Developmental Biology 329 (2009) 338–349

Contents lists available at ScienceDirect

Developmental Biology

j ourna l homepage: www.e lsev ie r.com/deve lopmenta lb io logy

Syndecan-1 regulates BMP signaling and dorso-ventral patterning of the ectodermduring early Xenopus development

Gonzalo H. Olivares, Héctor Carrasco, Francisco Aroca, Loreto Carvallo, Fabián Segovia, Juan Larraín ⁎Center for Aging and Regeneration, Center for Cell Regulation and Pathology, MIFAB, Faculty of Biological Sciences, P. Universidad Católica de Chile, Alameda 340, Santiago, Chile

⁎ Corresponding author. Fax: +56 2 686 2824.E-mail address: jlarrain@bio.puc.cl (J. Larraín).

0012-1606/$ – see front matter © 2009 Elsevier Inc. Aldoi:10.1016/j.ydbio.2009.03.007

a b s t r a c t

a r t i c l e i n f o

Article history:Received for publication 1 September 2008Revised 6 March 2009Accepted 6 March 2009Available online 18 March 2009

Keywords:Syndecan1HSPGBMP signalingDorso-ventral patterningEctodermChordin

Extracellular regulation of growth factor signaling is a key event for embryonic patterning. Heparan sulfateproteoglycans (HSPG) are among the molecules that regulate this signaling during embryonic development.Here we study the function of syndecan1 (Syn1), a cell-surface HSPG expressed in the non-neural ectodermduring early development of Xenopus embryos. Overexpression of Xenopus Syn1 (xSyn1) mRNA is sufficientto reduce BMP signaling, induce chordin expression and rescue dorso-ventral patterning in ventralizedembryos. Experiments using chordin morpholinos established that xSyn1 mRNA can inhibit BMP signaling inthe absence of chordin. Knockdown of xSyn1 resulted in a reduction of BMP signaling and expansion of theneural plate with the concomitant reduction of the non-neural ectoderm. Overexpression of xSyn1 mRNA inxSyn1 morphant embryos resulted in a biphasic effect, with BMP being inhibited at high concentrations andactivated at low concentrations of xSyn1. Interestingly, the function of xSyn1 on dorso-ventral patterning andBMP signaling is specific for this HSPG. In summary, we report that xSyn1 regulates dorso-ventral patterningof the ectoderm through modulation of BMP signaling.

© 2009 Elsevier Inc. All rights reserved.

Introduction

Tissue patterning in early embryonic development is specified bymorphogen gradients, among which are members of the BMP, Wntand Hh families. A key role for extracellular components inmorphogen gradient formation and interpretation has been demon-strated in fly and vertebrate embryos (Freeman and Gurdon, 2002;Lander et al., 2002; Tabata and Takei, 2004). The current under-standing of extracellular regulation of BMP signaling is particularlywell advanced, with many of the evolutionary conserved proteins thatdirectly interact with BMP having been identified. This set ofmolecules play a key role in the establishment of a very robustextracellular gradient of BMP signaling that patterns the germ layersin vertebrate and invertebrate embryos (De Robertis, 2006; O'Connoret al., 2006). Interestingly, many BMP binding proteins such aschordin/sog, twisted gastrulation (Tsg) and crossveinless2 (Cv2) havedual effects on BMP signaling and they can promote or inhibit BMPactivity depending on the cellular and molecular context (De Robertiset al., 2000; Larraín et al., 2001; Ambrosio et al., 2008; Bier, 2008;Serpe et al., 2008). This dual function could be useful for creation ofsharp borders along embryonic territories (De Robertis et al., 2000;Larraín et al., 2001).

Heparan sulfate proteoglycans (HSPG) also play an importantfunction in extracellular regulation of growth factor signaling duringembryonic development (Lin, 2004). HSPG are cell-surface and

l rights reserved.

extracellular matrix (ECM) components formed by a core proteinwith covalently attached heparan sulfate (HS) glycosaminoglycan(GAGs) chains. HS are linear polysaccharide chains made of repeateddisaccharide units composed of glucuronic acid and a glucosaminederivative. Depending on its core protein, HSPG can be classified indifferent families, including two cell-surface proteoglycans that areconserved from flies to vertebrates, glypicans and syndecans (Syn).Glypicans are GPI-anchored, while Syn are type I transmembraneproteins consisting of a single transmembrane domain and a shortconserved intracellular domain (Carey,1997). Themajority of theGAGsadded to glypicans and Syn ectodomains are of the HS type, althoughmodifications of Syn1 and 4 with chondroitin sulfate chains (CS) havebeen described (Carey, 1997). One Syn and two glypicans (dally anddally-like protein) genes are encoded in the Drosophila genome andfour Syn and at least six glypicans have been identified in vertebrates.

HS chains bind to a great variety of growth factors includingmembers of the TGFβ, Hh, Wnt and FGF families and are responsiblefor many HSPG functions (Lin, 2004). The role of HSPG, particularly ofSyn1 in FGF signaling has been extensively studied. HS chains can havea biphasic effect on FGF signaling; they can increase FGF binding to itscognate receptor or can reduce growth factor bioavailability throughsequestration, modulation of ligand stability and/or localization (Maliet al., 1993; Carey, 1997; Larraín et al., 1998; Esko and Selleck, 2002;Hacker et al., 2005; Bishop et al., 2007). A similar biphasic effect forHSPG on Shh activity has been described (Carrasco et al., 2005;Beckett et al., 2008).

Even though most HSPG functions are attributed to its HS chains,the core proteins also have important roles in cell–cell signaling

339G.H. Olivares et al. / Developmental Biology 329 (2009) 338–349

regulation. Syn1 modulates integrin activation and binding to theepithelial mitogen lacritin through extracellular domains on its coreprotein (Beauvais et al., 2004; Ma et al., 2006; McQuade et al., 2006).In addition, Syn2 and 4 control cell adhesion through extracellulardomains depleted of GAG chains (McFall and Rapraeger, 1998; Park etal., 2002). In the same venue, the core protein of glypican3 interactsand regulates Shh activity even when the GAG attachment sites aremutated (Capurro et al., 2005, 2008). Similarly, a glypican4 GAGmutant can regulate non-canonical Wnt signaling and gastrulation(Ohkawara et al., 2003) and a dally GAG mutant rescued somephenotypes of dally fly mutants (Kirkpatrick et al., 2006). In summary,HSPG regulates growth factor signaling through both GAG-dependentand -independent mechanisms.

Genetic studies in flies, worms, zebrafish and mice revealed theimportance of HSPG in cell signaling and morphogen gradientformation during early development, organ formation and adultphysiology (Lin, 2004; Hacker et al., 2005; Bulow and Hobert, 2006;Bishop et al., 2007). In the majority of these studies, genes encodingfor HS biosynthetic enzymes were mutated and the effect onembryonic development analyzed. HS biosynthetic enzymes arecommon for all members of the HSPG families; therefore, theexperimental approach mentioned was not informative enough toelucidate the functions of each specific proteoglycan. In order toaddress the role of each HSPG, the effect of mutations on genesencoding the core proteins has been studied, revealing that glypicansand its fly orthologues (dally, dally-like protein) regulate Wg, Shh, FGFand BMP/dpp signaling (Topczewski et al., 2001; Lin, 2004; Hacker etal., 2005; Song et al., 2005; Capurro et al., 2008).

To date, no function for Syn1 in early development has beenreported. At later stages Syn1 is involved in axon pathfinding in fliesand worms (Johnson et al., 2004; Steigemann et al., 2004; Rhiner etal., 2005) and is required for re-epithelialization after woundhealing in adult mice (Stepp et al., 2002). In addition, Wnt1 isunable to induce mammary tumors in Syn1 null-mice, suggestingthat Syn1 is required for Wnt1 signaling (Alexander et al., 2000). InXenopus, Syn2 is required for the establishment of left-rightasymmetry (Kramer and Yost, 2002) and Syn4 regulates gastrula-tion (Muñoz et al., 2006). During early development, Xenopussyndecan1 (xSyn1) is expressed in the animal cap at the blastulastage and later is excluded from the neural plate and restricted tothe non-neural ectoderm (Teel and Yost, 1996), similar to theexpression of BMP4 (Fainsod et al., 1994). Although exhibiting adynamic expression pattern, the role of xSyn1 during earlydevelopment has not been addressed.

This study aims to address the function of Syn1 during earlyXenopus development. Through gain- and loss-of-function experi-ments we have demonstrated that xSyn1 modulates BMP signalingwith a biphasic effect. At high concentrations xSyn1 mRNA blocksBMP activity and at low concentration it increases BMP signaling.This effect is specific for xSyn1 and seems to be independent of itsGAG chains. In addition, morpholino knockdown of xSyn1 producedan expansion of the neural plate and reduction of the non-neuralectoderm. In summary, we have found that xSyn1 regulates dorso-ventral patterning of the ectoderm through modulation of BMPsignaling.

