Interaction of human NAD(P)H:quinone oxidoreductase 1 (NQO1) with the tumor suppressor protein p53 in cells and cell-free systems.

NAD(P)H:quinone oxidoreductase 1 (NQO1) has been proposed to stabilize p53 via a redox mechanism involving oxidation of NAD(P)H as a consequence of the catalytic activity of NQO1. We report that treatment of HCT-116 human colon carcinoma cells with the NQO1 inhibitor ES936 had no effect on the levels of p53 protein. ES936 is a mechanism-based inhibitor of NQO1 that irreversibly blocks the catalytic function of the enzyme. This suggests that a redox mechanism involving NQO1-mediated NAD(P)H oxidation is not responsible for the stabilization of p53. We also examined the ability of the NQO1 protein to associate with p53 using co-immunoprecipitation experiments. Results from these experiments demonstrated co-immunoprecipitation of NQO1 with p53 and vice versa. The association between p53 and NQO1 was not affected by treatment of HCT-116 cells with ES936, demonstrating that the association was not dependent on the catalytic activity of NQO1. A comparison of isogenic HCT-116 p53+/+ and HCT-116 p53-/- cells demonstrated an interaction of NQO1 and p53 only in the p53+/+ cells. Experiments performed in an in vitro transcription/translation system utilizing rabbit reticulocyte lysates confirmed the interaction of NQO1 and p53. In these experiments a full-length p53 coding region was used to express p53 in the presence of recombinant NQO1 protein. An association of p53 and NQO1 was also observed in primary human keratinocytes and mammary epithelial cells. In studies where mdm-2 co-immunoprecipitated with p53, no association of mdm-2 with NQO1 was observed. These data demonstrate an association between p53 and NQO1 that may represent an alternate mechanism of p53 stabilization by NQO1 in a wide variety of human cell types.

The p53 gene is one of the major tumor suppressor genes in humans (1,2), and p53 mutations are one of the most common genetic events that occur in human cancers (3). When normal cells are subjected to stress signals, such as DNA damage or oxidative stress, p53 is activated, resulting in transcription of downstream genes that coordinate either growth arrest of the cell or apoptosis (4) preventing proliferation and clonal expansion of damaged cells. Mutations or deletions in p53 and a consequent loss of p53-dependent function leads to increased susceptibility to neoplasia (1). The identification of small molecules and proteins that increase p53 stability and, as a result, protect cells against cancer progression is an active area of current research (3). NAD(P)H:quinone oxidoreductase 1 (NQO1, DT-diaphorase, EC 1.6.99.2) 1 is a cytosolic flavoenzyme that catalyzes the two-electron reduction of a broad range of substrates (5). NQO1 is an obligate two-electron reductase and is classified as a detoxification enzyme primarily because of its ability to reduce quinone substrates directly to their less toxic hydroquinone derivatives bypassing the redox-cycling semiquinone radical (6,7). NQO1 can also function as an antioxidant enzyme reducing ubiquinone and vitamin E quinone to their antioxidant forms (8,9). We have characterized a polymorphism in NQO1 (NQO1*2) (10,11) and demonstrated that individuals homozygous for the NQO1*2 polymorphism (NQO1*2/*2) have no measurable NQO1 activity (12). The null phenotype of individuals carrying the homozygous NQO1*2 polymorphism is due to rapid proteasomal degradation of the mutant NQO1*2 protein (13). A lack of NQO1 protein due to the homozygous NQO1*2 polymorphism has been associated with an increased risk of various cancers including renal, urothelial, and cutaneous basal cell carcinomas (14,15). Of particular interest are five separate epidemiological studies that have identified the NQO1*2 polymorphism as a significant risk factor for development of leukemias of diverse origin (16 -20). NQO1-knock-out mice have been shown to be more susceptible to chemicalinduced skin cancer (21,22), and inducers of NQO1 have long been recognized as chemoprotective agents against a range of animal tumors (23,24).
Recently Asher et al. (25) proposed that NQO1, which is activated by many of the same stresses that activate p53, stabilized wild type p53 protein (25). Experiments were performed by examining p53 stability and function in cells transfected with human NQO1. Transfected wild type NQO1*1 protein, but not the mutant NQO1*2 protein, stabilized wild type p53 in HCT-116 cells (26). Experiments were also performed in the presence and absence of the competitive and relatively nonspecific NQO1 inhibitor dicumarol (27), which was found to increase proteasomal degradation of p53 and modulate p53-dependent apoptosis (25). Very recent work by the same group has demonstrated that transfected wild type NQO1, but not the mutant NQO1*2 protein, stabilizes wild type p53 in HCT-116 cells (26). A redox mechanism was proposed for p53 stabilization that relied upon NQO1-dependent NAD(P)H oxidation (25), although these authors have commented that physical interaction of p53 and NQO1 represents an attractive possibility as a mechanism of stabilization (26).
