Reciprocal regulation of p63 by C/EBP delta in human keratinocytes
Genetic experiments have clarified that p63 is a key transcription factor governing the establishment and maintenance of multilayered epithelia. Key to our understanding of p63 strategy is the identification of target genes. We perfomed an RNAi screening in keratinocytes for p63, followed by profiling analysis.
C/EBPδ, member of a family with known roles in differentiation pathways, emerged as a gene repressed by p63. We validated C/EBPδ as a primary target of ΔNp63α by RT-PCR and ChIP location analysis in HaCaT and primary cells. C/EBPδ is differentially expressed in stratification of human skin and it is up-regulated upon differentiation of HaCaT and primary keratinocytes. It is bound to and activates the ΔNp63 promoter. Overexpression of C/EBPδ leads to alteration in the normal profile of p63 isoforms, with the emergence of ΔNp63β and γ, and of the TA isoforms, with different kinetics. In addition, there are changes in the expression of most p63 targets. Inactivation of C/EBPδ leads to gene expression modifications, in part due to the concomitant repression of ΔNp63α. Finally, C/EBPδ is found on the p63 targets in vivo by ChIP analysis, indicating that coregulation is direct.
Our data highlight a coherent cross-talk between these two transcription factors in keratinocytes and a large sharing of common transcriptional targets.
KeywordsHaCaT Cell Primary Keratinocytes Sterile Alpha Motif Spinous Layer Primary Human Keratinocytes
p63 is a transcription factor -TF-homologous to the tumour suppressor p53 and to p73 . This class of proteins activate and repress genes as a result of binding to promoters and enhancer regions. Six isoforms can be found as a result of different transcription initiation sites and alternative splicing. p63 proteins contain -TA- or lack -ΔN- a transcriptional activation domain at the N-terminal, and the sterile alpha motif -SAM- domain, presumably a protein-protein interaction module. The resulting proteins have dissimilar transcriptional properties and, as a result, different biological behaviour. The paramount importance of p63 in development has been broadly illustrated by genetic experiments in different organisms. Mice lacking p63 die soon after birth with severe defects in limb, craniofacial and skin development [2, 3]. The major isoform present in keratinocytes and other epithelia -ΔNp63α- is essential for ectodermal development in zebrafish . In humans, several syndromes showing abnormalities in limbs, skin and annexes are caused by mutations in the p63 gene . In general, therefore, p63 is essential for the biology of multilayered epithelia.
Several TFs have an established role in keratinocytes programs , but few of them have so far been linked to p63 regulation. This topic is quite relevant since renewal of stem cells and ongoing active terminal differentiation most likely requires the progressive fine tuning of transctriptional programs, unlikely to be masterminded by a single TF.
C/EBPs are a family of six B-Zip TFs that activate and repress transcription under different differentiation and growth arrest conditions . C/EBPα and C/EBPβ are required for differentiation of adipocytes [8, 9, 10]. Inactivation of C/EBPε in KO mice leads to lack of natural killer cells . C/EBPδ KO has a mammary phenotype, with an alteration in the involution of the mammary glands upon lactation [12, 13]. Adipocyte differentiation is also impaired in cells that lack C/EBPδ when cultured in vitro, similarly to C/EBPβ .
By inactivating p63 in HaCat and primary cells, it was possible to investigate the roles of p63 in keratinocytes, notably the importance for differentiation and cell adhesion programmes [15, 16]. At the same time, however, p63 plays a major role in maintaining the proliferative potential of stem cells of the multilayered epithelia . Through the use of RNAi inactivation coupled to gene expression profiling, as well as ChIP on chip experiments, several labs have recently identified hundreds of p63 targets [15, 16, 17, 18, 19, 20]. Specifically, C/EBPδ emerged in the RNAi profiling of human HaCaT cells and primary keratinocytes as a target of p63 [18, 19]. This prompted us to validate it and characterize the p63-C/EBPδ connections by RT-PCR, ChIPs and immunofluorescence in human keratinocytes.
