Transcriptional Repression of Protein Kinase C (cid:1) via Sp1 by Wild Type p53 Is Involved in Inhibition of Multidrug Resistance 1 P-Glycoprotein Phosphorylation*

The protein kinase C (PKC) family consists of serine/ threonine protein kinases that play important roles in signal transduction, cell proliferation, and tumor formation. Recent studies found that PKCs are commonly overexpressed in human tumors, including soft tissue sarcoma (STS). Overexpression of PKCs contributes to invasion and migration of tumor cells and induction of angiogenesis. PKC can also phosphorylate the multidrug resistance (MDR) gene-encoded P-glycoprotein and induce MDR phenotype. Our previous studies showed that mutation of p53 enhanced STS metastasis and mediated the MDR phenotype. Restoring wild type (WT) p53 in STS cells containing mutant p53 sensitized the cells to chemotherapy. In the present study, we found that PKC (cid:1) protein expression is inhibited by WT p53 partly due to reduced PKC (cid:1) mRNA expression in STS cells, but p53 does not affect PKC (cid:1) mRNA stability. Deletion and mutation analysis of the PKC (cid:1) promoter fused to the luciferase reporter gene identified a Sp1 binding site ( (cid:2) 244/ (cid:2) 234) in the PKC (cid:1) promoter that is required for p53-mediated inhibition of PKC (cid:1) promoter activity. More importantly, PKC (cid:1) phosphorylates and activates MDR1 P-glycoprotein,

The protein kinase C (PKC) 1 family consists of a large number of serine/threonine kinases that are activated by extracellular signals. In mammalian cells, the PKC family is further divided into three subfamilies, conventional PKCs (␣, ␤I, ␤II, ␥), novel PKCs (␦, ⑀, , ), and atypical PKCs (, , ) (1). PKC is an additional PKC family member that was discovered recently and is known as protein kinase D (2). A distantly related PKC family includes three isoforms of the PKC-related kinases, known as PKN1, PKN2, and PKN3 (3). Although PKC family members share primary sequence similarities, the subfamily members have different enzymatic activation profiles. Specifically, conventional PKC isoforms are activated by 1,2-diacylglycerol and phosphatidylserine (PS) in a calcium-dependent manner, novel PKC members are activated only by 1,2-diacylglycerol and PS, and atypical PKC members are activated by PS alone as a co-factor (4). In resting cells, PKC predominantly resides in the cytosol in an inactive state and, upon stimulation, translocates as an active kinase to discrete subcellular locations, such as the plasma membrane, membranous vesicles, cytoskeleton, mitochondria, and nucleus (5).
PKC functions have been the subject of intense study for many years after identification of PKC as the intracellular receptor for tumor-promoting phorbol esters such as tissuetype plasminogen activator. Studies have shown that PKC activity is elevated in some human tumors when compared with that in adjacent normal tissues (6) and that elevated PKC activity is associated with increased metastatic or invasive potential in some human carcinoma cells (7). More recent studies indicated that PKC␣ plays an important role in promotion of tumor invasion, migration, enhanced vascular endothelial growth factor secretion, and development of the multidrug resistance (MDR) phenotype (8 -11). Because overexpression of PKCs has been found in many disorders, these kinases have become major targets for therapeutic intervention in a wide range of diseases, including cancers (12), and sensitization of tumors to radiotherapy and chemotherapy has been achieved with PKC inhibitors (13,14). Although many studies have focused on dysregulation of PKC functions, it is still unclear how expression of PKCs is regulated.
Mutation of the tumor suppressor gene p53 is commonly found in a wide variety of human tumors and is one of the most common genetic alterations in soft tissue sarcoma (STS), occurring in 30 -60% of these tumors (15). Interestingly, p53 missense mutations occur more frequently than non-sense mutations (16). Cancer cells harboring missense mutant (mut) p53 appear to have a gain of function with enhanced oncogenic properties and tumorigenic potential when compared with cells that merely lose p53 function through point mutations (16,17). Also, our previous study showed that loss of WT p53 function enhances metastasis of STS (18). Wild type (WT) p53 functions as a transcription factor and, by binding to specific DNA sequences, exerts its tumor suppressor activity by stimulating transcription of growth-inhibitory genes while preventing expression of proliferation-promoting genes (19). In contrast, mut p53 activates genes repressed by WT p53. Recent reports support this model by showing that transcription of the MDR1 gene and the proliferating-cell nuclear antigen is activated by mut p53 but repressed by WT p53 (20,21).
