p53 stimulates promoter activity of the sgk. serum/glucocorticoid-inducible serine/threonine protein kinase gene in rodent mammary epithelial cells.

sgk is a novel member of the serine/threonine protein kinase gene family that is transcriptionally regulated by serum and glucocorticoids in mammary epithelial cells. To functionally determine if the sgk promoter is regulated by the p53 tumor suppressor protein in mammary cells, a series of sgk promoter fragments with 5′-deletions were linked to the bacterial chloramphenicol acetyltransferase gene (sgk-CAT) and transiently cotransfected into nontumorigenic NMuMG or transformed Con8Hd6 mammary epithelial cells with p53 expression plasmids. Wild-type p53, but not mutant p53, strongly stimulated sgk promoter activity in both mammary epithelial cell lines. These effects were mediated by specific regions within the sgk promoter containing p53 DNA-binding sites. The sgk p53 sequence at −1380 to −1345 (site IV) was sufficient to confer p53-dependent transactivation to a heterologous promoter, and p53 was capable of binding to this sequence in vitro as assessed by gel shift analysis. In the nontumorigenic NMuMG epithelial cell line, cotransfection of wild-type p53 strongly stimulated the activities of both the sgk promoter and the well characterized p53-responsive p21/Waf1 promoter, whereas in Rat-2 fibroblasts, wild-type p53 repressed the basal activities of both promoters, revealing that sgk and p21/Waf1 are similarly regulated in a cell type-specific manner. Taken together, these results demonstrate that sgk is a new transcriptional target of p53 in mammary epithelial cells and represent the first example of a hormone-regulated protein kinase gene with a functionally defined p53 promoter recognition element.

sgk is a novel member of the serine/threonine protein kinase gene family that is transcriptionally regulated by serum and glucocorticoids in mammary epithelial cells. To functionally determine if the sgk promoter is regulated by the p53 tumor suppressor protein in mammary cells, a series of sgk promoter fragments with 5-deletions were linked to the bacterial chloramphenicol acetyltransferase gene (sgk-CAT) and transiently cotransfected into nontumorigenic NMuMG or transformed Con8Hd6 mammary epithelial cells with p53 expression plasmids. Wild-type p53, but not mutant p53, strongly stimulated sgk promoter activity in both mammary epithelial cell lines. These effects were mediated by specific regions within the sgk promoter containing p53 DNA-binding sites. The sgk p53 sequence at ؊1380 to ؊1345 (site IV) was sufficient to confer p53-dependent transactivation to a heterologous promoter, and p53 was capable of binding to this sequence in vitro as assessed by gel shift analysis. In the nontumorigenic NMuMG epithelial cell line, cotransfection of wild-type p53 strongly stimulated the activities of both the sgk promoter and the well characterized p53-responsive p21/ Waf1 promoter, whereas in Rat-2 fibroblasts, wild-type p53 repressed the basal activities of both promoters, revealing that sgk and p21/Waf1 are similarly regulated in a cell type-specific manner. Taken together, these results demonstrate that sgk is a new transcriptional target of p53 in mammary epithelial cells and represent the first example of a hormone-regulated protein kinase gene with a functionally defined p53 promoter recognition element.
An intricate network of protein kinases and phosphatases propagates various extracellular growth and differentiation signals from the plasma membrane into the nucleus, leading to changes in the phosphorylation status and the function of discrete sets of transcription factors. The catalytic activities of most protein kinases are regulated by specific interactions with regulatory proteins (1)(2)(3) and/or by phosphorylation (4,5). Recent studies have uncovered a newly emerging subfamily of serine/threonine protein kinase genes, including snk, sgk, plk, and fnk, that are predominantly regulated at the transcriptional level by hormone-and/or mitogen-induced pathways (6 -13). Our previous studies have identified the sgk (serum-and glucocorticoid-inducible protein kinase) serine/threonine protein kinase gene, which is transcriptionally regulated by serum and/or glucocorticoids in mammary epithelial cells and Rat-2 fibroblasts (12,13), as the second member of this subfamily of transcriptionally regulated protein kinase genes. sgk encodes a 49-kDa putative protein kinase that shares 45-55% homology with the catalytic domain of protein kinase C, the cAMP-dependent protein kinase A, the rac protein kinases, and the ribosomal protein S6 kinase (13). We have documented that sgk transcripts are expressed in a variety of adult rat tissues, with the highest expression in the thymus, lung, and ovary and detectable levels in the mammary gland and several other tissues (13). The cellular and tissue context strongly influences the expression of sgk since, depending on the cell type, different extracellular stimuli affect sgk transcription, or the same signals can regulate sgk expression with different kinetics. For example, in Rat-2 fibroblasts, induction of sgk gene transcription is an immediate-early response to serum that returns to near basal levels ϳ4 h after serum stimulation, whereas in mammary epithelial cells, sgk transcript levels remain at the induced levels for at least 48 h after their rapid induction by serum (12,13). In granulosa cells of the rat ovary, sgk expression is regulated by a combination of testosterone and folliclestimulating hormone (14), while in rat brain tissue, sgk expression is induced following injury to the central nervous system (15).
The cellular mechanisms governing the expression of sgk by diverse sets of extracellular stimuli are not well understood. To define the molecular details of this process and to uncover the transcriptional regulatory factors involved in modulating sgk gene expression, 4 kb 1 of the sgk promoter region upstream of the transcriptional start site was cloned from a rat genomic library (13). Sequence analysis of the sgk promoter region revealed a glucocorticoid response element (GRE), at approximately Ϫ1.0 kb, that by functional analysis is responsible for the glucocorticoid-stimulated transcription of sgk (13). The sgk promoter also contains a TATA box and Sp-1 elements as well as putative binding sites for a variety of transcriptional regulators such as the AP-1 complex, CCAAT/enhancer-binding protein, NF-B, GATA, and Ets-2, 2 which in other systems have been found to be important for transducing proliferation and/or differentiation signals (16 -20). A striking feature of the sgk promoter is the presence of multiple putative binding sites for the p53 tumor suppressor protein, which implicates the sgk gene as a direct transcriptional target of p53. The p53 protein is a transcriptional regulator (21,22) that plays an important role in cell cycle control, cellular differentiation, apoptosis, genomic stability, and response to DNA damage (23)(24)(25)(26).
