Figure 1.
Spatio-temporal expression of Dlx-Msx in developing limbs.
A-N, forelimbs. O-AB, hindlimbs, at E10.5 (A-H, O-V) and E11.5 (I-N, W-AB). Whole-mount X-gal staining on FLs and HLs from Dlx5+/−, Msx1+/− and Msx2+/− (lacZ+) heterozygous embryos are shown. Expression of Dlx6 was detected by WMISH on embryonic limbs at the same ages and shown. At E10.5 Msx2 and Msx1 are expressed in the AER and the anterior and posterior mesoderm of HLs and FLs. Dlx5 and Dlx6, at E10.5, are expressed in the AER of HLs and FLs and in the anterior limb mesoderm only of the FLs, but not of the HLs. At later stages (E11.5), Dlx5 and Dlx6 are then expressed in the anterior mesoderm of HLs. Black arrows indicate mesodermal expression. The AER is also indicated. Black asterisks indicate absence of expression. W’,W’’ histologic transversal sections of E11.5 HLs from Dlx5+/− embryos, stained with Xgal. Y’,AA’ histologic transversal sections of HLs from Msx1+/− (Y’) and Msx2+/− (AA’) embryos, to compare AER and mesodermal expression between these genes. Section planes and position are reported with red lines (in W, Y and AA). The extent of the Msx1-positive anterior and posterior mesoderm regions, based on the micrographs in AA’ and AC’, are indicated with dashed lines. A strong Dlx5-lacZ signal is detected in the AER (W’ and W’’), a weak Dlx5-lacZ signal, overlapping with the Msx1-lacZ and the Msx2-lacZ signal, is detected in the anterior mesoderm (W’’, indicated by black arrows).
Figure 2.
Reduction of Msx2 expression in Dlx5/Dlx6 DKO HLs.
A-P. Whole-mount X-gal staining to detect Msx1 and Msx2 expression in Dlx5;Dlx6 DKO. In the FL (A-D,I-L) no changes of expression is observed whereas in the HL (E-H, M-P), Msx2 expression is reduced in the AER and in the anterior limb mesoderm of the Dlx5;Dlx6 DKO HLs. Q,R. Sections of WT (Q) and Dlx5;Dlx6 DKO mutant HLs (R) hybridized in situ to detect Msx2, showing a drastically reduced Msx2 signal in the AER and in the underlying mesoderm, but not in a proximal mesoderm territory. S-U. Quantification of the expression of Dlx5, Dlx6, Msx1 and Msx2 mRNAs by qRT-PCR in HLs from Dlx5;Dlx6 DKO (S), Msx2−/− (T) and Msx1−/− (U), relative to WT. The results show a reduction of 45% of Msx2 expression in the Dlx5;Dlx6 DKO HLs compared to WT, but not of Msx1. Dlx5 and Dlx6 expression is downregulated in Msx1 KO HLs but not in Msx2 KO HLs. Expression of the knocked-out genes was also tested, as control, and always found to be reduced to undetectable levels (not shown).
Figure 3.
Skeletal preparations of the HLs of single and combined Msx;Dlx mutant animals.
Chondroskeletal preparation of the HLs of E14.5 embryos (micrographs on the left) and full skeletal preparation on newborn animals (micrographs on the right), representing single and combined Dlx;Msx mutant genotypes (indicated on the left). The HLs of Msx2+/−;Dlx5;Dlx6 DKO animals (E,F) display an aggravated ectrodactyly phenotype compared to Dlx5;Dlx6 DKO ones (C,D), with fusion of the external digit (1with 2, 4 with 5) and hypoplasia of the central digit. Msx2;Dlx5;Dlx6 TKO HLs (G,H) display a further aggravated ectrodactyly phenotype, with the external digits fused and extended towards the opposite (anterior-posterior) sides, and a complete absence of the central digit. The limbs of Msx1+/−;Dlx5;Dlx6 DKO mice (not shown) show ectrodactyly similar to that observed in Dlx5;Dlx6 DKO mice, whereas Msx1;Dlx5;Dlx6 TKO mice (I,J) show ectrodactyly and loss of skeletal elements deriving from the anterior mesoderm of the autopod and zeugopod, a phenotype seen in Msx1;Msx2 DKO mutant embryos (K). The stylopod shows no evident defects. The anterior-posterior orientation is shown. The numbers 1–5 indicate the digits (1 is the toe). Asterisks indicate hypoplasia or absence of skeletal structures. The drawings on the left schematically illustrate the Dlx-related (red elements) and the Msx-related (purple elements) skeletal defects, corresponding to the genotypes examined.