NQO1 is known to bind other proteins such as Hsp70 and Hsp40 (28). Consequently, we examined whether NQO1 could interact with p53 via a protein-protein interaction. In this work, we demonstrate that a novel suicide inhibitor of NQO1 developed in our laboratory, which irreversibly blocks the catalytic function of NQO1, has no effect on p53 stability. This suggests that a redox mechanism of stabilization is unlikely. However, for the first time, we demonstrate that NQO1 is able to physically associate with p53 suggesting that a proteinprotein interaction may be responsible for the stabilization of p53 by NQO1. In addition to the roles of NQO1 in direct detoxification of quinones and in antioxidant defense, the interaction of NQO1 with p53 may represent an additional mechanism that contributes to the chemoprotective activity of NQO1.
Cell Lines-The human colorectal carcinoma cell line HCT-116 was obtained from American Tissue Culture Collection (Manassas, VA). The cell lines HCT-116 p53ϩ/ϩ and HCT-116 p53Ϫ/Ϫ were obtained from Dr. B. Vogelstein, Johns Hopkins University (31). All HCT-116 cell lines were genotyped as homozygous wild type for the NQO1*1 allele (11). Cells were grown as monolayers at 37°C in 5% CO 2 in minimal essential medium supplemented with 10% (v/v) fetal bovine serum, 10 units/ml penicillin/streptomycin, and 2 mM L-glutamine. Second passage primary human epidermal keratinocytes (HEK) were obtained from the Vanderbilt Skin Disease Research Core. HEKs were isolated as described previously (32) and were cultured in EpiLife M-EPI-500 keratinocyte growth media (Cascade Biologics, Portland, OR) supplemented with human keratinocyte growth supplement S-001-5 (Cascade Biologics) and 0.06 mM CaCl 2 . Primary cultures of human mammary epithelial cells (HMECs) were obtained from Clonetics (San Diego, CA) and passaged according to manufacturer's instructions.
Co-immunoprecipitation-To examine the physical interaction between NQO1 and p53, we utilized three different co-immunoprecipitation methodologies: 1) immunoprecipitation with anti-p53 antibodies followed by NQO1 immunoblot analysis, 2) immunoprecipitation with anti-NQO1 antibodies (A180) cross-linked to protein A-Sepharose beads followed by p53 immunoblot analysis, and 3) immunoprecipitation with anti-NQO1 antibodies (A180) followed by immunoblot analysis with anti-p53 antibodies directly coupled to horseradish peroxidase.
Co-immunoprecipitation Studies in HCT-116 Cells-For co-immunoprecipitation experiments, an equal number of exponentially growing cells (5 ϫ 10 6 ) were harvested by trypsinization, washed in phosphatebuffered saline, and then lysed in freshly prepared RIPA buffer with protease inhibitors. Cells were lysed in RIPA buffer on ice for 15 min followed by centrifugation at 13,000 rpm for 5 min. Equal aliquots of supernatant (6 mg) were added to tubes containing anti-Bcl-2 (isotyped matched control, 2 g), anti-p53 (2 g), or anti-NQO1 (200 l) antibodies overnight at 4°C while gently agitating. Protein A/G-agarose beads (40 l) were then added, and the incubations were continued for an additional 90 min at 4°C. Protein A/G-agarose beads were collected by centrifugation and then washed four times with 50 mM Tris-HCl, pH 8.0, containing 150 mM NaCl, 1% (v/v) Nonidet P-40 followed by a single wash (1 ml) with 50 mM Tris-HCl, pH 8.0. The protein A/G-agarose beads were then suspended in 2ϫ Laemmli SDS sample buffer and heated to 75°C for 5 min in preparation for SDS-PAGE and subsequent immunoblot analysis.
