Benzo[a]pyrene activates the human p53 gene through induction of nuclear factor kappaB activity.

p53 is known to be recruited in response to DNA-damaging genotoxic stress and plays an important role in maintaining the integrity of the genome. In the present study, the effect of a potent lung cancer carcinogen, benzo[a]pyrene (B[a]P) on p53 expression was investigated. We showed that exposure of A549 and NIH 3T3 cells to B[a]P resulted in an increase in p53 mRNA levels and in p53 promoter activation, indicating that B[a]P-induced p53 expression is partly regulated at the transcriptional level. The p53 promoter region which extends from -58 to -43, overlapping the kappaB motif, is essential for both the p53 basal promoter activity and p53 promoter activation induced by B[a]P. Nuclear factor kappaB (NF-kappaB) proteins have been revealed to be activated in B[a]P-induced p53 expression. Activated NF-kappaB complexes were shown to contain predominantly p50 and p65 subunit components in A549 cells and p65 subunit in NIH 3T3 cells. In addition, the overexpression of IkappaBalpha completely inhibited NF-kappaB activation, p53 promoter transactivation and the stimulatory effect on p53 transcription induced by B[a]P. We therefore conclude that B[a]P transcriptionally activates the human p53 gene through the induction of NF-kappaB activity.

Lung cancer is one of the most prevalent cancers in the world, and its mortality is expected to remain very high for many years to come (1). Concurrently, the environmental air quality is deteriorating and the number of smokers has increased. Epidemiological studies over the past several decades have provided a considerable body of evidence linking lung cancer to a number of mutagens and carcinogens detected both in the environment and cigarette smoke.
Polycyclic aromatic hydrocarbon (PAH) 1 carcinogens, such as benzo[a]pyrene (B[a]P), are products of incomplete combustion of organic matter and are widespread in the environment. Carcinogenic and mutagenic effects of B[a]P have been well documented in human, animals, and mammalian cell systems (2,3). Most PAH require metabolic activation to vicinal bay-region or fjord-region dihydrodiol epoxides in order to express their carcinogenic activity and, in general, the PAHs are among the more potent known experimental carcinogens (4,5). Active metabolites bind covalently to DNA and thus result in DNA damage (4,6).
The continuous exposure of cells to exogenous and endogenous DNA-damaging stress is implicated in the pathogenesis of various cancers. A host of complex cellular DNA repair mechanisms have evolved to counteract the deleterious effects of DNA damage in humans (7). In addition to the existence of several detoxification systems, delays at specific stages of the cell cycle occur after DNA damage, most likely for the efficient and optimal removal of DNA lesions from the cellular genome before cell division (8,9). The apoptosis of cells has been shown to be necessary for the elimination of potential precursor cells escaping DNA repair (10). p53, a tumor suppressor protein, is believed to play an integral role in such cellular response pathways to DNA damage (9,11,12).
The p53 gene is one of the most commonly mutated genes identified in various types of human tumors, and the results of numerous studies suggest that the inactivation or abnormality of p53 is a critical step leading to neoplastic transformation (13). It is important to maintain the integrity of the genome (14). A loss of the p53 functions thus results in an enhanced frequency of genomic rearrangements or genomic instability (15,16) and also eliminates the growth arrest response or apoptosis induced by DNA-damaging genotoxic insults (12,17). In accordance with this function, p53 is known to be recruited in response to various DNA-damaging agents such as UV, ␥-irradiation, and anticancer drugs (12,18,19), and functions as a transcription factor to induce expression of various cellular genes involved in cell cycle control, such as a gene for p21, a cyclin-Cdk inhibitor that blocks cell cycle progression (20,21). The increased cellular p53 protein levels exposed to various genotoxic agents are due mainly to an increase in p53 protein stability rather than an increase in steady-state p53 mRNA levels (9,18,19). However, it has been suggested that an increase in p53 protein stability is not solely responsible for p53 recruitment in response to genotoxic stress, and it is more likely that p53 genotoxic stress response is a complex cellular process regulated at the transcriptional, mRNA stability level as well as the protein stability level (22,23).
