Transgenic Mice Overexpressing Protein Kinase Cδ in the Epidermis Are Resistant to Skin Tumor Promotion by 12-O-Tetradecanoylphorbol-13-acetate¹

The multistage model of mouse skin carcinogenesis is a useful system in which biochemical events unique to initiation, promotion, or progression can be studied and related to cancer formation (1, 2). TPA,⁴ a component of croton oil, is a potent mouse skin tumor promoter (3). TPA has been used extensively to study skin tumor promotion (1, 2). However, the exact molecular mechanisms of tumor promotion by TPA remain to be defined.

PKC, which is ubiquitous in eukaryotes, is a major intracellular receptor for TPA (4). PKC forms part of the signal transduction system involving the turnover of inositol phospholipids and is activated by DAG, which is produced as a consequence of this turnover (5). On the basis of the structural similarities and cofactor requirements, the PKC isoforms have been grouped into three subfamilies of enzymes: the conventional PKCs (α, βI, βII, and γ), which are dependent on Ca²⁺, PS, and DAG/TPA; the nPKCs (ε, δ, η, and θ), which require only PS and DAG/TPA; and the atypical PKCs (ι/λ and ζ), which retain PS dependence but have no requirement for Ca²⁺ or DAG/TPA for activation [PKCμ, which is usually classified as a nPKC, is not easily grouped with any of the other isoforms (6, 7)].

PKC is an important component of the signal transduction pathways controlling cell proliferation and tumorigenesis (1, 5, 8). The positive correlation between the affinity of different types of phorbol esters for PKC and their tumor-promoting efficacy suggests that the activation of PKC may be a critical step in the promotion of mouse skin tumor formation (9). Also, DAG, an endogenous activator of PKCs, promotes mouse skin tumor formation (10, 11). Taken together, these results implicate PKC activation as an essential step in mouse skin tumor promotion. However, several groups have demonstrated that repeated applications of TPA depress PKC activity and protein levels (12, 13, 14). These results indicate that both loss of PKC activity and degradation of PKC may be important for mouse skin tumor promotion by TPA. Studies on the effect of TPA treatment on epidermal PKCδ levels have produced variable results. Leibersperger et al. (15) detected a reduction in PKCδ protein levels after a single TPA treatment, as well as in TPA-induced papillomas. However, others observed little change in the total PKCδ protein levels in the epidermis or papillomas after single or repeated TPA treatments (12, 16, 17). Despite the lack of effect of a single treatment or repeated TPA treatments on epidermal PKCδ protein levels, the PKCδ kinase activity was significantly reduced by these treatments (12). Thus, a reduction in the level of epidermal PKCδ protein or activity may be important for tumor formation.

The distinct role that each individual PKC isoform plays in the signal transduction pathways to mouse skin tumor promotion by TPA is under extensive investigation but remains ambiguous. To evaluate the in vivo role of PKCδ in mouse skin tumor promotion by TPA, we generated transgenic mice overexpressing an epitope-tagged PKCδ (T7-PKCδ) under the control of the human keratin 14 promoter. The T7-PKCδ mice were healthy and fertile and exhibited only mild phenotypic alterations in the absence of TPA treatment. However, the T7-PKCδ mice were found to be extremely resistant to mouse skin tumor promotion by TPA.

TPA was purchased from Alexis Corp. (San Diego, CA). DMBA was purchased from Aldrich Chemical Company, Inc. (Milwaukee, WI). The anti-T7-Tag HRP-conjugated antibody was purchased from Novagen (Madison, WI). The restriction endonucleases were purchased from New England Biolabs (Beverly, MA). Genescreen was purchased from NEN Life Science Products (Boston, MA). The random-primed DNA labeling was purchased from Boehringer Mannheim (Indianapolis, IN). The acrylamide, bisacrylamide, SDS, 0.45 μm supported nitrocellulose membrane, Bio-Rad Protein Assay, Bio-Rad DC Protein Assay, and SDS-PAGE standards were purchased from Bio-Rad Laboratories (Hercules, CA). Monoclonal antibodies to PKCδ, ε, λ, μ, θ, and ζ were purchased from Transduction Laboratories (Lexington, KY). Polyclonal antibodies to PKCβI, βII, γ, ε, η, and actin were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). An affinity-purified anti-PKCα was purchased from Life Technologies, Inc. (Grand Island, NY). The immobilized protein A/G agarose was purchased from Pierce (Rockford, IL). The ECL and ECF Western blotting detection reagents and the protein kinase C enzyme assay system were purchased from Amersham Life Sciences Inc. (Arlington Heights, IL). FVB/NTacfBR mice (FVB/N), 7–9 weeks of age, were purchased from Taconic (Germantown, NY).

