Mechanisms of Development 77 (1998) 173–184
DAN is a secreted glycoprotein related to Xenopus cerberus E. Stanley a,b,1, C. Biben c,1, S. Kotecha d, L. Fabri e, S. Tajbakhsh f, C-C. Wang c, T. Hatzistavrou a, B. Roberts a , b, C. Drinkwater e, M. Lah e, M. Buckingham f, D. Hilton a,b, A. Nash e, T. Mohun d, R.P. Harvey c , g ,* a The Walter and Eliza Hall Institute of Medical Research, Post Office, Royal Melbourne Hospital, Parkville 3050, Australia The Cooperative Research Centre for Cellular Growth Factors, Post Office, Royal Melbourne Hospital, Parkville 3050, Australia c The Victor Chang Cardiac Research Institute, St. Vincent’s Hospital, 384 Victoria St, Darlinghurst 2010, Australia d National Institute for Medical Research, The Ridgeway, Mill Hill, London NW7 1AA, UK e Amrad Operations Pty Ltd., 576 Swan Street, Richmond 3121, Australia f Institut Pasteur, De´partement de Biologie Mole´culaire, 25 rue du Dr Roux, 75724 Paris, Cedex 15, France g The University of New South Wales, Kensington NSW 2033, Australia
b
Received 6 July 1998; revised version received 27 July 1998; accepted 28 July 1998
Abstract We report that DAN, a potential cell cycle regulator and tumour suppressor, is a secreted glycoprotein related to Xenopus cerberus. DAN, cerberus, its mouse relative Cer-1/cer-l/Cerberus-like/Cerr1, and the recently described factor DRM/Gremlin, appear to be members of the cystine knot superfamily, which includes TGFbs and BMPs. Like cerberus and mCer-1, DAN induced cement glands as well as markers of anterior neural tissue and endoderm in Xenopus animal cap assays, features of BMP signalling blockade. During mouse embryogenesis, Dan was expressed from E8.5 in cranial mesenchyme and somites, then later in limb and facial mesenchyme. The pattern in somites was highly dynamic, with transcripts initially localized to the caudal half of the nascent epithelial somite, then, after maturation, to sclerotomal cells adjacent to the neural tube. Dan was also expressed in the developing myotome. The expression domains include sites in which BMP inhibition is known to be important for development. Thus, DAN appears to be a secreted factor belonging to the cystine knot superfamily, and one of a growing number of antagonists acting to modulate BMP signalling during development. 1998 Elsevier Science Ireland Ltd. All rights reserved Keywords: Xenopus; Dan; Cystine knot
1. Introduction The rat Dan gene (also called NO3) was first identified as a mRNA down-regulated in v-src-transformed 3Y1 fibroblasts (Ozaki and Sakiyama, 1993) and has properties of a negative modulator of cell growth. Dan expression is downregulated in fibroblasts transformed with a variety of oncogenes (Ozaki and Sakiyama, 1993; Ozaki et al., 1996), whilst enforced expression of Dan in v-src-transformed cells appears to suppress the transformed phenotype
* Corresponding author. Tel.: +61 2 92958520; fax: +61 2 92958528; e-mail:
[email protected] 1 These authors contributed equally to this work.
0925-4773/98/$ - see front matter PII S0925-4773 (98 )0 0139-7
(Ozaki and Sakiyama, 1994; Sakiyama et al., 1994). Overexpression in non-transformed cells leads to a slower growth rate and delayed entry into S-phase (Ozaki et al., 1995). In addition to effects in cultured cells, a role for human DAN as a tumour suppressor protein has been suggested (Enomoto et al., 1994). DAN maps to the short arm of human chromosome 1, close to a region which is commonly deleted in neuroblastoma tumour samples. Moreover, DAN is grossly rearranged in a subset of tumours and its expression is variable and sometimes absent in neuroblastoma cell lines. However, its status as a tumour suppressor is still contentious, since recent data place DAN outside of the minimal region on chromosome 1 deleted in tumours (White et al., 1995).
1998 Elsevier Science Ireland Ltd. All rights reserved
174
E. Stanley et al. / Mechanisms of Development 77 (1998) 173–184
The mechanism by which DAN exerts its influence is currently unknown. The DAN protein contains ten cysteines, a proline-rich domain, two potential casein kinase II target sites and C-terminal (HX) repeats which bind metal ions (Ozaki and Sakiyama, 1993; Kondo et al., 1995). The potential for some cysteines to configure a C4 class zinc finger led to the suggestion that DAN is a transcription factor (Ozaki and Sakiyama, 1993). However, DAN shows some homology to the Xenopus laevis protein cerberus, a secreted cytokine expressed in the Spemann organizer region of the Xenopus gastrula (Bouwmeester et al., 1996), and to DRM (down-regulated in mos-transformed cells), a secreted protein identified, like DAN, because of its down-regulation in viral oncogene-transformed cells (Topol et al., 1997; Hsu et al., 1998). Furthermore, a mild growth inhibitory activity similar to that originally attributed to DAN was recently demonstrated in conditioned media from Dan-transfected cells (Sakiyama and Ozaki, 1997). These findings suggest that DAN might be secreted and exerts its effects as a cytokine. Xenopus cerberus is thought to function in specification of anterior structures during the gastrulation period of frog development, and when mis-expressed in embryos induces ectopic head-like structures which sometimes contain highly organized brain, cement gland, placodes, single eye, heart and liver (Bouwmeester et al., 1996). Formation of head structures in Xenopus appears to require simultaneous blockade of BMP and Wnt-stimulated signalling pathways (Glinka et al., 1997, 1998) and in embryo assays, cerberus can antagonize both of these processes (Bouwmeester et al., 1996; Glinka et al., 1997; Biben et al., 1998). A murine cerberus-like gene, mCer-1/cer-l/Cerberus-like/ Cerr1, expressed in embryos in a pattern remarkably similar to that of the frog gene, has recently been isolated (Belo et al., 1997; Thomas et al., 1997; Biben et al., 1998; Shawlot et al., 1998). Protein alignments suggest that cerberus and mCer-1 contain a cystine knot motif (Biben et al., 1998), the structural backbone of nerve growth factor (NGF), the transforming growth factor b family (TGFbs), plateletderived growth factor (PDGF), as well as potentially a variety of other secreted cytokines, structural proteins, inhibitors and toxins (Isaacs, 1995). In this paper we demonstrate directly that DAN is a secreted protein, supporting recent experiments by Harland and colleagues (Hsu et al., 1998). Protein alignments suggest that DAN, as well as DRM, are members of the cystine knot superfamily and like other members, DAN is secreted as a covalently-linked dimer. Furthermore, DAN has inductive properties similar to that of cerberus and mCer-1 in Xenopus animal cap assays, although may be a less potent inhibitor of BMP4. We have analyzed the expression of Dan using in situ hybridization and RNase protection. Dan was expressed broadly in adult tissues, but in a restricted and dynamic fashion in embryonic development, notably during somitogenesis and in craniofacial and limb mesenchyme. The pat-
terns are consistent with a role for DAN in modulating BMP signalling at select embryonic sites. Our data suggest that DAN is a signalling antagonist related to cerberus, modulating inductive processes and cell growth during embryonic and adult life.