Validation of C/EBPδ as a target of p63
C/EBPδ is regulated upon keratinocytes differentiation
C/EBPδ is differentially expressed in human skin
Regulation of ΔNp63 by C/EBPδ
To test this, we searched for C/EBP sites by using different algorithms in the human ΔNp63 promoter, and found a consensus of two juxtaposed boxes at -2400, in a region that shows considerable homology in the mouse gene (Fig. 4B). This led us to verify the activation of the ΔNp63 promoter using Luciferase constructs of different lengths, from -3000 to -200 relative to the transcriptional Start Site. As the promoter is known to be dependent from NF-Y, a ubiquitous activator [22, 23], we included it in our cotransfections assays as a positive control. These experiments were performed in U2OS, which provide a background that is free of the possible dominant self-regulating activity of p63 on its own promoter . We found that NF-Y and C/EBPδ singularly activated the -3000 contruct (Fig. 4C). Remarkably, addition of the two TFs together yielded a clear cooperative effect. However, shorter constructs (-1300, -400 and -200) did not show any significant increase upon C/EBPδ overexpression, while still retaining some NF-Y inducibility, most likely due to core promoter CCAAT boxes detailed in Refs.  and . The decrease in NF-Y inducibility with respect of the -3000 construct is probably due to a functional CCAAT box binding NF-Y in this region (See ChIP below). The simplest explanation for these results is that the -2400 C/EBP sites are vital for induction. To confirm that the activation observed is a primary phenomenon, we performed scanning ChIPs analysis on different regions of the C/EBPδ promoter in human KCs; Fig. 4D shows that C/EBPδ and NF-Y are found in the upstream region, in at least three different positions. Note that predicted and actual sites do not always ovelap (Fig. 4C and 4D). Taken together, these data indicate that (i) C/EBPδ binds and activates the ΔNp63 promoter in keratinocytes, (ii) point at the -2400 binding sites as important for regulation, (iii) confirm that NF-Y is a bona fide activator of the ΔNp63 promoter.
Consequences of C/EBPδ overexpression/inactivation on expression of p63 targets
RT-PCR analysis of keratinocytes in which C/EBPδ was inactivated showed an equally dramatic change in gene expression. In most cases, there is a perfect match with overexpression: genes that are activated, 24 or 48 hours after C/EBPδ transfections are down-regulated by RNAi, consistent with a positive role of C/EBPδ in their regulation. The exception is ESR, whose undetectable levels could not be further decreased by C/EBPδ inactivation. The genes that are repressed in overexpression are increased by RNAi, except EGF-R and Desmocollin 3, in which a dominant role of p63 can be envisaged. In summary, the C/EBPδ RNAi and overexpression experiments are consistent with this TF being important for the expression of many p63 targets.
p63 targets are bound by C/EBPδ in vivo
C/EBPδ-p63 targets are regulated during differentiation
The concerted modulation of specific gene expression programs is controlled by transcription factors. In specific pathways of response to external stimuli, one, or a few, TFs play a dominant role in this response. Matters are far more complex in multi-cellular systems involving terminal differentiation and the specification of exquisite features. What emerged is the concept of "master" regulators, whose role is to confer cell identity and to drive gene regulation programs accordingly. In the muscle system, for example, some of these masters -MyoD, Myogenin, MEF2- were identified and their gene regulation programs studied: they apparently show a preference for fellow TFs genes. A robust hierarchical cascade of specific programs is adjusted in such multisteps and multi-functions programs . Interestingly, this network is not restricted to TFs exquisitely expressed in muscle, since ubiquitous, or near ubiquitous TFs are an integral part of the network . Similarly, the recent identification of hundreds of p63 targets in various cell types [Reviewed in [28, 29, 30]] highlighted links with TFs that are known, or likely, to play an important role in skin biology .
The direct cross-talk between p63 and C/EBPδ, originally emerging from p63 RNAi screenings [18, 19] is confirmed by ChIP and C/EBPδ RNAi experiments. The decrease in ΔNp63α by inactivation of C/EBPδ strongly argues in favour of an activating role of C/EBPδ on ΔNp63 transcription, confirmed by overexpression; reporter constructs experiments and ChIPs point at important upstream sites at -2.4 Kb. Interestingly, the synergistic activation observed with NF-Y is a further argument in favour of the physiologic role of this particular member of the C/EBP family in ΔN p63 regulation: NF-Y, itself a poor activator, acts primarily in conjunction with a neighbouring TF [Reviewed in ]. In the MHC Class II promoters, for example, only the relevant RFX-5, and not other RFXs, is capable to synergize with NF-Y .