Materials and methods

Embryos and explants

Xenopus embryos obtained by in vitro fertilizationweremaintainedin 0.1× modified Barth saline solution (Sive et al., 2000, Moreno et al.,2005) and staged according to Nieuwkoop and Faber (1967).Ectodermal explants were excised at stage 9 and cultured in 1×modified Barth saline until sibling embryos reached the requiredstage. Dorsal and ventral marginal zones were dissected in 1×

modified Barth saline solution (Sive et al., 2000) and directly lysedfor RNA preparation. The dorsoanterior index (DAI) was estimated atstage 33 according to Kao and Elinson (1988). In situ hybridizationwas performed onwhole embryos as described (Lemaire and Gurdon,1994; Belo et al., 1997; Sive et al., 2000; http://www.hhmi.ucla.edu/derobertis/index.html). For lineage tracing of injected cells, 50 pg oflacZmRNAwas co-injected and visualized by Red-Gal staining. Digitalphotographs were taken with a Nikon Digital Sight DS-5M digitalcamera.

Morpholino oligos

AntisenseMO forXenopus laevis xSyn1were obtained fromGene ToolsLLC and consisted of the following sequences: xSyn1.1 MO, 5′-TTGTCCATCGTATCAGTGTGC-GCTG-3′; xSyn1.2 MO, 5′-ACAGGGCAACCA-ATAACTTGTCCAT-3′. Bmp4 and Bmp7 MOs (Reversade et al., 2005);Chordin MOs (Oelgeschlager et al., 2003); Noggin MO (Kuroda et al.,2004); β-catenin MO (Heasman et al., 2000); xSyn4 MOs (Muñoz et al.,2006) were as described. MOs were resuspended in sterile water to aconcentration of 1 mM. For xSyn1 MOs, a 1:1 ratio of both morpholinoswas used. A total of 4 nl (two times 2 nl) of morpholino solutions wasinjected at the two-cell stage or 8 nl (two times 4 nl or four times 2 nl) ofmorpholino solution injected at the four-cell stage.

RNA Injections

All mRNAs were synthesized in vitrowith SP6, T7 or T3 polymeraseusing the Message Machine kit (Ambion Inc.). For β-catenin MOrescue assays, capped mRNAs with or without additional MOs weremicroinjected into the equatorial region at four-cell stage.

BMP reporter assay in embryos

BMP activity was evaluated in embryos with two BMP-specificluciferase reporters: Msx2-luc (Sirard et al., 2000) and 4XBRE-luc(Korchynskyi and ten Dijke, 2002). For reporter assays, embryos weremicroinjected twice into animal region at two-cell stage with plasmidreporter and mRNAs and/or morpholinos. Ectodermal explants from15 embryos were excised at stage 9 and cultured in 1× modified Barthsaline solution until sibling embryos reached stage 11 and directlylysed in groups of five explants (using 20 μl Reporter Lysis Buffer/explant).

Protein analysis

Phospho-protein extracts were obtained as described (Birsoy etal., 2005) with some modifications. Briefly, protein extracts wereprepared from batches of 10 embryos using 10 μl homogenizationbuffer/embryo. Homogenization buffer for anti-phospho-Smad1blots was 20 mM Tris pH 8.0, 2 mM EDTA, 5 mM EGTA, 0.5% NP-40, 2 mM sodium orthovanadate, 100 mM NaF, 10 mM sodiumpyrophosphate containing protease inhibitors. Homogenates werespun down at 15,000 g for 10 min and the supernatant wascollected. 15 μl of samples were mixed with 7 μl of sample buffer.Samples were boiled and loaded on 10% SDS-PAGE gels andtransferred into PVDF membranes. The membranes were blockedfor 1–2 h at room temperature. Anti-Phospho-Smad1 antibody (CellSignaling Technology) was used at a dilution of 1:2000. As a loadingcontrol, membranes were stripped after signal detection, andincubated with anti-Smad1 antibody (Zymed) at a dilution of1:1000. Signal detection was carried out using the PIERCE ECLdetection system.

Heparan sulfate proteoglycans were detected as described (Stein-feld et al., 1996; Casar et al., 2004; Gutiérrez et al., 2006). Briefly, 50–100 ectodermal explants were obtained from embryos microinjectedwith mRNAs or MOs and at gastrula stage were lysed with 10 mM Tris

Fig. 1. xSyndecan-1 rescues dorsal structures in ventralized embryos. (A, B) xSyn1induces dorsal axes in Xenopus embryos. Four-cell stage embryos were microinjectedinto the vegetal region of one ventral blastomere with synthetic mRNA for xSyn1(400 pg). Embryos were allowed to develop until stage 22 and double axes quantitated.(C, D) xSyn1 rescue dorsal structures in βcat-morphant embryos. Two-cell embryoswere microinjected in each blastomere with MO-βcat (17 ng total) and after onecleavage, one blastomere was microinjected with different doses of xSyn1 syntheticmRNA (6–100 pg). Embryos were allowed to develop until stage 33 and DAI wasdetermined (Kao and Elinson, 1988). The amounts of xSyn1 used for rescue of βcat-morphant embryos have no effect on dorso-ventral patterning of wild type embryos.Note that as little as 6 pg of xSyn1 mRNA partially rescued dorsal structures and thatmore than 25 pg gave rise to hyperdorsalized embryos (blue). The graph is the averageof four independent experiments.

340 G.H. Olivares et al. / Developmental Biology 329 (2009) 338–349

pH 7.5, 150 mM NaCl, 0.5% Triton X-100 and protease inhibitors (2 μlhomogenization buffer/explant). Homogenates were spun down at15,000 g for 10 min and the supernatant was collected. 30 μl ofsamples were treated with 0.5 mU heparitinase obtained from Fla-vobacterium heparinum (Seikagaku, Japan) and separated by SDS-PAGE on a 4–15% gradient gel. Proteins were transferred on PVDF andsubjected to western blot analysis with anti-Δ-heparan sulfatemonoclonal antibody (3G10 epitope, US Biologicals), named anti-Δ-HS at a dilution of 1:2500. As a loading control, membranes werestripped after signal detection, and incubated with anti-β-tubulinantibody (Sigma) at 1:10,000.

RT-PCR

For RT-PCR analyses, RNA was pooled from ten ectodermalexplants, or three whole embryos. RT-PCR conditions and primersare described at http://www.hhmi.ucla.edu/derobertis/index.html.

DNA constructs

To obtain full length Syndecan1 from Xenopus, gastrula stage cDNAwasamplifiedusing twopairs of primers: thefirst pairwasdirected to the5′ coding sequence, containing an internal XhoI site: Fw 5′-ATGGA-CAAGTTACTGGCT-3′ and Rv 5′-TTCATCATGCCCAATAAGCA-3′, the secondpair was directed to the 3′ coding sequence, containing an internal XhoIsite and a silentmutation into the EcoRI site (bolded nucleotides) presentnear the stop codon: Fw 5′-TGGAATTTGAGGAGACAACTGA-3′ and Rv 5′-CTACGCGTAAAACTCCCG-3′. Each PCR fragment was subcloned intopGEM-T easy (Promega). Both clones were digested with EcoRI andXhoI and ligated into pCS2 to obtainpCS2-xSyn1. xSyn1was excised fromthepCS2withEcoRI and ligated intopGEM-TeasyandpcDNA3.We foundthat the mRNA prepared from pGEM-T easy was more active and wasused for all experiments. Flag-tagged full length xSyn1 was obtained byPCR using the primers: Fw 5′-GGATCCACGGATGTGAGCGTGAGA-3′ andRv 5′-CTTCTCATGATGACCTGC-3′ and the product was subcloned intopGEM-NChd-flag excised with BamHI/XhoI. A truncated form of xSyn1lacking the cytoplasmic domain (xSyn1-ΔC) was generated using thefollowingprimer 5′-CATGTGGAGCTCTCAATAGAGCATAAACCCAAC-3′. Forfunctional analysis of GAG chains of xSyn1, pcDNA3.1-flag-xSyn1was used as template for oligonucleotide directed mutagenesis(QuickChange Multi Site-Directed Mutagenesis Kit, Stratagene),which was used to generate a mutagenic construct of xSyn1 inwhich serines 34, 41, 43, 190, 244 and 497 were converted intoalanine (xSyn1-ΔGAG).

siRNA knockdown of Syn1 in C2C12 cells

siRNA against mouse Syn1 (siSyn1) (GenBankTM accessionnumber NM_011519.1, nucleotide annotation 1660GAGGTCTACTTTA-GACAACTT

1680) purchased from Ambion, Inc. was used in C2C12 cells

as described (McQuade et al., 2006). siSyn1 sequence was 5′-AAGUUGUCUAAAGUAGACCTC-3′. 70,000 cells/well were plated in asix-well plate. 24 h later, cells were transfected using LipofectA-MINE2000 (Invitrogen) containing a mix of 50 nM siSyn1; 1.5 ng/μlof BRE-luc and 0.1 ng/μl of control pRL Renilla luciferase plasmid inOptiMEM I transfection medium (Invitrogen) lacking serum andantibiotics. Control cells were transfected with a control oligonu-cleotide provided by the manufacturer. 24 h after transfection, cellswere trypsinized and 10,000 cells/well were plated in a 96-wellplate. Cells were serum starved for 12 h. The next day cells wereinduced with recombinant BMP4 (R&D Systems), and luciferaseactivity was measured 1 day later. Luciferase assays wereperformed using the Dual Luciferase Reporter Assay system(Promega) as described. In parallel, Syn1 protein expression wasanalyzed by western blot from homogenates of C2C12 transfectedwith siSyn1.