More recently, global phosphorylation and function of p53 were shown be inversely related to PKC activation (22). Furthermore, we found that STS cells expressing mut p53 demonstrated MDR phenotype, whereas restoration of WT p53 to these cells sensitized them to chemotherapy (23). Because PKC␣ is known to phosphorylate and activate the MDR-1 protein (11), we investigated whether WT p53 may revert to MDR phenotype through inhibition of PKC␣.
Northern Blot Analysis-Northern blotting was carried out as described previously (24). Briefly, total RNA was extracted by using RNAzol B reagent (Biotecx Laboratories) and separated on a 1% agarose gel. Separated RNA was transferred to a Hybond-N nylon membrane (Amersham Biosciences) in an aqueous hybridization solution as described previously (24). The hybridization probes included PKC␣ cDNA or glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA fragments radiolabeled with a random-primed labeling kit (Invitrogen) and purified with the QIAquick nucleotide removal kit (Qiagen). Blots were washed at high stringency (0.5ϫ saline sodium citrate (SSC), 0.1% SDS at 68°C) and exposed to Kodak BioMax film (Eastman Kodak Co.) at Ϫ70°C. Densitometric quantitation was performed with an Al-phaImager 2000 (AlphaInnotech).
Transient Transfection-Transfection was carried out with the Fu-GENE 6 transfection reagent (Roche Diagnostics) according to the manufacturer's instructions. After the addition of transfection mixtures, cells were incubated for another 4 h before they were cultured at different temperatures of either 32 or 38°C for 48 h.
Nuclear Runoff Assays-Nuclear runoff assays were performed as described previously (10) with modification. Briefly, matched pairs of SKLMS-1 and SKAla-2 cells were cultured at 32 or 38°C for 72 h. Nuclei were collected in lysis buffer (10 mM Tris-HCl, pH 7.4, 10 mM NaCl, 3 mM MgCl 2 , 0.35% sucrose, 0.5% Nonidet P-40) and centrifuged at 500 ϫ g for 3 min. Nuclei were then suspended in a reaction buffer (5 mM Tris-HCl, pH 8.0, 2.5 mM MgCl 2 , 150 mM KCl, 0.5 mM rATP/rCTP/ rGTP, 0.7 M rUTP, 2.5 mM dithiothreitol, 5 units of RNasin) in the presence of 100 Ci of [ 32 P]UTP and incubated at 30°C for 23 min. Total RNA was extracted with the addition of 40 g of a glycogen carrier and 0.8 ml of TRIzol reagent. Radiolabeled RNA pellets were collected using phenol/chloroform extraction and isopropyl alcohol precipitation before dissolution in 10 mM ␤-mercaptoethanol. Slot blots used for RNA hybridization were prepared with PKC␣ (2.5-5.0 g/slot), GAPDH (1 g/slot), and vector control (1 g/slot). cDNAs were denatured with 100 mM NaOH at 100°C for 10 min, neutralized with ammonium acetate to a final concentration of 1 M and directly blotted onto nylon membranes (ICN). Membranes were subsequently washed twice with 2ϫ SSC, baked at 80°C for 15 min, and prehybridized for 6 h at 65°C in hybridization buffer (50 mM PIPES, pH 6.5, 50 mM sodium phosphate, pH 7.0, 20 mM NaCl, 5% SDS, 2.5 mM EDTA, 50 g/ml denatured salmon sperm DNA). Hybridization was performed with equal amounts of labeled nuclear RNA (1 ϫ 10 6 cpm/ml) in hybridization buffer for 20 h at 65°C. Blots were then washed once at room temperature for 20 min and twice at 55°C for 20 min each in 1% SDS, 1ϫ SSC. Blots were then exposed to Kodak MR film. PKC␣ mRNA signal was normalized by GAPDH signal.