The BglII/SalI fragment from the pET-21c(+) vector, containing the T7-Tag open reading frame, was ligated to the NH₂ terminus of the mouse PKCδ cDNA in the pRSV-PKCδ vector to produce the pRSV-T7-PKCδ vector (18). The T7-PKCδ cDNA was ligated into the pGEM3Z-K14 β-globin vector by cloning fragments of T7-PKCδ from pRSV-T7-PKCδ into pGEM-I and pGL2-Basic multicloning sites to place BamHI compatible sites at the 5′ and 3′ ends of the of T7-PKCδ cDNA. These fragments were subcloned into the BamHI site of the pGEM3Z-K14 β-globin vectors to produce the pG3Z-K14-T7-PKCδ vector. The functional elements of pGEM3Z-K14-T7-PKCδ were isolated by partial digestion with EheI and HindIII and purified by electroelution and ion exchange chromatography after agarose gel electrophoresis. The purified K14-T7-PKCδ expression cassette was microinjected into the male pronuclei of one-cell fertilized embryos (FVB/N × FVB/N) by the University of Wisconsin’s Transgenic Mouse Facility. Genomic DNA from tail biopsies was digested with EcoRV, fractionated on a 0.7% agarose gel, and immobilized on a Genescreen. The radiolabeled EcoRV/BamHI fragment from pGEM3Z-K14 β-globin vector, encompassing ∼1 kb of the K14 promoter and the entire β-globin intron, was implemented as the probe for detection of integrated K14-T7-PKCδ DNA. The transgene was detected by Southern blot analyses using genomic DNA digested with EcoRV. An EcoRV/BamHI fragment from the pGEM3Z-K14 β-globin vector labeled with ³²P using the random-primed DNA labeling kit from Boehringer Mannheim was used as a probe.

Transgenic mice were maintained by mating transgenic siblings or mating transgenic mice with wild-type FVB/N mice. The generation of mice for the tumor promotion experiments was performed by mating F₁ and F₂ hemizygous T7-PKC δ mice with wild-type FVB/N mice. At 7–9 weeks of age, the dorsal skins of the mice were shaved 3–4 days before treatment, and those mice in the resting phase of their hair cycle were used for experimentation. The solutions of TPA and DMBA were prepared in acetone and were applied to the shaved backs of individual mice in a volume of 0.2 ml.

Mouse skin tumors were induced by the initiation-promotion regimen. For mouse skin tumor initiation, a single 100-nmol dose of DMBA in 0.2 ml of acetone was applied topically to the shaved backs of both male and female, line 16 and line 37, wild-type and transgenic mice. Two weeks after initiation, 5 nmol of TPA in 0.2 ml acetone or acetone alone was applied twice weekly to skin for the duration of the experiment (25 weeks). The tumor incidence and burden was observed weekly starting at 7 weeks of TPA promotion. The number of mice at 25 weeks of promotion for each experimental group was as follows: line 16 DMBA-TPA (wild type females, n = 15; transgenic females, n = 18; wild-type males, n = 15; transgenic males, n = 14); DMBA-TPA line 37 (wild type females, n = 19; transgenic females, n = 7; wild-type males, n = 11; transgenic males, n = 11); DMBA-acetone line 16 (wild-type females, n = 18; transgenic females, n = 20; wild-type males, n = 18; transgenic males, n = 15); and DMBA-acetone line 37 (wild-type females, n = 16; transgenic females, n = 9; wild-type males, n = 19; transgenic males, n = 14). Mice were removed from the experiment if they were wounded from fighting. Carcinomas were recorded grossly as downward-invading lesions, a subset of which was examined histologically, and malignancy was confirmed as invading the panniculus carnosus. Carcinoma-bearing mice were killed shortly after diagnosis. The Wilcoxan rank sum test was used to determine the value of P for the papilloma burden.