Results

Overexpression of xSyndecan1 mRNA rescues dorso-ventral patterning inXenopus ventralized embryos

To evaluate xSyn1 function, synthetic mRNA was prepared andinjected at the four-cell stage. We found that injection of 400 pg ofxSyn1 mRNA in one ventral and vegetal blastomere resulted in theformation of partial secondary axes (Figs. 1A and B, 70%, N=103).Axes with anterior structures (cement gland and eyes) were neverobserved and the embryos died when amounts higher than 1 ng wereinjected (data not shown).

Secondary axes in Xenopus embryos can be obtained by over-expression of components of the Wnt/β-catenin or Nodal pathwaysand by antagonizing the BMP signaling pathway (De Robertis et al.,

341G.H. Olivares et al. / Developmental Biology 329 (2009) 338–349

2000). Due to the fact that Syn1 is required for Wnt1 signaling in mice(Alexander et al., 2000) and HS chains are necessary for Wnt11signaling in Xenopus oocytes (Tao et al., 2005), we reasoned thatxSyn1 could induce double axes via increasing Wnt/β-cateninsignaling. To test this hypothesis we inhibited Wnt signaling in earlydevelopment by knockdown of β-catenin using morpholinos (Heas-man et al., 2000) and evaluated the effect of xSyn1 in this background.In the case that xSyn1 function to increase Wnt signaling, over-expression of xSyn1mRNAwould have no effect on embryos depletedof β-catenin. Xenopus embryos at the two-cell stage were micro-injected in both blastomeres with morpholino against β-catenin (MO-βcat, denominated βcat-morphant embryos in this manuscript) and atthe four-cell stage received a single injection of xSyn1 mRNA. Wefound that microinjection of xSyn1 mRNA was sufficient to rescuedorsal structures in βcat-morphant embryos (Figs. 1C and D). β-catenin knockdown resulted in completely ventralized embryos (80%DAI=0, N=140) as described before (Heasman et al., 2000).Microinjection of only 6 pg of xSyn1 mRNA partially restored dorsalstructures (60% DAI=1–2; 35% DAI=3–4, N=32) and higher dosesof the mRNA (50 pg) gave rise to hyperdorsalized embryos (30%

Fig. 2. xSyndecan-1 rescue of dorso-ventral patterning in a cell non-autonomous manner. (followed by one (1×, 40 pg of total mRNA) or four (4×, 80 pg of total mRNA) microinjectionstage 18, (H–K) 10 or (L) 20 andmolecular markers analyzed by in situ hybridization using prmolecular markers observed by in situ hybridization analysis was very penetrant. All βcat mN=12–15) and showed an expansion of sizzled (panel F, N=11). One injection of xSyn1 mR4× microinjections induce radial expression of sox2 and chordin in 18% (N=11) and 71% (NEmbryos at two-cell stage were microinjected with MO-βcat in both blastomeres, followed bdevelop, fixed at stage 27 and stained for β-galactosidase. (N) Section of one embryo, showinsee the insert). (O) xSyn1 rescue of Spemann Organizer specific markers. Embryos at the twoof xSyn1 mRNA in all blastomeres at the four-cell stage. RNA was isolated at stages 9, 10.5 a

DAI=8–10, N=124). In addition, a similar rescue effect was observedwhen a cytoplasmic deletion construct for xSyn1 was injected (seeSupplementary information fig. S1), indicating that the interaction ofSyn1 with intracellular signaling components is not required for itsdorsal rescue activity.

To further evaluate the ability of xSyn1 to rescue dorso-ventralpatterning, molecular markers were evaluated by in situ hybridizationand RT-PCR analysis. A single injection of xSyn1mRNA (40 pg) in βcat-morphant embryos was able to restore the expression of dorsalmarkers such as sox2, NCAM and α-actin (Figs. 2A–D and L) andventral markers including sizzled and msx1 (Figs. 2E–G and L) tonormal levels, but proved unable to do the same with cytokeratin (Fig.2L). No effect on Xbrawas observed, suggesting that xSyn1 has no roleon mesoderm formation (Fig. 2L). Of note, microinjection of thisamount of xSyn1 mRNA in wild type embryos had no effect in thelevels of molecular markers analyzed by RT-PCR (Fig. 2L, comparelanes 1 and 4). From these results we concluded that xSyn1 is active inthe absence of a functional Wnt pathway, indicating that its ability toinduce dorsal structures is independent of Wnt/β-catenin signaling.This was confirmed by the observation that overexpression of xSyn1

A–L) Embryos at two-cell stage were microinjected with MO-βcat in both blastomeres,s of xSyn1 mRNA at the four cell stage. Embryonic development was stopped at (A–G)obes against (A–D) sox2, (E–G) sizzled and (H–K) chordin or by (L) RT-PCR. The rescue oforphant embryos are completely devoid of sox2 and chordin expression (panels B and I,NA is 100% efficient in restoring dorso-ventral patterning (panels C, G and J, N=5–17).=7) of the embryos, respectively. (M, N) xSyn1 acts in a cell non-autonomous manner.y one microinjection of xSyn1 (40 pg) and lacZ (50 pg) at the four-cell stage, allowed tog that of all dorsal tissues only notochord was stained with β-galactosidase (for details-cell were microinjected with MO-βcat in both blastomeres, followed bymicroinjectionnd 11.5 and expression of organizer genes evaluated by RT-PCR.

342 G.H. Olivares et al. / Developmental Biology 329 (2009) 338–349

mRNA in βcat-morphant embryos was not able to restore expressionof Xnr3 and siamois (see Supplementary information fig. S2), twodirect targets of the Wnt/β-catenin pathway. In addition, we foundthat knockdown of β-catenin provides a sensitized background thatfacilitates the visualization of xSyn1 function.

To gain insight into the mechanism of double axes induction byxSyn1, we performed lineage tracing experiments. xSyn1 and β-galactosidase (lacZ) mRNAs were co-injected marginally in oneblastomere at the four-cell stage in βcat-morphant embryos. Embryoswere raised until the tadpole stage and stained for the presence oflacZ. The cells derived from the blastomere that received both mRNAs(xSyn1/lacZ) were found mainly in the notochord and headmesoderm (Figs. 2M, N), indicating that xSyn1 rescues dorso-ventralpatterning through a cell non-autonomous effect. These results werevery similar to those obtained in previous studies, where ventralizedembryos were rescued by overexpression of Spemann's organizergenes, such as noggin and chordin (Smith and Harland, 1992; Sasai etal., 1994). Due to the results discussed above, we decided to test theability of xSyn1 to induce Spemann's organizer genes. For this, weevaluated the expression of organizer genes in βcat-morphantembryos injected with xSyn1 mRNA. xSyn1 was able to rescue chor-din, goosecoid and ADMP, although with a slight delay in comparisontowild type embryos (Figs. 2H–K, and O). Additionally, xSyn1was ableto induce ectopic expression of chordin in wild type embryos (see

Fig. 3. Overexpression of xSyndecan1 inhibits BMP signaling. (A, B) Activation of BMPmicroinjected in one blastomere at the four-cell stage with synthetic mRNAs for xSyn1 (20 pembryos were raised until stage 25 and the presence of dorsal axes determined. The graph reinhibits Smad-1 phosphorylation. Wild type or βcat-morphant embryos at the four-cell stamRNA) with xSyn1 mRNA, embryos were cultured until stage 10.5, homogenized and phospextracts from 64-cell stage embryos were used (lane 1). Note that xSyn1was able to decreaseembryos (compare lane 3 with 4 and 5). Total Smad1/5/8 was measured as loading controlstage in all blastomeres with BMP reporter plasmid (Msx-luc, 30 pg), BMP4 (75 pg) and xluciferase activity determined.

Supplementary information fig. S3). In contrast, noggin was onlypartially rescued and the levels of cerberus were not re-established(Fig. 2O). These results indicate that xSyn1 can restore expression ofsome Spemann's organizer components, suggesting that the ability torescue dorso-ventral patterning can be mediated by induction ofSpemann's organizer genes such as chordin.

Overexpression of xSyndecan1 reduces BMP signaling

Another possibility to induce secondary axis is through reductionof BMP signaling. Since partial secondary axes induced by xSyn1 (Fig.1B) were similar to those obtained when BMP antagonists wereinjected (Larraín et al., 2000) and because the βcat-morphantembryos are more sensitive to respond to BMP inhibition (Reversadeet al., 2005), we decided to test whether xSyn1 regulates dorso-ventral patterning through a BMP-dependent mechanism. Sincerescue of dorsal structures in βcat-morphant embryos is a very robustassay, we decided to perform experiments in this sensitized back-ground. To test for the relationship between xSyn1 and BMP signaling,we analyzed the effect of BMP pathway activation in the ability ofxSyn1 to rescue dorsal structures. βcat-morphant embryos were co-injected at the four-cell stage with xSyn1 mRNA together withincreasing amounts of BMP4 or Smad1 mRNA, and the presence ofdorsal structures were evaluated at stage 25. We noted that BMP4 and

signaling counteracts rescue of dorsal axes by xSyn1. βcat-morphant embryos wereg) together with (A) BMP4 (serial dilution from 10–640 pg) or (B) Smad1 (37–150 pg),presents the average of three and two independent experiments, respectively. (C) xSyn1ge were microinjected one (1×, 40 pg of total mRNA) or four times (4×, 80 pg of totalhorylated Smad1/5/8 (P-Smad1/5/8) analyzed by western blot. As a negative control,Smad1 phosphorylation inwild type (compare lane 2 with 6 and 7) and βcat-morphant. (D) xSyn1 inhibits BMP reporter activity. Embryos were microinjected at the four-cellSyn1 (30–60 pg) or Chordin (48 pg) mRNA. Extracts were obtained at stage 10.5 and

343G.H. Olivares et al. / Developmental Biology 329 (2009) 338–349

Smad1 mRNAs were able to counteract the rescue of dorsal axes byxSyn1 in a dose-dependent manner (Figs. 3A, B).

