Protein Kinase C (PKC) δ Suppresses Keratinocyte Proliferation by Increasing p21Cip1 Level by a KLF4 Transcription Factor-dependent Mechanism*

PKCδ increases keratinocyte differentiation and suppresses keratinocyte proliferation and survival. However, the mechanism of proliferation suppression is not well understood. The present studies show that PKCδ overexpression increases p21Cip1 mRNA and protein level and promoter activity and that treatment with dominant-negative PKCδ, PKCδ-siRNA, or rottlerin inhibits promoter activation. Analysis of the p21Cip1 promoter upstream regulatory region reveals three DNA segments that mediate PKCδ-dependent promoter activation. The PKCδ response element most proximal to the transcription start site encodes six GC-rich DNA elements. Mutation of these sites results in a loss of PKCδ-dependent promoter activation. Gel mobility supershift and chromatin immunoprecipitation reveal that these DNA elements bind the Kruppel-like transcription factor KLF4. PKCδ increases KLF4 mRNA and protein level and KLF4 binding to the GC-rich elements in the p21Cip1 proximal promoter. In addition, KLF4-siRNA inhibits PKCδ-dependent p21Cip1 promoter activity. PKCδ increases KLF4 expression leading to enhanced KLF4 interaction with the GC-rich elements in the p21Cip1 promoter to activate transcription.

PKC isoforms have also been implicated in the regulation of keratinocyte proliferation (13,20,23,(33)(34)(35). This role is particularly important, because keratinocyte differentiation is associated with cessation of proliferation, and it would make mechanistic sense to have a common kinase activate both processes. A limited number of studies have examined the mechanism of nPKC regulation of keratinocyte proliferation. For example, the nPKC isoform PKC⑀ binds to and activates Fyn, a Src kinase, and this is associated with reduced keratinocyte proliferation (36). PKC⑀ forms a complex with cyclin E-cdk2-p21 Cip1 leading to phosphorylation of p21 Cip1 and cdk2 inhibition to reduce proliferation (37). In addition, PKC␣ suppresses proliferation in raft cultures of human keratinocytes by a mechanism that involves increased expression of p21 Cip1 (17).
This realization that PKC regulates cell proliferation points to the intriguing possibility that a single regulatory cascade may both increase differentiation and suppress proliferation, key processes that must proceed together during keratinocyte maturation. However, we have limited understanding regarding the mechanism of growth suppression. p21 Cip1 is an important regulator of cell cycle progression, and increased p21 Cip1 expression is associated with cessation of cell proliferation (38). Moreover, p21 Cip1 has been implicated as a key suppressor of proliferation progression in human keratinocytes (39 -44). We now describe a novel mechanism whereby PKC␦ suppresses keratinocyte proliferation. Our studies indicate that PKC␦ increases expression of the Kruppel-like factor KLF4, which interacts at GC-rich DNA elements in the proximal p21 Cip1 promoter to activate p21 Cip1 gene expression leading to cessation of cell proliferation.
Promoter Activity-For p21 Cip1 promoter activity analysis, 0.5 g of p21 Cip1 promoter reporter plasmid was mixed with 1 l of FuGENE 6 reagent diluted with 99 l of KSFM. The mixture was incubated at 25°C for 15 min and then added to a 50% confluent culture of primary human epidermal keratinocyte maintained in 2 ml of KSFM in a 9.6-cm 2 dish. For co-transfection experiments, with 0.5 g of p21 Cip1 promoter reporter plasmid and 0.5 g of PKC␦ expression plasmid were mixed, treated with FuGENE 6, and added to cells as indicated above. After 24 h, the cells were harvested, and the extracts were prepared for assay of luciferase activity.