It has recently been reported in a few studies that PAH and its metabolites result in a rapid accumulation of the p53 gene product in human and mouse cells (24,25). However, whether or not the increase of p53 protein resulting from exposure to carcinogens is also regulated at a transcriptional level is still unclear. In the present study, we found that the B[a]P induced-p53 protein accumulation also involved p53 promoter activation. Furthermore, we characterized what p53 promoter sequences are involved in B[a]P-induced p53 promoter activation and what cellular factors are involved. The results presented herein suggest the transcriptional activation of human p53 promoter by B[a]P-induced NF-B complexes.

EXPERIMENTAL PROCEDURES
Carcinogen and Antibodies-B[a]P, obtained from Sigma-Aldrich, was dissolved in dimethyl sulfoxide immediately prior to use and then was added directly to the cell culture medium as 1000ϫ stocks to obtain the desired final concentration. Polyclonal antibodies against IB␣ and the subunits of NF-B (p50, p65, p52, and c-Rel) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Monoclonal anti-p53 antibodies (Ab-6 and Ab-1) were obtained from Oncogene Science (Cambridge, MA).
Plasmids-The recombinant chloramphenicol acetyltransferase (CAT) vector, p53CATd325, which contains the human p53 gene promoter region Ϫ325 to ϩ12 (the 3Ј end of the fragment is an XbaI site and is about 12 bp downstream of the major transcription initiation site, the most 3Ј end of which is defined as ϩ1 in the present study) (22,26) was constructed using PCR. Human genomic DNA, derived from normal human lymphocytes, was used as a template, and PCR was performed with the following primers: sense, 5Ј-GTTAGGGTGTG-GATATTACG-3Ј; antisense, 5Ј-CAATCCAGGGAAGCGTGTCA-3Ј. The PCR product was digested with HindIII and XbaI and then ligated into pCAT Basic vector (Promega, Madison, WI), which contains the CAT reporter gene but lacks a eukaryotic promoter region. The orientation and sequence of the inserted fragment were confirmed by sequencing.
With p53CATd325 as a template, p53 promoter fragments from Ϫ135 to ϩ12 (d135), from Ϫ80 to ϩ12 (d80), and from Ϫ50 to ϩ12 (d50) with 5Ј HindIII and 3Ј XbaI ends were prepared by PCR and then were cloned into the pCAT Basic vector digested with HindIII and XbaI. Various linker-scanning mutations (mt 1 to mt 3) shown in Fig. 2 were introduced to p53CATd325 by an improved site-directed mutagenesis method using PCR (27).
Cell Culture, DNA Transfection, and Exposure to B[a]P-A human lung adenocarcinoma cell line, A549 (distal respiratory epitheliumlike), was maintained in RPMI 1640 medium containing 10% fetal bovine serum (CC Laboratories, Cleveland, OH). A mouse fibroblast cell line, NIH 3T3 was grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. For DNA transfection, cells were plated at a density of 5ϫ10 5 cells/60-mm diameter culture dish 24 h before transfection. The cells were transfected with 5 g of various CAT reporter plasmids and 1 g of a pSV-␤-gal control vector by LipofectAMINE reagent (Life Technologies Inc.) according to the manufacturer's instructions. At 20 h after transfection, the cells were treated for 24 h with B[a]P or cisplatin (CDDP) in a serum-free medium and then were harvested to determine their CAT activity and ␤-gal activity.
Reporter Plasmid Assays-After the transfection and exposure to B[a]P or CDDP, the cells were washed twice in PBS, harvested in 40 mM Tris-HCl, pH 7.5, 1 mM EDTA, 150 mM NaCl, resuspended in 100 l of 250 mM Tris-HCl, pH 7.8, and lysed by three freeze-thawing cycles. The preparation was microcentrifuged, and the supernatant (cell extract) was removed. To determine the transfection efficiencies, a portion of the cell extract was used to check the ␤-gal activity by a colorimetric assay using chlorophenol red-3-D-galactopyranoside (Roche Molecular Biochemicals) as described previously (28).
The CAT activity was assayed as previously reported (29). The cell extracts, normalized for ␤-gal activity, were heated to 65°C for 10 min and then were incubated with 1-deoxy[dichloroacetyl-a-14 C]chloramphenicol (Amersham Pharmacia Biotech) and acetyl CoA for 2 h and assayed for CAT activity by thin layer chromatography (Sigma-Aldrich). The CAT activity was visualized and quantified by a Fuji BAS 2000 phosphorimager (Fuji Photo Film, Odawara, Japan).
