Constitutive and DNA Damage Inducible Activation of pig3 and MDM2 Genes by Tumor-Derived p53 Mutant C277Y¹

The p53 gene is compromised in most human cancers by point mutation. Evidence is accumulating that these alterations frequently do not result in a complete loss of the sequence-specific transcriptional regulatory function of p53. Here, we describe the transcriptional activity of the p53 mutant C277Y isolated from a Ewing's sarcoma with high constitutive pig3 expression. Transient transfection of this mutant into a p53 null cell line resulted in activation not only of the pig3 but also of the MDM2 gene compatible with the presence of constitutively expressed MDM2 transcripts initiated from the P2 promoter in the p53-C277Y hemizygous Ewing's sarcoma cell line. Expression of endogenous pig3 and MDM2 genes was further enhanced on irradiation of this cell line. Here, suppression of p53-C277Y by RNAi reduced pig3 promoter activity, RNA, and protein expression. Reporter gene assays revealed that the potential of p53-C277Y to up-regulate MDM2 expression was similar to wild-type p53, whereas activation of the pig3 promoter was at least 5-fold increased over wild-type p53. The pentanucleotide microsatellite sequence present in exon 1 of the pig3 gene was found to be responsible for p53-C277Y-mediated activation. In concordance with a role of PIG3 protein for cell death, we showed residual apoptotic activity of p53-C277Y to which the described Ewing's sarcoma cell line was found to be resistant. p53-C277Y has previously been reported to bind to DNA with altered sequence specificity and to be unable to activate generic p53 target genes in yeast-based functional assays. Our results, therefore, show that a p53 mutant may behave differently when tested in its authentic cellular context.

The tumor suppressor protein p53 is a powerful transcription factor that plays an important role in cell cycle regulation, DNA repair, differentiation, senescence, and angiogenesis. It functions as a tetramer that can bind to about 300 different promoter elements in the human genome (1). p53 acts as both a sequence-specific activator as well as an unspecific repressor for many TATA box-containing genes (2). p53 is activated in response to cellular stress signals evoking either growth arrest or apoptosis. Disruption of p53 function by either mutation or protein inactivation through interaction with viral or cellular oncogenes is the most frequent alteration observed in human cancer (3). Most alterations are missense mutations that affect amino acid residues either directly contacting DNA or providing structural integrity to the core domain for sequence-specific DNA binding resulting in nuclear p53 accumulation and loss of transcriptional activity (4). Heterozygous germ-line p53 mutations predispose to a wide range of tumor types with early onset characterizing Li-Fraumeni syndrome (5). In addition to the loss of tumor suppressor function, some mutations may also result in a gain of oncogenic activity of p53 interfering with apoptosis or growth suppression (6) and increasing metastatic potential (7). The basis for the oncogenic function of mutant p53 is still not clear but, probably, both a dominant-negative effect toward wild-type (wt) p53 in the multimeric complex and possibly other p53 family members, as well as altered transcriptional activity on a specific subset of genes, may be involved (8). Individual mutations may differentially affect sequence-specific binding of p53 to distinct target genes, as has been shown for p21 and bax gene promoters (4, 9). For several p53 repressed genes, including PCNA (10), HSP70 (11), and MDR-1 (12), specific p53 mutations have been reported to result in their activation. Some p53 mutations result in DNA binding with altered structural preference as exemplified by the mutant p53-G245S which binds to non-B-DNA conformations (13). Phenotypically, introduction of different p53 mutants into p53-null H1299 lung adenocarcinoma cells resulted in increased chemoresistance (14). The p53-D281G mutant isolated from human prostatic cancer has been reported to switch the apoptotic function of the tumor suppressor to a Bcl2-insensitive cell death program (15). As reported recently, mutant p53 proteins promote gene amplification independently of their capacity to inactivate the wt p53 protein (16). This effect was found to be dependent on conservation of the ability of mutant p53 to interact with topoisomerase I, and, in the absence of transcriptional activity, was interpreted as a gain of function in destabilizing genomic integrity (17).

In a yeast-based mutagenesis screen, several p53 variants have been selected that bind to and transcriptionally activate from an altered DNA binding motif (TTT CATG AAA) but not from a wild-type site (18). Among them, C277Y has been known as a naturally occurring mutation in Ewing's sarcoma (19) and several other human tumors, including osteosarcoma, leiomyosarcoma, and carcinomas of the bladder, stomach, colon, breast, and liver, generally representing 0.3% of all mutations described in human cancers (20). In the yeast system, this mutant was capable of activating the waf1, but neither the human mdm2 gene (MDM2), cyclin G, gadd45, nor bax promoters (21). Here, we report that when tested in the authentic cellular background (Ewing's sarcoma), the C277Y mutant strongly activates MDM2 and the proapoptotic gene pig3.

