Nitric oxide induces phosphorylation of p53 and impairs ... - Nature

Oncogene (2003) 22, 2857–2868

& 2003 Nature Publishing Group All rights reserved 0950-9232/03 $25.00 www.nature.com/onc

ORIGINAL PAPERS

Nitric oxide induces phosphorylation of p53 and impairs nuclear export Nicole Schneiderhan1,3, Andreja Budde1,3, Yanping Zhang2 and Bernhard Bru¨ne*,1 1

Department of Cell Biology, Faculty of Biology, University of Kaiserslautern, Erwin-Schro¨dinger-Strasse, 67663 Kaiserslautern, Germany; 2Department of Molecular and Cellular Oncology, University of Texas, MD Anderson Cancer Center, Houston, TX 77030, USA; 3Department of Medicine IV-Experimental Division, Faculty of Medicine, University of Erlangen-Nu¨rnberg, 91054 Erlangen, Germany

The tumor suppressor p53 accumulates under diverse stress conditions and affects cell cycle progression and/or apoptosis. This has been exemplified for endogenously produced or exogenously supplied nitric oxide (NO) and thus accounts at least in part for pathophysiological signaling of that bioactive molecule, although detailed mechanisms remain to be elucidated. By using luciferase reporter assays, we show that NO stabilized a transcriptionally active p53 protein. Considering that p53 is targeted by murine double minute (Mdm2) for ubiquitination and subsequent proteasomal degradation and knowing that this interaction is impaired by, for example, UV-treatment with concomitant stabilization of p53 we questioned the p53/Mdm2 interaction in the presence of NO. Although p53 became phosphorylated at serine 15 under the impact of NO, coimmunoprecipitation with Mdm2 and ubiquitination remained intact, thus excluding any interference of NO with this pathway. The importance of N-terminal p53 phosphorylation was verified with p53 mutants where the first six serine residues have been converted to alanine, and which do not accumulate in response to NO. Regulation of p53 stability can be also achieved by affecting nuclear–cytoplasmic shuttling and it was presented that leptomycin B, an inhibitor of nuclear export, caused p53 accumulation. Cell fractionation and immunofluorescence staining following NO-treatment revealed predominant nuclear accumulation of p53 in close association with serine 15-phosphorylation, which suggests impaired nuclear–cytoplasmic shuttling. This was verified by heterokaryon analysis. We conclude that attenuated nuclear export contributes to stabilization and activation of p53 under the influence of NO. Oncogene (2003) 22, 2857–2868. doi:10.1038/sj.onc.1206431 Keywords: p53-ubiquitination; Mdm2; nucleo–cytoplasmic shuttling; leptomycin; heterokaryon analysis

Introduction The tumor suppressor p53 controls cellular homeostasis by affecting cell cycle progression and apoptosis with the notion that this role is altered in many tumors, thus *Correspondence: B Bru¨ne; E-mail: [email protected] Received 1 July 2002; revised 23 December 2002; accepted 29 January 2003

contributing to malfunction. Under normal conditions p53 exhibits an extreme short half-life and the protein amount is maintained at a low, often undetectable level. Upon exposure to stress, p53 is stabilized, predominantly by post-translational modification, such as phosphorylation (Meek, 1998), acetylation (Gu and Roeder, 1997) or glycosylation (Shaw et al., 1996). As a consequence, p53 becomes active as a transcription factor and promotes transcription of cell cycle regulating genes such as p21WAF1/CIP1.or mdm2 (Barak et al., 1993; El-Deiry et al., 1993; Wu et al., 1993) but also apoptotic ones, exemplified by bax or Fas (Miyashita and Reed, 1995). Specifically, Nitric oxide (NO) provokes induction of p21WAF1/CIPl or bax (Bru¨ne et al., 1999; Nakaya et al., 2000; Yang et al., 2000) as well as murine double minute (Mdm2) with the latter one in response to long-term NO-exposure (Nakaya et al., 2000). Alternatively, p53 may provoke transcriptional independent alterations in facilitating cell demise, that is, apoptosis. Special attention is given to one of the downstream targets, Mdm2, which exhibits an unique relation with p53. The mdm2 gene is transcribed and translated under the control of p53 and the Mdm2 protein binds p53 at the N-terminus to facilitate its functional inactivation (Oliner et al., 1992; Chen et al., 1996; Kubbutat et al., 1998) and degradation (Haupt et al., 1997; Kubbutat et al., 1997). It is established that the rapid turnover of p53 typically is achieved through the ubiquitin proteasome pathway (Maki et al., 1996). Thus, Mdm2 directly functions as a p53-specific E3-ligase, which covalently attaches ubiquitin moieties (Honda et al., 1997). Ubiquitination of p53 demands a conserved Mdm2 region of eight cysteine and one histidine amino acid that forms two zinc-binding sites, known as a RINGfinger motive. The RING-finger domain additionally is required to facilitate nuclear export of p53 (Boyd et al., 2000; Geyer et al., 2000). Blocking p53 nuclear export attenuates Mdm2-mediated p53 degradation and concomitantly stabilizes an active p53, which implies that degradation of p53, largely takes place in the cytoplasm (for reviews, see Gottifredi and Prives, 2001; Liang and Clarke, 2001; Zhang and Xiong, 2001a). Diverse signaling pathways are employed by different agents to stabilize p53. For example, UV abrogates the p53/Mdm2 interaction, concomitantly reducing p53-ubiquitination, which is followed by protein

No impairs nuclear export of p53 N Schneiderhan et al

2858

stabilization (Fuchs et al., 1998; Maki, 1999). In this context, phosphorylation of p53 at serine 15 is discussed to be responsible in destabilizing p53/Mdm2 interactions. However, causation is not proven in vivo under stress conditions and the role of p53 phosphorylation in attenuating the p53/Mdm2 interaction is controversially discussed (Shieh et al., 1997; Dumaz and Meek, 1999). Phosphorylation of p53 at several sites, for example, serine 15 is known for NO (Nakaya et al., 2000) as well as other stress conditions such as UV, IR or CdCl2treatment (Siliciano et al., 1997; Matsuoka and Igisu, 2001). However, whether the p53/Mdm2 interaction remains intact under the influence of NO needs to be investigated. Alternatively, leptomycin B attenuated chromosome region maintenance 1 (CRM1)/exportin1-mediated nuclear export of p53 and achieved p53-activation (Fukuda et al., 1997; Kudo et al., 1998). Zhang and Xiong (2001b) established a correlation between serine 15-phosphorylation after UV-treatment and reduced nuclear export of p53. Considering that NO emerged as a pathophysiological, endogenously produced mediator to stabilize a serine 15-phosphorylated p53 opened the possibility to explain NO-evoked p53 accumulation by suggesting impaired export. It is appreciated that the production of NO is an essential determinate in auto- and paracrine signaling. NO is generated under inflammatory conditions by constitutively expressed versus inducible isoenzymes and may serve as a cytotoxic molecule to produce cell demise along an apoptotic pathway. The ability of NO to promote a functional p53-response has been confirmed in murine and human systems (Messmer et al., 1994; Forrester et al., 1996) and has been extended to the observation that p53 in turn downregulates inducible NO-synthase (iNOS) expression through inhibition of the iNOS promoter. Although numerous reports confirmed that endogenously generated or exogenously delivered NO stabilized p53 to affect cell cycle progression and/or apoptosis (Messmer and Bru¨ne, 1996;

