Transactivation of the EGR1 Gene Contributes to Mutant p53 Gain of Function

It is well accepted that inactivation of the p53 tumor suppressor gene is an important event in the process of tumor development (1). Elimination of the normal functions of p53, one of the key players in the maintenance of genome stability, leads to accumulation of genetic aberrations that eventually cause malignant transformation of normal cells (2). A high percentage of human tumors maintain overexpression of various mutant forms of p53 (3). Furthermore, a growing number of epidemiologic surveys suggest a significant correlation between overexpression of specific mutant p53 forms and tumor prognosis (4). However, it remains unclear whether different p53 mutants initially act via a negative dominant mechanism, in which the protein expressed from the residual wild-type p53 (wtp53) allele is inactivated by the concomitantly expressed mutant p53, or whether mutant p53 acts primarily by a gain of function mechanism (5). The idea that mutant p53 functions by gain of function is supported by the observations that overexpression of mutant p53 forms in p53-null cells accentuated their transformed phenotype (6). Several p53 mutants were reported to increase tumorigenicity in mice (7, 8), mutation frequency (9), and metastatic potential (8). Interestingly, it was found that mutant p53 interferes with stress-induced apoptosis and overexpression of different p53 mutants in p53-null cells conferred increased resistance to chemotherapeutic drugs (10, 11, 12). However, anti-apoptotic activity mediated by p53 “core” mutants was relieved when these p53 mutants were also genetically modified at their NH₂-terminal domain (13), corresponding to the transactivation domain of the wtp53 protein (14).

The fact that mutant p53 facilitates transcription of transformation-related genes was shown in several studies. Different viral and cellular promoters were shown to be transactivated by mutant p53, including the MDR-1 (7, 15), c-myc (16), interleukin-6 (17), heat shock protein 70 (18), human epidermal growth factor receptor (19), and HIV-1 LTR promoters (20). Furthermore, transcription inhibitors were shown to counter the mutant p53 anti-apoptotic gain of function effect (12). It is presently unclear whether mutant p53 regulates transcription of particular genes by binding directly to specific regulatory DNA sequences, through interaction with other DNA-binding proteins, or both. Of note, chromatin immunoprecipitation analysis indicates that different p53 mutants bind different DNA sequences (21). In search of signature genes regulated by mutant p53, we used Affymetrix DNA microarrays to compare gene expression patterns of p53-null cells and their mutant p53-expressing derivatives. Induction of Early Growth Response 1 (EGR1) gene expression was one of the significant changes observed upon overexpression of mutant p53 in human prostate cancer– and lung cancer–derived cells. The EGR1 protein is a M_r 59,000 transcription factor involved in various biological functions including regulation of proliferation, growth, apoptosis, and angiogenesis (22, 23). The fact that this protein plays an important role in pathways that pertain directly to cancer progression, prompted us to focus on the analysis of the relationship between mutant p53 expression and the induction of EGR1. We found that mutant p53 protein can induce the transcription of EGR1 and elevate the expression of EGR1 downstream genes such as vascular endothelial growth factor (VEGF). Furthermore, mutant p53 can associate with EGR1 gene promoter DNA sequences. Knockdown of EGR1 in mutant p53-overproducing cells attenuates colony formation under stress conditions and renders these cells more vulnerable to genotoxic stress. Together, these findings suggest that EGR1 plays an important role in a mutant p53-regulated oncogenic pathway.

H1299 cells were obtained from American Type Culture Collection (Manassas, VA). H1299 cells were maintained in RPMI (Sigma, St. Louis, MO) supplemented with 10% fetal calf serum (Sigma). PC3 and HeLa cells were maintained in Dulbecco’s modified Eagle’s medium (Gibco, Invitrogen, Carlsbad, CA) supplemented with 10% fetal calf serum. SKBR3 cells were maintained in McCoy’s medium (Sigma) supplemented with 10% fetal calf serum.

Plasmids, pCMV-neo-Bam-p53R175H carrying either Arg or Pro at position 72, were obtained from B. Vogelstein (Johns Hopkins University School of Medicine, Baltimore, MD) and W. Kaelin (Dana-Farber Cancer Institute, Boston, MA), respectively. A pCDNA3-based expression plasmid for p53R175H and p53R175H (22, 23) was constructed by A. Bren (Weizmann Institute, Rehovot, Israel) and G. Blandino (Regina Elena Cancer Institute, Rome, Italy), respectively. Plasmid pM5neo-ecotropic-R, encoding the mouse ecotropic receptor, was provided by S. Benchimol (The Ontario Cancer Institute, Toronto, Canada). pBabe-puro-175H was constructed by removing the p53 open reading frame from pCMV-neo-Bam-p53R175H and inserting it into pBabe-puro.