During our studies on the function of the p53 pathway in Ewing's sarcoma family tumors (EFT; refs. 22, 23), we tested for the constitutive and activated expression of p53 downstream target genes on Northern blots. Among 16 EFT cell lines (7, 8, and 1 cell lines wild-type, mutant, and null for p53, respectively), STA-ET-2.2 showed a remarkably elevated constitutive pig3 expression. By real-time quantitative PCR, an 8- to 80-fold difference in pig3 RNA levels was found between STA-ET-2.2 cells and a p53 null (SK-N-MC) or a wt p53–expressing (VH64) EFT cell line, respectively an EFT cell line carrying a p53 R273C mutation (IARC-EW2; Fig. 1A). Using a probe (spanning MDM2 exons 3 to 5) that recognizes both basal (L-) and p53 inducible (S-) MDM2 transcripts, constitutive MDM2 RNA levels were found to be highly variable between the cell lines. On irradiation, all wt p53 cell lines showed pronounced inducibility of MDM2 gene activity peaking at 4 hours postirradiation (Fig. 1B). The subsequent decrease in MDM2 RNA levels cannot be attributed to the induction of apoptosis because VH64 cells are completely radio-resistant (23). In p53 null SK-N-MC cells, only a weak p53-independent increase over background levels was observed for MDM2 RNA with a maximum at around 7 hours postirradiation (Fig 1B). In contrast, in STA-ET-2.2 cells, MDM2 RNA gradually increased up to 24 hours postirradiation. Interestingly, reactivity with an exon 2 probe that is specific for MDM2 transcripts initiated from the p53 inducible P2 promoter (24) indicated constitutive P2 promoter activity in untreated STA-ET-2.2 cells that was further increased on X-irradiation (Fig. 1C). Expression of the pig3 gene was variably induced by X-ray treatment in the three wt p53 cell lines and absent from the p53 null cell line SK-N-MC. Remarkably, significant pig3 induction was observed in STA-ET-2.2 cells at 24 hours (Fig. 1B). In addition, basal waf1 RNA levels comparable to wt p53 cell lines and, on irradiation, a slight increase in waf1 expression was observed in this cell line, while bax inducibility was restricted to wt p53 cell lines only.

STA-ET-2.2 cells carry a hemizygous p53-C277Y mutation leading to nuclear accumulation of p53 (19) that has previously been shown to adopt an altered DNA binding specificity when tested on artificial p53 binding elements (18) and to retain some activity on the waf1 promoter in a yeast-based assay (21). Thus, we speculated that elevated constitutive and irradiation inducible activity of pig3 and MDM2 genes in STA-ET-2.2 cells may be assigned to the expression of p53-C277Y in this cell line.

To directly test this hypothesis, we transiently introduced p53-C277Y and, for comparison, wt p53, mutant p53-R273H, and empty expression vectors into the p53-null EFT cell line SK-N-MC and tested for pig3 and MDM2 expression 24 hours posttransfection. Because wt p53 overexpression results in apoptosis of SK-N-MC cells (23) that can be blocked by the general caspase inhibitory peptide z-VAD.fmk (22), the experiment was done both in the absence and the presence of the caspase inhibitor. Thus, we were able to monitor p53-induced gene expression that is not affected by apoptotic loss of cell viability. As illustrated in Fig. 2, transfection of p53-C277Y was similarly effective in pig3 gene induction as wt p53, while activation of MDM2 transcription was clearly detectable but weaker than in wt p53–transfected SK-N-MC. No effect on pig3 and MDM2 expression was observed after transfection of p53-R273H or empty vector. z-VAD.fmk treatment was capable of further increasing induced pig3 and MDM2 RNA levels in wt p53 but not in p53-C277Y–expressing cells. These results corroborate the potential of p53-C277Y to activate pig3 and MDM2 genes.

To confirm the link between constitutive pig3 gene activity and expression of p53-C277Y in STA-ET-2.2 cells, p53 siRNA was expressed from a retroviral small hairpin RNA expression vector (Fig. 3). Six days after infection, endogenous p53-C277Y protein levels were clearly reduced. Concomitantly, a significant decline in PIG3 RNA (about 10-fold) and protein expression was observed, providing further evidence for p53-C277Y-activated expression of pig3 in STA-ET-2.2 cells.

