior is at the right, (B′, C) sagittal sections, (C′) transversal section. S—somites, sm—somitogenic mesoderm, black arrowheads—neural folds, black arrow—notochord, white arrow—prechordal plate.

To knockdown Smurf1, we designed a Smurf1 morpholino oligonucleotide (MO) that would target Smurf1, but not the two Smurf2 alleles found in X. laevis (Fig. 2A). Notably, despite the pseudotetraploid nature of X. laevis, we were unable to detect in any X. laevis EST database Smurf1 transcripts that differ from the one we originally reported (Zhu et al., 1999). The Smurf1 antisense MO, but not a control MO, efficiently blocked Smurf1 translation in coupled transcription/translation reticulocyte lysate reactions (Fig. 2B). In addition, this MO inhibited Smurf1 production in Smurf1 mRNA-injected embryos up to 80% (Fig. 2C) and blocked secondary axis induction by ectopic wild-type (wt) Smurf1 (Figs. 2D–F). As an alternative approach, we employed a catalytically inactive Smurf1 Cys699Ala mutant (Smurf1CA), previously reported to act as a dominant-negative in mammalian C2C12 cells (Zhao et al., 2003). Consistent with its reported dominant-negative activity, Smurf1CA also blocked wt Smurf1-induced ectopic axes in Xenopus embryos (Fig. 2G).

Fig. 2.

Fig. 2

Open in a new tab

Characterization of Smurf1 MO and Smurf1CA effectiveness. (A) Alignment of Smurf1 MO sequence with Xenopus laevis (X.l.) Smurf1, zebrafish Danio rerio Smurf1 (zSmurf1) and two versions of X. laevis Smurf2. The translation start site is in bold. (B, C) The amount of Smurf1 protein produced by in vitro translation (B) or in the embryos injected with Smurf1 mRNA (C) is reduced in the presence of Smurf1 MO, but not control MO. (D–G) Smurf1 MO (F) and Smurf1CA (G) block secondary axis induction by the wild-type Smurf1 (E).

To determine regional requirements for Smurf1 in normal development, we injected the Smurf1 MO (details in Methods), or 0.1–0.2 ng Smurf1CA mRNA into blastomeres with well-known tissue fates. Injections targeting ventral blastomeres at the 4–8 cell stage, which are fated to form epidermis and ventro-posterior mesendodermal tissues (Moody, 1987), did not affect development (data not shown). However, animal–dorsal injections at the 4–8 cell stage, that mainly target prospective neural ectoderm (Moody, 1987), caused defective neural folding, microcephaly, and micropthalmy (Figs. 3A–D). Consistent with these early phenotypes, Smurf1 KD embryos reared to feeding tadpole stages 45–50 showed reduced eyes and pigmentation, and an overall smaller size (Fig. 4G). Defects in neural plate folding caused by either Smurf1 MO or Smurf1CA treatment were accompanied by loss of f-actin bundles at neural fold hinge points, revealed by phalloidin staining (Figs. 3F–I, arrows). Of note, injection of these blocking reagents into prospective dorsal mesoderm (by targeting dorsal–vegetal blastomeres at the 8-cell stage; Moody, 1987) did not cause any abnormalities. The onset and progression of gastrulation, as well as axis elongation occurred normally (data not shown). These first tests thus indicated a role for Smurf1 in neural, but not mesodermal development (a conclusion backed up by molecular analysis, below).

Fig. 3.

Fig. 3

Open in a new tab

Knockdown of endogenous Smurf1 results in neural defects. (A–D) Smurf1 MO (B) and Smurf1CA (D), but not control MO (A) or control β-gal mRNA (C) cause neural tube closure defects at neurula stages (top row) and microcephaly and microphtalmy at tailbud stages (bottom row). This phenotype is mimicked by 1–6 ng Smad1 mRNA overexpression (E). (F–I) F-actin accumulation at the neural fold hinge points at stage 18, revealed by phalloidin staining. Control embryos (F, H) have normal loop-shaped pattern. In Smurf1MO (G) or Smurf1CA embryos (I), f-actin staining is weak and the distance between the lateral hinge points is increased. Scale bar=0.1 mm.

Fig. 4.

Fig. 4

Open in a new tab

Rescue of Smurf1 MO and Smurf1CA embryos by wild-type Smurf1. (A–C, G) Neural folding defect (B) and overall phenotype of tadpoles caused by Smurf1MO is rescued by 0.1 ng of zSmurf1 mRNA (C, G). (D–F) Smurf1CA embryos are similarly rescued by X. laevis Smurf1.

Focusing on the neural defects, we performed rescue experiments as an important control for specificity of the knockdown reagents. Co-injection of wt X. laevis Smurf1or zebrafish Smurf1 (zSmurf1) mRNA rescued the neural folding and eye defects and improved the overall abnormal late phenotypes of Smurf1CA and Smurf1 MO embryos, respectively (Table 1 and Fig. 4). zSmurf1 is structurally and functionally homologous to X. laevis Smurf1, as determined by conservation of their primary protein sequence and ability to induce secondary axes when ectopically expressed in Xenopus embryos (Supplementary Fig. 1). Note that Smurf1 MO and zSmurf1 mRNA sequences have 9 base mismatches (Fig. 2A) and therefore should not interact with each other). Our results indicate that endogenous Smurf1 is required for neural, but not mesodermal development, particularly in closure of the anterior neural tube, and formation of the eyes and head.

Table 1.

Rescue of Smurf1 knockdown embryos by wt Smurf1

Na % NTDsb Na % eye defects
Uninjected 46 0 28 0
Control MO, 35 ng 34 0.9 Not scored
Smurf1 MO, 35 ng 44 77.3 24 58.3
Smurf1 MO, 35 ng+zSmurf1 mRNA, 0.1 ng 38 13.2 26 34.6
Uninjected 33 9.1 54 11.1
Smurf1CA mRNA, 0.1 ng 31 64.5 58 53.6
Smurf1CA mRNA, 0.1 ng+Smurf1 mRNA, 0.2 ng 32 9.4 56 28.6
Open in a new tab
a

N—number of embryos or eyes scored.

b

NTD (neural tube defects)-defective neural folding scored at stages 18–20.

Suppression of neural and up-regulation of epidermal genes in Smurf1 knockdown embryos

To understand the phenotypes in more detail, we asked next whether Smurf1 knockdown defects were accompanied by changes in the expression of neural genes. In situ hybridization on neurula stage embryos showed that a variety of regional neural markers were suppressed and/or mispatterned in Smurf1 MO and Smurf1CA embryos, including the forebrain marker Otx2, prospective eye marker Pax6, mid-hindbrain boundary marker En2 and rhombomeres 3 and 5 marker Krox20 (Figs. 5A–L). Similarly, the floorplate staining of a Shh target Nkx2.2 was reduced (our probe did not penetrate deep enough to stain the notochord, data not shown) (Figs. 5S–U). Importantly, the disruption of neural marker staining patterns was not associated with tissue death or lysis (Supplementary Fig. 2). In contrast to regional markers, the general neural markers N-CAM and Sox2 were not significantly changed, although N-CAM staining was weaker in some, but not all experiments (Figs. 5M–R, 6H–L, and data not shown).