Electroporation and siRNA-mediated Knockdown-Keratinocytes were electroporated with siRNA or plasmids using the Amaxa electroporator and the VPD-1002 nucleofection kit (Germany). For electroporation, keratinocytes were harvested with trypsin and replated 1 day prior to use. On the day of electroporation, 1 ϫ 10 6 of the replated cells were harvested with trypsin and resuspended in KSFM. The cells are collected at 2000 rpm, washed with 1 ml of sterile phosphate-buffered saline (pH 7.5), and suspended in 100 l of keratinocyte nucleofection solution. The cell suspension, which included 3 g of gene-specific siRNA, was mixed by gentle pipetting and electroporated using the T-018 settings. Warm KSFM (500 l) was added, and the suspension was transferred to a 21.3-cm 2 cell culture dish containing 3.5 ml of KSFM. When required, the cells were electroporated a second time with luciferase reporter or expression plasmid. This was accomplished by harvesting the cells with trypsin and resuspension in KSFM. The cells were collected, washed with PBS, and resuspended in nucleofection solution as above. The nucleofection suspension, which included 2 g of plasmid, was electroporated using the T-018 settings. The cells were plated and maintained for various times before the extracts were prepared for assay. Our electroporation method delivers nucleic acid reagents with greater than 90% efficiency.
Immunological Analysis-Equivalent amounts of protein were electrophoresed on a 4 -15% denaturing polyacrylamide gradient gel and transferred to nitrocellulose. The membranes were blocked, incubated with a specific primary antibody, washed, and exposed to an appropriate horseradish peroxidaseconjugated secondary antibody. Chemiluminescent detection was used to visualize secondary antibody binding. To assess intracellular p21 Cip1 distribution, total and nuclear extracts were prepared using NE-PER nuclear and cytoplasmic extraction reagent (Pierce). To monitor DNA synthesis, the cells were treated for 24 h with stimulus, and during the final 6 h prior to harvest, the cells were incubated with 10 M BrdU. BrdU uptake was monitored in fixed cells by indirect immunofluorescence.
p21 Cip1 mRNA Half-life-To analyze p21 Cip1 mRNA decay kinetics, keratinocytes were infected with 15 MOI of Ad5-EV or Ad5-PKC␦ for 24 h prior to the addition of 5 g/ml actinomycin D. At 0, 0.5, 1, 2, 3, and 4 h after actinomycin D addition, RNA was isolated and analyzed for p21 Cip1 and cyclophilin A mRNA content by quantitative RT-PCR using primers described in the previous section. The values for each mRNA level at each time point are the means Ϯ S.D. derived from triplicate quantitative RT-PCRs of independent samples. First order decay constants (k) were determined by nonlinear regression analysis (PRISM v3.03; GraphPad) of plots measuring the percent p21 Cip1 mRNA remaining versus time of actinomycin D treatment. p21 Cip1 mRNA decay constants were calculated based on the means Ϯ S.D. of n of independent time course experiments where n ϭ 3, permitting pair-wise statistical assessment using the Student's t test. Differences were considered significant if p Ͻ 0.05.
Gel Mobility Shift and Supershift-Human keratinocytes were infected Ad5-EV or Ad5-PKC␦ adenovirus (15 MOI), and at 24 or 48 h, the cells were washed with phosphate-buffered saline, and nuclear extract was prepared using NE-PER nuclear and cytoplasmic extraction reagent (Pierce). Binding of transcription factors to p21 Cip1 promoter sites was monitored by gel mobility shift assay. Five microgram of nuclear extract was incubated for 40 min at room temperature in a volume of 20 l containing 20 mM HEPES (pH 7.5), 10% glycerol, 50 mM KCl, 2 mM MgCl 2 , 0.5 mM EDTA, 0.5 mM DTT, 1 g/ml poly(dI-dC), 0.1 mg/ml bovine serum albumin, and 50,000 cpm radioactive, double-stranded, 32 P-end labeled Sp1(1/2) site oligonucleotide (5Ј-GGAGGGCGGTCCCGGGCGGCGC-3Ј, which encodes the Sp1 and Sp2 binding sites. The Sp1-1 and Sp1-2 binding sites from the human p21 Cip1 promoter region are indicated in bold and italic type (GenBank NG_009364). Although not shown, identical results were observed using a probe encoding p21 Cip1 promoter Sp1 sites 3, 4, 5, and 6 (5Ј-GGTCCCGCCTC-CTTGAGGCGGG CCCGGG CGGGGCGGTT-3Ј). For competition studies, a fixed molar excess of nonradioactive competitor oligonucleotide was added to the DNA binding reaction. For gel mobility supershift assay, 2 g of Sp1, Sp3, or KLF4 factor-specific antibody was added to the reaction mixture and incubated at 4°C for 3 h. The 32 P-labeled probe was then added, and the incubation was continued for an additional 40 min at 25°C Protein-DNA complexes were resolved in 4% polyacrylamide gels under nondenaturing conditions.