The p53 nuclear phosphoprotein can either positively or negatively regulate transcription in a gene-and tissue-specific manner. For example, the promoters of a variety of cellular genes associated with diverse p53-mediated responses are transcriptionally activated by p53 through p53-responsive elements in their promoter regions (27)(28)(29)(30)(31)(32)(33)(34)(35)(36)(37)(38). In contrast, p53 represses transcription of several genes that lack p53-responsive elements (39 -47). This repression response requires both the amino and carboxyl termini of p53 (48) as well as the oligomerization domain (49) and is likely due to interactions with components of core transcriptional machinery (44,50,51). Although not well characterized, for a few genes, transcriptional repression by p53 may involve its specific DNA binding capacity (42,45). Additionally, p53 has also been shown to regulate activity of certain genes either positively or negatively depending on specific cell type (27,52), suggesting the involvement of cell-specific factors or coregulators in p53-mediated transcriptional regulation.
In rat Con8Hd6 mammary tumor epithelial cells, from which the sgk protein kinase gene was cloned, treatment with glucocorticoids and serum suppresses cell growth and stimulates sgk expression (13). Thus, it appears that in these mammary tumor cells, the regulated transcription of sgk occurs concomitantly with suppression of growth. Recent studies are now emerging that strongly support a role for p53 in the homeostasis of mammary epithelial cell growth, although the cellular targets for p53 in mammary cells in the physiological context are largely unknown (53). Given the role of p53 as a transcription factor that inhibits cell cycle progression and the presence of multiple p53 DNA-binding sites in the sgk promoter, it is tempting to consider that sgk transcription may be regulated by p53. Besides the Con8Hd6 mammary tumor cells, sgk transcripts are also expressed and induced by glucocorticoids in nontumorigenic NMuMG mammary epithelial cells (13), thus providing a transformed (Con8Hd6) and a nontumorigenic (NMuMG) mammary epithelial cell line to compare the effects of p53. In this study, we show that p53 transactivates the sgk promoter in both NMuMG and Con8Hd6 mammary epithelial cells via a p53-responsive element located within the sgk promoter, thus demonstrating for the first time that a hormoneresponsive protein kinase gene is a direct transcriptional target of the p53 tumor suppressor protein.

EXPERIMENTAL PROCEDURES
Cells and Materials-NMuMG nontransformed mouse mammary epithelial cells were originally derived from normal glandular tissue of an adult NAMRU mouse (54). Cells were regularly cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 10 g/ml insulin, and gentamicin sulfate. Rat-2 fibroblasts and rat Con8Hd6 mammary epithelial tumor cells were cultured in Dulbecco's modified Eagle's medium/Ham's F-12 medium containing 10% calf serum. Cells were propagated at 37°C in humidified air containing 5% CO 2 , and the media were changed every 48 h. Dexamethasone was added to the cells to a final concentration of 1 M as indicated. Cell culture reagents such as Dulbecco's modified Eagle's medium, Dulbecco's modified Eagle's medium/Ham's F-12 medium, calf serum, fetal bovine serum, calcium-and magnesium-free phosphate-buffered saline, and trypsin/EDTA were supplied by BioWhittaker, Inc. (Walkersville, MD). Dexamethasone and insulin were purchased from Sigma. the CAT gene is driven by the cytomegalovirus (CMV) promoter. The GRE-CAT construct was a gift from Dr. Keith R. Yamamoto (Department of Biochemistry and Biophysics, University of California, San Francisco) and contains six copies of GRE sequence derived from mouse mammary tumor virus fused upstream of the alcohol dehydrogenase minimal promoter linked to the CAT reporter gene. The WWp-Luc reporter construct and the PG13-CAT and MG15-CAT reporter plasmids were kindly provided by Dr. Bert Vogelstein (Oncology Center and Program in Human Genetics and Molecular Biology, John Hopkins University, School of Medicine, Baltimore). The WWp-Luc construct contains Ϫ2.4-kb sequences of the p21/Waf1 promoter fused upstream of a luciferase reporter gene (38). The PG13-CAT reporter plasmid contains 13 copies of DNA sequence that binds wild-type p53 in vitro, whereas the MG15-CAT reporter bears 15 copies of a mutated sequence that precludes p53 binding (55). In both these constructs, CAT activity is driven by the early promoter of polyoma virus. The murine wild-type p53 expression plasmid CMVp53wt contains murine p53 cDNA sequences driven by the cytomegalovirus promoter. The murine mutant p53 expression plasmid CMVp53mt was isolated from Meth A fibrosarcoma, with point mutations at positions corresponding to residues 168 (Glu 3 Gly) and 234 (Met 3 Ile) (56,57), and is also driven by the cytomegalovirus promoter. Both these expression plasmids were a generous gift from Dr. Moshe Oren (The Weizmann Institute for Science, Rehovot, Israel). All other reagents utilized were of the highest available purity.