In order to have a more direct measure of the effect of xSyn1 in theBMP pathway, we evaluated the levels of Smad1 phosphorylation (P-Smad1). Wild type and βcat-morphant embryos were microinjectedat the four-cell stage, one or four timeswith xSyn1mRNA, harvested atgastrula stage and P-Smad1 levels evaluated. As expected, knockdownof β-catenin resulted in increased levels of Smad1 phosphorylation(Fig. 3C, compare lanes 2 and 3; Reversade et al., 2005), which werereduced by overexpression of xSyn1 (Fig. 3C, compare lanes 3, 4 and5). xSyn1 mRNAwas also able to reduce the levels of P-Smad1 in wildtype embryos (Fig. 3C, compare lanes 2, 6 and 7).

We also tested the effect of xSyn1 on the Msx-luciferase reporterthat responds to the BMP signaling pathway (Sirard et al., 2000). Four-cell stage embryos were co-injected four times into the equator withDNA for Msx-luciferase and xSyn1 mRNA, raised until stage 10.5 andluciferase levels measured. In agreement with the P-Smad1 analysis,we found that xSyn1 can reduce the levels of luciferase activity (Fig.3D), similar to the effect of chordin, a known BMP antagonist (Piccoloet al., 1996). Furthermore, the ability of xSyn1 to induce chordinexpression (Figs. 2H–K, O and see Supplementary information fig. S3)can also be considered a read-out of BMP inhibition. It has beendescribed that BMP4 signaling inhibits chordin expression in zebrafishand Xenopus embryos (Delaune et al., 2005; Maegawa et al., 2006),which we have confirmed by microinjection of morpholinos againstBMP4/7 in βcat-morphant embryos (see Supplementary informationfig. S4). Similarly, overexpression of xSyn1 increases chordin expres-

Fig. 4. xSyndecan-1 inhibits BMP signaling in a Chordin-independent manner. (A) xSyn1 reqthe four-cell stage with xSyn1 (25 and 100 pg) together with morpholino control (MOCo) or Cof embryos with dorsal structures. Graph summarized the average of two independent expsignaling in a Chordin-independent manner. Embryos were co-injected in both blastomeres iand xSyn1mRNA (15–60 pg). Ectodermal explants were excised at stage 9, developed until silevels in ectodermal explants were analyzed by RT-PCR. WE: whole embryo. (C) Effect of Chorco-injected four times at the four-cell stage with xSyn1 together with morpholino control (Mblot. (D) Effect of siSyn1 on BMP signaling in C2C12 myoblasts. C2C12 were co-transfected w20 h and luciferase activity determined. The graph shows the average of three independent

sion (see Supplementary information fig. S4) suggesting that rescue ofchordin expression can be a consequence of the ability of xSyn1 toblock BMP signaling. In summary, gain-of-function experimentsshowed that xSyn1 inhibits BMP signaling and functions upstreamto, or in-parallel with, BMP4 and Smad1.

The findings that overexpression of xSyn1 rescued dorsal struc-tures and blocked BMP signaling raised the question about whetherthese were direct or indirect effects of xSyn1. This is especiallyrelevant in view of the ability of xSyn1 to induce expression of chordin,a secreted protein known for its ability to rescue ventralized embryosthrough BMP antagonism (Sasai et al., 1994; Piccolo et al., 1996). Totest if the effect of xSyn1 in dorso-ventral patterning and BMPsignaling inhibition are dependent on chordin, we performedexperiments in the presence of chordin morpholinos (Oelgeschlageret al., 2003).

First, we tested the importance of chordin for the rescue of dorsalaxes by xSyn1 mRNA. We found that xSyn1 was no longer able torestore dorsal structures in chordin morphant embryos (Fig. 4A). Thiswas specific for chordin because morpholinos against noggin (Kurodaet al., 2004) had no effect (see Supplementary information fig. S5) inagreement with the observation that noggin expression was notrescued by xSyn1 (Fig. 2O). One possible explanation for therequirement of chordin is that xSyn1 can make a ternary complexwith chordin and cooperate in the binding of BMP, similarly to the roleof twisted gastrulation Tsg (Oelgeschlager et al., 2000). In favor of thishypothesis, chordin binds to HSPG on the cell surface, and depletion ofsulfated HS chains results in a reduction of its anti-BMP activity (Jasuja

uires Chordin to rescue dorsal structures. βcat-morphant embryos were co-injected athordin (MO-Chd). Rescue of dorsal axes was determined at stage 25 and expressed as %eriments. Green bar corresponds to βcat-morphant embryos. (B) xSyn1 inhibits BMPn the animal region at the two-cell stage with a BMP luciferase reporter (BRE-luc, 30 pg)bling embryos reached stage 11 and luciferase activity determined. Inset: chordinmRNAdin knockdown in the ability of xSyn1 to inhibit P-Smad1. βcat morphant embryos wereOCo) or Chordin (MO-Chd), raised until stage 10.5 and P-Smad1 evaluated by western

ith siCo or siSyn1 together with BRE-luc reporter, incubated with recombinant BMP4 forexperiments.

344 G.H. Olivares et al. / Developmental Biology 329 (2009) 338–349

et al., 2004). The formation of ternary complexes should result instrong cooperative effect as has been described for Tsg and chordin/sog (Oelgeschlager et al., 2000; Larraín et al., 2001; Scott et al., 2001).To test for cooperativity, we co-injected xSyn1 mRNA and chordinmRNA or protein and we evaluated dorsal phenotypes and molecularmarkers. For all the conditions tested, no cooperativity between thesetwo molecules has been observed (see Supplementary informationfig. S6 and data not shown). We concluded from these experimentsthat xSyn1 requires chordin to rescue dorsal structures but nocooperativity between these two molecules was demonstrated.

We then decided to evaluate the role of chordin in the ability ofxSyn1 to block BMP signaling. First, we took advantage of the factthat chordin is not expressed in the animal cap to test the effect ofxSyn1 in activation of a BMP reporter. In this case, we used the BRE-luciferase reporter, a synthetic promoter construct that is morespecific for BMP signaling than the Msx reporter (Korchynskyi andten Dijke, 2002). Embryos at the two-cell stage were injected twotimes in the animal pole with the BRE reporter together with xSyn1mRNA; animal caps were isolated at the blastula stage and luciferaselevels measured when sibling embryos arrived to stage 10.5. Wefound that xSyn1 efficiently reduces the levels of luciferase in theabsence of chordin expression (Fig. 4B). In addition, the effect ofchordin knockdown on the ability of xSyn1 to block P-Smad1 wastested. To evaluate this, βcat-morphant embryos were microinjectedat the four-cell stage with xSyn1 mRNA together with control orchordin morpholinos and the levels of P-Smad1 were determined atthe early gastrula stage. xSyn1 was still able to reduce the levels ofP-Smad1 in chordin morphant embryos (Fig. 4C, compare lanes 2and 4), though with less efficiency when compared to embryos notinjected with chordin morpholinos (Fig. 4C, compare lanes 3 and 4).These results indicate that xSyn1 can inhibit BMP activity in theabsence of chordin.

To further probe the chordin-independent effect of xSyn1, weperformed experiments in a cell culture assay. C2C12 cells weretransfected with the BMP reporters (Msx and BRE-luciferase) togetherwith siRNA against mouse syndecan-1 (McQuade et al., 2006). After2 days of transfection, cells were incubated with pure recombinantBMP4 for 20 h and the levels of mSyn1 and luciferase weredetermined. We found that Syn1 was reduced in siSyn1 transfectedcells and an increase in the BMP response was observed (Fig. 4D andsee Supplementary information fig. S7). These results suggested thatin a different context than the embryo, Syn1 can exert an inhibitoryeffect on BMP4, giving further support to the ability of Syn1 to inhibitBMP signaling.

In summary, from the aforementioned results we were able toconclude that xSyn1 requires chordin to restore dorso-ventralpatterning, but clearly demonstrated that xSyn1 inhibits BMPsignaling in a chordin-independent manner.

xSyndecan1 is required for proper BMP signaling in Xenopus embryos

The gain-of-function experiments demonstrated that xSyn1 mRNAis sufficient to inhibit BMP signaling but provide no evidence aboutthe necessity of xSyn1 for BMP signaling. To evaluate this, weperformed loss-of-function studies using morpholinos. Two morpho-linos complementary to the ATG region were designed, synthesizedand tested.

