Potential Interplay with the PKB/Akt Pathway and P14arf

FEBS Letters 583 (2009) 615–620

journal homepage: www.FEBSLetters.org

14-3-3 Binding to Pim-phosphorylated Ser166 and Ser186 of human Mdm2 – Potential interplay with the PKB/Akt pathway and p14ARF

Nicola T. Wood a, David W. Meek b, Carol MacKintosh a,* a MRC Protein Phosphorylation Unit, Sir James Black Centre, College of Life Sciences, University of Dundee, Dundee DD1 5EH, Scotland, UK b Biomedical Research Centre, Ninewells Medical School, University of Dundee, Dundee DD1 9SY, Scotland, UK article info abstract

Article history: Here we show that 14-3-3 proteins bind to Pim kinase-phosphorylated Ser166 and Ser186 on the Received 24 September 2008 human E3 ubiquitin ligase mouse double minute 2 (Mdm2), but not protein kinase B (PKB)/ Revised 24 December 2008 Akt-phosphorylated Ser166 and Ser188. Pim-mediated phosphorylation of Ser186 blocks phosphor- Accepted 11 January 2009 ylation of Ser188 by PKB, indicating potential interplay between the Pim and PKB signaling path- Available online 21 January 2009 ways in regulating Mdm2. In cells, expression of Pim kinases promoted phosphorylation of Ser166 ARF Edited by Angel Nebreda and Ser186, interaction of Mdm2 with endogenous 14-3-3s and p14 , and also increased the amount of Mdm2 protein by a mechanism that does not require Pim kinase activities. The implications of these ﬁndings for regulation of the p53 pathway, oncogenesis and drug discovery are dis- Keywords: 14-3-3 cussed. Cancer Mdm2 Structured summary: Phosphorylation MINT-6823587: PIM3 (uniprotkb:Q86V86) phosphorylates (MI:0217) MDM2 (uniprotkb:Q00987) by pro- Pim kinase tein kinase assay (MI:0424) MINT-6823623: MDM2 (uniprotkb:Q00987) physically interacts (MI:0218) with p14ARF (uniprotkb:Q8N7268N726) by coimmunoprecipitation (MI:0019) MINT-6823537: PKB (uniprotkb:P31749) phosphorylates (MI:0217) MDM2 (uniprotkb:Q00987) by protein kinase assay (MI:0424) MINT-6823574: PIM2 (uniprotkb:QP1W9) phosphorylates (MI:0217) MDM2 (uniprotkb:Q00987) by protein kinase assay (MI:0424) MINT-6823555: PIM1 (uniprotkb:P11309)P phosphorylates (MI:0217) MDM2 (uniprotkb:Q00987) by protein kinase assay (MI:0424)

1. Introduction by protein kinase B (PKB)/Akt, and clustered around a bipartite nuclear localization sequence (NLS), residues 178–184 [6], and an The tumor suppressor p53 is commonly deregulated in tumors nuclear export sequence (NES), residues 190–199 [7]. These phos- by mutation or deletion of p53 itself, and/or upregulation of mouse phorylations reportedly promote nuclear entry, allowing Mdm2 double minute 2 (Mdm2), an E3 ubiquitin ligase that targets p53 access to nuclear p53 [8–10]. The Mdm2/p53 complex is thought and other proteins for proteasomal degradation and also inhibits to shuttle back out to the cytoplasm for proteasomal degradation p53-dependent transcription [1,2]. Mdm2 is regulated by multisite of p53 [7,11]. phosphorylation [3–5], including phosphorylation of Ser166, Protein phosphorylation by PKB/Akt sometimes creates sites for

Ser186 and Ser188 (RRRAIS166ETEENSDELSGERQRKRHKS186DS188- binding to 14-3-3s, a family of dimeric proteins that dock onto spe- ISLSFDESLAL) that are variously reported to be phosphorylated ciﬁc phosphorylated residues and change their targets’ activities and interactions with other proteins [12]. Their dimeric structure allows 14-3-3s to bind simultaneously to two phosphorylated sites Abbreviations: Mdm2, mouse double minute 2; NES, nuclear export sequence; a minimum of 15 residues apart. These considerations prompted NLS, nuclear localization sequence; PKB, protein kinase B (also known as Akt); PP2A, us to test whether a 14-3-3 dimer might bind two of the reported protein phosphatase 2A. PKB-phosphorylated residues at Ser166, Ser186 and/or Ser188 * Corresponding author. Fax: +44 1382 223778. E-mail address: [email protected] (C. MacKintosh). on human Mdm2. We found that 14-3-3s do indeed bind