The EGR1 expression plasmid was constructed by cloning the EGR1 open reading frame, obtained by reverse transcription-PCR with specific primers (sense, CCC GGA TGG CCG CGG CCA AGG; and antisense, GGC CAT CTC CTC CTC CTG TCC) into pcDNA3.1/V5-His-TOPO using the pcDNA3.1/V5-His-TOPO TA expression kit (Invitrogen, Carlsbad, CA). EGR1-luc was constructed by cloning the EGR1 promoter, obtained by genomic PCR with specific primers (sense, GCC ACA CCC GGA AAG ACA C; and antisense, CTG GAC GAG CAG GCT GGA GAG), into the pGL3-Basic luciferase plasmid (Promega, Madison, WI).

Infections were carried out using a standard protocol. In brief, PC3 cells were stably transfected with pM5neo-ecotropic-R, encoding the mouse ecotropic receptor, and selected for 2 weeks in 0.4 mg/mL G418. Concurrently, 293T cells were transfected with the retroviral construct pBabe-puro-175H, using Fugene (Roche, Switzerland) according to the manufacturer’s instructions. Virus-containing culture supernatants were harvested 24 to 48 hours post transfection at 6-hour intervals and pooled together. The stably transfected PC3 cells were then infected with the filtered supernatants and selected for 48 hours in 2 μg/mL puromycin, resulting in cell populations stably expressing mutant p53R175H. Transfections were carried out using Fugene (Roche) at a ratio of 2 μg of DNA:4 μL of Fugene, according to the manufacturer’s instructions. H1299 cells were stably transfected with p53R175H followed by G418 selection (0.4 mg/mL) for 3 weeks to achieve single-cell clones.

PC3 cells, either noninfected or infected with mutant p53R175H, were plated on 6-cm dishes. Cells were then transfected with various plasmid combinations, with the aid of Fugene 6 (Roche), using a standard protocol. A plasmid encoding H2B-GFP was included in all transfections. Twenty-four hours after transfection, cells were treated with either 10 μmol/L etoposide or 1 μg/mL cisplatinum for either 48 or 72 hours. Cells were subsequently trypsinized and fixed in 80% methanol/20% PBS for at least 25 minutes. Cells were then rehydrated for at least 30 minutes in PBS, washed, resuspended in PBS containing 25 μg/mL propidium iodide (Sigma) and 10 μg/mL RNase A, and subjected to fluorescence-activated cell sorter-based cell cycle analysis. Only green fluorescent protein (GFP)-positive cells, expected to represent the successfully transfected subpopulation of the culture, were included in the analysis.

Nuclear extracts were prepared as described previously (24). Cells (10⁶∼10⁷) were washed twice with cold PBS and harvested. Cell pellets were resuspended by gentle pipetting in 400 μL of buffer A [10 mmol/L HEPES-KOH (pH 7.9), 1.5 mmol/L MgCl₂, 10 mmol/L KCl, 0.5 mmol/L dithiothreitol, 0.2 mmol/L phenylmethylsulfonyl fluoride (PMSF), and protease inhibitor mixture (Roche)]. After 15 minutes of incubation on ice, 25 μL of 10% NP40 were added, and vortexing was performed vigorously for 10 minutes. After centrifugation, cell pellets were resuspended in approximately 150 to 300 μL of buffer C [20 mmol/L HEPES-KOH (pH 7.9), 25% glycerol, 420 mmol/L NaCl, 1.5 mmol/L MgCl₂, 0.2 mmol/L EDTA, 0.5 mmol/L dithiothreitol, 0.2 mmol/L PMSF, and protease inhibitors mixture (Roche)]. Tubes were transferred to a rotating platform for 15 minutes at 4°C and then centrifuged, and the protein content of the supernatant was determined by the Bradford procedure. Aliquots were stored at −70°C.

Cells were lysed in passive lysis buffer (Promega). Lysate aliquots were resolved by SDS-PAGE on a 7.5% polyacrylamide gel, transferred to a nitrocellulose membrane, and probed sequentially with EGR1 antibody (sc-189; Santa Cruz Biotechnology, Santa Cruz, CA), a mixture of the p53-specific monoclonal antibodies PAb1801 and DO-1, and finally either antivinculin or antitubulin antibody (Sigma). Membranes were then reacted with secondary goat-antimouse and goat-antirabbit horseradish peroxidase–conjugated antibodies (1:10,000; Jackson) and developed using the ECL kit (Amersham Biosciences, Uppsala, Sweden).