First, we evaluated their effect on endogenous xSyn1 proteinlevels. For this, we used an antibody that recognizes the neo-epitopegenerated in the core protein of each HSPG after removal of the HSchains with heparitinase (anti-ΔHS). Importantly, this antibodyrecognizes all the HS modified core proteins present in the sample.Because xSyn1 mRNA is known to be expressed in the animal cap(Teel and Yost, 1996), both morpholinos were microinjected in theanimal pole in each blastomere right after appearance of thecleavage furrow. At stage 8, animal caps were isolated, cultured until

stage 11 and endogenous HSPG detected using the anti-ΔHSantibody. In two independent experiments, we have found thatthe morpholinos against xSyn1 specifically reduced the levels of twobands; one that migrated above the 200 kDa molecular weightmarker and the other one around 150 kDa (Fig. 5A, see arrow andarrowhead, respectively). Importantly, no effect on the other HSPGcore proteins was observed (Fig. 5A, see bracket). Treatment withmorpholinos against HSPG results in specific reduction of differentcore proteins when blots with anti-ΔHS antibody were performed.For example, injection of morpholinos against xSyn4 specificallyreduced the band of approximately 34 kDa (Muñoz et al., 2006)confirming the specificity of this experimental approach. Of note,the open reading frame of xSyn1 is almost twice the size of itsmouse orthologues, suggesting that it should run in SDS-PAGE withan apparent MW of approximately 150 kDa (mSyn1 runs in SDS-PAGE with an apparent MW of 75 kDa). This is in agreement withthe MW of the fast migrating band reduced by xSyn1 morpholinoknockdown (Fig. 5A, arrowhead). The slow migrating band reducedby xSyn1 morpholinos (Fig. 5A, arrow) can be explained because ofthe tendency of syn1 to form oligomers and run with apparentlyhigher molecular weights, something that has been reported formSyn1 (Ma et al., 2006).

Considering that overexpression of xSyn1 blocks BMP signalingand that xSyn1 and BMP4 are expressed in the ectoderm, wedecided to test the effect of xSyn1 morpholinos on the activation ofthe BRE-luciferase reporter in animal cap assays. Embryos wereinjected at the early two-cell stage with morpholinos against xSyn1together with BRE-luciferase, animal caps explanted, cultured untilstage 10.5 and luciferase activity was measured. We found thateach of the two different xSyn1-MO were able to reduce the levelsof luciferase (see Supplementary information fig. S8). Inhibition ofBRE activity was rescued by injection of xSyn1 mRNA (Fig. 5B),indicating that the effect of xSyn1 knockdown was specific.Moreover, knockdown of xSyn4, a member of the syndecan familyinvolved in gastrulation and non-canonical Wnt signaling (Muñozet al., 2006), was not able to reduce the levels of BRE-luciferase(Fig. 5C). In summary, xSyn1 knockdown is very specific andreliable (Eisen and Smith, 2008) because (1) endogenous levels ofxSyn1, but no other HSPG, were reduced; (2) two differentmorpholinos against xSyn1 gave the same effect; (3) BRE inhibitionwas rescued by xSyn1 mRNA, and (4) no inhibition of BRE activitywas observed when xSyn4, another member of the syndecanfamily, was knocked-down.

The reduction of the levels of BRE-luciferase in xSyn1 morphantembryos suggests that xSyn1 is required for proper activation of theBMP response. To determine the relationship between xSyn1 andother components of the BMP signaling we performed epistaticexperiments with Smad1 and BMP4. We found that Smad1 mRNAefficiently restores the levels of BRE-luciferase in animal caps whenxSyn1 was knocked-down (Fig. 5D), indicating that xSyn1 functionsupstream of intracellular components of the BMP signaling cascadegiving further confirmation to our findings shown in Fig. 3B. Then, wetested the effect of xSyn1 morpholinos on the ability of BMP4 toinduce the BRE reporter. In animal cap assays, injection of 10–20 pg ofBMP4 mRNA resulted in 6–7 fold induction of the BMP reporter, afterxSyn1 knockdown only a three fold induction was observed (Fig. 5E).This result indicated that BMP4 requires xSyn1 to signal, sugges-ting that Syn1 functions downstream BMP, perhaps as a cell-surfaceco-receptor.

Gain- and loss-of-function experiments for xSyn1 in Xenopusembryos produced an apparent paradox. Both, overexpression andknockdown resulted in reduced BMP signaling suggesting thatxSyn1 can have a biphasic effect depending on its concentration.Enhancement of growth factor signaling by overexpression of HSPGis more clearly visualized when tested in cells that express lowlevels of proteoglycans (Steinfeld et al., 1996). To directly

Fig. 5. xSyndecan1 is necessary for BMP signaling in Xenopus embryos. (A) Effect of xSyn1MO in endogenous protein levels. Embryos at the two-cell stage were microinjected in theanimal pole in both blastomeres with a mix of two morpholinos against xSyn1 (MOSyn1-1,2). Animal caps explants were excised at stage 9 and cultured until siblings reached stage10.5, homogenized, treated with heparitinase and HSPGs detected by western blot using α-ΔHS. A short (left) and long (right) exposure of the same blot are shown. Note thatMOSyn1(1,2) reduced the amount of two bands (arrow and arrowhead) that run similarly to xSyn1 expected molecular weight but do not interfere with other HSPG core proteins(bracket). Lanes 1,2 and 3,4 correspond to two independent experiments. Left panel: short exposure, right panel: long exposure. (B) Rescue of xSyn1 knockdown. Embryos were co-injected at the two-cell stage in both blastomeres with the BMP reporter (BRE-luc, 60 pg) together with MOSyn1(1,2) and xSyn1mRNA. Animal caps were excised at stage 8, cultureduntil stage 10.5 and luciferase activity measured. (C) Inhibition of BMP signaling is specific for xSyn1 knockdown. Embryos were injected with morpholinos against xSyn1 or xSyn4(MOxSyn4) together with BRE-luciferase and then analyzed as above. (D) Smad1 restores BMP signaling when xSyn1 is knockdown. (E) xSyn1 is required formaximal BMP signaling.(F) xSyn1 regulates BMP signaling in a biphasic manner. BMP activity was evaluated at different doses of xSyn1 mRNA (60 to 4 pg) in xSyn1 morphants embryos.

345G.H. Olivares et al. / Developmental Biology 329 (2009) 338–349

demonstrate a concentration dependent effect of xSyn1 on BMPsignaling we injected different amounts of xSyn1 mRNA in xSyn1morphant embryos, isolated animal caps and measured BREluciferase levels. We found that at high levels xSyn1 mRNA inhibitsBMP signaling (Fig. 5F), as shown before in wild type embryos (Figs.3C, D and 4B). Importantly, when lower levels of the mRNA wereinjected an increase (4-fold) in BMP signaling was observed (Fig.5F). This result indicates that xSyn1 can have a dual effect on BMPsignaling, at high concentrations it inhibits and at low levels itenhances BMP signaling.

xSyndecan1 is necessary for proper patterning of the ectoderm

During late gastrula and neurula stage, xSyn1 is expressed in thenon-neural ectoderm and is completely absent from the neural plate(Teel and Yost, 1996), similar to the expression of BMP4 (Fainsod et al.,1994). This expression pattern suggests a function for xSyn1 in dorso-ventral patterning of the ectoderm. To study the possible role of xSyn1in this process, xSyn1 morphant embryos were fixed at neurula stageand patterning of the neural plate and non-neural ectoderm studied byin situ hybridization for sox2 and cytokeratin, respectively. Reduction of

346 G.H. Olivares et al. / Developmental Biology 329 (2009) 338–349

xSyn1 resulted in the expansion of sox2 towards the ventral pole and aconcomitant reduction on the territories expressing cytokeratin (Figs.6A–F). At tailbud stages, xSyn1 morphant embryos showed defects onthe epidermis (Figs. 6G–H, see arrows). If allowed to develop further,these defects on the epidermis gets more extensive and tadpoles die(data not shown). In summary, these results indicate that xSyn1 isessential for proper dorso-ventral patterning of the ectoderm.

xSyndecan1 function is specific and independent of its GAG chains

One important question in proteoglycan biology is to understand ifdifferentHSPGhave specific functions andwhichdomains are involvedin those functions (core protein and/or GAG chains). To assess forspecificity, we compared the ability of xSyndecan4 (xSyn4), xGlypi-can4 (xGly4) and xSyn1 to rescue βcat-morphant embryos. The effectsof different HSPG were evaluated using two assays: rescue of dorsalstructures and reduction of Smad1 phosphorylation. xSyn1 over-expression was very efficient but no effect for xSyn4 and xGly4 wasobserved in these two assays (Figs. 7A and B). To discard the possibilitythat no effect with xSyn4 and xGly4was obtained because these HSPGwere not synthesized, we studied if overexpression of their mRNAsresulted in increased proteoglycan levels. For this, syntheticmRNAs foreach HSPG were microinjected in embryos at the four-cell stage, 100animal caps isolated, homogenized at stage 10.5, incubated withheparitinase and analyzed bywestern blotwith anti-ΔHS antibody.Wefound that the three HSPG were synthesized after mRNA over-expression (Fig. 7C). In conclusion, xSyn4 and xGly4, even thoughefficiently synthesized after mRNA overexpression, were unable torescue dorsal structures and to block BMP signaling, suggesting thatthese functions are specific for xSyn1 core protein or its GAG chains.

