|Home | About | Journals | Submit | Contact Us | Français|
Epidermal development and differentiation are tightly controlled processes that culminate in the formation of the epidermal barrier. A critical regulator of different stages of epidermal development and differentiation is the transcription factor p63. More specifically, we previously demonstrated that p63 is required for both the commitment to stratification and for the commitment to terminal differentiation. We now demonstrate that ΔNp63α, the predominantly expressed p63 isoform in postnatal epidermis, also plays a role in the final stages of epidermal differentiation, namely the formation of the epidermal barrier. We found that ΔNp63α contributes to epidermal barrier formation by directly inducing expression of ALOX12, a lipoxygenase which contributes to epidermal barrier function. Our data demonstrate that ΔNp63α directly interacts with the promoter of Alox12 in the developing epidermis. Furthermore, we found that the induction of Alox12 expression by ΔNp63α depends on intact p63 binding sites in the Alox12 promoter. Finally, we found that ΔNp63α can only induce Alox12 expression in differentiating keratinocytes, consistent with the role of ALOX12 in epidermal barrier formation.
The epidermis, the outermost layer of the skin, functions as the body’s primary barrier by protecting the organism from dehydration and environmental insults. The epidermal barrier, which is composed of the cornified envelope, the cornified lipid envelope, and extracellular lipids, is formed in a highly reproducible pattern during late stages of embryogenesis (1,2). Postnatally, the epidermal barrier is maintained by a complex epidermal differentiation program, which results in the constant production of the cellular and lipid components of the barrier (3). Compromised barrier function in prematurely born infants or in patients with certain inherited skin diseases often results in dehydration and increased susceptibility to infections (4–6).
Several members of the large family of lipoxygenases (LOX), lipid-peroxidizing enzymes which stereospecifically insert molecular oxygen into polyunsaturated fatty acids (7), have recently been implicated in epidermal barrier formation. Two LOXs which are critical for epidermal barrier formation are ALOX12B (also known as 12R-LOX) and ALOXE3 (also known as eLOX3). Inactivating mutations in ALOX12B and ALOXE3 are found in individuals with non-bullous congenital ichthyosiform erythroderma (NCIE), an ichthyosis characterized by hyperkeratosis and a disruption of the epidermal permeability barrier (8–10). In addition to ALOXE3 and ALOX12B, ALOX12 (also known as platelet-type 12-lipoxygenase [p12-LOX]), is expressed in mouse and human epidermis (11–14). Within mouse epidermis, ALOX12 is expressed in the granular layer and is involved in maintaining the epidermal water barrier as demonstrated by an increase in transepidermal water loss in Alox12 deficient mice (11). The relatively minor barrier defect observed in Alox12 deficient mice would suggest that ALOX12 has a minor role in epidermal barrier formation. An alternative explanation could be that compensatory pathways are induced that maintain barrier function in the absence of ALOX12.
In addition to mice with genetic abnormalities in LOXs, epidermal barrier formation is also compromised in mice which lack transcription factors that, directly or indirectly, induce expression of components of the epidermal barrier (3). For example, mice lacking the transcription factor p63 fail to establish an epidermal barrier, causing the mice to die perinatally due to dehydration (15–17). However, since p63-deficient mice completely fail to develop an epidermis, it remains unclear whether p63 directly contributes to epidermal barrier formation. This question is further complicated by the finding that alternative promoter usage and differential splicing cause the p63 gene to be transcribed into six isoforms, each of which may have unique roles during epidermal morphogenesis (18). For example, p63 isoforms containing a transactivation domain (TAp63) are required for the commitment to stratification during early stages of skin development (15). In contrast, p63 isoforms lacking this transactivation domain (ΔNp63) function after the commitment to stratification has occurred, and have important roles in cell adhesion, basement membrane formation, and terminal differentiation (19). Although ΔNp63 isoforms lack the N-terminal transactivation domain, they contain an internal transactivation domain, thus allowing ΔNp63 isoforms to activate gene expression (20,21).