Co-immunoprecipitation Studies in Human Primary Cell Cultures-Immunoblot analysis of proteins isolated from primary epithelial cell cultures were performed as described previously (32) using the anti-p53 monoclonal antibody DO-1 and the anti-NQO1 monoclonal antibody A180. For the experiments described in Fig. 3, an alternative methodology was employed where the anti-NQO1 antibody (A180) was crosslinked to protein A-Sepharose (PAS) to prevent interference with IgG heavy chain. For cross-linking, the PAS and antibodies were incubated overnight while mixing at 4°C. The antibody-bound PAS was washed twice with 500 mM sodium borate, pH 9.0 (NaB), resuspended in NaB containing 100 mM dimethyl pimelimidate (Pierce), and the pH readjusted to 8.2. The antibody-bound PAS was mixed with dimethyl pimelimidate for 2 h at 25°C, washed twice with 200 mM ethanolamine, pH 8.0, and then resuspended in ethanolamine and mixed for 2 h at 25°C. The cross-linked PAS was washed twice with phosphate-buffered saline and the incubation with dimethyl pimelimidate and subsequent steps repeated to achieve complete cross-linking of the antibodies to the PAS. After the final phosphate-buffered saline wash, the cross-linked, antibody-bound PAS was resuspended in phosphate-buffered saline, and chemical cross-linking was verified by immunoblot analysis prior to use in immunoprecipitation studies with primary cultured cell lines. To assure complete chemical cross-linking, DO-1-PAS was was processed by SDS-PAGE, and the presence of uncross-linked antibody was assessed by immunoblot analysis using a goat anti-mouse horseradish peroxidase-conjugated antibody (Pierce). In all DO1-PAS preparations used, free antibody was not detected by immunoblot analysis.
Co-immunoprecipitation Studies Using in Vitro Transcription/Translation Reactions-p53 in vitro transcription/translation was carried out using SP6 quick coupled transcription/translation system (TNT-RRL, Promega, Madison, WI) according to the manufacturer's instructions using 1 g of plasmid p53SP64 poly(A). Reactions (50 l) were carried out for 90 min at 32°C, after which a small aliquot of reaction mixture was removed, and p53 protein translation was monitored by immunoblot analysis. Reactions were terminated by the addition of 200 l of RIPA buffer containing protease inhibitors, and immunoprecipitation was carried out using anti-p53 monoclonal antibody PAb421 followed by NQO1 immunoblot analysis. The reverse immunoprecipitation was also performed using anti-NQO1 antibodies (A180/B771) followed by immunoblot analysis using anti-p53 antibodies (DO-1) directly coupled to horseradish peroxidase (see SDS-PAGE). The human wild type p53 coding region in pBSK-SN2 was a generous gift from Dr. Bert Vogelstein, Johns Hopkins University, Baltimore, MD and subcloned into the pSP64 poly(A) expression vector (Promega) using polymerase chain reaction (PCR) amplification of the full coding region. The following oligomers 5Ј-cccaagcttATGGAGGAGCCGCAGTCAGATCC-3Ј and 5Ј-tgctctagaAGATCAGTCTGAGTCAGGCCCTTCTG-3Ј containing HindIII and XbaI restriction sites, respectively, were used to amplify the p53 coding region. The temperature cycles used for this PCR amplification were as follows: 94°C for 5 min, 25 cycles of 94°C for 1 min, 63°C for 1 min, followed by 72°C for 1.5 min. After the final cycle the reaction was kept at 4°C. The p53SP64 construct was verified by complete sequencing of the coding region.
SDS-PAGE and Immunoblot Analysis-Proteins were separated by 12% SDS-PAGE and transferred to polyvinylidene difluoride membranes in 25 mM Tris, 192 mM glycine containing 20% (v/v) methanol at 110 volts for 1 h. Following transfer, membranes were placed overnight in blocking buffer (see above). Immunoblot analysis of co-immunoprecipitated proteins was performed as follows: 1) for NQO1 detection following p53 immunoprecipitation, membranes were probed first with 20 ml of blocking buffer containing 500 l of anti-NQO1 monoclonal antibodies (A180/B771) for 1 h at 27°C followed by 20 ml of blocking buffer containing HRP-conjugated secondary antibody (1:7,500) for 30 min at 27°C. 2) For p53 detection following NQO1 immunoprecipitation samples were prepared in 2-mercaptoethanol-free 2ϫ Laemmli buffer and heated to 75°C for 5 min. Membranes were probed with 20 ml of blocking buffer containing HRP-conjugated DO-1 diluted 1:2,000 for 45 min at 27°C. Protein bands were visualized using luminol-based enhanced chemiluminescence as described by the manufacturer (PerkinElmer Life Sciences).