To evaluate the role of the GAG chains, we mutated the potentialGAG attachment sites present in the core protein of xSyn1. xSyn1

Fig. 6. xSyndecan1 regulates dorso-ventral patterning of the ectoderm. (A–F) xSyn1 knockdocell stage were microinjected in both blastomeres with (A, D) control and (B, E) xSyn1 morpand (D, E) cytokeratin performed. Neural plate width was measured at the half of the antero-test showing that means vary significantly with (C) pb0.01 and (F) pb0.05, respectively. ThexSyn1 is necessary for epidermal integrity. Embryos at the two-cell stage were microinjecttailbud stage. Arrows point where epidermis is affected in xSyn1 morphant embryos.

contains six putative GAG sites, five of them are also present inmouse and rat Syn1 (Zhang et al., 1995; Teel and Yost, 1996).Surprisingly, a construct with the six putative GAG attachment sitesmutated was still able to rescue dorsal structures when injected inβcat-morphant embryos (Fig. 7D) and to block BMP signaling asefficiently as wild type xSyn1 (Figs. 7E, F). These results indicate thatthe ability of xSyn1 to restore dorso-ventral patterning is specific tothis cell-surface HSPG and is independent of the integrity of its GAGattachment sites.

Discussion

We have described, for the first time, a role for Syndecan1 in BMPsignaling and dorso-ventral patterning of the ectoderm during earlydevelopment of the Xenopus embryo. Through gain- and loss-of-function experiments we have demonstrated that xSyn1 can modulateBMP signaling. Overexpression of xSyn1 mRNA resulted in BMPinhibition and rescue of dorsal structures in ventralized embryos, aneffect that is independent of the presence of its six GAG attachmentsites. Morpholino knockdown reduced BMP activity in animal capsand expanded the size of the neural plate with the concomitantreduction of the non-neural ectoderm. Experiments in xSyn1morphant embryos showed that xSyn1 can have a biphasic effectdependent on its concentration; high levels can block and low levelscan enhance BMP signaling.

Because we found that overexpression of xSyn1 mRNA induceschordin expression, it was important to study if the ability of this HSPGto rescue axes and inhibit BMP was dependent on chordin. We clearlydemonstrated that restoration of dorso-ventral patterning wascompletely dependent on chordin (Fig. 4A), but contrary to that,xSyn1 can inhibit BMP signaling in the absence of chordin (Figs. 4B, C).These seemingly opposed results can be reconciled because inhibition

wn increased neural plate wide and reduced non-neural ectoderm. Embryos at the two-holinos (34 ng), raised until neurula stage and in situ hybridization against (A, B) sox2posterior axes with ImageJ. Statistical analysis with Kruskal–Wallis test and Dunn's postnumber of embryos used for each experimental point is indicated above each bar. (G, H)ed in both blastomeres with control and xSyn1 morpholinos (34 ng) and raised until

Fig. 7. xSyndecan1 function is specific and independent of the integrity of its GAG attachment sites. (A) Rescue of dorsal structures is specific for xSyndecan1. βcat-morphant embryosreceived one single microinjection at the four-cell stage with synthetic mRNAs for xSyn1, xSyn4 and xGly4 (5–80 pg) and the presence of dorsal structures was evaluated at stage 25.Average of three independent experiments. (B) Inhibition of BMP signaling is specific for xSyndecan1. βcat morphant embryos were radially microinjected with 50 pg of xSyn1, xSyn4and xGly4 and P-Smad1 levels evaluated in whole embryos at stage 10.5. (C) Biosynthesis of HSPG in Xenopus embryos. Embryos at four-cell stage were microinjected in allblastomeres with 80 pg of xSyn1, xSyn4 and xGly4, ectodermal explants were excised at stage 9 and cultured until siblings reached stage 10.5. Total amounts of HSPGs weredetermined by western blot using α-ΔHS that specifically recognized HSPG core proteins after heparitinase treatment. In embryos microinjetcted with xSyn1 mRNA an increase inthe amount of a low migration band (arrowhead) was observed (compare lanes 1 and 2). For xSyn4 (arrow) and xGly4 (asterisks) overexpression bands of the expected size weredetected. (D) xSyndecan1 GAGmutant can rescue dorsal structures. MO-βcat embryos weremicroinjected once at the four-cell stagewith 40–200 pg of xSyn1 and xSyn1-ΔGAGDNAconstructs and the presence of dorsal structures were evaluated at stage 25. Average of three independent experiments is shown. (E, F) xSyndecan1 GAG mutant can block BMPsignaling. Embryos were co-injected at the two-cell stage in both blastomeres with the BMP reporter (BRE-luc, 60 pg) xSyn1 and xSyn1ΔGAG mRNA in wild type (E) or xSyn1morphants (F). Animal caps excised at stage 8, cultured until stage 10.5 and luciferase activity was measured.

347G.H. Olivares et al. / Developmental Biology 329 (2009) 338–349

of BMP induces chordin (see Supplementary information fig. S4,Delaune et al., 2005; Maegawa et al., 2006). The fact that xSyn1 is notable to restore dorso-ventral patterning in chordin morphantembryos, together with its expression pattern (Teel and Yost, 1996)and the xSyn1 morpholino knockdown phenotypes, demonstratedthat xSyn1 has no role in Spemann Organizer function. Although thephysiological role of xSyn1 is not to establish dorso-ventral axes, it isimportant to highlight that the rescue of ventralized embryos is still avery powerful assay to identify genes involved in BMP signaling that

could then give rise to study the more physiological roles of thesenovel molecules.

The overexpression of xSyn1 mRNA in xSyn1 morphant embryosrevealed its biphasic effect. At high concentrations xSyn1 mRNAblocks BMP activity and at low concentration it increases BMPsignaling (Fig. 5F). The fact that increased levels of BMP4 counter-act the ability of xSyn1 to restore dorsal axis (Fig. 3A) suggests thatat high concentrations xSyn1 is upstream of BMP4, perhapsfunctioning as a BMP sequester. When xSyn1 levels were reduced

348 G.H. Olivares et al. / Developmental Biology 329 (2009) 338–349

using morpholinos, the opposite relationship between xSyn1 andBMP4 was observed. We found that BMP4 is less active when co-injected with xSyn1 morpholinos (Fig. 5E). This indicates thatxSyn1 is required to attain maximal BMP signaling, suggesting aco-receptor function for this cell surface HSPG. Loss-of-functionstudies in embryos only revealed the co-receptor function of BMPand its inhibitory effect was not detected. Contrary to this,reduction of Syn1 in C2C12 cells resulted in increased levels ofBMP signaling, revealing its inhibitory role (Fig. 4D). The differentroles of Syn1 in BMP signaling in embryos and C2C12 cells could beexplained by the different levels of expression of Syn1 in eachspecific cell-type.

The finding that xSyn1 can have positive and negative effects onBMP signaling depending on its concentration is reminiscent of othersecreted molecules involved in this pathway that have a dual orbiphasic effect on growth factor activity. One example is the case ofchordin/sog and Tsg that bind BMP4/dpp and form a ternarycomplex that, depending on the presence or absence of themetalloproteinase tolloid, can inhibit or enhance BMP signaling(Oelgeschlager et al., 2000; Larraín et al., 2001; Ross et al., 2001;Shimmi et al., 2005; Wang and Ferguson, 2005). Similarly, the BMP-binding protein Cv2 can promote or block BMP signaling dependingon its concentration and the presence or absence of other BMPbinding components (O'Connor et al., 2006; Ambrosio et al., 2008;Serpe et al., 2008). Biphasic modulators such as Tsg and Cv2 can be ofuse to produce spatial bistability leading to generation of sharp step-gradients and precise borders (Larraín et al., 2001; O'Connor et al.,2006; Serpe et al., 2008). Here, we introduced Syn1 as a new playerin the extracellular regulation of BMP signaling and showed that italso has a biphasic effect.

Biochemical and genetic studies have demonstrated that BMPbinds to HS chains and heparin (Ruppert et al., 1996; Groppe et al.,1998; Ohkawara et al., 2002; Lin, 2004), indicative of a signal-regulating interaction. Our finding that xSyn1 function in Xenopusembryos is independent of its GAG attachment sites agrees withrecent evidences of GAG independent functions for many HSPG(Ohkawara et al., 2003; Lin, 2004; Kirkpatrick et al., 2006; Capurro etal., 2008;). Of note, GAG-dependency changes during development,HS chains are completely required for BMP/dpp signaling in the wingimaginal disc but not for early embryonic dorso-ventral patterning(Haerry et al., 1997; Han et al., 2004; Bornemann et al., 2008). Fromthese observations and our results, we propose that Syndecan1 canmodulate BMP signaling in a GAG independent-manner.

The defects on the epidermis of xSyn1 morphant embryos suggestthat xSyn1 is essential for ectodermal development. This could beexplained because of defective patterning of the ectoderm or could berelated to the function of Syn1 in cell adhesion and its interactionwithintegrins (Beauvais et al., 2004). This late effect of xSyn1 is ofparticular interest because gene-targeting of Syn1 gives rise to micewith normal embryonic development but defective wound healing(Gallo et al., 1996; Park et al., 2001; Gotte et al., 2002; Stepp et al.,2002). Studying the effect of xSyn1 in Xenopus could provide aninteresting model to study ectodermal differentiation and woundhealing.

Competing financial interests

The authors declare that they have no competing financialinterests.

Acknowledgments

We thank R. Doger, J. Gutiérrez, M. Henderson, M. Labarca, Dr. H.Olguín, and members of the laboratory for critical reading of themanuscript. We thank Drs. J. Heasman, R.D. Sanderson, E. De Robertisand A. Hemmati-Brivanlou for providing reagents. G.H.O. was a

CONICYT Ph.D. fellow. L.C. was CONICYT postdoctoral fellow. Thiswork was supported by Fondo Nacional de Ciencia y Tecnología(FONDECYT 1030481).

