Role of Epiprofin, a zinc-finger transcription factor, in limb development

Abstract

The formation and maintenance of the apical ectodermal ridge (AER) is critical for the outgrowth and patterning of the vertebrate limb. In the present work, we have investigated the role of Epiprofin (Epfn/Sp6), a member of the SP/KLF transcription factor family that is expressed in the limb ectoderm and the AER, during limb development. Epfn mutant mice have a defective autopod that shows mesoaxial syndactyly in the forelimb and synostosis (bony fusion) in the hindlimb and partial bidorsal digital tips. Epfn mutants also show a defect in the maturation of the AER that appears flat and broad, with a double ridge phenotype. By genetic analysis, we also show that Epfn is controlled by WNT/b-CATENIN signaling in the limb ectoderm. Since the less severe phenotypes of the conditional removal of b-catenin in the limb ectoderm strongly resemble the limb phenotype of Epfn mutants, we propose that EPFN very likely functions as a modulator of WNT signaling in the limb ectoderm.

Figure 1. Expression pattern of Epiprofin during mouse limb bud development

Whole-mount in situ hybridization was performed on wild-type embryos of the indicated age with an Epfn antisense riboprobe. (A–D) At E9 and E9.5 Epfn is expressed in the presumptive fore- and hindlimb ectoderm as shown in the whole mounts and corresponding sections at the forelimb level as marked in the corresponding whole mount (B, D). At E10.5 (E) and E11.5 (F) Epfn expression is predominantly confined to the AER. At E12.5 (G) Epfn expression continues in the AER and extends into the dorsal and ventral ectoderm particularly at digital tips. At 14.5 (H) Epfn expression still occurs in the ectoderm of the digital tips. (I, J) show two consecutive serial sections (6 mm apart) through the presumptive forelimb of an E9 mouse embryo, showing Epfn expression in the limb field ectoderm (I), while Fgf8 expression has not yet been activated (J) at this stage.

Figure 2. Limb phenotype of Epiprofin mutant embryos

(A–H) External morphology (A–C) and skeletal pattern (E–H) of newborn control and mutant fore- and hindlimbs, as indicated. Note the soft tissue syndactyly in the forelimb (arrowhead in B) and the oligodactyly of four digits in the hindlimb (D and H). (I–L) Morphology of the digital tip of 6-week-old control (I) and mutant (K–L) limbs. (J) is a skeletal preparation of the broader distal tip of the ungeal phalanx in a newborn. The numbers in some panels indicate the identity of the digits.

Figure 3. Abnormal AER morphology and Fgf8 expression in the Epfn mutant limb

(A–D) Histological sections showing the AER morphology in a E11.5 wild type forelimb (A) and mutant fore- (B) and hindlimb (C–D). The mutant AER, delimited by arrowheads in (B) and (C), is expanded in the DV axis, lacks the typical elevation and sometimes protrudes into the mesenchyme (arrowheads in D). (E–L) are whole mount in situ hybridizations showing the expression pattern of Fgf8 in control and mutant limbs, as indicated at the top of each panel. The inserts show longitudinal sections at the level indicated in E and F.

Figure 4. Epiprofin and WNT/b-CATENIN signaling

(A–C) Whole mount in situ hybridization showing the expression pattern of Epfn in the hindlimb of wild type (A), gain-of-function mutant of b-catenin (Brn4Cre;DN-b-catenin) (B), and loss-of-function mutant of b-catenin (Brn4Cre;b -catenin ^flox/flox). The dashed line in (C) marks the contour of the mutant hindlimb that does not express Epfn. (D–E) Dkk1 pattern of expression is shown at E11 (top row) and E12 (bottom row) in WT and mutant limbs. Dkk1 expression is increased in the mutant distal hindlimb mesoderm as can be clearly appreciated in the longitudinal sections shown in the corresponding inserts. (F) The pattern of Lef1 expression is similar in normal and mutant limbs. (G) Axin2 expression was analyzed in compound mice expressing lacZ under the control of the Axin2 promoter. Axin2 expression is reduced in the mutant ectoderm and AER (arrows) compared with the control. Anterior views of the limbs are presented.

Figure 5. Epiprofin expression in the limb ectoderm is independent of FGF signaling

(A–D) Whole mount in situ hybridization for Epfn in E10 wild type and Msx2Cre;Fgf4;Fgf8 double mutants, as indicated. (E–F) Whole mount in situ hybridization for Epfn in an E9 Fgf10 mutant, showing normal expression in the limb ectoderm. (F) is a magnification of (E).

Figure 6. Dorsoventral patterning defects in Epiprofin mutant limbs

At E11.5, the expression domain of Lmx1b in the mutant hindlimb is properly restricted to the dorsal mesoderm (B–C), as in control limbs (A). At later stages (E14.5), occasional patches of ectopic Lmx1b expression, indicated by arrows, are detected in mutant (E–F), but not in wild type limbs (D). At E11.5, wild type and mutant limbs show similar expression pattern of Wnt7a in the dorsal ectoderm (G–H). (A) and (B) are lateral views, (C–F) show views from the tip of the limb, and (G–H) are longitudinal sections. The numbers in some panels indicate the identity of the digits.

Figure 7. Bmp4 expression and signaling in Epiprofin mutants

At E10.5, the expression of Bmp4 in the mutant fore and hindlimb is slightly downregulated (B) compared to wild type (A). At E11.5, this downregulation is clearer (C–D) shown in ventral views at the top and in anterior views at the bottom of the panel, as indicated. At E12.5, Bmp4 expression in the ectoderm appears patchy particularly in the hindlimbs shown in anterior and distal views (E, F). At E12.5, Msx2 expression is also downregulated in mutant limbs, confirming the reduction in BMP signaling at E12.5 (G–H).