Nakaya et al., 2000; Yang et al., 2000; or for review, see Bru¨ne et al., 2001), we still lack detailed knowledge of how NO causes these alterations. Considering NO as a physiological regulator of p53, we investigated mechanisms of p53 stabilization under the impact of NO. We questioned the p53/Mdm2 interaction, studied ubiquitination of p53, paid attention to post-translational modifications and examined nuclear–cytoplasmic shuttling of p53 by heterokaryon analysis.

Results NO provokes nuclear accumulation of p53 It was our attention to delineate mechanisms of p53 accumulation following NO exposure. In a first set of experiments, we followed the time-dependent accumulation of p53 in murine NIH 3T3 fibroblasts after exposure to 1 mm S-nitrosoglutathione (GSNO) (Figure 1a). Western blots revealed stabilization of p53 2 h following the addition of GSNO. p53 accumulation reached a maximum at 4 h, with declining protein levels afterwards. At 8 h, stabilized p53 disappeared, basically showing control values. However, the time–response was cell specific because p53 stabilization in RKO cells showed a slower response with maximal accumulation at 24 h (data not shown). As predicted for a transcription factor, p53 predominantly accumulated in the nuclei (Figure 1b). This was shown in RKO cells treated with 500 mm GSNO or with leptomycin B, an inhibitor of the CRM1-mediated nuclear export machinery. Cell fractionation confirmed the majority of p53 to reside in the nuclei after NOtreatment (compare lanes 2, 5 and 8) to a similar extent as found for leptomycin B (lanes 3, 6 and 9). Of course, cell fractionation is not 100% and one cannot expect a complete block in the nuclear export by any drug, which explains minor p53 amounts in the cytosol. In addition,

Figure 1 NO provokes stabilization of p53. (a) Time-dependent accumulation of p53 was detected by Western blotting in NIH 3T3 cells following exposure to 1 mm GSNO. Lane 1 shows controls without treatment, (b) Accumulation of p53, followed by Western blotting, in total lysate, nuclear versus cytoplasmic fractions of RKO cells following the treatment with 500 mm GSNO for 24 h or 5 ng/ ml LMB for 12 h. (c) Immunofluorescent staining of p53 in NIH 3T3 cells following stimulation with 1 mm GSNO for 3.5 h (image 2) versus control (image 1). Nuclear localization was proven by counterstaining with DAPI (images 3 and 4) Oncogene


2859

we proved nuclear accumulation of p53 in NIH 3T3 cells by immunofluorescence (Figure 1c). Anti-p53 staining showed clear nuclear accumulation of p53 confirmed by DAPI fluorescence. Thus, nuclear accumulation of p53 is a general feature of cells exposed to NO donors. NO stabilizes a transcriptionally active p53 Based on earlier reports, it appears that NO upregulates classical p53 downstream target genes such as mdm2, bax or p21WAF1/CIP1 although direct reporter gene activation in response to NO has not been reported. To ensure a transcriptionally active p53 under our conditions, we transfected NIH 3T3 cells with the p53dependent reporter PG13-Luc, containing 13 p53binding sites upstream of a polyoma-virus early promoter that directs transcription of luciferase and determined luciferase activity following the exposure to NO. Two chemically distinct NO donors such as GSNO and spermine-NONOate (SpNO) significantly increased luciferase activity (Figure 2a). The somewhat higher activity achieved with SpNO is reflected by stronger protein accumulation compared to GSNO under those experimental conditions (data not shown), which may be explained by different kinetics of NO release from these NO donors. As a control, cells were treated with proteasome inhibitor MG 132, which provoked accumulation of a transcriptionally active p53 by blocking its degradation (Lopes et al., 1997). As a further control, we ran assays in parallel by using MG15-Luc, a luciferase reporter construct with 15 mutated p53 binding sites, thus not allowing specific binding of the tumor-suppressor protein. With this construct, no p53dependent activation of luciferase activity appeared. Additionally, stimulation of RKO cells with 250 or 500 mm GSNO for different times resulted in increased Mdm2 protein expression (Figure 2b). Maximum Mdm2 level were observed after 8 h of GSNO in RKO cells. We

conclude that NO stabilized a transcriptionally active p53 that drives Mdm2 expression. NO does not interfere with thep53/Mdm2 interaction It is established that Mdm2, via protein–protein interactions, targets p53 for degradation by the ubiquitin proteasome pathway. Mdm2 facilitates polyubiquitination of p53 and assists in its nuclear–cytoplasmic shuttling. To see whether NO interferes with the p53/ Mdm2 interaction, thereby stabilizing p53, we transfected NIH 3T3 fibroblasts with an Mdm2 expression vector (pCHDM1A) with the intention to coimmunoprecipitate p53 in the presence or absence of NO (Figure 3). Under control conditions low levels of p53 co-precipitated with Mdm2 (Figure 3a, lane 1) with the notion that p53 was present at undetectable levels in cell extracts (input control). With GSNO, p53 accumulated in cell extracts (see input) and significant amounts of p53 co-precipitated with Mdm2 (Figure 3a, lane 2). Apparently, NO left the p53/Mdm2 complex intact. As a positive control for these experiments, we exposed cells to UV, a condition known to destabilize p53/Mdm2 interactions (Figure 3a, lane 4). As predicted, p53 accumulated following UV-treatment and the p53/Mdm2 interaction disappeared. In contrast, blocking the proteasome by MG 132 stabilized p53 and facilitated a strong p53/Mdm2 interaction (Figure 3a, lane 3). The situation found in RKO cells gave comparable results, although endogenous Mdm2 was used to coimmunoprecipitate p53. Significant amounts of stabilized p53, following an 8- and 24-h-lasting GSNOtreatment (see input) coimmunoprecipitated with Mdm2 (Figure 3b, lanes 2 and 3 versus controls shown in lane 1). Unexpectedly, when we reprobed the blot using an antibody that recognizes serine 15-phosphorylated p53, we found that such a form of p53 (i) accumulated after NO and (ii) was present in Mdm2 immunoprecipitates (Figure 3b). MG 132 stabilized p53 as well but

