Modulation of P53 C-Terminal Acetylation by Mdm2, P14 , And

Published OnlineFirst February 15, 2013; DOI: 10.1158/1535-7163.MCT-12-0904

Molecular Cancer Cancer Therapeutics Insights Therapeutics See related article by McCarthy et al., p. 352

Modulation of p53 C-Terminal Acetylation by mdm2, p14ARF, and Cytoplasmic SirT2

Ingeborg M.M. van Leeuwen1, Maureen Higgins2, Johanna Campbell2, Anna R. McCarthy1, Marijke C.C. Sachweh1, Ana Marín Navarro1, and Sonia Laín1,2

Abstract Acetylation of C-terminal lysine residues in the p53 tumor suppressor is associated with increased stability and transcription factor activity. The function, protein level, and acetylation of p53 are downregulated by mdm2, which in its turn is inhibited by the p14ARF tumor suppressor. Here, we show that p14ARF increases the level of p53 acetylated at lysine 382 in a nuclear chromatin-rich fraction. Unexpectedly, this accumulation of p53AcK382 is dramatically enhanced in the presence of ectopic mdm2. In light of these observations, we propose that p14ARF increases the binding of p53–mdm2 complexes to chromatin, thereby limiting the access of protein deacetylases to p53. Supporting this notion, we show that p53AcK382 can be deacetylated in the cytoplasm and that sirtuin SirT2 catalyzes this reaction. These results help understand why inhibition of both SirT1 and SirT2 is needed to achieve effective activation of p53 by small-molecule sirtuin inhibitors. Mol Cancer Ther; 12(4); 1–10. 2013 AACR.

Introduction mdm2 and subsequent proteasomal degradation (4). p53 is subjected to a variety of posttranslational SirT2 is primarily cytoplasmic, and it has been shown to modifications that modulate its function as a tumor reduce p53 acetylation at K382 in adenovirus-positive suppressor (1). Acetylation of p53 at C-terminal lysine HEK293 cells (6), where the adenoviral E1B55KD protein residues reduces binding to mdm2 and prevents ubi- is known to lead to the accumulation of large amounts quitination and subsequent degradation of p53 (2). In of p53 in a cytoplasmic juxtanuclear body (7). SirT2 has addition, C-terminal acetylation increases the affinity of also been implicated in regulating p53 in a study using p53 for DNA (3). Thus, p53 acetylation at the C-termi- HeLa cells (8), which are infected with human papillo- nus not only leads to stabilization but also promotes the mavirus (HPV). In HeLa cells, p53 degradation is gov- function of p53 as a transcription factor. Acetylation of erned by the action of the HPV protein E6 instead of p53 is carried out by several protein acetyltransferases mdm2. To date, there is no data about a possible effect including p300 and CREB-binding protein (CBP), of SirT2 on p53 acetylation in a non–virus-infected back- whereas deacetylation is catalyzed by protein deacety- ground. A recent study has reported that p53 is a substrate lases such as class I and II histone deacetylases (HDAC) for mammalian SirT3 in the mitochondria (9). Finally, and SirT1 of the sirtuin family of class III HDACs (4). SirT7, a nuclear sirtuin preferentially localized in the Sirtuins possess a high degree of homology in their nucleolus, has been shown to deacetylate p53 in vitro central regions (5) and catalyze deacetylation of lysine (10). The remaining members of the sirtuin family, SirT4, residues or ADP ribosylation reactions. Their activity is SirT5, and SirT6, have not been implicated in regulating þ dependent on NAD , a property that links sirtuin func- p53 acetylation (reviewed in ref. 11). tion to metabolic status. SirT1, which is primarily located mdm2 is the major negative regulator of the function of in the nucleus, deacetylates p53 at lysine 382, thereby p53 and therefore is regarded as an attractive target for enabling its ubiquitination by ubiquitin ligases such as cancer therapy (12). mdm2 binds directly to p53, obstruct- ing the transactivation domain of p53 and interferes with the function of p53 as a transcription factor (13). In addi- Authors' Affiliations: Department of 1Microbiology, Tumor and Cell tion, mdm2 contains a RING finger domain that catalyzes Biology, Karolinska Institutet, Stockholm, Sweden; and 2Centre for Oncol- ogy and Molecular Medicine, University of Dundee, Ninewells Hospital and the ubiquitination of p53 at lysine residues (14). This Medical School, Dundee, Scotland, United Kingdom ubiquitination leads to export of p53 from the nuclear Note: Supplementary data for this article are available at Molecular Cancer compartment and subsequent degradation. In addition, Therapeutics Online (http://mct.aacrjournals.org/). mdm2 has been reported to inhibit acetylation of p53 by Corresponding Author: Sonia Laín, Microbiology, Tumor and Cell Biology, protein acetyltransferases (15–17) as well as to promote Karolinska Institutet, Nobels vag€ 16, Stockholm 171 77, Sweden. Phone: deacetylation by HDACs (18). 46-8-52484603; Fax: 46-8-304-276; E-mail: [email protected] The p14ARF tumor suppressor is a small protein that doi: 10.1158/1535-7163.MCT-12-0904 binds directly to the acidic domain of mdm2. It constitutes 2013 American Association for Cancer Research. the major inhibitor of mdm2 and is also a very effective

www.aacrjournals.org OF1

van Leeuwen et al.