Mice were shaved and depilated before experimentation. The mouse skin was excised and scraped to remove the s.c. fat. The skin was either pulverized with a mortar and pestle under liquid N₂, or the epidermis was removed and homogenized. The ground skin or epidermis was homogenized with five volumes of PKC extraction buffer [20 mm Tris-HCl (pH 7.4), 0.3% Triton X-100, 2 mm EDTA, 10 mm EGTA, 0.25 m sucrose, 1 mm DTT, 10 μg/ml leupeptin, and 10 μg/ml aprotinin]. The homogenate was centrifuged at 100,000 × g for 60 min at 4°C, and the supernatant was the total PKC extract. Protein concentration in the total PKC extract was determined, and 50–100 μg of total PKC extract protein were fractionated on a 7.5 or 10% SDS-polyacrylamide gel. The proteins were transferred to 0.45 μm supported nitrocellulose membrane. The membrane was then incubated with the appropriate primary and secondary antibodies, and the detection signal was developed with ECL or ECF reagents (Amersham). Monoclonal antibodies to PKC α, βII, γ, δ, ε, θ, μ, λ, and ζ and T7-Tag or a polyclonal antibody to PKC η were used to detect the respective PKC isoforms at dilutions of 1:200, 1:1000, 1:500, 1:500, 1:500, 1:500, 1:2000, 1:500, 1:1000, 1:2000, and 1:1000, respectively.

The dorsal skin of mice was shaved and depilated 24 h before experimentation. The mice were euthanized, the dorsal skin was removed, and the epidermis was scraped off on ice with a razor blade. The epidermis from two to three mice were pooled and placed in 0.5 ml of IP lysis buffer [50 mm HEPES (pH 7.5), 150 mm NaCl, 10% glycerol, 1% Triton X-100, 1.5 mm MgCl₂, 10 μg/ml aprotinin, 10 μg/ml leupeptin, 1 mm phenylmethylsulfonyl fluoride, 200 μm Na₃VO₄, 200 μm NaF, and 1 mm EGTA], homogenized in a glass Teflon tissue homogenizer, and agitated for 30 min at 4°C. After centrifugation for 15 min at 14,000 rpm, the supernatant was used for immunoprecipitation. The lysate was preabsorbed with 5 μl of protein A/G agarose for 10 min at 4°C. Five μg of anti-T7-Tag antibody and 10 μl of protein A/G agarose were added to the lysate, and the volume of the lysate was adjusted to 1 ml with lysis buffer. The mixture was incubated for 2–4 h at 4°C with agitation. The immunoprecipitate was pelleted at 14,000 rpm in a microcentrifuge, washed, and resuspended in 300 μl of assay buffer [50 mm Tris-HCl (pH 7.4), 5 mm EDTA (pH 8.0), 10 mm EGTA (pH 7.9), 0.3% β-mercaptoethanol, 5 μg/ml aprotinin, 5 μg/ml leupeptin, and 50 μg/ml phenylmethylsulfonyl fluoride]. Twenty-five μl of the immunoprecipitate were assayed in kinase buffer containing 50 mm Tris-HCl (pH 7.4), 8 mm MgCl₂, 0.136 mm ATP, 0.2 μCi [γ-³²P]ATP, 100 μm EGFR peptide (ERKRTLRRL), 3 mm DTT, 34 μg/ml of Lα-phosphatidyl-l-serine, 3 μg/ml TPA, and 1 mm EGTA. The reaction was incubated at 37°C for 15 min, stopped with 10 μl of 300 mm H₃PO₄, spotted onto filter discs, washed with 75 mm H₃PO₄, and counted.

Mice were euthanized 72 h after the last TPA treatment. The uninvolved epidermis or papillomas were removed from four mice and pooled and placed in IP lysis buffer. The cleared supernatant was used for immunoprecipitation. The T7-PKCδ protein was precipitated with 5 μg of anti-T7-Tag antibody for 2–4 h at 4°C. The pelleted immunoprecipitate was resuspended in assay buffer and assayed as described above.

The tissue to be examined was excised promptly after euthanasia and immediately placed in 10% neutral buffered formalin. The tissue was fixed for 1 h in the formalin and then embedded in paraffin. Sections of 4 μm were cut for H&E staining or immunostaining. Deparaffinized slides were used for immunostaining.