0014-5793/$34.00 Ó 2009 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.febslet.2009.01.003 616 N.T. Wood et al. / FEBS Letters 583 (2009) 615–620 phosphorylated Mdm2, but this association is promoted not by amount that catalyzes phosphorylation of 1 nmol peptide sub- PKB phosphorylation, which we observe in vitro on Ser166 and strate in 1 min at 30 °C. Ser188, but by phosphorylation by oncogenic Pim kinases, which phosphorylate Ser166 and Ser186 [13]. In vivo, Pim kinases stabi- 2.5. Substrate purification and in vitro phosphorylation lize Mdm2 and promote phosphorylation-dependent interaction with endogenous 14-3-3s and p14ARF. GST-Mdm2 mini-proteins were overexpressed in E. coli BL21 cells (Life Technologies) by induction with 50–250 lM isopropyl- ß-D-thiogalactopyranoside at 26 °C for 16 h. Cells were sonicated, 2. Materials and methods lysates centrifuged and GST-fusion proteins purified as described above. Purified proteins were dialyzed against 50 mM Tris–HCl 2.1. Materials pH 7.5, 0.1 mM EGTA, 250 mM NaCl, 0.1% 2-mercaptoethanol, 270 mM sucrose, 0.03% Brij-35, 1 mM benzamidine, 0.2 mM PMSF Protein G-Sepharose was from GE Healthcare, tissue culture re- overnight at 4 °C and their identity confirmed by tryptic mass fin- agents from Biowhittaker and Life Technologies, pre-cast 4–12% gerprinting [16] and Western blotting. Purified proteins (2 lg) polyacrylamide Bis–Tris gels and associated buffers from Invitro- were phosphorylated in vitro with 400 mU/ml Pim kinases and gen. Unless stated, other chemicals were from Sigma–Aldrich or 1 U/ml PKB/Akt in 50 mM Tris–HCl pH 7.5, 1 mM EDTA pH 7.9, BDH. Mouse monoclonal SMP14 to detect Mdm2 and rabbit anti- 0.1% (v/v) 2-mercaptoethanol. Reactions were performed at 30 °C p14ARF were from Abcam, mouse anti-GST was from Sigma and for 30 min unless otherwise stated and terminated by addition of K19 pan anti-14-3-3 from Santa Cruz. Phosphospecific antibodies LDS sample buffer. to Mdm2 were described previously [5]. 2.6. Cell culture, lysis and immunoprecipitation 2.2. General methods and buffers Human embryonic kidney (HEK) 293 cells (European Collection DNA constructs for transfection of human cells were purified of Cell Cultures) were maintained in Dulbecco’s modified Eagle’s from transformed Escherichia coli DH5a using Qiagen Mega medium supplemented with 10% (v/v) fetal bovine serum. Cells plasmid purification kits. Lysis buffer contained 50 mM Tris–HCl were cultured on 10 cm diameter dishes and split to 10–20% con- pH 7.5, 1 mM EGTA, 1 mM EDTA, 1% (w/v) Triton X-100, 1 mM fluency the day prior to transfection. Cells were transiently trans- sodium orthovanadate, 50 mM sodium fluoride, 5 mM sodium fected with 5–10 lg plasmid DNA per dish using the DNA/ pyrophosphate, 0.27 M sucrose, 0.1% (v/v) 2-mercaptoethanol calcium phosphate co-precipitation method [17] and maintained and one ‘Complete’ protease inhibitor tablet (Roche) per 50 ml. in 5% CO2 at 37 °C for 16–30 h. Cells were serum starved for 5 h TBS contained 50 mM Tris–HCl pH 7.5, 500 mM NaCl (with 0.1% (down-regulating the PKB pathway; Pim kinases remain active in (v/v) Tween-20 included for TBS-T). Protein phosphatase 2A serum starved cells [14]) before lysis in 0.5–1 ml of ice-cold lysis (PP2A) was purified from bovine cardiac muscle by Julie Diplexcito buffer. Lysates were clarified by centrifugation at 13000Âg for and Jean Harthill. 10 min at 4 °C. For immunoprecipitations, 10 lg antibody was cou- pled to Protein G-Sepharose (30 ll of 50% slurry in PBS) by mixing 2.3. Plasmids for 1 h, and lysate (4 mg) added and mixed for 2 h at 4 °C. Cells were washed twice with TBS and twice with 50 mM Tris–HCl pH Human Pim-1 and Pim-2 mammalian expression constructs and 7.5, 1 mM EDTA, 0.1% (v/v) 2-mercaptoethanol, pelleting Sepharose murine Pim-3 constructs for mammalian and baculovirus expres- in between washes. sion were described previously [14], as were constructs for bacterial expression of glutathione S-transferase (GST)-Mdm2 2.7. Western blotting and 14-3-3 overlays mini-proteins [15]. The construct to express Mdm2 (NCBI Acc. BT007258) as an HA-tagged protein in mammalian cells was Proteins separated by SDS–PAGE were transferred to nitrocellu- prepared by subcloning an EcoR1 fragment from a full-length clone lose by wet electroblotting and processed for ECL detection (En- of Mdm2 in pCDNA3 into pCMV5.HA2. Mutations were introduced hanced Chemiluminescent reagent; Amersham Biosciences) with using KOD Hot Start DNA polymerase (Novagen). DNA was HRP-conjugated secondary antibodies (SMP14 immunoblots) or sequenced by The Sequencing Service, University of Dundee Odyssey Infrared Imaging (LiCor) with AlexaFluor or IR Dye conju- (www.dnaseq.co.uk). gates. Membranes were blocked in 5% Marvel milk powder/TBS prior to detection. Commercial antibodies were used according to 2.4. Enzymes and kinase assays the manufacturer’s instructions, diluted in 1% Marvel/TBS-T. Phos- phospecific antibodies were used as in [5]. Mouse anti-HA was GST-Pim-1 and GST-Pim-2 proteins were purified from used at 0.5 lg/ml and K19 (anti-14-3-3) at 1 lg/ml. 14-3-3 Over- lysates of transfected HEK293 cells by binding to glutathione- lays measure direct binding of digoxygenin-labeled 14-3-3 to Sepharose (Amersham Biosciences), washing extensively in wash phosphorylated target proteins on a nitrocellulose membrane, buffer (50 mM Tris–HCl pH 7.5, 250 mM NaCl, 0.03% Brij-35, and were performed as previously described [12]. 0.1 mM EGTA, 0.1% 2-mercaptoethanol, 0.2 mM PMSF, 1 mM benzamidine), and eluting in wash buffer containing 20 mM 3. Results glutathione. His-tagged Pim-3 and PKB/Akt (His-PKBa-S473D; residues 118–480 of the human protein) were purified in the 3.1. Phosphorylation by Pim kinases but not PKB promotes 14-3-3 Division of Signal Transduction Therapy from Sf9 insect cells on binding to Mdm2 in vitro Ni-NTA-Sepharose using standard techniques. PKB was activated by PDK1 (phosphoinositide-dependent kinase 1). The standard GST-fusions of four ‘mini-proteins’ spanning the functional substrate used to assay the Pim kinases was RSRMSSYPDRG domains of Mdm2 (termed MP1 [Mdm2 residues 1–110], MP2 (derived from murine Bad) as described [14]. PKB was assayed [108–200], MP3 [196–282] and MP4 [276–491]) were phosphory- against Crosstide (GRPRTSSFAEG). One mU of activity is the lated in vitro with PKB, Pim-1, Pim-2 and Pim-3. None were able to N.T. Wood et al. / FEBS Letters 583 (2009) 615–620 617