Isolation of Cytoplasmic and Nuclear Extracts-For the isolation of nuclear extracts, all procedures were performed on ice. Nearly confluent monolayers of A549 and NIH 3T3 cells, which had been treated with B[a]P or control medium for the appropriate time, were washed with phosphate-buffered saline (PBS), harvested by scraping into 1 ml of PBS, and pelleted at 5,000 rpm for 5 min. The pellet was washed twice with PBS, and then suspended in 0.4 ml of lysis buffer (10 mM HEPES, pH 7.9, 10 mM KCl, 0.1 mM EDTA, 1.5 mM MgCl 2 , 0.25% Nonidet P-40, 1 mM DTT, and 0.1 mM phenylmethylsulfonyl fluoride). The pellet was vortexed and centrifuged at 5,000 rpm for 2 min. The supernatant represented the cytoplasmic extract. The nuclear pellet was resuspended in 60 l of extract buffer (20 mM HEPES, pH 7.9, 420 mM NaCl, 0.1 mM EDTA, 1.5 mM MgCl 2 , 25% glycerol, 1 mM DTT, and 0.5 mM phenylmethylsulfonyl fluoride), and the nuclei were rocked for 30 min at 4°C. After centrifugation, the supernatant containing nuclear extract was removed and the protein concentration was determined. The nuclear extracts were then stored at Ϫ70°C until further use.
Electrophoretic Mobility-shift Assay-Electrophoretic mobility-shift assays were performed by incubating 10 g of nuclear extract in 10 l of binding buffer (10 mM Tris-HCl, pH 7.5, 50 mM NaCl, 0.5 mM DTT, 0.5 mM EDTA, 1 mM MgCl 2 , 4% glycerol, 0.25 mg/ml poly(dI-dC)) for 10 min at 4°C. Then, 35 fmol of 32 P-labeled oligonucleotide probe was added and the reaction mixture was incubated for 30 min at room temperature. For reactions involving competitor oligonucleotides, the unlabeled competitor and the labeled probes were premixed before addition to the reaction mixture. For supershift assays, the reaction mixture minus the probe was incubated with 2 l of specific antibodies for 1 h at 4°C. The 32 P-labeled oligonucleotide was then added and incubated for 30 min at room temperature. The samples were electrophoresed at 4°C for 2 h on a 4% nondenaturing polyacrylamide gel (30:1) in a Tris borate-EDTA buffer system. The dried gels were exposed to Kodak X-Omat film at Ϫ80°C.
Western Blot Analysis-Equal amounts of protein were denatured by heating to 95°C in SDS reducing buffer and were separated by electrophoresis on 10% SDS-polyacrylamide gels, and the proteins were electrophoretically transferred to Hybond-ECL nitrocellulose membranes (Amersham Pharmacia Biotech). The membranes were probed with polyclonal anti-IB␣ antibody or monoclonal anti-p53 antibody. Bound antibodies were detected using a ECL Western blotting analysis system (Amersham Pharmacia Biotech).
Northern Blot Analysis-The cells were grown for 24 h, then treated for 4, 8, or 24 h with B[a]P (10 M). At the end of the treatments, the total RNA was extracted by using Isogen (Nippon Gene, Toyama, Japan) according to the manufacturer's instructions. The total RNAs (20 g) from each treatment were denatured with formaldehyde-formamide and separated in a 1% agarose-formaldehyde gel. RNAs were then transferred to the Hybond-N nylon transfer membrane (Amersham Pharmacia Biotech). The membrane was hybridized with the full-length human p53 cDNA or a 0.83-kilobase PstI fragment of mouse p53 cDNA (30). The probe was labeled with [␣ 32 P]dCTP by random priming using the Megaprime DNA labeling system (Amersham Pharmacia Biotech), and purified through NICK spin columns (Amersham Pharmacia Biotech). A ␤-actin cDNA probe was later used to rehybridize the same membrane. The levels of p53 mRNA were quantified with the Fuji BAS 2000 phosphorimager and normalized to ␤-actin values.
Adenovirus Infection-Adenovirus infection was carried out essentially as described previously (28). Briefly, confluent A549 cells rinsed with serum-free medium twice were added to a recombinant adenovirus expressing IB␣ (a kind gift from Dr. Hikaru Ueno), which was diluted to a desired titer with serum-free medium. After incubation for 2 h at room temperature under gentle agitation, the medium containing an adenovirus was removed. The cells were washed and incubated in fresh growth medium for 3 days until assayed.