Figure 2 Transcriptional activation of p53. (a) NIH 3T3 cells were transiently transfected with a p53 reporter construct PG13-Luc or a control plasmid (MG15-Luc) to follow luciferase activity (RLU, relative light units) in response to NO- or MG 132-treatment. Stimulation was for 8 h with the concentrations indicated, (b) Western analysis of endogenous Mdm2 (SMP14 antibody) and p53 (DO-1 antibody) in RKO cells following treatment with 250 mm GSNO (lanes 2–4), 500 mm GSNO (lanes 5 and 6), 3 mm MG 132 (lane 7) or in controls (lane 1) Oncogene


2860

Figure 3 The interaction between p53 and Mdm2 after NO-treatment. (a) NIH 3T3 fibroblasts were transiently transfected with the Mdm2 expression vector pCHDM1A. Cells were then stimulated with 1 mm GSNO (4 h, lane 2), 5 mm MG 132 (4 h, lane 3), exposed to UV (96 J/m2) with a subsequent 8-h lasting incubation period (lane 4), or remained as controls (lane 1). Mdm2 (SMP14 antibody) was precipitated followed by immunodetection of either p53 (Pab122 antibody) or Mdm2 (SMP14 antibody). Input controls prior to immunoprecipitation (3% input) are shown for Mdm2 and p53. (b) RKO cells were treated with 500 mm GSNO (8 or 24 h, lane 2 versus 3) with MG 132 (3 mm, 6 h, lane 4) or remained as controls (lane 1). Endogenous Mdm2 was precipitated followed by detection of Mdm2 and p53. Serine 15-phosphorylation of p53 was detected by reprobing blots with a phosphoserine 15-specific p53 antibody. Input controls for Mdm2, p53 and serine 15-phosphorylated p53 are shown. Detection of tubulin (a) or GAPDH (b) served as loading controls

did not cause serine 15-phosphorylation (Figure 3b, lane 4 versus lanes 2 and 3). Taken together this results indicate that NO does not interfere with the p53/Mdm2 interaction despite phosphorylation of p53 at serine 15. Ubiquitination of p53 remained unchanged following NO-treatment In the ubiquitin proteasome pathway, p53 is polyubiquitinated by the E3-ligase activity of Mdm2 and concomitantly degraded. Considering that NO left the interaction between p53 and Mdm2 intact opened the possibility that NO blocked the E3-ligase activity of Mdm2, thus interfering with polyubiquitination and degradation of the protein. This hypothesis appeared logical considering that the cysteine-residue or the complexed Zn2+ of the RING-finger motive of Mdm2 can be targeted by NO. To address this question, we transfected NIH 3T3 cells with His-tagged ubiquitin, HA-tagged p53 and increasing amounts of Mdm2 expressing plasmids prior to exposure to spermine-NO, MG 132 or UV (Figure 4a). Expression of HA-tagged Oncogene

p53, endogenous p53 and Mdm2 was followed by Western analysis of aliquots of cell lysates (input) or following affinity purification of ubiquitinated p53 on Ni-NTA-beads (Figure 4a, upper panel). This allowed to compare not only protein levels of p53 but also to follow ubiquitination of the protein. Comparing control and spermine-NO-treated cells revealed a comparable degree of ubiquitin ladder formation with HA-p53 ubiquitinated isoforms extending from approximately 60 to 130 kDa. Having in mind that only ubiquitination of exogenous p53 is detected under these conditions, we conclude that NO does not interfere with ubiquitination reactions of exogenous p53. The amount of ubiquitination transiently increases with increasing Mdm2 amounts, that is, E3-ligase and decreases with high amounts of coexpressed Mdm2, which might be the consequence of faster proteasomal degradation. As experimental controls, we used MG 132 to increase p53 ubiquitination (Figure 4a, lane 13), whereas UV-treatment was employed to decrease ubiquitination (Figure 4a, lane 14). Cells, only transfected with His-tagged ubiquitin and exposed to 1 mm GSNO for 4 h served as a control


2861

Figure 4 Ubiquitination of p53 under the impact of NO. (a) NIH 3T3 cells were transiently transfected with His-tagged ubiquitin, HA-tagged p53 (0.6 mg) and increasing amounts (0–12 mg) of human Mdm2 (pCHDM1A) expression plasmids. At 24–28 h posttransfection cells were exposed to 0.5 mm spermine-NO (4 h, lanes 2, 4, 6, 8 and 10), 5 mm MG 132 (4 h, lane 13), UV (96 J/m2) followed by an 8-h lasting incubation period (lane 14), or remained as controls (lanes 1, 3, 5, 7, 9 and 12). Extracts from cells transfected with the ubiquitin expression plasmid only, followed by stimulation with 1 mm GSNO for 4 h are shown in lane 11. His-tagged p53 was purified on Ni-NTA beads as described in Experimental procedures and detected by Western blotting anti-HA. Input controls are shown in the lower panels for Mdm2 (SMP14 antibody), endogenous (endo) p53 (Pab122 antibody), exogenous (exo) HA-p53 (HA. 11 antibody) or b-tubulin as a loading control, (b) Ubiquitination of endogenous (endo) p53 in RKO following treatment with 500 mm GSNO (8 or 24 h, lanes 2 versus 3), 3 mm MG 132 (6 h, lane 4), UV (as described in Figure 4a) or in controls (lane 1). Western blotting with DO-1 allowed detection of ubiquitinated p53. Shorter exposition of the whole gel in the middle panel allowed to follow p53 stabilization. As a loading control GAPDH was detected (lower panel), (c) RKO cells were stimulated with 500 mm GSNO for 24 h, 5 ng/ml leptomycin B (LMB) for 12 h or left untreated. Ubiquitinated p53 was detected by Western blotting (DO-1 antibody) in total lysate (lanes 1–3) and the cytosolic (lanes 4–6) versus the nuclear (lanes 7–9) compartments. Mdm2 was detected by the SMP14 antibody. Distribution of the marker proteins b-tubulin and DNA-topoisomerase are indicated in the lower panels Oncogene


