XB-ART-17922Development August 1, 1996; 122 (8): 2359-66.
Xenopus mothers against decapentaplegic is an embryonic ventralizing agent that acts downstream of the BMP-2/4 receptor.
Dorsal-ventral patterning in vertebrate embryos is regulated by members of the TGF-beta family of growth and differentiation factors. In Xenopus the activins and Vg1 are potent dorsal mesoderm inducers while members of the bone morphogenetic protein (BMP) subclass pattern ventral mesoderm and regulate ectodermal cell fates. Receptors for ligands in the TGF-beta superfamily are serine-threonine kinases, but little is known about the components of the signal transduction pathway leading away from these receptors. In Drosophila the decapentaplegic protein (dpp), a homolog of vertebrate BMP-2 and BMP-4, functions in dorsal-ventral axial patterning, and a genetic screen for components involved in signaling by dpp has identified a gene named mothers against decapentaplegic (Mad). Mad encodes a unique, predicted cytoplasmic, protein containing no readily identified functional motifs. This report demonstrates that a gene closely related to Drosophila Mad exists in Xenopus (called XMad) and it exhibits activities consistent with a role in BMP signaling. XMad protein induces ventral mesoderm when overexpressed in isolated animal caps and it ventralizes embryos. Furthermore, XMad rescues phenotypes generated by a signaling-defective, dominant-negative, BMP-2/4 receptor. These results furnish evidence that XMad protein participates in vertebrate embryonic dorsal-ventral patterning by functioning in BMP-2/4 receptor signal transduction.
PubMed ID: 8756281
Article link: Development
Genes referenced: acta4 actc1 actl6a agtr1 arhgap26 bmp2 bmp4 bmpr1a dspp emb evx1 gdf1 h4c4 hoxa9 hoxb9 hoxc9 ncam1 nrp1 smad1 tbx2 tgfb1 twist1 wnt8a
Article Images: [+] show captions
|Fig. 1. Protein sequence comparison of Xenopus Mad with Drosophila Mad. The predicted Xenopus Mad (XMad) protein sequence is aligned with Drosophila Mad protein (DMad). The XMad cDNA encodes a predicted protein of 464 amino acids. XMad protein is 75% identical to DMad at the amino acid level (identical residues are indicated by vertical bars) and 85% similar when conservative amino acid substitutions are considered (double dots). Note the high degree of conservation in the N-terminal and C-terminal portions of the proteins. The GenBank accession number for XMad is U58834.|
|Fig. 2. Developmental expression of Xenopus Mad. (A) A developmental northern blot of total embryonic RNA shows that XMad is maternal and expressed at all stages of early development. Three transcripts are detected with the sizes (in kb) indicated on the left. The most abundant mRNA is 3.4 kb, and minor transcripts of 4.3 kb and 3.0 kb are also detected. It is not known whether these transcripts derive from closely related genes or splicing variants. Lane numbers correspond to developmental stages (Nieuwkoop and Faber, 1967): (7, 9) blastula, (11) gastrula, (15, 18) neurula, (26) tailbud tadpole and (38) swimming tadpole. (B) In situ hybridization shows that XMad transcripts are uniformly distributed in the early gastrula (stage 10). Dense staining covers the entire prospective ectoderm and marginal zone, but vegetal cells (the lighter region) do not stain efficiently by this procedure (Harland, 1991). A northern blot on isolated dorsal, ventral, animal and vegetal regions confirmed the uniformity of XMad expression at early stages (not shown). (C) A mid-gastrula embryo (stage 12) split sagitally reveals that XMad is expressed in the ectoderm and neurectoderm (arrows mark the ectodermal-mesodermal boundary), and expression in the underlying mesoderm is greater in the posterior, adjacent to the yolk plug (YP). This embryo is lightly pigmented and the brown line anterior to the yolk corresponds to involuted bottle cells. The embryo is positioned with the anterior to the left and dorsal at the top. (D) At tailbud tadpole stage 26 XMad expression is high in the central nervous system and head. (E) A close-up of the head of the embryo in D, highlighting XMad expression in the brain (b), eye (e) and head neural crest derivatives (mc, mandibular crest; hc, hyoid crest; abc, anterior branchial crest; pbc, posterior branchial crest). Expression in the otic vesicle, between the hyoid crest and anterior branchial crest, is also visible. Scale bars, 0.1 mm.|
