p63 in skin development and ectodermal dysplasias

Abstract

The transcription factor p63 is critically important for skin development and maintenance. Processes that require p63 include epidermal lineage commitment, epidermal differentiation, cell adhesion, and basement membrane formation. Not surprisingly, alterations in the p63 pathway underlie a subset of ectodermal dysplasias, developmental syndromes in which the skin and skin appendages do not develop normally. This review summarizes the current understanding of the role of p63 in normal development and ectodermal dysplasias.

Introduction

The epidermis, the outermost component of the skin, functions as the primary barrier between the organism and the environment. As such, the epidermis protects the organism from microbial, physical, and chemical assaults as well as from excessive water loss. The barrier function of the epidermis is established during embryogenesis and is maintained postnatally by stem cells, which are located in the basal layer of the interfollicular epidermis (Fuchs, 2009). When interfollicular stem cells divide, they give rise to daughter cells, termed transit amplifying (TA) cells. After a few rounds of cell division, TA cells permanently withdraw from the cell cycle, and move suprabasally to initiate terminal differentiation (Koster and Roop, 2007). This initial stage of terminal differentiation results in the formation of the spinous layer, the first suprabasal cell layer of the epidermis. Spinous keratinocytes subsequently differentiate into granular keratinocytes. Granular keratinocytes ultimately die when they form the stratum corneum, the outermost dead layer of the epidermis. The cells of the stratum corneum, termed corneocytes, replace their plasma membrane with a shell of cross-linked proteins. Together with extracellular lipids, these cross-linked proteins form the cornified cell envelope, the most important component of the epidermal water barrier (Rice and Green, 1977;Steven and Steinert, 1994). As described in more detail below, the transcription factor p63 is required for key events in epidermal development and differentiation, including epidermal lineage commitment, keratinocyte adhesion, basement membrane formation, epidermal differentiation, and barrier formation. Thus, it comes as no surprise that alterations in the p63 pathway lead to developmental disorders in which these processes are affected.

p63 in Developmental Disorders

Developmental disorders caused by alterations in the p63 pathway

As predicted from its critical function in the epidermis, abnormalities in the p63 pathway have been linked to several developmental disorders (Figure 1). First, p63 is mutated in several ectodermal dysplasias, a large group of developmental disorders in which ectodermal derivatives, such as the epidermis and its appendages, fail to develop normally (Rinne et al., 2007). Mutations in p63 were found to underlie several of these ectodermal dysplasias, including ectrodactyly, ectodermal dysplasia and cleft lip (EEC) (Celli et al., 1999) and ankyloblepharon ectodermal dysplasia and clefting (AEC or Hay-Wells syndrome) (McGrath et al., 2001). Interestingly, even though both of these disorders are caused by mutations in p63, the patients have distinct abnormalities. For example, severe skin erosions are common in AEC patients, whereas they are rare in EEC patients (Julapalli et al., 2009;Brunner et al., 2002). Conversely, limb abnormalities in EEC patients are generally severe, whereas they are relatively mild in AEC patients (Bree, 2009;Sutton et al., 2009;Brunner et al., 2002). These differences in phenotypic outcome are a reflection of the type of mutations found in EEC and AEC patients. Specifically, mutations in different functional domains of the p63 protein lead to different cellular defects, and thus to a different disease. In order to understand the etiology of these different but related diseases, we need to understand the functions of the different p63 proteins and protein domains.

Figure 1. p63 in developmental disorders.

Mutations in p63 or p63 target genes cause developmental disorders. Dominant mutations in p63 underlie AEC, EEC, and SHFM4, whereas dominant mutations in the p63 target genes P-cadherin, Dlx3, and Dlx5/6 underlie EEM, TDO, and SHFM1, respectively. Common features of EEC, EEM, and SHFM include severe limb abnormalities, including syndactyly and ectrodactyly (right image; image depicts an EEC patient). A characteristic feature of AEC patients is the presence of severe erosions, often located to the scalp (left image). Other ectodermal dysplasias caused by mutations in p63 include limb-mammary syndrome (LMS) and acro-dermato-ungual-lacrimal-tooth syndrome (ADULT). Images of AEC and EEC patients were provided by the National Foundation for Ectodermal Dysplasias (NFED).