2862

to eliminate crossreactivity of antibodies (Figure 4a, lane 11). In the following set of experiments, performed in RKO cells, we asked whether endogenous p53 shares regulatory features of exogenous p53 (Figure 4b). RKO cells were left untreated (Figure 4b, lane 1), incubated with GSNO for 8 or 24 h (Figure 4b, lanes 2 versus 3), exposed to MG 132 (Figure 4b, lane 4) or UV (Figure 4b, lane 5). Accumulation of p53 in response to all stimuli was followed in cell lysates by Western blotting using the DO-1 anti-p53 antibody. Detection of endogenous, ubiquitinated p53 showed increased ubiquitination in response to MG 132 and a decrease following UVtreatment (Figure 4b, lanes 4 and 5 versus controls shown in lane 1). Controls (Figure 4b, lane 1) revealed some degree of ubiquitination in close association with minor p53 accumulation. Following NO-treatment, p53 significantly accumulated in cell lysates after 8 and 24 h, but ubiquitination remained intact (Figure 4b, lanes 2 and 3 versus controls shown in lane 1). Comparing the relative amounts of ubiquitinated p53 to the level of stabilized protein suggests that the overall rate of ubiquitination remains constant under the impact of NO. Based on these results, we assume that NO neither affects ubiquitination of exogenous HA-tagged nor of endogenous p53. To exclude that NO may directly block enzymatic activities found in the core unit of the proteasomal degradation machinery, that is, the 20S-proteasome, we assessed formation of AMC from the substrate SucLLVY-AMC in the presence and absence of NO. Neither in RKO nor in NIH 3T3 cells, NO derived from GSNO or spermine-NO, interfered with degradation of Suc-LLVY-AMC under our experimental conditions (data not shown). Of course, MG 132 significantly impaired AMC formation, thus indicating attenuated proteasomal activity. To understand p53 accumulation under the impact of NO, we went on to examine nuclear–cytoplasmic shuttling of the tumor suppressor since ubiquitination of p53 in the nucleus versus cytoplasmic degradation is known. Therefore, we exposed RKO cells to GSNO or leptomycin B, which is known to stabilize p53 by blocking nuclear export (Figure 4c). Western blot analysis was used to detect ubiquitinated p53 in total RKO cell lysate (Figure 4c, lanes 1–3) and the cytoplasmic (Figure 4c, lanes 4–6) versus nuclear compartment (Figure 4c, lanes 7–9). DNA-topoisomerase and b-tubulin were chosen as marker enzymes to assure nuclear versus cytoplasmic separation. GSNO and leptomycin B predominantly led to an accumulation of ubiquitinated forms of p53 in the nucleus. Low amounts of less ubiquitinated p53 were detected in the cytosol as well. Apparently, NO and leptomycin B share the ability to accumulate p53 in the nucleus, opening the possibility that NO blocks nuclear export of ubiquitinated p53. It is assumed that Mdm2 may participate in nuclear– cytoplasmic shuttling of p53. Therefore, we investigated the cellular distribution of Mdm2 by Western analysis (Figure 4c, lanes 4–9). Results pointed to a Oncogene

homogeneous, nuclear–cytosolic distribution of Mdm2 and excluded an impact of NO in Mdm2 compartmentalization. NO induces N-terminal phosphorylation of p53 and attenuates nuclear–cytoplasmic shuttling Phosphorylation of p53 at multiple sites is established. We showed that after NO-treatment p53 is phosphorylated on serine 15 (Figure 3b). To examine this in more detail, we exposed RKO cells to GSNO for different times and analysed p53 using either an anti-serine 15-phosphorylated p53 antibody or an anti-p53 antibody that allowed recognition of total p53 (Figure 5a). As shown in Figure 5a, NO stabilized p53 with the further information that the protein is phosphorylated at serine 15. Similar results were seen when NIH 3T3 or U2OS cells were exposed to GSNO (data not shown). For comparison, it is indicated that MG 132 stabilized p53 without causing phosphorylation at serine 15 (Figure 5a, lane 5). To test whether phosphorylation of p53 in the N-terminus is of importance for its accumulation after NO-treatment, we transiently transfected U2OS cells with a p53 construct in which the N-terminal serine residues 6, 9, 15, 20, 33 and 37 (p53 D62–96 S>A) have been substituted by alanine (Figure 5b). Whereas ectopic expressed p53 D62–96 still accumulated under the impact of NO (Figure 5b, lanes 1 and 2), p53 D62—96 S>A did not (Figure 5b, lanes 3 and 4). However, inhibition of proteasomal degradation of p53 D62–96 S>A by MG 132 in the presence of NO still allowed p53 accumulation (Figure 5b, compare lanes 5–7) demonstrating regulatory properties of this mutant. Having established the importance of Nterminal phosphorylation of p53 for protein accumulation, we went on to access in a more direct way whether shuttling of p53 is abrogated under the impact of NO by performing heterokaryon analysis. To achieve this, we used U2OS cells that overexpress p53 because of adenoviral transfection (Figure 5c). U2OS cells were left untreated (upper panel), exposed to GSNO (middle panel) or stimulated with leptomycin B (lower panel), and concomitantly fused with adenoviral-hARFinfected Saos-2 cells. U2OS/Saos-2 heterokaryons were identified by phase contrast, while staining of hARF allowed to discriminate Saos-2 cells from hARFnegative U2OS cell. Migration of p53 from U2OS into Saos-2 nuclei was examined either by using the phosphoserine 15-specific p53 antibody or by analysing total p53. In controls, a portion of p53 shuttled from U2OS into Saos-2 nuclei, indicating that export of p53 from p53-positive into p53-negative cells occurred. Following GSNO-treatment serine 15-phosphorylated p53 significantly increased in U2OS (Figure 5c). However, shuttling of serine 15-phosphorylated p53 from U2OS nuclei to Saos-2 nuclei was completely attenuated under the influence of NO. These findings were mimicked by using leptomycin B that blocks nuclear export of p53 by disrupting the nuclear export complex. Leptomycin B promoted serine


2863

Figure 5 NO attenuates nuclear export of serine 15-phosphorylated p53. (a) Western analysis of serine 15-phosphorylated and total p53 in RKO cells following treatment with 500 mm GSNO (lanes 2 and 3), 5 ng/ml LMB (lane 4), 3 mm MG 132 (lane 5) or in controls (lane 1). An antiphosphoserine 15 specific versus DO-1 (total p53) antibody was used, (b) U2OS cells were transiently transfected with p53 D62–96 (lanes 1 and 2) or p53 D62–96 S>A (lanes 3–7) expressing plasmids, stimulated for 12 h with 750 mm GSNO (lanes 2, 4 and 6) or left untreated (lanes 1, 3 and 5). As a positive control p53 D62–96 S>A transfected cells were costimulated for 12 h with 750 mm GSNO and 5 mm MG 132 (lane 7). p53 was detected by using the 1801 anti-p53 antibody. The position of endogenous p53 is indicated by a star. The arrow indicates transiently expressed p53 at lower size because of deletion of amino acids 62–96. (c) U2OS cells (expressing adenoviral transfected p53) and Saos-2 cells (expressing adenoviral transfected hARF) were cocultured (upper panel), stimulated for 6 h with either 1 mm GSNO (middle panel) or 10 ng/ml LMB (lower panel). Nuclear–cytoplasmic shuttling of total p53 (anti-p53, FL-393) or serine 15-phosphorylated p53 (anti-ser15-P-p53) was examined by heterokaryon assays. Saos-2 cells were identified by antibodies directed against hARF

15-phosphorylation of p53 in U2OS cells, and blocked shuttling of serine 15-phosphorylated p53 from U2OS nuclei to Saos-2 nuclei. We conclude from these findings that NO, in analogy to leptomycin B, attenuates nuclear–cytoplasmic shuttling of serine 15-phosphorylated p53.