Appendix A. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.ydbio.2009.03.007.

References

Alexander, C.M., Reichsman, F., Hinkes, M.T., Lincecum, J., Becker, K.A., Cumberledge, S.,Bernfield, M., 2000. Syndecan-1 is required for Wnt-1-induced mammarytumorigenesis in mice. Nat. Genet. 25, 329–332.

Ambrosio, A.L., Taelman, V.F., Lee, H.X., Metzinger, C.A., Coffinier, C., De Robertis, E.M.,2008. Crossveinless-2 Is a BMP feedback inhibitor that binds chordin/BMP toregulate Xenopus embryonic patterning. Dev. Cell 15, 248–260.

Beauvais, D.M., Burbach, B.J., Rapraeger, A.C., 2004. The syndecan-1 ectodomainregulates alphavbeta3 integrin activity in human mammary carcinoma cells. J. CellBiol. 167, 171–181.

Beckett, K., Franch-Marro, X., Vincent, J.P., 2008. Glypican-mediated endocytosis ofhedgehog has opposite effects in flies and mice. Trends Cell Biol. 18, 360–363.

Belo, J.A., Bouwmeester, T., Leyns, L., Kertesz, N., Gallo, M., Follettie, M., De Robertis,E.M., 1997. Cerberus-like is a secreted factor with neutralizing activity expressedin the anterior primitive endoderm of the mouse gastrula. Mech. Dev. 68, 45–57.

Bier, E., 2008. Intriguing extracellular regulation of BMP signaling. Dev. Cell 15, 176–177.Birsoy, B., Berg, L., Williams, P.H., Smith, J.C., Wylie, C.C., Christian, J.L., Heasman, J., 2005.

XPACE4 is a localized pro-protein convertase required for mesoderm induction andthe cleavage of specific TGFbeta proteins in Xenopus development. Development132, 591–602.

Bishop, J.R., Schuksz, M., Esko, J.D., 2007. Heparan sulphate proteoglycans fine-tunemammalian physiology. Nature 446, 1030–1037.

Bornemann, D.J., Park, S., Phin, S., Warrior, R., 2008. A translational block to HSPGsynthesis permits BMP signaling in the early Drosophila embryo. Development 135,1039–1047.

Bulow, H.E., Hobert, O., 2006. The molecular diversity of glycosaminoglycans shapesanimal development. Annu. Rev. Cell Dev. Biol. 22, 375–407.

Capurro, M.I., Xiang, Y.Y., Lobe, C., Filmus, J., 2005. Glypican-3 promotes the growth ofhepatocellular carcinoma by stimulating canonical Wnt signaling. Cancer Res. 65,6245–6254.

Capurro, M.I., Xu, P., Shi, W., Li, F., Jia, A., Filmus, J., 2008. Glypican-3 inhibits hedgehogsignaling during development by competing with patched for hedgehog binding.Dev. Cell 14, 700–711.

Carey, D.J., 1997. Syndecans: multifunctional cell-surface co-receptors. Biochem. J. 327,1–16.

Casar, J.C., Cabello-Verrugio, C., Olguin, H., Aldunate, R., Inestrosa, N.C., Brandan, E., 2004.Heparan sulfate proteoglycans are increased during skeletal muscle regeneration:requirement of syndecan-3 for successful fiber formation. J. Cell. Sci. 117, 73–84.

Carrasco, H., Olivares, G.H., Faunes, F., Oliva, C., Larraín, J., 2005. Heparan sulfate proteoglycansexert positive and negative effects in Shh activity. J. Cell. Biochem 96, 831–838.

De Robertis, E.M., 2006. Spemann's organizer and self-regulation in amphibianembryos. Nat. Rev. Mol. Cell Biol. 4, 296–302.

De Robertis, E.M., Larraín, J., Oelgeschlager, M., Wessely, O., 2000. The establishment ofSpemann's organizer and patterning of the vertebrate embryo. Nat. Rev. Genet. 1,171–181.

Delaune, E., Lemaire, P., Kodjabachian, L., 2005. Neural induction in Xenopus requiresearly FGF signalling in addition to BMP inhibition. Development 132, 299–310.

Eisen, J.S., Smith, J.C., 2008. Controlling morpholino experiments: don't stop makingantisense. Development 135, 1735–1743.

Esko, J.D., Selleck, S.B., 2002. Order out of chaos: assembly of ligand binding sites inheparan sulfate. Annu. Rev. Biochem. 71, 435–471.

Fainsod, A., Steinbeisser, H., De Robertis, E.M., 1994. On the function of BMP-4 inpatterning the marginal zone of the Xenopus embryo. EMBO J. 13, 5015–5025.

Freeman, M., Gurdon, J.B., 2002. Regulatory principles of developmental signaling.Annu. Rev. Cell Dev. Biol. 18, 515–539.

Gallo, R., Kim, C., Kokenyesi, R., Adzick, N.S., Bernfield, M., 1996. Syndecans-1 and -4 areinduced during wound repair of neonatal but not fetal skin. J. Invest. Dermatol. 107,676–683.

Gotte, M., Joussen, A.M., Klein, C., Andre, P., Wagner, D.D., Hinkes, M.T., Kirchhof, B.,Adamis, A.P., Bernfield, M., 2002. Role of syndecan-1 in leukocyte-endothelialinteractions in the ocular vasculature. Invest. Ophthalmol. Vis. Sci. 43, 1135–1141.

Groppe, J., Rumpel, K., Economides, A.N., Stahl, N., Sebald, W., Affolter, M., 1998.Biochemical and biophysical characterization of refoldedDrosophilaDPP, a homologof bone morphogenetic proteins 2 and 4. J. Biol. Chem. 273, 29052–29065.

Gutiérrez, J., Osses, N., Brandan, E., 2006. Changes in secreted and cell associatedproteoglycan synthesis during conversion of myoblasts to osteoblasts in responseto bone morphogenetic protein-2: role of decorin in cell response to BMP-2. J. CellPhysiol. 206, 58–67.

Hacker, U., Nybakken, K., Perrimon, N., 2005. Heparan sulphate proteoglycans: thesweet side of development. Nat. Rev. Mol. Cell Biol. 6, 530–541.

Haerry, T.E., Heslip, T.R., Marsh, J.L., O'Connor, M.B., 1997. Defects in glucuronatebiosynthesis disrupt wingless signaling in Drosophila. Development 124,3055–3064.

349G.H. Olivares et al. / Developmental Biology 329 (2009) 338–349

Han, C., Belenkaya, T.Y., Khodoun, M., Tauchi, M., Lin, X., Lin, X., 2004. Distinct andcollaborative roles of Drosophila EXT family proteins in morphogen signalling andgradient formation. Development 131, 1563–1575.

Heasman, J., Kofron, M., Wylie, C., 2000. Beta-catenin signaling activity dissected in theearly Xenopus embryo: a novel antisense approach. Dev. Biol. 222, 124–134.

Jasuja, R., Allen, B.L., Pappano, W.N., Rapraeger, A.C., Greenspan, D.S., 2004. Cell-surfaceheparan sulfate proteoglycans potentiate chordin antagonism of bone morphoge-netic protein signaling and are necessary for cellular uptake of chordin. J. Biol.Chem. 279, 51289–51297.

Johnson, K.G., Ghose, A., Epstein, E., Lincecum, J., O'Connor, M.B., Van Vactor, D., 2004.Axonal heparan sulfate proteoglycans regulate the distribution and efficiency of therepellent slit during midline axon guidance. Curr. Biol. 14, 499–504.

Kao, K.R., Elinson, R.P., 1988. The entire mesodermal mantle behaves as Spemann'sorganizer in dorsoanterior enhanced Xenopus laevis embryos. Dev. Biol. 127, 64–77.

Kirkpatrick, C.A., Knox, S.M., Staatz, W.D., Fox, B., Lercher, D.M., Selleck, S.B., 2006. Thefunction of a Drosophila glypican does not depend entirely on heparan sulfatemodification. Dev. Biol. 300, 570–582.

Korchynskyi, O., ten Dijke, P., 2002. Identification and functional characterization ofdistinct critically important bone morphogenetic protein-specific response ele-ments in the Id1 promoter. J. Biol. Chem. 277, 4883–4891.

Kramer, K.L., Yost, H.J., 2002. Ectodermal syndecan-2 mediates left-right axis formationin migratingmesoderm as a cell-nonautonomous Vg1 cofactor. Dev. Cell 2, 115–124.

Kuroda, H., Wessely, O., De Robertis, E.M., 2004. Neural induction in Xenopus:requirement for ectodermal and endomesodermal signals via chordin, noggin,beta-catenin, and cerberus. PLoS Biol. 2, 632–634.

Lander, A.D., Nie, Q., Wan, F.Y., 2002. Do morphogen gradients arise by diffusion?Dev. Cell 2, 785–796.

Larraín, J., Carey, D.J., Brandan, E., 1998. Syndecan-1 expression inhibits myoblastdifferentiation through a basic fibroblast growth factor-dependent mechanism.J. Biol. Chem. 273, 32288–32296.

Larraín, J., Bachiller, D., Lu, B., Agius, E., Piccolo, S., De Robertis, E.M., 2000. BMP-bindingmodules in chordin: a model for signalling regulation in the extracellular space.Development 127, 821–830.