We used the FVB/N inbred mouse strain for transgenesis. The FVB/N mice are an excellent strain for the production of transgenic mice (19). A detailed TPA dose-response study was performed with the FVB/N mice to determine the appropriate dose of TPA for mouse skin tumor promotion. The mice were initiated by topically applying a single 100-nmol dose of DMBA to the dorsal skin and then were treated twice weekly with different doses of TPA. The first papillomas appeared after 8 weeks of promotion with 5 and 10 nmol of TPA at incidences of 4 and 33%, respectively. After 11 weeks of promotion, mice treated with 2 nmol of TPA developed papillomas with an incidence of 4%. At the end of 19 weeks of promotion with 10, 5, or 2 nmol of TPA, the incidence was 100, 94, and 50%, respectively (Fig. 1,A). The papilloma burden was 15.0 ± 0.8, 6.6 ± 0.8, and 0.8 ± 0.2 at 19 weeks of promotion with 10, 5, or 2 nmol of TPA, respectively (Fig. 1,B). No papillomas were observed after treatment of the initiated skin with the vehicle acetone or 0.1 or 1.0 nmol of TPA (Fig. 1 B). The FVB/N mice also exhibited a high incidence of carcinomas. The first carcinomas appeared at 15 weeks with 10 nmol of TPA. At the end of 19 weeks of tumor promotion, the carcinoma incidence was 92, 52, and 4% with 10, 5, or 2 nmol of TPA, respectively (data not shown).

As a prelude to the analysis of the role PKCδ in skin tumor promotion, we determined the PKC isoform composition in the FVB/N mouse skin. PKCα and PKCβII were the only conventional Ca²⁺-dependent isoforms identified in the FVB/N skin extracts. The presence of PKCβI in the skin could not be confirmed because of the cross-reactivity of the anti-PKCβI antibodies with PKCβII (data not shown). The nPKC isoforms δ, ε, η, θ, and μ were all detected in the mouse skin. The atypical PKC isoforms λ and ζ were also detected in skin extracts from FVB/N mice (Fig. 1 C).

The human keratin 14 promoter was used to direct expression of PKCδ to the basal cells of the epidermis (20). The human keratin 14 promoter has been successfully used to overexpress several proteins in the mouse epidermis such as TGF-α, tumor necrosis factor-α, and keratinocyte growth factor (21, 22, 23). The mouse cDNA for PKCδ, tagged at its NH₂ terminus with 21 amino acids from the T7 bacteriophage major coat protein, was inserted at the BamHI site of the pGEM3Z-K14 β-globin vector to create the pGEM3Z-K14-T7-PKCδ vector (Fig. 2 A). This construct expressed active T7-PKCδ protein in in vitro PKC immunocomplex kinase assays after transient transfection into cultured cells (data not shown).

The linear K14-T7-PKCδ expression cassette was removed from the pGEM3Z-K14-T7-PKCδ by digestion with EheI and HindIII (Fig. 2,A). Transgenic mice were generated from one-cell embryos that were microinjected with the K14-T7-PKCδ expression cassette (University of Wisconsin Biotechnology Center Transgenic Animal Facility). A Southern analysis of this DNA was performed by treatment of the genomic DNA with the endonuclease EcoRV and by hybridizing the fractionated, membrane-bound DNA with a probe containing ∼1 kb of the 3′ end of the K14 promoter and the β-globin intron from the pG3Z-K14-β globin vector. The Southern analysis identified nine potential founder mice (Fig. 2 B).

The founders were bred to wild-type FVB/N mice to produce F₁ mice that were examined for expression of the transgene by immunoblot analysis (Fig. 2,C). Individual adult F₁ carrier mice from each line of potential T7-PKCδ mice were euthanized, the skin was removed, and total Triton X-100 soluble PKC extracts were made. Expression of the transgene was detected by immunoblotting with a monoclonal antibody to the T7 epitope of T7-PKCδ or a monoclonal antibody to PKCδ. Two mouse lines with high (line 16) and low (line 37) T7-PKCδ expression levels were selected for further expansion (Fig. 2 C). Offspring of founders 3 and 32 also exhibited expression of the transgene and a similar phenotype but were not expanded for further study (data not shown).