To determine the role of ΔNp63α in the epidermis, we previously generated epidermal-specific inducible ΔNp63 knockdown (ΔNp63 i-kd) mice (22). In these mice, a hairpin consisting of sequences specific for ΔNp63 isoforms was placed under the control of the “gene-switch” system. The regulator of the “gene-switch” system (Glp65) is activated by progesterone antagonists such as RU486, and was placed under control of the keratin 14 promoter (K14.Glp65), thus targeting expression of Glp65 to the basal layer of the epidermis. In mice carrying both components of the inducible system, topical treatment with RU486 results in expression of the ΔNp63 hairpin. This hairpin is subsequently processed into ΔNp63-specific siRNAs which degrade endogenous ΔNp63 transcripts. Using this system, we found that downregulating ΔNp63 in the epidermis resulted in hyperproliferation, impaired terminal differentiation, and basement membrane abnormalities (22). Furthermore, most ΔNp63 i-kd mice died after a few days of transgene activation, presumably due to excessive water loss caused by impaired epidermal barrier function.
Although these experiments demonstrated a role of p63 in epidermal barrier formation, it is currently not known how p63 regulates this process. Here, we identify Alox12 as the first direct target gene of p63 with a role in epidermal barrier formation. We found that, in the developing epidermis, ΔNp63α interacts with p63 response elements located within the Alox12 promoter, and that this interaction is required for the induction of Alox12. Furthermore, our data demonstrate that ΔNp63α can only induce Alox12 expression in differentiating keratinocytes, consistent with the role of ALOX12 in epidermal barrier formation.
Epidermal-specific inducible ΔNp63 knockdown mice (ΔNp63 i-kd) were previously generated and characterized (22). To downregulate ΔNp63 expression in the basal layer of the epidermis, ΔNp63 i-kd and control newborn mice were treated topically with 1mg/ml RU486 for 5 consecutive days. Skin samples were obtained at day 6. All experiments involving mice were performed with approval from the Institutional Animal Use and Care Committee (IACUC).
Total RNA was extracted using RNeasy Kits (Qiagen) and cDNA was synthesized using the High-capacity cDNA Archive Kit (Applied Biosystems). An assays-on-demand TaqMan probe for Alox12 was also obtained from Applied Biosystems. For PCR amplification of cDNA, we used Taqman Universal PCR Master Mix and the Opticon2 System (MJ Research). The amount of 18S RNA in every sample was used to normalize Alox12 mRNA levels. The relative amount of each mRNA was determined by the comparative CT method. Statistical significance was determined using two tailed t-tests.
Protein was extracted from skin of newborn ΔNp63 i-kd and control mice after 5 days of topical RU486 treatment. Western blotting was performed with rabbit anti-ALOX12 (Santa Cruz), mouse anti-p63 (4A4; Santa Cruz), and goat anti-β-actin (Santa Cruz) antibodies. After incubation with HRP-conjugated secondary antibodies (Sigma), protein bands were visualized using SuperSignal West Pico Substrate (Pierce).
Primary keratinocytes were isolated from newborn mice and cultured as previously described (23). Keratinocytes were transduced with adenoviruses encoding ΔNp63α or GFP at an m.o.i. (multiplicity of infection) of 10. After transduction, keratinocytes were cultured in 50% conditioned media (CM) containing 0.05mM Ca2+. The Ca2+ concentration in the media was increased to 0.1mM after 24 hours to induce keratinocyte differentiation. Keratinocytes were harvested for RNA extraction 24 hours after this switch in Ca2+ concentration.
