Figure 1.
GATA and FOG factors are expressed in the limb bud mesenchyme, loss of GATA6 from limb bud mesenchyme results in hindlimb preaxial polydactyly.
(A) Expression of GATA1-6 and FOG1/2 factors relative to GAPDH was assayed by RT-qPCR in anterior (Ant) or posterior (Post) regions of E11.5 mouse forelimbs or hindlimbs. *p<0.05, **p<0.01, ***p<0.001; Student's test. Error bar indicates standard deviation, n = 3. In each case, the transcript abundance (relative to GAPDH) in the anterior forelimb of the first embryo examined (out of 3) was set to equal 1.0; the transcript abundance (relative to GAPDH) in either other regions of this embryo or in other embryos was normalized to this arbitrary set point. (B) Whole mount in situ hybridization analysis of GATA6 expression in E10.5 (a, d), E11.5 (b, e), and E12.5 (c, f) WT forelimbs (a–c) or hindlimbs (d–f) performed with a 3′ UTR mouse GATA6 probe [28]. Arrows point to the location of GATA6 expression. At least 5 embryos of each stage were analyzed and representative samples are shown. (C) Alcian Blue/Alizarin Red staining of the hindlimb of either a P21 control mouse (left) or a Prx1-Cre; GATA6fl/fl mouse (right) is displayed. At least 20 animals of Prx1-Cre; GATA6fl/fl genotype were analyzed and preaxial polydactyly was observed with 100% penetrance in these mutants (representative animal with 7 hindlimb digits is shown in the right panel). Anterior digits are on the left side of each image.
Figure 2.
Loss of GATA6 in the hindlimb bud induces ectopic expression of hedgehog responsive genes.
Whole mount in situ hybridization analysis of gene expression in E11.5 hindlimb buds from either control (A–Q) or Prx1-Cre; GATA6fl/fl (A'–Q') embryos. Ectopic gene expression is indicated by red arrows. At least 4 Prx1-Cre; GATA6fl/fl limbs were analyzed for each in situ probe and representative limbs are shown.
Figure 3.
Transient forced expression of GATA6 throughout the developing limb bud represses induction of Shh, Gli1, and Grem1, and results in loss of digits.
(A) Whole mount in situ hybridization analysis of gene expression in E10.5 forelimb and E11.5 hindlimb buds from either control (a–d, e–h) or Prx1-Cre; ROSA26-rtTA; GATA6-Tg (a'–d', e'–h') transgenic mouse embryos (that had been administered doxycycline beginning at E4.5); at least 6 limbs were analyzed for each probe, representative samples are shown. (B) RT-qPCR analysis of gene expression in E11.5 fore- or hindlimb buds isolated from either a transgenic (Tg) Prx1-Cre; ROSA26-rtTA; GATA6-Tg embryo (that had been administered doxycycline beginning at E4.5) or a littermate control (Cont) embryo. The value (i.e., transcript examined relative to GAPDH) of the 1st sample in each graph (FL control) was set to 1.0; and the other samples normalized to that value for each of the graphs. (C) Alcian Blue/Alizarin Red staining of either P0 control (a, b) or Prx1-Cre; ROSA26-rtTA; GATA6-Tg (a', b') mouse forelimbs (a, a') and hindlimbs (b, b') (that had been exposed to doxycycline from E4.5 until E11.5). At least 10 P0 Prx1-Cre; ROSA26-rtTA; GATA6-Tg animals were analyzed and digit loss was observed with 100% penetrance in both forelimbs and hindlimbs in these mutants; milder mutants are shown.
Figure 4.
Conditional deletion of GATA6 does not significantly alter Gli3 processing in the hindlimb bud.
(A) Western blot analysis of Gli3FL and Gli3R steady state levels in anterior (lanes 1 and 3) or posterior (lanes 2 and 4) halves of E11.5 hindlimb buds isolated from either control (lanes 1 and 2) or Prx1-Cre; GATA6fl/fl (lanes 3 and 4) mouse embryos. α–Tubulin is used as a loading control. (B) Quantification of Gli3FL/Gli3R ratios (by densitometry of Western Blots) in anterior or posterior halves of E11.5 hindlimb buds isolated from either littermate controls (lanes 1 and 2) or Prx1-Cre; GATA6fl/fl (lanes 3 and 4) animals, 12 embryos of each genotype were analyzed. Error bar indicates standard deviation, n = 12.
Figure 5.
GATA6 binds to regulatory regions that control the expression of the mouse Gli1 gene and blocks expression of reporters driven by this sequence. Loss of Gli1 cannot rescue polydactyly induced by conditional deletion of GATA6 in mouse hindlimbs.