An intriguing finding is the differential regulation of p63 splicing isoforms in keratinocytes upon C/EBPδ overexpression. Normally, these cells express exclusively ΔNp63α and it is currently unclear whether TA isoforms are induced upon differentiation in vitro, since different results were reported [Reviewed in  and ]. Overexpression of C/EBPδ led to the appearance of ΔNβ and γ, and -later- of the TA isoforms. Thus, C/EBPδ acts not only to support the activity of the ΔN promoter in general, but, directly or indirectly, to impinge on an isoform switch that modifies the p63 portfolio. ΔNp63β and γ induction was also observed upon activation and differentiation of cells of the corneal limbus , and a role of TA isofoms in differentiation has been proposed . The switch could have a far reaching effect on expression of p63 targes, particularly in the upper epidermal layers, since the ΔN and TA isoforms have often quite dissimilar transcriptional effects.
How modification of the splicing is accomplished is unclear, at the moment; it has become recently evident, however, that regulation of transcription and RNA processing are highly coordinated events and that specific transcription factors -and cofactors- are known to play a role in pre-mRNA processing . Interestingly, one of the cofactors involved is CARM1 , an arginine methyltransferase which modifies other factors promoting exon-skipping . CARM1 is activated by ΔNp63α  and it was shown to be involved in other differentiation systems . Thus, it is possible that p63 regulators, such as C/EBPδ, impact on factors loaded on the ΔN promoter and travel along the gene with PolII complexes to adjust splicing events.
Our results indicate that many, perhaps most of the p63 targets are coregulated by C/EBPδ. An intersection between C/EBPs and p53 family members was noticed before. Typical p53-induced pro-apoptotic targets are not activated in the absence of C/EBPδ in KO mammary cells and anti-apoptotic genes are not repressed . The two factors are apparently engaged in a positive feed-back loop, with many genes commonly controlled: Cyclin D1, Bfl1, SGP2, Gas1, Bak and IGFBP5. In addition, expression of C/EBPδ is induced upon Vitamin D treatment in prostate cells, leading to a dramatic arrest in clonal expansion , presumably through VDR, another p63 target . Given the pivotal role of p63 in prostate development , the interplay between p63 and C/EBPδ could be both direct and indirect, through VDR. Note that the overlap between p53 and p63 targets in the skin is large, but not absolute [17, 18].
HaCaT cells are an epithelial line derived from the back of adult human skin that spontaneusly underwent immortalization in vitro ; they exhibit normal differentiation and have provided a valuable tool for studying regulation of keratinization in human cells. The interpretation of data concerning p63 in HaCaT is often complicated by the presence of two mutated p53 alleles , which are known to impinge on the p63 pathways. This is apparently ot the case for C/EBPδ, but for other p63 targets we are analyzing, there are dramatic differences in regulation between HaCat and KCs (S.B., R.M, in preparation). In light of the overlap between the targets of the two TFs, this implies that a mutated p53 might subvert the tumor suppressive and pro-differentiation role of C/EBPδ.
Based on current knowledge derived mostly from RNAi experiments, a tentative view of the influence of p63 and C/EBPδ is presented in Fig. 5. The picture is complicated by the effect of p63 inactivation through C/EBPδ RNAi and by the differential role of TAp63  activated at late time points. With two exceptions, EGF-R and DSC3, all genes show a consistent behaviour in overexpression and RNAi. Genes that are repressed by p63, ETAR, AIM2, T-Cadherin, JunB, and should be up-regulated by p63 decrease, are rather decreased, an indication that they are mostly rely on C/EBPδ. Some of these -AIM2, T-Cadherin, JunB- limit growth. A role of C/EBPδ as a tumor suppressor was established by the finding that knock-out MEFs show genomic instability, impaired contact inhibition and reduced serum dependance . For the p63-repressed c-Jun, up-regulation by C/EBPδ could either be because of removal of p63 or repression by C/EBPδ, as indicated by overexpression at late time points. As for genes activated by ΔNp63, such as EGF-R, TGFβ-RII and E-cadherin, the latters could be under joint control; EGF-R, which drops after C/EBPδ and p63 RNAi, could decrease after C/EBPδ overexpression at 48 hours through the increase of TAp63 isoforms, shown to be strong repressors of the EGF-R promoter. This would be consistent with a growth control and pro-differentiative role of both TAp63 and C/EBPδ. Similarly, removal of the repressive ΔNp63 [See Ref. ] by C/EBPδ RNAi is consistent with an increase in PCNA mRNA and with a small, late decrease in overexpression. In conclusion, it is clear that coregulation of p63/p53 and C/EBPδ is part of a larger program that controls cell-cycle progression and differentiation. To reconstruct the network, it will now be necessary to obtain unbiased information about the panel of C/EBPδ targets in keratinocytes, through the use of RNAi and ChIP on chip technology.