Determination of PKC␣ mRNA Half-Life-SKLMS-1 and SKAla-2 cells were grown at 32°C and 38°C, respectively, for 72 h. Total RNA was isolated at various times (0, 12, and 24 h) after the addition of the transcription inhibitor actinomycin D (5 g/ml). Twenty micrograms of total RNA was analyzed by Northern blotting. The PKC␣ signal was normalized according to the basal level. The half-life of the mRNA was determined with the use of regression curves (28).
P-gp Phosphorylation in SKLMS-1 and Temperature-sensitive mut p53 Cells-P-gp phosphorylation in [ 32 P]P i -labeled SKLMS-1 and SKAla cells was determined as described previously (29). Briefly, nearconfluent cell monolayers grown in 75-cm 2 flasks were labeled with [ 32 P]P i after the cells were washed with phosphate-free buffered saline and phosphate-free medium. Six milliliters of phosphate-free medium containing 10 mM HEPES, pH 7.3, and 0.6 mCi of [ 32 P]P i was added to cell monolayers for 3 h at 37°C or 32°C as indicated. The labeled cells were washed 3 times on ice with 2 ml of ice-cold phosphate-buffered saline, scraped from the plates in lysis buffer (50 mM Tris-HCl, 140 mM NaCl, 0.1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 5 mM NaF, 1 mM sodium vanadate, 2 mM EDTA, 10 g/ml leupepstatin, 10 g/ml pepstatin A, 200 g/ml aprotinin, 0.5 mM phenylmethylsulfonyl fluoride; 1 ml/plate), and lysed on ice for 20 min. Lysates were centrifuged at 13,800 ϫ g for 15 min, and the supernatants were collected.
Immunoprecipitation of P-gp was carried out according to a described previously method (29). Briefly, 5 g of the anti-P-gp antibody C 219 (Centocor) was added to 200 l of the supernatants described above and incubated overnight in a sample rotator. Fifty microliters of protein A-Sepharose 4B (Amersham Biosciences) was added to the supernatants and incubated in the sample rotator for 30 min. The beads were collected by centrifugation for 1 min and washed once in 1 ml of NaCl and 1% Nonidet P-40 and twice in an extraction buffer containing 1 M urea. The immune complexes were dissociated from the beads by incubation in 100 l of sample buffer for 15 min at 30°C. Samples were then electrophoresed on 6% acrylamide gels, and autoradiography was performed.
PKC␣ Promoter Deletion and Mutation Constructs-Human PKC␣ promoters (Ϫ1571/ϩ77 and Ϫ227/ϩ77) were generously provided by Dr. Robert Glazer (Georgetown University). All other constructs were made by amplifying the PKC␣ promoter with the reverse primer 5Ј-GGAGAGTCGGGCTGGTGCTG-3Ј and the forward primers 5Ј- ). PKC␣ promoter with Sp1 binding site mutation (Ϫ243/Ϫ239) construct was made by site-directed mutagenesis with the QuikChange site-directed mutagenesis kit (Stratagene). The following oligonucleotide 5Ј-CCGCCGCCGCCGCCGC-CGTTTTTTCCCCTTGCCC-3Ј was used to mutate the Ϫ244/Ϫ234 Sp1 site from GCCGCCTCCCC to GTTTTTTCCCC. The site-specific mutation was confirmed by DNA sequencing. The PKC␣ promoter deletion and mutation constructs were cloned into a TA cloning vector, and their orientation was checked using restriction enzyme digestion and sequencing. The PKC␣ promoter deletion constructs in the TA vector were digested with KpnI/XhoI and subcloned into pGL3-basic vector containing the firefly luciferase gene (Promega).
Reporter Gene Assays-Reporter gene assays were performed as described previously (15). Briefly, SKLMS-1 cells were seeded at a density of 2 ϫ 10 5 cells/well in 6-well plates and incubated in complete medium overnight at 37°C. Cells were subsequently treated with Ad-p53 or Ad-LacZ and further incubated for 48 h. Using FuGENE 6, cells were then cotransfected with 1 g of pGL3-basic containing the various PKC␣ deletion constructs and 0.5 g of the internal control pSV40-␤galactosidase. Twenty-four hours later, SKLMS-1 were harvested, and luciferase activity was determined and normalized to ␤-galactosidase activity (15).