By encoding two different N-termini (TA and ΔN) and multiple C-termini (α, β, γ, δ, and ε), the p63 gene generates multiple protein isoforms (Yang et al., 1998;Mangiulli et al., 2009). All p63 isoforms contain identical DNA binding and oligomerization domains. Whereas mutations in the ΔN N-terminus or the α C-terminus cause AEC, mutations in the DNA binding domain cause EEC (Figure 2) (Rinne et al., 2008;McGrath et al., 2001;Celli et al., 1999). Nevertheless, the precise molecular pathway alterations that are triggered by these mutations are not yet understood.

Figure 2. Mutations in different domains of p63 cause different ectodermal dysplasias.

The p63 gene is expressed as multiple isoforms. Isoforms depicted here are the full-length isoform, TAp63α, and the predominantly expressed p63 isoform in late embryonic and postnatal epidermis, ΔNp63α. Mutations in the DNA binding domain, common to all p63 isoforms, can cause EEC or SHFM4. Mutations in the ΔN N-terminus or the SAM domain (only present in the α C-terminus) underlie AEC. For an overview of p63 mutations found in ectodermal dysplasia patients, please see these references (Rinne et al., 2007;Rinne et al., 2008). TA: transactivation domain, oligo: oligomerization domain, SAM: sterile alpha motif.

Another developmental disorder caused by p63 mutations is Split Hand Foot Malformation (SHFM4) (Ianakiev et al., 2000). SHFM is a non-syndromic developmental disorder, in which patients have severe limb abnormalities (ectrodactyly and syndactyly), but no other developmental defects (Duijf et al., 2003). Five independent chromosomal loci have been associated with SHFM (SHFM1-SHFM5). p63 was mapped to one of these loci, SHFM4, and mutations in p63 were found to underlie SHFM associated with this locus (Figures 1 and 2). Even though causative mutations have not been identified in the remaining four loci, candidate genes for SHFM have been mapped to these loci. For example, the SHFM1 locus harbors the DLX5 and DLX6 genes (Scherer et al., 1994). Interestingly, the limb phenotype of mice that lack both Dlx5 and Dlx6 phenocopies that of SHFM patients (Merlo et al., 2002;Robledo et al., 2002). Further, p63 directly induces Dlx5 and Dlx6 expression, suggesting that SHFM1 and SHFM4 are caused by alterations in the same genetic pathway (Lo Iacono et al., 2008).

In addition to Dlx5 and Dlx6, mutations in other p63 target genes were also found to underlie several developmental disorders. For example, mutations in the p63 target gene P-cadherin cause ectodermal dysplasia, ectrodactyly, and macular dystrophy (EEM) (Shimomura et al., 2008;Kjaer et al., 2005). Limb abnormalities in patients with EEM show a remarkable similarity to those in patients with EEC (Figure 1). Further, both EEM and EEC patients have hair abnormalities. Because P-cadherin was previously found to be involved in hair follicle morphogenesis, these findings suggest that P-cadherin is an important mediator of p63 function during limb and hair development (Jamora et al., 2003). Another target gene of p63 that has been linked to a developmental disorder is Dlx3, mutations in which cause Tricho-Dento-Osseus (TDO) syndrome (Radoja et al., 2007;Price et al., 1998). In TDO syndrome, the mutant Dlx3 proteins have a dominant-negative function towards the wild type Dlx3 proteins, thus impairing the function of Dlx3 (Duverger et al., 2008). Interestingly, although p63 induces Dlx3 expression in basal keratinocytes, Dlx3 degrades p63 in suprabasal cell layers, suggesting that these proteins participate in an intricate feedback loop (Di Costanzo et al., 2009). Taken together, these data indicate that mutations in p63 itself and in genes transcriptionally regulated by p63 can lead to developmental disorders with similar phenotypes