Plasmid Constructions-The construction of the pϪ4.0sgk-CAT plasmid, which contains sgk promoter sequences (Ϫ4000 to ϩ51) cloned upstream of the CAT gene in the vector pBLCAT3, has been described previously (13). The various sgk promoter-CAT deletions were generated utilizing the Erase-a-Base system (Promega, Madison, WI). The procedure involved controlled digestions of the double-digested (HindIII and AflII) parental construct pϪ4.0sgk-CAT with exonuclease III/S1 nuclease. The opposing 5Ј-protruding end of the HindIII site was protected by filling in the 3Ј-recessed end with ␣-phosphorothioate dNTP. The deleted fragments were religated with T4 DNA ligase (New England Biolabs Inc.), and positive clones were sequenced to determine the end points of deletions. The reporter construct sgk p53tk-CAT was created by subcloning a double-stranded oligonucleotide spanning 5Јflanking gene sequence Ϫ1380 to Ϫ1345 (5Ј-CCTGCCCAACTCAGGCT-GCCTCCTGCGACTTGCCT-3Ј) with the HindIII site at the 5Ј-end and with the BamHI site at the 3Ј-end into HindIII/BamHI sites of the plasmid HSVtk-CAT, which contains the thymidine kinase (tk) promoter (Ϫ105 to ϩ57) upstream of the CAT gene (58). The plasmid constructs were confirmed by DNA sequencing.
Transfection Methods-NMuMG mammary epithelial cells from logarithmically growing cultures in 60-or 100-mm tissue culture plates were transfected by the calcium phosphate precipitation method (59). The total amount of DNA used in CaPO 4 transfections for CAT assays was held constant at 20 g, and in appropriate transfections, the total DNA was adjusted to this amount using the empty CAT vector plasmid pBLCAT3. For luciferase assays, cells were plated in 60-mm tissue culture plates, and a total of 4 g of DNA (0.8 g of WWp-Luc and 3.2 g of either wild-type or mutant p53-encoding plasmids) were used for transfections. Con8Hd6 mammary tumor cells and Rat-2 fibroblasts were transfected by electroporation using previously described procedures (13). Typically, cells were transfected with 10 g of reporter plasmid and 10 g of the appropriate expression plasmid, with the total amount of DNA adjusted to 30 g using the promoterless pBLCAT3 vector DNA. Wherever indicated, the cell cultures were treated with 1 M dexamethasone for 24 h prior to harvesting the cells. Transfections were performed in triplicates and repeated at least three times.
CAT and Luciferase Reporter Gene Assays-At 40 -48 h post-transfection, the cells were harvested for CAT and luciferase assays, and the protein content of the cell extracts was estimated with the Bradford procedure (60). A quantitative nonchromatographic assay (61) was used to measure CAT activity in the cell extracts as detailed elsewhere (13). Gel Mobility Shift Assays-Preparation of nuclear extracts from NMuMG mammary cells was based on the method of Dignam et al. (62). The protein contents in the nuclear extracts were determined by the p53-responsive Hormone-inducible Protein Kinase Gene Bradford procedure (60). The sequences of the various oligonucleotides (sense) used for DNA binding studies are as follows: sgk p53 site IV, 5Ј-CCTGCCCAACTCAGGCTGCCTCCTGCGACTTGCCT-3Ј (corresponding to sequence Ϫ1380 to Ϫ1345 within the sgk promoter); sgk p53 site III, 5Ј-AGAGGGGCAGGCATCAGGGCAAGGGTATTGACTG-3Ј (spanning the sequence between Ϫ1155 and Ϫ1125 within the sgk promoter); sgk p53 site II, 5Ј-GCGGACGGAGGCAGAGTGCATCCCG-GCAGG-3Ј (representing region Ϫ285 to Ϫ255 within the sgk promoter); and sgk p53 site I, 5Ј-GAAGCCTCGGTTCTGCCTGGGGCGACTG-GCACGAA-3Ј (present in the sequence between Ϫ235 and Ϫ205 relative to the transcriptional start site of the sgk promoter). The p53 consensus oligonucleotide sequence designated as Con p53 is 5Ј-GGA-CATGCCCGGGCATGTCC-3Ј, a palindromic p53 recognition site to which p53 binds with high affinity (63). The mutant p53-binding sequence is referred to as Mut p53, bearing the sequence 5Ј-GACAT-GGGGCCCCATGTC-3Ј. The core CCCGGG nucleotides are inverted, and p53 is unable to bind to this sequence (63). Unlabeled oligonucleotides corresponding to sgk p53 sites IV to I, Con p53, Mut p53, and a nonspecific sequence were utilized as competitor DNA in gel shift assays. All the oligonucleotides were synthesized by a Model 394 synthesizer in the Cancer Research Laboratory Microchemical Facility of the University of California, Berkeley.
Radiolabeling of 5Ј-ends of the appropriate oligonucleotides was carried out in the presence of equal amounts (10 pmol) of sense and antisense strands, [␥-32 P]ATP (7000 Ci/mmol; ICN Biomedicals Inc.), and T4 polynucleotide kinase (Boehringer Mannheim) at 37°C for 30 min, followed by annealing of the labeled strands. The annealing procedure for generating either labeled or unlabeled DNA involved adding 0.1 M NaCl (0.1 volume) to equal amounts of sense and antisense strands, heating for 10 min at 70°C, and gradually cooling to room temperature. The free unincorporated nucleotides and single-stranded DNA were separated from the end-labeled double-stranded DNA by native polyacrylamide gel electrophoresis (8% gel). The labeled doublestranded oligonucleotide was excised and eluted in 400 l of TE buffer (10 mM Tris, 1 mM EDTA) and 40 l of 3 M sodium acetate, pH 5.0, for 3 h, followed by ethanol precipitation, rinsing in 70% ethanol, and resuspension in TE buffer. The radioactive oligonucleotide probes were stored at Ϫ70°C.