Increasing amounts of each DNA sample were mounted onto a nitrocellulose membrane. The membrane was washed with PBS, dehydrated at 60°C overnight, and then washed twice in PBS and blocked with 5% skim milk in PBS for 3 hours. Nuclear extracts were prepared as described and diluted to a protein concentration of 0.5 mg/mL in gel-shift buffer [12.5 mmol/L Tris-HCl (pH 7.9), 3.1 mmol/L MgCl₂, 25 mmol/L KCl, 0.5 mmol/L dithiothreitol, 10% glycerol, 0.25 mmol/L EDTA, 0.2 mmol/L PMSF, protease inhibitors mixture (Roche), and 1 mg/mL pGL3-Basic plasmid DNA for blocking]. Nuclear extracts were incubated overnight at 4°C with the membrane, followed by one 5-minute wash in PBS-Tween. The membrane was next incubated for 1 hour with a p53-specific polyclonal serum under standard Western blot analysis conditions, followed by three consecutive 5-minute washes in PBS-Tween and incubation for 30 minutes with horseradish peroxidase–conjugated goat-antirabbit secondary antibody. Membranes were developed with an ECL kit (Amersham Biosciences).

Formaldehyde (Merck, Darmstadt, Germany) was added directly into cell culture medium to a final concentration of 1%. Fixation proceeded at room temperature for 10 minutes and was stopped by addition of glycine to a final concentration of 0.125 mol/L. Dishes were rinsed with cold PBS and incubated with 5 mL of 20% trypsin-EDTA (Gibco) in PBS. Cells were removed by scraping, collected by centrifugation, and washed in cold PBS. Cells were incubated on ice for 20 minutes in 5 mmol/L Pipes (pH 8.0), 85 mmol/L KCl, and protease inhibitors mixture; nuclei were dounced and collected by microcentrifugation at 4,000 rpm, resuspended in nuclei lysis buffer [1% SDS, 10 mmol/L EDTA, 50 mmol/L Tris-HCl (pH 8.1), and protease inhibitors mixture], and incubated on ice for 10 minutes. Samples were sonicated on ice to an average length of 600 to 1,000 bp and then microfuged at 14,000 rpm. The chromatin solution was precleared by incubation with protein A beads for 2 hours at 4°C. Before use, protein A beads were blocked with 1 μg/mL sheared herring sperm DNA and 4% bovine serum albumin for at least 4 hours at 4°C. Precleared chromatin from 2.5 × 10⁷ cells was diluted 1:10 in dilution buffer (0.01%SDS, 1% Triton X-100, 1.2 mmol/L EDTA,167 mmol/L NaCl, and protease inhibitors mixture) and divided into 600-μL aliquots. Each aliquot was mixed with 30 μL of protein A beads cross-linked by dimethyl pimelimidate to anti-p53 polyclonal antibody and rotated at 4°C for 12 hours. Beads subjected to a similar process in the absence of antibody were used as negative control. Immunoprecipitates were washed twice with dilution buffer, twice with washing buffer [100 mmol/L Tris-Cl (pH 9.0), 500 mmol/L LiCl, 1% NP40, 1% deoxycholic acid, and PMSF], and once with Tris EDTA buffer. Elution of immune complexes was carried out by addition of 50 μL of elution buffer (50 mmol/L NaHCO₃ and 1% SDS). The samples were diluted again to 0.5 mL, and a second round of immunoprecipitation, washing, and elution was performed under the same conditions. Proteins and RNA were removed by addition of 10 μg of RNase A per sample for 30 minutes, followed by addition of 30 μg of proteinase K for 2 hours at 42°C. Cross-links were reversed by incubation at 65°C overnight. DNA was extracted once with phenol:chloroform and once with chloroform and precipitated with 0.1 volume of 3 mol/L NaAc and 2.5 volume of ethanol. Pellets were collected by microcentrifugation, resuspended in 30 μL of water, and analyzed by PCR. Total input samples were resuspended in 100 μL of water and then diluted 1:10. PCR reactions contained 2 μL of immunoprecipitated or diluted total input, 50 ng of each primer (EGR1 forward, 5′-GCG GTA CCG GGC AGC ACC TTA TTT GGA G; EGR1 reverse, 5′-GCG GTA CCC ACT CCC GGT TCG CTC TCA; glyceraldehyde-3-phosphate dehydrogenase (GAPDH) forward, 5′-GTA TTC CCC CAG GTT TAC AT; and GAPDH reverse, 5′-TTC TGT CTT CCA CTC ACT CCT), 10% dimethyl sulfoxide, and Ready mix PCR master mix (Promega) in a total volume of 20 μL. After 35 cycles of amplification, PCR products were run on a 1.5% agarose gel and analyzed by ethidium bromide staining.