Function of mutant p63 proteins

All disorders caused by p63 mutations identified to date are inherited in an autosomal dominant fashion. However, it remains unclear how the mutant p63 proteins exert their dominant function. Because loss of one p63 allele does not cause ectodermal dysplasia in mice or humans, it appears that haplo-insufficiency is not the underlying disease mechanism (van Bokhoven et al., 2001;Mills et al., 1999;Yang et al., 1999). Recent studies have provided some insight into the differential regulation of target genes by mutant p63 proteins expressed in patients. For example, whereas p63 proteins carrying AEC mutations can activate the Dlx5 and Dlx6 promoters, p63 proteins carrying EEC or SHFM mutations cannot, potentially providing an explanation for the lack of limb abnormalities in patients with AEC (Lo Iacono et al., 2008). However, these studies did not determine whether mutant p63 proteins affect the function of wild type p63 proteins when they are co-expressed, as they are in ectodermal dysplasia patients. Thus, the function of mutant p63 proteins in the cellular context they are normally expressed in is largely unknown. Further, the molecular basis for the differential target gene activation by different mutant p63 proteins remains unclear. However, it is likely that differences in DNA binding characteristics contribute to the differential activation of p63 target genes by different mutant p63 proteins. In fact, even though mutations in the p63 DNA binding domain only occur in SHFM and EEC, alterations in DNA binding characteristics are also predicted for some AEC mutants due to conformational changes of the mutant p63 proteins (McGrath et al., 2001). Another possible explanation for the differences in target gene activation may be the differential interaction of mutant p63 proteins with co-factors required for target gene activation or repression.

Mouse models

To further understand the role of p63 in developmental disorders, animal models which mimic these disorders are imperative. A mouse model that mimics the skin fragility phenotype observed in AEC patients was recently generated by downregulating ΔNp63 in the epidermis (Koster et al., 2007). Both in mice with reduced epidermal ΔNp63 expression and in AEC patients, skin erosions were characterized by suprabasal epidermal proliferation, delayed terminal differentiation, and basement membrane abnormalities (Koster et al., 2009). Together, these abnormalities likely contribute to the skin fragility observed in AEC patients. Consistent with the finding that skin fragility of AEC patients can be mimicked by downregulating ΔNp63, mutant ΔNp63α proteins expressed in AEC patients function as dominant-negative molecules, thereby preventing the induction of p63 target genes. Whether the additional developmental defects in AEC patients are also caused by a dominant-negative role of the mutant p63 proteins, or whether mutant p63 proteins also possess a gain-of-function role, remains to be determined. Since the skin phenotypes of mice with reduced ΔNp63 expression and AEC patients are indistinguishable, this mouse model will be a valuable tool for testing therapeutic approaches aimed at treating skin erosions in AEC patients.

p63 in epidermal development and maintenance

The human developmental disorders caused by alterations in the p63 pathway clearly demonstrate the importance of p63 in skin and appendage development. To dissect the genetic pathways controlled by p63, in vitro and in vivo models have been employed (see below). These models have elucidated critical roles for p63 in various key steps required for epidermal development and homeostasis (Figure 3).

Figure 3. Role of p63 in the epidermis.

The epidermis is a stratified epithelium, which consists of several layers of keratinocytes (indicated on left). When interfollicular stem cells, which reside in the basal layer, divide, they give rise to transit amplifying cells which constitute most of the basal layer. When transit amplifying cells initiate terminal differentiation, they withdraw from the cell cycle and move suprabasally, forming the spinous layer. Further differentiation results in the formation of the granular and cornified cell layers. p63 is involved in various processes required for epidermal development, differentiation, and homeostasis (indicated on right). Genes involved in these processes are indicated in the text.