Discussion Here, we investigated mechanisms of p53 stabilization under the impact of NO. NO provoked accumulation of a transcriptionally active and N-terminal-phosphorylated p53 with predominant nuclear localization. Oncogene


2864

Stabilized and serine 15-phosphorylated p53 still bound its negative regulator Mdm2 and Mdm2-evoked polyubiquitination of p53 remained intact. The importance of N-terminal phosphorylation for NO-elicited p53 accumulation was verified by using p53 D62–96 S>A, which failed to respond to NO. Based on cell fractioning and heterokaryon analysis, we suggest that NO prevents, at least in part, nuclear–cytoplasmic shuttling of p53, which correlates with N-terminal, that is, serine 15phosphorylation and results in nuclear protein stabilization/activation. NO affects cell cycle progression and/or initiates apoptosis in part by a p53-dependent pathway (Messmer and Bru¨ne, 1996). This is established for various murine and human cell systems (Bru¨ne et al., 1999). Considering the still growing number of situations that are characterized by induction of the high NO-output system, that is, iNOS expression, one may envision that the ability of NO to modulate p53 is of pathophysiological relevance. Besides a basic knowledge describing p53 accumulation in response to NO, detailed mechanistic insights remained open. However, we need to increase our knowledge on these signaling pathways to understand numerous reports showing alterations of cell cycle progression and/or apoptosis in association with NO-evoked p53 accumulation. Cell fractioning and fluorescence microscopy revealed predominantly nuclear localization of p53 following NO-treatment with the notion that p53 is transcriptionally active. Transactivation of p53 downstream targets such as p21 or Mdm2 by NO has been observed earlier (Nakaya et al., 2000), but whether p53 was directly responsible for this transactivation was not clear. Stabilization of p53 and activation of a luciferase reporter construct is also elicited by MG 132, a compound known to block proteasomal degradation of p53. Further experiments established that NO and leptomycin B provoked phosphorylation of p53 at serine 15. This is in line with current concepts proposing posttranslational modification of p53 at either the N- or Cterminus to affect stability regulation in general, with serine 15-phosphorylation being a prerequisite for the transcriptional activity of p53 in particular (Dumaz and Meek, 1999). Studies with p53 N-terminal mutations at serines 15, 20, 33 or 37 indicated multiple and functionally overlapping phosphorylation sites to control p53 activity in response to DNA damage (Shieh et al., 1997; Dumaz and Meek, 1999; Hirao et al., 2000; Turenne et al., 2001). Whether kinases such as ATR or DNA-PK, known to be involved in post-translational modification of p53, are engaged in transmitting the NO signal needs further investigations. However, one can exclude the necessity of ATM, since it is shown that in ATM/ cells p53 still accumulates under the impact of NO (Wang et al., 2002). Kim et al. (2002) showed the requirement of p38 kinase to achieve phosphorylation of p53 at serine 15 after NO-treatment in chondrocytes. In contrast, our investigations with the p38 kinase inhibitor SB203580 and the MEK1 inhibitor PD98059 could not prevent NO-mediated p53 accumulation and phosphorylation at serine 15 in RKO or U2OS Oncogene

cells. Along that line previous reports (Callsen and Bru¨ne, 1999) also excluded the involvement of these kinases to account for p53 accumulation in response to NO donors in macrophages. Thus, a cell type-specific contribution of MAPK-family members exists. Phosphorylation of p53 not only affects transcriptional activation of p53 but also is considered to be a master regulator of p53/Mdm2 complex formation. Based on in vitro experiments, it is predicted that phosphorylation of p53 at serine 15 is incompatible with p53/Mdm2 interactions, thus facilitating protein dissociation with concomitant p53 stabilization (Shieh et al., 1997; Pise-Masison et al., 1998). However, based on luciferase assays and in vitro interaction studies, Dumaz and Meek (1999) noticed that serine 15 modification of p53 alone did not dissociate p53 from Mdm2 although modification of serine 15 stimulates the transcriptional activity of the tumor suppressor. This situation is reflected by NO-treatment when p53/Mdm2 complex formation remained intact in NIH 3T3 fibroblasts as well as in the human RKO cells despite observing enhanced luciferase activity. Apparently, different stimuli employ divergent pathways to accumulate p53. In cell culture, we showed that NO stabilized p53 that is phosphorylated at serine 15 and activates p53 with Mdm2 remaining bound. In contrast, although expected based on literature information, UV behaved differently. UV-elicited serine 15-phosphorylation of p53 increases Mdm2 degradation and therefore stabilizes p53 (Maki et al., 1996; Fuchs et al., 1998). Recently, it was reported that an initial, however transient, drop in Mdm2 accounts for early accumulation of p53 in response to NO, whereas in the second phase p53 remained stabilized although Mdm2 increases above controls because of transactivation of the mdm2 gene (Wang et al., 2002). Our results confirm an increased Mdm2 expression and p53 stabilization under the influence of NO especially at later time points. While Wang et al. (2002) noted difficulties in explaining p53 accumulation under conditions of elevated Mdm2 expression, our study offers an explanation based on nuclear trapping of p53. Importantly, recent studies suggest that Mdm2 also affects p53 localization (Tao and Levine, 1999a; Boyd et al., 2000; Geyer et al., 2000). It is proposed that binding of Mdm2 escorts p53 out of the nucleus (for reviews, see Liang and Clarke, 2001; Zhang and Xiong, 2001a) with the notion that the RING-finger domain of Mdm2 rather than the nuclear export sequence (NES) is required (Boyd et al., 2000; Geyer et al., 2000). However, fractionation of cells after NO-treatment showed equal distribution of Mdm2 in the nuclear and cytoplasmic compartment, which implies that compartmentalization of Mdm2 is not affected by NO. Another player in p53 stability regulation is the tumor suppressor p14ARF, which stabilizes p53 by attenuating nuclear export of Mdm2 (Tao and Levine, 1999b). An active role of p14ARF in facilitating NO-elicited p53 accumulation was excluded in studies by Wang et al. (2002) by still observing p53 accumulation in response to NO in p14ARF-negative fibroblasts.


