Recent studies have identified roles for C/EBPβ in cellular survival and tumorigenesis, however, the mechanisms through which C/EBPβ regulates these processes are not fully understood. Previously, we demonstrated that C/EBPβ−/− mice are resistant to carcinogen-induced skin tumorigenesis and in response to topical carcinogen treatment display a 17-fold increase in keratinocyte apoptosis compared to wild-type mice. Here, we have investigated the mechanisms through which C/EBPβ regulates apoptosis in response to carcinogenic stress. Analysis of carcinogen-treated C/EBPβ−/− mouse skin revealed a striking increase in the number of p53 immunopositive keratinocytes in the epidermis of C/EBPβ−/− mice compared to wild-type mice and this increase was temporally associated with a concomitant anomalous increase in apoptosis. The increased levels of p53 were functional as Mdm2, Bcl-2, C/EBPα and p21 were differentially regulated in the epidermis of carcinogen-treated C/EBPβ−/− mice. The increase in p53 protein was not associated with an increase in p53 mRNA levels. To determine whether p53 is required for the increased apoptosis in C/EBPβ−/− mice, C/EBPβ/p53 compound knockout mice were generated. Carcinogen-treated C/EBPβ/p53 compound knockout mice did not display increased apoptosis demonstrating p53 is required for the proapoptotic phenotype in C/EBPβ−/− mice. Our results demonstrate that altered keratinocyte survival in C/EBPβ−/− mice results from aberrant regulation of p53 protein and function and indicate C/EBPβ has a role in the negative regulation of p53 protein levels in response to carcinogen-induced stress.
CCAAT/enhancer binding proteins (C/EBPs) are members of the basic leucine zipper (bZIP) class of transcription factors and are involved in fundamental processes including cell growth, differentiation, survival, inflammation, immune response and metabolism (Ramji and Foka, 2002). C/EBPβ is expressed in many tissues, including epidermis (Maytin and Habener, 1998; Oh and Smart, 1998) where it plays a role in the early stages of stratified squamous differentiation (Zhu et al., 1999). Recent studies have identified a role for C/EBPβ in tumor development (Zhu et al., 2002), senescence (Sebastian et al., 2005), cell survival (Buck et al., 2001; Zhu et al., 2002; Wessells et al., 2004; Li et al., 2005) and cellular transformation (Zhu et al., 2002). In terms of cell transformation, C/EBPβ can cooperate with Ras12V to transform NIH3T3 cells (Zhu et al., 1999) and more recently, it has been shown that cell-cycle-dependent phosphorylation of C/EBPβ on Ser64 and Thr189 is required to promote Ras-induced transformation of NIH3T3 cells (Shuman et al., 2004).
C/EBPβ-deficient mice are completely refractory to skin tumorigenesis induced by carcinogens such as 7,12-dimethylbenz[a]anthracene (DMBA) that produce skin tumors containing oncogenic Ras mutations (Zhu et al., 2002). Importantly, apoptosis is dramatically elevated in epidermal keratinocytes of DMBA-treated C/EBPβ-null mice compared to similarly treated wild-type mice (Zhu et al., 2002) suggesting that C/EBPβ regulates cell survival by suppressing apoptosis in response to DNA damage/oncogenic stress. Studies utilizing epidermis-specific C/EBPβ knockout mice (keratin 5-Cre;C/EBPβfl/fl mice) demonstrated that C/EBPβ exerts a keratinocyte intrinsic role in cell survival in response to carcinogen treatment (Sterneck et al., 2005).