C/EBPδ belongs to a family of TFs important for several differentiation pathways, including adipocytes, liver and haematopoietic lineages . Specifically, it has been associated to the early phases of adipocyte differentiation, together with C/EBPβ. It is induced upon several environmental changes that lead to growth arrest [47, 48, 49, 50, 51] and serum withdrawal in mammary cells and lung epithelia cells . Overexpression leads to growth arrest in mammary and prostate epithelial cell lines [53, 54, 55]. The growth arrest features are common to other C/EBPs [Reviewed in ]. Our findings that C/EBPδ is up-regulated during cell-cycle exit and differentiation of immortalized HaCaT and primary keratinocytes, and the decrease in expression of PCNA, a proliferative marker, are in agreement with these results.
Nothing is known about C/EBPδ expression during epidermal development of mouse skin. Despite careful examination, the available antibodies are unable to stain mouse cells in immunofluorescence analysis [B.T, R.M., unpublished]. In humans, little variation of expression of C/EBPδ was detected in the different phases of maturation of skin annexes, the hair follicle and the sebaceous gland, unlike other members of the family . At the opposite, we observe here that C/EBPδ staining is not uniform in human interfollicular skin, suggesting that the reciprocal interplay with p63 is complex. Co-expression is seen in selected basal cells, which is in line with the activating role of C/EBPδ on the ΔN promoter; in other cells of the same layer, strongly positive for p63, C/EBPδ fades: this is also consistent with the negative role of p63 on C/EBPδ expression. There is a coexpression and a balance in the spinous layer, and in the upper granular layer, with cells exiting from a proliferation status and terminally differentiating, C/EBPδ is prominent. In transfection assays, Desmocollin 3, which is expressed in basal cells, is transactivated by C/EBPδ and C/EBPβ, the suprabasal Desmocollin 1 by C/EBPδ and C/EBPα . We confirm that both these genes require C/EBPδ expression, as determined by RNAi experiments. Thus our results are consistent with C/EBPδ playing a dual role in skin differentiation, both in very early and late stages. In the former, it could limit, through its growth suppressive properties the growth potential of early progenitors. In the later stages, it may coordinate cell-cycle exit and induce differentiation markers. Our observations could now be extended in other systems, notably in the mammary gland and in prostate, in which both TFs have been singularly been shown to play key roles.
One caveat to the C/EBPδ role in skin is the fact that KO mice have no apparent alterations. However, this might be due to redundancy with C/EBPβ: it was reported that C/EBPβ is expressed in the cytoplasm of basal keratinocytes and becomes nuclear in the spinous layer . C/EBPβ KO mice have skin hyperplasia with downplay of keratin 1 and 10 expression  and C/EBPβ is important for keratinocyte survival . Of particular significance are the data recently reported by the group of Vinson on a transgenic model expressing a dominant negative C/EBP -termed A-C/EBP- in basal keratinocytes: these mice have hyperplasia of the basal epidermis and increased apoptosis of the upper layers . p53 and pro-apoptotic markers are induced and C/EBPβ dramatically reduced; as A-C/EBP is not C/EBPβ-specific, C/EBPδ might contribute to the observed phenotype. Given the well known complexity within this family, through homo- and heterodimerizations, production of dominant negatives through the use of internal AUG within C/EBPβ, much work lies ahead to establish their role in the different layers of epidermis.