activator of the transcription factor function of p53 (DMSO). The following antibodies were used: p53 [DO1, (reviewed in ref. 19). p14ARF is very rich in arginine mouse monoclonal (ref. 32), and CM-1, rabbit polyclonal residues and has an isoelectric point of 12.91, which is (ref. 33)], p53 acetylated at K382 (BioLegend #614202, higher than the 11.54 of histone H3. There are several rabbit polyclonal, and AbCam #ab75754, rabbit mono- possible models attempting to explain the mechanism by clonal), mdm2 [4B2, mouse monoclonal (34)], p14ARF which p14ARF has such a strong influence on both p53 (Santa Cruz #53392, mouse monoclonal), b-galactosidase stability and activity. p14ARF has been proposed to seques- (AbCam #ab616, rabbit polyclonal), a-tubulin (Sigma ter mdm2 in the nucleolar compartment, inhibit export of #T6199, mouse monoclonal), 20S proteasome subunit p53 from the nucleus, and/or downregulate p53 ubiqui- a-6 (Affiniti #PW8100, mouse monoclonal), histone H3 tination by mdm2 (20–22). These models, although not (AbCam #ab1791, rabbit polyclonal), histone H3 acetylat- incompatible, have fuelled considerable debate since their ed at K9 (AbCam #ab12179, mouse monoclonal), histone proposal more than a decade ago (23, 24). H4 (Upstate, #07-108, rabbit polyclonal), SirT2 (AbCam Here, we present new studies on the effects of mdm2 #ab51023, rabbit monoclonal), and paxillin (AbCam and p14ARF on the stability, localization, ubiquitination, #ab32084, rabbit monoclonal). and acetylation of p53. We found that expression of ectopic mdm2 caused an increase in p53 acetylation at Transient transfections and plasmids lysine 382 when ectopic p14ARF was also present. We Cells were seeded on 10-cm tissue culture dishes and propose that mdm2–p14ARF complexes sequester p53 transfected using the calcium phosphate precipitation in a nuclear chromatin-rich fraction. As a consequence, method, as described previously (21). At least 1 hour p53 is protected from the action of protein deacetylases before transfection, the medium was replaced by fresh that are not linked to chromatin, such as cytoplasmic high-glucose DMEM supplemented with 10% FBS only. deacetylases. Supporting this model, we show that For H1299 cells, the medium was changed back to RPMI- p53KAc382 is a substrate for the cytoplasmic deacetylase based medium 18 hours after transfection. Cells were SirT2. This is the first time that p53 deacetylation in harvested for Western blot analysis (see below) 36 hours the cytoplasm has been shown. These findings provide posttransfection. The following plasmids were used: an insight into why inhibition of both SirT1 and SirT2 pcDNA3 constructs expressing p14ARF (kind gift from seems to be required to achieve effective p53 activation K.H. Vousden, The Beatson Institute for Cancer Research, with small-molecule sirtuin inhibitors (25). Finally, our Glasgow, United Kingdom), human wild-type p53, p53- observations are in line with previously suggested tu- R273H, p53-K320T, p53-6KR, SirT1-H363Y, and SirT2- mor suppressor roles for mdm2 (26), and the recently H187Y; pCMV constructs expressing human mdm2 described positive effect of mdm2 on p53-dependent (hdm2, kind gift from A.J. Levine, Institute for Advanced transcription when the E3 ubiquitin ligase function of Study, Princeton, NJ) and hdm2-C464A (21); and pCOC- mdm2 is impaired (27). X2 constructs expressing mdm2 (kind gift from M. Oren, The Weizmann Institute of Science, Rehovot, Israel) and Materials and Methods mdm2 N-terminal fragments (residues 1–244, 1–258, and Cells, reagents, and antibodies 1–342; ref. 35). The vector expressing hdm2EVEAAA MCF7 cells [purchased from American Type Culture mutant was generated by standard site-directed muta- Collection (ATCC)] and ARN8 cells (28) were grown in genesis of the pCMV-hdm2 plasmid. Vectors expressing high-glucose Dulbecco’s Modified Eagle’s Medium HAUSPcys and His6-tagged ubiquitin were kind gifts (DMEM) supplemented with FBS and penicillin/strepto- from D.P. Xirodimas (The University of Dundee, Dundee, mycin. H1299 cells (purchased from ATCC) were culti- UK) and S. Mittnacht (University College of London, vated in RPMI supplemented in the same manner. H1299- London, UK), respectively. BSK and pcDNA3 empty pcDNA3 control cells and H1299-SirT2–overexpressing vectors were used to equalize the total amount of DNA. cells are described previously (29). Cells were grown at b-Galactosidase was used as a control for transfection, 37 C and 5% CO2 in a humidified atmosphere. The viability, and loading efficiency. authors did not authenticate the cell lines for the present study. However, H1299s are p53-null and express high Nuclear extraction levels of endogenous p14ARF, whereas MCF7 cells bear Cell cultures with washed twice with PBS and then the wild-type p53 that can be activated by nutlin-3 treatment. cells were scraped into 1 mL of PBS and harvested by These features were checked on a regular basis, and cell centrifugation. The cells were resuspended in homogeni- lines were also routinely tested for mycoplasma (MycoA- zation buffer (10 mmol/L Tris-HCl, pH 7.5, 10 mmol/L lert Detection Kit, Lonza). Leptomycin B (LMB) was NaCl, and 1.5 mmol/L MgCl2 supplemented with prote- obtained from Cancer Research UK and stored at 80C ase inhibitor cocktail) and lysed in a Dounce homogenizer as a 10 mmol/L solution in absolute ethanol. Etoposide, fitted with a loose pestle. Samples were centrifuged at nutlin-3 (racemic), EX527, and MG132 were purchased 16,100 g, and the pellets (nuclear chromatin-rich, Nch) from Sigma. Tenovins (tnv1, tnv6, and tnvD3) were syn- and the supernatants (cytoplasmic soluble, Cs) transferred thesized as described previously (30, 31) and stored at to separate tubes. The Nch fractions were washed in 20C as 40 mmol/L solutions in dimethyl sulfoxide homogenization buffer. Samples for Western blot analysis

OF2 Mol Cancer Ther; 12(4) April 2013 Molecular Cancer Therapeutics

p53 C-Terminal Acetylation in the Nucleus and Cytoplasm

(see below) were prepared with 1 SDS sample buffer nant SirT2 prepared as described and components from (Invitrogen). the Fluor de Lys Fluorescent Assay System (Enzo Life Sciences kits BML-AK555 and BML-AK556). SirT2 was Analysis of cell extracts by Western blotting used at a final concentration of 0.01 mg/mL and the Western blot analysis was carried out as described FdL-SirT1/T2 substrates at 15 mmol/L. The FdL-SirT1 previously (36). Dithiothreitol (Sigma #D06032) was substrate is a peptide consisting of residues 379–382 of added to the samples to a final concentration of 100 human p53 [Arg-His-Lys-Lys(Ac)] whereas the FdL-SirT2 mmol/L. Proteins were separated on NuPAGE 4%–12% peptide is composed of residues 317–320 of human p53 Bis–Tris gels (Invitrogen) in 1 NuPAGE MOPS buffer [Gln-Pro-Lys-Lys(Ac)]. (Invitrogen #NP000102) supplemented with NuPAGE Assay using membrane-bound acetylated p53. Whole- antioxidant (Invitrogen #NP0005). The mobility of the cell extracts were separated by electrophoresis using proteins of interest was determined using molecular 4%–12% Bis–Tris gels (Invitrogen) and transferred to weight markers (SeeBlue Plus2, Invitrogen #LC5925). PVDF membranes (Millipore). Membrane pieces con- Protein transfer to polyvinylidene difluoride (PVDF) taining the protein substrates of interest were cut and membranes (Millipore #IPVH00010) was done with 1 incubated in 10 Complete EDTA-free protease inhib- NuPAGE transfer buffer (Invitrogen #NP0006). Following itor (Roche), 527 mmol/L Tris pH 7.5, 5 mmol/L dithio- incubation with primary and horseradish peroxidase- threitol, 750 mmol/L NaCl in the absence or presence þ conjugated secondary antibodies, chemiluminiscent sig- of SirT2 (0.023 mg/mL), and 5 mmol/L NAD for nals were detected with ECL solution (Amersham Bios- 90 minutes at 37C. Relevant antigens were detected ciences). Adobe Photoshop was used to adjust the by chemiluminescence following incubation with pri- brightness and contrast of scanned blot images. All lanes mary and horseradish peroxidase-conjugated second- were treated in the same manner. ary antibodies.