C/EBPβ is important for the survival of numerous cell types in response to various stressors. For example, C/EBPβ has a prosurvival function in Myc/Ras transformed macrophages through an autoregulatory pathway involving IGF-1 (Wessells et al., 2004). C/EBPβ also has a prosurvival/antiapoptosis function in CCl4-treated hepatic stellate cells (Buck et al., 2001). In Wilms’ tumors, or nephroblastomas, C/EBPβ is expressed at high levels in relapsing and metastatic tumors and blocking C/EBPβ expression in a Wilms’ tumor cell line results in spontaneous apoptosis (Li et al., 2005). Increased expression of C/EBPβ is associated with human breast, colorectal and ovarian tumorigenesis (Sundfeldt et al., 1999; Rask et al., 2000; Bundy and Sealy, 2003). In addition, cyclin D1 can interact with C/EBPβ to alter gene expression and there is evidence that this interaction is required for the unique patterns of gene expression observed in human cancers that overexpress cyclin D1 (Lamb et al., 2003). Collectively, these studies indicate that C/EBPβ has a significant role in the proliferation/survival of a variety of human and rodent tumor cells.
In the current study we investigated the mechanisms through which C/EBPβ regulates cell survival in response to carcinogenic stress. We observed that altered keratinocyte survival in C/EBPβ−/− mice results from aberrant regulation of p53 protein and function and our results indicate that C/EBPβ has a critical role in the negative regulation of p53 protein levels in response to carcinogen-induced stress.
C/EBPβ−/− mice display a 17-fold increase in apoptosis in epidermal keratinocytes at 24 h following a single topical treatment with DMBA compared to similarly treated wild-type mice (Zhu et al., 2002). To begin to define this apoptotic response in C/EBPβ−/− mice and to determine whether the increase in apoptosis involves p53, a time course for DMBA-induced apoptosis and p53 protein expression in C/EBPβ−/− and wild-type mice was conducted. Apoptotic keratinocytes were identified in hematoxylin and eosin (H&E) stained skin sections and were characterized by the simultaneous presence of pyknotic nuclei, eosinophilic cytoplasm and some loss of cellular attachment. A representative H&E stained skin section from DMBA-treated wild-type and C/EBPβ−/− mice (24 h post-DMBA treatment) is shown in Figure 1a and b. Numerous apoptotic keratinocytes were detected in DMBA-treated C/EBPβ−/− mice while apoptotic cells were scarce in similarly treated wild-type mice. Similar results, both in terms of numbers and location of apoptotic keratinocytes, were obtained in wild-type and C/EBPβ−/− mice utilizing terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling (TUNEL) staining (Figure 1c and d). Within the interfollicular epidermis, most (>95%) of the apoptotic keratinocytes appeared in the basal layer. Within the follicular epidermis, increased numbers of apoptotic keratinocytes were observed in the infundibulum (Figure 1e) and isthmus region of the hair follicles of DMBA-treated C/EBPβ−/− mice and while an increase was also observed in wild-type mice, it occurred to a lesser degree. As shown in Figure 1f, the number of interfollicular basal apoptotic keratinocytes was quantified every 4 h post-topical DMBA treatment up to 32 h in wild-type and C/EBPβ−/− mice. The overall apoptotic response of C/EBPβ−/− mice was significantly different from that of wild-type mice (two factor analysis of variance (ANOVA) P<0.05). Apoptosis was significantly increased in DMBA-treated C/EBPβ−/− mice at 16, 20 and 24 h post-DMBA treatment compared to similarly treated wild-type mice. The number of apoptotic keratinocytes in C/EBPβ−/− mice peaked at 24 h post-DMBA treatment and this number corresponds to ∼2% of total interfollicular basal keratinocytes and represents a 60-fold increase in apoptotic keratinocytes in DMBA-treated C/EBPβ−/− mice. In comparison, the maximum increase in the number of apoptotic keratinocytes in DMBA-treated wild-type mice was only fivefold. After 24 h the number of apoptotic keratinocytes in C/EBPβ−/− mice decreased sharply which is likely due to the clearance (cutaneous/systemic absorption, metabolism and excretion) of topically applied DMBA from the epidermis. At 28 and 32 h the levels of apoptotic keratinocytes were similar in both genotypes.