Fig. 1. GST-Mdm2 MP2 phosphorylated in vitro by Pim-1 binds 14-3-3s: (a) schematic representation showing full-length mouse double minute 2 (Mdm2) and the four GST- Mdm2 mini-proteins (MP1–4) used as substrates and (b) 14-3-3 Overlays of Mdm2 mini-proteins dephosphorylated by protein phosphatase 2A (PP2A) (50 mU/ml) and phosphorylated in vitro by the kinases indicated. Except for the first lane (no treatment), other substrates were pre-incubated with protein phosphatase 2A at 30 °C for 30 min and the reaction stopped by adding 50 lM microcystin-LR, prior to adding 10 mM MgAcetate/0.1 mM ATP and the kinase indicated and incubating for a further 30 min at 30 °C. Phosphorylation reactions were stopped by the addition of LDS sample buffer. bind 14-3-3s following phosphorylation by PKB, whereas 14-3-3s kinases and was phosphorylation-dependent, as no interaction did bind PKB-phosphorylated E3 ubiquitin ligase Nedd4-2 was observed with kinase-dead Pim-1 or Pim-2 (kinase-dead (Fig. 1). In contrast, phosphorylation by any of the three Pim Pim-3 was unavailable at the time of study) (Fig. 2b). While all kinases promoted 14-3-3 binding to GST-mini-protein 2, which three Pim kinases phosphorylate Ser166 and Ser186, consistent harbors the serine-flanked NLS and NES motifs. with recent findings [13], Pim-3 also phosphorylated Ser188 (Fig. 2b). Together, these data indicate that phosphoSer186 is 3.2. PKB and Pim kinases differentially phosphorylate serine residues essential for 14-3-3 binding and explain why PKB is not able to flanking the NLS promote this interaction. In an in vitro time course, Ser186 was maximally phosphory- Using phosphospecific antibodies [5], PKB and Pim-1 were both lated by Pim-1 within 1 min, whereas phosphorylation of Ser166 found to phosphorylate Ser166, and PKB also phosphorylated and 14-3-3 binding both increased gradually over 30 min Ser188 whereas Pim-1 phosphorylated Ser186 (Fig. 2a). Binding (Fig. 2c). Both phosphoSer166 and phosphoSer186 are required of 14-3-3s was observed specifically after incubation with Pim for optimal 14-3-3 binding, as the interaction was significantly