The DNA binding reactions (20 l) contained nuclear extract proteins (10 g), 0.5 ng of 32 P-labeled (5 ϫ 10 4 cpm) DNA probe, poly(dI-dC) (500 ng), and 7 l of 2 ϫ binding buffer (20% glycerol, 20 mM Hepes, pH 7.9, 50 mM KCl, 6.25 mM MgCl 2 , 0.5% Nonidet P-40, 0.2 mM EDTA, 4 mM spermidine) and were allowed to incubate at 4°C for 20 min. For competition experiments, a 100-fold excess of the indicated unlabeled double-stranded oligonucleotides was added prior to the addition of radiolabeled DNA probe. In some cases, the reaction mixtures were preincubated with 100 ng of specific anti-p53 antibodies (PAb421) for 15 or 60 min before the addition of radiolabeled probes. The protein-DNA complexes were resolved on a 4% native polyacrylamide gel (80:1 acrylamide/bisacrylamide) in 0.25 ϫ TAE running buffer (0.04 M Tris acetate, 1 mM EDTA, pH 7.4) containing 1 mM EDTA and 0.05% Nonidet P-40 at 4°C at 180 V. The gels were usually prerun for 2 h at 4°C. The protein-DNA complexes in the dried gels were visualized by autoradiography using Amersham Hyperfilm.
Western Blotting-Nuclear extracts were prepared essentially as described for gel mobility shift assays (62). Thirty micrograms of nuclear protein extracts from untransfected NMuMG cells were electrophoretically separated on 7.8% SDS-polyacrylamide gel. Following electrophoresis at 80 V, the proteins were transferred to nitrocellulose membrane (Micron Separations, Westborough, MA), blocked overnight at 4°C with TBST blocking solution (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.05% Tween 20) containing 5% nonfat dry milk, and then incubated overnight with primary anti-p53 monoclonal antibody PAb421 at a 1:100 dilution in TBST blocking solution containing 1% nonfat dry milk at 4°C. The secondary antibody used was directed against mouse IgG conjugated to horseradish peroxidase (Zymed Laboratories, Inc., South San Francisco, CA) at a 1:10,000 dilution in TBST blocking solution containing 1% nonfat dry milk and incubated for 1 h. The signal was detected by enhanced chemiluminescence on Hyperfilm ECL (Amersham Corp.) according to the manufacturer's instructions. Parallel sets of samples were subjected to electrophoresis, and equivalent protein loading was demonstrated by Coomassie Blue staining of the protein gel.

RESULTS
Transfection of Wild-type p53 Stimulates the Activity of the sgk Gene Promoter in NMuMG and Con8Hd6 Mammary Epi-thelial Cells-Transcription of the sgk gene is highly induced by the synthetic glucocorticoid dexamethasone both in Con8Hd6 mammary tumor epithelial cells and in nontumorigenic NMuMG mammary epithelial cells. This response is due to a functional GRE located at Ϫ1.0 kb within the sgk promoter (13). 3 Sequence analysis of the sgk promoter has further revealed several putative binding sites for the p53 protein within the 1500-bp fragment just upstream of the transcription initiation site. To test if the sgk promoter is a transcriptional target of the p53 protein, nontumorigenic NMuMG and transformed Con8Hd6 mammary epithelial cells were transfected with a sgk promoter-CAT chimeric reporter plasmid containing Ϫ1428 bp of sgk promoter sequences fused to the CAT reporter gene (Ϫ1428sgk-CAT) alone or along with expression plasmid encoding murine wild-type p53 protein. This sgk promoter fragment contains the GRE as well as four putative p53-binding sites. Transfected cells were treated with or without dexamethasone for 24 h, and cell extracts were assayed for CAT-specific activity. As shown in Fig. 1, dexamethasone stimulated the activity of the Ϫ1428sgk-CAT reporter plasmid in both NMuMG and Con8Hd6 mammary cells. Cotransfection of wild-type p53 elicited a 40-fold stimulation of the activity of the Ϫ1428sgk-CAT reporter plasmid in nontumorigenic NMuMG cells (Fig. 1, upper panel). The absolute level of sgk promoter activity was significantly increased in both dexamethasone-treated and untreated NMuMG cells transfected with wild-type p53. Furthermore, dexamethasone induced a mild increase in sgk promoter activity in the presence of wild-type p53. In Con8Hd6 mammary tumor cells, wild-type p53 stimulated a 6-fold increase in basal CAT activity that was approximately equal to the level of the induced promoter activity observed in cells treated with dexamethasone alone (Fig. 1, lower panel). No further increase in sgk promoter activity was observed in dexamethasonetreated Con8Hd6 cells transfected with wild-type p53.
Parallel sets of mammary cells were transfected with a GRE-CAT reporter plasmid that contains six copies of GREs fused upstream of the CAT reporter gene as a control for glucocorticoid responsiveness or with the CMV-CAT reporter plasmid, which is known to be transcriptionally repressed by p53 in a variety of cell types (46). As shown in Fig. 2, dexamethasone strongly stimulated GRE-CAT activity in both cell types (left panel), while cotransfection of wild-type p53 strongly repressed CMV-CAT reporter gene activity (right panel) under conditions in which p53 induced sgk promoter activity. Taken together, our results show that wild-type p53 strongly activates the glucocorticoid-responsive sgk promoter in both nontumorigenic (NMuMG) and transformed (Con8Hd6) mammary epithelial cells in a dexamethasone-independent manner.
Mapping of p53-responsive Regions within the sgk Promoter by Deletion Analysis-To determine the region of the sgk promoter that confers the responsiveness to wild-type p53, a series of 5Ј-progressive deletions in the sgk promoter were generated by controlled exonuclease III digestions. The corresponding sgk promoter-CAT chimeric reporter plasmids (sgk-CAT) were constructed to contain varying lengths of sgk promoter sequences that all terminate at ϩ51 in the sgk gene. The resulting sgk-CAT constructs had four (Ϫ1428sgk-CAT), two (Ϫ681sgk-CAT, Ϫ303sgk-CAT), one (Ϫ236sgk-CAT), or no (Ϫ190sgk-CAT) putative p53 DNA-binding sites in the sgk promoter region (Fig.  3A, lower panel). NMuMG or Con8Hd6 cells were cotransfected with the sgk-CAT reporter plasmids alone or together with expression vectors for either murine wild-type p53 or mutant p53 with two point mutations at amino acid residues 168 and 234, which renders the mutant deficient in DNA binding (64).