Fig. 5.

Fig. 5

Open in a new tab

Smurf1 knockdown disrupts neural gene expression. Anterior neural marker Otx2 (A–D), prospective eye marker Pax6 (E–H), rhombomere 3 and 5 marker Krox20 and the marker of mid-hindbrain boundary En2 (I–L) at late neurula stage 18. Pan-neural marker N-CAM at stage 18 (M–P) and stage 14 (Q, R). (S–U) Shh target gene in the neural plate, Nkx2.2, at stage 17. All are anterior views.

Fig. 6.

Fig. 6

Open in a new tab

BMP-responsive genes are up-regulated by blocking Smurf1. (A–D) The zone of Msx1, a marker of the neural plate border, is wider in Smurf1 MO and Smurf1CA embryos (B, D, arrowheads), than in control embryos (A, C). (E–G) Expression pattern of the epidermal marker, Xep, encroaches into the placodal zone (F, arrow), but loses its neural domain (asterisk) in Smurf1CA embryos. Its normal pattern is rescued by wt Smurf1 (G). Pink color demarcates lineage tracing by β-gal. (H–L) Double staining of control, Smurf1 MO and Smurf1CA embryos with Xep (purple, arrows) and Sox2 (cyan) and rescue of the Xep pattern by wt Smurf1 (M) and Smad6 (N). All are anterior views.

In contrast to marker reduction above, expression domains of both the neural border marker Msx1 (Figs. 6A–D, arrowheads) and the epidermal marker Xep (Figs. 6E,F,H–L, arrows) were expanded in Smurf1 KD embryos. Notably, Xep also has a neural domain of expression (Vasiliev et al., 1997) that was lost in Smurf1 KD embryos (Figs. 6E–G, asterisk), but the normal pattern of Xep expression was restored by co-injection of 0.2 ng of wt Smurf1 mRNA (Figs. 6G, M). Altogether, these data indicate that in Smurf1 KD embryos a moderate conversion of neural to non-neural cell fate takes place. Moreover, this does not seem to be mediated by defects in mesoderm patterning, as none of the mesoderm markers analyzed was affected in Smurf1CA and Smurf1 MO embryos (Fig. 7 and data not shown).

Fig. 7.

Fig. 7

Open in a new tab

Mesodermal markers are not affected by blocking Smurf1. General mesodermal marker XBra (A, B), the marker of the prospective head mesoderm Frzb (C, D) and the marker of the Organizer and notochord Chd (E–H) are expressed normally in Smurf1CA embryos. (A–D) Vegetal view, dorsal side up. (E–H) Dorsal view, animal pole up. Representative embryos are shown.

Smad1 overexpression mimics Smurf1 knockdown

We hypothesized that the embryonic phenotype generated by Smurf1 knockdown was caused by stabilization of one or more of Smurf1 targets that, under normal conditions, would be degraded due to Smurf1-mediated ubiquitylation. If so, then overexpression of a potential target(s) would be expected to mimic Smurf1 knockdown phenotype. Several Smurf1 targets have been identified (see Introduction), of which the BMP/ Smad1 pathway is arguably the most critical known effector of neuroectodermal and ectodermal induction and patterning (Wilson et al., 1997; Harland, 2000; De Robertis and Kuroda, 2004). We tested the ability of several potential targets to mimic Smurf1 KD. Notably, Smad1 (Fig. 3E), but not RhoA (Supplementary Fig. 3) caused nearly identical defects as Smurf1 KD.

We then tested to what extent would the neural-to-epidermal conversion take place in Smad1 overexpressing embryos. Although 6 ng Smad1 mRNA potently induced Xep and Msx1 in the neural plate (Figs. 8E, H), the effects of 1 ng Smad1 mRNA were very similar to Smurf1 knockdown. Specifically, N-CAM was only slightly suppressed (Fig. 8B), whereas Xep (Fig. 8D, arrow) and Msx1 (Fig. 8G, arrowheads) were up-regulated at the periphery of the neural plate, and the neural domain of Xep disappeared (Figs. 8C, D asterisk). In spite of just a moderate conversion of the neural plate cells into epidermis, neural folding movements were strongly inhibited, as revealed by a wider N-CAM staining pattern (Fig. 8B) and wider unstained area in Xep and Msx1-stained embryos (Figs. 8D, G), in agreement with the overall phenotype (Fig. 3E). These data demonstrate a significant degree of correlation between the effects of Smurf1 knockdown and moderate Smad1 overexpression, which points to the BMP/Smad1 pathway as a primary physiological target of Smurf1 in Xenopus embryos.

Fig. 8.

Fig. 8

Open in a new tab

A low dose of Smad1 overexpression mimics Smurf1 knockdown. N-CAM staining is slightly reduced (A, B), while Xep (D, arrow) and Msx1 (G, arrowheads) are up-regulated at the periphery of the neural plate in embryos injected with 1 ng Smad1 mRNA. As a positive control, 6 ng Smad1 mRNA strongly induces Xep (E, arrow) and Msx1 (H, arrowheads) within the neural plate. Notice the wider neural plate in all Smad1-overexpressing embryos (B, D, E, G, H). All are anterior views at stage 14 (A, B) or 17 (C–H).

Smad6 rescues Smurf1 knockdown embryos

To test whether BMP/Smad1 pathway is activated in neuroectoderm of Smurf1 KD embryos, we attempted to rescue the Smurf1 KD phenotype by Smad6, an inhibitory Smad specific for the BMP pathway that acts by disrupting Smad1/5–Smad4 complex formation (Hata et al., 1998). We found that indeed, co-injection of 1 ng Smad6 mRNA rescued the neural folding defects in Smurf1 MO embryos (Figs. 9A, C) and Smurf1CA embryos (Figs. 9B, D), and Sox2 and Xep expression patterns in Smurf1CA embryos (Fig. 6N). These results demonstrate that the defects caused by loss of Smurf1 function result from enhanced BMP/Smad1 signaling in the neural plate.

Fig. 9.

Fig. 9

Open in a new tab

A Smad1/5 antagonist, Smad6, rescues the Smurf1 knockdown phenotype. (A, B) Representative stage 18 (A) and stage 17 (B) embryos. (C, D) Summary of Smurf1 MO (C) and Smurf1CA (D) rescue by Smad6 (n—number of embryos).

