Scotin: A new p63 target gene expressed during epidermal differentiation.

p63, a member of the p53 family, is transcribed from two different promoters giving rise to two different proteins: TAp63 that contains the N-terminal transactivation domain and DeltaN that lacks this domain. In this article we describe a new target gene Scotin induced by TAp63 during epithelial differentiation. This gene was previously isolated as a p53-inducible proapoptotic gene and the protein is located in the endoplasmic reticulum and in the nuclear membrane. Scotin expression is induced in response to endoplasmic reticulum (ER) stress in a p53 dependent or independent manner. We detected Scotin upregulation in primary keratinocyte cell lines committed to differentiate. In this paper we also show that Scotin is expressed in the supra basal layer of the epidermis in parallel with TAp63, but not DeltaNp63 expression. We conclude that Scotin is a new p63 target gene induced during epithelial differentiation, a complex process that also involves ER stress induction.

The transcription factor p63, belonging to the p53 gene family [1], is transcribed from two different promoters, generating two classes of proteins: one containing an N-terminal transactivation domain (TAp63) and another that lacks this domain (ΔN) [2-4] In addition, alternative splicing at the 3′ end of the transcripts (TAp63 or ΔNp63) generates three different C-terminal splicing variants: α,β,γ. ΔNp63 can act as dominant-negative suppressor of the TA isoforms [5] and is also endowed with its own transcription-activating function [6-8]. During the last few years, evidence has accumulated on the important role of p63 in the epidermal differentiation program. Mice lacking p63 show severe limb, craniofacial, and skin defects and die soon after birth [9-12]. Genetic complementation of p63−/− mice with TAp63α, ΔNp63α or both suggest that ΔNp63 is important for maintaining the proliferative potential of the basal layer, whereas TAp63 contributes by acting synergistically and/or subsequently to ΔNp63 to allow differentiation [13]. Indeed, TAp63 is capable of inducing growth differentiation factor 15 (GDF15) expression after it has been recruited to the proximal p53/p63 binding site located on the GDF15 promoter and this factor plays a critical role in the regulation of keratinocyte differentiation. [14]. Little is known about the p63, and especially TAp63, target genes relevant for epidermal formation and differentiation, but Notch signaling seems to be important for the commitment of embryonic keratinocytes to terminal differentiation [15]. p63 acts as a selective modulator of Notch 1-dependent transcription and also transactivates the Notch ligand, JAGG-1 [16,17].

In this paper we demonstrate that during epidermal differentiation p63 is capable of inducing Scotin gene expression. This gene was previously identified as a p53-inducible proapoptotic gene located in the endoplasmic reticulum (ER) and in the nuclear membrane. Scotin can induce apoptosis, independently of p53, in response to ER stresses [18]. Endoplasmic reticulum stress can be triggered by a number of changes in normal ER function (accumulation of unfolded, misfolded or excessive protein, ER lipid or glycolipid imbalances etc.) and as a consequence of impaired ER function, luminal calcium is released from the ER stores and apoptotic signaling is triggered [19]. Our laboratory previously demonstrated that TAp73α, another member of the p53 family, is capable of inducing Scotin and ER stress [20,21]. TAp73α was also able to induce GADD153/CHOP, a bZIP transcription factor, induced under ER stress conditions [22-24]. GADD153 is implicated in programmed cell death induced by ER stresses and is upregulated during the differentiation of normal mammary epithelial cells [25].

In this work we have investigated the correlation between Scotin induction, mediated by increased TAp63 expression in the upper layers of the differentiating epidermis, during the complex process that leads to formation of the cornified envelope.

Materials and methods

RTqPCR. Real-time PCR was performed on an ABI-7500 SDS instrument (Applied Biosystem, Foster City, CA, USA) using the Platinum SYBR Green qPCR SuperMix UDG without ROX (Invitrogen Life Technologies, Carlsbad, CA n°cat: 11733–046) in a total volume of 25 μl. The PCR reactions were performed using the following primers: hScotF1 (+) (5′-TTG GAG GCT GAG GAT AAG GGG-) hScotR1 (−) (5′-TGT GGA GCG AGG AAA GGT GTG-) hGADD153F (+) (5′-GCT TCT CTG GCT TGG CTG ACT-) hGADD153R (−) (5′-CCA GGG AGC TCT GAC TGG AA-), hActinF (+) (5′-AAA GAC CTG TAC GCC AAC A-) h ActinR (−) (5′-CGG AGT ACT TGC GCT CAG-) hTAp63F (+) (5′-GGA CTG TAT CCG CAT GCA G) hTAp63R (+) (GAG CTG GGC TGT GCG TAG), hDNp63F (GAA GAA AGG ACA GCA GCA TTG AT), hDNp63R (−) (GGG ACT GGT GGA CGA GGA G). The reaction was performed using the following PCR program: 50 °C for 2 min 95 °C for 3 min followed by 40 cycles of 94 °C for 20 ′′ and 59 °C for 40′′.