The relative levels of immunoprecipitable T7-PKCδ kinase activity in the epidermis of transgenic mice (lines 16 and 37) are shown in Fig. 2,D. In this experiment, the epidermis was scraped from the skin of two mice and homogenized in a 1% Triton X-100 lysis buffer, and the soluble fraction was used for immunoprecipitation with the anti-T7-Tag antibody. The immunoprecipitates were assayed for PKC activity in the presence or absence of the PKC activators PS and TPA. The average level of PS- and TPA-stimulated T7-PKCδ kinase activity in line 16 mice was 3-fold greater than line 37 mice. Male and female mice had similar T7-PKCδ kinase activity within each transgenic line (Fig. 2 D). Immunoblots of the total extracts probed with the anti-PKCδ antibody displayed an 8-fold increase in PKCδ protein levels in line 16 mice and a 2-fold increase in line 37 mice when compared with the endogenous level of PKCδ protein (data not shown).

The expression of transgenes regulated by the human K14 promoter was not limited to the epidermis. Tissue samples were isolated from a transgenic line 16 F₁ mouse and were extracted with a PKC extraction buffer containing 0.3% Triton X-100. Probing the membrane-bound PKC extracts with the anti-T7 antibody identified expression of T7-PKCδ in the thymus and stomach in addition to the skin (Fig. 2 E). Expression of the K14 promoter has been observed previously in these tissues (22).

Neither transgenic line 16 nor line 37 exhibited any significant phenotypic abnormalities (Fig. 3,A). The presence of the transgene did not affect litter size or the sex ratio of the litters. Neither pups nor adult mice exhibited any gross phenotypic abnormalities in the dorsal epidermis. H&E-stained dorsal skin sections demonstrated no difference in the morphology of the wild-type and the transgenic mouse epidermis (Fig. 3, C and D). The only phenotype that was observed in all four of the lines was hyperkeratosis of the tail. H&E-stained sections from the untreated tail skin of wild type and T7-PKCδ mice were examined. T7-PKCδ mice exhibited hyperplasia of the tail epidermis (Fig. 3, D and E). The onset of the hyperkeratosis in the tail epidermis occurred around 2–4 weeks of age. This phenotype exhibited an incomplete penetrance, which varied over time. Adult mice monitored weekly over a 2-month period exhibited an average penetrance of 38 and 40% in line 16 females and males, respectively. In line 37, the penetrance of the hyperkeratosis was 25 and 50% for females and males, respectively. The hyperkeratosis was temporary, persisting from <1 week to several weeks, depending on the mouse.

Overexpression of T7-PKCδ in the epidermis may induce compensatory alterations in the levels of other PKC isoforms that could modulate the effects of elevated PKCδ levels. The relative levels of various PKC isoforms in the epidermis of wild-type and line 16 T7-PKCδ mice were determined by immunoblot analysis. The protein levels of epidermal PKCα, ε, λ, μ, and ζ were not affected in T7-PKCδ mice (Fig. 4). However, the levels of PKCβII and PKCθ protein were reduced in T7-PKCδ mice. Additionally, the level of PKCη protein was elevated in the epidermis from T7-PKCδ mice (Fig. 4).

The T7-PKCδ transgenic mice were evaluated for their sensitivity to TPA skin tumor promotion. In these experiments, transgenic T7-PKCδ mice and wild-type littermates were initiated by application of a single 100-nmol dose of DMBA and were then treated twice weekly with 5 nmol of TPA in acetone. No papillomas were observed in any of the mice initiated with DMBA and promoted with acetone alone (data not shown). The reduction in both papilloma and carcinoma burden for both male and female T7-PKCδ line 16 mice was dramatic (Fig. 5, B and D; Table 1). The papilloma burden was reduced relative to the wild-type mice by 72% in the male T7-PKCδ mice and by 74% in the female mice after 25 weeks of tumor promotion (Fig. 5,D; Table 1). A second experiment with the line 16 mice elicited similar reductions in papilloma burden in the transgenic mice (data not shown). The line 16 T7-PKCδ mice also exhibited a delay in tumor incidence relative to the wild-type control mice (Fig. 5,C; Table 1). Both male and female FVB/N mice surpassed 50% papilloma incidence by the eighth week of promotion, whereas the T7-PKCδ mice did not exceed 50% incidence until weeks 12 and 13 of promotion for females and males, respectively. Tumor incidence of 100% was achieved by the tenth week of promotion for both male and female wild-type mice. In contrast, the female T7-PKCδ mice did not reach 100% incidence over the 25 weeks of tumor promotion, whereas the male T7-PKCδ mice reached 100% incidence only after 23 weeks of promotion (Fig. 5,C; Table 1). The T7-PKCδ line 37 mice, which expressed 4-fold less T7-PKCδ than line 16, were tested for their responsiveness to mouse skin tumor promotion by TPA. In the experiment with the line 37 mice, both male and female transgenic mice exhibited no significant reduction in papilloma burden or incidence compared with the wild-type mice (Fig. 5, E and F).