RESULTS
p53 Protein Levels Do Not Change in HCT-116 Cells in Response to Inhibition of NQO1-We have recently developed ES936 as a suicide inhibitor of NQO1 (29). ES936 inhibits NQO1 at low concentrations in an irreversible manner and has advantages over dicumarol, since the latter is a competitive rather than an irreversible inhibitor of NQO1 and is known to inhibit many other enzymes (27,35). Treatment of HCT-116 cells with ES936 had little effect on the stability of p53 (Fig.  1A), suggesting that inhibition of the catalytic function of NQO1 does not affect p53 stability. Inhibition of greater than 99% of NQO1 catalytic activity in HCT-116 cells by ES936 (100 nM) was verified by NQO1 activity assays (Fig. 1B). Additional control experiments demonstrated that treatment of HCT-116 cells with ES936 (100, 250, or 500 nM) had little effect on NQO1 protein levels (Fig. 1C).
Co-immunoprecipitation of NQO1 and p53-To examine a potential interaction between NQO1 and p53 proteins in HCT-116 cells, we utilized a co-immunoprecipitation approach. Experiments were performed using p53 antibodies for immunoprecipitation and subsequent immunoblotting with antibodies to NQO1 ( Fig. 2A). In Fig. 2B, we demonstrate in HCT-116 cells that pretreatment with ES936, a suicide inhibitor of NQO1, had little effect on the association of p53 and NQO1, suggesting that the interaction of the two proteins is not dependent on the catalytic activity of NQO1. An isogenic HCT-116 cell line without p53 (p53Ϫ/Ϫ) has been derived by deletion of the p53 allele (31), and an interaction of NQO1 and p53 could be detected only in HCT-116 p53ϩ/ϩ and not in HCT-116 p53Ϫ/Ϫ cells (Fig. 2C). Immunoblot analysis of HCT-116 p53ϩ/ϩ and p53Ϫ/Ϫ cell lines confirmed the absence of p53 in the p53Ϫ/Ϫ cell line (Fig. 2D). These data also demonstrated that both HCT-116 p53ϩ/ϩ and HCT-116 p53Ϫ/Ϫ cells had marked levels of NQO1 as indicated by immunoblotting (Fig. 2D). Although HCT-116 cells from ATCC and HCT-116 p53ϩ/ϩ cells obtained as part of the p53 isogenic pair of cell lines (see "Materials and Methods") would be expected to be essentially identical, we also verified that ES936 had little effect on p53 or NQO1 protein levels in the HCT 116 p53ϩ/ϩ cell line (Fig. 2E), confirming the data in Fig. 1. Inhibition of NQO1 catalytic activity by ES936 was verified by activity measurements at the indicated time points and was 90% at 1 min and greater than 99% at all subsequent time points.
The experiments in Fig. 2 were performed using a p53 anti-body for immunoprecipitation. The reverse immunoprecipitation using an NQO1 antibody for immunoprecipitation and a p53 antibody for immunoblotting performed in the same manner resulted in co-migration of IgG heavy chain and interference with the p53 signal. To overcome this problem we utilized immunoprecipitating antibodies directly coupled to agarose beads, which prevents IgG heavy chain co-migration with p53. Using this technique (Fig. 3), an association of NQO1 and p53 could be detected in HCT-116 cells and also in primary HEKs and primary HMECs. Note the slower migrating species of p53 protein present in both the HEKs and HMECs. The shift in migration is likely due to differential phosphorylation of p53 in the primary cultures of normal cells versus the transformed cell line HCT-116. We have used an additional approach of employing an anti-NQO1 antibody for immunoprecipitation followed by immunoblotting using an anti-p53 antibody (DO-1) directly coupled to HRP. This technique eliminates the need for the use of a secondary antibody and thus prevents interference of IgG heavy chain with p53. Interaction of NQO1 and p53 was also evident in HCT-116 cells using this method (data not shown). The association of NQO1 with p53 was confirmed in coimmunoprecipitation experiments in which p53 was generated in an in vitro transcription/translation system (RRL). Expression of p53 in the RRL system using a p53 coding region inserted into the SP64 poly(A) plasmid was confirmed by immunoblotting (Fig. 4A). Recombinant NQO1 (0.5 g) was added FIG. 1. p53 and NQO1 protein levels and NQO1 activity following treatment of HCT-116 cells with ES936. A, p53 and ␤-actin protein levels were measured by immunoblot analysis in HCT-116 cells (50 g lysate) following treatment with the NQO1 mechanism-based inhibitor ES936 (100, 250 nM) for 2 and 8 h. B, NQO1 activity was measured using the rate of dicumarol-sensitive reduction of DCPIP in HCT-116 cells following treatment with ES936 (100 nM) for 2 h. C, NQO1 and ␤-actin protein levels were measured by immunoblot analysis in HCT-116 cells (50 g of lysate) following treatment with the NQO1 mechanism-based inhibitor ES936 (0 -500 nM) for 2 h. to the RRL system as purified protein and its presence confirmed also by immunoblotting (Fig. 4A). Immunoprecipitation was performed using an anti-p53 antibody (PAb421) followed by immunoblotting using anti-NQO1 monoclonal antibodies (Fig. 4B). Recombinant NQO1 was added to the RRL system either before or after synthesis of p53, and in either case an association between NQO1 and p53 was observed (Fig. 4B). An association of NQO1 and p53 was also apparent in a reverse immunoprecipitation employing either of two monoclonal antibodies to NQO1 (B771 or A180) followed by immunoblotting with anti p53 antibody DO-1 directly coupled to HRP (Fig. 4C).