Primary keratinocytes, cell cultures, and imaging. Primary keratinocytes were isolated according to Yuspa et al. (1989) from the skin of new born mice (CD-1 mouse strain). The skin was floated overnight in trypsin/EDTA at 4 °C and primary keratinocytes were isolated. Cells were cultured on collagen-coated dishes in medium supplemented with 0.05 mM Ca2+, while 1.2 mM Ca2+ was used to differentiate primary keratinocytes. TAp63α and ΔNp63α-inducible Saos-2 cells were cultured as described by [26]. HaCat cells were maintained in DMEM with 10% FBS. Embryos were fixed as described by [13]. The following primary antibodies were used: monoclonal anti-p63 (Ab4, Neomarkers, Fremont, California, USA; 1/300), rabbit polyclonal anti-mouse Scotin (JC105; dilution 1/150).

Cell cultures (HaCat and Saos-2) were plated on 20 mm glass coverslips at a density of 30,000 cells/cm2 or 25,000 cells/cm2. After 24 h Saos-2 transfected cells or HaCat cells, committed to differentiate (1.2 mM Ca2+ and 0.1% FBS), were fixed as described in Gressner et al. (2005). Cells were then incubated for 1 h at RT with the following antibodies: monoclonal anti-GADD153 (Clone B3, Santa Cruz; 1/100 dilution), and polyclonal anti-loricrin (Covance; 1/200 dilution). Fluorescence was evaluated by confocal microscopy (Nikon, C1 on Eclipse TE200; EZC1 software).

Transfection and Western blots. Cells were transiently transfected (Lipofectamine 2000 Invitrogen, Life Technologies, Carlsbad, CA) and Western blots were performed as described by Candi et al. (2006) using the following primary antibodies: monoclonal anti-p63 (Ab4 clone 1/500), polyclonal anti-tubulin (H-235, Santa Cruz, 1:1000 dilution) polyclonal anti-K10 (Covance; 1/500 dilution), polyclonal anti-mouse Scotin (JC105; 1:1000 dilution), polyclonal anti-human Scotin (H105; dilution 1/1000) and polyclonal Involucrin (Covance; 1:1000 dilution). Proteins were detected using the ECL method.

Results

Since p53 and TAp73α are able to induce Scotin expression and endoplasmic reticulum (ER) stress [18,20], we evaluated the presence of ER stress during the epidermal differentiation program in relation to the ability of p63 to activate Scotin expression. To investigate the ability of p63 to trigger Scotin expression, we performed RTqPCR experiments (Fig. 1A) using Tet-On cell lines which can be induced to express TAp63α, ΔNp63α or p53 (positive control) in the absence of interfering endogenous p53. Saos-2 cells are p53 null and the expression of endogenous p63 is only weakly detectable with RT-PCR techniques. This allows complete exclusion of the contribution of p53 and minimizes the effect of endogenous p63 expression. The results confirmed the ability of p53 [18] and TAp63α to induce Scotin expression. ΔNp63α does not induce Scotin expression; rather ΔNp63α appears to reduce endogenous Scotin mRNA (Fig. 1A). This data has been confirmed by immunofluorescence on TAp63α transfected Saos-2 which shows Scotin expression only in cells where expression of TAp63 is clearly detectable (e.g. transfected; Fig. 1B). The induced Scotin appears as aggregates in the ER indicating the presence of ER stress, as previously shown [18,20]. These aggregates could be detected following TAp63α expression but not after the induction of ΔNp63α (data not shown).

Fig. 1

TAp63α induces Scotin expression (A) Saos-2 cells stably transfected with doxycycline-inducible vectors containing p53, TAp63α or ΔNp63α were induced with doxycycline for 6 and 12 h. The over-expression of the selected p53 family members were confirmed by Western blot. RTqPCR shows that p53 and TAp63α expression, not ΔNp63α, triggers Scotin induction. (B) Saos-2 cells were transfected (12 and 24 h) with a TAp63 expression vector and were then immunostained for p63 (green) and Scotin (red). In TAp63 transfected cells, Scotin accumulates in the peri-nuclear area (possibly in the Golgi or endoplasmic reticulum). In contrast, p63−/− cells showed cytoplasmic Scotin staining. Bar = 20 μm.

Since ER stress can be confirmed by the evaluation of the presence of the GADD153 transcription factor we tested GADD153 expression in the epithelial HaCat cell line, which expresses endogenous p63. HaCat cells were induced to differentiate, using keratin 10 (expressed in the epidermal suprabasal layer) as a marker of differentiation (Fig. 2A). RTqPCR has been used to demonstrate that the induction of TAp63α expression (Fig. 2B) is accompanied by the induction of Scotin and GADD153 transcripts during the differentiation program. The former gene is upregulated by more than 50% (Fig. 2C) while expression of the latter is more than doubled (Fig. 2D). The Supplementary Fig. 1B clearly demonstrates also that in differentiating HaCat cell line, there is a nuclear relocalization of GADD153. Another marker of terminal differentiation, loricrin, was used to confirm the differentiation of the HaCat cells (S1, lower panels).