ChIP Assay-ChIP assays were conducted as described (52) with minor modification. Briefly, human keratinocytes were cross-linked with 1.42% formaldehyde at room temperature for 15 min followed by quenching with 125 mM glycine. The cells were collected by centrifugation, and cell pellets were washed twice with ice-cold PBS containing histone deacetylase inhibitors. Cross-linked cells were collected by centrifugation and lysed in 150 l of lysis buffer (50 mM Tris-HCl, pH 8, 10 mM EDTA, 1% SDS, 1 mM PMSF, 20 mM sodium butyrate, and protease inhibitors). The samples were chilled on ice, and DNA was sheared using a Branson Sonifier (three 30-s pulses on ice at 40% amplitude with 30 s between pulses) to produce fragments of 1,000 bp. Radioimmune precipitation assay buffer (400 l) containing protease inhibitors and histone deacetylase inhibitors was added, and the chromatin was centrifuged at 12,000 ϫ g for 10 min. Aliquots were used for generating input DNA and for antibody immunoprecipitation. ChIP grade antibodies were added to Dynabeads protein A and incubated for 2 h at 4°C with rotation at 40 rpm. Sheared chromatin was added, and mixture was incubated at 4°C overnight with rotation. The chromatin-antibody complex was washed twice with radioimmune precipitation assay buffer, and 40 l of Chelex 100 slurry (10% w/v) was added to the washed beads, and samples were boiled for 10 min. The samples were then treated with proteinase K for 30 min at 55°C, and the samples were boiled for 10 min. Enrichment of KLF4-associated DNA sequences in immunoprecipitated samples and input samples was detected by quantitative RT-PCR using sequence specific primers and LightCycler 480 SYBR Green I Master mix. ChIP primers were as follows: p21 Cip1 proximal promoter located at nucleotides Ϫ150/Ϫ4 (5Ј-GCTGGGCAGCCAGGAGCCTG-3Ј, 5Ј-CTG-CTCACACCTCAGCTGGC-3Ј) and p21 Cip1 distal promoter located at nucleotides Ϫ827/Ϫ673 (5Ј-CTGCTGCAACCAC-AGGGATTTCTT-3Ј, 5Ј-TGTTGATTGTCACATGCTT-CCGGG-3Ј).

PKC␦ Increases p21 Cip1 mRNA Level and Promoter Activity-
Although overexpression of PKC␦ in epidermis in vivo suppresses cell division (53), few studies have assessed the underlying mechanism. The present studies focus on p21 Cip1 as a PKC␦ target and potential mediator of growth suppression. Fig.  1A shows that PKC␦ expression produces a concentration-dependent increase in p21 Cip1 promoter activity. Rottlerin, a PKC␦ inhibitor, inhibits this increase, but Go6976, an inhibitor of PKC␣ and ␤, does not (Fig. 1B). We next tested the impact of dominant-negative PKC␦ and PKC␦-siRNA on p21 Cip1 promoter activity. As shown in Fig. 1 (C and D), expression of dominant-negative PKC␦ or treatment with PKC␦-siRNA reduces p21 Cip1 promoter basal activity. The immunoblot in Fig. 1E confirms a PKC␦-siRNA-dependent partial knockdown of PKC␦. We have previously shown that PKC␦-Tyr-311 phosphorylation is required for optimal PKC␦ activity (31). We therefore tested the ability of the PKC␦-Y311F mutant to increase p21 Cip1 promoter activity. As shown in Fig. 1F, PKC␦-Y311F produces a reduced increase in promoter activity as compared with wild-type PKC␦. Taken together, these findings suggest that PKC␦ regulates p21 Cip1 expression.