Analysis of CAT-specific activity revealed that cotransfection of wild-type p53 caused a 23-fold stimulation of the promoter activity of Ϫ1428sgk-CAT in NMuMG mammary epithelial cells, whereas a significant reduction in the extent of induction by wild-type p53 was observed in the sgk-CAT constructs containing progressively shorter sgk promoter fragments (Fig. 3B). Transfection of the mutant p53 expression plasmid had no effect on sgk promoter fragments except for a mild stimulatory effect on the reporter plasmid containing the Ϫ1428-bp sgk fragment. A similar pattern of activation only by wild-type p53, but not mutant p53, was observed in transfected Con8Hd6 mammary tumor cells, although the magnitude of induction of sgk-CAT deletion constructs by wild-type p53 was less compared with that of NMuMG cells (Fig. 3B). The p53-mediated stimulation of sgk promoter activity in both nontumorigenic NMuMG and transformed Con8Hd6 mammary cells appeared to target a similar region of the sgk 5Ј-upstream sequence that contains two p53 DNA-binding sites between Ϫ681 and Ϫ1428 of the sgk promoter (p53 sites III and IV) likely to be responsible for most of the p53 response.
sgk p53 DNA-binding Site IV Is Sufficient to Confer Positive Transcriptional Effects of p53 to a Heterologous Promoter-Of the two putative p53-binding sites in the p53-responsive region of the sgk promoter, p53 site IV between Ϫ1380 and Ϫ1345 bp is most homologous to the consensus p53-binding site. To test if this p53 recognition site alone is sufficient to render the positive transcriptional regulation observed with the sgk promoter to a heterologous promoter, an oligonucleotide corresponding to sgk p53 site IV was inserted into a CAT vector upstream of sequence Ϫ105 to ϩ50 of the thymidine kinase minimal promoter to form sgk p53tk-CAT. NMuMG mammary cells were cotransfected with either a sgk p53tk-CAT or a tk-CAT reporter plasmid that does not contain p53 site IV sequences along with expression vectors for either wild-type or mutant p53. Cells were also transfected with two other reporter plasmids, PG13-CAT, which contains 13 copies of DNA sequences that bind p53 in vitro (55) fused upstream of the CAT reporter gene, or MG15-CAT, which contains 15 copies of mutated p53-binding sequence. These reporter plasmids were either transfected alone or cotransfected with either wild-type or mutant p53encoding plasmids. Measurement of CAT activity demonstrated that wild-type p53 stimulated, by ϳ8-fold, the activity of the sgk p53tk-CAT reporter plasmid in NMuMG epithelial cells, while mutant p53 had no effect on sgk p53tk-CAT activity

FIG. 2. Transactivation and transrepression of control reporter plasmids by dexamethasone and wild-type p53 in NMuMG and Con8Hd6 cells.
NMuMG or Con8Hd6 cells were transfected with 10 g of synthetic GRE-CAT reporter plasmid containing six copies of GREs linked to a minimal promoter (left panel), and the total amount of DNA was adjusted appropriately with the empty CAT vector plasmid pBLCAT3 as described under "Experimental Procedures." Following transfection, cells were treated either with (ϩ) or without (Ϫ) dexamethasone (Dex; 1 M) for 24 h. Both cell types were also transfected with the CMV-CAT reporter plasmid (10 g) alone (Ϫ) or together with (ϩ) 10 g of expression plasmids for wild-type p53 (wp53; right panel). Forty hours post-transfection, cells were harvested for measurement of CAT activity and calibrated to protein levels as described for Fig. 1. Values denote means Ϯ S.D. from triplicate samples derived from three separate experiments. (Fig. 4). Transfection of the minimal promoter-containing tk-CAT reporter plasmid only resulted in a minor stimulation by wild-type p53 in NMuMG cells while remaining at basal levels under all other conditions. As expected, cotransfection of wildtype p53 stimulated, up to 5-fold, the activity of only the wildtype p53-responsive PG13-CAT reporter plasmid, but not the mutant MG15-CAT reporter plasmid (Fig. 4). Mutant p53 had negligible effects on both of the reporter plasmids. These results demonstrate that the putative p53 site IV in the sgk promoter is a functional p53 recognition element that can confer positive transcriptional regulation of sgk promoter activity by wild-type p53 in normal NMuMG epithelial cells.
Interaction of p53 with the sgk p53 Site IV Sequence by Gel Shift Analysis-Since the 35-bp p53 site IV sequence present between Ϫ1380 and Ϫ1345 bp of the sgk promoter is sensitive to transcriptional regulation by p53, gel shift experiments were utilized to examine the formation of nuclear protein-DNA complexes. Nuclear extracts isolated from NMuMG epithelial cells were incubated with a 32 P-labeled oligonucleotide corresponding to Ϫ1380 to Ϫ1345 bp of the sgk promoter, and protein-DNA complexes were resolved by native polyacrylamide gel electrophoresis. Incubation of nuclear extracts derived from the NMuMG epithelial cells with the sgk p53 DNA probe (site IV) resulted in the formation of three distinct protein-DNA complexes, A, B, and C (Fig. 5, left panel). To determine the specificity of these interactions, a series of competition experiments were carried out using different types of unlabeled oligonucleotides. The specific complexes were completely eliminated by the addition of a 100-fold molar excess of specific unlabeled sgk p53 site IV DNA, but not by same amount of nonspecific unrelated DNA (nonspecific DNA) or mutated consensus p53-binding site DNA (Mut p53). When excess unlabeled DNA corresponding to the p53 consensus sequence (Con p53) was used as a competitor, only protein-DNA complex C was efficiently competed off, FIG. 4. sgk p53 site IV (؊1380 to ؊1345) confers positive regulation by p53 to a heterologous promoter. NMuMG cells were transfected with the sgk p53tk-CAT reporter plasmid (10 g) containing sgk p53 sequences (Ϫ1380 to Ϫ1345) linked to a heterologous thymidine kinase minimal promoter, with the tk-CAT vector lacking these sequences, with the PG13-CAT reporter construct containing 13 copies of the p53-responsive element, or with the MG15-CAT reporter plasmid containing 15 copies of mutated p53 sequences either in the absence or presence of cotransfected expression plasmids for wild-type p53 (wt p53; 10 g) or mutant p53 (mt p53; 10 g). CAT activity was determined in cell lysates as described for Fig. 1 (upper panel) and expressed with respect to protein content. Experiments were performed in triplicates, and the reported values denote means Ϯ S.D. derived from three separate transfections.