Establishment of the epidermal fate

p63 was initially identified based on its sequence homology with the tumor suppressor gene p53 (Yang et al., 1998). However, unlike p53, p63 is expressed in a tissue-specific manner, and is primarily expressed in stratified epithelia, such as the epidermis. Further, whereas p53-deficient mice are born without major abnormalities, ablating p63 from the germline of mice led to striking developmental abnormalities. Mice lacking p63 were born with a shiny translucent skin, and without any appendages, such as teeth, hair follicles, and mammary glands (Mills et al., 1999;Yang et al., 1999). The severe skin phenotype observed in p63-deficient mice was found to be caused by a complete failure of epidermal commitment, the process in which the surface ectoderm adopts an epidermal fate (Koster et al., 2004;De Rosa et al., 2009). During normal epidermal development, commitment to the epidermal lineage involves the repression of the non-epidermal keratin pair K8/K18 and the induction of the epidermal keratin pair K5/K14 (Jackson et al., 1981;Byrne et al., 1994). Interestingly, p63 is required both for the suppression of K8/K18 and for the induction of K5/K14. This was initially illustrated by the observation that K8 and K18 are aberrantly expressed in the surface epithelium of mice with a germline deletion of p63 as well as in reconstructed human epidermis with decreased p63 expression levels (Koster et al., 2004;Truong et al., 2006;De Rosa et al., 2009). Conversely, ectopic expression of p63 in cultured cells, single-layered epithelia, or embryonic stem cells led to the expression of K5 and K14 (Medawar et al., 2008;Koster et al., 2004;Romano et al., 2009;Aberdam et al., 2008). In addition to controlling the expression of keratins, p63 also directly represses two cell cycle inhibitors Ink4a and Arf (Su et al., 2009b). This repression is also critical for epidermal commitment as demonstrated by the finding that epidermal lineage commitment, indicated by the induction K5 and K14 expression, in p63-deficient mice can be rescued by simultaneously ablating either Ink4a or Arf (Su et al., 2009b). However, even though epidermal commitment is restored in these rescued mice, the epidermis does not mature normally as demonstrated by the heterogeneous expression of differentiation markers and the apparent absence of a stratum corneum (Su et al., 2009b). Thus, epidermal differentiation requires the activation of additional genetic pathways that are not downstream of Ink4a and Arf.

In addition to the failure to develop an epidermis, p63-deficient mice also fail to develop appendages. Because appendage development requires extensive epithelial-mesenchymal interactions, the defects in appendage development in p63-deficient mice are likely to be, in part, secondary to the failure to develop an epidermis. However, p63 also has a direct role in appendage development by inducing genes that are required for the development of different appendages, such as Dlx5/6 and P-cadherin (Lo Iacono et al., 2008;Shimomura et al., 2008). This direct role of p63 in appendage development may provide an explanation for the existence of ectodermal dysplasias caused by p63 mutations in which appendages, but not the epidermis, are affected. In this subset of ectodermal dysplasias, mutant p63 proteins might still be able to activate p63 target genes required for normal development of the interfollicular epidermis, but they may fail to activate genes required for appendage development.

p63 in basal keratinocytes

In addition to its role in epidermal lineage commitment, p63 is also critical for epidermal differentiation and barrier formation, both during epidermal morphogenesis and postnatally (Figure 3). Within late embryonic and postnatal epidermis, the highest levels of p63 expression are observed in the basal layer, where it is predominantly expressed as the ΔNp63α isoform (Yang et al., 1998;Liefer et al., 2000). The basal layer is generally thought of as a homogeneous population of cells. However, it is actually composed of cells that are progressing along the differentiation pathway, including interfollicular stem cells, young TA cells, and mature TA cells. Whereas young TA cells are regularly cycling, mature TA cells have a limited proliferative capacity and are ready to embark on the terminal differentiation program (Lehrer et al., 1998). p63 may have very different functions in the various cell types within the basal layer. In fact, this would explain apparent contradictory results that have been obtained regarding the role of p63 in proliferation of basal keratinocytes. For example, p63 was found to maintain proliferation of basal keratinocytes, in part by repressing various cell cycle inhibitors, including p21, 14-3-3σ, Ink4a, and Arf, as well as by inducing the expression of genes required for cell cycle progression, including ADA and FASN (Truong et al., 2006;Lefkimmiatis et al., 2009;Sbisa et al., 2006;D’Erchia et al., 2006;Westfall et al., 2003;Su et al., 2009b). In apparent contrast, downregulating p63 in basal keratinocytes of mouse epidermis led to increased proliferation, demonstrating that p63 is also required for cell cycle exit (Koster et al., 2007). This induction of cell cycle exit is mediated, in part, by the direct induction of p57^Kip2, a cyclin-dependent kinase inhibitor that is induced when keratinocytes undergo terminal differentiation (Beretta et al., 2005;Martinez et al., 1999). Further, p63 directly represses the expression of genes required for cell cycle progression including cyclin B2 and cdc2 (Testoni and Mantovani, 2006). This apparent controversy is most easily explained by postulating that p63 maintains proliferation in early TA cells, whereas it induces cell cyle exit in mature TA cells. This difference in p63 function may be caused by different expression levels and/or interaction with co-factors that regulate p63 function.