To assure that this regulation is biologically meaningful, we monitored the impact of PKC␦ expression and knockdown on endogenous p21 Cip1 mRNA and protein level. PKC␦ expression increases ( Fig. 2A) and PKC␦ knockdown reduces p21 Cip1 mRNA and protein level (Fig. 2B). To further make the case that this is transcriptional regulation and is not due to changes in mRNA stability, we monitored the impact of increased PKC␦ expression on p21 Cip1 mRNA turnover. Although PKC␦ expression substantially increases p21 Cip1 mRNA ( Fig. 2A), p21 Cip1 mRNA half-life is not changed in cells expressing basal versus elevated levels of PKC␦ (t1 ⁄ 2 ϭ 2.2 in Ad5-EV cells versus 1.94 in Ad5-PKC␦ cells) (Fig. 2C). Taken together, these studies strongly argue that PKC␦ controls p21 Cip1 level via transcriptional mechanisms. It is interesting that PKC␦-siRNA reduces p21 Cip1 protein level by more than 80%, indicating that tonic PKC␦ activity is required to maintain endogenous p21 Cip1 expression.
PKC␦ Stimulates Nuclear Accumulation of p21 Cip1 -The above studies indicate that PKC␦ drives p21 Cip1 expression; however, it is important to assess the impact on activity. Nuclear accumulation is an index of p21 Cip1 activity (54). Fig.  3A, shows that the total p21 Cip1 level increases, and this is associated with increased nuclear levels. Although the increase in nuclear content is largely a response to the overall increase in p21 Cip1 level, there is also an impact on subcellular distribution. This is apparent when cells are stained to assess p21 Cip1 location. We detect cytoplasmic p21 Cip1 in 96% of Ad5-EV-infected cells. This cytoplasmic staining is evident in the upper panels of Fig. 3B. In contrast, in Ad5-PKC␦ infected cells, cytoplasmic staining is observed in only 1% of cells (lower panels). This consistent observation suggests that in addition to increasing p21 Cip1 level, PKC␦ promotes movement into the nucleus. As shown in Fig. 3C, these changes are associated with a reduction in BrdU incorporation from 45.2% in empty virus-infected to 1.3% in PKC␦-expressing cells, indicative of reduced cell proliferation in the presence of increased PKC␦.
PKC␦ Regulation of p21 Cip1 Promoter Activity-To study the mechanism of the PKC␦-dependent increase in p21 Cip1 expression, we sought to identify PKC␦ response elements in the p21 Cip1 promoter. The p21 Cip1 promoter upstream regulatory region encodes six GC-rich sites in the proximal promoter that are reported to bind Sp1 transcription factors (Sp1, Sp2, etc.), and two p53 binding sites in the distal promoter (Fig. 4A) (55). We constructed a p21 Cip1 promoter truncation series and monitored the ability of PKC␦ to activate expression of each construct. As shown in Fig. 4B, PKC␦ produces a three-peak pattern of promoter activation. To our knowledge, this is a novel pattern of activation. PKC␦ response regions are located at nucleotides Ϫ251/Ϫ60, Ϫ2001, and Ϫ2326. This complicated FIGURE 1. PKC␦ activates p21 Cip1 gene expression. A, human keratinocytes were transfected with the 0.5 g of p21-2326 luciferase reporter plasmid and the indicated level of PKC␦ expression vector. After 24 h, the cells were harvested, and lysates were assayed for luciferase activity. B, PKC␦ activation of p21 Cip1 promoter activity is reduced by PKC␦ inhibitor. Keratinocytes were transfected with 0.5 g of p21-2326 and 0.5 g of PKC␦ expression vector and after 24 h were treated with rottlerin or Go6976. After an additional 24 h, extracts were prepared for assay of luciferase activity. C, keratinocytes were transfected with 0.5 g of p21-2326 and the indicated concentration of empty vector (EV) or dominant-negative (dn) PKC␦-encoding plasmid, and after 24 h, extracts were prepared for assay of luciferase activity. D, keratinocytes were electroporated with 3 g of control or PKC␦-siRNA/1 ϫ 10 6 cells. At 48 h, the cells were electroporated with 2 g of endotoxin-free p21-2326, and after an additional 24 h, extracts were prepared and assayed for luciferase activity. E, keratinocytes were treated with control or PKC␦-siRNA as indicated above, and after 48 h, extracts were prepared for detection of PKC␦ and ␤-actin. A similar reduction was observed at 72 h (not shown). F, keratinocytes were transfected with 0.5 g of p21-2326 and 0.5 g of plasmid encoding PKC␦-wt or PKC␦-Y311F, and after 24 h, extracts were prepared for luciferase activity assay. PKC␦-wt and PKC␦-Y311F are expressed at identical levels (not shown) (31). All of the values are the means Ϯ S.D., and similar results were observed in a minimum of three experiments.
pattern suggests that the p21 Cip1 promoter encodes multiple PKC␦-responsive elements. The fact that these regions are separated by nonresponsive regions suggests the presence of interspersed enhancer and suppressor elements. This arrangement is typical of complex promoters (56). In the present manuscript, we focus on the element located within nucleotides Ϫ251/Ϫ60.
