Figure 1.
The expression of Dmrt2, visualised by whole mount in situ hybridization.
Embryos were hybridized with a Dmrt2 RNA probe. (A) an E8.5 embryo, (B) an E9.0 embryo, (C, F) E9.5 embryos, (D) an E10.5 embryo, (E) an E11.5 embryo. C′ and C″ are magnified views of C. Dmrt2 transcripts are present in all somites until E11.5 when they are only detected caudally. (F) A Pax3IRES-nlacZ/+ embryo at E9.5 hybridized with a Dmrt2 probe (blue) and treated with Red-gal to reveal β-galactosidase (β-gal) expression from the Pax3IRES-nlacZ allele (red). Dmrt2 expression was also detected in proximal forelimb buds and branchial arches (arrowheads in D). (G–I) Transverse sections of the embryos shown in F. Equivalent axial levels are also indicated in C′ and C″. White dotted lines outline the dermomyotome. Dmrt2 expression is observed in the epaxial domain of immature somites (C″, G), and then throughout the Pax3 positive dermomyotome (C′, H). In more mature anterior somites, the expression of Dmrt2 in the epaxial and hypaxial edges of the dermomyotome has decreased (I). S0, presomitic mesoderm prior to the first somite; NT, neural tube; Ep, epaxial somite; Hyp, hypaxial somite.
Figure 2.
The expression of Dmrt2 is controlled by Pax3.
(A–C) Whole mount in situ hybridization with a Dmrt2 probe of Pax3GFP/+ (A), Pax3PAX3-FKHR-IRESnlacZ/GFP (Pax3PAX3-FKHR/GFP, B) and Pax3Pax3-En-IRESnlacZ/+ (Pax3Pax3-En/+, C) embryos at E9.5. Pax3 expression is shown as GFP fluorescence (inserts in A, B), or as Salmon-gal staining (insert in C) of trunk somites. A′–C′ show enlargements of the posterior somite regions used for sections. (D–L) Sections at equivalent axial levels showing Dmrt2 transcripts in the dermomyotome of control Pax3GFP/+ (D–F), gain of function Pax3PAX3-FKHR/GFP (G–I) or partial loss of function Pax3Pax3-En/+ (J–L) embryos. Dmrt2 transcripts are higher in gain of function embryos, where they are now detectable throughout the dermomyotome, whereas when Pax3 activity is altered they are barely detectable in the epaxial domain of immature somites (J) and detectable in a restricted region of more mature somites (L). Arrowheads indicate the full extent of the dermomyotome. PSM, presomitic mesoderm; S0, PSM prior to the first somite.
Figure 3.
A conserved 286 bp element 5′ of the Dmrt2 gene directs Pax3-dependent expression in the somite.
(A) A schematic representation of the mouse Dmrt genomic locus showing overall sequence conservation (blue bars) between human, rat, chicken and Xenopus (green bars). (B) The nucleotide sequence of the conserved 286 bp Dmrt2 element (red circle in A) in mouse (the position on ch19; 25728951 to 25729236, about 18 kb 5′ of Dmrt2) and comparison with a homologous region of human, chick and Xenopus genomes, with asterisks indicating conserved nucleotides between all these species. Five putative Pax3 binding sites are framed in black and red. (C) Electrophoretic mobility shift assays (EMSA) performed with oligonucleotides conjugated with biotin, containing putative Pax3 binding sites (1–5) as indicated in (B), incubated with extracts of HEK293 cells, with (+) or without (−) a Pax3 expression vector, or oligonucleotides without biotin (+dsDNA), including mutated oligonucleotides (+mut-dsDNA). The Pax3 site in the 145 bp regulatory element at −57.5 kb from the Myf5 gene provides a positive control (Myf5-145; [9]). Significant binding is seen on site2 when Pax3 protein is present. For supershift assays, monoclonal Pax3 antibody (DSHB) was used (+anti-Pax3); +anti-Flag provides a negative control. Arrowheads show the band due to binding of the Pax3 protein. NS; non-specific. (D) Chromatin immunoprecipitation (ChIP) analysis of Pax3 binding to the 286 bp sequence in vivo was performed with chromatin prepared from somites of E9.5 embryos (without head, neural tube, and internal tissues). Upper panels show the evaluation of sheared genomic DNA (1/10 input) with two sets of Dmrt2 primers, for the potential regulatory sequence (286) and for an upstream control sequence (+20 k) without Pax3 binding sites (results not shown), in left and right panels respectively. As an additional positive or negative control, whole genome sheared genomic DNA, with (whole genome) or without proteinase K (No Proteinase K) treatment is shown (In the absence of proteinase K cross-linked chromatin blocks the PCR reaction). Lower panels show ChIP (IP) with two different Pax3 antibodies (αPax3, Bajard et al. (2006), Lagha et al. (2008)) and control rabbit serum. Left lower bands indicate Pax3 binding to the 286 bp sequence, not seen with the control +20 kb sequence (right panels). (E) A transient transgenic embryo (E9.5) with the 286 bp Dmrt2 fragment, fused to the TK promoter and nlacZ reporter shows dermomyotome expression. (F) The transgene, with mutated Pax3 binding site2, shows loss of β-galactosidase activity in the dermomyotome (F′), but some ectopic expression in the myotome and ventral somite (F″), seen in more mature somites in the non-mutated transgenic embryo (E″). Observations on transgenic embryos are summarised in Table 1.