Fig. 2. Identiﬁcation of phosphorylation sites required for 14-3-3 binding: phosphospeciﬁc immunoblots and 14-3-3 overlays of GST-Mdm2-MP2 after phosphorylation in vitro by (a) protein kinase B (PKB) or Pim-1, (b) wild-type or kinase-dead Pim kinases, (c) Pim-1, with reactions terminated at the time points indicated, and (d) GST-Mdm2-MP2 and alanine substitution mutants after in vitro phosphorylation with Pim-1. Mdm2 is shown as a loading control. 618 N.T. Wood et al. / FEBS Letters 583 (2009) 615–620 inhibited by a substitution with alanine at either site (Fig. 2d). Intriguingly, when Ser186 could not be phosphorylated (in mutants harboring the Ser186Ala substitution), Pim-1 weakly phosphorylated Ser188, although this did not compensate for the lack of phosphoSer186 with respect to 14-3-3 binding.

3.3. Phosphorylation of Ser186 by Pim-1 blocks subsequent phosphorylation of Ser188 by PKB

In vitro phosphorylation of GST-mini-protein 2 with Pim-1 blocked subsequent phosphorylation of Ser188 by PKB (Fig. 3a). This was speciﬁcally due to phosphorylation of Ser186 since the Ser186Ala mutant (phosphorylated at Ser166) could still be phosphorylated by PKB at Ser188. However, when Ser166 and Ser188 were ﬁrst phosphorylated by PKB, Pim-1 could still phosphorylate Ser186 (Fig. 3b).

3.4. Expression of Pim kinases stabilizes Mdm2 in vivo but binding to endogenous 14-3-3s and p14ARF is phosphorylation-dependent

Pim kinases are controlled by regulated expression of the con- stitutively active enzymes [18]. The stimuli that induce Pim kinase expression in vivo also trigger more-widespread changes however. Therefore, to test the specific effects of Pim kinases on Mdm2 in vivo, constructs encoding wild-type or kinase-dead (inactive) Pim kinases were co-expressed in HEK293 cells alongside full-length Mdm2. Strikingly, Pim kinase expression induced a significant in- Fig. 4. Co-expression with Pim-1 leads to elevated levels of HA-Mdm2 in cells: Western crease in the levels of Mdm2 protein (Fig. 4), an effect that was blots showing recombinant Pim-1 and Mdm2 levels in lysates of HEK293 cells that were transfected to express HA-Mdm2 alone (wild-type or alanine substitution independent of phosphorylation at Ser166 and Ser186 since it mutants), or HA-Mdm2 plus wild-type or kinase-dead GST-Pim-1. Similarly, was observed with Mdm2 mutants harboring Ser to Ala substitu- enhanced Mdm2 protein levels were seen for cells that co-expressed FLAG-tagged tions at these positions and upon co-expression of both kinase- Pim kinases (wild-type and kinase-dead), but not cells that co-expressed GST (data dead and wild-type Pim kinases. 14-3-3 binding however, was not shown). phosphorylation-dependent; phosphospecific antibodies revealed that only Mdm2 co-expressed with active, wild-type Pim kinases was phosphorylated on Ser166 and Ser186 and endogenous 14- 3-3s co-purified specifically with Pim-phosphorylated wild-type Mdm2 (Fig. 5). We also checked whether other known interactors of Mdm2 (including p53, the transcriptional co-activator p300, the Mdm2-related protein MdmX and the tumor suppressor p14ARF)