Immunofluorescence staining and cell imaging Cells were seeded on 2-well glass slide chambers Results (Nunc), incubated for 24 hours, and then treated as indi- mdm2 increases p53 C-terminal acetylation in the cated. After rinsing twice with PBS, cells were fixed with presence of p14ARF ice-cold methanol–acetone (1:1), rehydrated in serum-free According to the most accepted model, p53 is mono- DMEM, and then incubated with anti-p53 (DO1 diluted ubiquitinated by mdm2 in the nuclear compartment in serum-free DMEM) antibody for 1 hour at room tem- andthenexportedtothecytoplasmwhereitis perature. Thereafter, cells were washed 4 times and incu- polyubiquitinated and subsequentlydegradedbythe bated with secondary fluorescently labeled antibody proteasome (37). However, some degree of p53 poly- (FITC donkey anti-mouse IgG, Jackson ImmunoResearch ubiquitination and degradation can also occur in the Laboratories) in the dark for 45 minutes at room temper- nuclear compartment (38). We conducted cytoplasmic ature. Cells were counterstained with Hoechst (Sigma and nuclear fractionations based on the classic Dignam #B1155) for 1 minute, and then the slides were mounted method (39), which separates cytoplasmic proteins such using Hydromount (National Diagnostics #HS106). Cells as a-tubulin and paxillin from chromatin-bound pro- were visualized using an Axiovert 200M microscope from teins such histones H3 and H4. Using this method, we Zeiss powered with the Volocity software from Improvi- observed that p53, as well as its ubiquitinated forms, sion and a Hamamatsu Orca ER camera. accumulates primarily in the cytoplasmic compartment upon a short treatment with the proteasome inhibitor SirT2 expression and purification MG132 (Fig. 1A and B). This suggests that the degra- SirT2 cDNA (residues 34–356) was cloned into the dation of ubiquitinated forms of p53 occurs mainly in PGEX-6P-1 vector (GE Life Sciences) and overexpressed the cytoplasm. However, we detected a substantial in BL21(DE3)T1R bacterial cells (Sigma) as a gluta- leakage of nuclear p53, and in particular ubiquitinated thione S-transferase (GST) fusion protein. Purification p53, during sample preparation, as evidenced in Fig. of the GST-SirT2 was achieved using glutathione 1C. Therefore, we conclude that the Dignam extraction sepharose (GE LifeSciences) followed by cleavage of method separates a fraction containing mainly cyto- the fusion protein using PreScission protease (GE Life- plasmic p53, but also soluble nuclear p53, and a nuclear Sciences). SirT2 was separated from GST via glutathione chromatin-enriched fraction containing insoluble p53 sepharose chromatography (GSTrap 4B, GE Life- that is likely to be tightly bound to nucleic acids. We Sciences) and the SirT2 containing fractions pooled and named these fractions Cs and Nch, respectively. In further purified via gel filtration chromatography agreement with this interpretation, the nuclear export (HiLoad 16/60 Superdex 75, GE Healthcare). inhibitor and p53 activator LMB (40) has no effect on the Cs and Nch distribution of p53 obtained by the Dignam Sirtuin activity assays method(Fig.1D).However,amongtheextraction Fluorescence-based assay. Deacetylation assays were methods tested, the Dignam protocol provided the best conducted as described previously (29) using recombi- separation.

www.aacrjournals.org Mol Cancer Ther; 12(4) April 2013 OF3

van Leeuwen et al.

Figure 1. Localization of ubiquitin-conjugated and unconjugated forms of p53. Cells were transiently transfected with plasmids expressing p53, p53-R273H, human mdm2 (hdm2), His6-tagged ubiquitin, and bacterial b-galactosidase. Cytoplasmic (Cs) and nuclear (Nch) fractions were separated as described in Materials and Methods, and protein levels analyzed by Western blotting. MCF7 cells (A) and H1299 cells (B) were treated with 20 mmol/L MG132 for 1 hour before harvesting or left untreated. Blots were probed with antibodies against p53, b-galactosidase, histone H3, and alpha-6 ARF proteasomal subunit. C, the ubiquitinated forms of p53 become strongly linked to the Nch fraction in the presence of hdm2 and p14 . H1299 cells were transfected with plasmids expressing p53-R273H and hdm2 or with empty control vectors. Cs and Nch fractions from transfected cells were mixed as indicated and, thereafter, the resulting samples were subjected to a new centrifugation to separate the Cs and Nch fractions again. D, H1299 cells expressing ectopic p53-R273H were also transfected with plasmids expressing hdm2 wild-type or hdm2-EVEAAA (residues 250–252inhdm2acidic domain substituted by alanine, which prevents polyubiquitination and degradation of p53). Where indicated, cells were also transfected with plasmids expressing p14ARF and a dominant-negative HAUSP construct (HAUSPcys) or treated with 20 nmol/L LMB for 3 hours.