Figure 4.
Dmrt2 is required for the maturation of the epaxial somite, where Myf5 is first expressed.
Whole mount in situ hybridization of Dmrt2+/− (A) and Dmrt2−/− (B) embryos at E9.5, with a Myf5 probe. (C–F) Sections of embryos shown in (A, B) at different axial levels - left hand panels show more mature somites; immature somites where the epaxial myotome is beginning to form are shown in right hand panels. The onset of Myf5 expression is perturbed in the absence of Dmrt2. (G, H) Whole mount in situ hybridization of Dmrt2+/− (G) and Dmrt2−/− (H) embryos with a myogenin (Myog) probe, showing a striking delay of myogenin expression in the absence of Dmrt2. (I, J) Whole mount in situ hybridization of Dmrt2+/− (1) and Dmrt2−/− (J) embryos showing Pax3 transcripts.
Figure 5.
Dmrt2 binding sites in the Myf5 epaxial enhancer are essential for normal activity in vivo.
(A) The sequence of the Myf5 epaxial enhancer (EpExt), which drives early Myf5 expression in the epaxial domain of newly formed somites. TCF/LEF binding sites are boxed in green, the Gli1 binding site in red, and four putative Dmrt2 binding sites in black. (B–D) EMSA performed with oligonucleotides conjugated with biotin, containing Dmrt2 binding sites (1–4) as indicated in A, incubated with extracts of HEK293 cells, with (+) or without (−) a Dmrt2HA expression vector, or oligonucleotides without biotin (dsDNA) (including mutated oligonucleotides; mut-dsDNA). In the supershift assay (D), binding was disrupted by adding anti-HA (Monoclonal; Roche) to HA-tagged Dmrt2, whereas control anti-Flag antibody had no effect. The arrowheads show the binding between the oligos and the tagged protein. NS; non-specific. (E, F) Transient transgenic analysis to examine the role of the Dmrt2 binding sites in the Myf5 EpExt enhancer. The transgenes, containing mutated Dmrt2 binding sites1, 3, 4, showed reduced β-galactosidase activity in developing somites with either the TK (E) or Myf5-BA (branchial arch) promoter region (F). The BA element that directs transgene expression to the branchial arches, provides a positive control (arrowheads in F). Observations on transgenic embryos are summarised in Table 1.
Figure 6.
Overexpression of Dmrt2 accelerates myogenesis.
(A) Luciferase assay in NIH3T3 cells with the Myf5EpExt fragment with (yellow bars) or without (green bars) mutated Dmrt2 binding sites 1, 3, 4 (mutMyf5EpExt) as used in Figure 5, and increasing concentrations of a Dmrt2-expression vector. As a positive control, β-catenin and Gli1 expression vectors were used (blue bars). Progressive increase in Dmrt2 expression, leads to increased activation of the Myf5EpExt which is lost when Dmrt2 binding sites 1, 3, 4 are mutated. (B) The transgene for Dmrt2 conditional over-expression, in which Cre recombination removes the CAT sequence leading to Dmrt2 and reporter Tomato red expression under the strong CAG promoter. (C, C′, D, D′) Transgenic embryos, containing the CAG-floxedstop-Dmrt2-IRES-tdTomato (CAG-LSL-Dmrt2) transgene, crossed with PGK-Cre mice (left in C, C′), and Pax3Cre/+ mice (left in D, D′). Ectopic Dmrt2 expression is seen in PGK-Cre or Pax3Cre/+ expressing embryos reflecting Cre expression (arrowheads in C, D). (E–I) Whole mount in situ hybridization of transgenic control or Pax3-Cre activated embryos at E9.5. Myf5 (E.H) and myogenin (F, I) expression are up-regulated in developing somites of embryos overexpressing the Dmrt2 transgene in Pax3 positive cells (arrowheads in E, F) compared to controls (H, I). (G, J) Sections showing somites of control or activated Dmrt2 transgenic embryos at the same axial level, treated by immunohistochemistry with a laminin antibody. Laminin is detected in the premature myotome, where it is not normally present at this developmental stage. (arrowheads in G). (K, L) Whole mount X-gal staining of Myf5nlacZ/+ embryos crossed onto the Dmrt2 expressing transgenic line, with (K) or without (L) activation by Pax3-Cre, show perturbed location of β-galactosidase activity. This is shown on sections at the axial levels indicated in K and L, in which Pax3 activation of the Dmrt2 transgene results in more extensive dermomyotome (arrows in K′, K″) and early myotome (K′) expression (not seen in the control (L′, L″). Arrowheads indicate somite extent.
Figure 7.
A schematic representation of the regulatory network presented in this paper.
Direct targets are shown in heavy type. Indirect targets are indicated by a discontinuous arrow. The myogenin promoter depends on a critical Ebox [47],[48] targeted by Myf5. α6β1 integrin expression depends on Myf5, in cells entering the myotome [39]. This receptor interacts with Laminin1 as a ligand, leading to formation of the basement membrane of the myotome. Laminin expression is modulated by Dmrt2 (see also [25]).
Table 1.
Transient transgenic embryos and lines (where indicated) (at E9.5, unless otherwise indicated).