To determine whether the increased aberrant apoptotic response in the C/EBPβ−/− mice is associated with alterations in p53 levels, immunohistochemical staining for p53 in DMBA-treated wild-type and C/EBPβ−/− mouse skin was conducted. In vehicle treated mice of both genotypes the number of p53 immunostained cells was low and they displayed weak nuclear staining indicating low levels of p53 protein (Figure 2a). However, in DMBA-treated C/EBPβ−/− mice numerous interfollicular basal keratinocytes displayed intense nuclear p53 staining while similarly treated wild-type mice displayed relatively infrequent p53 staining in basal cells of the epidermis (Figure 2a). Increased numbers of p53-positive keratinocytes were also observed in the hair follicles of DMBA-treated C/EBPβ−/− mice compared to similarly treated wild-type mice (Figure 2b). The number of p53-positive basal keratinocytes in DMBA-treated C/EBPβ−/− and wild-type mice was quantified over the 32 h time course (Figure 2c). The response of C/EBPβ−/− mice was significantly different from that of wild-type mice (two factor ANOVA P<0.05). The number of p53-positive interfollicular keratinocytes was significantly increased at 4, 8, 16, 20 and 24 h post-DMBA treatment in C/EBPβ−/− mice compared to wild-type mice. These results indicate that p53 levels are deregulated in C/EBPβ−/− mice in response to DMBA treatment. The time course for the DMBA-induced increase in p53-positive keratinocytes (Figure 2c) and the time course for apoptosis (Figure 1f) in C/EBPβ−/− mice is similar suggesting a causal relationship between the increased p53 levels and apoptosis.
To determine whether the DMBA-induced increase in p53 protein levels in C/EBPβ−/− keratinocytes is associated with increased p53 activity, we examined p53-regulated downstream gene products. Western blot analysis of epidermal lysates prepared from DMBA-treated wild-type and C/EBPβ−/− mice demonstrated that p53 levels were preferentially increased in C/EBPβ−/− epidermis (Figure 3a) confirming our immunohistochemical observations of differential induction of p53 protein. The increase in p53 levels as detected by immunoblot analysis appears modest compared to the increased numbers of p53-positive cells in DMBA-treated C/EBPβ−/− mouse epidermis (Figure 2c). However, it should be noted that epidermal lysates contain protein derived from suprabasal and basal cells that express low p53 protein levels and this produces a large dilution of the p53-positive basal keratinocytes, which account for approximately 7% of basal keratinocytes. Thus, a 10-fold increase in p53 in 7% of the cells would result in less than a -twofold induction upon immunoblot analysis of total epidermal protein. p21 and C/EBPβ are both known to be positively regulated by p53 in response to DNA damage. As shown in Figure 3a, both p21 and C/EBPα protein levels were induced to higher levels by DMBA treatment in C/EBPβ−/− epidermis compared to similarly treated wild-type mice. Immunohistochemical staining for epidermal p21 revealed a significant increase in the number of p21-positive cells in DMBA-treated C/EBPβ−/− mice (Figure 3b) and this increase is similar to that observed for p53 immunopositive cells (Figure 2c). The large increase in p21 immunopositive cells provides additional evidence that immunoblot analysis of total epidermal protein results in a dramatic under representation of the actual changes. Bcl-2 has antiapoptotic activity and is negatively regulated by p53. As shown in Figure 3a, Bcl-2 protein levels were decreased in the epidermis of DMBA-treated C/EBPβ−/− mice compared to wild-type mice (Figure 3a). Collectively, these results demonstrate that the increased p53 levels observed in the epidermis of DMBA-treated C/EBPβ−/− mice are associated with increased p53 function. Our immunohistochemical staining results indicate that a select minor population of keratinocytes within the epidermis is responding to carcinogen treatment with increased p53 protein and p53 target gene expression. Analysis of total epidermal protein/mRNA for p53 and p53 target gene products results in a significant under representation of the response of target cells to carcinogen-induced stress due to a poor signal to noise ratio.