For chromatin immunoprecipitation experiments, embryonic skin was isolated from 20 E14.5 or E15.5 embryos and collected in D-PBS on ice. Tissue samples were homogenized and fixed in 1% formaldehyde. Cross-linking was stopped by adding a final concentration of 0.125M glycine. After washing, cells were resuspended in 1ml lysis buffer (1% SDS and 10mM EDTA) and sonicated. Chromatin was immunoprecipitated with 3μg of anti-p63α (H129; Santa Cruz) or anti-K14 antibodies. Immunoprecipitated samples were used to perform PCR with primers surrounding p63RE-Alox12: FW 5′-TAG CTG ATC CAA ACG CAC AC and RV 5′-GGG CCA GGT CCA AAA CTT TA. GAPDH was amplified as control to confirm equal genomic DNA content in the immunoprecipitated samples using primers FW: 5′-CCA ATG TGT CCG TCG TGG AT and RV: 5′-TGC TGT TGA AGT CGC AGG AG.
A 450bp fragment of the Alox12 promoter containing a putative p63 response element (p63RE-Alox12) was cloned into the pGL3-basic vector (Promega) using primers: FW 5′-GTG TTT GGG GAC CAC AGA T and RV 5′-GGC TCC CTC TGG CCT CTC A. Mutations in p63RE-Alox12 were introduced using the QuikChange® Site-Directed Mutagenesis Kit (Stratagene) according to manufacturer’s instructions. The sequences of the p63 binding site were changed from CAGG to CCCG and from CTAG to CTCC.
Keratinocytes were nucleofected using Amaxa Biosystems Nucleofector ®II and the Amaxa Kit for Primary Mammalian Epithelial Cells (1×106 cells per cuvette, program M5). Keratinocytes were nucleofected with reporter constructs (1μg), a ΔNp63α expression construct (3μg), and a pCMV-βgal plasmid (0.5μg). Keratinocytes were plated on collagen IV-coated plates in 50% CM containing 0.05mM Ca2+ and incubated at 37°C. After 16 hours, keratinocytes were washed and fed with 50% CM containing 0.1mM Ca2+. Cells were harvested 36 hours later and analyzed for luciferase and β-galactosidase activity using Dual-Light® System (Applied Biosystems) and Dynex Technologies Revelation MLX luminometer. Average reporter gene activities and standard deviations were determined based on three independent experiments. Statistical significance was determined using two tailed t-tests.
To identify novel ΔNp63α target genes, we performed microarray analyses and compared gene expression profiles of ΔNp63 i-kd and control skin of newborn mice that were treated for 5 days with RU486. We previously reported that IκB kinase-alpha (Ikkα) was downregulated in ΔNp63 i-kd skin and determined that IKKα is a direct target gene of ΔNp63α in the developing epidermis (22). Since IKKα functions downstream of ΔNp63α, genes that are downregulated in both ΔNp63 i-kd skin and in Ikkα−/− skin are unlikely to be direct ΔNp63α target genes. To identify ΔNp63α target genes that function independent of IKKα, we also performed microarray analyses on skin RNA isolated from Ikkα−/− and control mice (24). By performing a comparative analysis of the two sets of microarray data, we identified Alox12 as a transcript that was significantly downregulated in ΔNp63 i-kd skin, but not in Ikkα−/− skin. Real-time RT-PCR analysis and Western blot analysis confirmed that Alox12/ALOX12 expression was downregulated in ΔNp63 i-kd skin, when compared to control skin (Fig. 1A, B). Interestingly, expression levels of Aloxe3 and Alox12b were unchanged in ΔNp63 i-kd skin (data not shown). Together with the finding that downregulating ΔNp63 in human keratinocytes results in reduced ALOX12 expression in organotypic cultures derived from these cells (25), these data suggest that Alox12 is a target gene of ΔNp63 in the skin.
Since ΔNp63α is the predominantly expressed p63 isoform in newborn mouse epidermis, the reduced expression of ALOX12 in ΔNp63 i-kd skin is most likely due to the reduction of ΔNp63α expression. To test whether ΔNp63α can induce Alox12 expression, we overexpressed ΔNp63α in primary keratinocytes using an adenovirus (Ad-ΔNp63α). Keratinocytes were cultured under proliferating conditions (0.05mM Ca2+) or under conditions that promote differentiation (0.1mM Ca2+), and Alox12 expression levels were determined using Real Time RT-PCR. Interestingly, we found that whereas ΔNp63α did not affect Alox12 expression in keratinocytes cultured under low Ca2+ conditions, Alox12 expression was induced by ΔNp63α in keratinocytes cultured under high Ca2+ conditions (Fig. 1C). Surprisingly, baseline Alox12 transcript levels were comparable in keratinocytes cultured under low or high Ca2+ conditions. Since ALOX12 protein is only produced in differentiating keratinocytes, these data suggest that ALOX12 protein is either not produced or not properly stabilized in undifferentiated keratinocytes.