FIG. 3. p53-mediated activation of sgk promoter-CAT deletions in Con8Hd6 and NMuMG cells.
A, the upper panel shows the sequences of the four putative p53 recognition elements within the Ϫ1428-bp sgk promoter fragment and are designated as sites I-IV. The base pairs underlined denote nucleotides sharing homology with the p53 consensus sequence. The lower panel illustrates the location of the putative p53-binding sites and the GRE in the context of the Ϫ1428sgk-CAT reporter construct. The arrows and corresponding numbers refer to the 5Ј-end points of the sgk promoter deletions incorporated into the Ϫ1428sgk-CAT, Ϫ681sgk-CAT, Ϫ303sgk-CAT, Ϫ236sgk-CAT, and Ϫ190sgk-CAT reporter plasmids. B, the indicated sgk-CAT reporter plasmids (10 g) were transfected alone or cotransfected with 10 g of expression vectors for either wild-type p53 (wt p53) or mutant p53 (mt p53) into either Con8Hd6 cells (left panel) or NMuMG cells (right panel) as described in the text. The 5Ј-sgk promoter fragments used are shown on the far right, and the corresponding numbers refer to the 5Ј-end points of deletions of the various constructs. Forty hours post-transfection, cells were assayed for CAT activity as described for Fig. 1, and the CAT activity was normalized to protein levels. Data represent means Ϯ S.D. obtained from at least three separate transfections, each carried out in triplicate.
while the two slower migrating complexes, A and B, were not affected (Fig. 5, left panel). The specificity of binding of p53 with the sgk p53 site IV sequence was also examined by preincubating the nuclear extracts for 1 h with the PAb421 anti-p53 monoclonal antibody. The p53 monoclonal antibody disrupted the formation of protein-DNA complex C (Fig. 5, left  panel, PAb421). In addition to the sgk p53 site IV sequence, three other putative p53-binding sites designated as site I (Ϫ235 to Ϫ205), site II (Ϫ285 to Ϫ255), and site III (Ϫ1155 to Ϫ1125) exist within Ϫ1428 bp of the sgk promoter that differ from each other and from the p53 consensus sequence to varying extents. Competition experiments were carried out to test if the different sgk p53 sequences can specifically compete for p53 binding with the radiolabeled site IV DNA probe. As also shown in Fig. 5 (right panel), the specific complex formed (No competitor) can be competed off with a 100-fold molar excess of unlabeled sgk p53 sequences spanning site III, II, or I, but not with an unrelated DNA sequence (nonspecific DNA). Gel shift analysis using NMuMG nuclear extracts and radiolabeled oligonucleotides corresponding to sgk p53 sites I-III revealed the specific formation of protein-DNA complex C, as observed with site IV, with each putative sgk p53-binding site, although the protein binding appeared to be weaker compared with that of site IV (data not shown).
To functionally test for the presence of p53 in nuclear extracts, a parallel set of gel shift experiments was carried out using a radiolabeled oligonucleotide corresponding to the consensus p53 DNA-binding site. In contrast to the three protein-DNA complexes detected with the sgk DNA site IV probe in NMuMG extracts (Fig. 5), incubation of the p53 consensus probe with these nuclear extracts yielded a single specific complex (Fig. 6, No competitor). Competition experiments established the specificity of binding. For example, the addition of excess specific unlabeled Con p53 fragment completely abolished the single band, whereas excess unrelated nonspecific DNA or DNA encoding a mutated p53-binding site was incapable of competing off the protein-DNA complex (Fig. 6). More importantly, the specific complex using radiolabeled Con p53 was competed off with excess unlabeled sgk p53 site IV DNA. When the nuclear extracts were preincubated with p53 antibodies for a short time (15 min), a partial supershift of the specific complex was observed, whereas upon longer incubation (60 min) with p53-specific antibodies, a completely supershifted complex was obtained. The presence of p53 in these nuclear extracts was analyzed by Western blotting. As also shown in Fig. 6, detectable amounts of p53 were present in these nuclear extracts, accounting for the specific protein-DNA complex formation. These results show that p53 present in the nuclear extracts of NMuMG epithelial cells can bind specifically to the sgk p53 site IV sequence or the consensus p53 DNA-binding site. In addition, the different putative sgk p53binding sequences (sites III, II, and I) are all homologous to some extent to the site IV sequence as these three sites can efficiently compete for specific binding of p53 with the radiolabeled sgk p53 site IV DNA probe.