To quantitatively address the transcriptional potential of p53-C277Y, we did reporter gene assays in comparison to wt p53 and to the R273H mutant using the wt p53 responsive promoters from MDM2, pig3, and rgc genes in SK-N-MC cells lacking endogenous p53 expression. Co-transfection of reporter and p53 expression constructs with a green fluorescent protein (GFP) expression plasmid and subsequent flow cytometric assessment of GFP expression indicated similar transfection efficiencies and viability of transfected cells when measured 24 hours posttransfection. At 44 hours, the proportion of dead cells significantly increased in wt p53–transfected cells due to extensive apoptosis (data not shown and Fig. 7). Fig. 4A shows equal expression levels of p53 proteins in the presence of co-transfected reporter vectors. As illustrated in Fig. 4B, reporter gene assays done at 24 hours posttransfection showed similar activation of pig3 and rgc reporter constructs by wt p53, whereas MDM2 promoter activity was approximately 10 times higher. The activities of mutant p53 proteins toward these promoters were evaluated relative to wt p53 24 and 44 hours after transfection (Fig. 4C). In concordance with the results obtained for the endogenous genes (Fig. 2), a strong induction of MDM2 and pig3 promoter-driven luciferase activity by the C277Y but not by the R273H mutant was observed. Significantly, induction of pig3 promoter activity at 24 hours after transfection by p53-C277Y was even 5 times higher than by the wild-type protein. MDM2 promoter activity in response to p53-C277Y closely approached wt p53–mediated activation (about 75% of wt). In contrast, the rgc response element was activated by neither of the two mutants. A further increase in relative MDM2 and pig3 promoter-driven reporter gene activities in C277Y versus wt p53–transfected cells at 44 hours has to be assigned at least in part to increased cellular mortality of wt p53–transfected cells. Cellular stress evoked by X-irradiation further contributed to increased pig3 promoter induction by p53-C277Y but did not affect the activities of either MDM2 or rgc promoters in transient transfection assays (Fig. 4D).

In addition, we tested the ability of synthetic small inhibitory RNA to p53 (siRNA-p53) to modulate pig3 promoter activity in STA-ET-2.2 cells (Fig. 5). A pig3 promoter-driven luciferase reporter vector was co-transfected with either synthetic siRNA-p53 or, for control, an irrelevant mismatched siRNA into STA-ET-2.2 cells. In concordance with previously published results (25) 30 hours post transfection, a clear reduction in p53 protein levels was observed with the specific siRNA but not with the irrelevant siRNA (not shown). Concomitantly, a 3-fold reduction in luciferase activity was observed in siRNA-p53–transfected cells. Taken together, these results are in concordance with a direct activation of MDM2 by p53-C277Y and imply a gain of transcriptional function of this mutant on the pig3 promoter.

The pig3 promoter region contains a putative p53 binding site between residues −328 and −312 that has previously been considered to be responsible for pig3 responsiveness to wt p53 (26). More recently, a microsatellite sequence between positions +442 and +516 has been suggested to mediate regulation of the gene by p53 (27). To test which portion of the pig3 upstream region is responsible for p53-C277Y-dependent activation, we compared four different reporter gene constructs in co-transfection experiments with expression plasmids for p53-C277Y, wt p53, and p53-R273H (Fig. 6). While the activity of p53-R273H in the reporter assay was about 100- to 170-fold lower than that of wt p53 with any of the constructs, p53-C277Y induced luciferase activity 4- to 12-fold higher than wt p53. The highest reporter activities were obtained using a 1.4-kb promoter fragment containing both the canonical p53 binding motif and the pentanucleotide microsatellite sequences. Mutation of the canonical p53 binding motif in the construct Pig3-delA did not significantly affect responsiveness to either wt p53 or p53-C277Y, indicating that this sequence is not involved in the regulation of the pig3 gene by either wild-type or mutant p53. Confirming previously published results (27), the microsatellite sequence comprised by 15 repeats of the TGYCC pentanucleotide motif was sufficient for reporter gene activation by wt p53. Intriguingly, p53-C277Y induced luciferase activity 12 times better than wt p53 from this construct. These results establish that p53-C277Y mediates pig3 gene activation via the microsatellite contained in exon 1 of the gene.