Fig. 5. Interaction of Mdm2 with 14-3-3s and p14ARF is phosphorylation-dependent: (a) phosphospecific immunoblots and 14-3-3 overlay of HA-Mdm2 immunoprecipi- tated from extracts of transfected HEK293 cells co-expressing either wild-type or Fig. 3. Sequential phosphorylation of MP2 by PKB and Pim kinases: residues kinase-dead Pim-1 and (b) Western blots of endogenous 14-3-3s and p14ARF in anti- phosphorylated in vitro by either (a) Pim-1 followed by PKB or (b) PKB followed HA immunoprecipitates extracts of from transfected HEK293 cells co-expressing by Pim-1. Reactions were started with the first kinase for 30 min followed by the wild-type or double alanine mutant HA-Mdm2 with wild-type or kinase-dead second kinase for a further 30 min. Pim-1. N.T. Wood et al. / FEBS Letters 583 (2009) 615–620 619 were affected by Pim-phosphorylation and/or 14-3-3 binding. We though Mdm2 levels are increased by co-expression of both found that p14ARF selectively co-purified with Pim-phosphorylated wild-type or kinase-dead (inactive) Pim-1. We suggest therefore wild-type Mdm2 (Fig. 5). that this up-regulatory effect may be independent of Pim-1 phosphorylation at Ser166 and Ser186 and ARF binding to Mdm2; it 4. Discussion is also independent of transcriptional feedback control from other components of the p53 pathway since it is observed even when Our findings connect three types of signaling component, Mdm2 is expressed from a heterologous promoter. The functional Mdm2, Pim kinases and 14-3-3 proteins, which are each linked significance of ARF binding specifically to Pim-phosphorylated to the p53 pathway and cancers. Pim kinases are overexpressed Mdm2, and indeed the precise role of 14-3-3s in this complex in hematological malignancies and other cancers, with impacts therefore, remain to be determined. on Mdm2 and the p53 pathway [13]. 14-3-3s have central regula- Finally, finding that both wild-type and kinase-dead Pims pro- tory roles in processes that are deregulated in cancers. Generally, mote elevated levels of Mdm2 in cells ([13] and this study) has 14-3-3/target interactions promote cell survival events and several implications for the Pim kinases as a drug targets for hematological 14-3-3 isoforms are overexpressed in various tumors. In contrast, malignancies. Clearly, it will be critical to determine whether or the 14-3-3sigma isoform is frequently deleted in tumors, and has not drug-inhibited Pims can also elevate cellular Mdm2 levels. special roles in DNA damage responses, inhibits cell cycle progression, and is a p53-inducible gene [19,20]. Recently, 14-3-3sigma Acknowledgments was reported to bind to the ring domain of Mdm2, destabilize Mdm2 and promote its cytoplasmic localization, thereby stabiliz- This work was supported by the UK Medical Research Council ing p53 [21]. 14-3-3sigma is not present in the HEK293 cells used and the companies who have supported the Division of Signal in our study, though these cells do contain the other six 14-3-3 Transduction Therapy at the University of Dundee, namely Astra- isoforms (theta, gamma, beta, eta, zeta, epsilon) [22]. It will be Zeneca, Boehringer Ingelheim, GlaxoSmithKline, Merck and Co., important to determine the selectivity of 14-3-3 isoforms for bind- Merck KGaA and Pfizer. Thanks to Claire Balfour for tissue culture ing to the ring domain [21] and Pim-phosphorylated Ser166 and support, Carol Hogan for helpful discussions, and Andrew Macdon- Ser186 (this study). 14-3-3s have also been reported to impact ald for providing Pim constructs and performing some kinase on Mdm2 levels by binding to PKB/Akt-phosphorylated sites on assays. MdmX, which stabilizes MdmX itself and also Mdm2 [23]. A recent study [24] set a precedent for Pim-mediated phosphor- References ylation and 14-3-3-binding of residues flanking an NLS on the KIP1 cyclin-dependent kinase inhibitor p27 , inhibiting NLS binding [1] Laurie, N.A. et al. (2006) Inactivation of the p53 pathway in retinoblastoma. to the importins required for nuclear import. Whether 14-3-3 Nature 444, 61–66. binding to Pim-phosphorylated pSer166 and pSer186 on Mdm2 [2] Toledo, F. and Wahl, G.M. (2007) MDM2 and MDM4: p53 regulators as targets in anticancer therapy. Int. J. Biochem. Cell. Biol. 39, 1476–1482. influences its subcellular localization remains to be determined. [3] Guerra, B., Gotz, C., Wagner, P., Montenarh, M. and Issinger, O.G. (1997) The For Mdm2, it is PKB/Akt that has been proposed to play a role in carboxy terminus of p53 mimics the polylysine effect of protein kinase CK2- translocating Mdm2 to the nucleus in