2. Results 2.1. DAN is a putative cystine knot protein Weak homology between DAN and cerberus was previously noted (Bouwmeester et al., 1996), although its significance was not explored. We aligned the predicted amino acid sequence of mouse DAN with that of frog cerberus, mouse Cer-1 and rat DRM, and found that the homology was restricted to a cysteine-rich domain (Fig. 1). Across this region, DAN showed 27% amino acid identity with cerberus, 22% with mCer-1 and 27% with DRM, considerably less than the 56% shared by cerberus and mCer-1, and the 98% by mouse and human DAN (Enomoto et al., 1994). However, the configuration of cysteines in DAN matched that of cerberus, mCer-1, DRM and other putative cystine knot proteins such as human Norrie disease protein (NDP), mouse intestinal mucins (IMRP), human von Willebrand Factor (vWF8) and human TGFb2 (Fig. 1). The crystal structures of NGF, PDGF and TGFb2 reveal that six of their cysteines configure a knot-like topology, the so-called cystine-knot (Meitinger et al., 1993). A structural glycine present in all demonstrated and putative cystine knot proteins (Meitinger et al., 1993; starred in Fig. 1), is also conserved in DAN. DAN carries four additional cysteines in addition to those proposed to form the knot (A–D; Fig. 1), and these can be found in analogous positions in mucins and NDP, with three (A–C) being present in mCer-1, cerberus and DRM (Biben et al., 1998). The cysteine used for dimerization in TGFbs (Isaacs, 1995) is not present in DAN (see Section 3). Database screening identified additional hypothetical peptides with homology to DAN, encoded within open reading frames of expressed sequence tags (ESTs). For comparison, we have included in Fig. 1 the sequence of one such peptide, a conceptual translation of EST AA289245 from the Genbank database. Its multiple cysteines are spaced in a manner that suggests it may also be a member of the cystine knot superfamily. The protein shows 28% amino acid identity with DAN over the cysteine-rich region, with the glycine characteristic of the family conserved. DAN, cerberus, mCer-1, DRM and the EST-encoded protein appear to define a novel branch of the cystine knot superfamily and may have a related function. 2.2. DAN is secreted as a N-glycosylated homodimer All characterized members of the cystine knot superfamily are secreted proteins. DAN contains 21 hydrophobic
E. Stanley et al. / Mechanisms of Development 77 (1998) 173–184
amino acids at its N-terminus, reminiscent of a signal peptide. To test for secretion, the mouse Dan cDNA was inserted into the expression plasmid pEFBOS I FLAG (Mizushima and Nagata, 1990; Biben et al., 1998) to create pEFBOS DAN-cFLAG. DAN expressed from this vector carried a FLAG epitope at its C-terminus. DAN-cFLAG protein was detected in culture supernatants of COS and SR-3Y1 cells after transfection with expression vector (see Section 4). Under denaturing and reducing conditions, COS cell-derived DAN-cFLAG migrated predominantly as
175
a single species of 29 kDa (Fig. 2A), larger than the 19 kDa predicted from the open reading frame of the Dan cDNA (Ozaki et al., 1996). While the negatively charged FLAG epitope, known to retard migration of tagged proteins in PAGE, may account for some of this discrepancy, DANcFLAG would also be expected to be N-glycosylated if correctly secreted. Treatment of DAN-cFLAG with N-glycosidase F reduced the apparent molecular mass to 25 kDa, indicative of N-glycosylation (Fig. 2A). Glycosylation most likely occurred at asparagine 38 (N38) within the single N-
Fig. 1. Alignment of the amino acid sequence of the putative cysteine-knot region of mouse DAN with sequences from similar regions of Xenopus cerberus (xCer) (Bouwmeester et al., 1996); mCer-1 (Belo et al., 1997; Thomas et al., 1997; Biben et al., 1998); human EST AA289245; rat DRM (Topol et al., 1997); human Norrie disease protein (NDP)(Berger et al., 1992; Chen et al., 1992); human von Willebrand factor (vWF8) (Bonthron et al., 1986); mouse intestinal mucin related protein (IMRP, assembled from ESTs with the annotation ‘similar to SW:MUCL-RAT P98089’) and human TGFb (Madisen et al., 1988). The cysteines are designated according to Meitinger et al. (1993) with cysteines 1–6 forming the knot and cysteines A-D potentially involved in intramolecular or intermolecular disulphide bonds (see text). The cysteine used for dimerization in TGFß is shown in dark blue. The glycine conserved among cystine knot proteins is indicated with a star.
176
E. Stanley et al. / Mechanisms of Development 77 (1998) 173–184
gests a homodimer. Whether DAN is secreted as a homodimer in vivo remains to be established. 2.3. DAN has an anti-BMP-like activity in Xenopus animal cap assays
Fig. 2. (A) Western blotting analysis of DAN-cFLAG protein secreted from COS cells before (−) and after (+) treatment with N-glycosidase F. DAN protein was detected using the M2 anti-FLAG antibody. (B) Western blot analysis of SR-3Y1 cell-derived DAN-cFLAG separated by SDS PAGE under reducing (+) and non-reducing (−) conditions. The position and size of molecular weight standards is shown on the left of each panel. (C) Sequence of the N-terminal region of mouse DAN as predicted by the cDNA. The arrowhead indicates the N-terminal residue of SR-3Y1 cellderived DAN-cFLAG, as determined by amino acid sequence analysis. The processed hydrophobic signal peptide is underlined and the determined sequence bolded. Cysteine 1 of the cystine-knot is indicated (1) and asparagine 38, the putative site of N-glycosylation (see text), is marked with a star.