The p14ARF tumor suppressor is a highly positively expressing human mdm2 together with p14ARF was much charged protein with a high afﬁnity for chromatin. As higher than the levels observed in counterparts expres- can be seen in Fig. 1D, in the presence of ectopic p14ARF, sing p14ARF alone (Fig. 2A). The large increase most p53 forms accumulate in the Nch compartment, but in p53AcK382 induced by combining ectopic mdm2 and this does not occur when a dominant-negative mutant p14ARF occurred independently of the transcriptional of the HAUSP p53 deubiquitinase (41) is added to the activity of p53 (Fig. 2B). system. Furthermore, leakage of ubiquitin-conjugated p53 A strong joint effect of p14ARF and mdm2 on the level of from the Nch fraction is substantially diminished in the p53 acetylation also occurs with the hdm2–C464A mutant presence of p14ARF (Fig. 1C). (Fig. 2A), indicating that ubiquitin and nedd8 ligase activ- When an antibody against p53 acetylated at lysine 382 ities of mdm2 are dispensable in this context. The hdm2– was used, we observed that the expression of ectopic C464A construct has a critical mutation in the RING ﬁnger p14ARF leads to a slight increase in the levels of p53AcK382 domain of mdm2 that impairs ubiquitin/nedd8 conjuga- in the Nch fraction (Fig. 2A and C). In addition, we noted tion. However, the regions of mdm2 that interact with ARF that the level of p53AcK382 in Nch fractions from cells p14 are essential, as shown by the weak effect of the

OF4 Mol Cancer Ther; 12(4) April 2013 Molecular Cancer Therapeutics

p53 C-Terminal Acetylation in the Nucleus and Cytoplasm

Figure 2. Effect of p14ARF on the levels of p53 acetylated at lysine 382. H1299 cells were transfected with plasmids expressing p53, human mdm2 (hdm2), ARF p14 , and b-galactosidase. Cytoplasmic (Cs) and nuclear (Nch) fractions were separated as described in Materials and Methods, and protein levels analyzed by Western blotting. A, comparison of the effect of p14ARF on acetylated p53 in the presence or absence of hdm2. The hdm2-C464A construct is defective for both ubiquitination and neddylation. B, effect of p14ARF on the level of acetylation of mutant p53. The p53-R273H mutant is transcriptionally inactive. The p53-6KR construct, where the 6 C-terminal lysines have been mutated to arginine, serves to assess the speciﬁcity of the antibody against p53AcK382. C, cells were transfected with plasmids expressing 3 different mdm2 deletion mutants: D11 (residues 1–342), D18 (residues 1–258), or D4 (residues 1–244). The D11 and the D18 mdm2 deletion mutants, but not the D4 mutant, bind to p14ARF (35). D, cells were transfected as indicated and incubated with 20 mmol/L MG132 for 1 hour before harvesting or left untreated. a-Tubulin and paxillin were used as cytoplasmic markers and histone H3 as a nuclear marker. deletion mutant mdm2-D4 on the p14ARF-mediated Deacetylation of p53 by SirT2 increase in the level of acetylated p53 in the nucleus (Fig. According to Fig. 2, p14ARF causes accumulation of D 2C). Unlike mdm2 and hdm2-C464A, mdm2- 4doesnot acetylated p53 in the Nch fraction and this is enhanced bind to p14ARF (35). In addition, as can be seen in Fig. 2D, by mdm2. However, on the basis of existing experimental the action of p14ARF is not substantially enhanced by data, mdm2 is unlikely to have a direct positive effect MG132. These observations indicate that the impact of on the acetylation of p53 at the biochemical level even p14ARF on p53 acetylation is not solely due to stabiliza- in the presence of p14ARF. For instance, it has been tion of p53 and requires binding of p14ARF to mdm2. shown that mdm2 inhibits p300-mediated p53 acetylation To conﬁrm that the Nch fraction does indeed contain by forming a ternary complex with these 2 proteins mainly chromatin and chromatin-bound proteins, we (15) and that mdm2–HDAC1 complexes enhance p53 carried out a standard chromatin extraction. From this deacetylation (18). experiment (Supplementary Fig. S1A), it can be seen that As mentioned in the introduction, SirT1 is an important the amount of acetylated p53 bound to chromatin is p53 deacetylase. However, as shown in Supplementary strongly increased by p14ARF and mdm2 co-expression. Figs. S1B and S2, results obtained by overexpressing SirT1 This is consistent with the results obtained by analyzing or a dominant-negative mutant SirT1 suggest that SirT1 the Cs and Nch fractions. is not the only factor modulating p53 deacetylation in

www.aacrjournals.org Mol Cancer Ther; 12(4) April 2013 OF5

van Leeuwen et al.

Figure 3. Identification of p53AcK382 as a SirT2 substrate. A, ARN8 cells were treated with 1 mmol/L EX527 and 20 mmol/L etoposide for 6 hours to raise the levels of acetylated p53. Whole-cell extracts were separated by electrophoresis and transferred to PVDF membranes. Membrane pieces containing the þ protein substrates of interest were cut and incubated in the absence or presence of SirT2 (0.023 mg/mL) and 5 mmol/L NAD for 90 minutes at 37C. B, SirT2 deacetylates both SirT1 and SirT2 Fluor de Lys (FdL) fluorogenic substrates derived from acetylated lysine residues within p53. The FdL p53- 382 and p53-320 peptides contain acetylated residues K382 and K320, respectively. Recombinant SirT2 was incubated with 15 mmol/L of each substrate and 1 mmol/L NADþ and deacetylation detected by measuring the fluorescence generated. C, H1299 were transiently transfected with plasmids expressing p53 (wild-type), SirT2, and b-galactosidase and incubated with the indicated doses of tenovin-6 for 6 hours. Whole-cell extracts were analyzed by Western blotting.