|Fig. 3. XMad and DMad ventralize Xenopus embryos. Synthetic mRNAs encoding control (pGem vector) or Mad sequences were injected into the equatorial region of two dorsal or two ventral blastomeres at the 4-cell blastula stage, and phenotypes were scored at tadpole stage 40. Dorsal (A) or ventral (B) injections of control (pGem vector) mRNA resulted in normal embryos. (C) Dorsal injection of XMad mRNA caused severe ventralization. (D) Ventral injection of XMad mRNA caused posterior thickening and a slight reduction in the tail. The average dorso-anterior index, a measure of the degree of dorsal and anterior mesodermal patterning (Kao and Elinson, 1989), was for each group: A, DAI=5 (n=18); B, DAI=5 (n=18); C, DAI=0 (n=18); D, DAI=5 (n=20). (E) Expression of DMad in dorsal blastomeres ventralized the embryos (DAI=1.9, n=20), while ventral expression yielded relatively normal embryos (DAI=5, n=21; not shown). (F) Northern blot analysis of muscle actin (a dorsal mesodermal marker) and aT1 globin (a ventral mesodermal marker) gene expression in the control and XMadinjected embryos shown in A-D. Lanes a-d correspond to embryos in A-D. Note the loss of muscle actin and the significant increase in globin expression in embryos injected dorsally with XMad mRNA (lane c). This indicates that overexpression of XMad protein in the presumptive dorsal mesoderm respecified its fate to that of ventral mesoderm. Ventral expression of XMad boosted globin expression slightly (lane d), but dorsal or ventral injections of vector mRNA had no effect (lanes a and b). Histone H4 mRNA (Perry et al., 1985) was scored as a control for RNA loading.|
|Fig. 4. XMad and DMad induce mesoderm in animal caps. The panel shows an RT-PCR analysis of mesodermal marker gene expression in animal caps injected with 3.0 ng of mRNA for pGem (C), DMad (DM), XMad (XM) or Xenopus BMP-4 (B4). The last two lanes show RT-PCR products from stage-18 embryonic cDNA synthesized in the presence (emb RT+) or absence (emb RT-) of reverse transcriptase to control for cDNA synthesis and DNA contamination, respectively. Note that expression of DMad, XMad and BMP-4 induced each of the ventro-posterior mesoderm markers assayed. EF1-a was scored as a positive control for cDNA synthesis. Animal caps were harvested at stage 30 (tadpole) to score aT1 globin expression, stage 11 (mid-gastrula) to score Xwnt 8 and Xtwist, and stage 18 (neurula) to score Xhox-3 and Xlhbox6. The eF1-a signal shown corresponds to that of stage 11 cDNA, but all other cDNA samples treated with reverse transcriptase were positive.|
|Fig. 5. XMad rescues dominant-negative BMP2/4 receptor phenotypes. The panels display phenotypes of whole embryos (A-C) and isolated ventral marginal zones (VMZs, D-E) expressing the dominant-negative BMP-2/4 receptor (Graf et al., 1994), alone or in combination with XMad. 50 pg of each mRNA were injected into the marginal zone of two ventral blastomeres at the 4-cell stage. VMZs were explanted at stage 10.5, and embryos and VMZ explants were scored at stage 40. (A) Expression of the dominant-negative BMP- 2/4 receptor (tBR) from injected mRNA resulted in tadpoles with secondary axial structures in 33% of the cases (n=18). A typical secondary axis is indicated by the arrows. (B) Co-expression of XMad together with tBR resulted in 100% normal embryos (n=18). (C) Injection of control (pGem) mRNA resulted in 100% normal embryos (n=21). (D) VMZs expressing tBR elongated and developed pigmented melanocytes, a neural derivative. (E) VMZs from embryos co-expressing tBR and XMad were rescued to normalcy and formed oblong ‘belly pieces’ like control VMZs (F). (F) Injection of control (pGem) mRNA into VMZs resulted in a typical wild-type VMZ morphology. (G) A northern blot on RNA from the VMZs shown in D-F. VMZs expressing tBR (lane d) developed dorsal mesoderm, as revealed by the expression of muscle actin, and they lacked ventral mesoderm (red blood) as reflected by the absence of aT1 globin expression. When XMad was co-expressed with tBR (lane e) dorsalization of the VMZ was reversed; the explants lacked muscle and expressed globin, similar to control VMZs (lane f). Cytoplasmic actin mRNAs cross-hybridize with the muscle actin probe and migrate as two bands above the muscle-specific message, providing a positive control for RNA loading in the gel.|
|Fig. 6. Neural induction by tBR expression in animal cap ectoderm is inhibited by co-expression of XMad. Lanes 1-4 correspond to an RTPCR analysis of cDNA from animal caps injected at the 2-cell stage with 2 ng pGem RNA (lane C),1.0 ng XMad mRNA (lane XM), 50 pg tBR mRNA (lane tBR), and 50 pg tBR mRNA plus 50 pg XMad mRNA (lane tBR + XM). Caps were cut at stage 8 and harvested at neurula stage 18. Note the induction of N-CAM and NRP1 (a neuralenriched ribonucleoprotein; Richter et al., 1990) mRNAs when BMP signals were blocked by tBR. This effect was suppressed by coexpression of XMad. Some background expression of NRP1 is normal. Furthermore, XMad alone did not induce neural tissue (lane XM). Lanes 5 and 6 are positive and negative controls for RT-PCR as in Fig. 2. EF1-a is a positive control for cDNA synthesis.|
|smad1 (SMAD family member 1) gene expression in Xenopus laevis embryo, assayed via in situ hybridization, NF stage 26, lateral view, anterior left, dorsal up.|