glycosylation consensus site present in the DAN sequence (NIT; amino acids 38–40; Fig. 2C). Consistent with this assignment, the identity of amino acid 38 could not be determined during N-terminal sequencing of DAN-cFLAG purified from SR-3Y1 cells using the Edman method (see Section 4), an outcome indicative of post-translational modification. N-terminal amino acid sequencing also revealed processing after the 16th N-terminal hydrophobic amino acid (Fig. 2C), presumably the result of signal peptide cleavage. In this regard, DAN appears to differ from mCer-1, from which a pro-peptide also appears to be cleaved (Biben et al., 1998). Cystine knot proteins are secreted as homo or heterodimers (Isaacs, 1995). Under denaturing and non-reducing conditions, purified DAN-cFLAG migrated with a molecular mass of approximately 65 kDa, suggesting secretion as a dimer (Fig. 2B). While secreted DAN may be heterodimeric, the unambiguous N-terminal amino acid sequence data obtained from SR-3Y1-purified material, strongly sug-
In view of the structural similarity between mouse DAN and Xenopus cerberus, we compared their activities in Xenopus animal cap assays. In this system, cerberus and its mouse homologue mCer-1 induce cement glands and markers of anterior neural tissue and endoderm in the absence of mesoderm induction (Bouwmeester et al., 1996; Belo et al., 1997; Biben et al., 1998), features of BMP or Wnt signalling blockade. Synthetic mRNAs encoding DAN and cerberus were microinjected into Xenopus embryos at the two-cell stage. Animal pole regions (animal caps) were explanted at the blastula stage and cultured until siblings reached early tadpole stages. In a proportion of explants, DAN induced the formation of discrete cement gland-like structures which produced sticky exudate, although with an efficiency only 25% of that of cerberus (data not shown). To examine in more detail the tissues induced by DAN in animal caps, we assayed the expression of a number of tissue-specific and region-specific markers using reverse transcriptase-polymerase chain reaction (RT-PCR). DAN and cerberus induced an identical profile of markers (Fig. 3). These included the pan-neural marker N-CAM (Kintner and Melton, 1987), the anterior marker Otx-2 (Pannese et al., 1995; Gammill and Sive, 1997), the cement gland-specific marker CG13 (Jamrich and Sato, 1989; Sive et al., 1989), the cardiac and foregut markers Nkx2.5 and Nkx2.3 (Tonissen et al., 1994; Evans et al., 1995) and the endoderm-enriched marker endodermin (Edd) (Sasai et al., 1996). The hindbrain marker Krox20 (Bradley et al., 1993) and markers of ventro-lateral mesoderm (T4 globin and XeHAND) were not induced (Banville and Williams, 1985; Sparrow et al., 1998). The capacity of DAN to promote the formation of dorsoanterior tissues in animal caps may be indicative of inhibition of BMP or Wnt signalling. Inhibition of BMP signalling by mis-expression of dominant-negative forms of BMP receptor chains, or with natural inhibitors chordin, noggin and follistatin, leads to induction of a suite of markers similar or identical to those induced by DAN, cerberus and mCer-1 (see Section 3; Hemmati-Brivanlou and Melton, 1997). The recently-described Wnt inhibitor, dickkopf (dkk), displays a similar activity (Glinka et al., 1997). Chordin, noggin and follistatin, as well as a distinct Wnt inhibitor, Frzb/Frzb-1 (Leyns et al., 1997; Wang et al., 1997) are known to inhibit signalling by binding directly to their cognate ligands in the extracellular space (Piccolo et al., 1996; Zimmerman et al., 1996; Fainsod et al., 1997; Leyns et al., 1997; Wang et al., 1997). To examine whether DAN was able to antagonize BMP signalling in a manner analogous to that of cerberus and mCer-1, we co-expressed DAN or cerberus with BMP4 in animal caps by co-injecting equimolar
E. Stanley et al. / Mechanisms of Development 77 (1998) 173–184
amounts of their respective mRNAs. As shown previously (Bouwmeester et al., 1996; Biben and Harvey, 1997), injection of BMP4 RNA alone led to induction of ventro-lateral mesoderm, as judged by expression of T4 globin and XeHAND (Fig. 3). With equimolar cerberus and BMP4, expression of markers induced exclusively by cerberus (Otx-2, NCAM, Nkx2.5, Nkx2.3, CG13, Edd) and those induced exclusively by BMP4 (T4 globin, XeHAND) were
177
reduced, indicative of mutual antagonism. However, when DAN and BMP4 were co-expressed, there was little if any effect on the expression of BMP-specific markers (Fig. 3, lanes 5 and 6). Nevertheless, from the reverse perspective, the DAN-induced markers (Otx-2, NCAM, Nkx2.5, Nkx2.3, CG13, Edd) were strongly reduced in the presence of BMP4, showing that BMP4 could antagonize the effects of DAN. The data suggest that DAN, like cerberus, has anti-BMP activity, although is somewhat less stable or less potent against BMP4 (see Section 3). 2.4. Expression of Dan in adult tissues Among a limited range of adult rat and human tissues analyzed previously for Dan expression, robust levels were found in the lung, with lower levels in the kidney, brain, intestine, stomach and spleen (Ozaki and Sakiyama, 1993; Enomoto et al., 1994). We analyzed Dan expression in adult mouse tissues by RNase protection, using a cyclophilin probe to control for RNA integrity (Fig. 4). Dan was expressed in all tissues analyzed, with the exception of the liver, although the levels here may be below detection with the RNase protection assay. Levels in the pancreas and salivary gland were also low. The highest levels were seen in the lung, confirming previous findings, as well as in the uterus and bladder. 2.5. Dan expression during embryogenesis
Fig. 3. DAN has an anti-BMP-like activity in Xenopus animal caps. RTPCR analysis of markers induced in animal caps at stage 28 after injection of mRNAs encoding XCer (lane 2), DAN (lane 3) and BMP4 (lane 6), or after coinjection of 1:1 molar ration of XCer:BMP4 (lane 4) or DAN:BMP4 (lane 5), compared to uninjected controls (lane 1) and whole stage 25 embryos (lane 7). CG13 is a marker of cement gland (Jamrich and Sato, 1989; Sive et al., 1989). Nkx2.3 and Nkx2.5 are expressed in cardiac progenitors and anterior pharyngeal endoderm (Tonissen et al., 1994; Evans et al., 1995). NCAM is a pan-neural marker (Kintner and Melton, 1987). Otx2 is expressed in midbrain, forebrain, placodes, cement gland and anterior mesoderm, identifying tissues with anterior character (Pannese et al., 1995; Gammill and Sive, 1997). Krox20 is expressed in rhombomeres 3 and 5 of the hindbrain (Bradley et al., 1993). Edd expression is highest in endoderm (Sasai et al., 1996).T4 globin identifies blood, a ventral mesodermal derivative (Banville and Williams, 1985). XeHAND marks cardiac and vascular smooth muscle progenitors in lateral mesoderm (Sparrow et al., 1998; #683). Equality of mRNA input in the RT-PCRs was assessed by the expression of EF-1 alpha (EF1a) (Krieg et al., 1989).
Dan expression during mouse embryogenesis was analyzed by in situ hybridization using a wholemount protocol and digoxygenin-labelled cRNA sense and antisense probes (see Section 4). No expression was detected between embryonic days 6.5 and 8.0 (E6.5-E8; data not shown). Expression first became apparent around E8.5 at the onset of organogenesis, when transcripts were seen in newly formed somites and in head mesenchyme ventral to the midbrain (Fig. 5A). At E9.5, Dan expression continued to be evident in the nascent somite region, now towards the caudal end of the embryo, and also in more mature somites (Fig. 5B). At this stage, expression in head mesenchyme encompassed both the mid and hindbrain regions. Expression was also seen in the first branchial (mandibular) arch. At E10.5, transcripts were seen throughout the somite array, in the mandibular arch and its maxillary process, and in developing forelimb buds, predominantly in their dorsoanterior aspect (Fig. 5C,D). At E12.5, Dan was expressed in somites, strongly in the facial region, and in dorsal and ventral domains within forelimb and hindlimb buds (Fig. 5E,F). 2.6. Dan expression in somites Histological sectioning was performed to examine the cellular detail of Dan expression. The pattern in somites was found to be highly dynamic. In E8.5–9.0 embryos
178
E. Stanley et al. / Mechanisms of Development 77 (1998) 173–184
Fig. 4. RNase protection analysis of Dan expression in RNA extracted from adult mouse tissues. The positions of the protected fragments representing mouse Dan (mDan) and the cyclophilin internal control are shown. Note that the protected Dan fragment runs only slightly above that of the undigested cyclophilin probe. Dan is expressed in all adult tissues with the exception of the liver. Expression in the pancreas is weak but can be detected in longer exposure.
viewed as wholemounts, Dan was expressed in the nascent somite region in two or more stripes (Fig. 5A,B). Sections of E9.5 embryos revealed that transcripts were localized predominantly to the caudal region of the most recently formed somite (Fig. 6A) and more weakly in a stripe within the presomitic mesoderm. Interestingly, this pattern appears to complement that of mCer-1, expressed in the cranial half of nascent somites (Biben et al., 1998). These distributions would appear to reflect the specification of cranial and caudal domains with somites, which in turn influence neural crest, spinal column and vertebral patterning (Biben et al., 1998; Tajbakhsh and Sporle, 1998).