our model system. Therefore, to explain our data about D3 could increase the levels of p53 and p53AcK382 in the levels of acetylated p53 in the Nch fraction, we hypo- the Cs fraction. This was indeed the case (Fig. 4A and B) thesized that p14ARF, through its ability to retain and, as expected, this effect was particularly visible when p53–mdm2 complexes in the nucleus, could be protecting a p53 construct with mutations in its nuclear localization p53 from deacetylation by cytoplasmic deacetylases. signal NLS-I was used (p53-K320T, Fig. 4C). The NLS-I Thus, we set out to test whether SirT2, a deacetylase domain is a key for the migration of p53 into the nucleus known to be located primarily in the cytoplasmic com- (43). Furthermore, the increase in p53AcK382 in the Cs partment (42), had any effect on p53 acetylation. fraction in response to tenovin-D3 is reduced in cells There is previous evidence that SirT2 can affect p53 overexpressing SirT2 (Fig. 4C). In contrast, tenovin-1 and acetylation (6, 8). However, these studies were conducted -6 lead to an increase in both nuclear and cytoplasmic in cell lines in which viral proteins govern the degrada- p53AcK382 (Fig. 4D). tion of p53. Furthermore, these studies did not establish In agreement with previous reports, in our system, whether acetylated p53 is a substrate for SirT2. Here, we SirT2 is mainly cytoplasmic (Fig. 4A and C), whereas show that SirT2 can deacetylate full-length p53 directly SirT1 is primarily nuclear (Supplementary Fig. S3). SirT2 in a biochemical assay. This can be seen in Fig. 3A, where is actively exported from the nucleus in a Crm1-depen- membrane-bound acetylated p53 was incubated in the dent manner and, following LMB treatment, SirT2 accu- þ presence and absence of puriﬁed SirT2 and NAD .In mulates rapidly in the nucleus (44). However, in our addition, SirT2 deacetylates a p53 peptide containing experiments, we were not able to detect an effect of LMB acetylated K382 (Fig. 3B). on the distribution of SirT2 or p53 (Fig. 4C). This is Tenovin-6 is a small-molecule inhibitor of both SirT1 possibly due to leakage of non–chromatin-bound proteins and SirT2 (29) that can increase the levels of p53 acetylat- from the Nch fraction during the extraction procedure. ed at K382 as shown in Fig. 3C. Supporting that inhibition The data presented in Figs. 3 and 4 indicate that SirT2 of SirT2 by tenovin-6 contributes to this effect, overex- is involved in the deacetylation of p53 in the cyto- pression of SirT2 reduces p53 acetylation level (Fig. 3C). plasmic compartment. Further supporting this postu- Tenovin-D3 is a novel tenovin analogue that shows selec- lation, we observed that in cells stably transfected with tivity for SirT2 (31). As expected for a SirT2 inhibitor, a plasmid expressing SirT2 the levels of p53AcK382 a tenovin-D3 increases the levels of acetylated-K40 -tubu- in the Cs fraction are reduced while the levels of lin (31). However, unlike tenovin-6, tenovin-D3 has p53AcK382 in the Nch compartment, where SirT2 levels reduced activity against SirT1 and is very poor at increas- are lower, remain unaffected by SirT2 expression (Fig. ing p53 levels (31). Here, we examined whether tenovin- 5A). Accordingly, expression of the deacetylase-

OF6 Mol Cancer Ther; 12(4) April 2013 Molecular Cancer Therapeutics

p53 C-Terminal Acetylation in the Nucleus and Cytoplasm

Figure 4. Chemical SirT2 inhibition increases the levels of acetylated p53 in the cytoplasm. A, H1299 cells were transfected with plasmids expressing p53 and b-galactosidase and treated with the indicated doses of tenovin-D3 for 16 hours. Cytoplasmic (Cs) and nuclear (Nch) fractions were separated as described in Materials and Methods, and protein levels analyzed by Western blotting. B, p53 accumulates in the cytoplasm in response to tenovin-D3. MCF7 cells were preincubated with 20 mmol/Ltenovin-D3(orDMSO)for1hourandthen2mmol/L nutlin-3 was added for 5 hours. Nutlin-3 is a small molecule that selectively stabilizes p53 (54), and it was used here to raise p53 levels. p53 was detected by indirect immunofluorescence using the DO1 anti-p53 antibody and anti-mouse IgG-FITC. Nuclei were stained with Hoechst. C, H1299-pcDNA3 control cells and H1299 cells stably overexpressing SirT2 were transfected with p53-K320T (i.e., a p53 mutant bearing a defective nuclear localization signal NLS-I) and treated with 2nmol/LLMBor15mmol/L tenovin-D3 for 16 hours as indicated. Vertical lines between lanes indicate that they are from the same film but have been rearranged in the image. D, H1299 cells were transfected with plasmids expressing p53, BSK, and b-galactosidase and treated with 10 mmol/L tenovin-1 or tenovin-6 for 16 hours. defective SirT2-H187Y dominant-negative mutant led to downregulates p53 acetylation by protein acetyltrans- an increase in p53AcK382 (Fig. 5B). Possibly because of ferases (15–17) and favors deacetylation of p53 by HDACs thepresenceofsomeectopicSirT2-H187Yinthenucleus such as HDAC1 (18). Therefore, it is not unexpected that as a consequence of overexpression (Fig. 5B), we also there is an enhanced accumulation of p53 and p53AcK382 observed a slight increase in p53-dependent transcrip- in whole-cell extracts in the presence of mdm2 inhibitor tion in the presence of this SirT2 mutant (Fig. 5C). An p14ARF (16, 45). equivalentmutantofSirT1hadamuchstrongereffect According to published data, p14ARF expression (Fig. 5C). reverses the negative effect of mdm2 on p300-mediated acetylation of p53 (16). In addition, p14ARF activates p53 Discussion acetyltransferases such as p300, thus resulting in further Recent reports show that the acetylation status of p53 stabilization and acetylation of p53 (45). Here, we confirm plays a crucial role in the modulation of p53 levels and that p14ARF does indeed increase p53 acetylation. Corre- transcription factor activity (4). For example, acetylation spondingly, we show that this accumulation of of C-terminal lysines in p53 is associated with stability and p53AcK382 occurs in chromatin-rich fractions, which increased transcription factor activity of this tumor sup- explains further why p14ARF is a very efficient activator pressor. mdm2 impairs p53 transcription factor activity, of p53-dependent transcription. The p14ARF-mediated catalyzes p53 ubiquitination, promotes its export from the accumulation of p53 and acetylated p53 in chromatin nucleus, and targets it for degradation. In addition, mdm2 fractions confirms and completes previous studies,

www.aacrjournals.org Mol Cancer Ther; 12(4) April 2013 OF7

van Leeuwen et al.