We identified a direct and mutual link between p63 and a member of the C/EBP family of transcription factors, C/EBPδ. The following relevant findings are the reported: (i) C/EBPδ expression is confined to keratinocytes and it is up-regulated in cells committed to differentiation; it is increased in in vitro differentiation systems, both of HaCaT and primary keratinocytes. (ii) C/EBPδ activates transcription of the ΔNp63 promoter, and it appears to be part of the mechanisms that control p63 splicing. (iii) It is involved in regulation of other p63 targets regulated during differentiation, as verified by RNAi, overexpression and ChIP assays. We therefore suggest that the mutual link between these TFs is important for the correct differentiation of keratinocytes.
Cells and culture conditions
First passage primary human keratinocytes -KCs- were derived from breast of healthy individuals and grown on a feeder-layer of lethally irradiated 3T3 cells in DMEM F12 added of Insulin (5 μg/ml), EGF-R (10 ng/ml) hydrocortisone (0.4 μg/ml), T3 (2 nM), Cholera toxin (0.1 nM) and transferrin (5 μg/ml). HaCaT were grown in DMEM or in low calcium medium when assayed for differentiation, which was added with 1.4 mM CaCl2 in 0.1% serum conditions. Primary KCs were differentiated by adding CaCl2 (1.4 mM final concentration) in the presence of 10% foetal calf serum.
RT-PCR and transfections
HaCaT cells were transiently trasfected using Oligofectamine or Lipofectamine 2000 (Gibco-BRL, USA) for 3 hours with 150 ng/cm2 of human p63 siRNA oligonucleotide (ACAATTTCATGTGTAACAGCA) which targets aminoacids 265–272 in the central DNA-binding domain of p63. After incubation overnight in DMEM, transfection was repeated for 3 additional hours. RNA was extracted from HaCaT cells using an RNA-Easy kit (Qiagen). 2.5 × 105 first passage primary human keratinocytes from healthy individuals were similarly transfected with Nucleofector (Amaxa, D) according to the Manufacturer' conditions with the off-target siRNA control oligos (5'-AUGAACGUGAAUUGCUCAA-3', 5'-UAAGGCUAUGAAGAGAUAC-3', 5'-AUGUAUUGGCCUGUAUUAG-3', 5'-UAGCGACUAAACACAUCAA-3'; Dharmacon D-00181001), or with three different oligonucleotides targeting human C/EBPδ (S1: GAUGCAGCAGAAGUUGGUGuc; S2: GACUCAGCAACGACCCATuu; S3: GGAAAAGACUGAGCAUGCUuu).
RNA was extracted 24 or 48 hours after transfections. For cDNA syntesis, 4 μg of RNA were used with M-MLV-RT kit (Invitrogen, USA). Semi-quantitative PCR analysis were performed with specific primers [see additional file 1].
In Figure 4, 2 × 105 U2OS were transfected with Lipofectamine (Gibco-BRL) using 1,2 μg of reporter plasmids, 200 ng of C/EBPδ, 70 ng of the NF-YA, NF-YB, NF-YC plasmids, and carrier for a total DNA concentration of 2 μg. Six independent transfections in duplicate were performed.
ChIP analysis were carried out with the method described in Ref.  with an anti-C/EBPδ antibody (Active Motif #39006), anti NF-YB (Diagenode), anti-p53 (Ab7, Calbiochem) and anti-p63 abs (SantaCruz SC137, 4A4 DAKO, Diagenode). For the oligos used see additional file 1.
Western blot and Immunofluorescence
Western blot analysis was performed using standard procedure with a Pierce secondary antibody and detection system. Skin sections derived from thighs of healthy donors are shown in Figure 3. Cryopreserved human skin sections were fixed in 4% fresh paraformaldehyde for 10' and incubated o/n with the primary antibodies anti-p63  and anti-C/EBPδ (Active Motif #39006). After 1 hour incubation with the fluorocrome-conjugated secondary antibodies, the slides were stained with DAPI and mounted in Vectashield. Fluorescence was analysed with a Leica confocal microscope.