To determine whether the anomalous increase in apoptosis in the C/EBPβ−/− mice is dependent upon p53, we crossed C/EBPβ−/− mice with p53-deficient mice to produce C/EBPβ/p53 compound knockout mice. C/EBPβ/p53 compound knockout mice as well as C/EBPβ−/−, p53−/− and wild-type littermates were treated with a single dose of DMBA and the number of apoptotic keratinocytes was quantified. As shown in Figure 4, DMBA-treated C/EBPβ−/− mice displayed their characteristic elevated levels of apoptosis. In contrast, DMBA-treated C/EBPβ/p53 compound knockout mice displayed low levels of apoptosis similar to that observed in DMBA-treated wild-type mice. Thus, the deletion of p53 in C/EBPβ−/− mice (C/EBPβ−/−; p53−/−) rescues keratinocytes from DMBA-induced apoptosis. These results demonstrate that p53 is an absolute requirement for the anomalous increase in apoptosis in C/EBPβ−/− keratinocytes.
Our results reveal that p53 protein levels are deregulated in C/EBPβ−/− epidermis in response to DMBA and that this deregulation is responsible for the increased aberrant apoptotic response in C/EBPβ−/− mice. However, it is not known how C/EBPβ represses p53 levels in response to carcinogen stress. As C/EBPβ can function as a transcriptional repressor of certain genes we measured p53 mRNA levels with quantitative TaqMan reverse transcription–polymerase chain reaction (RT–PCR). Basal levels of p53 mRNA in epidermis of C/EBPβ−/− mice were ∼2-fold greater than that in wild-type mice suggesting C/EBPβ has a repressor function in the regulation of basal p53 mRNA expression (Figure 5a). However, treatment of wild-type and C/EBPβ−/− mice with DMBA did not significantly induce p53 mRNA in either genotype (Figure 5a) indicating that the increased p53 protein levels in DMBA-treated C/EBPβ−/− epidermis (Figure 2c) are not due to the upregulation of p53 mRNA levels. These results suggest C/EBPβ has a role in the regulation of p53 protein levels independent of regulation of p53 mRNA levels. Mdm2 is a p53 target gene and major negative regulator of p53 protein levels. If C/EBPβ is involved in the positive regulation of Mdm2 levels, then C/EBPβ deficiency would result in decreased Mdm2 levels and increased p53 protein levels. We measured Mdm2 mRNA levels with quantitative TaqMan RT–PCR and conducted immunohistochemical staining for Mdm2 protein. As shown in Figure 5b, Mdm2 mRNA levels were greater in the DMBA-treated mice and we observed increased numbers of Mdm2 immunopositive cells with strong nuclear staining in DMBA-treated C/EBPβ−/− mice compared to similarly treated wild-type mice (Figure 5c). Collectively, these results indicate p53 protein levels and function are increased in C/EBPβ−/− mice despite elevated levels of Mdm2.
Previously, we demonstrated that C/EBPβ−/− mice display elevated levels of apoptosis in response to topical DMBA treatment and that these mice are completely refractory to skin tumorigenesis induced by carcinogens such as DMBA (Zhu et al., 2002). These studies suggested the possibility that C/EBPβ is required for the survival of DMBA-induced tumor precursor cells. Accordingly, C/EBPβ deficiency would result in apoptosis of the tumor precursor cells and account for the resistance to skin tumor development in carcinogen-treated C/EBPβ−/− mice. Thus, understanding the proapoptotic response in C/EBPβ−/− mice has important implications for early tumor development. In the current study we demonstrated that the abnormal increase in apoptosis in DMBA-treated C/EBPβ−/− keratinocytes requires p53 and is due to the aberrant upregulation of p53 levels and function. Our results indicate that in wild-type mice, C/EBPβ suppresses carcinogen-induced apoptosis in keratinocytes through repression of p53 protein levels/function. In the absence of C/EBPβ, carcinogen treatment results in the derepression of p53 protein levels resulting in increased p53 protein levels, increased p53-regulated gene expression and increased C/EBPβ−/− keratinocyte apoptosis. We propose that the proapoptotic phenotype in C/EBPβ−/− mice results from a defect in the ability of C/EBPβ−/− keratinocytes to suppress p53-mediated apoptosis in response to carcinogen-induced stress. DMBA produces oncogenic mutations in Ras in keratinocytes of treated mouse skin. Currently, it is not known whether the increase in DMBA-induced apoptosis in C/EBPβ−/− keratinocytes is in response to DNA damage and/or oncogenic Ras-induced oncogenic stress.