Figure 5. Genetic manipulation of SirT2 activity affects the levels of acetylated p53. A, H1299 pcDNA3 control cells and H1299 cells stably overexpressing SirT2 were transfected with plasmids expressing p53 (wild-type or 6KR mutant). Cytoplasmic (Cs) and nuclear (Nch) fractions were separated as described in Materials and Methods, and protein levels analyzed by Western blotting. b-Galactosidase expression from a control vector was monitored as a control for transfection efﬁciency and viability. B, H1299-pcDNA3 control cells and H1299 cells stably overexpressing SirT2 were transfected with plasmids expressing a p53 mutated at the nuclear signal NLS-I (p53-K320T) and a deacetylase-inactive SirT2 construct (SirT2-H187Y). b-Galactosidase expression from a control vector was monitored as a control for transfection efﬁciency and viability. C, MCF7 cells were transfected with the RGC-DFos-LacZ p53- dependent reporter construct, plasmids expressing dominant-negative mutant sirtuins SirT1-H363Y and SirT2-H187Y, and a control reporter plasmid expressing luciferase under the control of a SV40 promoter. b-Galactosidase activity was measured 24 hours after transfection and values normalized using the luciferase readings. Error bars correspond to SD (n ¼ 4).

suggesting that p14ARF inhibits the export of p53 from the tin-enriched compartment in the presence p14ARF,an nucleus (20). However, it was striking that the levels of event that requires the presence of mdm2 to bridge the ARF p53AcK382 in the Nch fraction reached in the presence of interaction between p53 and p14 . As a consequence, both p14ARF and mdm2 were higher than those observed p14ARF–mdm2 complexes protect p53 from deacetylases, in the absence of mdm2. This observation was confirmed including cytoplasmic deacetylases such as SirT2. Our by 2 different chromatin extraction methods (Fig. 2A and vision of the modulation of p53 C-terminal acetylation Supplementary Fig. S1A). Hence, contrary to expecta- by mdm2, p14ARF, and SirT2 is illustrated in Fig. 6. tions, our results show that the effect of p14ARF on p53 Whether binding of mdm2–p14ARF to p53 can also prevent acetylation at lysine 382 is enhanced, rather than dimin- deacetylation of p53 by other cytoplasmic deacetylases ished, by the presence of mdm2. Given this paradoxical remains to be determined. Similarly, deacetylation of p53 observation, we hypothesize that p14ARF increases the in the nucleus could also be mediated by deacetylases association of p53–mdm2 complexes to chromatin and other than SirT1. In this regard, p53 has been shown to be that this limits the accessibility to p53 of protein deace- regulated by several sirtuin and non-sirtuin HDACs (see tylases, including cytoplasmic deacetylases. Introduction and ref. 4). Here, we show for the first time that p53 can be dea- Previous studies have reported possible tumor sup- cetylated in the cytoplasm and that SirT2 is involved in pressor roles for mdm2 (26, 46). Interestingly, an intact p53 deacetylation in this compartment. SirT2 expression p53/mdm2/p14ARF axis may be needed to detect an has previously been implicated in regulating p53 (6, 8); antitumor effect of mdm2 (47). Altogether our work on however, these studies were conducted with whole-cell acetylated p53 also suggests that under conditions where extracts from virally infected HEK293 and HeLa cells p14ARF expression is induced, mdm2 could play a positive where p53 modulation is strongly influenced by viral role in the activation of p53. We analyzed p53-dependent factors. Our data are the first to characterize the effect of transcription in the presence of p14ARF and increasing SirT2 on p53 in a virus-free background. Furthermore, amounts of ectopic mdm2. However, this experiment did unlike previous reports, we provide evidence that SirT2 not lead to an increase in p53 transcriptional activity by deacetylates p53 directly. mdm2 (Supplementary Fig. S4). This is not unexpected in In summary, we propose a model in which all forms of this type of assays using overexpressed proteins. One p53 (including acetylated p53) accumulate in the chroma- reason could be that binding of mdm2 to p53 suffices to

OF8 Mol Cancer Ther; 12(4) April 2013 Molecular Cancer Therapeutics

p53 C-Terminal Acetylation in the Nucleus and Cytoplasm

In addition to contributing to a further understanding of the regulation of p53 acetylation, the results presented here also provide an insight into why inhibition of both SirT1 and SirT2 is necessary to achieve effective activation of p53 by small-molecule sirtuin inhibitors such as the tenovins. This is in full agreement with results published previously using sirtuin inhibitors unrelated to the tenovins (25). It is worth mentioning here that tenovin-6, which has antitumor activity in vivo as a single agent (29, 48), has been recently shown to induce p53 in imatinib-resistant leukemic stem cells and achieve cure in a murine model for chronic myelogenous leukemia (CML) in combination with imatinib (49, 50). Finally, these results show that p53 acetylation in the cytoplasm is modulated by a metabolism-sensitive protein deacetylase. Now that roles for cytoplasmic p53 in cell death are becoming consolidated (51–53), our observation that it is possible to protect p53 with small molecules in this cellular compartment may have ther- apeutic implications and be important in elucidating the transcription-independent effects of p53.

Disclosure of Potential Conflicts of Interest No potential conflicts of interest were disclosed. Figure 6. Regulation of different forms of p53 by mdm2 and p14ARF in cytoplasmic and nuclear fractions. In this model, the mdm2–p14ARF Authors' Contributions complex sequesters p53 in the nuclear (chromatin-rich) fraction, Conception and design: S. Lain thereby protecting p53 from deacetylation and degradation in the Development of methodology: A.R. McCarthy, S. Lain cytoplasm. For simplicity, modulation of p53 by other sirtuin and Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): I.M.M. van Leeuwen, J. Campbell, A.R. non-sirtuin deacetylases has not been included in this figure. McCarthy, M.C.C. Sachweh, A. Marıń Navarro Analysis and interpretation of data (e.g., statistical analysis, biostatis- tics, computational analysis): I.M.M. van Leeuwen, S. Lain Writing, review, and/or revision of the manuscript: I.M.M. van Leeuwen, reduce p53-dependent transcription even when the level M.C.C. Sachweh, S. Lain Administrative, technical, or material support (i.e., reporting or orga- of p53 acetylation is high. We speculate that additional nizing data, constructing databases): I.M.M. van Leeuwen, M. Higgins, mechanisms are likely to be required to enable a positive J. Campbell, S. Lain role for mdm2 in p53-dependent transcription in our Study supervision: S. Lain system. In a recent groundbreaking publication, Clegg Grant Support and colleagues (27) obtained enhanced expression of p53 This study was supported by grants from various funding agencies: I. target genes in mouse embryonic fibroblasts from an M.M. van Leeuwen (Association for International Cancer Research, AICR mdm2–C462A knock-in mouse model. This extremely grant #10-0043), M. Higgins and J. Campbell (Cancer Research UK, grant #C8/A6613), A.R. McCarthy (Swedish Research Council, Vetens- interesting observation shows that in this model system, kapsradet), M.C.C. Sachweh (Karolinska Institutet), A. Marıń Navarro there are mechanisms that make it possible for mdm2 (Erasmus Programme), and S. Lain (Karolinska Institutet and the Swedish to enhance p53 transcriptional activity when mdm2 E3 Cancer Society, Cancerfonden). The costs of publication of this article were defrayed in part by the ubiquitin ligase function is impaired. We believe that payment of page charges. This article must therefore be hereby marked our results showing that mdm2 promotes p53 acetyla- advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate tion in chromatin-rich fractions when p14ARF is present this fact. ARF (p14 inhibits ubiquitination of p53 by mdm2) contrib- Received September 13, 2012; revised January 8, 2013; accepted January ute to elucidating such mechanisms further. 22, 2013; published OnlineFirst February 15, 2013.