Note added in proofs
After submission of this manuscript, Barbaro et al. (J. Cell Biology 177, 1027–1049, 2007) reported that C/EBPδ and p63 are coexpressed in stem cells of the corneal limb epithelium and share common gene targets. These authors also resported that C/EBPδ binds to the ΔNp63 promoter.
We wish to thank A. Costanzo, H. Van Bokhoven, R. Smart, M. Chidgey, J. DeWille for gift of reagents, S. Pozzi for help with human KCs experiments. We thank S. Rodeghiero and U. Fascio at the CIMAINA facility for skilful assistance with confocal microscopy. M.A. Viganò was supported by MIUR-FIRB Giovani Ricercatori and UE-EPISTEM contracts. This work was supported by Telethon, Fondazione Cariplo and UE-EPISTEM (RM).
- 4.Bakkers J, Hild M, Kramer C, Furutani-Seiki M, Hammerschmidt M: Zebrafish DeltaNp63 is a direct target of Bmp signaling and encodes a transcriptional repressor blocking neural specification in the ventral ectoderm. Dev Cell 2002, 2: 617-627. 10.1016/S1534-5807(02)00163-6.CrossRefPubMedGoogle Scholar
- 11.Yamanaka R, Barlow C, Lekstrom-Himes J, Castilla LH, Liu PP, Eckhaus M, Decker T, Wynshaw-Boris A, Xanthopoulos KG: Impaired granulopoiesis, myelodysplasia, and early lethality in CCAAT/enhancer binding protein epsilon-deficient mice. Proc Natl Acad Sci USA 1997, 94: 13187-13192. 10.1073/pnas.94.24.13187.PubMedCentralCrossRefPubMedGoogle Scholar
- 12.Gigliotti AP, Johnson PF, Sterneck E, DeWille JW: Nulliparous CCAAT/enhancer binding protein delta (C/EBPdelta) knockout mice exhibit mammary gland ductal hyperlasia. Exp Biol Med 2003, 228: 278-285.Google Scholar
- 24.Smith C, Zhu K, Merritt A, Picton R, Youngs D, Garrod D, Chidgey M: Regulation of desmocollin gene expression in the epidermis: CCAAT/enhancer-binding proteins modulate early and late events in keratinocyte differentiation. Biochem J 2004, 380: 757-765. 10.1042/BJ20040077.PubMedCentralCrossRefPubMedGoogle Scholar
- 30.Barbieri CE, Pietenpol JA: Transcriptional Programs Regulated by p63 in Normal Epithelium and Tumors. CellCycle 2007, 6: 246-254.Google Scholar
- 40.Ikezoe T, Gery S, Yin D, O'Kelly J, Binderup L, Lemp N, Taguchi H, Koeffler HP: CCAAT/enhancer-binding protein delta: a molecular target of 1,25-dihydroxyvitamin D3 in androgen-responsive prostate cancer LNCaP cells. Cancer Res 2005, 65: 4762-4768. 10.1158/0008-5472.CAN-03-3619.CrossRefPubMedGoogle Scholar
- 45.Candi E, Rufini A, Terrinoni A, Dinsdale D, Ranalli M, Paradisi A, De Laurenzi V, Spagnoli LG, Catani MV, Ramadan S, Knight RA, Melino G: Differential roles of p63 isoforms in epidermal development: selective genetic complementation in p63 null mice. Cell Death Differ 2006, 13: 1037-47. 10.1038/sj.cdd.4401926.CrossRefPubMedGoogle Scholar
- 54.Hutt JA, O'Rourke JP, DeWille J: Signal transducer and activator of transcription 3 activates CCAAT enhancer-binding protein delta gene transcription in G0 growth-arrested mouse mammary epithelial cells and in involuting mouse mammary gland. J Biol Chem 2000, 275: 29123-29131. 10.1074/jbc.M004476200.CrossRefPubMedGoogle Scholar
- 57.Bull JJ, Muller-Rover S, Chronnell CM, Paus R, Philpott MP, McKay IA: Contrasting expression patterns of CCAAT/enhancer-binding protein transcription factors in the hair follicle and at different stagesof the hair growth cycle. J Invest Dermatol 2002, 118: 17-24. 10.1046/j.0022-202x.2001.01629.x.CrossRefPubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.