Our results indicate that the aberrant increase in p53 protein levels induced by carcinogen treatment in C/EBPβ−/− epidermis occur independent of changes in p53 mRNA levels. Regulation of p53 protein levels and p53 transcription activity is complex and approximately 10 regulatory feedback loops have been identified (Harris and Levine, 2005). One key regulator of p53 protein levels is Mdm2. Mdm2 negatively regulates p53 protein by mediating nuclear export and subsequent ubiquitylation and proteosomal degradation and also decreases p53 transcriptional activity by direct binding to p53 (Honda et al., 1997; Freedman and Levine, 1998). We initially hypothesized that if C/EBPβ were involved in the positive regulation of Mdm2 expression, then loss of C/EBPβ in DMBA-treated C/EBPβ−/− epidermis would likely result in decreased Mdm2 levels and increased p53 protein levels. However, our results show the opposite, that is, Mdm2 mRNA and protein are increased in carcinogen-treated C/EBPβ−/− mice compared to carcinogen-treated wild-type mice. Thus, despite increased Mdm2 protein in C/EBPβ−/− epidermis there is increased levels of functional p53 as determined by increased p53-target gene expression and apoptosis. These results suggest that the activity of Mdm2 protein is attenuated in C/EBPβ−/− epidermis in response to carcinogen-induced stress. Several mechanisms are known to modulate Mdm2 activity including post-translational modification of p53 and Mdm2 as well as interactions with cellular factors including p19Arf, MdmX and Arf-BP1 (Brooks and Gu, 2006). Acetylation of C-terminal lysines on p53 blocks Mdm2 induced ubiquitylation of p53 and p53 degradation (Ito et al., 2002). ATM and/or ATR-dependent phosphorylation of Mdm2 (Maya et al., 2001) as well as p53 (Shieh et al., 1999) blocks p53-Mdm2 interaction and stabilizes p53. The tumor suppressor p19Arf, promotes Mdm2 degradation, sequesters Mdm2 in nucleolus and blocks nucleocytoplasmic shuttling of Mdm2 resulting in the accumulation of p53 (Kamijo et al., 1998; Pomerantz et al., 1998; Zhang et al., 1998; Tao and Levine, 1999). MdmX can form dimers with Mdm2 and stabilize p53 (Jackson and Berberich, 2000). It is possible that C/EBPβ may regulate genes involved in post-translational modifications of p53 such as those involved in p53 phosphorylation, methylation, acetylation, ubiquitylation or sumoylation all of which can influence the stability and activity of the p53 protein (Appella and Anderson, 2001). Future studies in our laboratory will examine the expression of regulators of p53 and Mdm2 proteins as well as specific post-translational modifications in p53 and Mdm2 in C/EBPβ−/− and wild-type epidermis in response to carcinogen-induced stress.
DNA damage and oncogenic stress, among other types of cellular stress, induce and activate p53 and based on the integration of incoming stress signals, cells will determine whether to undergo growth arrest, senescence or apoptosis (Vogelstein et al., 2000; Vousden and Lu, 2002). At the peak of apoptosis in DMBA-treated C/EBPβ−/− epidermis, we observed that approximately one of four cells with increased p53 levels underwent apoptosis compared to only one of 12 in similarly treated wild-type mice. These results suggest that in response to carcinogen-induced stress the cellular decision in p53-positve C/EBPβ-deficient keratinocytes is heavily weighted in favor of cell death. As discussed above, this result could be due to specific post-translational modifications and/or protein interactions of p53 that enhance its activity and are differentially regulated in the two genotypes.