Enhanced p53 Expression in A549 and NIH 3T3 Cells Treated by B[a]P-Although the results of previous studies
showed that the cellular p53 protein levels increase upon exposure to B[a]P and its metabolites, and suggested that this increase is due mainly to an increase in p53 protein stability (24,31,32), it has not been studied as to whether transcriptional activation is also involved in p53 recruitment in response to carcinogen treatment. To investigate the mechanisms leading to p53 protein accumulation in human lung adenocarcinoma cells after carcinogen treatment, the A549 cells that retain a wild-type p53 gene were exposed to B[a]P for increasing periods of time under serum-free conditions. Western blotting showed an increased p53 expression following stimulation of A549 cells with B[a]P (Fig. 1A). 2.1-or 5.1-fold increase in p53 protein levels was detected after 4 or 8 h of treatment; they remained elevated even after 24 h (Fig. 1A). NIH 3T3 cells exposed to B[a]P also showed similar results in a Western blotting analysis (Fig. 1C). A Northern blotting analysis revealed an increased accumulation of p53 mRNA (2.5-fold) in B[a]P-treated A549 cells at 8 and 24 h (Fig. 1B), which was most likely responsible, in part, for the increased p53 protein levels. The discrepancy between p53 protein and mRNA levels after the treatment for 4 h may be the result of an increased stability of the protein, consistent with the previous findings (23,24). CDDP, an anti-cancer agent that was reported to result in an increase in p53 mRNA (22), was used as a positive control. A similar stimulatory effect with B[a]P on p53 mRNA increase was observed in the A549 cells treated with the CDDP (data not shown). The p53 mRNA levels in NIH 3T3 cells also increased in response to B[a]P treatment (Fig. 1D), although the extent was less than in A549 cells.
Activation of p53 Promoter by B[a]P-The above described results prompted us to examine whether B[a]P activates the p53 gene promoter. A previous study identified that an 85-bp fragment spanning the p53 promoter region, the 3Ј end of which is about 10 bp downstream of the major transcriptional initiation sites, is all that is required for full promoter activity (26), and this region is also required for genotoxic stress-responsive activities (22). However, we used the p53 promoter-CAT fusion plasmid containing a 325-bp p53 promoter region, the 3Ј end of which is also about 10 bp downstream of the major transcriptional initiation sites as shown in Fig. 2. Previous studies have demonstrated that the activity of this 325-bp region is nearly the same as that of 2.4-kilobase promoter (22,26), and this region includes all transcription factor binding motifs discovered upstream of the major transcriptional initiation sites (for a review, see Ref. 33). The p53CATd325 reporter plasmid was transfected into A549 and NIH 3T3 cells, and CAT activity was measured after exposure to B[a]P or CDDP. As shown in Fig. 3, B[a]P and CDDP were able to induce p53 promoter activation. 30 M B[a]P was able to induce the p53 promoter-driven CAT gene expression 3-4-fold above the control level in A549 and NIH 3T3 cells.

Identification of a Novel p53 Promoter Element Involved in Basal and B[a]P-induced Promoter
Activation-Furthermore, the p53 promoter-CAT reporter genes that contain 135-and 80-bp p53 promoter region (pCATd135 and pCATd80; Fig. 2) were transfected into A549 cells and showed 2.0-and 2.3-fold increases of CAT activity, in the basal level, respectively, and an enhanced response to B[a]P. However, further deletion to Ϫ50 (pCATd50) resulted in a 3.2-fold reduction in the basal level and complete unresponsiveness to B[a]P treatment (Fig.  4). Essentially similar results were also obtained with NIH 3T3 cells (data not shown). These results indicate that the ciselements required for basal and B[a]P stress-responsive activities reside within this 80-bp region of the p53 promoter and further suggest that the p53 promoter region from Ϫ325 to Ϫ135 contains a region that represses the p53 proximal promoter activity.