At the end of 25 weeks of tumor promotion in the T7-PKCδ line 16 mice, the carcinoma incidence in the female mice was 78 and 37% for the wild-type and transgenic mice, respectively. Similarly, the carcinoma incidence for the males in this experiment was 45% for the wild-type mice and 7% for the transgenic mice. With respect to the wild-type controls, the average latency period for carcinoma development was delayed by 3 weeks for both male and female T7-PKCδ line 16 transgenic mice. Carcinoma development in the tumor promotion experiment with the T7-PKCδ line 37 mice, as compared with wild-type littermates, was not consistently altered by the presence of the transgene compared with wild type controls (data not shown).

To determine whether papilloma formation in the T7-PKCδ mice required the loss of T7-PKCδ, the levels of T7-PKCδ activity and total PKCδ protein were examined at the end of the 25 weeks of tumor promotion. Seventy-two h after the last treatment, four mice from each treatment group were euthanized, and the dorsal, uninvolved epidermis and papillomas were removed. The samples were solubilized in IP lysis buffer, immunoprecipitated with the anti-T7-Tag antibody, and assayed for PKC activity with an EGFR peptide substrate. In two separate experiments, T7-PKCδ kinase activity could be detected in extracts from all of the treatment groups (Table 2). A comparison of the T7-PKCδ kinase activity in the dorsal epidermis from acetone or TPA-treated mice revealed that TPA treatment modestly reduced the level of immunoprecipitable T7-PKCδ kinase activity after chronic TPA treatment (Table 2). Additionally, the level of T7-PKCδ kinase activity in papillomas as compared with the levels in the uninvolved TPA-treated epidermis was not consistently altered.

These extracts were also examined for the levels of total PKCδ protein present in each treatment group (Fig. 6). The level of PKCδ remained constant in the epidermis of wild-type mice even after 25 weeks of biweekly TPA treatment. In the epidermal extracts from the TPA-treated T7-PKCδ mice, minor reductions in the level of total PKCδ were observed with respect to the levels of PKCδ in the acetone-treated controls. Expression of endogenous PKCδ in papillomas from wild-type mice were consistently reduced to extremely low levels as compared with the levels in the uninvolved epidermis of TPA-treated mice. Unlike papillomas examined from wild-type mice, total PKCδ protein was maintained in papillomas from T7-PKCδ mice at variable levels as compared with the surrounding, uninvolved epidermis (Fig. 6).

PKCδ has been shown to be an important component of the signaling pathways involved in the regulation of cell growth, differentiation, and apoptosis (24). Overexpression of PKCδ in NIH 3T3 cells and CHO cells reduced their growth rates (25, 26). Furthermore, expression of a constitutively active PKCδ inhibited colony formation of both wild-type and c-Ha-ras-transformed NIH 3T3 cells (27). Additionally, the transformation of c-Src overexpressing rat fibroblasts could be dramatically enhanced by the inhibition of PKCδ using pharmacological inhibitors or a dominant-negative PKCδ mutant (28). Specific cleavage of PKCδ by caspase-3 occurs during UV or ionizing radiation-induced apoptosis. This indicates that PKCδ may play a central role in the execution of apoptosis (29, 30). However, conclusive evidence defining the in vivo role of individual PKC isoforms in tumor promotion has been elusive (1, 8). Using directed overexpression of PKCδ in the mouse epidermis, we found that the PKCδ isoform inhibited tumor promotion by TPA. Thus, PKCδ may be a key regulator of mouse skin tumor promotion by TPA.

The expression of elevated levels of PKCδ in the epidermis of FVB/N mice did not alter the normal regulation of skin development or differentiation. The lack of phenotypic alterations may have occurred for several reasons. The endogenous PKCδ is readily detectable in the mouse epidermis (15, 16). This suggests that a high level of PKCδ protein may be required for normal epidermal differentiation and that the PKCδ elevation attained in our transgenic lines was not sufficient to disrupt epidermal differentiation. The T7-PKCδ transgene exhibited little constitutive activity; thus, activation of T7-PKCδ would have been dependent on endogenous activators, like DAG. The level of DAG, or other activators, is most likely regulated upstream of T7-PKCδ (5). Hence, the rate-limiting step in PKCδ signaling in normal epidermal homeostasis may not be at the level of PKCδ protein but at the level of activators present.