Larraín, J., Oelgeschlager, M., Ketpura, N.I., Reversade, B., Zakin, L., De Robertis, E.M.,2001. Proteolytic cleavage of chordin as a switch for the dual activities of twistedgastrulation in BMP signaling. Development 128, 4439–4447.

Lemaire, P., Gurdon, J.B., 1994. A role for cytoplasmic determinants in mesodermpatterning: cell-autonomous activation of the goosecoid and Xwnt-8 genes alongthe dorsoventral axis of early Xenopus embryos. Development 120, 1191–1199.

Lin, X., 2004. Functions of heparan sulfate proteoglycans in cell signaling duringdevelopment. Development 131, 6009–6021.

Ma, P., Beck, S.L., Raab, R.W., McKown, R.L., Coffman, G.L., Utani, A., Chirico, W.J.,Rapraeger, A.C., Laurie, G.W., 2006. Heparanase deglycanation of syndecan-1 isrequired for binding of the epithelial-restricted prosecretory mitogen lacritin. J. CellBiol. 174, 1097–1106.

Maegawa, S., Varga, M., Weinberg, E.S., 2006. FGF signaling is required for beta-catenin-mediated induction of the zebrafish organizer. Development 133, 3265–3276.

Mali, M., Elenius, K., Miettinen, H.M., Jalkanen, M., 1993. Inhibition of basic fibroblastgrowth factor-induced growth promotion by overexpression of syndecan-1. J. Biol.Chem. 268, 24215–24222.

McFall, A.J., Rapraeger, A.C., 1998. Characterization of the high affinity cell-bindingdomain in the cell surface proteoglycan syndecan-4. J. Biol. Chem. 273,28270–28276.

McQuade, K.J., Beauvais, D.M., Burbach, B.J., Rapraeger, A.C., 2006. Syndecan-1 regulatesalphavbeta5 integrin activity in B82L fibroblasts. J. Cell Sci. 119, 2445–2456.

Moreno, M., Muñoz, R., Aroca, F., Labarca, M., Brandan, E., Larraín, J., 2005. Biglycan is anew extracellular component of the chordin-BMP4 signaling pathway. EMBO J. 24,1397–1405.

Muñoz, R., Moreno, M., Oliva, C., Orbenes, C., Larraín, J., 2006. Syndecan-4 regulatesnon-canonical Wnt signalling and is essential for convergent and extensionmovements in Xenopus embryos. Nat. Cell Biol. 8, 492–500.

Nieuwkoop, P.D., Faber, J., 1967. Normal Table of Xenopus laevis (Daudin). North Holland,Amsterdam.

O'Connor, M.B., Umulis, D., Othmer, H.G., Blair, S.S., 2006. Shaping BMP morphogengradients in the Drosophila embryo and pupal wing. Development 133, 183–193.

Oelgeschlager, M., Larraín, J., Geissert, D., De Robertis, E.M., 2000. The evolutionarilyconserved BMP-binding protein twisted gastrulation promotes BMP signalling.Nature 405, 757–763.

Oelgeschlager, M., Kuroda, H., Reversade, B., De Robertis, E.M., 2003. Chordin is requiredfor the Spemann organizer transplantation phenomenon in Xenopus embryos. Dev.Cell 4, 219–230.

Ohkawara, B., Iemura, S., ten Dijke, P., Ueno, N., 2002. Action range of BMP is defined byits N-terminal basic amino acid core. Curr. Biol. 12, 205–209.

Ohkawara, B., Yamamoto, T.S., Tada, M., Ueno, N., 2003. Role of glypican 4 in theregulation of convergent extension movements during gastrulation in Xenopuslaevis. Development 130, 2129–2138.

Park, P.W., Pier, G.B., Hinkes, M.T., Bernfield, M., 2001. Exploitation of syndecan-1shedding by Pseudomonas aeruginosa enhances virulence. Nature 411, 98–102.

Park, H., Kim, Y., Lim, Y., Han, I., Oh, E.S., 2002. Syndecan-2 mediates adhesion andproliferation of colon carcinoma cells. J. Biol. Chem. 277, 29730–29736.

Piccolo, S., Sasai, Y., Lu, B., De Robertis, E.M., 1996. Dorsoventral patterning in Xenopus:inhibition of ventral signals by direct binding of chordin to BMP-4. Cell 86,589–598.

Reversade, B., Kuroda, H., Lee, H., Mays, A., De Robertis, E.M., 2005. Depletion of Bmp2,Bmp4, Bmp7 and Spemann organizer signals induces massive brain formation inXenopus embryos. Development 132, 3381–3392.

Rhiner, C., Gysi, S., Frohli, E., Hengartner, M.O., Hajnal, A., 2005. Syndecan regulates cellmigration and axon guidance in C. elegans. Development 132, 4621–4633.

Ross, J.J., Shimmi, O., Vilmos, P., Petryk, A., Kim, H., Gaudenz, K., Hermanson, S., Ekker, S.C., O'Connor, M.B., Marsh, J.L., 2001. Twisted gastrulation is a conservedextracellular BMP antagonist. Nature 410, 479–483.

Ruppert, R., Hoffmann, E., Sebald, W., 1996. Human bone morphogenetic protein 2contains a heparin-binding site which modifies its biological activity. Eur. J.Biochem. 237, 295–302.

Sasai, Y., Lu, B., Steinbeisser, H., Geissert, D., Gont, L.K., De Robertis, E.M., 1994. Xenopuschordin: a novel dorsalizing factor activated by organizer-specific homeobox genes.Cell 79, 779–790.

Scott, I.C., Blitz, I.L., Pappano, W.N., Maas, S.A., Cho, K.W., Greenspan, D.S., 2001.Homologues of twisted gastrulation are extracellular cofactors in antagonism ofBMP signalling. Nature 410, 475–478.

Serpe, M., Umulis, D., Ralston, A., Chen, J., Olson, D.J., Avanesov, A., Othmer, H., O'Connor,M.B., Blair, S.S., 2008. The BMP-binding protein Crossveinless 2 is a short-range,concentration-dependent, biphasic modulator of BMP signaling in Drosophila. Dev.Cell 14, 940–953.

Shimmi, O., Umulis, D., Othmer, H., O'Connor, M.B., 2005. Facilitated transport of a Dpp/Scw heterodimer by Sog/Tsg leads to robust patterning of the Drosophilablastoderm embryo. Cell 120, 873–886.

Sirard, C., Kim, S., Mirtsos, C., Tadich, P., Hoodless, P.A., Itie, A., Maxson, R., Wrana, J.L.,Mak, T.W., 2000. Targeted disruption in murine cells reveals variable requirementfor Smad4 in transforming growth factor beta-related signaling. J. Biol. Chem. 275,2063–2070.

Sive, H.L., Grainger, R.M., Harland, R.M., 2000. Early Development of Xenopus laevis: ALaboratory Manual. Cold Spring Harbor Laboratory Press, New York.

Smith, W.C., Harland, R., 1992. Expression cloning of noggin, a new dorsalizing factorlocalized to the Spemann organizer in Xenopus embryos. Cell 70, 829–840.

Song, H.H., Shi, W., Xiang, Y.Y., Filmus, J., 2005. The loss of glypican-3 induces alterationsin Wnt signaling. J. Biol. Chem. 280, 2116–2125.

Steigemann, P., Molitor, A., Fellert, S., Jackle, H., Vorbruggen, G., 2004. Heparan sulfateproteoglycan syndecan promotes axonal and myotube guidance by slit/robosignaling. Curr. Biol. 14, 225–230.

Steinfeld, R., Van Den Berghe, H., David, G., 1996. Stimulation of fibroblast growth factorreceptor-1 occupancy and signaling by cell surface-associated syndecans andglypican. J. Cell Biol. 133, 405–416.

Stepp, M.A., Gibson, H.E., Gala, P.H., Iglesia, D.D., Pajoohesh-Ganji, A., Pal-Ghosh, S.,Brown, M., Aquino, C., Schwartz, A.M., Goldberger, O., Hinkes, M.T., Bernfield, M.,2002. Defects in keratinocyte activation during wound healing in the syndecan-1-deficient mouse. J. Cell Sci. 115, 4517–4531.

Tabata, T., Takei, Y., 2004. Morphogens, their identification and regulation. Development131, 703–712.

Tao, Q., Yokota, C., Puck, H., Kofron, M., Birsoy, B., Yan, D., Asashima, M., Wylie, C.C., Lin,X., Heasman, J., 2005. Maternal wnt11 activates the canonical wnt signalingpathway required for axis formation in Xenopus embryos. Cell 120, 857–871.

Teel, A.L., Yost, H.J., 1996. Embryonic expression patterns of Xenopus syndecans. Mech.Dev. 59, 115–127.

Topczewski, J., Sepich, D.S., Myers, D.C., Walker, C., Amores, A., Lele, Z., Hammerschmidt,M., Postlethwait, J., Solnica-Krezel, L., 2001. The zebrafish glypican knypek controlscell polarity during gastrulation movements of convergent extension. Dev. Cell 1,251–264.

Wang, Y.C., Ferguson, E.L., 2005. Spatial bistability of Dpp-receptor interactions duringDrosophila dorsal-ventral patterning. Nature 434, 229–234.

Zhang, L., David, G., Esko, J.D., 1995. Repetitive Ser-Gly sequences enhance heparansulfate assembly in proteoglycans. J. Biol. Chem. 270, 27127–27135.