Cells were seeded in 24-well culture dishes. Each well was transfected with a reporter plasmid expressing the firefly luciferase gene under the transcriptional control of the different gene promoters, together with increasing amounts of various expression plasmids and β-galactosidase (β-gal) plasmid. Luciferase activity was assayed 48 hours post transfection. Each plasmid combination was transfected into three identical wells. Luciferase assays were performed using (D)-luciferin (Roche). Luminescence was determined with the aid of a Rosys-Anthos Lucy 3 luminometer. The luciferase values were normalized to β-gal activity.

Total RNA was extracted using the Trireagent kit (Macherey Nagel, Germany). Two μg of each RNA sample were reverse transcribed using Moloney murine leukemia virus reverse transcriptase (Promega, Madison, WI) and random hexamer primers. Real-time PCR was performed on an ABI 7000 machine (Applied Biosystems) using Sybr Green PCR mastermix (Applied Biosystems), EGR1-specific primers (sense, TTT GCC AGG AGC GAT GAA C; antisense, CCG AAG AGG CCA CAA CAC TT) and p53-specific primers (sense, CCC AAG CAA TGG ATG ATT TGA; antisense, GGC ATT CTG GGA GCT TCA TCT). cDNA levels were normalized to GAPDH amplified with appropriate primers (sense, ACC ACA GTC GCC ATC AC; antisense, TCC ACC ACC CTG TTG CTG TA).

Isogenic mutant p53 producer cell lines were established from well-characterized p53-null cell lines. To that end, we used H1299 lung adenocarcinoma cells and PC3, a prostate cancer–derived cell line. Mutant p53-overexpressing cell lines were generated by either transfection or infection of the p53-null parental cells with a vector encoding mutant p53 protein. We chose to focus on p53R175H, a common p53 mutant previously shown to possess a marked anti-apoptotic gain of function in H1299 cells (11). Single cell–cloned H1299 derivatives and PC3 polyclonal pools were selected in the presence of the corresponding drugs, and levels of p53 protein were determined by Western blot analysis (Fig. 1 A and B). Ten individual H1299-derived single-cell clones expressing p53R175H were selected for additional analysis. PC3-derived cell populations were analyzed as polyclonal pools without additional subcloning.

To search for specific mutant p53-regulated genes, gene expression patterns of p53-null and derivative mutant p53-producing H1299 cells were analyzed. Differential expression patterns were determined by the use of Affymetrix DNA microarrays. To avoid clonal variability and improve the statistical strength of the study, equal amounts of RNA obtained from individual clones were pooled together before cDNA synthesis. cDNA was subsequently prepared and subjected to hybridization with DNA microarrays. Data obtained were analyzed by the Affymetrix analysis criteria. Genes that are either up-regulated or down-regulated after expression of mutant p53 were identified (Table 1) and postulated to represent genes directly affected by mutant p53 as well as genes whose expression is modulated by secondary events occurring after mutant p53 expression. One of the genes whose expression was consistently and prominently induced by mutant p53 was EGR1, and this gene was therefore chosen for additional study. Quantitative reverse transcription-PCR analysis confirmed that EGR1 mRNA levels were indeed elevated to a similar extent in both H1299 and PC3 cells as a consequence of mutant p53 overexpression (Fig. 1 C). It should be noted, however, that this increase in basal levels of EGR1 mRNA was not always easy to detect: Although in some experiments it was quite prominent, in others, it was only minor or even undetectable (data not shown). As discussed later, a more unequivocal effect of mutant p53 on EGR1 mRNA levels became apparent under conditions that lead to stronger stimulation of EGR1 gene expression.

To examine the possibility that up-regulation of EGR1 expression is mediated by transactivation of the EGR1 promoter by mutant p53, we cloned the EGR1 promoter sequence upstream to a luciferase reporter gene and transfected it into cells together with mutant p53 expression plasmids. H1299 cells were transfected with the EGR1-luciferase plasmid and increasing concentrations of R175H and several other tumor-associated p53 hot spot mutants. As seen in Fig. 2,A (left panel), mutants R175H, R248W, R273H, and R281G elicited a dose-dependent increase in luciferase activity, whereas no increase was noticed with H179E. For most mutants, comparable protein levels were produced in the transfected cells (Fig. 2 B). These results suggest that p53R175H, as well as most of the common cancer-associated p53 mutants, can up-regulate EGR1 expression via transactivation of the EGR1 promoter.