2865

The remaining interaction of p53 with its E3-ligase Mdm2 after NO does not necessarily guarantee an efficient p53-ubiquitination. The RING-finger motive of Mdm2 forms two interleaved zinc-binding sites and it is known that NO induces Zn2+ release in vitro from metallothionein (Kroncke et al., 1994) or increases intracellular free Zn2+ (Berendji et al., 1997). If NO targets the RINGfinger motive to release Zn2+, one would expect attenuation of polyubiquitination of p53. However, we noticed an unaltered ubiquitination pattern of either exogenous HA-tagged or endogenous p53 under NOtreatment. To validate our test system, we confirmed that UV irradiation, indeed, decreased p53-ubiquitination that can be explained by attenuated p53/Mdm2 interactions as seen in coimmunoprecipitation experiments. Furthermore, enhanced ubiquitination of p53 resulted from the addition of MG 132, that is, blocked proteasomal degradation, which caused stabilization and activation of the tumor-suppressor protein. Additional experiments excluded that NO directly blocked the core proteolytic activity of the 20S-proteasome. Therefore, we conclude that NO achieves accumulation of polyubiquitinated p53 without impairing p53/Mdm2 interactions. Nuclear export of p53 may not only depend on Mdm2 considering that p53 contains an NES in its tetramerization domain as well (Stommel et al., 1999). It is suggested that this sequence is masked upon tetramerization induced by phosphorylation of p53 at serine 392. NO, besides inducing phosphorylation of serine 15, causes serine 392-phosphorylation as well, which may contribute to an altered export behavior of p53 under NO (Nakaya et al., 2000). However, heterokaryon analysis with a p53 mutant, were the last 15 or 30 Cterminal amino acids have been removed, showed an unaltered p53 export (data not shown), excluding a major contribution of serine 392-phosphorylation in p53 shuttling. Export of p53 from the nucleus is blocked by leptomycin B, suggesting a CRM1-dependent mechanism (Stommel et al., 1999). Leptomycin B interferes with complex formation between the nuclear export receptor CRM1, RanGTP and NEScontaining proteins, by covalent modification of CRM1 (Fornerod et al., 1997; Ossareh-Nazari et al., 1997). To date, it remains open whether p53 is directly targeted by CRM1 because Mdm2 contains a leucine-rich NES as well and mutation of this site diminished Mdm2mediated p53 degradation, that is, nuclear export (Roth et al., 1998). This opens the possibility that CRM1 targets rather Mdm2 than p53. However, Stommel et al. (1999) showed that p53, which fails to bind Mdm2 (p53 L14Q/F19F or L22Q/W23S) underwent nuclear– cytoplasmic shuttling. To prove unequivocally alterations in p53 shuttling heterokaryon analysis is the method of choice. By using this method, we obtained evidence that NO attenuated nuclear–cytoplasmic shuttling of serine 15-phosphorylated p53. Experimentally, our results are substantiated by using leptomycin B, an established blocker of p53 nuclear export.

Serine 15-phosphorylation has been correlated with attenuated nuclear export and it remains to be established whether this striking correlation accounts for a cause–effect relation (Zhang and Xiong, 2001b). Evidence that NO may affect import or export of p53 emerged recently (Davis et al., 2000). It was shown that bleomycin stabilized p53 in mouse macrophages in the cytoplasm and that an additional, short-lasting NO-pulse provoked predominant nuclear localization of p53. Blocking p53 export from the nucleus provides a rational explanation for unaltered p53/Mdm2 interactions and polyubiquitination of p53. Mechanistically, leptomycin B targets an active cysteine residue in CRM1 that blocks formation of the export protein complex. A similar mechanism may apply for NO as well, considering the ability of NO-related redox species to nitrosylate thiol groups (Stamler et al., 1992). This working hypothesis will be tested in future. Our study provides evidence that NO accumulates p53 by attenuating the nuclear export and thereby defines a novel signaling pathway of NO. This helps to understand how NO transmits a signal that in turn affects cell cycle progression and/or apoptosis.

Materials and methods Materials SpNO was obtained from Biotrend (Cologne, Germany). GSNO was synthesized as described by Hart (1985). MG 132, cycloheximide, polyethylene glycol 8000 and DAPI (40 ,6diamino-2-phenylindole) were obtained from Sigma-Aldrich (Steinheim, Germany). MG 132 was dissolved in DMSO. Medium and supplements were obtained from Biochrom (Berlin, Germany) or PAA (Linz, Austria). Cell culture Murine NIH 3T3 fibroblasts, human U2OS cells (both expressing wild-type p53) and human Saos-2 cells (p53 negative) were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal calf serum, 100 U/ml penicillin and 100 mg/ml streptomycin. RKO cells were kept in RPMI with supplements as specified for DMEM. Cells were cultured at 371C and 5% CO2. Plasmids and viruses For the description of plasmids, we refer to original articles: for PG13-Luc and MG15-Luc, see Kern et al. (1992); for pMT 107 (His6-tagged ubiquitin), see Treier et al. (1994); for HAp53, see Marin et al. (1998); for pCHDM1A (human Mdm2), see Chen et al. (1995). Recombinant adenoviruses are described in Zhang and Xiong (2001b). p53 D62–96 (D62–96) and p53 D62–96 S>A (Nala) are described in Hengstermann et al. (1998). Transfection and luciferase assay NIH 3T3 cells were transfected with SuperFect (Quiagen, Hilden, Germany) according to the manufacturer’s protocol. Cells were seeded 1 day prior to transfection at 60% confluency. Transfection mixture was made of DNA Oncogene


2866 and SuperFect at 1 mg/7 ml. Cell treatment was started 20–25 h post-transfection. For luciferase assay, we followed instructions provided by the distributor (Promega, Mannheim, Germany).

Immunofluorescence Cells were grown on coverslips. Following treatment, cells were fixed with 3% paraformaldehyde (Serva, Heidelberg, Germany) for 15 min on ice. Cells were permeabilized for 10 min with ice-cold 0.2% Triton X-100, blocked with FCS for 20 min at room temperature and incubated overnight at 41C with 10 ng/ml anti-p53 antibody (clone Pab122, protein Apurified hybridoma supernatant, kindly provided by Professor Stahl, Homburg/Saar, Germany). Cells were washed three times with PBS and incubated for 2 h at 371C with a rabbit anti-mouse Cy3-conjugated antibody (Jackson Immuno Research Laboratories/Dianova, Hamburg, Germany). After washing with PBS, cells were stained for 5 min with 1 mg/ml DAPI and mounted in Mowiol (Calbiochem-Novabiochem, Schwalbach, Germany). Analysis was with a Leica microscope (Leica, Heerbrugg, Switzerland).