Role of the GC-rich DNA Elements-As noted in Fig. 4A, the p21 Cip1 promoter encodes six GC-rich DNA elements, previously characterized as Sp1 factor binding sites (Sp1-1, Sp1-2 etc.), clustered in the proximal promoter between nucleotides Ϫ120/Ϫ50 (46). We assessed whether these sites are required for PKC␦-dependent regulation. We challenged full-length p21 Cip1 promoter constructs encoding GC-rich site mutations with PKC␦. Fig. 4C shows that mutation of the Sp1-1, Sp1-2, Sp1-3, or Sp1-4 sites partially reduces PKC␦-stimulated activity but that elimination of all six Sp1 sites, Sp1(⌬1-6), is required to eliminate the response. This suggests that GC-rich element interacting proteins may drive PKC␦-dependent p21 Cip1 transcription. A number of transcription factors and co-factors have been described as interacting at the proximal p21 Cip1 promoter to regulate transcription, including Sp1 and Sp3 (55), the ␤-helix-loop-helix factor, AP4 (57), and the histone deacetylase, p300 (55). Kruppel-like transcription factors also interact at these sites (58). Fig. 5A shows that PKC␦ expression increases p21 Cip1 protein level, and the level of hKLF4. In contrast, there is no change in Sp1, Sp3, AP4, or p300 level. The hKLF4 increase is associated with a parallel increase in KLF4 mRNA (Fig. 5B).
Transcription Factor Interaction with GC-rich Elements-These findings suggest that KLF4 may mediate the PKC␦-dependent increase in p21 Cip1 promoter activity. We therefore FIGURE 2. PKC␦ controls p21 Cip1 mRNA and protein level. A, keratinocytes were infected with 15 MOI of Ad5-EV or Ad5-PKC␦, and after 24 h, extracts were prepared for detection of PKC␦ and p21 Cip1 mRNA and protein. The mRNA abundance values are the means Ϯ S.D. B, PKC␦ knockdown reduces p21 Cip1 mRNA and protein level. Keratinocytes were electroporated with 3 g of the indicated siRNA, and after 48 h, the PKC␦ and p21 Cip1 mRNA and protein levels were monitored. Similar results were observed in three separate experiments. C, PKC␦ does not alter p21 Cip1 mRNA half-life. Keratinocytes were infected with 15 MOI of Ad5-EV or Ad5-PKC␦ for 24 h. RNA was harvested for measurement of p21 Cip1 mRNA level at 0 -4 h after the addition of 5 g/ml actinomycin D (actD). The log linear plot was used to determine first order decay constants and half-life as described under "Experimental Procedures." FIGURE 3. PKC␦ affects p21 Cip1 nuclear level. Keratinocytes were infected with 15 MOI of Ad5-EV or Ad5-PKC␦ adenovirus. A, at 24 h after Ad5-PKC␦ infection, total and nuclear extracts were prepared to detect p21 Cip1 . We electrophoresed four times more total than nuclear extract (based on cell equivalents). B, at 24 h post-infection, the cells were fixed and stained with anti-p21 Cip1 (green) and Hoechst (nucleus, blue). The p21 Cip1 panels are merged p21 Cip1 /Hoechst images. Cytoplasmic p21 Cip1 staining was observed in 96 Ϯ 3% of Ad5-EV and 1 Ϯ 0.5% of Ad5-PKC␦-infected cells (mean Ϯ S.D., n ϭ 300). assessed whether KLF4 interacts with the GC-rich p21 Cip1 proximal promoter-binding elements. Fig. 6A reveals a shifted band when 32 P-Sp1(1/2) probe is incubated with nuclear extract. An important finding is that the mobility of this band increases in extracts prepared from PKC␦-expressing cells, suggesting an altered composition or structure of the binding complex. Accumulation of this faster migrating band is time-dependent; it is variably present in cells treated for 24 h with PKC␦ but is always present when cells are exposed to PKC␦ for 48 h (Fig.  6A). Fig. 6B shows appropriate self-competition of radioinert probe against 32 P-Sp1(1/2), which indicates that the binding is specific. We next monitored Sp1, Sp3, and KLF4 interaction with 32 P-Sp1(1/2) by gel mobility supershift. Incubation of the extract with anti-KLF4 results in a loss of the supershifted band. This suggests that KLF4 interacts at this transcription site (Fig.   6C). In contrast, we could not demonstrate interaction of Sp1 or Sp3 by this method. We next assessed whether these proteins interact with the proximal promoter region (Ϫ150/Ϫ4) using chromatin immunoprecipitation. This method is extremely useful, because it can demonstrate interaction in situ in intact cells (59). Fig. 6D shows low level KLF4 interaction in empty virus-infected cells and a 2.5-fold increase in this interaction in PKC␦-expressing cells. In contrast, KLF4 does not bind to the Ϫ827/Ϫ673 region, which lacks GC-rich binding sites.