Cell Type-specific Activation or Repression of the sgk Promoter by Wild-type p53 Protein-Previous studies have shown that the magnitude and kinetics of sgk gene expression can be differentially regulated in epithelial cells versus fibroblasts (12). Since p53 has been shown to differentially regulate target promoters in a cell type-specific manner (27,52), the cell typespecific transcriptional effects of wild-type p53 protein on sgk FIG. 5. Gel shift analysis of binding of p53 with the sgk p53 site IV sequence. Nuclear extracts (10 g) prepared from NMuMG cells were incubated with 32 P-end-labeled double-stranded DNA probes representing region Ϫ1380 to Ϫ1345 of the sgk promoter (site IV). The indicated reaction mixtures contained no extract; no unlabeled competitor DNA (No competitor); a 100-fold molar excess of double-stranded oligonucleotides corresponding to sgk p53 site IV, the consensus p53binding site (Con p53), an unrelated DNA (nonspecific DNA), or a mutant p53-binding site (Mut p53); or a 100-fold molar excess of unlabeled competitor DNA sgk p53 site I (Ϫ235 to Ϫ205), sgk p53 site II (Ϫ285 to Ϫ255), or sgk p53 site III (Ϫ1155 to Ϫ1125). Binding reactions were preincubated for 1 h with 100 ng of anti-p53 antibodies (PAb421). The protein-DNA complexes formed were separated by 4% native polyacrylamide gel electrophoresis and visualized by autoradiography. Arrows indicate the positions of protein-DNA complexes and free probe.
FIG. 6. p53 present in nuclear extracts as determined by Western blotting interacts with p53 consensus sequences in NMuMG cells as assessed by gel shift analysis. Nuclear extracts (10 g) prepared from NMuMG cells were incubated with 32 P-end-labeled double-stranded DNA probes representing the p53 consensus sequence (Con p53). The indicated reaction mixtures contained no extract; no unlabeled competitor DNA (No competitor); or a 100-fold molar excess of double-stranded oligonucleotides corresponding to sgk p53 site IV, the consensus p53-binding site (Con p53), an unrelated DNA (nonspecific DNA), or a mutant p53-binding site (Mut p53). Binding reactions were preincubated for 15 or 60 min with 100 ng of anti-p53 antibodies (PAb421). The protein-DNA complexes formed were separated by 4% native polyacrylamide gel electrophoresis and visualized by autoradiography. Arrows indicate the positions of protein-DNA complexes and free probe. Nuclear protein extracts (30 g) from NMuMG cells were electrophoretically fractionated, blotted onto nitrocellulose filters, and analyzed for p53 protein by Western blotting as described under "Experimental Procedures" (inset). The protein molecular mass standards are shown on the left. The arrow indicates the position of the p53-specific band.
promoter activity were examined in NMuMG mammary epithelial cells as compared with Rat-2 fibroblasts. We also examined the effects of p53 on the p21/Waf1 cell cycle inhibitor gene, which is a well characterized transcriptional target of p53 and contains p53-binding sites within its promoter (38), for comparison with the sgk promoter. To determine if sgk and p21/Waf1 promoter activities are similarly regulated by p53 in a cellspecific manner, NMuMG mammary epithelial cells or Rat-2 fibroblasts were transfected either with the Ϫ1428sgk-CAT reporter plasmid or the WWp-Luc reporter plasmid, which contains Ϫ2.4 kb of human p21/Waf1 promoter sequences linked to a luciferase reporter gene. Cells were cotransfected with expression vectors for either wild-type or mutant p53, and reporter gene activities were monitored 40 h post-transfection. As shown in Fig. 7, cotransfection of wild-type p53 strongly stimulated both sgk and p21/Waf1 promoter activities in NMuMG epithelial cells and repressed the activities of both promoters in transfected Rat-2 fibroblasts. The activity of the Ϫ1428sgk-CAT reporter plasmid was unaffected by mutant p53. However, the mutant p53 used for these studies repressed p21/Waf1 promoter activity in Rat-2 fibroblasts (Fig. 7) and had a partial stimulatory effect in transfected NMuMG epithelial cells. These results show that wild-type p53 exerts both positive and negative cell type-specific transcriptional regulation on sgk promoter activity in a manner generally similar to the well characterized p53-responsive p21/Waf1 promoter. Moreover, the cell type-specific positive or negative transcriptional effects of p53 appear to be a more general effect on target genes containing p53 DNA-binding sites.

DISCUSSION
The p53 tumor suppressor protein has been proposed to mediate its diverse biological effects in response to various external stimuli by stimulating or repressing the transcription of particular sets of target genes. Our results demonstrate that, both in NMuMG nontumorigenic mammary epithelial cells and Con8Hd6 mammary tumor cells, one such transcriptional target of wild-type p53 is the sgk serum/glucocorticoid-inducible protein kinase gene. In both mammary cell types, wild-type p53 strongly stimulated basal sgk promoter activity in transiently transfected cells in a manner independent of glucocorticoids. Cotransfection of wild-type p53 reduced the dexamethasone responsiveness of the sgk promoter, which was likely due to near maximal stimulation of sgk-CAT reporter plasmid activity by p53. In Con8Hd6 mammary tumor cells, dexamethasone induces a G 1 cell cycle arrest under conditions in which sgk gene expression is rapidly stimulated (13), implicating a role for sgk in the growth suppression response. The transcriptional stimulation of sgk promoter activity by p53 further supports this idea and represents the first example of a hormone-regu-lated protein kinase gene that contains a functionally defined p53-responsive element in its promoter region.