The only abnormality observed in the T7-PKCδ mice was the sporadic hyperproliferation and hyperkeratosis of the tail epidermis. This type of localized phenotypic alterations has been observed by several investigators who have used this heterologous human K14 promoter for transgene production (21, 31). Mice overexpressing TGF-α from the human K14 promoter exhibited thickening of the epidermis in the ear, tail, footpads, rectal epithelium, foreskin, scrotum, and vaginal epithelium (21). Targeted expression of the human papillomavirus type-16 genome with the K14 promoter led to ear and face hyperplasia with dysplasia and papillomatosis with complete penetrance (31). A simple explanation for these regional alterations in these transgenic mouse lines may be that the affected sites are areas prone to mechanical irritation (21, 32). The sporadic nature of the dry tail phenotype of the T7-PKCδ mice suggests that irritation may elicit this phenotype. Other investigators have suggested mechanical irritation as the probable cause for focal, regional phenotypes in mice harboring epidermal transgenes (21, 31, 32) Another potential cause for this restricted phenotype could be regional skin sensitivities to the different transgenes. This has been observed with the EGF treatment of adult mice, resulting only in the hyperproliferation in the epidermis of the tail and the footpads of treated mice (33).

Recent investigations have demonstrated cross-regulation between different PKC isoforms. Overexpression of PKCα in lymphoid cell lines specifically induces a reduction in PKCδ mRNA and protein (34). Analysis of the PKC signaling cascade, resulting in the activation of the c-fos promoter in mouse mammary epithelial cells, revealed that PKCλ was upstream of PKCε and PKCζ was downstream of PKCε (35). Thus, overexpression of an individual PKC isoform may affect the regulation of other isoforms. Using immunoblot analysis, we found the levels of epidermal PKCα, ε, λ, μ, and ζ were unaltered by overexpression of PKCδ, whereas the levels of PKCβII and PKCθ were reduced and PKCη was elevated. Thus, overexpression of PKCδ does not induce major alterations in the PKC isoform composition of the epidermis. The specific modulation PKCβII, η, and θ, but not the other isoforms, suggests that the effects of elevated epidermal PKCδ are specific and not an artifact of the model system.

The results of mouse skin tumor promotion with the T7-PKCδ mice were dramatic. The T7-PKCδ line 16 mice averaged a 73% reduction in papilloma burden for both male and female mice. The appearance of tumors on the T7-PKCδ line 16 mice was also delayed by 4 weeks on average. The carcinoma incidence was also reduced in the line 16 mice. The low number of papillomas and the delay in their formation in the T7-PKCδ line 16 mice suggest that a second event may need to occur to allow the formation of papillomas in the line 16 mice. The reduction in carcinoma incidence may be a direct effect of PKCδ on papilloma progression or simply the result of the reduced papilloma burden. Additional studies need to be performed to determine whether this is a direct or indirect effect of PKCδ overexpression. The T7-PKCδ line 37 mice exhibited lower levels of T7-PKCδ protein and activity compared with line 16 and did not display any alterations in sensitivity to mouse skin tumor promotion. The inhibition of tumor promotion in the line 16 mice and the lack of alterations in the line 37 mice imply that the inhibition of papilloma formation by treatment with DMBA/TPA requires a threshold level of PKCδ activity.

Reduced levels of PKCδ protein and activity have been observed after single or multiple TPA treatments of the mouse skin (12, 15). The level of PKCδ protein was also found to be significantly reduced in papillomas (15). However, these reductions in PKCδ protein levels were not consistently observed (12, 16, 17). The level of T7-PKCδ activity and total PKCδ protein levels were analyzed at the end of the 25-week carcinogenesis experiment to determine whether loss of PKCδ was important for papilloma formation in the wild-type and transgenic mice. In the T7-PKCδ line 16 mice, the level of T7-PKCδ kinase activity in the TPA-treated epidermis was only modestly reduced compared with acetone-treated transgenic mice. T7-PKCδ kinase activity was detected in papillomas from transgenic mice, and the level of activity did not change consistently compared with the activity in the surrounding, uninvolved epidermis. However, examination of the PKCδ protein levels in wild-type mice demonstrated a consistent loss of endogenous PKCδ protein in the papillomas from wild-type mice. In contrast, PKCδ was detected in papillomas from T7-PKCδ mice at variable levels. The observed loss of detectable PKCδ from wild-type papillomas implies that decreased PKCδ protein normally facilitates papilloma formation. The retention of T7-PKCδ activity in papillomas formed in transgenic mice suggests that the loss of T7-PKCδ activity is not necessary for the formation of papillomas on the transgenic mice. The observed delay in papilloma incidence and reduction in papilloma burden suggests that a secondary event was needed for papilloma formation in the transgenic mice. Activation of other signal transduction pathways may allow the cells to overcome the inhibition of papilloma formation by elevated PKCδ levels.