Endogenous Smurf1 is required for neural induction in animal pole ectoderm and ventral patterning

Since Smurf1 knockdown affects both neural folding and patterning in whole embryos, we attempted to determine whether blocking Smurf1 in animal caps affects neural differentiation independent of apparent cell movements, using neural differentiation assays with isolated animal cap ectoderm. The epidermal and neural fates of primary ectoderm are specified by differential activation of the BMP/Smad1 pathway: elevated Smad1 signaling triggers epidermal differentiation, while absent or sub-threshold Smad1 signaling allows neural differentiation (Wilson et al., 1997; Harland, 2000; De Robertis and Kuroda, 2004). We activated neural differentiation in animal caps by injecting 0.5 ng or 1 ng mRNA of the secreted BMP antagonist, Chordin. This potently induced N-CAM in the presence of control MO, but co-injection of the Smurf1 MO greatly reduced both the number of N-CAM positive caps and intensity of N-CAM expression (Figs. 10A, B). Likewise, Smurf1CA suppressed induction of neural and cement gland markers by Chordin in animal caps, as scored by quantitative RT-PCR on N-CAM, XAG1 and Xagr2 genes (Fig. 10C). Thus, defects in neural differentiation are unlikely due to effects of Smurf1 KD on cell movements.

Fig. 10.

Fig. 10

Open in a new tab

Blocking Smurf1 inhibits response to Chordin and enhances response to BMP. (A, B) Smurf1 MO, but not control MO, blocks neural induction by Chordin in animal caps, as scored by in situ hybridization with N-CAM probe. (A) Combined data from two experiments (n—number of animal caps), (B) representative animal caps from each treatment. (C) Similarly, Smurf1CA suppresses expression of N-CAM and two cement gland markers, XAG1 and Xagr2, in Chordin mRNA-injected animal caps, as measured by quantitative RT-PCR at stage 25. (D) Smurf1CA also inhibits secondary axis induction by Chordin mRNA injected in the ventral marginal zone (n—number of embryos). (E) Likewise, embryos injected with 1 ng Chordin mRNA in the animal pole are strongly dorsalized, which is reversed by co-injection of Smurf1CA mRNA. (F) Quantitative RT-PCR on animal caps injected with BMP4 mRNA, with or without Smurf1CA. Expression of both early (Vent1 and Wnt8, scored at stage 10.5) and late BMP4 target genes (GATA6 and αT4-globin, scored at stage 25) is enhanced by Smurf1CA. Panels C and F were repeated twice on independent cDNA with similar results.

Besides the ectoderm, we were also able to reveal endogenous Smurf1 activity in ventral marginal zone (VMZ) tissues, which also express Smurf1 but at a lower level than in the DMZ (Fig. 1). We found that Smurf1CA was able to inhibit the ability of Chordin to induce secondary axial structures when the two were co-injected in the VMZ (Fig. 10D). Smurf1CA could also counteract dorsalization of whole embryos by Chordin mRNA injection into the animal pole at the two-cell stage (Fig. 10E). These findings reveal an underlying activity for endogenous Smurf1 in ventral/posterior tissues that is exhibited when tissues are sensitized with a BMP inhibitor (also refer to Discussion).

These data are consistent with elevation of endogenous BMP/Smad1 signaling when Smurf1 is knocked down in embryos. We propose that the function of Smurf1 in Xenopus embryos is to impose a limit on the BMP/Smad1 signaling in the ectoderm that is sufficient for epidermal differentiation of non-neural ectoderm but low enough to permit neural induction in the dorsal ectoderm.

Blocking Smurf1 enhances tissue responses to BMP signals

Animal caps provide a convenient way to analyze interactions among genetic pathways. We have proposed that Smurf1 is an antagonist of Smad1 and that Smurf1 knockdown sensitizes ectodermal cells to the BMP/Smad1 signals. To further test the latter possibility, we examined the expression of known BMP4 target genes in animal caps injected with a limiting dose of BMP4 mRNA in the presence or absence of Smurf1CA. BMP levels in isolated animal cap ectoderm are already above the threshold required for the epidermal specification, as evident by differentiation of animal cap explants into epidermis (Grunz and Tacke, 1989). However, by boosting BMP levels further, one can induce ventral mesoderm in the animal caps (Jones et al., 1992; Hemmati-Brivanlou and Thomsen, 1995). We therefore tested whether blocking endogenous Smurf1 can cooperate with BMP4/Smad1 signaling to induce higher expression levels of BMP target genes. We found that minor amounts of ventrolateral mesoderm were induced by injection of limiting doses of BMP4 mRNA (0.2 ng) or Smad1 mRNA (1.0 ng) alone, but when Smurf1CA mRNA was coinjected with either BMP4 or Smad1, we observed significant elevation of early BMP target genes marking ventro-lateral mesoderm (Vent1 and Wnt8, scored at stage 10.5) and late markers of blood (GATA6 and αT4-globin scored at stage 25; Fig. 10F and data not shown). These results show that blocking Smurf1 in the ectoderm can counteract BMP inhibition by antagonists such as Chordin, and also make ectoderm more sensitive to BMP/Smad1 signals. These effects are consistent with elevation of BMP/Smad1 signaling in animal pole ectoderm when endogenous Smurf1 is inhibited. We test this proposition more directly, next.

Phospho-Smad1 levels are elevated in proneural ectoderm of Smurf1 knockdown embryos

To directly test whether blocking Smurf1 stimulates BMP/Smad1 signaling in the ectoderm, we compared the levels of the activated, phosphorylated form of Smad1/5 (P-Smad1/5) in prospective neural ectoderm (NE) excised from wt and Smurf1CA embryos (Fig. 11). We dissected the NE from 10 to 15 embryos at stage 10.25 and performed Western blot analysis on the total proteins, staining simultaneously with antibodies against P-Smad1/5, β-tubulin (loading control) and Smurf1/Smurf1CA (expression control). Since there exists no evidence of Smad5/8 expression in early X. laevis development, protein staining likely represents just Smad1. We found that stage 10.25 NE from Smurf1CA embryos contained P-Smad1 levels that were on average threefold higher than NE from wt embryos, and similar P-Smad1 levels were produced by 1 ng Smad1 mRNA injected on the dorsal side, substantiating our earlier, phenotype-based finding that this dose of Smad1 mRNA most closely imitates Smurf1 knockdown (see Figs. 3E, 8). However, these levels of P-Smad1 did not reach levels as high as seen in the ventral ectoderm (VE), which were fivefold over control. This likely explains why the epidermal program is not significantly initiated within the neural plate of Smurf1CA and Smurf1 MO embryos, or embryos injected with 1 ng Smad1 mRNA (Figs. 6, 8D, G). A dorsal–ventral gradient of P-Smad1 has been observed in the animal hemisphere directly (Faure et al., 2000; Schohl and Fagotto, 2002). Our knockdown of Smurf1 almost eliminates this Smad1 signaling difference between dorsal and ventral tissues. Importantly, the P-Smad1 present in Smurf1CA ectoderm is restored to normal low levels by co-injection of wt Smurf1 (Fig. 11). These experiments directly demonstrate that blocking Smurf1 function in Xenopus embryos results in elevated phosphorylation of endogenous Smad1, which is likely the basis of the phenotypic and molecular defects described above.