Deletion analysis of the sgk promoter revealed that the p53responsive region observed in both transfected mammary epithelial cell lines harbors two putative p53-binding sites, sgk p53 site IV (Ϫ1380 to Ϫ1345) and sgk p53 site III (Ϫ1155 to Ϫ1125). Transfection assays demonstrated that the 35-bp fragment corresponding to sgk p53 site IV alone was functional and that p53 specifically interacted with these sequences in vitro. We have also observed specific binding of p53 to the sgk p53 site III sequence (data not shown). Previous studies have defined two copies of the 10-bp motif 5Ј-RRRC(A/T)(T/A)GYYY-3Ј separated by 0 -13 bp (63) as the p53 consensus sequence. The sgk p53 site IV sequence (5Ј-ccTGCCCAACTCAGGCTGCCTC-CTGCGACTTGCCT-3Ј) as well as the site III sequence (5Ј-agagGGGCAGGCATCAGGGCAAGGGTATTGACTg-3Ј) exhibit strong homology to the p53 consensus sequence compared with the other two p53-binding sites, I and II, outside of the functionally defined p53-responsive region. These two sites may account for the minor activation of sgk-CAT reporter plasmids containing shorter regions of sgk promoter activity in transfected NMuMG cells. A wide range of degeneracy in p53 recognition sequences exists as a variety of p53-activated promoters contain p53-binding sites different from the p53 consensus sequence. For instance, the p53-responsive element within the human T-cell leukemia virus type I enhancer is composed of a GC-rich element (5Ј-GCCCTGACGTGTCCCC-3Ј) (65). Other p53-responsive elements that differ from the consensus sequence include the two copies of the weak p53binding element (5Ј-GGGCGGAGTTA-3Ј) within the SV40 origin of replication (66) and the p53 sequences defined from mouse genomic clones (5Ј-GACACTGGTCACACTTGGCTGCT-TAGGAAT-3Ј) (67). In this regard, although sgk p53 site I and II sequences are not strongly related to the p53 consensus sequence, p53 was capable of specifically binding to these sequences in vitro (data not shown), and these sequences efficiently competed with the sgk p53 site IV sequence for binding as assessed by gel shift assays.
In contrast to the strong p53-dependent stimulation of sgk promoter activity in mammary cells, wild-type p53 evoked a repression of sgk promoter activity in Rat-2 fibroblasts. More importantly, wild-type p53 regulated the activity of the p21/ Waf1 gene promoter, a well characterized transcriptional target of p53 (38), in a cell type-specific manner generally similar to the effects observed on sgk promoter activity. This result further establishes the sgk gene as a biologically relevant transcriptional target of p53. The exact mechanism involved in the p53-mediated down-regulation of cellular promoters is not well understood, although it has been proposed that repression by FIG. 7. Cell type-dependent activation or inhibition of the sgk and p21/ Waf1 promoters by p53. NMuMG cells or Rat-2 fibroblasts were transfected either with the Ϫ1428sgk-CAT reporter plasmid alone (left panel) or with the WWp-Luc reporter plasmid alone (right panel) or cotransfected with expression plasmids encoding either wild-type p53 (wt p53) or mutant p53 (mt p53). CAT or luciferase activity was estimated in cell extracts as described in the text and normalized to protein levels. Data shown are means Ϯ S.D. obtained from three separate transfections, each carried out in triplicate.
p53 occurs by interaction with components of the general transcriptional machinery (44,50,51) and, in some cases, may entail binding of p53 to specific DNA sequences (42,45). In fact, the cell type-specific positive or negative regulation of genes by p53 appears to be confined to genes containing p53-binding sites. For example, the creatine kinase gene and proliferating cell nuclear antigen gene promoters, which contain functional p53-binding sites within their promoter regions, are also either positively or negatively regulated in a cell type-specific manner (27,29,52). In contrast, a CMV-CAT reporter that lacks p53binding sites and is known to be repressed by p53 (46) was suppressed by p53 in both mammary epithelial cells and Rat-2 fibroblasts (data not shown). A variety of viral and cellular proteins (68,69) and also certain transcription factors (70) have been shown to interact with p53 and to modulate its function (71)(72)(73)(74). Thus, depending on the cellular context, p53 is able to affect either transactivation or transrepression of the sgk gene, or the p21 gene, presumably in conjunction with cell typespecific regulatory factors.
Many transcriptional target genes of p53 are involved in important cellular processes such as growth regulation (e.g. fos, jun, myc, cyclin A, and p21/Waf1) (38,40,75,76) and apoptosis (e.g. bcl-2 and bax) (33,42). The cell type-specific growth suppression effects of p53 appear to be related to the ability of p53 to act as a specific transcriptional activator and, in certain tissues, its ability to interact with the MDM2 protein (77). Many studies have implicated p53 to be important for the growth control of mammary epithelial cells. For example, approximately one-fourth of all sporadic breast cancers express mutant forms of p53 containing deletions or mutations (78), whereas other mammary tumors are characterized as having an allelic loss of the short arm of chromosome 17, which includes p53 (79,80). Consistent with a role for p53 in the evolution of mammary tumors, a deficiency of p53 accelerates mammary tumorigenesis in Wnt-1 transgenic mice (81). Other evidence suggests that p53 mutations are important in the early preneoplastic stages of mammary tumorigenesis (77). In addition, loss of p53 protein in certain types of mammary cells exposed to ␥-irradiation (53) or as a result of human papillomavirus E6-induced degradation (82) is accompanied by immortalization of these cells and failure to arrest in the G 1 phase of the cell cycle.
In our study, wild-type p53, but not mutant p53, strongly stimulated sgk promoter activity in both nontumorigenic and transformed mammary epithelial cells. In this context, functional characterization of a positive p53 recognition element within the promoter of this protein kinase gene implicates that a subset of p53-mediated responses in mammary cells, such as on the control of cell proliferation or programed cell death after exposure to DNA-damaging agents (83), may be mediated by positive or negative changes in cellular phosphorylation cascades and/or activities of potential sgk substrates. In mammary tumor epithelial cells, we propose that sgk may be involved in the glucocorticoid-regulated growth suppression response and that activation of sgk by p53 may represent a distinct, but complementary, growth regulatory pathway. A key focus of our future approaches will be to understand the functional roles and biological context by which p53 targets the sgk gene in mammary epithelial cells and if activation of the sgk promoter requires mammary-specific nuclear factors or coactivators to interact with the p53-DNA complex.