As shown in Fig. 2, the 80-bp p53 core promoter region contains potential binding motifs for NF-B (34) and basic helix-loop-helix transcription factors (35). We introduced three linker-scanning mutations in this p53 core promoter region (Fig. 2), and p53 promoter-CAT reporters containing these mutated 325-bp p53 promoter regions were examined for their basal promoter activities and responsiveness to B[a]P treatment. The results shown in Fig. 4 indicate that the mutations (mt 1 and mt 2) in the p53 promoter region from Ϫ58 to Ϫ43, overlapping the B motif, result in 4.4-and 3.2-fold reduction in the basal promoter activity, respectively, and severely impair the p53 promoter activation induced by B[a]P. A mutation (mt 3) affecting a potential helix-loop-helix transcription factor binding motif did not have any significant deteriorative effect on the basal as well as on p53 B[a]P stress-responsive promoter activity. These results thus indicate that the p53 promoter region extending from Ϫ58 to Ϫ43, overlapping the B motif, is essential for both the p53 basal promoter activity and p53 promoter activation induced by B[a]P.
B[a]P Activates NF-B in A549 and NIH 3T3 Cells-The results from a deletion analysis of p53 gene promoter region in the present and previous study (34) prompted us to determine whether NF-B proteins are involved in B[a]P-induced p53 expression. We first examined the induction of NF-B complexes in B[a]P-treated A549 and NIH 3T3 cells. Gel shift assays performed with nuclear extracts from A549 cells showed that stimulation with B[a]P at 10 M induced a strong NF-B DNA binding activity within 4 h and maintained until 24 h after B[a]P treatment. This induction was observed when the nuclear extracts were incubated with a canonical NF-B consensus probe (data not shown) or with a NF-B p53 probe derived from Ϫ39 to Ϫ60 of the p53 promoter (wt NF-B p53 probe) (Fig. 5A). A similar stimulating effect of B[a]P on NF-B DNA binding activity was also observed in NIH 3T3 cells (Fig. 5B).
Competition studies were performed with a 50-fold excess of the unlabeled wt NF-B p53 probe, an oligonucleotide probe with mutations in the p53 NF-B site (mt NF-B p53 probe) and the canonical NF-B consensus probe. As shown in Fig. 6, the protein binds NF-B consensus site specifically, as the binding activity was fully inhibited by the unlabeled wt NF-B p53 probe and by the canonical NF-B consensus probe but not by the mt NF-B p53 probe. A further characterization of the nuclear proteins binding to the NF-B motif was obtained by preincubating the nuclear extracts with antibodies against the p50, p65, p52, and c-Rel NF-kB subunits. Both p50 and p65 antibodies partially shifted the B[a]P-induced complex in A549 cells (Fig. 6A), whereas p52 and c-Rel antibodies did not. These data indicate that B[a]P activates NF-kB p50 and p65 proteins in A549 cells. Analysis of specific subunit components in the complex revealed the presence of p65 protein in NIH 3T3 cells (Fig. 6B).
Overexpression of IB␣ Abrogates B[a]P-induced Transcription of p53-A recombinant adenovirus expressing IB␣ (rAd.IB␣) was used to infect A549 cells. After 3 days of infection with the recombinant adenovirus, the A549 cells expressed high levels of IB␣ in a m.o.i.-dependent manner, based on the Western blotting findings (Fig. 7A). The amounts expressed FIG. 2. The structure of 5 deletion and linker-scanning mutants of p53 promoter-CAT reporter genes. The human p53 promoter sequence from Ϫ30 to Ϫ60 and potential transcription factor binding motifs, B (35) and helix-loop-helix (36), were shown. The most 3Ј end of major transcription initiation sites for the human p53 gene as determined by Tuck and Crawford (28) is tentatively defined as ϩ1 in the present study. Underlined sequences in p53CATmt 1, p53CATmt 2, and p53CATmt 3 represent the mutated positions. were estimated to be about 60 -70 times higher as compared with the endogenous protein. The stimulation of the infected cells at day 3 with B[a]P for 6 h did not detectably decrease the IB␣ levels. In transient transfection assays, B[a]P stimulation in A549 cells infected with the rAd.IB␣ did not induce any significant p53 promoter activation as compared with the level of activation in the control cells (Fig. 7B; compare with Fig. 3). An electrophoretic mobility-shift assay analysis demonstrated that the overexpression of the IB␣ led to a complete suppression of B[a]P-induced NF-B activation (Fig. 7C). No stimulatory effect of B[a]P on p53 mRNA levels was observed in the A549 cells infected with the rAd.IB␣ (data not shown). DISCUSSION p53 recruitment in response to various genotoxic stresses is an important cellular response to maintain the integrity of the genome (14). After metabolic activation, carcinogens, such