Nuclear Extract Preparation and Electrophoretic Mobility Shift Assay-The nuclear extracts from SKLMS-1 cells were prepared accor-ding to the method of Andrews and Faller (30). DNA binding assays for Sp1 were performed with 15 g of nuclear extracts as described by Shie et al. (31). 32 P-Labeled double-strand wild type Sp1 oligonucleotides (5Ј-CCGCCGCCGCCGCCGCCGCCGCCTCCCCTTGCCC-3Ј) containing the Sp1 binding site in the human PKC␣ promoter and mutant Sp1 oligonucleotides (5Ј-CCGCCGCCGCCGCCGCCGTTTTTTCCCCTTGC-CC-3Ј) were used as probes. The competition was performed with a 100-fold excess of unlabeled wild type or mutant Sp1 oligonucleotides. Supershift experiments were performed with anti-Sp1 antibody (Santa Cruz Biotechnology, Santa Cruz, CA). The reactions were analyzed on 4% polyacrylamide gels containing 0.25ϫ Tris borate/EDTA buffer (20 mM Tris, 20 mM boric acid, 0.5 mM EDTA, pH 8.0).

RESULTS
WT p53 Inhibits PKC␣ Expression-To investigate the effects of p53 on PKC␣ in STS, we first analyzed PKC␣ expression in STS cells with different p53 functional status by Western blot analysis. PKC␣ was highly expressed in SKLMS-1 cells, which contain mut p53, and at very low levels in U2-OS cells, which contain WT p53 (Fig. 1A). Similarly, high PKC␣ expression was found in HT1080 cells, which are deficient in p53, and Saos-2 cells, which are p53 null.
To further define the role of p53 in modulation of PKC␣ expression, we used the SKAla cells, which are SKLMS-1 cells that express the Ala-143 temperature-sensitive mut p53 (26). The Ala-143 point mutation allows for production of WT p53 at 32°C but mut p53 at 38°C. Western blot analysis of PKC␣ in these SKAla cells showed that PKC␣ expression was higher when p53 was in the mutant conformation at 38°C than when it was in the WT conformation at 32°C (Fig. 1B), whereas the parental sarcoma cells SKLMS-1 and vector control cells SK-Neo showed no significant changes in PKC␣ expression when they were shifted between 32 and 38°C culture conditions. Therefore, the difference in PKC␣ expression between 32 and 38°C in these SKAla cell lines was induced specifically by the change in functional p53 status, not by the temperature variation itself. We used the p21expression level as a marker for functional WT p53.
To further determine the effect of WT p53 on PKC␣ expression in a p53 null background, we also transfected the Saos-2 cell line with the Ala-143 temperature-sensitive p53 mutant expression vector and observed the effect of WT and mut p53 on PKC␣ expression. Six hours after transfection, the transfected cells were shifted to 32 or 38°C and cultured for an additional 36 h, then PKC␣ expression was analyzed by Western blotting. PKC␣ expression was inhibited when the transfected Saos-2 cells were cultured at 32°C with WT p53 function that induced p21 expression (Fig. 1C). In contrast, PKC␣ level remained high when the transfected Saos-2 cells were cultured at 38°C with mut p53 function, and the induction of p21 expression was negligible. These results clearly indicated that WT p53 inhibits PKC␣ expression.
p53 Inhibits PKC␣ mRNA Expression-To determine whether p53 inhibited PKC␣ expression at the mRNA level, we performed Northern blot analysis of PKC␣ mRNA level using a PKC␣-specific probe (American Type Culture Collection) in the SKLMS-1, SKNeo, and SKAla cell lines cultured at 32 or 38°C. In SKAla-1, SKAla-2, and SKAla-3 cells, PKC␣ mRNA (ϳ9.3 kilobases) was low at 32°C with the WT p53 conformation but high at 38°C with the mut p53 conformation. No comparative changes were found in either parental SKLMS-1 or vector control SKNeo cells ( Fig. 2A). A similar result was found in Saos-2 cells transfected with Ala-143 and incubated at 32 versus 38°C (Fig. 2B). In addition, SKAla-2 expressing the temperature-sensitive mut p53 showed a time-dependent inhibition of PKC␣ mRNA expression when shifted from 38 to 32°C and vice versa (Fig. 2C). Therefore, p53 inhibits PKC␣ expression at the mRNA level.