In transverse sections, it was evident that Dan expression within presomitic mesoderm and nascent somites was most prominent in a zone adjacent to the neural tube (Fig. 6B). As newly-formed somites matured, expression was at first down-regulated, then progressively re-established (Fig. 5B). In somites re-initiating expression, the caudal restriction seen in nascent somites appeared to be lost, transcripts now being localized approximately centrally within somite boundaries and, as before, adjacent to the neural tube (Fig. 6C). In older somites in which sclerotomal and dermomyotomal compartments were clearly evident, Dan expression was seen predominantly in sclerotome next to the neural tube (Fig. 6D). However, this pattern changed rapidly as somites matured still further. Expression in sclerotome became progressively restricted, while new expression was established in the developing myotome. Sections through the tail of an E12.5 embryo revealed Dan transcripts located predominantly in myotome, with no apparent restriction along the antero-posterior (A/P) axis (Fig. 6E). We have also analyzed Dan expression in E11.0 embryos carrying a single Myf5-nlacZ knockin allele (Tajbakhsh et al., 1996). Although the nlacZ insertion disrupts the Myf5 gene, heterozygote animals show no apparent phenotype. Expression of the allele highlights the myotomal compartment of the developing somite. In wholemount embryos costained for Dan expression and b-galactosidase activity, the weaker Dan staining in myotome was obscured by the LacZ staining, but non-myotomal staining in a periodic pattern along the A/P axis was still evident (Fig. 7A,B). Sectioning confirmed co-expression of Dan and LacZ in myotome, as well as Dan expression in sclerotomal cells adjacent to the
Fig. 5. Expression of mDan during embryogenesis as assessed by wholemount in situ hybridization using an antisense Dan probe. Hybridizations with sense probe showed no specific signal (data not shown). (A) Lateral view of an E8.5 embryo showing early expression of Dan in cephalic mesenchyme (cm) and the newly formed somites (ns). (B) E9.5 embryo showing Dan expression in most somites. Expression is also seen within the first branchial arch (ba). (C) Dorsal view of an E10.5 embryo highlighting Dan expression in somites (sm) and the dorsoanterior aspect of the forelimb bud (lb). (D) Lateral view of the E10.5 embryo shown in C, highlighting transcription of Dan within head mesenchyme, now underlying the midbrain and hindbrain, and increased expression in the branchial region (ba) and the dorsal aspect of the forelimb bud (lb). (E) Ventral view of a decapitated E12.5 embryo showing strong expression of Dan within the dorsal and ventral aspects of the forelimb (fl) and initial expression in the hindlimb (hl). (F) Head of the embryo shown in E (E12.5) showing high expression of Dan in the facial primordia. Fig. 6. Histological sections of embryos hybridized in wholemount with a Dan probe. (A) Parasagittal section through the presomitic mesoderm (psm) and the region of nascent somite formation of an E 9.5 embryo, highlighting Dan expression in the caudal aspect of the last formed somite (bracket). (B) Slightly oblique cross section through the presomitic mesoderm (psm) of an E9.5 embryo showing Dan expression in cells proximal to the neural tube (nt). That only one side shows staining, reflects the anterior/posterior periodicity of expression in this region, as seen in (A). (C) A section through the trunk region of an E9.5 embryo highlighting Dan expression in epithelial somites as they re-initiate expression (see text). Transcripts are confined to the medial half of the somites (s), where in contact with the neural tube (nt). (D) Section through more mature somites of the embryo shown in (C), where Dan expression is restricted to the medial part of the sclerotome (sc) proximal to the neural tube (nt), but does not encompass the dorsal lip of the dermomyotome. (E) A horizontal section through the tail of an E12.5 embryo showing Dan transcripts in myotomes (m). Note absence of expression in sclerotome (sc), except in caudal-most segments. (F) Section through the forelimb bud (lb) of an E12.5 embryo showing Dan expression in subepithelial mesenchymal expression of dorsal (upper) and ventral (lower) domains. (G) Section through the branchial region of an E10.5 embryo showing localized Dan expression in endodermal epithelium of pharyngeal pouches 1 and 2 (pp), and mesenchyme of the branchial arches 1 and 2 (ba1, ba2). (H) The same embryo as shown in G, showing Dan expression in cephalic mesenchyme subjacent to the midbrain (mb) and hindbrain (hb), with expression absent under the forebrain (fb). In this section, expression is also evident in branchial arch (ba) mesenchyme. (I) Section through the facial primordia of an E12.5 embryo emphasizing the sub-epithelial nature of Dan expression in this region (arrows), as in limb buds. Fig. 7. (A,B) Co-detection of Dan and Myf5-nlacZ expression in E11 embryos. (A) Dan expression in axial muscle precursors is obscured by nlacZ activity driven by Myf5 regulatory elements (arrows). However stripes corresponding to regionalized sclerotomal expression are visible (arrow heads). (B) Higher magnification of the caudal region of the embryos shown in A. (C,D). Expression pattern of Dan in the limb buds (lb) of E12.5 wildtype (C) or Splotch mutant (D) embryos. No difference in the pattern was detected, confirming Dan expression in local limb mesenchyme.
E. Stanley et al. / Mechanisms of Development 77 (1998) 173–184
central portion of the neural tube (not shown). Each stripe of Dan expression in sclerotome was situated between the maturing dorsal root ganglia, a region that will form the neural arches of the vertebrae (Verbout, 1985). Expression was absent in the vertebral body. 2.7. Dan expression in limbs and head mesenchyme In limbs, both wholemounts and sections showed expression in dorsal and ventral sub-epithelial domains (Figs. 5E
and 6F). This region becomes populated with myoblasts that migrate into the limb bud from the hypaxial myotomal compartment of adjacent somites (Tajbakhsh and Buckingham, 1994). To test whether Dan was expressed in muscle or local limb mesenchyme, we examined expression in limbs of Splotch mutant embryos, in which myoblast migration is blocked due to the loss of function of the Pax3 gene (Bober et al., 1994; Goulding et al., 1994; Yang et al., 1996). The Dan pattern in limbs was identical in homozygous and normal siblings (Fig. 7C,D), indicating that expression was
5
6
179
7
180
E. Stanley et al. / Mechanisms of Development 77 (1998) 173–184
established autonomously in dorsal and ventral domains within local limb mesenchyme. Within facial primordia, Dan expression was first established at E9.5 in the mesenchyme of the mandibular arch (Fig. 5B,D). At E10.5, expression occurred in the mandibular arch and its maxillary process, with sections showing transcripts also in the endoderm of adjacent pharyngeal pouches 1 and 2 (Fig. 6G). By E12.5, expression was seen in a sub-epithelial layer of facial mesenchyme (Fig. 6I). Sections also confirmed Dan expression in head mesenchyme situated beneath the midbrain and hindbrain at E9.5 (Fig. 6H).