Impact of KLF4 on p21 Cip1 Promoter Activity-We next assessed the functional impact of KLF4 on p21 Cip1 promoter activity. Cells were transfected with the p21 Cip1 promoter reporter and full-length hKLF4, a hKLF4 mutant lacking the zinc finger domain, hKLF(1-388), and a mutant encoding only the zinc finger domain, hKLF4(335-470) (58). Treatment with hKLF4(1-470) (wild type) increases transcription, but the inactive mutants, hKLF4(1-388) and hKLF4(335-470), do not (Fig.  7A). hKLF4(1-470) and hKLF4(1-388) were confirmed to be expressed at similar levels by immunoblot, thereby confirming that the difference p21 Cip1 promoter activation is not due to a difference in KLF4 expression (Fig. 7A). Because anti-KLF4 binds to KLF4 within amino acids 1-180, expression of hKLF4(335-470) could not be confirmed. To further confirm the biological relevance of this response, we show that the level of hKLF4 that we delivered to the cells in these experiments is comparable with that observed following stimulation with PKC␦ (Fig. 7A). In addition, simultaneous expression of hKLF4(1-470) with PKC␦ augments the PKC␦-dependent increase (Fig. 7B). Consistent with these findings, reducing KLF4 level with siRNA reduces p21 Cip1 protein level and p21 Cip1 promoter activity (Fig. 7C). Moreover, KLF4 knockdown is not associated with changed PKC␦ level, suggesting that the reduced p21 Cip1 promoter activity is not due to feedback reduction of PKC␦ level.
We next examined whether KLF4 activation of p21 Cip1 promoter activity requires the proximal GC-rich elements. KLF4 . PKC␦ activation of p21 Cip1 promoter involves multiple elements. A, schematic of p21 Cip1 promoter upstream regulatory region showing the p53 and GC-rich (Sp1) (nucleotides Ϫ120/Ϫ50) response elements (25). The distances are in nucleotides relative to the transcription start site. B, keratinocytes were transfected with 0.5 g of each p21 Cip1 promoter luciferase reporter plasmid in the presence of 0.5 g of PKC␦ or empty (ϪPKC␦) expression vector. After 24 h, the cells were harvested, and extracts were assayed for luciferase activity. C, keratinocytes were transfected, incubated, and assayed for luciferase activity as above. The respective plasmid names identify the mutated Sp1 sites (e.g. ⌬1, ⌬2, ⌬1-6, etc.). We observed similar results in each of four experiments. expression increases activity of the full-length wild-type p21 Cip1 promoter (p21-2326) (Fig. 8B). Basal and KLF4-stimulated promoter activity is increased for some (⌬2) and decreased for other (⌬1, ⌬3, and ⌬4) Sp1 site deletion mutants. However, deletion of all six sites (⌬1-6) results in a complete loss of KLF4-dependent regulation.