Heterokaryon assay Basically, we followed a modified protocol of Roth et al. (1998). The day before fusion, U2OS and Saos-2 cells were infected with Ad-p53 or Ad-hARF, respectively. Cells were trypsinized, mixed at a 1 : 1 ratio, seeded in a six-well dish at 80–90% confluency and covered with medium. At 18–20 h postinfection, the coculture was stimulated with GSNO or leptomycin B (kindly provided by Dr Yoshida Minoru, Tokyo, Japan) for 6 h. At 20 min before fusion, medium was exchanged with DMEM containing 10% FCS and 50 mg/ml cycloheximide. To induce fusion, cells were washed with PBS, covered with PBS/50% polyethylene glycol 8000 for 3 min at 371C, washed with PBS and incubated with DMEM/10% FCS/50 mg/ml cycloheximide for 1 h at 371C. Thereafter, cells were fixed with 3% paraformaldehyde for 10 min at room temperature, permeabilized with 0.2% Triton X-100 (5 min, 41C) and blocked for 30 min at room temperature with 0.5% BSA. Cells were incubated with the following primary antibodies: anti-p53 (0.1 mg/well, FL-393), antiphospho-p53 (serine 15) (0.15 ml/well, New England Biolabs, Frankfurt am Main, Germany) and anti-p14ARF (0.4 mg/well, clone 14P02 NeoMarkers, Fermont, CA, USA) for 1 h at room temperature. Cells were washed, incubated for 60 min at room temperature with a mixture of fluorescein isothiocyante (FITC)-conjugated donkey anti-rabbit, rhodamine red-X-conjugated donkey anti-goat or AMCA (7-amino-4-methylcoumarin-3-acetic acid)-conjugated donkey anti-mouse (all Jackson Immuno Research, West Grove, PA, USA) and covered with mounting solution followed by documentation using an inverted fluorescence microscope (Olympus).

Western blot analysis and antibodies For Western blotting, the following antibodies were used: antiMdm2 (SMP14, Santa Cruz, Heidelberg, Germany), anti-p53 (DO-1, Oncogene, Schwalbach, Germany, or 1801, Dianova) or purified Pab122 hybridoma supernatant, antiphospho-p53 (serine 15) (New England Biolabs, Frankfurt am Main, Germany), anti-HA (Clone 16B12/HA.11, Babco, Freiburg, Germany), anti-b-tubulin (Boehringer, Mannheim, Germany), Oncogene

anti-DNA-topoisomerase a, b, (StressGen, Victoria, Canada). Secondary antibodies used were: goat anti-rabbit IgG and goat anti-mouse IgG (Promega, Mannheim, Germany), sheep antimouse IgG (Amersham Pharmacia Biotech, Braunschweig, Germany) and goat anti-mouse IgG2b human adsorbed (Southern Biotechnologies, AL, USA), all conjugated to horseradish peroxidase. Antigen–antibody complexes were detected by ECL (Amersham Pharmacia Biotech, Braunschweig, Germany).

Coimmunoprecipitation NIH 3T3 were transfected with plasmids pCHDM1A expressing human Mdm2. At 20–24 h post-transfection, cells were stimulated for times indicated followed by cells lysis (lysis buffer: 20 mm Tris-HCl, pH 7.9, 5 mm MgCl2, 0.2 mm EDTA, 20% glycerol, 150 mm NaCl, 1 complete protease inhibitors (Roche, Mannheim, Germany), 1 mm NaF, 1 mm orthovanadate, 0.5% NP-40) on ice for 30 min. Lysates were centrifuged at 100 000 g for 45 min at 41C and equal protein amounts of each sample were incubated with agarose-bound anti-Mdm2 antibodies (SMP14, Santa Cruz, Heidelberg, Germany) at 41C overnight. Beads were washed three times for 20 min with lysis buffer (see above), boiled in SDS-loading buffer and subjected to 7.5% SDS–PAGE. After blotting on to nitrocellulose, filters were incubated with antibodies of interest.

Detection of ubiquitinated p53 For details, see Treier et al. (1994). Briefly, NIH 3T3 cells (150 mm dish) were transfected with plasmids expressing HAp53, His-ubiquitin (pMT 107) and Mdm2 (pCHDM1A). Subsequently, cells were harvested and lysed in nondenaturing buffer (50 mm Tris, pH 8.0, 5 mm EDTA, 150 mm NaCl, 0.5% NP-40, and 1 mm PMSF) to analyse inputs, or in denaturing lysis buffer (6 m guanidinium–HCl, 0.1 m Na2HPO4/NaH2PO4, pH 8.0) for binding to Ni-NTA beads (Quiagen, Hilden, Germany), which was carried out at 41C over night. Beads were washed with high stringency, proteins were precipitated, separated by 7.5% SDS–PAGE, blotted and ubiquitinated p53 was detected by incubation with anti-HA antibodies.

Cell fractioning 106 RKO cells were harvested, washed and resuspended in 100 ml hypotonic buffer (10 mm HEPES, pH 7.9, 30 mm NaCl, 0.1 mm EDTA, 1.5 mm MgCl2, 0.5 mm PMSF, 1 complete protease inhibitors, 1 mm NaF, 1 mm orthovanadate, 0.05% NP-40). Cell lysis was controlled by a microscope and nuclei were spun down at 400 g, 41C, 4 min. Supernatants were transferred to fresh tubes and centrifuged at 18 000 g for 10 min at 41C to recover the cytosolic fraction. Nuclei were washed three times with 300 ml of 50 mm Tris-HCl, pH 8.3, 5 mm MgCl2, 0.1 mm EDTA, 40% glycerol and lysed in 10 mm HEPES, pH 7.9, 350 mm NaCl, 0.1 mm EDTA, 1.5 mm MgCl2, 0.5 mm PMSF, 1 complete protease inhibitors, 1 mm NaF, 1 mm orthovanadate, 1.5% Triton X-100 at 41C for 30 min. Nuclear lysates were centrifuged (41C, 10 min, 18 000 g) to obtain nuclear fractions.

Statistics Experiments were conducted at least three times. Figures show representative results.


2867 Abbreviations NO, nitric oxide; GSNO, S-nitrosoglutathione; SpNO, spermine-NONOate; DAPI, 40 ,6-diamino-2-phenylindole; CRM1, chromosome region maintenance 1; Mdm2, murine double minute; NES, nuclear export sequence. Acknowledgements We thank M Decker and B Rogge for technical assistance. Thanks go to B Vogelstein for providing p53 reporter

plasmids, D Bohmann for providing the His-ubiquitin expression plasmid, C Jost for providing the HA-p53 construct and K. Vousden as well as J Chen for providing the Mdm2 construct and M Scheffner for providing p53 D62–96 and p53 D62–96 S>A plasmids. We also thank M Yoshida for his generous gift of LMB. This work was supported by the Deutsche Forschungsgemeinschaft (BR999), Wilhelm SanderFoundation and Boehringer Ingelheim Fonds for a travel award to NS.