To further understand the role of Alox12 in the epidermis, we determined the expression pattern of Alox12 during skin morphogenesis. To this end, we performed Real-Time RT-PCR analysis for Alox12 on skin RNA isolated from ICR embryos beginning at embryonic day (E) 10.5 through term. We found that Alox12 transcript levels were first detectable at E13.5, while robust expression was observed starting at E14.5 (Fig. 1D). Alox12 remains expressed at later developmental stages and postnatally, consistent with previous reports [Fig. 1D and (11,14)].
We next asked whether ΔNp63α induces Alox12 expression by binding to a p63 response element (p63-RE) in the promoter region of Alox12. Although p63-RE remain poorly defined, it is generally accepted that p63 interacts with degenerate p53-RE (26,27). However, the amount of degeneracy that allows for p63 binding remains unclear (28,29). To identify regulatory elements within the Alox12 promoter, we initially compared a 10kb genomic region upstream of the mouse and human Alox12 genes using the Consite software (http://asp.ii.uib.no:8090/cgi-bin/CONSITE/consite). These comparisons led to the identification of a conserved region within the Alox12 promoter, which contained a putative p63-RE. This p63-RE (p63RE-Alox12) is located approximately 170bp upstream from the first exon of Alox12 (Fig. 3A) and is conserved in the human ALOX12 promoter.
To determine whether ΔNp63α physically interacts with p63RE-Alox12, we performed chromatin immunoprecipitation (ChIP) assays. Since Alox12 expression levels are high at E14.5 and E15.5 (Fig. 1D), we isolated chromatin from skin samples obtained at these developmental stages. To immunoprecipitate protein-DNA complexes, we used anti-p63α or control (anti-K14) antibodies. Since ΔNp63α is the predominant p63 isoform expressed in E14.5 and E15.5 skin, DNA fragments immunoprecipitated with the anti-p63α antibody most likely correspond to fragments that are bound by ΔNp63α. PCR amplification of these DNA fragments with primers surrounding p63RE-Alox12, revealed an increased recovery of promoter fragments containing p63RE-Alox12 (Fig. 2), demonstrating that ΔNp63α directly interacts with the promoter of Alox12 in E14.5 and E15.5 embryonic skin.
To determine if ΔNp63α induces Alox12 expression by interacting with p63RE-Alox12, we performed luciferase reporter gene assays. To this end, we cloned a 450bp fragment of the Alox12 promoter containing p63RE-Alox12, into a luciferase reporter vector. Since ΔNp63α functions in keratinocytes, luciferase reporter assays were performed in mouse primary keratinocytes. After nucleofection with the reporter constructs, keratinocytes were cultured in media containing a high (0.1mM) concentration of Ca2+ to maintain the cells in a differentiating state. Consistent with our finding that ΔNp63α can induce endogenous Alox12 expression only in keratinocytes cultured under high Ca2+ conditions, we found that ΔNp63α could activate the Alox12 reporter in keratinocytes cultured under high Ca2+ conditions (Fig. 3B). To determine if activation of the Alox12 reporter was dependent on p63RE-Alox12, we mutated the core sequences of p63RE-Alox12 in the context of the Alox12 reporter construct (p63RE-Alox12-Mut) (Fig. 3A). We found that ΔNp63α was unable to activate p63RE-Alox12-Mut, demonstrating that ΔNp63α-mediated p63RE-Alox12 activation is dependent on the interaction of ΔNp63α with p63RE-Alox12 (Fig. 3B). Taken together, these data suggest that in E14.5 and E15.5 embryonic skin ΔNp63α induces Alox12 expression by directly interacting with the Alox12 promoter.