We reported previously (13) that core mutant p53 protein with additional mutations at residues 22 and 23 in the transactivational domain (p53R175H,22-23) was unable to block drug-induced apoptosis, suggesting that such p53 mutants require an intact transactivation domain for their anti-apoptotic “gain of function” activity. It was therefore of interest to examine whether such transcription-deficient p53 mutants have also lost their capacity to transactivate the EGR1 promoter sequence. As seen in Fig. 2,A (right panel), overexpression of p53R175H,22-23 in two different cell lines did not elicit a significant increase in luciferase activity, even though it was expressed at levels comparable with those of p53R175H (Fig. 2,B). This implies that the activation of the EGR1 promoter by mutant p53 requires the integrity of the transactivation domain. Of note, the EGR1promoter behaved very differently from classical p53-responsive promoters, which are transactivated preferentially by wtp53 but not by mutant p53 (Fig. 2,C). A Western blot representing the relative p53 protein levels in this experiment is shown in Fig. 2 D.

The increase in EGR1 mRNA levels and EGR1-luciferase activity suggests that mutant p53 specifically transactivates the EGR1 gene. We therefore wished to determine whether mutant p53 could interact physically with EGR1 promoter sequences. To that end, chromatin immunoprecipitation analysis was performed on H1299 cells transfected with either p53R175H or wtp53; nontransfected cells served as a control. As seen in Fig. 2,E, EGR1 promoter sequences were selectively immunoprecipitated with mutant p53. Thus, mutant p53 is selectively associated in vivo with the EGR1 promoter. A markedly fainter signal was obtained with wtp53; whether this reflects a real, albeit weak interaction between the EGR1 promoter and wtp53 is presently under investigation. To further confirm the results in cells expressing endogenous mutant p53, we performed chromatin immunoprecipitation analysis on SKBR3 cells naturally harboring the R175Hp53 mutant. As seen in Fig. 2 F, EGR1promoter sequences were immunoprecipitated specifically by the anti-p53 antibodies but not by the irrelevant control anti-hemagglutin antibodies.

Western blot analysis was performed to confirm that the increase in EGR1 mRNA is also accompanied by a corresponding increase in EGR1 protein. It should be noted however, that detection of EGR1 protein in nonstressed cells is difficult (25, 26). Therefore, to facilitate additional analysis, we subjected p53R175H-producing cells and their empty vector controls to serum starvation for 24 hours, followed by serum re-addition. As seen in Fig. 3,A, EGR1 protein levels increased gradually within the 1st hour after serum addition. This increase was significantly more pronounced in cells expressing p53R175H compared with control H1299 cells (Fig. 3 A). Thus, mutant p53 increases not only EGR1 mRNA but also EGR1 protein levels.

To confirm that the effect of mutant p53 on EGR1 levels can be seen also under other stress conditions that lead to EGR1 elevation, we treated cells with desferrioxamine, which is a hypoxia-mimetic drug; phorbol ester 12-myristate 13-acetate (phorbal myristate acetate or 12-O-tetradecanoylphorbol-13-acetate), which is a PKC activator; and several DNA damaging agents (cisplatin, etoposide, and doxorubicin) and monitored the relative levels of EGR1 protein. As seen in Fig. 3 B and C, under all of those conditions, cells expressing mutant p53 showed significantly elevated EGR1 levels compared with cells that do not express any p53.

Our findings indicate that EGR1 is a specific mutant p53 target gene. To find out whether EGR1 induction is a downstream event that contributes to mutant p53 gain of function activity, we asked whether knockdown of EGR1 expression by RNA interference (RNAi) would compromise such gain of function activity. To that end, an appropriate EGR1-specific sequence was inserted into the pSUPER plasmid, which directs the synthesis of small interfering RNA (siRNA) in mammalian cells. Similar plasmids directing the synthesis of either p53 siRNA carrying three mismatches (mp53 RNAi) or mouse p63 siRNA served as specificity controls. Transfection of the EGR1 siRNA plasmid (EGR RNAi) caused a specific, albeit partial, down-regulation of endogenous EGR1 mRNA (Fig. 4,A) and protein (Fig. 4,B) in cells overexpressing mutant p53; no such down-regulation was seen when either pSUPER empty vector (data not shown) or mp53 RNAi (Fig. 4 A and B) was used as a control.