3. Discussion 3.1. Structure of the Dan protein In this paper we have investigated the significance of the amino acid homology between mammalian DAN and Xenopus cerberus. DAN is a putative cell cycle regulator and tumour suppressor (Enomoto et al., 1994), while cerberus is a secreted cytokine involved in anterior embryonic patterning (Bouwmeester et al., 1996). Although weak, we found that the homology between DAN and cerberus was restricted to a cysteine-rich region in which the arrangement of cysteines and other key amino acids resembled that present in a protein motif called the cystine knot, the structural backbone of a superfamily of secreted proteins that includes TGFbs and BMPs. This homology, and our finding that DAN is a secreted dimeric glycoprotein, strongly suggests that DAN, cerberus, its mouse relative Cer-1, as well as the related proteins DRM and AA389245, constitute a novel branch of the cystine knot superfamily. 3.2. Biological activity of DAN and related proteins Dan was originally isolated as a gene down-regulated upon transformation of fibroblasts with v-src. It is also down-regulated after transformation with a variety of other oncogenes and subsequent studies have documented its negative effects on growth of transformed and non-transformed cells (Ozaki and Sakiyama, 1993, 1994; Ozaki et al., 1995, 1996). Normal 3Y1 fibroblasts over-expressing Dan, for example, show retarded growth and accumulate in the G1 phase of the cell cycle (Ozaki et al., 1995). It is interesting to note that the related protein, DRM, which has significant homology to DAN over the putative cystine knot region (see Fig. 1), was also identified as a gene down-regulated in transformed cells, in this case with the v-mos oncogene (Topol et al., 1997). Our finding that DAN is a secreted glycoprotein strongly suggests that its cell cycle effects occur via an autocrine secretion loop. Furthermore, the activity of DAN in Xenopus animal cap assays suggests that DAN is an inhibitor of BMP signalling. DAN was able to induce a suite of markers
in animal caps similar or identical to those induced by a range of synthetic and natural BMP inhibitors, acting at different stages in the signalling pathway (Piccolo et al., 1996; Zimmerman et al., 1996; Fainsod et al., 1997; Hemmati-Brivanlou and Melton, 1997; Frisch and Wright, 1998). A common feature of these agents is that they neuralize and anteriorize animal cap ectoderm. The effects are thought to reflect the natural role played by BMPs in establishing ectodermal cell fate, as well as the induction of a default anterior neural state when BMP signalling is inhibited (Hemmati-Brivanlou and Melton, 1997). In animal caps, endogenous BMP4 is robustly expressed and the predominant cellular fate is that of epidermis (HemmatiBrivanlou and Melton, 1997). However, BMPs appear to act as morphogens in the determination of a variety of tissue fates, including ectoderm (Dosch et al., 1997; Wilson et al., 1997), with the level of active BMP dictated by an interplay between ligand and antagonist. In animal caps, induction of anterior neural tissue would appear to be at one extreme end of the spectrum, that of total BMP signalling inhibition. While Dan has characteristics of a BMP antagonist in animal caps, it was a poor inhibitor, when compared to cerberus, of the ventro-lateral mesoderm-inducing activity of co-injected BMP4. On the other hand, co-injected BMP4 strongly inhibited induction of anterior neural markers by DAN. One interpretation is that DAN is less stable then cerberus, or is a less potent inhibitor of BMP4. It is possible, however, that the induction of these markers by DAN reflects a totally different sort of activity. Inhibition of Wnt signalling in caps with a dominant-negative version of Wnt-8 or the natural Wnt-inhibitor dkk-1, also leads to induction of anterior neural markers (Hoppler et al., 1996; Glinka et al., 1998) and specification of anterior structures in vivo appears to require simultaneous blockade of both BMP and Wnt signalling (Glinka et al., 1997). Furthermore, cerberus has been shown to have anti-Wnt activity (Glinka et al., 1997). Recently published experiments by Harland and colleagues shed light on this issue (Hsu et al., 1998). They have independently demonstrated that DAN is a secreted protein and that DAN, the human cerberus relative hCER, as well as a related factor Gremlin (the Xenopus cognate of DRM), block BMP2 signalling by direct binding to ligand. In addition to inhibiting BMP2, cerberus (although not DAN or Gremlin) can inhibit signalling from activin and Xnr-2, other members of the TGFb superfamily. While the particular affinities of DAN for different BMP and TGFß members have not been quantified nor compared to cerberus and other members, our data collectively suggest that DAN, cerberus and DRM/Gremlin constitute a novel branch of the cystine knot superfamily and inhibit signalling from cytokines of the TGFb superfamily (including BMPs), potentially with distinct potencies and/or specificities. Other possible activities for DAN, however, remain to be explored.
E. Stanley et al. / Mechanisms of Development 77 (1998) 173–184
3.3. Expression of Dan during development Dan was found to be expressed in virtually all the adult tissues analyzed by RNase protection assays. During embryonic development, however, expression was regionalized and dynamic. Transcripts were detected principally in four zones: in cephalic mesenchyme underlying the midbrain and hindbrain, in forming and maturing somites, in dorsal and ventral domains within developing limbs, and in the first branchial arch, then facial mesenchyme. Although speculative at this point (see above), we can rationalize Dan expression at these sites in terms of a role in BMP inhibition. Within limbs and face, for example, Dan expression occurred in sub-epithelial mesenchyme. The common feature of these zones is that they are invaded by migratory myoblasts before their differentiation into skeletal muscle. In the case of facial muscles, myoblasts derive from prechordal and paraxial mesoderm (Trainor et al., 1994), while in the case of limbs they derive from the hypaxial compartment of somites at the level of the limbs (Milaire, 1976; Christ and Ordahl, 1995). It is clear from our analysis of Splotch mutant embryos, in which myoblast migration is blocked, that Dan was not principally expressed in myoblasts themselves. It has been shown that BMP4 expressed in lateral mesoderm inhibits myogenesis in the lateral somite (Hirsinger et al., 1997); hence, DAN expression in the myogenic region of limbs and face may serve to protect invading myoblasts from the inhibitory effects of BMPs, expressed in adjacent regions to promote chondrogenesis and regulate cell proliferation and/or death (Francis et al., 1994; Francis-West et al., 1995; Marazzi et al., 1997; Zou et al., 1997). Noggin, also a BMP inhibitor, is expressed in the limb in a dynamic fashion, and a role in modulating the chondrogenic and other activities of BMPs in the limb has been demonstrated by the knockout phenotype in mice (Brunet et al., 1998). An analogous function for DAN in modulating BMP activity within the developing somite can also be proposed. While the recent focus has been on the influence of BMP-4 in somite patterning, because of its expression in lateral mesoderm, the developing somite is potentially subject to the influences of several different BMPs expressed in different tissues, for example BMP3 and 7 in notochord, BMP4 in ventral mesoderm, and BMP4 and 7 in dorsal ectoderm (McMahon et al., 1998). BMP4 expressed from lateral mesoderm delays the elaboration of myogenesis in lateral, but not medial somite (Hirsinger et al., 1997). The medial somite appears to be protected from the inhibitory effects of laterally-expressed BMP4 by factors or signals arising in axial tissues, and this inhibition appears to be essential for both myogenic and sclerotomal differentiation (Pourquie et al., 1993; Hirsinger et al., 1997; Reshef et al., 1998). One such protective factor could be noggin, itself expressed in the dorso-medial region of the somite, as well as notochord and floorplate (Hirsinger et al., 1997; Capdevila and Johnson, 1998; McMahon et al., 1998; Reshef et al., 1998). Dan
181
is also expressed in the medial portion of the somite from its earliest stages, although in contrast to noggin, is more prominent ventrally, suggesting an early role in sclerotomal specification. As somites matured, Dan was also expressed in the forming myotome, potentially protecting it from the known inhibitory effects of BMP. As demonstrated in both chick and mouse, BMPs can, if unimpeded, act at long range across the whole somite, inducing lateral markers medially and totally inhibiting myogenesis (Hirsinger et al., 1997; McMahon et al., 1998). Since multiple BMPs are expressed in the vicinity of the somite, differences in potency, specificity and timing of expression between individual BMP/TGFb inhibitors may add complexity and subtlety to how and where morphogen gradients are established. This may influence how signalling through a single pathway is used for multiple purposes within the confines of an embryonic structure containing only a few hundred cells or less.