as PAHs, bind to DNA and form predominantly covalent carcinogen-DNA adduct, and thus express their carcinogenic and genotoxic activity (3,4,6). Previous studies have focused on the modulation of PAH mediated p53 protein accumulation posttranscriptionally. In this paper, we presented evidence that PAH stimulated p53 accumulation may also be induced transcriptionally. In our experiments, the p53 mRNA levels increased 1.6 -2.5-fold after the cells were treated with B[a]P, while the degree of p53 protein induction was more than 1.6 -2.5-fold, thus indicating that p53 is regulated at the protein level as well as the mRNA level. B[a]P exposure to the cells activated the p53 gene promoter through the binding of NF-B proteins to the B site. Activated NF-B complexes were shown to contain predominantly p50 and p65 subunit components in A549 cells and p65 subunit in NIH 3T3 cells. In addition, the overexpression of IB␣ completely inhibited NF-B activation, The effect of PAHs, a ubiquitous environmental pollutant, on p53 expression was recently investigated (23,24,31,32). p53 protein expression was first reported to be closely correlated with B[a]P-DNA adducts in carcinoma cell lines by Ramet et al. (31). They found the activation of B[a]P in both A549 lung cancer and MCF-7 breast adenocarcinoma cell lines containing wild-type p53 and formation of BPDE-DNA adduct were followed by an increase in p53 protein expression. The possible reasons for the increase in p53 expression are, like other DNA-damaging agents, the stabilization of p53 protein in the cells (24,31,32,36) and the stimulation of p53 transcription (22,23).
Since the p53 genotoxic stress response is a complex cellular process and may be regulated at the transcriptional level (22,23), PAH-induced p53 expression may also be regulated in the same manner. Indeed, our results revealed that B[a]P stimulated the p53 expression transcriptionally through the p53 promoter activation, thus suggesting the existence of another pathway regulating p53 expression. Our results are supported by the conclusion of Sun et al. and Hellin et al. (22,23), although both groups used anti-cancer drugs as genotoxic agents.
The previous study by Tuck and Crawford identified the human p53 promoter region (from Ϫ78 to ϩ10) required for full basal p53 promoter activity (26). The results presented in our study indicate that this p53 promoter region is sufficient for p53 B[a]P stress-responsive promoter activity. Similar results have been reported by Sun et al. (22), in which they also found this region to be sufficient for p53 genotoxic stress-responsive promoter activity. This region contains potential binding motifs for several important transcription factors, including NF-B and helix-loop-helix. The present study identified that a novel p53 promoter element involved in basal and B[a]P-induced promoter activation resides in the region from Ϫ58 to Ϫ43, overlapping the kB motif. B[a]P induced the p53 promoter through the binding of NF-B proteins to the B site.
NF-B is a rapidly inducible transcription factor involved in the response of various stimuli, including cytokines, activators of protein kinase C, viral infections, and oxidants. Following its intracellular activation, NF-B regulates the expression of many genes that code for cytokines, growth factors, acute phase response proteins, and cellular receptors and thus modulates the various cellular responses to the stimuli. NF-B complexes bind DNA as dimers constituted from a family of proteins, Rel/NF-B family. The family contains the proteins p50 (NF-B1), p52 (NF-B2), p65 (RelA), RelB, and c-Rel (Rel) (37). The Rel/NF-B family of proteins share an amino-terminal ϳ300amino acid domain (Rel homology domain), including DNA binding and dimerization domains and the nuclear translocation signal, which is most likely the binding site for IB. In unstimulated cells, NF-B is found in the cytoplasm and is bound to IB (IB␣ and IB␤), which thus prevents it from translocating to the nuclei. When these cells are stimulated, specific kinases phosphorylate IB, causing its rapid degradation by proteasomes. The release of NF-B from IB results in the translocation of NF-B into the nucleus, where it binds to specific sequences in the promoter regions of target genes.
Previous studies have identified the p53 responsive element that overlaps with an NF-B binding site. This B site is responsible for the induction of p53 transcription by transfection with p65, or treatment with either tumor necrosis factor-␣ or daunomycin, an anti-cancer agent (23,34,38). We herein demonstrated that, in human lung epithelial cells and mouse fibroblast cells, human p53 promoter was activated by NF-B after exposure to B[a]P, thus supporting the above conclusion.