References 1. Kruse JP, Gu W. Modes of p53 regulation. Cell 2009;137:609–22. 5. Martinez-Pastor B, Mostoslavsky R. Sirtuins, metabolism, and cancer. 2. Poyurovsky MV, Katz C, Laptenko O, Beckerman R, Lokshin M, Ahn J, Front Pharmacol 2012;3:22. et al. The C terminus of p53 binds the N-terminal domain of MDM2. Nat 6. Jin YH, Kim YJ, Kim DW, Baek KH, Kang BY, Yeo CY, et al. Sirt2 Struct Mol Biol 2010;17:982–9. interacts with 14-3-3 beta/gamma and down-regulates the 3. Luo J, Li M, Tang Y, Laszkowska M, Roeder RG, Gu W. Acetylation of activity of p53. Biochem Biophys Res Commun 2008;368: p53 augments its site-speciﬁc DNA binding both in vitro and in vivo. 690–5. Proc Natl Acad Sci U S A 2004;101:2259–64. 7. Liu Y, Shevchenko A, Berk AJ. Adenovirus exploits the cellular aggre- 4. Brooks CL, Gu W. The impact of acetylation and deacetylation on the some response to accelerate inactivation of the MRN complex. J Virol p53 pathway. Protein Cell 2011;2:456–62. 2005;79:14004–16.

www.aacrjournals.org Mol Cancer Ther; 12(4) April 2013 OF9

van Leeuwen et al.