order to down-regulate catalyzed MDM2 phosphorylation. Oncogene 14, 2683–2688. [4] Hay, T.J. and Meek, D.W. (2000) Multiple sites of in vivo phosphorylation in the p53 [8–10]. Although there is general consensus that PKB/Akt MDM2 oncoprotein cluster within two important functional domains. FEBS phosphorylates Mdm2 at Ser166 [8–10,25], the second PKB/Akt Lett. 478, 183–186. site has been proposed to be either Ser186 [9,10] or Ser188 [5] Milne, D., Kampanis, P., Nicol, S., Dias, S., Campbell, D.G., Fuller-Pace, F. and Meek, D. (2004) A novel site of AKT-mediated phosphorylation in the human [5,25]. While we confirm Ser166 and Ser188 as in vitro PKB/Akt MDM2 onco-protein. FEBS Lett. 577, 270–276. sites, we failed to detect phosphoSer188 on tagged Mdm2 under [6] Olson, D.C., Marechal, V., Momand, J., Chen, J., Romocki, C. and Levine, A.J. conditions where other PKB/Akt substrates were phosphorylated (1993) Identification and characterization of multiple mdm-2 proteins and mdm-2–p53 protein complexes. Oncogene 8, 2353–2360. in cells (not shown), which could, perhaps, be due to hierarchical [7] Roth, J., Dobbelstein, M., Freedman, D.A., Shenk, T. and Levine, A.J. (1998) interactions between phosphorylated sites on Mdm2. For example, Nucleo-cytoplasmic shuttling of the hdm2 oncoprotein regulates the levels of here we found that phosphorylation of Ser186 by Pim-1 in vitro the p53 protein via a pathway used by the human immunodeficiency virus rev blocks subsequent phosphorylation of Ser188 by PKB/Akt. This protein. Embo J. 17, 554–564. [8] Ashcroft, M., Ludwig, R.L., Woods, D.B., Copeland, T.D., Weber, H.O., MacRae, observation is noteworthy in view of the proximity of sites phos- E.J. and Vousden, K.H. (2002) Phosphorylation of HDM2 by Akt. Oncogene 21, phorylated by Pim kinases and PKB, both to each other, and to 1955–1962. functional motifs governing subcellular localization. Indeed, it is [9] Mayo, L.D. and Donner, D.B. (2001) A phosphatidylinositol 3-kinase/Akt pathway promotes translocation of Mdm2 from the cytoplasm to the nucleus. tempting to speculate that any interplay between these pathways Proc. Natl. Acad. Sci. USA 98, 11598–11603. (one leading to 14-3-3 binding, one not) may represent a potential [10] Zhou, B.P., Liao, Y., Xia, W., Zou, Y., Spohn, B. and Hung, M.C. (2001) HER-2/neu regulatory switch controlling nucleo-cytoplasmic shuttling of induces p53 ubiquitination via Akt-mediated MDM2 phosphorylation. Nat. Cell Biol. 3, 973–982. Mdm2. [11] Tao, W. and Levine, A.J. (1999) Nucleocytoplasmic shuttling of oncoprotein Here, Pim-phosphorylated Mdm2 specifically co-purified not Hdm2 is required for Hdm2-mediated degradation of p53. Proc. Natl. Acad. Sci. only 14-3-3s, but also the tumor suppressor p14ARF. ARF sup- USA 96, 3077–3080. [12] Pozuelo Rubio, M., Geraghty, K.M., Wong, B.H., Wood, N.T., Campbell, D.G., presses aberrant cell growth in response to oncogenic stimulation Morrice, N. and Mackintosh, C. (2004) 14-3-3-Affinity purification of over 200 by various antagonistic interactions that release p53 from the human phosphoproteins reveals new links to regulation of cellular inhibitory effects of Mdm2, directly inhibiting Mdm2 ubiquitin metabolism, proliferation and trafficking. Biochem. J. 379, 395–408. [13] Hogan, C., Hutchison, C., Marcar, L., Milne, D., Saville, M., Goodlad, J., Kernohan, ligase [26,27], and, in certain circumstances, sequestering Mdm2 N. and Meek, D. (2008) Elevated levels of oncogenic protein kinase Pim-1 in the nucleolus [11,28]. ARF binding also stabilizes Mdm2 [29] induce the p53 pathway in cultured cells and correlate with increased Mdm2 and is required for the Pim-induced upregulation of Mdm2 in mantle cell lymphoma. J. Biol. Chem. 283, 18012–18023. observed by Hogan et al. [13]. While our coimmunoprecipitation [14] Macdonald, A., Campbell, D.G., Toth, R., McLauchlan, H., Hastie, C.J. and Arthur, J.S. (2006) Pim kinases phosphorylate multiple sites on Bad and promote 14-3- data are not inconsistent with the hypothesis that expression of 3 binding and dissociation from Bcl-XL. BMC Cell Biol. 7, 1. the Pim-1 oncogene induces ARF and that ARF is required for [15] Kurki, S., Peltonen, K., Latonen, L., Kiviharju, T.M., Ojala, P.M., Meek, D. and increased Mdm2 expression/stability, we find that ARF co-purifies Laiho, M. (2004) Nucleolar protein NPM interacts with HDM2 and protects tumor suppressor protein p53 from HDM2-mediated degradation. Cancer Cell specifically with Mdm2 co-expressed with wild-type Pim-1 even 5, 465–475. 620 N.T. Wood et al. / FEBS Letters 583 (2009) 615–620