Fig. 11.

Fig. 11

Open in a new tab

Phospho-Smad1 levels are elevated in Smurf1CA embryonic neuroectoderm. Dorsal (prospective neural) ectoderm was cut from stage 10.25 embryos injected with indicated mRNAs and analyzed by Western blotting (A). (B) Quantitation of relative P-Smad1/5 levels summarized from four experiments (n—number of experiments, error bars—standard deviation).

Discussion

To summarize our findings, we have observed that blocking endogenous Smurf1 suppresses neural folding and neural patterning in whole embryos, inhibits neural induction and enhances BMP sensitivity in animal caps, and increases endogenous P-Smad1 levels in the prospective neural ectoderm. All of these effects are mimicked by moderate Smad1 overexpression and can be rescued by inhibitory Smad6. We conclude that the function of endogenous Smurf1 in Xenopus embryos is to limit BMP/Smad1 signaling in the proneural ectoderm. Therefore, Smurf1 acts as an intracellular BMP/Smad1 signaling antagonist that likely synergizes with other anti-BMP mechanisms operating on the dorsal side of the embryo, such as Spemann organizer-derived BMP antagonists and their cofactors (De Robertis and Kuroda, 2004; Wills et al., 2006; Lebreton and Jones, 2006), a Smad4 ubiquitin ligase, Ectodermin (Dupont et al., 2005) and β-catenin dependent transcriptional repression of the BMP4 gene (Baker et al., 1999).

The role of Smurf1 in X. laevis development

Previous work on Drosophila showed that dSmurf is essential for embryonic dorso-ventral patterning and hindgut development, and that it acts by spatially and temporally restricting DPP signaling (Podos et al., 2001). Information about the potential developmental function of vertebrate Smurf1, however, has been limited to overexpression studies in Xenopus, where Smurf1 was shown to dorsalize mesoderm and neuralize ectoderm (Zhu et al., 1999; Zhang et al., 2001), and the phenotype of Smurf1 KO mouse (Yamashita et al., 2005), which revealed a somewhat surprising lack of embryonic defects considering that Smurf1 is expressed during gastrula, neurula and later stages of mouse development (Yamashita et al., 2005; G.H.T., unpublished data). However, Yamashita and colleagues (2005) observed a two to threefold increase in Smurf2 transcription in Smurf1 KO mice, which possibly compensated for the loss of Smurf1 during mouse embryonic development. Preliminary tests of Smurf1;Smurf2 double KO mice (Yamashita et al., 2005) indicated embryonic lethality, so the issue of whether or not Smurf1 has a function in vertebrate embryogenesis has remained unresolved.

In our present study, we used Xenopus embryos to investigate the developmental role of Smurf1. The stereotypical fate map of the Xenopus embryo allowed us to target two different loss-of-function reagents, a Smurf1 MO and a dominant-negative mutant, into blastomeres that would give rise to different tissues. We found that Smurf1 knockdown perturbs development of the nervous system, but not the dorsal mesoderm, where Smurf1 might share redundant functions with Smurf2, which has a similar expression pattern and target proteins (Zhang et al., 2001; and our unpublished data). Alternatively, we may be unable at present to effectively block Smurf1 in the dorsal mesoderm (see below). In regard to the neural development, we found that two different aspects require endogenous Smurf1, namely neural patterning and neural folding. Notably, maternal-zygotic Drosophila dSmurf mutants also display both patterning and morphogenetic defects, the latter revealed by the failure of the hindgut to narrow, elongate and undergo closure (Podos et al., 2001). Thus, the ability of Smurfs to regulate both cell fate and cell movement seems to be evolutionarily conserved.

Smurf1 function in mesoderm induction and patterning

Smurf1 knockdown reagents targeted to the mesoderm of intact embryos did not cause obvious defects, but partially blocking BMP signaling in the embryo sensitizes the mesoderm to Smurf1 knockdown: blocking Smurf1 reverses dorsalization and secondary axis induction by Chordin in whole embryos (Figs. 10D, E), and Smurf1 inhibition in animal caps ventralizes their response to dorsal mesoderm-inducing doses of activin (Supplementary Fig. 3). A similar phenomenon was observed in a Chordin knockdown study (Oelgeschlager et al., 2003), where Chordin inhibition by an MO only moderately ventralized whole embryos, whereas it completely blocked mesoderm induction in activin-treated animal caps or secondary axis formation in host embryos implanted with Chordin MO-treated Spemann Organizers. Smurf1 knockdown by itself in isolated animal caps, or the DMZ or VMZ of whole embryos, did not affect development. This might be because even with Smurf1 knockdown, BMP signaling in these tissues is not elevated sufficiently to alter their normal fates. However, when animal caps or VMZ tissues are challenged with activin or BMP inhibitors, respectively, elevated BMP signaling levels caused by partial Smurf1 knockdown is sufficient to affect fate. The question of whether Smurf1 plays a role in dorsal marginal zone development remains unresolved by our present study. Currently available knockdown techniques may be unable to completely block Smurf1 function in Xenopus dorsal mesoderm, possibly because of maternal Smurf1 protein, or because knockdown of Smurf1 in the DMZ/Spemann Organizer is insufficient to boost the absent or low levels of BMP signaling in that region (due to multiple BMP inhibitors in those tissues). Generation and characterization of hypomorphic, rather than null Smurf1;Smurf2 double KO mouse embryos (Yamashita et al., 2005), should help resolve this issue, assuming that Smurf1 function in amphibians and mammals is conserved.

The molecular target of embryonic Smurf1

Smurf1 binds a variety of well-characterized and also novel proteins (Barrios-Rodiles et al., 2005; our unpublished Yeast Two Hybrid data). Stabilization of one or more of these potential substrates could be responsible for Smurf1 KD phenotype. Analysis of how WW domain deletion mutants of Smurf1CA affect Xenopus development (Supplementary Fig. 4) suggests a mechanism of Smurf1CA dominant-negative activity, and helps limit the identity of Smurf1 targets relevant to the present study. Smurf WW domains are known to interact with one or more PY motifs in partner proteins and substrates (Pickart, 2001a). We found that deleting either of the two WW domains in Smurf1CA abolishes its dominant-negative activity, suggesting that Smurf1CA functions by binding a PY-containing substrate(s), thereby protecting it from degradation by endogenous Smurf1. Among proposed Smurf1 substrates, Smad1/5/8, Smad6, Smad7 and MEKK2 possess a PY motif, but Alk6, RhoA or Par6 (that bridges RhoA to Smurf1) does not. Thus, RhoA, an otherwise logical candidate for cell movement effects (Copp et al., 2003), is unlikely to be involved in Smurf1 KD phenotypes. This is supported experimentally by our observation that RhoA overexpression does not mimic Smurf1 knockdown (Supplementary Fig. 3).