p53 Inhibits PKC␣ mRNA Expression through Transcriptional Repression Not Affecting PKC␣ mRNA Stability-Inhibition of PKC␣ mRNA by p53 may result from either transcriptional repression or reduced mRNA stability. To investigate the mechanism by which p53 inhibits PKC␣ mRN〈 expression, we examined the rate of transcription of PKC␣ by nuclear runoff assays. Although there was not a significant difference in PKC␣ transcription between 38 and 32°C in parental SKLMS-1 cells, we reproducibly detected an approximate 30% reduction in expression of PKC␣ transcript at 32°C (WT p53) compared with that at 38°C (mut p53) in SKAla-2 cells (Fig.  3A). Therefore, reduced transcription may at least partly account for the decrease in PKC␣ mRNA level by p53.
We next investigated whether p53 also inhibits PKC␣ mRNA stability in SKLMS-1 and SKAla cells. New RNA synthesis was blocked by treating cells with actinomycin D (5 g/ml). Total RNA was extracted at different time points after treatment. Northern blot analysis using the PKC␣-specific probe showed that there was not a significant difference in the measured half-life of PKC␣ mRNA from both SKLMS-1 and SKAla-2 cells cultured at 32 and 38°C (Fig. 3B). Thus, the inhibited RNA expression level at 32°C compared with that at 38°C in SKAla-2 cells was not due to a decreased mRNA stability.
WT p53 Inhibits PKC␣ Promoter Activity via Sp1 Binding Site-To further understand the mechanisms by which WT p53 inhibits PKC␣ transcription, we fused the 1.6-kilobase PKC␣ promoter to the firefly luciferase reporter gene (Fig. 4A) to determine the effect of WT p53 on PKC␣ promoter activity. We preinfected SKLMS-1 cells with various doses of adenoviruses expressing WT p53 (Ad-p53) or expressing the LacZ gene (Ad-LacZ) for 48 h and subsequently transiently transfected the cells with the PKC␣ promoter-luciferase report construct (Ϫ1571/ϩ77); we assayed reporter gene expression 48 h later. Compared with Ad-LacZ-infected cells, Ad-p53-infected cells showed markedly inhibited PKC␣ promoter activity at different cell:virus particle ratios in a viral particle concentrationdependent manner (Fig. 4B).
To localize the promoter region responsible for WT p53mediated suppression of PKC␣ transcription, we generated a series of PKC␣ promoter deletion constructs that were composed of various lengths of 5Ј-promoter sequences fused to the luciferase gene (Fig. 4A). We transiently transfected the PKC␣ promoter-luciferase constructs and pGL3-basic vector (vector control) into SKLMS-1 cells preinfected with the Ad-p53 or Ad-LacZ (1000 virus particles/cell). The promoter activities of the various deletion constructs were not inhibited by Ad-LacZ, but they were markedly inhibited by Ad-p53 (Fig. 4C). However, the inhibitory effect of WT p53 was abolished when PKC␣ promoter was deleted down to Ϫ227, which also had dramatically reduced basal promoter activities. The data indicated that the Ϫ260/ϩ77 construct is sufficient to confer WT p53-induced inhibition and that this 34-bp promoter region upstream to the transcription start site, from Ϫ260 to Ϫ227, may contain the p53-responsive element. As expected, the promoterless luciferase pGL3-basic vector showed no promoter activity and could not be repressed by WT p53 (data not shown).