Because pig3 gene activation has been reported to be associated with the apoptotic function of p53 (28), we tested the ability of ectopically overexpressed p53-C277Y to induce cell death in four different EFT cell lines that differ in their endogenous p53 gene status. The apoptotic activity was tested 44 hours posttransfection in comparison to wt p53, to the DNA binding mutant p53-R273H, to the transactivation domain double mutant p53-22Q23S, and to the empty vector control (Fig. 7). Although some variability in the cell death-inducing activity of p53 expression constructs was observed in cell lines sensitive to wt p53, p53-C277Y showed the highest apoptotic activity among the three p53 mutants tested, which was significantly increased over the empty vector control. In contrast, in concordance with our previously published results (23), STA-ET-2.2 cells expressing endogenous p53-C277Y were completely resistant to all constructs, including wt p53. Insensitivity of this cell line to the deleterious effects of p53 accumulation is likely due to high Bcl2 and low APAF1 levels in this cell line (23) and may be envisaged as a protection mechanism against the residual apoptotic activity of p53-C277Y considering constitutively high p53 protein levels in this cell line (29).

In contrast to other tumor suppressors, which are usually completely abrogated by large deletions or frameshift mutations in cancer cells, point mutations resulting in single amino-acid substitution prevail for p53. Some of these variants retain the ability to activate a subset of p53 target genes. Most of these transcriptional “gain of function” mutations are considered to confer growth advantage to tumor cells and to contribute to cancer progression. Examples are p53-C174Y, which was shown to induce the fos promoter restricting growth suppression of transfected Saos-2 cells but failing to transactivate the MDM2, waf1, bax, and cyclin G promoters (30), and p53-D281G active on the human epidermal growth factor (EGFR) and human multiple drug resistance genes (MDR-1) and inducing increased expression of endogenous NF-κB and PCNA proteins (31). Using a yeast-based screening assay, two classes of p53 mutants retaining partial transactivation activity have been defined (21). The first group showed sequence-specific DNA binding activity to p21, and occasionally to SCS and cyclin G promoters even at physiologic temperature (37°C), whereas the second group of mutants with amino acid substitution mostly positioned in the β-scaffold of the p53 protein retained transcriptional activity only at significantly sub-physiologic temperatures. p53-C277Y belongs to the first group affecting an amino acid residue that directly contacts DNA and that regulates DNA binding affinity in a redox state-dependent manner (32). Replacement of cystein 277 by either histidine or tyrosine results in altered DNA binding specificity as shown previously using artificial promoters (18). Interestingly, however, the amino acid residue at position 402 in Caenorhabditis elegans p53 corresponding to mammalian C277 is a tyrosine. It is intriguing to speculate that the replacement of an amino acid at a specific DNA contact site during evolution did not completely change target selectivity of p53. We show here that p53-C277Y is capable of transactivating MDM2 and pig3 in the authentic tumor cell milieu, albeit with delayed kinetics, although in the yeast-based assay, this mutant did not activate these promoters when tested at 20°C, 30°C, and 37°C.²

In another study using a yeast transactivation assay, 16 of 77 tumor-derived p53 mutants displayed preferential binding to p21, but not to bax or pig3 promoters (33). p53-C277Y clearly differs from this group of mutants in that it strongly activated the pig3 promoter and that the transactivating ability was further increased on a DNA damaging stimulus. Szak et al. (26) defined the pig3 promoter as a low-affinity p53 binding site that may require interaction with additional factors possibly explaining delayed kinetics of pig3 induction by wt p53 (34). Here, the p53 binding element was located by chromatin immunoprecipitation-PCR to a region between nucleotides −441 and −166 containing a putative p53 binding site (−328 to −309; ref. 28). Recently, however, a pentanucleotide polymorphic microsatellite sequence further downstream (+442 to +516) has been shown to bind to p53 and to be necessary and sufficient to mediate pig3 gene activation (27). Our reporter gene assays not only confirm this observation for wt p53 but extend it to p53-C277Y which activated 12 times better than wt p53 from the isolated microsatellite.

It has been proposed that the inability of tumor-derived p53 mutants to induce apoptosis is correlated with a deficiency in pig3 gene activation, suggesting a direct involvement of PIG3 in p53-mediated cell death (27). In fact, we show residual apoptotic activity of p53-C277Y in transient transfection assays. These results suggest that tumors harboring this mutant have to be protected from p53-induced cell death by additional genetic alterations conferring resistance. In fact, the two cell lines, STA-ET-2.1 (not shown) and STA-ET-2.2, established from the tumor, from which p53-C277Y was isolated, were completely resistant to wt p53–induced cell death. Interestingly, these two cell lines were the only ones of a series of seven EFT cell lines showing a significantly increased Bcl2/Bax expression ratio and a low level of APAF1 expression (23).