We have built a case that the BMP pathway, at the level of Smads, is the major physiological target of Smurf1 during early Xenopus development, but we cannot formally exclude the possibility that another Smurf1 substrate(s) contributes to the observed phenotypes. However, even if such substrates exist, they do not need to be invoked to explain the effects of Smurf1 knockdown. Altered BMP signaling and specifically elevated P-Smad1 activity are sufficient to explain the effects we observe when endogenous Smurf1 is blocked. Since Smads are targeted by Smurf1 via their WW domains, and neither Smad5 nor Smad8 has not been identified in X. laevis, we suggest that Smad1 is the principal endogenous target of Smurf1 in early development of the X. laevis embryo.

Our data show that Smurf1 is required in dorsal ectoderm to restrict BMP/Smad1 signaling to a level that permits neural development. A similar restrictive function has recently been proposed for Ectodermin, a Smad4-specific ubiquitin ligase that limits TGF-β signaling in the ectoderm (Dupont et al., 2005). We suggest that ubiquitin-mediated destruction of intracellular components of the BMP pathway by ubiquitin ligases, such as Smurf1, provides another level of BMP antagonism in addition to extracellular BMP inhibitors secreted from the Spemann organizer. The fact that Smurf1 is expressed in a gradient across the animal pole and marginal zone of the gastrula, with highest levels on the prospective dorsal/anterior side, reinforces this model.

The timing of Smurf1 function

In Smurf1 KD embryos, regional neural markers are affected more than general neural markers (Figs. 5, 6). This suggests that Smurf1 antagonism of BMP/Smad1 signaling becomes essential after general neural induction has taken place, and a variety of evidence supports this possibility. BMP4 and BMP7 are expressed abundantly in proximity to the neural tube in the epidermis and prechordal plate, and in the dorsal neural tube itself (Liem et al., 1995; Hartley et al., 2001). Active P-Smad1/5/8 protein is also directly detectable at the periphery of the anterior neural plate (Wawersik et al., 2005). Furthermore, neural patterning is disrupted in transgenic Xenopus embryos expressing BMP4 under Pax6 promoter (i.e. in the neural plate; Hartley et al., 2001). These data indicate that successful neural development requires BMP/Smad1 pathway suppression continuously after gastrulation. Thus, as development progresses and the neural tube extends and expands anteriorly, BMP inhibition by Organizer- and mesoderm-derived antagonists may become less effective, while local and cell-autonomous mechanisms, such as Smad1 ubiquitylation and dephosphorylation (Knockaert et al., 2006), take over.

Regulation of cell movements by Smurf1 and Smad1

Besides fate and pattern, Smurf1 knockdown affects the organization and behavior of neural plate cells, manifested by defective closure of the anterior neural tube and disruption of factin networks at neural fold hinge points (Fig. 3). Previous studies implicate BMP/Smad1 signaling in regulation of cell movements, independent of cell fate, with examples including zebrafish gastrulation (Myers et al., 2002), pre-migratory neural crest cell delamination from the dorsal neural tube (Sela-Donenfeld and Kalcheim, 1999) and ganglion formation by enteric neural crest cells (Goldstein et al., 2005). The molecular mechanism of how the BMP/Smad1 pathway regulates cell movement, however, remains unknown.

Regulation of cell movement independent of cell fate is commonly accomplished by two non-canonical Wnt pathways, the Wnt/Ca2+ pathway and the Wnt/planar cell polarity (PCP) pathway, whose effectors include Ca2+-binding proteins, actin cytoskeleton regulators, e.g. Rho and Rac, and the Shh pathway (Kuhl et al., 2000; Copp et al., 2003; Ueno and Greene, 2003; Zohn et al., 2003; Klein and Mlodzik, 2005; Park et al., 2006). Moreover, several actin-binding and/or regulating proteins have been specifically implicated in cranial neural tube closure, such as Shroom, vinculin, profilin, Mena, MARCKS and RhoGAP (reviewed in Copp et al., 2003), and knockdown of Shroom in Xenopus embryos produces a phenotype almost identical to Smurf1 knockdown (Haigo et al., 2003). One way BMP/Smad1 signaling might regulate cell movements is transcriptional regulation of a gene(s) involved in actin cytoskeletal dynamics. In zebrafish embryos, high levels of BMP signaling suppress transcription of the non-canonical Wnt ligands Wnt5a and Wnt11 (Myers et al., 2002). However, Wnt5a and Wnt11 are not expressed in Xenopus prospective neural ectoderm nor suppressed in Smurf1 KD embryos as a whole (data not shown), so this mechanism is unlikely to account for the Smurf1 KD phenotype. Moreover, in Smurf1 KD embryos, we measured the expression levels of a number of other actin regulators and components of the non-canonical Wnt pathways (Glypican4, Xfz3, Xfz8, Dishevelled, Strabismus, Shroom, MARCKS and Prickle) by quantitative RT-PCR, but did not observe significant deviation in their levels compared to wt embryos (data not shown). Sonic Hedgehog (Shh) signaling also regulates neural plate patterning and folding, so we analyzed expression of Shh and one of its target genes, Nkx2.2, by in situ hybridization. We found Shh expression mostly unchanged (data not shown), and Nkx2.2 was down-regulated in Smurf1 KD embryos that displayed abnormal neural folding (Figs. 5S–U). However, lack of a tight correlation between Nkx2.2 suppression and folding defects indicates that abnormal Shh signaling is probably not responsible for the described neural tube closure defects. If there is a BMP/Smad1 transcriptional target involved in Xenopus neurulation, it remains to be identified.

Alternatively, Smad1 may be involved at a post-transcriptional level in cytoskeletal regulation. Several cytoplasmic Smad1-interacting factors have recently been identified that potentially regulate actin cytoskeleton and/or cell movements. These include Dishevelled1 (a Wnt signal transducer), filamin (an f-actin binding protein), Par3 (a cell/embryo polarity regulator) and Erbin (which links the tyrosine kinase receptor ErbB2 to the cytoskeleton and desmosomes) (Sasaki et al., 2001; Moustakas et al., 2001; Warner et al., 2003; Zwijsen et al., 2003). Additionally, several members and regulators of the Rho family interact with Smad1, Smad4 and Alk6 in a co-immunoprecipitation screen (Barrios-Rodiles et al., 2005). It would be an intriguing and novel finding if Smad1 operates as both a transcription factor and a cytoplasmic actin regulator, analogous to the dual functions of β-catenin.