To further identify the p53-responsive element for the PKC␣ inhibition in this 34-bp PKC␣ promoter region, we mutated the only known transcription factor binding site in this region, the Sp1 site (Ϫ244/Ϫ234, GCCGCCTCCCC) (32), as shown in Fig.  4D, and evaluated the transcriptional activity of this mutated construct. The Ϫ260 mut PKC␣ promoter with Sp1 site mutation retained basal promoter activity as the wild type Ϫ260 construct. However, the Ϫ260 mut with Sp1 site mutation was not inhibited by p53 as significantly as was the wild type Ϫ260 reporter gene (17 versus 62%) (Fig. 4E). To directly test Sp1 binding to the Sp1 site in this region, we also performed electrophoretic mobility shift assay using the 32 P-labeled oligonucleotides of the 34-bp PKC␣ promoter region. Although no specific Sp1 binding was detected using the mutant Sp1 probe (Fig. 4F, lane 1), a specifically shifted band can be detected in SKLMS-1 nuclear extracts using wild type Sp1 containing oligonucleotides (Fig. 4F, lane 2), which was supershifted by anti-Sp1 antibody (Fig. 4F, lane 3). The specific binding of Sp1 could be abolished by unlabeled wild type oligonucleotides (Fig.  4F, lane 4). These data indicated that WT p53 repression of the FIG. 1. WT p53 inhibits PKC␣ protein expression in human STS cells. A, 200 g of total cell lysates obtained from SKLMS-1 (mut p53), U2-OS (WT p53), Saos-2 (p53 null), and HT1080 (p53-deficient) cells were analyzed by Western blotting with an anti-PKC␣ antibody. ␤-Actin was used as a loading control, and p21 was used as WT p53 function marker. B, protein obtained from SKLMS-1, SKNeo, SKAla-1, SKAla-2, and SKAla-3 cells was analyzed by Western blotting. SKAla cells are SKLMS-1 cells transfected with Ala-143 temperature-sensitive p53 mutant expressing vector and cultured at 32°C with WT p53 or 38°C with mut p53, respectively. C, Saos-2 cells were transiently transfected with an Ala-143 temperature-sensitive mut p53 expressing vector and cultured at 32 and 38°C, respectively, as described under "Experimental Procedures." Cell lysates were prepared, and Western blotting was performed as described in A.
PKC␣ promoter activity is mediated through the Sp1 binding site located in the Ϫ244/Ϫ234 region of the PKC␣ promoter (Fig. 4D).
PKC␣ Inhibition-induced Decrease in Phosphorylation of MDR1 P-glycoprotein-Previous reports showed that PKC␣ is selectively overexpressed in human STS, and PKC␣ catalyzes MDR1 P-glycoprotein phosphorylation and activation, which contributes to the MDR phenotype (29). To investigate the biological significance of p53-mediated PKC␣ transcriptional repression, we examined the effect of PKC␣ inhibition by p53 on phosphorylation of MDR1 P-glycoprotein and expression. We found a lower level of phosphorylation of MDR1 P-glycoprotein in SKAla-1, SKAla-2, and SKAla-3 cells when they were cultured at 32°C as compared with that at 38°C (32/38°C ratio was 76.5, 38.5, and 48.7%, respectively). There was a slight decrease in total P-glycoprotein level when cells were cultured at 32°C (32/ 38°C ratio was 92, 81.9, and 86%, respectively) that was consistent with our previous report that wild type p53 can inhibit MDR1 expression (23) (Fig. 5A). No significant difference was detected when SKLMS-1 cells and vector control SKNeo cells were shifted between 32 and 38°C (Fig. 5A). The MDR-1 phosphorylation status correlated well with the p53-modulated PKC␣ expression shown in Fig. 1A. These data suggested that inhibition of PKC␣ by wild type p53 plays an important role in reducing the phosphorylation of P-glycoprotein.