mdm-2 Does Not Associate with NQO1 in HCT-116 Cells-Interaction of mdm-2 with p53 leads to rapid proteasomal degradation of p53, and detection of the p53-mdm-2 interaction is facilitated during proteasomal inhibition. In Fig. 5A, we demonstrate interaction of p53 with mdm-2 after treatment of HCT-116 cells with the proteasomal inhibitor MG132. No interaction of NQO1 with mdm-2 could be detected in HCT-116 cells either in the presence or the absence of MG132 (Fig. 5B). MG132 treatment had no effect on the levels of mdm-2 or NQO1 in HCT-116 cells as determined by immunoblot analysis (data not shown). DISCUSSION In this work, we demonstrate that wild type p53 associates with NQO1 in a variety of human tumor cells, in primary FIG. 4. Co-immunoprecipitation of recombinant NQO1 with p53 generated in an in vitro transcription/translation system. A, immunoblot analysis for p53 and NQO1 prior to immunoprecipitation of in vitro transcription/translation reactions. *, rhNQO1 (0.5 g) added before p53 synthesis; **, rhNQO1 (0.5 g) added after p53 synthesis. B, NQO1 was detected by immunoblot analysis following immunoprecipitation of p53 (PAb 421) generated in an in vitro transcription/translation system. C, in the reverse immunoprecipitation, p53 was detected by immunoblot analysis following immunoprecipitation of NQO1 with anti-NQO1 monoclonal antibodies (A180 or B771). human cell types, and in cell-free systems. Experiments were performed using a co-immunoprecipitation approach employing an anti-p53 monoclonal antibody followed by immunoblotting with an anti-human NQO1 antibody. Reverse immunoprecipitations were also performed employing either an NQO1 antibody directly coupled to agarose beads or an anti-NQO1 monoclonal antibody followed by immunoblotting with a p53 antibody directly coupled to HRP. Both of these techniques avoided any co-migration of IgG heavy chain with p53. Initially, HCT-116 cells were used to demonstrate the proteinprotein interaction of NQO1 and p53 using both forward and reverse immunopreceipitations. We also utilized an isogenic pair of HCT-116 cell lines differing only in their p53 status to explore the association of NQO1 and p53. NQO1 could be detected by immunoblotting in a p53 immunoprecipitate in HCT-116 p53ϩ/ϩ cells but not in HCT-116 p53Ϫ/Ϫ cells. An interaction of p53 and NQO1 was confirmed in a cell-free system where p53 was expressed in a coupled transcription/translation system containing recombinant hNQO1 protein. Importantly the interaction of NQO1 and p53 could also be detected in two primary human cell types, human epidermal keratinocytes and human mammary epithelial cells, suggesting that this protein-protein association occurs in a wide variety of human cell types. The interaction of NQO1 and p53 in HMECs was less efficient than in other cell types with similar levels of p53 and NQO1. This suggests that interaction of NQO1 and p53 may be cell type-specific and may reflect the predominance of other p53 interacting proteins in certain cell types. Previous work performed on the stabilization of p53 by NQO1 had suggested a redox mechanism of stabilization dependent on the catalytic function of NQO1 (25). In our experiments ES936, a specific suicide inactivator of NQO1, had no effect on the stability of p53. These data suggest that the physical association of NQO1 with p53 should be considered as an alternate mechanism of NQO1-mediated stabilization of p53 (Scheme 1).