Since the cloning of p63 in 1998, many putative target genes of ΔNp63α have been identified. These target genes include genes required for keratinocyte adhesion, keratinocyte proliferation, keratinocyte terminal differentiation, and basement membrane formation (19). Intriguingly, ΔNp63α induces these target genes at different developmental stages, suggesting that induction of target gene expression by ΔNp63α requires transcriptional co-activators which are specifically expressed at the appropriate developmental time points. Although some co-activators of TAp63 isoforms have been identified (30,31), co-activators that cooperate with ΔNp63α during epidermal development remain to be identified. The earliest known gene induced by ΔNp63α during epidermal morphogenesis is PERP, a desmosomal component which is critical for cell adhesion in the epidermis (32). PERP is induced by ΔNp63α shortly after surface epithelial cells have adopted an epidermal fate, a process which occurs around E9.5 in mice. Subsequently, at E11.5, ΔNp63α induces a critical component of the embryonic basement membrane, FRAS1 (22). When commitment to terminal differentiation occurs, around E15.5, ΔNp63α induces the expression of IKKα, which is required for cell cycle withdrawal during spinous layer development (22,33). Finally, ΔNp63α also synergizes with Notch signaling components to induce expression of the terminal differentiation marker keratin 1 (34). Although other putative ΔNp63α target genes have been identified, most of these were identified using cell lines derived from tissues where p63 does not normally function, such as liver and osteosarcoma-derived cell lines. Therefore, it remains to be determined whether these genes are regulated by p63 during epidermal development and/or differentiation.
We now add Alox12, which encodes a LOX involved in epidermal barrier formation, to the list of genes induced by ΔNp63α (11). During epidermal development, ΔNp63α directly interacts with the promoter region of Alox12, and this interaction is required for the induction of Alox12 by ΔNp63α. Like epidermal terminal differentiation, epidermal barrier formation is a process which is dependent on extracellular Ca2+. In the epidermis, the extracellular Ca2+ concentration is highest in the granular layer, where components of the epidermal barrier are produced (35). Furthermore, the establishment of the Ca2+ gradient in the developing epidermis occurs concurrently with the acquisition of epidermal barrier function (36). The instructive role for Ca2+ in epidermal barrier formation was established by the finding that Ca2+ is essential for the expression of genes that are required for barrier formation (37–39). Consistent with these findings, we found that ΔNp63α could only induce Alox12 expression in keratinocytes that were cultured in media containing high levels of Ca2+. Since ΔNp63α cannot induce Alox12 expression in keratinocytes cultured under low Ca2+ conditions, Alox12 is likely to represent another target gene of ΔNp63α which requires transcriptional co-activators for its expression.
Paradoxically, unlike ALOX12, the expression of ΔNp63α declines during epidermal differentiation. However, consistent with previous reports (40), we found that differentiating keratinocytes still contain ~25% of the amount of ΔNp63α protein that is present in undifferentiated keratinocytes (unpublished observations). This level of ΔNp63α protein is clearly functional since it cooperates with Ca2+ to induce Alox12 expression in differentiating keratinocytes, where ΔNp63α and ALOX12 are co-expressed.
In summary, we have identified a novel role for ΔNp63α in directly inducing the expression of Alox12, a gene involved in epidermal barrier formation. Our finding that ΔNp63α only induces Alox12 expression in primary keratinocytes that are cultured under differentiating conditions is consistent with the known role for ALOX12 in epidermal barrier formation.
We thank Dr. Frank McKeon for providing mouse p63 expression constructs. We also would like to thank Dr. Peter J. Koch for his constructive comments on the manuscript. This work was supported by NIH grants to DRR (AR052263, CA52607, and AR47898) and MIK (AR054696), and a research grant from the National Foundation for Ectodermal Dysplasias (NFED) to DRR and MIK.