KLF4 Is Required for PKC␦-dependent Growth Suppression and p21 Cip1 Induction-We also examined the impact of KLF4 knockdown on the PKC␦-dependent increase in p21 Cip1 and reduction in cell number. Fig. 9A shows that PKC␦ overexpression increases hKLF4 level and that this increase is suppressed by hKLF4-siRNA. Consistent with our previous studies, keratinocyte cell number increases 2.5-fold during a 3-day growth time (Fig. 9B). This increase is enhanced by treatment with KLF4-siRNA and suppressed by PKC␦ overexpression. In addition, knockdown of KLF4 partially reverses the PKC␦-dependent reduction in cell number. The PKC␦-dependent growth suppression is associated with a PKC␦-dependent increase in p21 Cip1 mRNA. Treatment with hKFL4-siRNA partially attenuates this increase (Fig. 9C). To further confirm the role of PKC␦ and hKLF4 in regulating p21 Cip1 , we show that treatment with PKC␦-siRNA reduces KLF4 and p21 Cip1 protein level and that p21 Cip1 expression can be restored by virus-mediated expression of hKLF4 (Fig. 10).
Various studies indicate that PKC␦ also regulates keratinocyte proliferation. PKC␦ expression in HaCaT keratinocytes reduces cell proliferation (23), and PKC␦ also increases keratinocyte susceptibility to apoptosis (23,32,61). However, little information is available regarding the mechanism whereby PKC␦ suppresses proliferation. In the present report, we identify that p21 Cip1 is a PKC␦ target. BrdU uptake is an index of cell progression through the S phase and therefore cell proliferation. We show that PKC␦ expression reduces the percentage of BrdU-positive cells from 45 to 1.3%. This change is associated with a substantial increase in p21 Cip1 mRNA and p21 Cip1 protein. Rottlerin, which inhibits the PKC␦ isoform, suppresses this increase. Because rottlerin inhibits other targets in addition to PKC␦ (77), we also show that treatment with PKC␦-siRNA and expression of dominant-negative PKC␦ suppresses the response. In contrast, treating with PKC␣/␤ inhibitor, Go6976, does not reduce the increase, suggesting that PKC␣ and ␤ do not regulate p21 Cip1 expression. Phosphorylation of PKC␦ at tyrosine 311 is associated with increased activity (31). We show that the phosphorylation-defective PKC␦ mutant, PKC␦-Y311F, has reduced ability to increase p21 Cip1 promoter activity, further suggesting it has a key role. We further show that p21 Cip1 accumulates in the nucleus of PKC␦-expressing cells, suggesting that PKC␦ may stimulate nuclear translocation.
KLF4 Mediates the PKC␦ Stimulus-Sp1 factors and p53 are important regulators of p21 Cip1 expression (45,55). Two p53 response elements are located in the distal p21 Cip1 promoter, and six GC-rich elements are present in the proximal promoter (nucleotides Ϫ140/Ϫ60) (46,55). Although p53 induces p21 Cip1 expression in some systems (45), initial studies indicated that the increase we observe is not mediated by p53 (not shown). However, promoter truncation and mutagenesis experiments indicate that the GC-rich elements are required for regulation. These elements interact with Sp1 and Kruppel-FIGURE 6. KLF4 interacts with GC-rich elements in the p21 Cip1 proximal promoter. Keratinocytes were infected with 15 MOI Ad5-EV or Ad5-PKC␦. A, after 24 or 48 h, nuclear extracts were prepared, incubated with Sp1(1/2)-32 P, and electrophoresed on a 4% acrylamide nondenaturing gel. The short arrow indicates the major band, and the long arrow indicates the fast mobility band observed in cells expressing elevated PKC␦ levels. B, control extracts was prepared as above and incubated with Sp1(1/2)-32 P in the presence of the indicated molar excess of radioinert probe. The arrow indicates band migration. C, extracts prepared as above were incubated with IgG, Sp1, Sp3, and KLF4 specific antibodies for 3 h at 4°C, for 40 min with Sp1(1/2)-32 P and then electrophoresed. D, ChIP assay was performed as described under "Experimental Procedures" using p21 Cip1 promoter-derived PCR primers encoding the indicated range of nucleotides. NE, nuclear extract; EV, empty vector; FP, free probe.
like transcription factors (46,58). We observe increased Kruppel-like factor 4 expression and binding to the GC-rich elements in PKC␦-expressing cells. Moreover, wild-type KLF4 increases p21 Cip1 promoter activity, but activity is not regulated by inactive KLF4 mutants. We had anticipated that Sp1 factors may be important in this regulation, but Sp1 and Sp3 did not interact with the p21 Cip1 promoter GC-rich elements.