The ability of PKCδ to suppress papilloma formation implies that its activation may block epidermal keratinocyte proliferation and/or transformation. These results are consistent with the role of PKCδ in in vitro cell culture studies that have shown PKCδ to be an inhibitor of cell growth and transformation. Overexpression of PKCδ inhibited the growth of fibroblasts, vascular smooth muscle cells, and endothelial cells by delaying passage through different phases of the cell cycle, depending on the cell type (25, 26, 36, 37). Inhibition of vascular smooth muscle cell proliferation by elevated PKCδ levels correlated with decreased expression of cyclins D1 and E (36). In the mouse epidermis, TPA-induced proliferation correlated with up-regulation of both cyclin D1 and cyclin E (38). Additionally, the homozygous deletion of cyclin D1 reduced the papilloma burden in response to DMBA-TPA treatment of the mouse skin (39). These proteins may be important mediators of the TPA response mediated by PKCδ in the mouse epidermis (Fig. 7). To begin to examine the role of PKCδ in the regulation of epidermal proliferation, the effect of a single treatment with 5 nmol of TPA on epidermal hyperplasia was examined over a 48-h period. No differences in hyperplasia were detected between wild-type and transgenic epidermis after TPA treatment (data not shown). Thus, the effects of PKCδ overexpression in the skin may be more complex than simply blocking cell proliferation; however, further study will be required to prove this.

Transient cotransfections of PKCδ with AP-1-regulated reporter genes in NIH3T3 cells have demonstrated that PKCδ can activate the AP-1 response elements (27). Furthermore, the activation of AP-1 by PKCδ positively correlated with its inhibition of activated c-Ha-ras-mediated transformation of NIH3T3 cells. In mouse keratinocytes, induction of terminal differentiation by TPA correlates with the activation of AP-1 (40, 41). Induction of terminal differentiation in initiated cells would prevent their expansion into papillomas. Thus, activation of AP-1 may also be important for tumor suppression in T7-PKCδ mice (Fig. 7).

Introduction of an activated c-Ha-ras gene into primary mouse keratinocytes leads to a block in their ability to differentiate in response to elevated Ca²⁺ and increased susceptibility to papilloma formation in skin grafts (42, 43). The transformation of keratinocytes by c-Ha-ras overexpression correlates with increased levels of TGF-α and reduced PKCδ activity (44, 45, 46). The reduction in PKCδ activity coincides with tyrosine phosphorylation of PKCδ by the Src tyrosine kinases c-Src or c-Fyn. This appears to be mediated by TGF-α activation of the EGFR signaling pathway (46, 47). Because activation of the c-Ha-ras locus is the initiating event in ∼90% of the papillomas formed on DMBA-treated skin, down-regulation of PKCδ by this mutation may be important for papilloma formation (48). Elevation of the intracellular level of PKCδ may prevent complete suppression of PKCδ activity. This retention of PKCδ activity may then be able to prevent the development of papillomas from initiated cells (Fig. 7).

Activation of PKC by TPA must impart a positive growth signal to promote tumor formation in wild-type mice. Because PKCδ appears to inhibit papilloma formation, other isoforms may be transducing this positive signal. This supports the hypothesis that individual PKC isoforms have unique roles in cell growth regulation. Furthermore, the ability of PKCδ to inhibit papilloma formations suggests that PKCδ may truly be an in vivo suppressor of tumorigenic transformation.

We thank Ashok Rajput, Sarah Bourguinon, Rachelle Stenzel, and Kathy Helmuth for excellent technical support. We also thank Dr. Elaine Fuchs at the University of Chicago for providing the pGEM3Z-K14 β-globin vector.