References Barak Y, Juven T, Haffner R and Oren M. (1993). EMBO J., 12, 461–468. Berendji D, Kolb-Bachofen V, Meyer KL, Grapenthin O, Weber H, Wahn V and Kroncke KD. (1997). FEBS Lett., 405, 37–41. Boyd SD, Tsai KY and Jacks T. (2000). Nat. Cell Biol., 2, 563– 568. Bru¨ne B, von Knethen A and Sandau KB. (1999). Cell Death Differ., 6, 969–975. Bru¨ne B, von Knethen A and Sandau KB. (2001). Cell. Signal., 13, 25–33. Callsen D and Bru¨ne B. (1999). Biochemistry, 38, 2279–2286. Chen J, Lin J and Levine AJ. (1995). Mol. Med., 1, 142–152. Chen J, Wu X, Lin J and Levine AJ. (1996). Mol. Cell. Biol., 16, 2445–2452. Davis DW, Weidner DA, Holian A and McConkey DJ. (2000). J. Exp. Med., 192, 857–869. Dumaz N and Meek DW. (1999). EMBO J., 18, 7002–7010. El-Deiry WS, Tokino T, Velculescu VE, Levy DB, Parsons R, Trent JM, Lin D, Mercer WE, Kinzler KW and Vogelstein B. (1993). Cell, 75, 817–825. Fornerod M, Ohno M, Yoshida M and Mattaj IW. (1997). Cell, 90, 1051–1060. Forrester K, Ambs S, Lupold SE, Kapust RB, Spillare EA, Weinberg WC, Felley-Bosco E, Wang XW, Geller DA, Tzeng E, Billiar TR and Harris CC. (1996). Proc. Natl. Acad. Sci. USA, 93, 2442–2447. Fuchs SY, Adler V, Buschmann T, Wu X and Ronai Z. (1998). Oncogene, 17, 2543–2547. Fukuda M, Asano S, Nakamura T, Adachi M, Yoshida M, Yanagida M and Nishida E. (1997). Nature, 390, 308–311. Geyer RK, Yu ZK and Maki CG. (2000). Nat. Cell Biol., 2, 569–573. Gottifredi V and Prives C. (2001). Science, 292, 1851–1852. Gu W and Roeder RG. (1997). Cell, 90, 595–606. Hart TW. (1985). Tetrahedron Lett., 26, 2013–2016. Haupt Y, Maya R, Kazaz A and Oren M. (1997). Nature, 387, 296–299. Hengstermann A, Whitaker NJ, Zimmer D, Zentgraf H and Scheffher M. (1998). Oncogene, 17, 2933–2941. Hirao A, Kong YY, Matsuoka S, Wakeham A, Ruland J, Yoshida H, Liu D, Elledge SJ and Mak TW (2000). Science, 287, 1824–1827. Honda R, Tanaka H and Yasuda H. (1997). FEBS Lett., 420, 25–27. Kern SE, Pietenpol JA, Thiagalingam S, Seymour A, Kinzler KW and Vogelstein B. (1992). Science, 256, 827–830. Kim SJ, Hwang SG, Shin DY, Kang SS and Chan JS. (2002). J. Biol. Chem., 277, 33501–33508. Kroncke KD, Fehsel K, Schmidt T, Zenke FT, Dasting I, Wesener JR, Bettermann H, Breunig KD and Kolb-

Bachofen V. (1994). Biochem. Biophys. Res. Commun., 200, 1105–1110. Kubbutat MH, Jones SN and Vousden KH. (1997). Nature, 387, 299–303. Kubbutat MH, Ludwig RL, Ashcroft M and Vousden KH. (1998). Mol. Cell. Biol., 18, 5690–5698. Kudo N, Wolff B, Sekimoto T, Schreiner EP, Yoneda Y, Yanagida M, Horinouchi S and Yoshida M. (1998). Exp. Cell Res., 242, 540–547. Liang and Clarke MF. (2001). Eur. J. Biochem., 268, 2779–2783. Lopes UG, Erhardt P, Yao R and Cooper GM. (1997). J. Biol. Chem., 272, 12893–12896. Maki CG, Huibregtse JM and Howley PM. (1996). Cancer Res., 56, 2649–2654. Maki CG. (1999). J. Biol. Chem., 274, 16531–16535. Marin MC, Jost CA, Irwin MS, DeCaprio JA, Caput D and Kaelin Jr WG. (1998). Mol. Cell. Biol., 18, 6316–6324. Matsuoka M and Igisu H. (2001) Biochem. Biophys. Res. Commun., 282, 1120–1125. Meek DW. (1998). Cell. Signal., 10, 159–166. Messmer UK, Ankarcrona M, Nicotera P and Bru¨ne B. (1994). FEBS Lett., 355, 23–26. Messmer UK and Bru¨ne B. (1996). Biochem. J., 319, 299–305. Miyashita T and Reed JC. (1995). Cell, 80, 293–299. Nakaya N, Lowe SW, Taya Y, Chenchik A and Enikolopov G. (2000). Oncogene, 19, 6369–6375. Oliner JD, Kinzler KW, Meltzer PS, George DL and Vogelstein B. (1992). Nature, 358, 80–83. Ossareh-Nazari B, Bachelerie F and Dargemont C. (1997). Science, 278, 141–144. Pise-Masison CA, Radonovich M, Sakaguchi K, Appella E and Brady JN. (1998). J. Virol., 72, 6348–6355. Roth J, Dobbelstein M, Freedman DA, Shenk T and Levine AJ. (1998). EMBO J., 17, 554–564. Shaw P, Freeman J, Bovey R and Iggo R. (1996). Oncogene, 12, 921–930. Shieh SY, Ikeda M, Taya Y and Prives C. (1997). Cell, 91, 325–334. Siliciano JD, Canman CE, Taya Y, Sakaguchi K, Appella E and Kastan MB. (1997). Genes Dev., 11, 3471–3481. Stamler JS, Simon DI, Osborne JA, Mullins ME, Jaraki O, Michel T, Singel DJ and Loscalzo J. (1992). Proc. Natl. Acad. Sci. USA, 89, 444–448. Stommel JM, Marchenko ND, Jimenez GS, Moll UM, Hope TJ and Wahl GM. (1999). EMBO J., 18, 1660–1672. Tao W and Levine AJ. (1999a). Proc. Natl. Acad. Sci. USA, 96, 3077–3080. Tao W and Levine AJ. (1999b). Proc. Natl. Acad. Sci. USA, 96, 6937–6941. Treier M, Staszewski LM and Bohmann D. (1994). Cell, 78, 787–798. Oncogene


2868 Turenne GA, Paul P, Laflair L and Price BD. (2001). Oncogene, 20, 5100–5110. Wang X, Michael D, de Murcia G and Oren M. (2002). J. Biol. Chem., 277, 15697–15702. Wu X, Bayle JH, Olson D and Levine AJ. (1993). Genes Dev., 7, 1126–1132.

Oncogene

Yang F, von Knethen A and Bru¨ne B. (2000). J. Leukoc. Biol., 68, 916–922. Zhang Y and Xiong Y. (2001a). Cell Growth Differ., 12, 175–186. Zhang Y and Xiong Y. (2001b). Science, 292, 1910–1915.