Fig. 2

Scotin and GADD153 are upregulated in differentiating HaCat cell line. (A) HaCat cells were induced to differentiate for 0 and 6 days. The Western blot shows increasing expression of keratin 10, compared to that of tubulin, during the differentiation program. (B) Quantitative PCR experiments were performed on RNA extracted from differentiating HaCat cells (0 and 6 days). TAp63α is induced at the transcriptional level together with Scotin (C) and GADD153 (D); the latter gene doubling its expression during the differentiation program.

To further investigate the importance of Scotin during epidermal differentiation we analyzed its expression in the mouse. To facilitate the analysis of Scotin expression, the skin of new born mice has been used, since it is hairless although the epidermis is fully differentiated. The results demonstrate that Scotin is expressed in the upper layers of the epidermis (where TAp63 is present), but is completely absent from the basal layer, where ΔNp63α is highly expressed (Fig. 3A). We used a pan-p63 antibody and ΔNp63α is almost 99% of total p63 expression in the skin; therefore TAp63α expression is not detectable in the upper layers of epidermis with this antibody. Western blots performed on mouse keratinocytes committed to differentiate, as demonstrated by involucrin positivity (Fig. 3C), revealed that the strong upregulation of Scotin during differentiation (Fig. 3B) is accompanied by a reduction in ΔNp63α expression level.

Fig. 3

Scotin and GADD153 expression in mouse skin sections (A) Scotin is expressed only in the suprabasal layer of the epidermis (red), but it is not expressed in the basal layer where ΔNp63 (Ab4 clone) is present (green). (B) Western blot experiments performed on mouse keratinocyte lysates induced to differentiate (0, 1, 3, and 5 days) revealed Scotin upregulation at the protein level. Conversely ΔNp63 expression is progressively reduced compared to tubulin expression, during the differentiation program. (C) Involucrin increases, as expected, its expression in differentiating mouse keratinocytes (from 0 to 5 days). Western blot of tubulin was used as a loading control. Bar = 20 μm.

In order to confirm our data in man, we tested Scotin expression in a primary epithelial cell line (NHEK) induced to differentiate in vitro; RTqPCR analysis showed that Scotin expression doubles by day 5 during differentiation (Fig. 4A), and this phenomenon is clearly paralleled by increased expression of TAp63α (Fig. 4B) with a reduction in ΔNp63α expression level (Fig. 4C). Western blot analysis performed on keratinocyte lysates shows that there is almost a fifty per cent induction of Scotin at the protein level during epidermal differentiation (Fig. 4D and E). Again K10 demonstrates that NHEK cells fully undergo the differentiation process.

Fig. 4

Scotin expression in NHEK cells (A) NHEK cells were induced to differentiate and RNA was used to perform RTqPCR. Scotin doubles its expression (from 0 to 5 days) in differentiating human keratinocytes. (B) TAp63α is upregulated at the transcriptional level while (C) ΔNp63α is down-regulated as expected in differentiating NHEK cells. (D) Western blot experiments performed on human keratinocyte lysates showed Scotin induction at day 3 and 5 of differentiation. This is paralleled by ΔNp63 down-regulation (Ab4 clone antibody) and keratin 10 upregulation. (E) The densitometric analysis, that is representative of few experiments, shows that Scotin expression is increased by almost 50% in differentiating NHEK cells.

Discussion

Here, we describe a novel p63 target gene induced during epidermal differentiation. Scotin is expressed in the nuclear membrane and in the endoplasmic reticulum. It has also been reported that Scotin triggers ER stress and apoptosis, and is known to be induced by p53 and TAp73 α. Only the TA, but not the ΔNp63 isoform, is able to induce Scotin expression in Saos-2 cells. We have also shown that transfection of TAp63α, not ΔNp63α, in Saos-2 cells is capable of inducing the expression of GADD153 (Supplementary Fig. 1A). In vitro experiments on HaCat cells committed to differentiate show transcriptional induction of both Scotin and GADD153, which was confirmed ex vivo in differentiating mouse keratinocytes, where strong induction of Scotin expression has been reported. The correlation between epidermal differentiation and Scotin expression has also been confirmed in new born mouse skin, where both Scotin and TAp63α are expressed in the suprabasal layer. Conversely Scotin is not detectable in the basal layer where ΔNp63α is highly expressed. This data suggests that the increasing level of TAp63α or the increase in the TAp63α/ΔNp63α ratio in the upper epidermal layers, could activate Scotin expression during epidermal differentiation in vivo. Experiments performed on primary human epidermal cells (NHEK) committed to differentiate confirmed Scotin induction at both transcriptional and protein level. In this system Scotin induction is less evident than in ex vivo mouse keratinocytes. This difference is probably due to a reduced capability/potential of NHEK cells to differentiate in vitro, confirmed by their slower decrease in ΔNp63α expression, compared with mouse keratinocytes.

In conclusion these data demonstrate that during epidermal differentiation there is induction of ER stress as shown by induction of Scotin and GADD153 expression. Scotin is known to activate Caspase 3 in cultured cells. Since Caspase 3 is activated by Notch1 during the commitment of embryonic keratinocytes to terminal differentiation, ER stress could be one upstream signal that leads to this caspase activation.