July 1, 2002;
Repression through a distal TCF-3 binding site restricts Xenopus myf-5 expression in gastrula mesoderm.
The development of skeletal muscle
in the vertebrate embryo
is controlled by a transcriptional cascade that includes the four myogenic regulatory factors Myf-5, MyoD
, and MRF4
. The dynamic expression pattern of myf-5 during myogenesis is thought to be consistent with its role during early determination of the myogenic lineage. To study the factors and mechanisms, which regulate myf-5 transcription in Xenopus, we isolated a genomic DNA clone containing 4858 bp of Xmyf-5 5'' flanking region. Using a transgenic reporter assay, we show here that this genomic contig is sufficient to recapitulate the dynamic stage- and tissue
-specific expression pattern of Xmyf-5 from the gastrula
bud stages. For the primary induction of myf-5 transcription, we identify three main regulatory elements, which are responsible for (i) activation in dorsal mesoderm
, (ii) activation in ventral mesoderm
, and (iii) repression in midline mesoderm
, respectively. Their combined activities define the two-winged expression domain of myf-5 in the preinvoluted mesoderm
. Repression in midline mesoderm
is mediated by a single TCF binding site located in the 5'' end of the -4.8 kbp sequence, which binds XTcf-3 protein in vitro. Endogenous Wnt signaling in the lateral mesoderm
is required to overcome the long-range repression through this distal
TCF site, and to stimulate myf-5 transcription independently from it. The element for ventral mesoderm
activation responds to Activin. Together, these results describe a regulatory mosaic of repression and activation, which defines the myf-5 expression profile in the frog gastrula
[+] show captions
Fig. 2. Time course of pFLCFP reporter gene expression in transgenic embryos. Transcripts of the endogenous Xmyf-5 or the GFP reporter gene were detected by whole mount in situ hybridization with Xmyf-5 or GFP antisense RNA probes, respectively. (A–E) Expression pattern of endogenous Xmyf-5 RNA at (A) stage 10−, (B) stage 11, (C) stage 11.5, (D) stage 19, (E) stage 31; (F–J) expression of the pFLGFP transgene in pigmented embryos at the same stages. The expression pattern of GFP transgene was very similar to that of the endogenous myf-5, although a faint positive signal was detected above the dorsal lip in wide type embryos (A), but not in transgenic embryos at stage 10− (F). (A–C,F–H) Vegetal views of gastrulae with the dorsal side facing up; (D,I) dorsal views of neurulae; (E,J) lateral views of tadpoles. Neurulae and tadpoles: anterior to the left.
Fig. 1. Molecular cloning of Xmyf-5 upstream DNA. (A) The isolated 4858 bp genomic DNA fragment contains upstream sequences, 5′ untranslated region and part of the open reading frame of the Xmyf-5 gene. The TATA box, OLS motif and a putative TCF binding site are indicated. The transcription initiation site is indicated by arrow. (B) Diagram of Xmyf-5 promoter deletion constructs used in luciferase assays and transgenic experiments.
Fig. 3. Truncation analysis of the Xmyf-5 upstream region. Representative views of stage 11.5 transgenic embryos carrying the following transgenes: (A) p-295GFP, no significant signal was detected; (B) p-727GFP, subtle and non-specific GFP expression was extending throughout the animal hemisphere; (C,D) An arch-shaped expression pattern of GFP was present in both p-1778GFP and p-3384GFP transgenic embryos. Sections through the dorsal midline show that the positive signal is above the dorsal lip (inset). (E,F) GFP-positive signal extended ventrally and formed a ring surrounding the blastopore in p-4223GFP and p-4531GFP transgene embryos. All panels show vegetal views with dorsal side up, except (B) (lateral view, dorsal to the right).
Fig. 4. Activin inducibility of Xmyf-5 reporter genes. (A) Experimental scheme. Twenty-five picograms of the various myf-5–luciferase reporter constructs were injected into the animal pole of 2-cell stage embryos. Animal caps were dissected from stage 8.5 embryos and treated either immediately with 50 ng/ml Activin for 1 h or cultured in 1× MBS. At stage 12.5, explants were lysed, and luciferase activities were measured. (B) Fold of induction of the indicated myf-5 reporter constructs by Activin. For each construct, RLU of Activin treated caps was divided by RLU of untreated caps after normalization to internal standard (see Section 4). (C) Luciferase activities of the various constructs injected into the dorsal marginal zone of both cells at stage 2. RLU, relative light units.
Fig. 5. The distal TCF binding site mediates local repression of myf-5 transcription. (A–C) Representative transgenic embryos carrying the δTcfGFP construct during gastrulation (vegetal views, dorsal side up). GFP RNA was detected in the dorsal region at stage 10− (A); subsequently, the signal extended ventrally along the blastopore lip (B, stage 10.5), until it finally formed a full circle at stage 11.5 (C); arrowhead indicates blastopore lip pigmentation. (D) Luciferase assays of episomal reporter gene injections. Embryos were injected dorsolaterally at the 2-cell stage with 25 pg plasmids of pFLLuc or δTcfLuc, then harvested at stage 12.5 for measuring luciferase activities. RLU, relative light units, normalized to internal standard.
Fig. 6. Xenopus TCF3 binds to the distal Tcf-motif in the myf-5 gene in vitro. The DNA fragment from position −4739 to −4761 of the Xmyf-5 genomic DNA clone was radiolabeled and then incubated with nuclear extracts from midgastrula (stage 12.5) embryos. (A) A DNA–protein complex (C) was resolved by gel electrophoresis, which is marked by an arrow. The formation of this complex was inhibited by pre-incubation with unlabelled competitor DNA of wild type sequence (WT), but not by pre-incubation with point-mutated TCF-binding site (Mutant). The numbers above the gel indicate molar excess of unlabelled competitor DNA over labeled DNA probe. (B) Nuclear extracts from myc-tag Xtcf3-injected embryos form a DNA–protein complex (C) with the same mobility rate as nuclear extracts from un-injected embryos (Non-injected). Pre-incubation of the Xtcf3-injected extracts with Anti-myc-tag antibody, but not with unrelated IgG, produces a supershift complex (C*). No supershift complex was formed in extracts from un-injected embryos, irrespective of the antibody added.
Fig. 7. Positive Wnt input into Xmyf-5 expression. One hundred picograms of Frzb mRNA or 50 pg Lef1-EnR mRNA were microinjected laterally into one blastomere of 2-cell stage embryos. Embryos were processed for whole mount in situ hybridization at stage 12 with antisense myf-5 (A–C) or at stage 11.5 for GFP (D) probes. All embryos are shown in vegetal views with the dorsal/anterior side pointing up. Uninjected control embryo shows the normal Xmyf-5 expression pattern (A). Xmyf-5 activity in the dorsal–lateral mesoderm was reduced where Frzb (B) or Lef1-EnR was injected (C). In δTcfGFP transgenic embryos, GFP expression was locally abolished at the site of Frzb-injection (D). The asterisk demarcates weak blue staining for nuclear galactosidase, whose RNA was co-injected as lineage tracer.