One of the major mechanisms of degradation of p53 involves association with mdm-2, which targets p53 for proteasomal degradation (36,37). A p53-mdm-2 interaction could be observed after treating cells with the proteasomal inhibitor MG132, but no interaction of NQO1 with mdm-2 could be observed either in the presence or absence of proteasomal inhibitors. The NQO1*1 wild type protein is not subject to proteasomal degradation (13), and as expected, pretreatment of HCT-116 cells with MG132 did not result in any detectable interaction of NQO1 with mdm-2. These data suggest that NQO1 does not interact with mdm-2; therefore, the potential exists for competition of mdm-2 and NQO1 for p53. In addition to interacting with p53, we have previously demonstrated that wild type NQO1 is able to associate with Hsp70 and Hsp40 but not Hsp90 (28). p53 also interacts with proteins of the Hsp70 and Hsp40 families (38), and therefore, one possibility to consider is that p53, Hsp70/40, and NQO1 co-exist in a multiprotein complex. However, data obtained in the present study demonstrate that recombinant exogenous NQO1 when added to RRLs can interact with p53 transcribed and translated using a plasmid expression system. In previous work, we have shown that Hsp70 interacts with early, immature forms of NQO1 and does not interact with the mature protein (28). Since Hsp70 does not interact with the mature recombinant form of NQO1 (28), the participation of Hsp70 in the protein complex of p53 and the mature form of recombinant NQO1 added to the RRL system can be ruled out. Whether other proteins participate in the complex containing NQO1 and p53 remains to be elucidated (Scheme 1).
Proteins other than mdm2 are known to interact with p53 (39,40), so NQO1 is certainly not unique in its propensity to associate with p53. Such proteins include CBP/p300 (41), HIF1␣ (42), ZBP-89 (43), WOX-1 (44), protein kinase 2 (45), nuclear actin (46), ARF (47), PIAS/Sumo (48), and heat shock proteins (38,49). Protein-protein interactions may lead to either stabilization or degradation of p53 (39,40), and our data taken together with the previous work of Asher et al. (25,26) FIG. 5. Co-immunoprecipitation of MDM-2 with p53 but not NQO1 in HCT-116 cells treatment with proteasome inhibitor MG132. A, mdm-2 was detected by immunoblot analysis following immunoprecipitation of p53 from HCT-116 cells pretreated with MG132 (25 M) for 6 h. In the absence of MG132 pretreatment no association of mdm-2 with p53 was observed. B, no mdm-2 protein was observed by immunoblot analysis following immunoprecipitation of NQO1 from HCT-116 cells grown in the presence and absence of MG132 (25 M) for 6 h. suggests that NQO1 may be another important protein involved in stabilization of p53. Different stress signals utilize multiple pathways to stabilize p53 such as down-regulation of mdm2, modification of mdm2 activity by ARF binding, phosphorylation, or regulation of the localization of p53 and/or mdm-2 (50). The growing list of proteins that interact with p53 may well reflect multiple mechanisms of p53 regulation that differ depending on the particular stress response.
An important question that arises is whether the known metabolic functions of NQO1, the ability to detoxify quinones and protection against oxidative stress because of its role in ubiquinone and vitamin E metabolism, could be responsible for the protective effects of NQO1 against such a wide range of human solid tumors and leukemias (16 -20). Interestingly in recent work, disruption of the NQO1 gene in mice led to the development of myelogenous hyperplasia, and NQO1 null mice were found to have markedly decreased bone marrow p53 content relative to wild type animals (51). Given the wide range of cancers that have been associated with a lack of NQO1 due to the NQO1*2 polymorphism, it is conceivable that NQO1 is functioning via a non-catalytic mechanism to protect against neoplasia. There is precedent for proteins normally considered as metabolic enzymes playing other roles in the cell. For example, glutathione S-transferase has recently been demonstrated to physically associate with the c-Jun-N-terminal kinase (JNK1) leading to inhibition of kinase activity (52) and modulation of c-Jun-N-terminal kinase signaling and cellular proliferation (53,54).
In summary, we report a physical association of p53 and NQO1, which we propose as a non-catalytic mechanism influencing p53 stability. This may provide a mechanism for the increased susceptibility of individuals lacking NQO1 due to the NQO1*2 polymorphism to various forms of cancer.