Next, we asked whether knockdown of EGR1 expression has an effect on mutant p53 biological activities associated with growth control and apoptosis. In the first set of experiments, we compared the efficiency of soft agar colony formation by H1299 cells subjected to various genetic manipulations. Cells were transfected either with an empty vector or with p53R175H, with or without EGR1 RNAi. Two days post transfection, cells were seeded in soft agar, and colonies were counted 1 month later. As seen in Fig. 5,A (top panel), expression of mutant p53 in H1299 cells elevated significantly the number of colonies, whereas transfection of EGR1 RNAi compromised this elevation. To confirm that the RNAi expression plasmid was functional in this system, we performed real-time PCR analysis of EGR1 mRNA 2 days after transfection. As shown in Fig. 5 A (bottom panel), levels of EGR1 mRNA were significantly reduced in H1299 175H cells transfected with the EGR1 RNAi plasmid.

To further assess the role of EGR1 in mutant p53 gain of function, the effect of EGR1 knockdown on the plating efficiency of tumor cells overexpressing mutant p53 was measured. As seen in Fig. 5 B, overexpression of mutant p53 in p53-null H1299 cells resulted in a marked increase in plating efficiency. Importantly, EGR1 RNAi reversed to a great extent the enhancement of colony formation by mutant p53. This further supports the conjecture that elevation of EGR1 expression by mutant p53 plays a role in the gain of function activities of mutant p53, as reflected in the ability of cells to grow under stress.

One of the well-documented facets of mutant p53 gain of function is an increased resistance of tumor cells to stress-induced apoptosis (11, 12, 13). In view of the proposal that EGR1 is a downstream target of mutant p53, we investigated the effect of EGR1 expression levels on cell survival under different kinds of stress, particularly those that result in elevation of EGR1 (see Fig. 3).

Parental PC3 cells or PC3 cells overexpressing p53R175H were transfected with a plasmid expressing either EGR1 siRNA or a siRNA control plasmid (mouse p63 siRNA). A plasmid encoding H2B-GFP was included in all transfections to enable identification of successfully transfected cells. After DNA damage (1 μg/mL cisplatinum for 48 hours), GFP-positive cells were analyzed by fluorescence-activated cell sorter for cell cycle distribution. As seen in Fig. 6 A, cells expressing mutant p53 underwent apoptosis at consistently lower rates than parental p53-null cells. On the other hand, both parental and mutant p53-expressing cells exhibited an increase in apoptosis after transfection with an EGR1 RNAi plasmid. Of note, EGR1 knockdown in mutant p53 overexpressors restored apoptosis to a rate comparable with that seen in parental p53-null cells expressing basal levels of EGR1.

To determine whether EGR1 can contribute to cell survival also in tumor cells harboring endogenous mutant p53, we used SKBR3 human breast cancer-derived cells carrying a mutation at position 175 of their endogenous p53. Knockdown of either p53 or EGR1 led to a significant increase in cell death when the cultures were treated with increasing concentrations of desferrioxamine (Fig. 6 B).

Real-time reverse transcription-PCR analysis confirmed that EGR1 mRNA in transfected cells was down-regulated by both EGR1 RNAi and p53 RNAi, but not by RNAi directed specifically against mouse p63, serving as an irrelevant control (Fig. 6,C). In contrast, p53 mRNA levels were reduced only when a p53 RNAi plasmid was transfected (Fig. 6 D). It is noteworthy that the apparent extent of reduction is most certainly an underestimate of the true knockdown efficacy, because the transfection efficiency was well below 100%, and thus the RNA analyzed is derived from a mixture of transfected and nontransfected cells.

Moreover, as seen in Fig. 7 (top panel), down-regulation of either p53 or EGR1 in SKBR3 cells resulted in a significant increase in cell death after exposure to DNA-damaging anticancer drugs; the observed increase is probably an underestimate, because transfection efficiency was no more than 50%, as assessed with the aid of a cotransfected GFP expression plasmid (data not shown). Real-time reverse transcription-PCR analysis confirmed that the relative EGR1 mRNA levels were reduced when cells were transfected with EGR1 RNAi or p53 RNAi plasmids compared with cells transfected with control RNAi plasmid (Fig. 7, bottom panel). Here, too, the actual extent of knockdown is likely to be markedly higher than apparent from the data, owing to the limited transfection efficiency. These data confirm the notion that the endogenous mutant p53 in such tumor cells contributes, at least to some extent, to increased resistance of the cells to chemotherapeutic agents. Taken together, our observations strongly suggest that EGR1 knockdown effectively decreases the anti-apoptotic gain of function effect of mutant p53. Hence, the induction of EGR1 expression by mutant p53 is proposed to play a significant role in this gain of function.