(A) GATA6 binding sites identified by EMSA analysis in the promoter of the mouse Gli1 gene. Positions of GATA6 binding sites are indicated relative to the transcription start site of Gli1. Oligos 1–6 used to perform electrophoretic mobility shift essay (EMSA) are indicated. Positions of qPCR primers used in ChIP experiments are shown (P1–P3), P3 is a control pair of primers located ∼6 kb downstream of Gli1 start site. (B) EMSA employing in vitro translated mouse GATA6 with oligos from the Gli1 promoter (displayed in A). Arrows point to the position of either the GATA6-oligo complexes or to the unbound labeled probes. (C) Chromatin IP to assay GATA6 interaction with the Gli1 promoter in limb buds isolated from either E11.5 Prx1-Cre; ROSA26-rtTA; GATA6-Tg animals, that had been pulsed with doxycycline from E4.5 until harvest, versus limb buds isolated from E11.5 Prx1-Cre; GATA6fl/fl mice. Positions of ChIP primers on the Gli1 promoter are displayed in A. Chromatin was sheared to between 500–1000 bp prior to ChIP with anti-GATA6 antibody. *p<0.05; Student's test. Error bar indicates standard deviation, n = 3. (D) GATA6 blocks Gli1-dependent activation of Gli1-luciferase. NIH3T3 cells were co-transfected with a ∼3.6 kb Gli1-firefly luciferase reporter plus expression vehicles encoding either Gli, GATA6, or FOG2, as indicated. Firefly luciferase units were normalized to the expression of co-transfected SV40-renilla luciferase to obtain Relative Luciferase Units (RLU). *p<0.05, **p<0.01, ***p<0.001; Student's test. Error bar indicates standard deviation, n = 3. (E) Alizarin Red/Alcian blue staining of the hindlimbs of P0 mice of the indicated genotypes. At least 5 animals of each genotype were analyzed; representative samples are shown.
Figure 6.
GATA6 binds to regulatory regions that control the expression of the mouse Shh gene and blocks expression of reporters driven by this sequence. Conditional deletion of Shh from the mouse limb bud rescues polydactyly induced by conditional loss of GATA6 in the mouse hindlimbs.
(A) GATA6 binding sites identified by EMSA in the conserved region (ShhE) of the mouse Shh limb bud enhancer ZRS. Position of GATA6 binding sites are indicated relative to the first HindIII site in the ZRS sequence [16]. Oligos 7 and 8 used to perform EMSA are indicated. Positions of qPCR primers used in CHIP experiments are shown (P4–P6), P6 is a control pair of primers located ∼6 kb upstream of the ShhE sequence. (B) Electrophoretic mobility shift assays of in vitro translated GATA6 with oligos from the ShhE region of the ZRS enhancer. Arrows point to the position of either the GATA6-oligo complexes or to the unbound labeled probes. (C) Chromatin IP to assay GATA6 interaction with the conserved region (ShhE) of the mouse Shh limb bud enhancer (ZRS) in limb buds isolated from either E11.5 Prx1-Cre; ROSA26-rtTA; GATA6-Tg animals, that had been pulsed with doxycycline from E4.5 until harvest, versus limb buds isolated from E11.5 Prx1-Cre; GATA6fl/fl mice. Positions of primers on the ShhE enhancer are displayed in A. Chromatin was sheared to between 500–1000 bp prior to ChIP with anti-GATA6 antibody. **p<0.01; Student's test. Error bar indicates standard deviation, n = 3. (D) GATA6 and FOG2 cooperatively block Hand2/Hoxd13-mediated induction of a luciferase reporter driven by the conserved region of the ZRS enhancer (ShhE) that drives Shh expression in the limb bud. NIH3T3 cells were co-transfected with an ShhE-firefly luciferase reporter plus expression vehicles encoding either Hand2, Hoxd13, GATA6-WT, GATA6-DNA binding domain mutant (GATA6DBDmut), or FOG2, as indicated. Firefly luciferase units were normalized to the expression of co-transfected SV40-renilla luciferase to obtain Relative Luciferase Units (RLU). *p<0.05, **p<0.01, ***p<0.001; Student's test. Error bar indicates standard deviation, n = 3. (E) Alizarin Red/Alcian blue staining of the hindlimbs of P0 mice of the indicated genotypes. At least 5 animals of each genotype were analyzed; representative samples are shown. (F) Whole mount in situ hybridization analysis of Gli1 expression in E11.5 Prx1-Cre; GATA6fl/fl; Shhfl/fl embryos vs. E11.5 Prx1-Cre; GATA6fl/fl; Shhfl/+ embryos. Four limbs of each genotype were analyzed; representative images are shown. Ectopic Gli1 expression is noted by the red arrow.
Figure 7.
A model for GATA6 action in the anterior region of developing hindlimb buds.
GATA6 is expressed in the anterior portion of the developing mouse hindlimb bud (displayed in top figures), and specifically represses polydactyly (PD) in the hindlimb by directly blocking anterior expression of Shh (most likely in a FOG-dependent fashion), and thus represses ectopic expression of both direct Shh transcriptional targets (Gli1 and Grem1) in anterior limb bud mesenchyme and indirect targets (FGF4/8) in the anterior AER. In hindlimb buds engineered to lack GATA6 (displayed in bottom figures), ectopic expression of Shh in the anterior mesenchyme induces expression of both Gli2/3 FL and Gli1 which activate expression of the BMP antagonist Grem1, which in turn derepresses expression of FGF4/8 in the anterior region of the apical ectodermal ridge. Thus, as a consequence of deletion of GATA6 in Prx1-Cre; GATA6fl/fl animals, the hindlimbs of these animals display preaxial polydactyly. Expressed genes are indicated by black font; knocked-out or repressed genes are indicated by grey font.