In terms of basal p53 mRNA expression, we observed that basal p53 mRNA levels were significantly increased in untreated C/EBPβ−/− mouse epidermis compared to untreated wild-type epidermis suggesting that C/EBPβ has a role in the negative regulation of basal p53 mRNA levels. Consistent with this notion are several recent studies that have identified a repressor function for C/EBPβ in the regulation of gene expression (Corbi et al., 2000; Burkart et al., 2005; Di-Poi et al., 2005; Sankpal et al., 2005). Moreover, we have identified putative C/EBPβ binding sites in the p53 proximal promoter. Additional studies are required to determine whether C/EBPβ is a direct repressor of basal p53 expression.
In summary, our results demonstrate that p53 protein levels are deregulated in DMBA-treated C/EBPβ−/− mice and we have provided genetic evidence that p53 is required for the proapoptotic phenotype of C/EBPβ−/− mice. Our results indicate that C/EBPβ has a role in the negative regulation of p53. While additional studies are required to understand how C/EBPβ represses p53, our current study reveals a novel link between C/EBPβ, p53 and apoptosis. C/EBPβ may be a potential molecular target for cancer therapy as blocking C/EBPβ function in a p53 proficient tumor cell may result in apoptosis and tumor regression.
Materials and methods
The C/EBPβ−/− mice used in this study have been described (Sterneck et al., 1997). The C/EBPβ−/− and wild-type mice were generated by mating C/EBPβ+/− females to C/EBPβ+/− males (C57BL/6;129/sv). Male p53−/− mice (C57BL/6) were crossed with female C/EBPβ+/− mice; F1 p53+/−C/EBPβ+/− mice were crossed to generate the four genotypes used experimentally. The mice were genotyped using the following PCR primers: for p53 wild-type (Forward: TGCCCTGTGCAGTTGTGGGT CA, Backward: ATTTCCTTCCACCCGGATAAGATG); p53 knockout (Forward: ATGACTGCCATGGAGGAGT CACAGTC, Backward: TTTACGGAGCCCTGGCGCTC GATGT); C/EBPβ wild-type (Forward: AGCCCCTACCTG GAGCCGCTCGCG, Backward: GCGCAGGGCGAACGG GAAACCG); and C/EBPβ knockout (Forward: GCTCCA GACTGCCTGGGAAAAG, Backward: GGCCCGGCTA GACAGTTACACG). The hair of the dorsal skin of the mice (7–11 weeks old) was clipped with electric clippers at least 2 days before each experiment. Mice were treated with a single dose (400 or 200 nmol) of DMBA in 200 μl acetone or 200 μl acetone alone as vehicle control.
Detection of apoptotic cells/TUNEL
Wild-type and C/EBPβ mice were treated with 400 nmol DMBA (at least three mice/group) and at each time point after treatment the treated dorsal skin was excised and fixed for 24 h in 10% neutral buffered formalin phosphate. Four different areas of dorsal skin were taken, processed, embedded in paraffin, and 5 μm sections were cut and stained with H&E. More than 4000 basal keratinocytes were examined for each mouse. Apoptotic keratinocytes in the interfollicular stratum basale were scored based on the simultaneous presence of the following three criteria: dark pyknotic nuclei, cytoplasmic eosinophilia, and some loss of cellular attachment. Data are presented as the number of apoptotic keratinocytes/cm length of mouse skin. TUNEL assay on skin sections was conducted using DeadEnd Fluorometric TUNEL System (Promega, Madison, WI, USA) following the manufacturer's protocol.