To assure that the decreased phosphorylation of MDR1 Pglycoprotein was due to specific inhibition of PKC␣ by wild type p53 but not other downstream effects of the wild type p53, we treated the SKAla-1 and SKAla-2 cells growing at 38°C with PKC␣ inhibitor Gö6976 (Calbiochem) (33) to inhibit PKC␣ kinase activity. After 24 h of treatment, the phosphorylation of MDR1 P-glycoprotein was dramatically inhibited, which is associated with lower PKC␣ kinase activity, as indicated by decreased PKC␣ phosphorylation, whereas there was no significant change in total PKC␣ and total P-glycoprotein (Fig. 5B). The results clearly showed that inhibition of PKC␣ itself is sufficient to decrease phosphorylation of P-glycoprotein. Taken together, our data demonstrated that wild type p53 inhibits PKC␣ expression, which subsequently leads to decreased MDR1 P-glycoprotein phosphorylation and, thus, increased chemosensitivity; loss of p53 function results in increased Pglycoprotein phosphorylation because of the increased PKC␣ level, which contributes to chemo-resistance. DISCUSSION The functions of PKCs in signal transduction, cell proliferation, and tumor formation have been extensively investigated (34). Recent studies have shown that PKCs are also involved in tumor migration through regulation of matrix metalloproteinases and their inhibitors (9), potentiation of vascular endothelial growth factor induction (10), and proliferation of tumor cells (35). Because overexpression of PKCs has been found in many diseases, PKCs have become one of the major targets for interventions of tumor progression (12). However, the precise mechanism of PKC regulation is largely unknown. There have been only two studies of transcriptional control of PKCs; one FIG. 2. Wt p53 inhibits PKC␣ mRNA expression in human STS cells. A, 30 g of total RNA from SKLMS-1, SKNeo, SKAla-1, SKAla-2, and SKAla-3 cells was analyzed by Northern blotting. Northern blotting analysis of GAPDH was used as a loading control. B, 30 g of total RNA from Saos-2 cells, which were transfected with Ala-143 temperature-sensitive mutant p53-expressing vectors as described in Fig. 1C, was analyzed by Northern blotting. C, temperature-sensitive p53 mutant SKAla-2 cells were cultured at 32 or 38°C for 72 h. The cells were then shifted from 32 to 38°C or from 38 to 32°C, and total RNA was prepared on days 0, 2, 4, and 6 and analyzed by Northern blotting for PKC␣ expression.
showing that 1,25-dihydroxyvitamin D3 transcriptionally regulates PKC␤ (36) and the other showing that fibroblast growth factor-2 up-regulates PKC⑀ mRNA 1.6 -3.0-fold and PKC␦ mRNA 1.7-fold (37). To the best of our knowledge our study has provided a novel link between WT p53 tumor suppressor and the inhibition of PKC␣, a critical signaling component involved in multiple aspects of oncogenesis.
One question has arisen. Which cis-DNA element is responsible for the WT p53 inhibition effect on PKC␣ promoter activity? Because there is no conservative p53 DNA binding sequence in the PKC␣ promoter, other unknown cis (or trans) elements or protein factors may be responsible for the suppres- FIG. 3. Wt p53 Inhibits PKC␣ mRNA expression through transcriptional repression not affecting PKC␣ mRNA stability. A, PKC␣ mRNA expression was analyzed by nuclear runoff assay. Nuclei from SKLMS-1 and SKAla-2 cells that were cultured at 32 and 38°C, respectively, were collected, and transcription rates were determined by nuclear runoff assay as described under "Experimental Procedures." PKC␣ signals were normalized to respective GAPDH signals, and the PKC␣ mRNA inhibition rate was calculated by comparing the PKC␣ signal at 32°C with WT p53 with that at 38°C with mut p53. Data are representative of at least three independent experiments. B, PKC␣ mRNA half-life was not affected by p53 statues. SKLMS-1 and SKAla-2 cells cultured at 32 and 38°C, respectively, were treated with the transcription inhibitor actinomycin D (5 g/ml). Total RNA was isolated at various time points (0, 12, and 24 h) and analyzed by Northern blotting. The PKC␣ signal was normalized to the basal level (at 0 h), and the half-life of the PKC␣ mRNA was determined with the regression curves. Mean Ϯ S.E. values from three independent experiments. sion of PKC␣ transcription by WT p53. Using a series of 5Ј deletion constructs of the PKC␣ promoter, we found that a 34-bp element (Ϫ260 to Ϫ227) contains a cis-element responsible for the repression of the PKC␣ promoter by WT p53. The region of the human PKC␣ promoter is GC-rich (88%) and contains a Sp1 binding site (Ϫ244/Ϫ234) (32). Sp1 protein is a well known transcription factor that is involved in the control of transcription of many important genes (38). Transcriptional repression of other genes by p53 was shown via p53 binding to Sp1 that prevented Sp1 binding to the target promoter region (39,40). Also, p53 can form a heterocomplex with Sp1 and inhibit Sp1 activity in the TF-1 human erythroleukemia cell line (40). Therefore, Sp1 binding site may be responsible for inhibition of PKC␣ by p53. Our data showed that mutation of the Sp1 binding site (Ϫ243/Ϫ239) in the PKC␣ promoter resulted in the loss of p53 inhibitory effect on PKC␣ in SKLMS-1 cells without loss of its basal promoter activity. We conclude that p53 represses PKC␣ transcription through the Sp1 binding site in the promoter.