VEGF, a major regulator of the response to hypoxia, was previously reported to be up-regulated by mutant p53 (27, 28). At least according to some studies, VEGF expression is also positively regulated by EGR1, albeit probably not in a direct manner (26). We therefore compared the induction of VEGF mRNA by desferrioxamine in H1299 cells and in their mutant p53-overexpressing derivatives. Although basal VEGF mRNA levels were not appreciably higher in cells overexpressing mutant p53, the extent of VEGF induction by desferrioxamine was remarkably higher in these cells compared with the parental, p53-null cells (Fig. 8,A). The up-regulation of VEGF expression by mutant p53 is in agreement with earlier findings (27, 28). Furthermore, in SKBR3 cells expressing endogenous p53R175H, knockdown of either the endogenous mutant p53 or EGR1 resulted in a mild but reproducible drop in the concentration of VEGF protein secreted into the cell culture medium (Fig. 8,B). The effect of mutant p53 knockdown was somewhat more pronounced than that of EGR1 knockdown; this might imply that EGR1 up-regulation may be only one of several mediators of the effect of mutant p53 on VEGF expression. These data are consistent with the notion that induction of EGR1 gene expression by mutant p53 mediates, at least in part, the ability of mutant p53 to up-regulate VEGF promoter activity. This is further supported by DNA binding analysis: As shown in Fig. 8 C, the p53R175H protein did not bind to a plasmid carrying the VEGF promoter while exhibiting good interaction with the CD95/Fas promoter, which was recently found to be a target for binding and transcriptional repression by mutant p53 (29). Hence, the VEGF promoter is not targeted directly by mutant p53. Our finding that RNAi-mediated down-regulation of EGR1 expression leads to a reduction in the expression of VEGF is in agreement with a previous report (26) that down-regulation of EGR1 in mice leads to a significant reduction in VEGF after hypoxic stress. It should be noted, however, that other studies do not support the conclusion that EGR1 is a significant regulator of VEGF expression (30, 31). This discrepancy might imply that EGR1 can regulate VEGF expression only in certain cell types or under specific experimental conditions.

All together, the data discussed above strongly suggest that mutant p53 employs EGR1 as a mediator of some of its downstream gain of function tumor-promoting effects.

In the present study, we examined the involvement of EGR1 in mutant p53 gain of function. Using expression microarray analysis, EGR1 was identified as one of the prominent genes up-regulated in tumor cells overexpressing the p53R175H mutant. This up-regulation, seen with a number of different cancer-associated p53 hotspot mutants, was also confirmed at the EGR1 protein level. Moreover, promoter analysis and chromatin immunoprecipitation assays indicated that the EGR1 gene is a direct transcriptional target of p53R175H. The EGR1 protein induced by p53R175H mutant p53 is biochemically active, as manifested by its ability to mediate the up-regulation of VEGF expression (Fig. 8), as well as of two other EGR1 target genes, fibronectin 1 and TGF-β1 (data not shown). It is of particular note that the two p53 mutants most active in up-regulation of the EGR1 promoter, R248W and D281G (Fig. 2 A), were also found to be the most potent in biological in vitro and in vivo assays for mutant p53 gain of function (7). Although only correlative, this might imply that the oncogenic gain of function of mutant p53 is closely coupled with its ability to up-regulate EGR1 gene expression. For the sake of simplicity, we chose to focus on data from experiments employing p53R175H. However, a parallel albeit somewhat more limited analysis was also performed with another common p53 mutant, p53R273H; overall, p53R273H was found to behave qualitatively in a similar manner to p53R175H with regard to EGR1 activation and its functional consequences (data not shown). It is of note that p53R175H and p53R273H represent different classes of p53 mutants: p53R175H is a gross conformational mutant, whereas p53R273H is a DNA contact site mutant. Our conclusions may thus be relevant to a broad range of p53 mutants implicated in human cancer.