Acknowledgments

This work was supported by NIH grant R01HD032429 to G.H.T., and a Stony Brook Institute for Cell and Developmental Biology Predoctoral Fellowship to E.M.A. We thank Drs. T. Bouwmeester, B. Herrmann, T. Sargent, Y. Sasai, J. Wallingford and A. Zaraisky for plasmids, and members of the Thomsen, Holdener and Sirotkin labs for advice and discussion.

Footnotes

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ydbio.2006.08.009.

Publisher's Disclaimer: This article was originally published in a journal published by Elsevier, and the attached copy is provided by Elsevier for the author’s benefit and for the benefit of the author’s institution, for non-commercial research and educational use including without limitation use in instruction at your institution, sending it to specific colleagues that you know, and providing a copy to your institution’s administrator. All other uses, reproduction and distribution, including without limitation commercial reprints, selling or licensing copies or access, or posting on open internet sites, your personal or institution's website or repository, are prohibited. For exceptions, permission may be sought for such use through Elseviers permissions site at: http://www.elsevier.com/locate/permissionusematerial

References

  1. Baker JC, Beddington RSP, Harland RM. Wnt signaling in Xenopus embryos inhibits BMP4 expression and activates neural development. Genes Dev. 1999;13:3149–3159. doi: 10.1101/gad.13.23.3149. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Barrios-Rodiles M, Brown KR, Ozdamar B, Bose R, Liu Z, Donovan RS, Shinjo F, Liu Y, Dembowy J, Taylor IW, Luga V, Przulj N, Robinson M, Suzuki H, Hayashizaki Y, Jurisica I, Wrana JL. High-throughput mapping of a dynamic signaling network in mammalian cells. Science. 2005;307:1621–1625. doi: 10.1126/science.1105776. [DOI] [PubMed] [Google Scholar]
  3. Bryan B, Wrighton K, Wu G, Feng XH, Liu M. Ubiquitination of RhoA by Smurf1 promotes neurite outgrowth. FEBS Lett. 2005;579:1015–1019. doi: 10.1016/j.febslet.2004.12.074. [DOI] [PubMed] [Google Scholar]
  4. Chizhikov VV, Millen KJ. Roof plate-dependent patterning of the vertebrate dorsal central nervous system. Dev Biol. 2005;277:287–295. doi: 10.1016/j.ydbio.2004.10.011. [DOI] [PubMed] [Google Scholar]
  5. Copp AJ, Greene NDE, Murdoch JN. The genetic bases of mammalian neurulation. Nat Rev Genet. 2003;4:784–793. doi: 10.1038/nrg1181. [DOI] [PubMed] [Google Scholar]
  6. Dale L, Wardle FC. A gradient of BMP4 activity specifies dorsal–ventral fates in early Xenopus embryos. Semin Cell Dev Biol. 1999;10:319–326. doi: 10.1006/scdb.1999.0308. [DOI] [PubMed] [Google Scholar]
  7. De Robertis EM, Kuroda H. Dorsal–ventral patterning and neural induction in Xenopus embryos. Annu Rev Cell Dev Biol. 2004;20:285–308. doi: 10.1146/annurev.cellbio.20.011403.154124. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Dupont S, Zacchigna L, Cordenonsi M, Soligo S, Adorno M, Rugge M, Piccolo S. Germ-layer specification and control of cell growth by Ectodermin, a Smad4 ubiquitin ligase. Cell. 2005;121:87–99. doi: 10.1016/j.cell.2005.01.033. [DOI] [PubMed] [Google Scholar]
  9. Ebisawa T, Fukuchi M, Murakami G, Chiba T, Tanaka K, Imamura T, Miyazono K. Smurf1 interacts with transforming growth factor-β type I receptor through Smad7 and induces receptor degradation. J Biol Chem. 2001;276:12477–12480. doi: 10.1074/jbc.C100008200. [DOI] [PubMed] [Google Scholar]
  10. Faure S, Lee MA, Keller T, ten Dijke P, Whitman M. Endogenous patterns of TGFβ superfamily signaling during early Xenopus development. Development. 2000;127:2917–2931. doi: 10.1242/dev.127.13.2917. [DOI] [PubMed] [Google Scholar]
  11. Goldstein AM, Brewer KC, Doyle AM, Nagy N, Roberts DJ. BMP signaling is necessary for neural crest cell migration and ganglion formation in the enteric nervous system. Mech Dev. 2005;122:821–833. doi: 10.1016/j.mod.2005.03.003. [DOI] [PubMed] [Google Scholar]
  12. Grunz H, Tacke L. Neural differentiation of Xenopus laevis ectoderm takes place after disaggregation and delayed reaggregation without inducer. Cell Differ Dev. 1989;1989:207–211. doi: 10.1016/0922-3371(89)90006-3. [DOI] [PubMed] [Google Scholar]
  13. Haigo SL, Hildebrand JD, Harland RM, Wallingford JB. Shroom induces apical constriction and is required for hingepoint formation during neural tube closure. Curr Biol. 2003;13:2125–2137. doi: 10.1016/j.cub.2003.11.054. [DOI] [PubMed] [Google Scholar]
  14. Harland RM. In situ hybridization: an improved whole-mount method for Xenopus embryos. Methods Cell Biol. 1991;36:685–695. doi: 10.1016/s0091-679x(08)60307-6. [DOI] [PubMed] [Google Scholar]
  15. Harland R. Neural induction. Curr Opin Genet Dev. 2000;10:357–362. doi: 10.1016/s0959-437x(00)00096-4. [DOI] [PubMed] [Google Scholar]
  16. Hartley KO, Hardcastle Z, Friday RV, Amaya E, Papalopulu N. Transgenic Xenopus embryos reveal that anterior neural development requires continued suppression of BMP signaling after gastrulation. Dev Biol. 2001;238:168–184. doi: 10.1006/dbio.2001.0398. [DOI] [PubMed] [Google Scholar]
  17. Hata A, Lagna G, Massague J, Hemmati-Brivanlou A. Smad6 inhibits BMP/Smad1 signaling by specifically competing with the Smad4 tumor suppressor. Genes Dev. 1998;15:186–197. doi: 10.1101/gad.12.2.186. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Hemmati-Brivanlou A, Thomsen GH. Ventral mesodermal patterning in Xenopus embryos: expression patterns and activities of BMP-2 and BMP-4. Dev Genet. 1995;17:78–89. doi: 10.1002/dvg.1020170109. [DOI] [PubMed] [Google Scholar]