In summary, p53-C277Y has been known for its altered DNA binding specificity on artificial promoters in in vitro tests. We provide the first evidence for constitutive and inducible transcriptional activity of p53-C277Y on MDM2 and a gain of function of this mutant on pig3, in the authentic tumor cell context that so far has escaped detection in yeast-based assays.

The cell line SK-N-MC was kindly provided by J. Biedler (Memorial Sloan Kettering Cancer Center, New York). The cell lines VH64, RM82, and WE68 were kindly supplied by F. Van Valen (Department of Pediatrics, University of Muenster, Germany), FPBH by M. Vetterlein (Institute of Tumour Biology and Cancer Research, University of Vienna, Austria), and SAL2 by G. Hamilton (Department of Surgery, University of Vienna, Austria). TC252 and IARC-EW2 were generously provided by T. Triche (Department of Pathology, Children's Hospital, Los Angeles, CA) and G.M. Lenoir (International Agency for Research on Cancer, Lyon, France), respectively. SK-ES1 and RDES were obtained from the American Type Culture Collection (Rockville, MD). The STA-ET series of cell lines was established at the Children's Cancer Research Institute (Vienna, Austria). The p53 gene status of these cell lines as indicated in Fig. 1A has previously been described (19). All EFT cell lines were grown on fibronectin-coated plastics in RPMI 1640 supplemented with 10% fetal calf serum at 37°C.

The luciferase reporter plasmid pBR322pI-LUC-pig3 carrying an approximately 6-kb–long pig3 gene promoter fragment (∼−5600 to ∼+550) was kindly provided by C. Polyak (Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA; ref. 28). Reporter plasmids carrying either a 1.4-kb promoter fragment from −861 to +546 (pGL3PIG3), or a derivative of this fragment from which the canonical p53 binding site between −330 and −309 has been deleted by oligonucleotide-directed mutagenesis (delA), or a microsatellite sequence comprised by 15 repeats of the TGYCC motif between +442 and +516 of the pig3 gene, have been generously provided by Matthias Dobbelstein (Institute of Virology, Philipps University, Marburg, Germany; ref. 27). The plasmid carrying the MDM2 gene promoter pmdm2-luc was kindly supplied by J. C. Bourdon (Department of Surgery and Molecular Oncology, Ninewells Hospital and Medical School, University of Dundee, Dundee, United Kingdom). The luciferase reporter plasmid carrying the rgc promoter was prepared by insertion of a double-stranded 22-oligomer TTTGCCTTGCCTGGACTTGCCTTT to the KpnI/MluI site within the plasmid pGL3-Basic. A GFP expression plasmid was constructed by insertion of the cytomegalovirus promoter into the multiple cloning site of pEGFP-1 (Clontech Laboratories, Palo Alto, CA). All mammalian expression vectors were based on pCMVNeo (35). The wt p53 and p53-R273H expression plasmids pC53-SN3 and pC53-4.2N3 were kindly supplied by B. Vogelstein (John Hopkins Oncology Center, Baltimore, MD). The p53-22Q23S encoding vector pRc/CMV-hp53 has been obtained from A. Levine (Princeton University, Princeton, NJ). The C277Y mutant that was cloned from the STA-ET-2.2 cell line (18) was inserted for expression into the BamH1 site of pCMV-Neo.

The sequence of the synthetic siRNA against p53 has previously been described (25). The sequence of the control irrelevant siRNA EF4M was CACCCACGGUCCUUCACAC with dTdT 3′ overhangs on both strands. The p53 specific retroviral small hairpin RNA expression vector pSR.p53 was generated by cloning the siRNA expression cassette from pSUPER.p53, purchased from OligoEngine (Seattle, WA), into pSUPER.retro (OligoEngine).

DNA (and DNA + siRNA) transfer into logarithmically growing EFT cells (0.5 to 1.0 × 10⁶ cells) was accomplished in six-well plates by lipofection of Qiagen purified plasmid DNA (Hilden, Germany) using LipofectAMIN PLUS reagent (Invitrogen, Paisley, United Kingdom) followed by 3 hours incubation in serum-free Optimem medium according to the manufacturer's instructions. Cells were washed with PBS and lysed 24 or 44 hours after transfection and luciferase activity was measured using Bright-Glo Luciferase Assay System (Promega, Madison, WI).