4. Experimental procedures 4.1. DNA constructs The complete coding region of Dan was amplified by PCR from a murine lung cDNA library (Stratagene, catalogue number 936307) with oligonucleotide primers Dan2 (5′AGCTGGCGCGCCTGCTTTGGGTCCTGGTGG-3′) and Dan3 (5′-AGCTACGCGTGTCCTCAGCCCCCTCTTCCTC3-′). PCR was performed with 50 ng of each primer in the presence of 200 mM dNTPs, 2.5 mM MgCl2, 0.2 units of Taq polymerase (GIBCO) in the manufacturer’s buffer. Reaction conditions were as follows: 4 cycles of 95°C (20 s); 50°C (30 s); 72°C (60 s), followed by 26 cycles of 95°C (20 s); 60°C (30 s); and 72°C (60 s). The product was gel purified and digested with AscI and MluI prior to insertion into the AscI site of the mammalian expression vector pEFBOS I FLAG (Mizushima and Nagata, 1990; Biben et al., 1998). DAN protein produced from this vector (pEFBOS DAN-cFLAG) carried a FLAG epitope fused to its C-terminus. 4.2. DAN protein expression and purification pEFBOS DAN-cFLAG vector was co-transfected with a puromycin resistance plasmid into SR-3Y1 fibroblasts using Lipofectamine (GIBCO) according to manufacturer’s instructions. Following selection with puromycin (Sigma; 25 mg/ml), resistant colonies were picked and expanded. Culture supernatants were assayed for DAN-cFLAG using the M2 anti-FLAG antibody (Kodak Eastman) immobilized on a Biosensor 2000 (Pharmacia). DAN-cFLAG was purified from conditioned medium of the highest producing clone using M2 affinity resin (Kodak Eastman) and elution with FLAG peptide (Kodak Eastman; 60 mg/ml) in Trisbuffered saline. Column fractions were monitored by
182
E. Stanley et al. / Mechanisms of Development 77 (1998) 173–184
Western blotting using M2 antibody. DAN was dialyzed to remove FLAG peptide, buffer exchanged into 20 mM Tris, 0.02% Tween 20 and further purified by ion exchange using a Mono Q 5/5. DAN eluted at approximately 0.3 M NaCl. Fractions were monitored on the Biosensor 2000 and by SDS-PAGE. For N-glycosylation analysis, COS cell-derived DAN-cFLAG was treated with N-glycosidase F (Boehringer), as per manufacturer’s instructions, before PAGE and Western blotting using the M2 antibody. N-terminal sequence was determined on a Hewlett Packard N terminal sequencer using Edman degradation chemistry. 4.3. Xenopus laevis embryo procedures Embryo production, staging, microinjection and manipulation were as previously described (Biben et al., 1998). 0.5–1.0 ng of synthetic mRNA was microinjected into the animal pole region of fertilized eggs. DAN-cFLAG template was generated by subcloning an XbaI fragment from pEFBOS DAN-cFLAG (see above) into the EcoRV site of transcription vector pT7TS (Cleaver et al., 1996). Cerberus template was a PCR-generated coding fragment cloned into pCS2 (Biben et al., 1998). mRNA injections, isolation of animal caps and RT-PCR analysis of marker expression were performed as previously described (Biben et al., 1998). 4.4. In situ hybridization and beta-galactosidase staining Embryos were collected and processed for in situ wholemount hybridization as described, using digoxygeninlabelled riboprobes (Wilkinson, 1992; Biben and Harvey, 1997). Dan probes were transcribed from pBluescript SK + carrying PCR-derived fragments corresponding to nucleotides 539–1660 (C-terminal 21 amino acids and entire 3′ untranslated region) or 68–604 (coding region only) of the mouse cDNA sequence (Genbank D50263) (Ozaki et al., 1996). Staining for b-galactosidase using X-gal followed by in situ hybridization was carried out as described (Tajbakhsh et al., 1997). 4.5. RNase protection analysis RNA from adult tissues was prepared as described (Chomzynski and Saachi, 1987), prior to a second round of purification using the CsCl2 cushion method (Sambrook et al., 1989). RNase protections were performed as described (Ambion Instruction manual RPA II; catalogue #1410). Dan probe corresponded to nucleotides 553–715 of the published sequence (Ozaki et al., 1996) cloned into the HincII site of pBluescript SK+ (Stratagene). The control cyclophilin probe (Hasel and Sutcliff, 1990) was synthesized from a template obtained from Ambion. Dan and cyclophilin probes were included in the same reaction, with Dan in approximately 50 fold excess.
Acknowledgements We thank David Elliott for help with transfections and Steven Mihajlovic for histology. This work was supported by the National Health and Medical Research Council of Australia (NH&MRC), the National Heart Foundation of Australia, the Human Frontiers Science Program and Amrad Operations Pty Ltd. E.S. holds an NM&MRC C.J. Martin Fellowship. C-C.W. holds an Australian Development Cooperation Fellowship. References Banville, D., Williams, J.G., 1985. The pattern of expression of the Xenopus laevis tadpole alpha-globin genes and the amino acid sequence of the three major tadpole alpha-globin polypeptides. Nuc. Acids Res. 13, 5407–5421. 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 expressed in the anterior primitive endoderm of the mouse gastrula. Mech. Dev. 68, 45–57. Berger, W., Meindl, A., van de Pol, T.J., Cremers, F.P., Ropers, H.H., Doerner, C., Monaco, A., Bergen, A.A., Lebo, R., Warburg, M., Zergollern, L., Lorenz, B., Gal, A., Bleeker-Wagemakers, E.M., Meitinger, T., 1992. Isolation of a candidate gene for Norrie disease by positional cloning. Nat. Genet. 1, 199–203. Biben, C., Harvey, R.P., 1997. Homeodomain factor Nkx2-5 controls leftright asymmetric expression of bHLH eHand during murine heart development. Genes Dev. 11, 1357–1369. Biben, C., Stanley, E., Fabri, L., Kotecha, S., Rhinn, M., Drinkwater, C., Lah, M., Wang, C.-C., Nash, A., Hilton, D., Ang, S.-L., Mohun, T., Harvey, R.P., 1998. Murine cerberus homologue mCer-1: a candidate anterior patterning molecule. Dev. Biol. 194, 135–151. Bober, E., Franz, T., Arnold, H.-H., Gruss, P., Tremblay, P., 1994. Pax-3 is required for the development of limb muscles - a possible role for the migration of dermomyotomal muscle progenitor cells. Development 120, 603–612. Bonthron, D., Orr, E.C., Mitsock, L.M., Ginsburg, D., Handin, R.I., Orkin, S.H., 1986. Nucleotide sequence of the pre-pro-von Willebrand factor cDNA. Nucleic Acids Res. 14, 7125–7127. Bouwmeester, T., Kim, S.-H., Sasai, Y., Lu, B., De Robertis, E.M., 1996. Cerberus is a head-inducing secreted factor expressed in the anterior endoderm of Spemann’s organizer. Nature 382, 596–601. Bradley, L.C., Snape, A., Bhatt, S., Wilkinson, D.G., 1993. The structure and expression of the Xenopus Krox-20 gene: conserved and divergent patterns of expression in rhombomeres and neural crest. Mech. Dev. 40, 73–84. Brunet, L.J., McMahon, J.A., McMahon, A.P., Harland, R.M., 1998. Noggin, cartilage morphogenesis, and joint formation in the mammalian skeleton. Science 280, 1455–1457. Capdevila, J., Johnson, R.L., 1998. Endogenous and ectopic expression of noggin suggests a conserved mechanism for regulation of bmp function during limb and somite patterning. Dev. Biol. 197, 205–217. Chen, Z.Y., Hendriks, R.W., Jobling, M.A., Powell, J.F., Breakefield, X.O., Sims, K.B., Craig, I.W., 1992. Isolation and characterization of a candidate gene for Norrie disease. Nat. Genet. 1, 204–208. Chomzynski, P., Saachi, N., 1987. Single step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162, 156–159. Christ, B., Ordahl, C.P., 1995. Early stages of chick somite development. Anat. Embryol. 191, 381–396. Cleaver, O.B., Patterson, K.D., Krieg, P.A., 1996. Overexpression of the tinman-related genes XNkx-2.5 and XNkx-2.3 in Xenopus embryos results in myocardial hyperplasia. Development 122, 3549–3556.