Mouse skins were fixed in 10% neutral buffered formalin phosphate and embedded in paraffin. After deparaffinization, tissue sections (5 μm) were subjected to antigen retrieval with citrate buffer in 95°C degree water bath for 20 min followed by treatment with 0.1% H2O2 to block endogenous peroxidase activity. For Mdm2 immunostaining antigen retrieval was conducted using Dako Antigen Retrieval Buffer in 95°C water bath for 30 min. Sections were incubated with either rabbit polyclonal anti-p53 antibody (1:1000) (Santa Cruz Biotechnology, Santa Cruz, CA, USA, sc-6243), anti-p21 antibody (1:1000) (Santa Cruz Biotechnology sc-471) or anti-mdm2 antibody (1:500) (Santa Cruz sc-965) at 4°C for 20–24 h after blocking with normal goat serum. This was followed by incubation with biotinylated goat anti-rabbit IgG at room temperature for 30 min. Detection was made with Vectastatin Elite ABC kit (Vector Laboratories, Burlingame, CA, USA) and 3, 3′-diaminobenzidine (BioGenex Laboratories, San Ramon, CA, USA) as the chromagen following the manufacturer's protocol. The sections were counterstained with hematoxylin, dehydrated and mounted. p53, p21 and Mdm2-positive basal keratinocytes were scored in 3 sections/mouse for at least 3 mice/group and data are expressed as p53-, p21- and Mdm2-positive cells/cm length of mouse skin.
A two-factor ANOVA with interaction was conducted on the square root transformed numbers of apoptotic cells and on log transformed numbers of p53-positive cells. The factors were genotype and time. Cell numbers were transformed to reduce heterogeneity of the error variances. When interaction was significant, single factor ANOVA was were conducted to compare genotypes at each fixed time point.
Preparation of epidermal homogenates
The hair on the dorsal skin of the mice was clipped with an electric clipper at least 2 days before each experiment. Mice were treated, killed and the shaved dorsal skin was removed. The skin was spread on an index card and immediately immersed in liquid nitrogen. The epidermis was scraped from the dermis with a surgical scalpel and placed in radioimmunoprecipitation assay buffer (1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS), 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 1 protease inhibitor cocktail (Roche, Basel, Switzerland) and 1 mM sodium orthovanadate in phosphate-buffered saline). The samples were sonicated on ice, vortexed and centrifuged at 14 000 g for 10 min at 4°C. Supernatants were stored at −80°C until use. Protein concentration was determined using Bio-Rad DC protein assay reagent.
Equal amounts of protein were dissolved in SDS sample buffer, boiled and separated by SDS–polyacrylamide gel electrophoresis (PAGE). The separated proteins were transferred to an Immobilon-P membrane (Millipore, Billerica, MA, USA). Following incubation in blocking buffer, the membranes were probed with antibody for C/EBPα (sc-61), C/EBPβ (sc-520), p53 (sc-6243), p21 (sc-397) from Santa Cruz, Bcl-2 (554279, Becton Dickinson, San Jose, CA, USA) or β-actin (A5441, Sigma, St Louis, MO, USA). The membranes were washed and then probed with a horseradish peroxidase-linked secondary antibody (Amersham, Piscataway, NJ, USA). Detection was made with an enhanced chemiluminescence reagent (Amersham) followed by exposure of the membrane to film.
Quantitative real-time RT–PCR
Total RNA from DMBA-treated (400 nmol) wild-type and C/EBPβ epidermis was extracted using Tri Reagent (Sigma) 20 h after treatment and purified with RNeasy Mini Kit (Qiagen, Chatsworth, CA, USA). cDNA was synthesized using ImProm-II reverse transcription (RT) system (Promega) following the manufacturer's protocol. cDNA from 50 ng total RNA was used as template to perform quantitative real-time PCR using mouse p53, Mdm2 or 18S TaqMan Gene Expression Assay (Applied Biosystems, Foster City, CA, USA) and TaqMan Universal PCR Master Mix (Applied Biosystems) in the ABI Prism 7000 Sequence Detection System. Samples were run in triplicate and prepared according to the manufacturer's protocol. The expression levels for all samples were normalized to the endogenous control 18S (Applied Biosystems). Data were analysed using the Comparative CT Method and presented relative to the wild-type acetone control. N=4–8 mice/genotype/treatment.
We would like to thank Dr Cavell Brownie for help with the statistical analysis. This work was supported by a grant CA46637 from the National Cancer Institute.