Figure 4.
Expression of Bmp4 in Dlx5;Dlx6 DKO limbs.
A-D. Detection of Bmp4 mRNA by WMISH on WT (left) and Dlx5;Dlx6 DKO mutant (right) limbs, at E11. FLs are on the top, HLs are on the bottom. E,F. In situ detection of Bmp4 mRNA in the pharyngeal arches region of WT (left) and Dlx5;Dlx6 DKO mutant (right), at E11, as a control for RNA preservation. G-L. Detection of Gremlin mRNA in FLs (G,H) and HLs (I-L) of WT (left) and Dlx5;Dlx6 DKO mutant embryos (right), at E10.5. While Bmp4 expression in the anterior mesoderm of the FLs (A,B) or the HLs (C,D) is unchanged (green arrowheads), expression in the central wedge of the AER of mutant embryos is diminished in the HLs, but not in the FLs (red arrows in D). Gremlin expression is unchanged both in the FLs (G,H) and in the HLs (I-L) of Dlx mutant embryos (red arrowheads). Genotypes and probes are reported on the top. The Anterior-Posterior (A-P) orientation is indicated. M,N. whole-mount photographs documenting the reduced size of mutant embryos and justifying the slightly reduced size of the mutant limbs, often observed. O. Quantification of Bmp2 and Bmp4 mRNAs by qRT-PCR in the anterior and posterior halves of HLs from embryos with the genotype indicated on the top of each graph, compared to WT.
Figure 5.
Binding of DLX5 on predicted conserved elements close to BMP2 and BMP4. A.
Location of predicted conserved Dlx binding sites in regions of the mammalian genome around the BMP2 (top) and BMP4 (bottom) loci. Sites are indicated with colour vertical bars, the chromosomal position and coordinates are shown. The mammalian genomic conservation is reported on the bottom. With the exception of B2-RE3, all elements fall within stretches of conserved sequences. B. Sequences and alignment of the predicted Dlx binding elements. The sequence corresponding to the PWM is shown in red. Dashed lines indicate the degenerated part of the binding sequence. C. Western blot analysis to demonstrate expression of DLX5-myc and DLX5-Q184P-myc proteins in U2Os cells. The molecular weight of the detected proteins is indicated on the left. D. ChIP analyses on the predicted Dlx binding sites in the human U2Os cells, transfected with the DLX5-myc and DLX5-Q178P-myc expression vector, or with the control empty vector. The input chromatin (positive control) is shown on the left, ChIP with or without anti-myc are shown in the mid panels. ChIP with and without anti-H4Ac on the same elements are shown on the right.
Figure 6.
A dynamic model for Dlx-Msx-Bmp functional interactions during HL development.
Schematic drawing to summarize our results and illustrate our model of functional interaction between Dlx5;Dlx6, Msx1, Msx2, Bmp2 and Bmp4. On the top, a scheme of the limb bud, the AER (in light blue color) and the mesoderm (in pink color) is reported. Below, the proposed dynamic model of gene regulations, shown for an Early (E9.5–E10, on the left) and a Late (E10–E10.5, on the right) phases of HL development, using the same color code as above. The anterior mesenchyme is framed with a dotted black box; the Ant-Post and Prox-Dist directions are shown. Bmp2 and Bmp4 are placed at the interface between the AER and the Ant Mes, to indicate that these are diffusible signaling molecules.