X-ray treatment of EFT cell lines (6 Gy) was done using a Phillips RT100 irradiation device (1.7 mm aluminum filter, 100 kV) at 12.5 Gy/minute. Irradiation of the transfected cells was done 12 hours after transfection. After treatment, cells were incubated at 37°C for the indicated times. Cytotoxicity was assessed by a vital dye exclusion test. Because dead and dying cells tend to detach from the culture plate, the floating and attached cells were separately collected, stained with trypan blue, and counted.

RNA extraction and Northern blotting were done according to standard procedures. The probes for human MDM2 exons 3 to 5 were generated by PCR amplification using primers ex3F (TGTCAATACCAACATGTCTG) with ex5R (CTGTGCTCTTTCACAGAGAAG). The probe for human MDM2 exon 2 was generated by PCR amplification using primers ex2F (CTGTGTTCAGTGGCGATTGGA) with ex2R (GGATCAGCAGAGAAAAAGTGG). The pig3 gene probe was kindly supplied by Dale Haines (The Fels Institute for Cancer Research and Molecular Biology, Philadelphia, PA). Where indicated, cells were incubated in the presence of 0.2 mol/L Z-Val-Ala-DL-Asp-fluoromethyl ketone (z-VAD.fmk; Bachem, Bubendorf, Switzerland) to monitor the expression of p53 downstream genes in the absence of p53-induced apoptosis.

For quantification of pig3 gene expression, real-time quantitative PCR was done using pig3-forward (ACCGCCTTCCAGCTGTTACA) and pig3-reverse (CACTCAGTCCTGCATGGATTAGC) primers. PCR reactions were set up in a total volume of 25 μL, including 5 μL template cDNA (corresponding to 100 ng), 12.5 μL TaqMan Universal Master Mix (Applied Biosystems, Foster City, CA), 900 nmol/L of each of the primers, and 200 nmol/L FAM-labeled probe (CTTGTGGGAAATGTTCAGGCTGGAGACTATG). Amplification was carried out using the ABI Prism 7700 or 7900 Sequence Detecting System (Applied Biosystems) for a total of 50 cycles. After an initial denaturation step for 10 min at 95°C, each cycle consisted of denaturation for 15 seconds at 95°C, and annealing and primer extension for 60 seconds at 60°C.

For protein expression analysis, cells were harvested 24 hours posttransfection and lysed in a buffer containing 7 mol/L urea, 100 mmol/L DTT, 50 mmol/L HEPES (pH 7.6), 100 mmol/L NaCl, 0.05% Triton X-100, 50 mmol/L NaF, and protease inhibitor mixture (Roche, Vienna, Austria). One microgram of total protein was separated by 10% SDS-PAGE, transferred onto a nitrocellulose membrane (Schleicher & Schuell, Keene, NH), and probed with the DO-1 monoclonal antibody kindly provided by B. Vojtesek (Masaryk Memorial Cancer Institute, Brno, Czech Republic). The antibody to PIG3 (Ab-1) was purchased from Oncogene Research Products (Boston, MA).

Fluorescence-activated cell sorter analyses were done with a FACSCalibur (BD, Sunnyvale, CA). Twenty-four or 44 hours after co-transfection with pCMV-EGFP, p53, and reporter luciferase plasmid, both adherent and floating cells were harvested and the percentage of GFP-positive viable cells was evaluated.

DNA content in transiently p53 or control-transfected cells was determined as previously described (23). Briefly, 44 hours after transfection adherent cells were combined with cells that had detached from the plastic and were fixed with 70% ice-cold ethanol for at least 1 hour. After rehydration and washing in PBS, cells were stained with antibodies CM1 or DO1 (a kind gift of B. Vojtesek, Masaryk Memorial Cancer Institute, Brno) for detection of p53 or with an antibody to β-galactosidase (Z3781, Promega), followed by propidium iodide (50 μg/mL). FITC-conjugated secondary anti-rabbit (F205) or anti-mouse (F313) antibodies were from DAKO (Vienna, Austria). DNA content in FITC-positive and -negative cells was assessed by two-color fluorescence-activated cell sorter analysis.

We thank M. Dobbelstein for providing several pig3 reporter vectors and helpful discussion and F. Watzinger for help with real-time PCR.