p63 in differentiation

Whereas ΔNp63α expression is high in basal keratinocytes, its expression is reduced to approximately 25% in suprabasal keratinocytes (King et al., 2006). The rapid downregulation of ΔNp63α in suprabasal keratinocytes appears to be mediated by several processes. First, ΔNp63α transcripts are degraded by microRNA-203 (Yi et al., 2008;Lena et al., 2008). As predicted, microRNA-203 is specifically expressed in suprabasal keratinocytes whereas it is virtually absent from basal keratinocytes (Sonkoly et al., 2007;Yi et al., 2008;Lena et al., 2008). Depletion of microRNA-203 was found to result in suprabasal proliferation of keratinocytes (Yi et al., 2008). However, whether the increased proliferation was due to the observed increase in p63 expression or due to ectopic expression of other microRNA-203 target genes was not investigated. In addition to ΔNp63α transcript degradation, ΔNp63α protein is also actively degraded in suprabasal keratinocytes. This is mediated by several processes including the targeting of ΔNp63α for degradation through the proteosomal pathways by the E3 ubiquitin ligase Itch and p14^Arf (Rossi et al., 2006;Vivo et al., 2009). Finally, the p63 target gene Dlx3 promotes ΔNp63α degradation in suprabasal keratinocytes (Di Costanzo et al., 2009).

Together, the above-described mechanisms are responsible for the rapid functional inactivation of p63 in suprabasal keratinocytes. However, despite this active degradation of ΔNp63α in suprabasal cell layers, the remaining ΔNp63α protein is sufficient to control important aspects of keratinocyte differentiation. For example, ΔNp63α directly induces IKKα, a critical mediator of cell cycle exit during keratinocyte differentiation (Marinari et al., 2008;Koster et al., 2007). Further, ΔNp63α induces expression of the terminal differentiation marker K1 (Ogawa et al., 2008;Nguyen et al., 2006). Finally, ΔNp63α is also important for the formation of the epidermal barrier by inducing at least two genes that are required for barrier formation, Claudin 1 and Alox12 (Lopardo et al., 2008;Kim et al., 2009). Interestingly, despite the ability of ΔNp63α to induce genes required for terminal differentiation and barrier formation, it does not induce these genes in the basal layer. The most likely explanation for this observation is that co-factors required for the induction of terminal differentiation genes are absent from basal keratinocytes.

p63 in cell-cell adhesion

In addition to regulating epidermal development and differentiation, p63 is also important for cell-cell adhesion within the epidermis. A general role for p63 in mediating cell-cell adhesion was identified by downregulating p63 in cultured mammary cells (Carroll et al., 2006). Downregulating p63 in this system led to impaired cell adhesion, as well as to reduced expression of several components of desmosomes, the multi-protein complexes that connect keratinocytes (Cheng and Koch, 2004;Carroll et al., 2006). In addition, p63 was found to directly regulate the expression of the desmosomal component Perp (Ihrie et al., 2005). In the absence of Perp expression, mice display severe intra-epidermal blistering further underscoring the importance of Perp in cell adhesion. Consistent with these findings, cell adhesion defects have also been reported in AEC patients, a subset of who display aberrant Perp expression (Payne et al., 2005;Beaudry et al., 2009).