8. Li Y, Matsumori H, Nakayama Y, Osaki M, Kojima H, Kurimasa A, et al. inhibitor of sirtuin SirT2, increases p21 expression in a p53-inde- SIRT2 down-regulation in HeLa can induce p53 accumulation via p38 pendent manner. Mol Cancer Ther 2013;12:352–60. MAPK activation-dependent p300 decrease, eventually leading to 32. Bartkova J, Bartek J, Vojtesek B, Lukas J, Rejthar A, Kovarik J, et al. apoptosis. Genes Cells 2011;16:34–45. Immunochemical analysis of the p53 oncoprotein in matched primary 9. Li S, Banck M, Mujtaba S, Zhou MM, Sugrue MM, Walsh MJ. p53- and metastatic human tumours. Eur J Cancer 1993;29A:881–6. induced growth arrest is regulated by the mitochondrial SirT3 deace- 33. Midgley CA, Fisher CJ, Bartek J, Vojtesek B, Lane D, Barnes DM. tylase. PLoS ONE 2010;5:e10486. Analysis of p53 expression in human tumours: an antibody raised 10. Vakhrusheva O, Smolka C, Gajawada P, Kostin S, Boettger T, Kubin T, against human p53 expressed in Escherichia coli. J Cell Sci et al. Sirt7 increases stress resistance of cardiomyocytes and prevents 1992;101:183–9. apoptosis and inflammatory cardiomyopathy in mice. Circ Res 34. Chen J, Marechal V, Levine AJ. Mapping of the p53 and mdm-2 2008;102:703–10. interaction domains. Mol Cell Biol 1993;13:4107–14. 11. van Leeuwen I, Lain S. Sirtuins and p53. Adv Cancer Res 2009;102: 35. Midgley CA, Desterro JM, Saville MK, Howard S, Sparks A, Hay RT, 171–95. et al. An N-terminal p14ARF peptide blocks Mdm2-dependent ubiqui- 12. Brown CJ, Lain S, Verma CS, Fersht AR, Lane DP. Awakening guardian tination in vitro and can activate p53 in vivo. Oncogene 2000;19:2312– angels: drugging the p53 pathway. Nat Rev Cancer 2009;9:862–73. 23. 13. Momand J, Zambetti GP, Olson DC, George D, Levine AJ. The mdm-2 36. van Leeuwen IM, Higgins M, Campbell J, Brown CJ, McCarthy AR, oncogene product forms a complex with the p53 protein and inhibits Pirrie L, et al. Mechanism-specific signatures for small-molecule p53 p53-mediated transactivation. Cell 1992;69:1237–45. activators. Cell Cycle 2011;10:1590–8. 14. Honda R, Tanaka H, Yasuda H. Oncoprotein MDM2 is a ubiquitin ligase 37. Li M, Brooks CL, Wu-Baer F, Chen D, Baer R, Gu W. Mono- versus E3 for tumor suppressor p53. FEBS Lett 1997;420:25–7. polyubiquitination: differential control of p53 fate by Mdm2. Science 15. Kobet E, Zeng X, Zhu Y, Keller D, Lu H. MDM2 inhibits p300-mediated 2003;302:1972–5. p53 acetylation and activation by forming a ternary complex with the 38. Xirodimas DP, Stephen CW, Lane DP. Cocompartmentalization of p53 two proteins. Proc Natl Acad Sci U S A 2000;97:12547–52. and Mdm2 is a major determinant for Mdm2-mediated degradation of 16. Ito A, Lai CH, Zhao X, Saito S, Hamilton MH, Appella E, et al. p300/ p53. Exp Cell Res 2001;270:66–77. CBP-mediated p53 acetylation is commonly induced by p53-acti- 39. Carey MF, Peterson CL, Smale ST. Dignam and Roeder nuclear extract vating agents and inhibited by MDM2. EMBO J 2001;20:1331–40. preparation. Cold Spring Harb Protoc 2009;2009:pdb prot5330. 17. Jin Y, Zeng SX, Dai MS, Yang XJ, Lu H. MDM2 inhibits PCAF (p300/ 40. Lain S, Midgley C, Sparks A, Lane EB, Lane DP. An inhibitor of nuclear CREB-binding protein-associated factor)-mediated p53 acetylation. J export activates the p53 response and induces the localization of Biol Chem 2002;277:30838–43. HDM2 and p53 to U1A-positive nuclear bodies associated with the 18. Ito A, Kawaguchi Y, Lai CH, Kovacs JJ, Higashimoto Y, Appella E, et al. PODs. Exp Cell Res 1999;248:457–72. MDM2-HDAC1-mediated deacetylation of p53 is required for its deg- 41. Li M, Chen D, Shiloh A, Luo J, Nikolaev AY, Qin J, et al. Deubiquitination radation. EMBO J 2002;21:6236–45. of p53 by HAUSP is an important pathway for p53 stabilization. Nature 19. Sherr CJ. Divorcing ARF and p53: an unsettled case. Nat Rev Cancer 2002;416:648–53. 2006;6:663–73. 42. Frye RA. Phylogenetic classification of prokaryotic and eukaryotic 20. Tao W, Levine AJ. Nucleocytoplasmic shuttling of oncoprotein Hdm2 Sir2-like proteins. Biochem Biophys Res Commun 2000;273:793–8. is required for Hdm2-mediated degradation of p53. Proc Natl Acad Sci 43. Shaulsky G, Goldfinger N, Ben-Ze'ev A, Rotter V. Nuclear accumula- U S A 1999;96:3077–80. tion of p53 protein is mediated by several nuclear localization signals 21. Xirodimas D, Saville MK, Edling C, Lane DP, Lain S. Different effects of and plays a role in tumorigenesis. Mol Cell Biol 1990;10:6565–77. p14ARF on the levels of ubiquitinated p53 and Mdm2 in vivo. Onco- 44. North BJ, Verdin E. Interphase nucleo-cytoplasmic shuttling and gene 2001;20:4972–83. localization of SIRT2 during mitosis. PLoS One 2007;2:e784. 22. Weber JD, Taylor LJ, Roussel MF, Sherr CJ, Bar-Sagi D. Nucleolar Arf 45. Sekaric P, Shamanin VA, Luo J, Androphy EJ. hAda3 regulates sequesters Mdm2 and activates p53. Nat Cell Biol 1999;1:20–6. p14ARF-induced p53 acetylation and senescence. Oncogene 23. Llanos S, Clark PA, Rowe J, Peters G. Stabilization of p53 by 2007;26:6261–8. p14ARF without relocation of MDM2 to the nucleolus. Nat Cell Biol 46. Hamard PJ, Manfredi JJ. Mdm20s dilemma: to degrade or to translate 2001;3:445–52. p53?Cancer Cell 2012;21:3–5. 24. Rodway H, Llanos S, Rowe J, Peters G. Stability of nucleolar 47. Brown DR, Thomas CA, Deb SP. The human oncoprotein MDM2 versus non-nucleolar forms of human p14(ARF). Oncogene 2004; arrests the cell cycle: elimination of its cell-cycle-inhibitory function 23:6186–92. induces tumorigenesis. EMBO J 1998;17:2513–25. 25. Peck B, Chen CY, Ho KK, Di Fruscia P, Myatt SS, Coombes RC, et al. 48. Marshall GM, Liu PY, Gherardi S, Scarlett CJ, Bedalov A, Xu N, et al. SIRT inhibitors induce cell death and p53 acetylation through targeting SIRT1 promotes N-Myc oncogenesis through a positive feedback loop both SIRT1 and SIRT2. Mol Cancer Ther 2010;9:844–55. involving the effects of MKP3 and ERK on N-Myc protein stability. 26. Manfredi JJ. The Mdm2-p53 relationship evolves: Mdm2 swings both PLoS Genet 2012;7:e1002135. ways as an oncogene and a tumor suppressor. Genes Dev 2010;24: 49. Li L, Wang L, Wang Z, Ho Y, McDonald T, Holyoake TL, et al. Activation 1580–9. of p53 by SIRT1 inhibition enhances elimination of CML leukemia stem 27. Clegg HV, Itahana Y, Itahana K, Ramalingam S, Zhang Y. Mdm2 RING cells in combination with imatinib. Cancer Cell 2012;21:266–81. mutation enhances p53 transcriptional activity and p53-p300 interac- 50. Yuan H, Wang Z, Li L, Zhang H, Modi H, Horne D, et al. Activation of tion. PLoS ONE 2012;7:e38212. stress response gene SIRT1 by BCR-ABL promotes leukemogenesis. 28. Blaydes JP, Hupp TR. DNA damage triggers DRB-resistant phos- Blood 2012;119:1904–14. phorylation of human p53 at the CK2 site. Oncogene 1998;17: 51. Vaseva AV, Marchenko ND, Ji K, Tsirka SE, Holzmann S, Moll UM. p53 1045–52. opens the mitochondrial permeability transition pore to trigger necro- 29. Lain S, Hollick JJ, Campbell J, Staples OD, Higgins M, Aoubala M, et al. sis. Cell 2012;149:1536–48. Discovery, in vivo activity, and mechanism of action of a small-mol- 52. Galluzzi L, Morselli E, Kepp O, Vitale I, Pinti M, Kroemer G. Mitochon- ecule p53 activator. Cancer Cell 2008;13:454–63. drial liaisons of p53. Antioxid Redox Signal 2011;15:1691–714. 30. McCarthy AR, Pirrie L, Hollick JJ, Ronseaux S, Campbell J, Higgins M, 53. Maiuri MC, Galluzzi L, Morselli E, Kepp O, Malik SA, Kroemer G. et al. Synthesis and biological characterisation of sirtuin inhibitors Autophagy regulation by p53. Curr Opin Cell Biol 2010;22:181–5. based on the tenovins. Bioorg Med Chem 2012;20:1779–93. 54. Vassilev LT, Vu BT, Graves B, Carvajal D, Podlaski F, Filipovic Z, et al. in 31. McCarthy AR, Sachweh MCC, Higgins M, Campbell J, Drummond vivo activation of the p53 pathway by small-molecule antagonists of CJ, Van Leeuwen IMM, et al. Tenovin-D3, a novel small-molecule MDM2. Science 2004;303:844–8.

OF10 Mol Cancer Ther; 12(4) April 2013 Molecular Cancer Therapeutics

Modulation of p53 C-Terminal Acetylation by mdm2, p14ARF, and Cytoplasmic SirT2

Ingeborg M.M. van Leeuwen, Maureen Higgins, Johanna Campbell, et al.

Mol Cancer Ther Published OnlineFirst February 15, 2013.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-12-0904

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2013/02/15/1535-7163.MCT-12-0904.DC1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/early/2013/03/28/1535-7163.MCT-12-0904. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.