E. Stanley et al. / Mechanisms of Development 77 (1998) 173–184 Dosch, R., Gawantka, V., Delius, H., Blumenstock, C., Niehrs, C., 1997. Bmp-4 acts as a morphogen in dorsoventral mesoderm patterning in Xenopus. Development 124, 2325–2334. Enomoto, H., Ozaki, T., Takahashi, E., Nomura, N., Tabata, S., Takahashi, H., Ohnuma, N., Tanabe, M., Iwai, J., Yoshida, H., Matsunaga, T., Sakiyama, S., 1994. Identification of human DAN gene, mapping to the putative neuroblastoma tumor suppressor locus. Oncogene 9, 2785–2791. Evans, S., Yan, W., Murillo, M.P., Ponce, J., Papalopulu, N., 1995. tinman, a Drosophila homeobox gene required for heart and visceral mesoderm specification, may be represented by a family of genes in vertebrates: XNkx-2.3, a second homologue of tinman. Development 121, 3889– 3899. Fainsod, A., Deissler, K., Yelin, R., Marom, K., Epstein, M., Pillemer, G., Steinbeisser, H., Blum, M., 1997. The dorsalizing and neural inducing gene follistatin is an antagonist of BMP-4. Mech. Dev. 63, 39–50. Francis, P.H., Richardson, M.K., Brickell, P.M., Tickle, C., 1994. Bone morphogenetic proteins and a signalling pathway that controls patterning in the developing chick limb. Development 120, 209–218. Francis-West, P.H., Robertson, K.E., Ede, D.A., Rodriguez, C., IzpisuaBelmonte, J.C., Houston, B., Burt, D.W., Gribbin, C., Brickell, P.M., Tickle, C., 1995. Expression of genes encoding bone morphogenetic proteins and sonic hedgehog in talpid (ta3) limb buds: their relationships in the signalling cascade involved in limb patterning. Dev. Dyn. 203, 187–197. Frisch, A., Wright, C.V., 1998. XBMPRII, a novel Xenopus type II receptor mediating BMP signaling in embryonic tissues. Development 125, 431–442. Gammill, L.S., Sive, H., 1997. Identification of otx2 target genes and restrictions in ectodermal competence during Xenopus cement gland formation. Development 124, 471–481. Glinka, A., Wu, W., Delius, H., Monaghan, A.P., Blumenstock, C., Niehrs, C., 1998. Dickkopf-1 is member of a new family of secreted proteins and functions in head induction. Nature 391, 357–362. Glinka, A., Wu, W., Onichtchouk, D., Blumenstock, C., Niehrs, C., 1997. Head induction by simultaneous repression of Bmp and Wnt signalling in Xenopus. Nature 389, 517–519. Goulding, M., Lumsden, A., Paquette, A.J., 1994. Regulation of Pax-3 expression in the dermomyotome and its role in muscle development. Development 120, 957–971. Hasel, K.W., Sutcliff, J.G., 1990. Nucleotide sequence of a cDNA coding for mouse cyclophilin. Nucleic Acids Res. 18, 4019. Hemmati-Brivanlou, A., Melton, D., 1997. Vertebrate embryonic cells will become nerve cells unless told otherwise. Cell 88, 13–17. Hirsinger, E., Duprez, D., Jouve, C., Malapert, P., Cooke, J., Pourquie, O., 1997. Noggin acts downstream of Wnt and Sonic Hedgehog to antagonize BMP4 in avian somite patterning. Development 124, 4605–4614. Hoppler, S., Brown, J.D., Moon, R.T., 1996. Expression of a dominantnegative Wnt blocks induction of MyoD in Xenopus embryos. Genes Dev. 10, 2805–2817. Hsu, D.R., Economides, A.N., Wang, X., Eimon, P.M., Harland, R.M., 1998. The Xenopus dorsalizing factor Gremlin identifies a novel family of secreted proteins that antagonize BMP activities. Mol. Cell 1, 673– 696. Isaacs, N.W., 1995. Cystine knots. Curr. Opin. Struct. Biol. 5, 391–395. Jamrich, M., Sato, S., 1989. Differential gene expression in the anterior neural plate during differentiation. Development 105, 779–786. Kintner, C.R., Melton, D.A., 1987. Expression of Xenopus N-CAM RNA in ectoderm is an early response to neural induction. Development 99, 311–325. Kondo, K., Ozaki, T., Nakamura, Y., Sakiyama, S., 1995. DAN gene product has an affinity for Ni2 + . Biochem. Biophys. Res. Comm. 216, 209–215. Krieg, P.A., Varnum, S.M., Wormington, W.M., Melton, D.A., 1989. The mRNA encoding elongation factor 1-alpha (EF-1 alpha) is a major transcript at the mid-blastula transition in Xenopus. Dev. Biol. 133, 93–100.