KLF4 is an interesting member of the Kruppel-like factor family of regulators (78). KLF4 is unusual compared with other Kruppel-like factors in that it can activate or suppress transcription in a context-dependent manner (78). KLF4 has a role in epidermis where it is expressed in suprabasal cells (79). Its presence is essential for epidermal barrier formation (79), because KLF4 knock-out mice do not form a competent barrier and do not survive (80), and mice that overexpress KLF4 in epidermis display enhanced barrier formation (81). KLF4 also has a role in regulating proliferation in tissues such as the oral epithelium (79) and colon (82). However, only limited information is available regarding the mechanism of this regulation. In  ). At 24 h, the cells were harvested, and extracts were assayed for luciferase activity. To monitor KLF4 levels, extracts were prepared at 24 h post-transfection, and hKLF4 was detected by immunoblot. To compare the level of KLF4 in KLF4-and PKC␦-overexpressing cells, keratinocytes were infected with 15 MOI of the indicated virus or transfected with 0.5 g of hKLF4 expression plasmid, and after 24 h, extracts were prepared for KLF4 immunoblot. B, keratinocytes were transfected with 0.5 g of luciferase reporter, 0.5 g of hKLF4(1-470) expression plasmid, and 0.5 g of PKC␦ expression vector. At 24 h, cell extracts were prepared for luciferase assay. C, keratinocytes were electroporated with 3 g of the indicated siRNA and at 48 h transfected with 0.5 g of luciferase reporter plasmid. After an additional 24 h, extracts were prepared for luciferase activity assay and immunoblot. the colonic epithelium, KLF4 reduces proliferation by suppressing cyclin D1 level (82) via a mechanism that involves Sp1 binding to the cyclin D1 gene promoter (83). Our studies are novel in that the link PKC␦ activation to increased KLF4 binding to the p21 Cip1 promoter and increased p21 Cip1 expression. It is important to note that multiple regulatory mechanisms control p21 Cip1 level in cells (45,45,55). Our present studies suggest an important role for KLF4 in mediating the PKC␦-dependent increase in p21 Cip1 . However, it would be naïve to suggest that KLF4 is the only transcription factor that mediates this regulation, and it is very likely that other transcription factors have a context-dependent role.

Coordinate Regulation of Keratinocyte Proliferation and
Differentiation-Previous studies indicate that a PKC␦, Ras, MEKK1, MEK3, and p38␦/ERK cascade increases AP1 and Sp1 factor expression and nuclear accumulation and that these factors bind to elements in the involucrin promoter to drive gene expression (Fig. 11) (27,28,32,63,66). Our present studies suggest that this same cascade increases KLF4 expression and binding to the p21 Cip1 promoter to increase p21 Cip1 expression and thereby suppress cell proliferation (Fig. 11). This regulation is particularly interesting, because it indicates a mechanism whereby PKC␦ can coordinately control keratinocytes differentiation and proliferation. This appears to make mechanistic sense because cessation of proliferation and initiation of differentiation must occur simultaneously during epidermal development. These findings are also interesting because a limited number of studies have examined the impact of PKC isoforms on keratinocyte proliferation. PKC⑀ binds to and activates Fyn, a Src kinase, and this is associated with reduced keratinocyte proliferation (36). PKC⑀ also forms a complex with cyclin E-cdk2-p21 Cip1 leading to phosphorylation of p21 Cip1 and cdk2 inhibition to reduce proliferation (37). In addition, PKC␣ suppresses proliferation in raft cultures of human keratinocytes via a mechanism that involves increased expression of p21 Cip1 (17). The present studies identify a new mechanism whereby PKC␦ regulates both proliferation and differentiation. Additional studies will be necessary to gain insight regarding the molecular details of this regulation.   . PKC␦ coordinate regulation of keratinocytes differentiation and proliferation. PKC␦ initiates a cascade the coordinately increases keratinocyte differentiation and suppresses keratinocyte proliferation. Active PKC␦ increases AP1 and Sp1 transcription factor level and nuclear accumulation. These factors bind to the distal regulatory region of the hINV promoter to drive transcription and increase involucrin expression (i.e. differentiation) (29,66). Active PKC␦ also increases KLF4 level and binding to GC-rich elements in the p21 Cip1 proximal promoter where it acts to increase p21 Cip1 expression to suppress proliferation.