Nevertheless, these results are in conflict with those reported by Sun et al. (22), who showed that NF-B, while inducible by various anticancer drugs, does not play a major role in human p53 promoter activation induced by genotoxic agents. This discrepancy could result from using different cell types that may cause a lack of essential cellular factors (11,22,23). Indeed, it has previously been reported that daunomycin failed to trigger p53 gene expression in myeloid leukemia cells, indicating that anthracycline drugs might have distinct p53 induction activities in different cell types (11).
It is established that the dimer composition of the NF-B complex determines its fine DNA binding specificity, giving rise to selective transcriptional activation or attenuation. Transcriptional activation of specific sets of genes will primarily depend on various dimer combinations being activated distinctly, or whether their relative amounts in cell types and tissues are subject to regulation (37,39). Analysis of subunits present in the activated complexes from B[a]P-treated A549 cells indicated the presence of both p65 and p50 components, which together constitute the predominant transcriptionally active NF-B dimer combination. We failed to find the presence of other subunits except p65 in the activated complexes from B[a]P-treated NIH3T3 cells. There are several possible explanations for the failure of identification of p50 subunit in NIH 3T3 cells. First, it is reported that p50 is barely detectable in quiescent NIH 3T3 cells that are derived by incubation of cells in serum-free medium (40). The p50 subunit in NIH 3T3 cells binding to NF-B sequence may be too small to be detected by supershift assay in the present study. Second, binding ability of Rel/NF-B dimers can be modulated by the B sites, nearby DNA sequences, other DNA-binding proteins, and other unclear factors (37). Even using same cells' nuclear extract and identical canonical NF-B consensus probe discrepant results were also observed (41)(42)(43)(44). Third, another reason might be dependent on the cell line tested. There is a report suggesting at least two NIH 3T3 cell lines of different origins circulating in the world (45). Indeed, some researchers identified p50 in NIH 3T3 cells binding to NF-B sequence (46), while others did not (47). Nevertheless, since p50 lacks a potent transcriptional activation domain, it does not generally activate transcription as homodimers in vivo (37). Transfection of p50 subunit did not activate transcription of p53 promoter (34). p65 subunit contains a more potent transactivation domain than p50 subunit, and expression of p65 alone, or of p65, in combination with other subunits, result in more significant increase in NF-Bdependent transcription than other subunits (37,39). Therefore, the p65 subunit in the activated complexes is responsible mainly for transcriptional activity of NF-B.
NF-B activation in response to various stimuli is generally considered to occur rapidly within minutes (37). In the present study we found that the NF-B activation by B[a]P was induced in 2-4 h and then was sustained for up to 24 h. The time course of NF-B activation by B[a]P is concordant with the results when NF-B was activated by daunorubicin and UV irradiation (38,48). Bender et al. (48) recently found two totally different sequentially occurring mechanisms of UV-induced NF-B activation and IB degradation. They reported that the early IB␣ degradation at 30 min to 6 h is not initiated by UV-induced DNA damage. It does not involve the phosphorylation of IB by IB kinase. On the other hand, IB degradation and NF-B activation at late time points, 15-20 h after UV irradiation, requires the phosphorylatable IB␣. This late mechanism thus clearly supports our results.
When IB␣ was overexpressed in A549 cells, the B[a]Pinduced NF-B activation was suppressed and thus inhibited the p53 promoter activation in transient transfection assays and the stimulatory effect on p53 mRNA levels. These data provide direct evidence that the functional inhibition of NF-B alters the B[a]P-induced p53 promoter activation and support the view that B[a]P-mediated p53 transcription is NF-B-dependent.
In summary, the present study revealed that a potent lung carcinogen, B[a]P, increased the p53 expression at both the protein and mRNA levels. It activated the p53 promoter in two different cell lines, thus suggesting that the induced p53 was, at least partly, regulated transcriptionally. We identified that the p53 promoter region extending from Ϫ58 to Ϫ43, overlapping the B motif, is essentially required for both p53 basal promoter activity and p53 promoter activation induced by B[a]P. Furthermore, NF-B proteins are also identified to be involved in the B[a]P-induced p53 expression.