PKC␣ also plays an important role in the development of MDR by increasing the transcriptional activity of the MDR gene 1. Transfection of PKC␣ into MCF-7 breast cancer cells has been shown to induce the MDR phenotype (41). MDR phenotype is a well characterized, multifactorial process of tumor resistance to anticancer agents. One key mediator of the MDR phenotype is the MDR1 gene, which encodes the ATP-dependent drug-efflux pump, MDR1 P-glycoprotein. Overexpression of MDR1 P-glycoprotein is associated with drug resistance in several tumors, including high grade osteosarcoma (42). MDR1 P-glycoprotein was found be highly expressed in 48% of primary sarcomas and 64% of metastatic sarcomas (43). To effectively function as a drug-efflux pump, P-glycoprotein must FIG. 4. Localization of the PKC␣ promoter region responsible for WT p53-mediated suppression of PKC␣ transcription. A, PKC␣ promoter deletion constructs were generated by polymerase chain reaction followed by cloning as described under "Experimental Procedures". B, dose-dependent inhibition of PKC␣ promoter-luciferase activity by Ad-p53. SKLMS-1 cells were pretreated with various doses of Ad-p53 or Ad-LacZ for 48 h. The cells were then cotransfected with 1 g of the PKC␣ promoter-luciferase construct (Ϫ1571/ϩ77) with 0.5 g of pSV40-␤-galactosidase. Twenty-four hours later the luciferase activity was measured as described under "Experimental Procedures." C, PKC␣ promoter deletion constructs were transiently transfected into SKLMS-1 cells that were pretreated with Ad-p53 or Ad-LacZ (1000 virus particles/cell), and luciferase activities were determined on cells extracts and normalized to ␤-galactosidase activity. Values are expressed as the percentage of the relative luciferase activity (100%) in cell extracts from untreated SKLMS-1 cells transfected with PKC␣ promoter constructs. Mean Ϯ S.E. of duplicate of three independent assays. *, p Ͻ 0.05. D, Sp1 binding site of the PKC␣ promoter sequence responsible for p53-mediated inhibition. E, relative luciferase activity was measured in wild type Ϫ260/ϩ77 and mutant Ϫ260/ϩ77 PKC␣ constructs as in C. F, identification of specific Sp1 binding to the 34-bp PKC␣ promoter region by electrophoretic mobility shift assay. Nuclear extracts from SKLMS-1 cells were incubated with radiolabeled mutant Sp1 oligonucleotides (lane 1) or wild type Sp1 oligonucleotides alone (lane 2) or plus anti-Sp1 antibody (lane 3). SKLMS-1 nuclear extracts were also preincubated with 100ϫ unlabeled wild type Sp1 oligonucleotides and then were incubated with radiolabeled wild type Sp1 oligonucleotides (lane 4). The reactions were analyzed on 4% polyacrylamide gels containing 0.25ϫ Tris borate/EDTA buffer. Ab, antibody. be phosphorylated by PKC␣ (44), which suggests that PKC␣ is indirectly involved in the development of the MDR phenotype. Our current study shows that in addition to direct down-regulation of MDR1 P-glycoprotein by wild type p53 (23), another important mechanism by which WT p53 reverts the MDR phenotype and exerts its anticancer activity is that transcriptional repression of PKC␣ by WT p53 leads to decreased phosphorylation of MDR1 P-glycoprotein. This inhibitory effect on PKC␣ and subsequent decreased phosphorylation of P-glycoprotein by WT p53 is consistent with our previous reports that reintroduction of WT p53 sensitizes STS to chemotherapeutic agents and inhibits STS growth in vitro and in vivo (23,45).
In summary, we demonstrated here that restoration of WT p53 into human leiomyosarcoma cells containing mut p53 markedly suppresses PKC␣ transcription through the Sp1 binding site located Ϫ244/Ϫ234 in the promoter upstream of transcription start site. Importantly, inhibition of PKC␣ contributes to the decreased phosphorylation of MDR1 P-glycoprotein. Our data suggest that the use of molecular-based therapies targeting mut p53 and PKC␣ will be a new effective strategy for reverting the MDR phenotype and controlling the progression of STS.