183
Leyns, L., Bouwmeester, T., Kim, S.-H., Piccolo, S., De Robertis, E.M., 1997. Frzb-1 Is a secreted antagonist of wnt signaling expressed in the Spemann organizer. Cell 88, 747–756. Madisen, L., Webb, N.R., Rose, T.M., Marquardt, H., Ikeda, T., Twardzik, D., Seyedin, S., Purchio, A.F., 1988. Transforming growth factor-beta 2: cDNA cloning and sequence analysis. DNA 7, 1–8. Marazzi, G., Wang, Y., Sassoon, D., 1997. Msx2 is a transcriptional regulator in the BMP4-mediated programmed cell death pathway. Dev. Biol. 186, 127–138. McMahon, J.A., Takada, S., Zimmerman, L.B., Fan, C.-M., Harland, R.M., McMahon, A.P., 1998. Noggin-mediated antagonism of BMP signaling is required for growth and patterning of the neural tube and somite. Genes Dev. 12, 1438–1452. Meitinger, T., Meindl, A., Bork, P., Rost, B., Sander, C., Haasemann, M., Murken, J., 1993. Molecular modelling of the Norrie disease protein predicts a cystine knot growth factor tertiary structure. Nat. Genet. 5, 376–380. Milaire, J., 1976. Contribution of somite cells to the development of posterior limb buds in mice. Arch. Biol. (Liege) 87, 315–343. Mizushima, S., Nagata, S., 1990. pEF-BOS, a powerful mammalian expression vector. Nucleic Acids Res. 18, 5322. Ozaki, T., Ma, J., Takenaga, K., Sakiyama, S., 1996. Cloning of mouse DAN cDNA and its down-regulation in transformed cells. Jpn. J. Cancer Res. 87, 58–61. Ozaki, T., Nakamura, Y., Enomoto, H., Hirose, L., Sakiyama, S., 1995. Overexpression of DAN gene product in normal rat fibroblasts causes a retardation of the entry into S phase. Cancer Res. 55, 895–900. Ozaki, T., Sakiyama, S., 1993. Molecular cloning and characterization of a cDNA showing negative regulation in v-src-transformed 3Y1 rat fibroblasts. Proc. Natl. Acad. Sci. USA 90, 2593–2597. Ozaki, T., Sakiyama, S., 1994. tumor-suppressive activity of no3 gene product in v-src-transformed rat 3y1 fibroblasts. Cancer Res. 54, 646– 648. Pannese, M., Polo, C., Andreazzoli, M., Vignali, R., Kablar, B., Barsacchi, G., Boncinelli, E., 1995. The Xenopus homologue of Otx2 is a maternal homeobox gene that demarcates and specifies anterior body regions. Development 121, 707–720. 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. Pourquie, O., Coltey, M., Teillet, M.-A., Ordahl, C., Le Douarin, N., 1993. Control of dorsoventral patterning of somitic derivatives by notochord and floorplate. Proc. Natl. Acad. Sci. USA 90, 5242–5246. Reshef, R., Maroto, M., Lassar, A.B., 1998. Regulation of dorsal somite cell fates: BMPs and Noggin control the timing and pattern of myogenic regulatory expression. Genes Dev. 12, 290–303. Sakiyama, S., Ozaki, T. 1997 Dan gene (Review in Japanese). Gan to Kagaku Ryoho. Jpn. J. Cancer Chemother. 24, 1465–1470. Sakiyama, S., Ozaki, T., Enomoto, H., 1994. Molecular cloning and characterization of a cDNA showing tumour-suppressive activity in v-srctransformed 3Y1 rat fibroblasts. Advan. Enzyme Regul. 34, 247–255. Sambrook, J., Fritsch, E.F., Maniatis, T., 1989. Molecular Cloning: A Laboratory Manual, Second Edition. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Sasai, Y., Lu, B., Piccolo, S., De Robertis, E.M., 1996. Endoderm induction by the organizer-secreted factors chordin and noggin in Xenopus animal caps. EMBO J. 15, 4547–4555. Shawlot, W., Deng, J.M., Behringer, R.R., 1998. Expression of the mouse cerberus-related gene, Cerr1, suggests a role in anterior neural induction and somitogenesis. Proc. Natl. Acad. Sci. USA 95, 6198–6203. Sive, H., Hattori, K., Weintraub, H., 1989. Progressive determination during formation of the anteroposterior axis in Xenopus laevis. Cell 58, 171–180. Sparrow, D.B., Kotecha, S., Towers, N., Mohun, T.J., 1998. Xenopus eHand: a marker for the developing cardiovascular system of the embryo that is regulated by bone morphogenetic paroteins. Mech. Dev. 71, 151–163.
184
E. Stanley et al. / Mechanisms of Development 77 (1998) 173–184
Tajbakhsh, S., Buckingham, M.E., 1994. Mouse limb muscle is determined in the absence of the earliest myogenic factor myf-5. Proc. Natl. Acad. Sci. USA 91, 747–751. Tajbakhsh, S., Rocancourt, D., Buckingham, M., 1996. Muscle progenitor cells failing to repond to positional cues adopt non-myogenic fates in myf-5 null mice. Nature 384, 266–270. Tajbakhsh, S., Rocancourt, D., Cossu, G., Buckingham, M., 1997. Redefining the genetic hierarchies controlling skeletal myogenesis: Pax-3 and Myf-5 act upstream of MyoD. Cell 89, 127–138. Tajbakhsh, S., Sporle, R., 1998. Somite development: constructing the vertebrate body. Cell 92, 9–16. Thomas, P., Brickman, J.M., Popperl, H., Krumlauf, R., Beddington, R.S.P., 1997. Axis. Duplication and Anterior Identity in the Mouse Embryo In Pattern Formation during Development, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, Vol. LXII, pp. 115–125. Tonissen, K.F., Drysdale, T.A., Lints, T.J., Harvey, R.P., Krieg, P.A., 1994. XNkx2.5, a Xenopus Gene Related to Nkx-2.5 and tinman: Evidence for a Conserved Role in Cardiac Development. Dev. Biol. 162, 325–328. Topol, L.Z., Marx, M., Laugier, D., Bogdanova, N.N., Boubnov, N.V., Clausen, P.A., Calothy, G., Blair, D.G., 1997. Identification of drm, a novel gene whose expression is suppressed in transformed cells and which can inhibit growth of normal but not transformed cells in culture. Mol. Cell. Biol. 17, 4801–4810. Trainor, P.A., Tan, S.-S., Tam, P.L., 1994. Cranial paraxial mesoderm: regionalization of cell fate and impact on craniofacial development in mouse embryos. Development 120, 2397–2408.
Verbout, A.J., 1985. The development of the vertebral column. Adv. Anat. Embryol. Cell Biol. 90, 1–122. Wang, S., Krinks, M., Lin, K., Luyten, F.P., Moos, J.M., 1997. Frzb, a secreted protein expressed in the spemann organizer, binds and inhibits Wnt-8. Cell 86, 757–766. White, P.S., Maris, J.M., Beltinger, C., Sulman, E., Marshal, H.N., Fujimori, M., Kaufman, B.A., Biegel, J.A., Allen, C., Hilliard, C., Valentine, M.B., Look, A.T., Enomoto, H., Sakiyama, S., Brodeur, G.M., 1995. A region of consistent deletion in neuroblastoma maps within chromosome 1p36.2–36.3. Proc. Natl. Acad. Sci. Usa 92, 5520–5524. Wilkinson, D.G., 1992. Whole mount in situ hybridization of vertebrate embryos. In: Wilkinson, D.G. (ed.), In Situ Hybridization: A Practical Approach, Oxford University Press, Oxford, pp. 75–84. Wilson, P.A., Lagna, G., Suzuki, A., Hemmati-Brivanlou, A., 1997. Concentration-dependent patterning of the Xenopus ectoderm by BMP4 and its signal transducer Smad1. Development 124, 3177–3184. Yang, X.M., Vogan, K., Gros, P., Park, M., 1996. Expression of the met receptor tyrosine kinase in muscle progenitor cells in somites and limbs is absent in Splotch mice. Development 122, 2163–2171. Zimmerman, L.B., De Jesus-Escobar, J.M., Harland, R.M., 1996. The Spemann organizer signal noggin binds and inactivates bone morphogenetic protein 4. Cell 86, 599–606. Zou, H., Wieser, R., Massague, J., Niswander, L., 1997. Distinct roles of type I bone morphogenetic protein receptors in the formation and differentiation of cartilage. Genes Dev. 11, 2191–2203.