  19. Hill CS. TGF-β signaling pathways in early Xenopus development. Curr Opin Genet Dev. 2001;11:533–540. doi: 10.1016/s0959-437x(00)00229-x. [DOI] [PubMed] [Google Scholar]
  20. Jones CM, Lyons KM, Lapan PM, Write CVE, Hogan BLM. DVR-4 (Bone Morphogenetic Protein-4) as a posterior-ventralizing factor in Xenopus mesoderm induction. Development. 1992;115:639–647. doi: 10.1242/dev.115.2.639. [DOI] [PubMed] [Google Scholar]
  21. Kaneki H, Guo R, Chen D, Yao Z, Schwarz EM, Zhang YE, Boyce BF, Xing L. Tumor necrosis factor promotes Runx2 degradation through up-regulation of Smurf1 and Smurf2 in osteoblasts. J Biol Chem. 2006;281:4326–4333. doi: 10.1074/jbc.M509430200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Klein T, Mlodzik M. Planar cell polarization: an emerging model points in the right direction. Annu Rev Cell Dev Biol. 2005;21:155–176. doi: 10.1146/annurev.cellbio.21.012704.132806. [DOI] [PubMed] [Google Scholar]
  23. Knockaert M, Sapkota G, Alarcon C, Massague J, Brivanlou AH. Unique players in the BMP pathway: small C-terminal domain phosphatases dephosphorylate Smad1 to attenuate BMP signaling. Proc Nat Acad Sci USA. 2006;103:11940–11945. doi: 10.1073/pnas.0605133103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Kofron M, Demel T, Xanthos J, Lohr J, Sun B, Sive H, Osada S-I, Wright C, Wylie C, Heasman J. Mesoderm induction in Xenopus is a zygotic event regulated by maternal VegT via TGFβ growth factors. Development. 1999;126:5759–5770. doi: 10.1242/dev.126.24.5759. [DOI] [PubMed] [Google Scholar]
  25. Kuhl M, Sheldahl LC, Park M, Miller JR, Moon RT. The Wnt/Ca2+ pathway: a new vertebrate Wnt signaling pathway takes shape. Trends Genet. 2000;16:279–283. doi: 10.1016/s0168-9525(00)02028-x. [DOI] [PubMed] [Google Scholar]
  26. Lebreton S, Jones CM. Emergence of organizer function: a lot of “stuff” involved. Semin Cell Dev Biol. 2006;17:110–116. doi: 10.1016/j.semcdb.2006.01.002. [DOI] [PubMed] [Google Scholar]
  27. Liang Y-Y, Lin X, Liang M, Brunicardi FC, ten Dijke P, Chen Z, Choi K-W, Feng X-H. dSmurf selectively degrades decapentaplegic-activated MAD, and its overexpression disrupts imaginal disc development. J Biol Chem. 2003;278:26307–26310. doi: 10.1074/jbc.C300028200. [DOI] [PubMed] [Google Scholar]
  28. Liem KF, Jr, Tremml G, Roelink H, Jessell TM. Dorsal differentiation of neural plate cells induced by BMP-mediated signals from epidermal ectoderm. Cell. 1995;82:969–979. doi: 10.1016/0092-8674(95)90276-7. [DOI] [PubMed] [Google Scholar]
  29. Lutz M, Knaus P. Integration of the TGF-β pathway into the cellular signaling network. Cell Signal. 2002;14:977–988. doi: 10.1016/s0898-6568(02)00058-x. [DOI] [PubMed] [Google Scholar]
  30. Moody SA. Fates of the blastomeres of the 16-cell stage Xenopus embryo. Dev Biol. 1987;119:560–578. doi: 10.1016/0012-1606(87)90059-5. [DOI] [PubMed] [Google Scholar]
  31. Moren A, Imamura T, Miyazono K, Heldin CH, Moustakas A. Degradation of the tumor suppressor Smad4 by WW and HECT domain ubiquitin ligases. J Biol Chem. 2005;280:22115–22123. doi: 10.1074/jbc.M414027200. [DOI] [PubMed] [Google Scholar]
  32. Moustakas A, Souchelnytskyi S, Heldin C-H. Smad regulation in TGFβ signal transduction. J Cell Sci. 2001;114:4359–4369. doi: 10.1242/jcs.114.24.4359. [DOI] [PubMed] [Google Scholar]
  33. Murakami G, Watabe T, Takaoka K, Miyazono K, Imamura T. Cooperative inhibition of bone morphogenetic protein signaling by Smurf1 and inhibitory Smads. Mol Biol Cell. 2003;14:2809–2817. doi: 10.1091/mbc.E02-07-0441. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Myers DC, Sepich DS, Solnica-Krezel L. BMP activity gradient regulates convergent extension during zebrafish gastrulation. Dev Biol. 2002;243:81–98. doi: 10.1006/dbio.2001.0523. [DOI] [PubMed] [Google Scholar]
  35. Nieuwkoop PD, Faber J. Normal Table of Xenopus laevis (Daudin) North Holland: Amsterdam; 1967. [Google Scholar]
  36. Novoselov VV, Alexandrova EM, Ermakova GV, Zaraisky AG. Expression zones of three novel genes abut the developing anterior neural plate of Xenopus embryo. Gene Expr Patterns. 2003;3:225–230. doi: 10.1016/s1567-133x(02)00077-7. [DOI] [PubMed] [Google Scholar]
  37. Oelgeschlager M, Kuroda H, Reversade B, De Robertis EM. Chordin is required for the Spemann organizer transplantation phenomenon in Xenopus embryos. Dev Cell. 2003;4:219–230. doi: 10.1016/s1534-5807(02)00404-5. [DOI] [PubMed] [Google Scholar]
  38. Ozdamar B, Bose R, Barrios-Rodiles M, Wang H-R, Zhang Y, Wrana JL. Regulation of the polarity protein Par6 by TGFβ receptors controls epithelial cell plasticity. Science. 2005;307:1603–1609. doi: 10.1126/science.1105718. [DOI] [PubMed] [Google Scholar]
  39. Park TJ, Haigo SL, Wallingford JB. Ciliogenesis defects in embryos lacking inturned or fuzzy function are associated with failure of planar cell polarity and Hedgehog signaling. Nat Genet. 2006;38:303–311. doi: 10.1038/ng1753. [DOI] [PubMed] [Google Scholar]
  40. Pickart CM. Mechanisms underlying ubiquitination. Annu Rev Biochem. 2001a;70:503–533. doi: 10.1146/annurev.biochem.70.1.503. [DOI] [PubMed] [Google Scholar]
  41. Pickart CM. Ubiquitin enters the new millennium. Mol Cell. 2001b;8:499–504. doi: 10.1016/s1097-2765(01)00347-1. [DOI] [PubMed] [Google Scholar]
  42. Podos SD, Hanson KK, Wang Y-C, Ferguson EL. The dSmurf ubiquitin–protein ligase restricts BMP signaling spatially and temporally during Drosophila embryogenesis. Dev Cell. 2001;1:567–578. doi: 10.1016/s1534-5807(01)00057-0. [DOI] [PubMed] [Google Scholar]
