© 1999 Macmillan Magazines Ltd p63 is essential for regenerative proliferation in limb, craniofacial and epithelial development Annie Yang*, Ronen Schweitzer², Deqin Sun³, Mourad Kaghad§, Nancy Walker§, Roderick T. Bronsonk, Cliff Tabin², Arlene Sharpe³¶, Daniel Caput§, Christopher Crum³ & Frank McKeon* Departments of * Cell Biology and ² Genetics, Harvard Medical School, and ¶ Immunology Research Division, ³ Department of Pathology, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts 02115, USA § Sano® Biorecherche, Innopole BP 137, 31676 Labege Cedex, France k US Department of Agriculture, Human Nutrition Research Center on Aging, and Department of Pathology, Tufts University School of Veterinary Medicine, Boston, Massachusetts 02111, USA ......................................................................................................................... The p63 gene, a homologue of the tumour-suppressor p53 (refs 1±5), is highly expressed in the basal or progenitor layers of many epithelial tissues 1 . Here we report that mice homozygous for a disrupted p63 gene have major defects in their limb, craniofacial and epithelial development. p63 is expressed in the ectodermal surfaces of the limb buds, branchial arches and epidermal appendages, which are all sites of reciprocal signalling that direct morphogenetic patterning of the underlying meso- derm. The limb truncations are due to a failure to maintain the apical ectodermal ridge, a strati®ed epithelium, essential for limb development. The embryonic epidermis of p63 -/- mice undergoes an unusual process of non-regenerative differen- tiation, culminating in a striking absence of all squamous epithelia and their derivatives, including mammary, lacrymal and salivary glands. Taken together, our results indicate that p63 is critical for maintaining the progenitor-cell populations that are necessary to sustain epithelial development and morphogenesis. p63 shows remarkable structural similarity to p53 and to the related p73 gene 6 . Unlike p53, however, several messenger RNAs are transcribed from two different promoters of the p63 gene, yielding both transactivating (TA-p63) and non-transactivating (DN-p63) isotypes 1 (Fig. 1a). Embryonic stem cells bearing a p63 allele lacking exons 6, 7 and 8 were produced by homologous recombination 7 and used to generate mice heterozygous for this mutation (Fig. 1b, c). These deleted exons comprise most of the DNA-binding domain that is common to all known p63 isotypes, and are presumably essential for the function of both the TA- and DN-p63 proteins. Immunoblotting of whole-embryo extracts with the 4A4 anti-p63 antibody 1 showed that p63 -/- embryos lacked species of relative molecular mass 80K, likely to be the DN-63a isotype, present in both wild-type and heterozygote specimens (Fig. 1d). p63 -/- mice derived from heterozygous crosses die within a day of birth, owing to desiccation and maternal neglect secondary to severe develop- mental malformations. A comparison of wild-type and mutant newborn mice and late-stage embryos revealed striking craniofacial abnormalities, limb truncations, and a complete absence of an epidermis and related appendages, including hair follicles, vibrissae, and tooth primordia (Fig. 1f±j). Defects in limb morphogenesis in p63 mutants were evident as early as E9.5 and progressed through to E12 (Fig. 2a), a period of limb-bud formation dependent on ectodermal±mesenchymal sig- nalling interactions 8,9 . During this interval, p63 expression was evident in the limb buds, branchial arches and the overall ectoderm (Fig. 2b). In particular, the limb buds showed intense staining in the apical ectodermal ridge (AER), a region of specialized, multilayered ectoderm which is essential for growth and patterning of the underlying mesenchyme 8±10 (Fig. 2b). The ectodermal surfaces of the branchial arches also showed strong expression of p63 RNA. This ectoderm is important in directing the morphogenesis of craniofacial skeletal and soft-tissue structures 11 , both of which are hypoplastic in p63-de®cient mice. In accord with the p63 transcript expression patterns, the 4A4 anti-p63 monoclonal antibody revealed strong nuclear staining of ectodermal cells of E11 embryos, especially those comprising the AER of the limb bud (Fig. 2c) and the maxillary and mandibular branchial ectoderm (results not shown). The high levels of p63 expression in the AER suggested a potential role for p63 in maintaining the structure or function of this specialized ectoderm. We therefore performed serial sectioning of E11 p63 -/- and control embryos to determine the integrity of the AER. Control sections showed the characteristic, strati®ed ecto- derm present along the interface of dorsal and ventral surfaces of the limb bud. Corresponding sections of p63 -/- specimens provided no evidence of a discernible AER structure. Instead, mutant limb buds had a single-layered epithelium at the distal tip, indistinguishable from the surrounding ectoderm (Fig. 2c). Fibroblast growth factor 8 (FGF-8) can recapitulate the morpho- genetic activities of the ectoderm that are necessary for the initiation and maintenance of limb outgrowth, and is a key indicator of the AER's functional integrity 12,13 . We analysed FGF-8 expression in p63-mutant embryos by whole mount in situ hybridization. Sig- ni®cantly, a weak FGF-8 signal was detected on the ventral surface of p63 -/- forelimb buds at E9.5, before the formation of an AER. From E9.5 onwards, however, when the distal ventral ectoderm thickens and begins the strati®cation process that ultimately gives rise to an AER 14 , FGF-8 expression was progressively reduced in the p63- mutant limb bud compared with wild-type littermates (Fig. 2d). FGFs secreted by AER cells induce limb outgrowth by maintaining the underlying mesenchyme, known as the progress zone, in an undifferentiated, highly proliferative state 9 . Expression of progress- zone markers, such as Msx1, thus serves as an indicator for AER activity 15,16 . AER activity was indeed reduced in the p63-mutant embryo, as indicated by the reduced Msx1 expression in the progress zone (Fig. 2e). There was a remnant of FGF-8 expression in mutant limb buds that was con®ned to the ventral surface, even at later stages when differential growth of the dorsal and ventral ectoderm normally results in a shift of FGF-8 to the distal tip 14 . One possibility, then, for the failure of AER formation in p63 -/- mice was a disruption of dorsal/ventral (D±V) patterning in the limb bud. We therefore examined the expression of Lmx1b (ref. 17), a dorsal mesenchyme marker, and of Wnt7a (refs 18, 19), a dorsal ectoderm marker, in mutant limb buds. Both Lmx1b and Wnt7a showed a more ventral spatial distribution than is seen in the wild type (Fig. 2f, and data not shown). A de®ned, albeit ventrally displaced D±V border was present in mutant limb buds, as judged by the absence of Lmx1b and Wnt7a in the ventral domain. Together, these results indicate that D±V patterning itself was not perturbed in the mutant embryos. Instead, they suggest that the lack of a proper AER in p63 -/- limb buds results from a failure of the ectoderm to undergo the growth and differentiation that give rise to this strati®ed epithelium. This inability to maintain the structural integrity of the AER, in turn, probably underlies the defects in limb morphogenesis observed in p63 -/- mice. p63 is strongly expressed in the basal, or progenitor, cells of several epithelial structures present in the epidermis, cervix, uro- genital tract, prostate and breast 1 . The importance of p63 function in epithelial differentiation was revealed by the remarkable absence of skin and related appendages in late embryonic and newborn p63- de®cient mice (Fig. 1). Histological analysis con®rmed the absence of a strati®ed epidermis in p63 -/- mice, as they lack the characteristic structure of basal, suprabasal and corni®ed layers, and associated hair follicles (Fig. 3a). In addition, tissues derived from the same stem cells that give rise to the epidermis, including mammary, sebaceous, lacrymal and salivary glands, are also absent in the p63 -/- letters to nature 714 NATURE | VOL 398 | 22 APRIL 1999 | www.nature.com