p63 in basement membrane formation and cell-basement membrane adhesion

In addition to its role in mediating cell-cell adhesion, p63 is also important for regulating the adhesion of keratinocytes to the basement membrane. Adhesion of keratinocytes to the basement membrane is mediated by integrins, a family of transmembrane receptors for the basement membrane protein laminin (Larsen et al., 2006). p63 was found to induce expression of several of these integrin subunits, including integrin α3 (Kurata et al., 2004;Carroll et al., 2006). Further, p63 is also important for the formation of the basement membrane, as demonstrated by the basement membrane defects in mice with reduced ΔNp63 expression as well as in AEC patients (Koster et al., 2009;Koster et al., 2007). p63 controls basement membrane formation, at least in part, by directly inducing the expression of the basement membrane component Fras1 (Koster et al., 2007). Interestingly, in mice and humans, Fras1 deficiency causes a severe embryonic blistering phenotype, demonstrating the importance for Fras1 in basement membrane integrity (Vrontou et al., 2003;McGregor et al., 2003). Further, in prostate, the basement membrane associated with p63 expressing cells is thicker and more uniform than the basement membrane associated with basal cells that do not express p63 (Liu et al., 2009). Together, these findings suggest a critical role for p63 in forming and maintaining the basement membrane.

p63 in stem cells

In addition to TA cells and differentiating keratinocytes, p63 is also expressed in stem cells of several tissues, including the skin. However, a role for p63 in these stem cells has remained contentious. At least three different stem cell populations exist in the skin: interfollicular stem cells, bulge stem cells, and skin-derived precursors (SKPs). Of these, the interfollicular stem cells and the bulge stem cells regenerate the epidermis and hair follicle, respectively. Although either of these stem cell types can contribute to the epidermal or hair follicle lineage in response to injury, they only contribute to their own lineage under homeostatic conditions (Ito et al., 2005;Claudinot et al., 2005). A functional role for p63 in interfollicular and bulge stem cells was first proposed based on the phenotype of p63-deficient mice. Some investigators attributed the absence of epidermal morphogenesis in these mice to a premature depletion of stem cells (Yang et al., 1999). Further, colony forming assays using keratinocytes with reduced p63 expression levels, resulted in the formation of smaller colonies, which has been interpreted as evidence for stem cell depletion (Senoo et al., 2007). However, whether this apparent lack of proliferation affected stem cells and/or TA cells was not investigated.

Recently, it was demonstrated that TAp63 isoforms are also expressed in skin-derived precursors (SKPs), dermal precursor cells that reside in the dermal sheath and dermal papilla (Fernandes et al., 2004;Su et al., 2009a). Interestingly, using a conditional TAp63 knockout mouse model, it was found that ablation of TAp63 from both the dermal and epidermal compartments led to skin blistering and accelerated aging (Su et al., 2009a). However, ablation of TAp63 specifically in the epidermis did not recapitulate these phenotypes, suggesting that TAp63 exerts its role in SKPs and not in the epidermis (Su et al., 2009a). However, because ablation of TAp63 specifically from SKPs is currently not technically feasible, the ultimate proof for a role of TAp63 in SKPs has yet to be generated.

Summary

The involvement of abnormalities in the p63 pathway in a subset of developmental disorders underscores the importance of p63 for normal skin and appendage development. Indeed, animal models and cell culture studies have identified roles for p63 in various processes, including epidermal lineage commitment, epidermal differentiation, cell adhesion, and basement membrane formation. However, the function of mutant p63 proteins expressed in ectodermal dysplasia remains poorly understood. Whereas mutant p63 proteins in some ectodermal dysplasias, such as AEC, cause skin fragility, mutant p63 proteins expressed in other ectodermal dysplasias, such as EEC, cause limb abnormalities. These differences are likely caused by differential activation of p63 target genes by the different mutant p63 proteins. For example, whereas EEC mutants may be able to normally activate p63 target genes involved in skin development and differentiation, they may fail to activate target genes required for limb development. Further studies into the role of p63 in skin and appendage development as well as into the molecular role of mutant p63 proteins will be necessary for a better understanding of the role of p63 in ectodermal dysplasias and other developmental disorders.

Abbreviations

TA cells

Transit amplifying cells

AEC

Ankyloblepharon ectodermal dysplasia and clefting

EEC

Ectrodactyly ectodermal dysplasia and cleft lip

SHFM

Split hand foot malformation

EEM

Ectrodactyly ectodermal dysplasia and macular dystrophy