EGR1 is a protein encoded by an immediate early growth response gene, which is rapidly and transiently induced by various factors including growth factors, DNA damaging agents, and stress mediators (32). Up-regulation of EGR1 expression may result in apparently contradictory activities including mitogenesis (33, 34), differentiation (35, 36), tumor suppression (37), apoptosis, and protection from apoptosis (38, 39). It was suggested that EGR1 exerts its activity by regulating the expression of different genes involved in various pathways. EGR1 regulates the insulin-like growth factor-II, platelet-derived growth factor-A and platelet-derived growth factor-B genes, which are known to be involved in cell proliferation (40, 41, 42); BCL-2, fibronectin, and nuclear factor-κB, which are associated with survival and cell differentiation (43, 44, 45); as well as p53, PTEN, and tumor necrosis factor-α, which are involved in apoptosis (39, 46, 47). Additional relevant targets are VEGF and tissue factor, which are associated with tumor progression (32, 48), and p57/KIP2 and TGFβ1, which induce growth inhibition in a cell type–dependent manner (49, 50). The fact that EGR1 itself is up-regulated by several growth factors and oncogenes supports the notion that it functions as a growth-promoting protein. Although its induction is generally transient and greatly dependent on the nature of the various inducers, it appears to be sustained in a high proportion of prostate cancer cell lines and tumors, suggesting that EGR1 stimulates tumor cell growth in certain types of cancer (51). Indeed, EGR1 overexpression promotes growth in several systems, including kidney and endothelial cells (52). In contrast, in breast, lung, and brain tumors, EGR1 is down-regulated and when re-expressed results in growth suppression and apoptosis (53, 54, 55). EGR1 was also observed to be required for apoptosis in normal and tumor cells and to stimulate differentiation (56). It seems that EGR1 lies at a convergence point, and its effects depend on the signals transduced and the cell context.

Because EGR1 is strongly and frequently up-regulated in prostate tumors, several studies have addressed the importance of this protein in the development of prostate cancer. It was shown that in the prostate tumor mouse model (TRAMP mice), tumor progression depends on EGR1 expression (57). Studies of human prostate tumors show strong correlation between the level of EGR1 expression and tumor grade (51). Furthermore, after hypoxia in tumor cell masses that become depleted of oxygen, EGR1 is induced and transactivates growth factors, proteases, and matrix molecules that are responsible for tumor cell proliferation, survival, and development of new blood vessels that determine tumor progression (30, 58, 59). Altogether, although the contribution of EGR1 to cancer is complex and seemingly even contradictory, it is evident that in some types of cancer, it has a substantial positive role in tumor promotion.

As suggested by our findings, induction of EGR1 contributes in a number of ways to the gain of function effects of mutant p53, and down-regulation of EGR1 attenuates the oncogenic and anti-apoptotic effects of overexpressed mutant p53. Thus, in some tumors naturally overexpressing mutant p53, the predicted induction of EGR1 is likely to contribute to tumor progression. This may pertain particularly to types of cancer in which elevated EGR1 expression is known to be associated with tumor progression, such as prostate cancer (22). In this regard, it is of note that many of the experiments described here used PC3 prostate carcinoma cells. Our data raise the possibility that a similar functional relationship between mutant p53 gain of function and EGR1 activation may also exist in some cases of additional tumor types, as represented by the H1299 and SKBR3 cell lines.

The picture appears to be practically opposite in tumor cells retaining wtp53 expression. In fact, numerous studies indicate that EGR1 can enhance the growth inhibitory and anti-apoptotic effects of wtp53 (46, 60). EGR1 can directly activate the p53 gene promoter and induce an increase in p53 mRNA (46). In cells retaining wtp53 expression, this might restrict the growth-promoting and anti-apoptotic activities of EGR1 and serve as an effective functional negative feedback loop. In this regard, it is noteworthy that EGR1 appears to exert pro-oncogenic effects in lung cancer cells overexpressing mutant p53 (this study), whereas elevated expression of EGR1 was found to correlate positively with apoptosis in H460 human lung carcinoma cells (61), which express wtp53.

Our findings suggest that the contribution of EGR1 is reversed when tumor cells undergo p53 gene mutations. In such cells, mutant p53 will actually drive a positive feedback loop. The activation of the EGR1 promoter by mutant p53 will lead to elevated EGR1 protein, which in turn will activate the p53 promoter and contribute to the maintenance of constitutive high levels of mutant p53.

It is noteworthy that EGR1 can stimulate not only p53 expression but also p73 expression, and the latter may contribute to the pro-apoptotic effects of EGR1 (62). Furthermore, joint activation of wtp53 and p73 by EGR1 may have a synergistic apoptotic outcome, because these two p53 family members cooperate in the transactivation of pro-apoptotic target genes (63). Conversely, in tumor cells naturally overexpressing mutant p53, the possible induction of p73 by EGR1 is expected to be functionally neutralized through the physical binding and subsequent inactivation of p73 by the mutant p53 protein (10, 64, 65). Hence, one important aspect of p53 mutations in cancer may be the switching of EGR1 from a transformation-restrictive, tumor-suppressing molecule into an overt positive contributor to oncogenesis.

In conclusion, our data reveal a novel connection between p53 and EGR1, and imply that p53 status in a given tumor may serve as a critical determinant in directing EGR1 either toward antiproliferative effects in cells expressing wtp53 or toward tumor-promoting effects in cells harboring mutant p53.