  43. Sasaki A, Masuda Y, Ohta Y, Ikeda K, Watanabe K. Filamin associates with Smads and regulates transforming growth factor-β signaling. J Biol Chem. 2001;276:17871–17877. doi: 10.1074/jbc.M008422200. [DOI] [PubMed] [Google Scholar]
  44. Schohl A, Fagotto F. β-catenin, MAPK and Smad signaling during early Xenopus development. Development. 2002;129:37–52. doi: 10.1242/dev.129.1.37. [DOI] [PubMed] [Google Scholar]
  45. Sela-Donenfeld D, Kalcheim C. Regulation of the onset of neural crest migration by coordinated activity of BMP4 and Noggin in the dorsal neural tube. Development. 1999;126:4749–4762. doi: 10.1242/dev.126.21.4749. [DOI] [PubMed] [Google Scholar]
  46. Suzuki C, Murakami G, Fukuchi M, Shimanuki T, Shikauchi Y, Imamura T, Miyazono K. Smurf1 regulates the inhibitory activity of Smad7 by targeting Smad7 to the plasma membrane. J Biol Chem. 2002;277:39919–39925. doi: 10.1074/jbc.M201901200. [DOI] [PubMed] [Google Scholar]
  47. Tao Q, Yokota C, Puck H, Kofron M, Birsoy B, Yan D, Asashima M, Wylie CC, Lin X, Heasman J. Maternal Wnt11 activates the canonical Wnt signaling pathway required for axis formation in Xenopus embryos. Cell. 2005;120:857–871. doi: 10.1016/j.cell.2005.01.013. [DOI] [PubMed] [Google Scholar]
  48. Ueno N, Greene ND. Planar cell polarity genes and neural tube closure. Birth Defects Res C Embryo Today. 2003;69:318–324. doi: 10.1002/bdrc.10029. [DOI] [PubMed] [Google Scholar]
  49. Vasiliev OL, Lukyanov SA, Belyavsky AV, Kazanskaya OV, Zaraisky AG. A novel marker of early epidermal differentiation: cDNA subtractive cloning starting on a single explant of Xenopus laevis gastrula epidermis. Int J Dev Biol. 1997;41:877–882. [PubMed] [Google Scholar]
  50. von Bubnoff A, Cho KWY. Intracellular BMP signaling regulation in vertebrates: pathway or network? Dev Biol. 2001;239:1–14. doi: 10.1006/dbio.2001.0388. [DOI] [PubMed] [Google Scholar]
  51. Wang H-R, Zhang Y, Ozdamar B, Ogunjimi AA, Alexandrova E, Thomsen GH, Wrana JL. Regulation of cell polarity and protrusion formation by targeting RhoA for degradation. Science. 2003;302:1775–1779. doi: 10.1126/science.1090772. [DOI] [PubMed] [Google Scholar]
  52. Warner DR, Pisano MM, Roberts EA, Greene RM. Identification of three novel Smad binding proteins involved in cell polarity. FEBS Lett. 2003;539:167–173. doi: 10.1016/s0014-5793(03)00155-8. [DOI] [PubMed] [Google Scholar]
  53. Wawersik S, Evola C, Whitman M. Conditional BMP inhibition in Xenopus reveals stage-specific roles for BMPs in neural and neural crest induction. Dev Biol. 2005;277:425–442. doi: 10.1016/j.ydbio.2004.10.002. [DOI] [PubMed] [Google Scholar]
  54. Wills A, Harland RM, Khokha MK. Twisted gastrulation is required for forebrain specification and cooperates with Chordin to inhibit BMP signaling during X. tropicalis gastrulation. Dev Biol. 2006;289:166–178. doi: 10.1016/j.ydbio.2005.10.022. [DOI] [PubMed] [Google Scholar]
  55. Wilson L, Maden M. The mechanisms of dorsoventral patterning in the vertebrate neural tube. Dev Biol. 2005;282:1–13. doi: 10.1016/j.ydbio.2005.02.027. [DOI] [PubMed] [Google Scholar]
  56. Wilson PA, Lagna G, Suzuki A, Hemmati-Brivanlou A. Concentration-dependent patterning of the Xenopus ectoderm by BMP4 and its signal transducer Smad1. Development. 1997;124:3177–3184. doi: 10.1242/dev.124.16.3177. [DOI] [PubMed] [Google Scholar]
  57. Xanthos JB, Kofron M, Wylie C, Heasman J. Maternal VegT is the initiator of a molecular network specifying endoderm in Xenopus laevis. Development. 2001;128:167–180. doi: 10.1242/dev.128.2.167. [DOI] [PubMed] [Google Scholar]
  58. Yamashita M, Ying S-X, Li C, Cheng SY, Deng C, Zhang YE. Ubiquitin ligase Smurf1 controls osteoblast activity and bone homeostasis by targeting MEKK2 for degradation. Cell. 2005;121:101–113. doi: 10.1016/j.cell.2005.01.035. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Ying S-X, Hussain ZJ, Zhang YE. Smurf1 facilitates myogenic differentiation and antagonizes the bone morphogenetic protein-2-induced osteoblast conversion by targeting Smad5 for degradation. J Biol Chem. 2003;278:39029–39037. doi: 10.1074/jbc.M301193200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Zhang Y, Chang C, Gehling DJ, Hemmati-Brivanlou A, Derynck R. Regulation of Smad degradation and activity by Smurf2, an E3 ubiquitin ligase. Proc Natl Acad Sci. 2001;98:974–979. doi: 10.1073/pnas.98.3.974. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Zhao M, Qiao M, Oyajobi BO, Mundy GR, Chen D. E3 ubiquitin ligase Smurf1 mediates core-binding factor ±1/Runx2 degradation and plays a specific role in osteoblast differentiation. J Biol Chem. 2003;278:27939–27944. doi: 10.1074/jbc.M304132200. [DOI] [PubMed] [Google Scholar]
  62. Zhao M, Qiao M, Harris SE, Oyajobi BO, Mundy GR, Chen D. Smurf1 inhibits osteoblast differentiation and bone formation in vitro and in vivo. J Biol Chem. 2004;279:12854–12859. doi: 10.1074/jbc.M313294200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Zhu H, Kavsak P, Abdollah S, Wrana JL, Thomsen GH. A SMAD ubiquitin ligase targets the BMP pathway and affects embryonic pattern formation. Nature. 1999;400:687–693. doi: 10.1038/23293. [DOI] [PubMed] [Google Scholar]
  64. Zohn IE, Chesnutt CR, Niswander L. Cell polarity pathways converge and extend to regulate neural tube closure. Trends Cell Biol. 2003;13:451–454. doi: 10.1016/s0962-8924(03)00173-9. [DOI] [PubMed] [Google Scholar]
  65. Zwijsen A, Verschueren K, Huylebroeck D. New intracellular components of bone morphogenetic protein/Smad signaling cascades. FEBS Lett. 2003;546:133–139. doi: 10.1016/s0014-5793(03)00566-0. [DOI] [PubMed] [Google Scholar]

RESOURCES