ΔNp63 knockout mice reveal its indispensable role as a master regulator of epithelial development and differentiation

Abstract

The transcription factor p63 is important in the development of the skin as p63-null mice exhibit striking defects in embryonic epidermal morphogenesis. Understanding the mechanisms that underlie this phenotype is complicated by the existence of multiple p63 isoforms, including TAp63 and ΔNp63. To investigate the role of ΔNp63 in epidermal morphogenesis we generated ΔNp63 knock-in mice in which the ΔNp63-specific exon is replaced by GFP. Homozygous ΔNp63(gfp/gfp) animals exhibit severe developmental anomalies including truncated forelimbs and the absence of hind limbs, largely phenocopying existing knockouts in which all p63 isoforms are deleted. ΔNp63-null animals show a poorly developed stratified epidermis comprising isolated clusters of disorganized epithelial cells. Despite the failure to develop a mature stratified epidermis, the patches of ΔNp63-null keratinocytes are able to stratify and undergo a program of terminal differentiation. However, we observe premature expression of markers associated with terminal differentiation, which is unique to ΔNp63-null animals and not evident in the skin of mice lacking all p63 isoforms. We posit that the dysregulated and accelerated keratinocyte differentiation phenotype is driven by significant alterations in the expression of key components of the Notch signaling pathway, some of which are direct transcriptional targets of ΔNp63 as demonstrated by ChIP experiments. The analysis of ΔNp63(gfp/gfp) knockout mice reaffirms the indispensable role of the ΔN isoform of p63 in epithelial biology and confirms that ΔNp63-null keratinocytes are capable of committing to an epidermal cell lineage, but are likely to suffer from diminished renewal capacity and an altered differentiation fate.

Fig. 1.

Gross morphology and histological analysis of mice with a targeted deletion of ΔNp63. Matings between ΔNp63^gfp/+ animals generated wild-type and heterozygous mice that are phenotypically normal, whereas ΔNp63^gfp/gfp animals display truncated limbs and the absence of a mature stratified skin epidermis and appendages at all time points examined. Other affected epithelial structures include the oral cavity. The dashed line demarcates the boundary between the epidermis and the dermis. Scale bars: 75 μm.

Fig. 2.

Dorsal and ventral skin sections of E13.5-18.5 wild-type and ΔNp63^gfp/gfp animals stained for K5 and K8. Antibody staining for K5 (red) and K8 (green) reveals that epidermal keratinocytes in the ΔNp63^gfp/gfp mice express the simple epithelial marker K8, which is absent in the wild-type animals (compare right and left panels). Some ΔNp63^gfp/gfp keratinocytes co-express K5 and K8 (middle panel, yellow staining). Scale bars: 37.5 μm.

Fig. 3.

Accelerated keratinocyte differentiation in the absence of ΔNp63. Dorsal skin sections of E15.5 wild-type and ΔNp63^gfp/gfp mice stained for various markers associated with the keratinocyte stratification (K5) and differentiation (K10) program reveal that ΔNp63-null keratinocytes are able to progress through each program. Examination of markers associated with late differentiation (Lor, loricrin; Fil, filaggrin; Inv, involucrin) reveal that ΔNp63-null keratinocytes undergo premature terminal differentiation as compared with wild-type keratinocytes. Nuclei are stained with DAPI. Scale bars: 37.5 μm.

Fig. 4.

Loss of key components of the extracellular matrix in the epidermis of ΔNp63-null mice. Dorsal skin sections from E15.5 ΔNp63^gfp/gfp mice reveal a dramatic reduction in several crucial basement membrane proteins, including laminin α5 (Lamin5) and collagen IV (COLIV), with modest to moderate changes in E-cadherin (E-Cad), HSPG, nidogen and integrin β1 (β1) expression. Sections are co-stained for K5 as indicated and nuclei are counterstained with DAPI. Scale bars: 37.5 μm.

Fig. 5.

Loss of ΔNp63 affects the expression of hemidesmosome and desmosome junction proteins. Immunofluorescence staining of dorsal and ventral skin sections at E15.5 reveals reduced expression of the hemidesmosomal components integrin α6 and β4 in ΔNp63^gfp/gfp mice. Alterations to various desmosome junction proteins are also seen. All sections are co-stained for K5 and counterstained with DAPI. Dsg1, desmoglein 1; Dsg2, desmoglein 2; Pkp2, plakophilin 2. Scale bars: 37.5 μm.

Fig. 6.

Effects of ΔNp63 loss on components of the Notch signaling pathway. Dorsal skin sections at E15.5 reveal increased expression levels of Notch2 and Notch3, whereas Hes1 and Notch1 appear modestly reduced, in ΔNp63^gfp/gfp compared with wild-type mice. Insets show higher magnification views. Sections are co-stained for K5 and counterstained with DAPI. Scale bars: 37.5 μm.

Fig. 7.

Binding of ΔNp63 to the Notch3 gene. (A) Position of the putative p63 binding site in the human NOTCH3 gene as previously identified by ChIP-seq analysis of human primary keratinocytes using the 4A4 and H129 anti-p63 antibodies (black boxes). These sequences correlate with regions of the genome that are enriched for regulatory markers in NHEK cells, including DNase hypersensitive sites. The p63-ChIP sequence regions are also associated with enhancers (H3K4me1 and H3K4me2) and open chromatin structures (H3K9ac and H3K27ac) (red dashed box). H3K4me1, monomethylated histone H3 lysine 4; H3K4me2, dimethylated histone H3 lysine 4; H3K9ac, acetylated histone H3 lysine 9; H3K27ac, acetylated histone H3 lysine 27. Arrow indicates the transcription start site. (B) Genome Vista alignment of a ∼10 kb segment of the Notch3 gene showing sequence conservation between human and mouse. Black arrow indicates the conserved DNA segment containing the p63 binding site. Peaks represent percent conservation. Blue boxes represent exons. Light-blue shaded box represents the 5′ UTR. (C) ChIP-qPCR results using the RR14 and H129 antibodies in mouse keratinocytes confirms specific binding of p63 to this region and not to the IgG control after normalization to a control genomic locus. Data are represented as mean ± s.d. ^*P<0.001, Student’s t-test.