[16] Campbell, D.G. and Morrice, N.A. (2002) Identiﬁcation of protein [24] Morishita, D., Katayama, R., Sekimizu, K., Tsuruo, T. and Fujita, N. (2008) Pim phosphorylation sites by a combination of mass spectrometry and solid kinases promote cell cycle progression by phosphorylating and down- phase Edman sequencing. J. Biomol. Technol. 13, 119–130. regulating p27Kip1 at the transcriptional and posttranscriptional levels. [17] Graham, F.L. and van der Eb, A.J. (1973) Transformation of rat cells by DNA of Cancer Res. 68, 5076–5085. human adenovirus 5. Virology 54, 536–539. [25] Feng, J., Tamaskovic, R., Yang, Z., Brazil, D.P., Merlo, A., Hess, D. and Hemmings, [18] Qian, K.C. et al. (2005) Structural basis of constitutive activity and a unique B.A. (2004) Stabilization of Mdm2 via decreased ubiquitination is mediated by nucleotide binding mode of human Pim-1 kinase. J. Biol. Chem. 280, 6130– protein kinase B/Akt-dependent phosphorylation. J. Biol. Chem. 279, 35510– 6137. 35517. [19] Wilker, E.W., Grant, R.A., Artim, S.C. and Yaffe, M.B. (2005) A structural basis [26] Honda, R. and Yasuda, H. (1999) Association of p19(ARF) with Mdm2 inhibits for 14-3-3sigma functional speciﬁcity. J. Biol. Chem. 280, 18891–18898. ubiquitin ligase activity of Mdm2 for tumor suppressor p53. Embo J. 18, 22– [20] Wilker, E.W. et al. (2007) 14-3-3sigma Controls mitotic translation to facilitate 27. cytokinesis. Nature 446, 329–332. [27] Midgley, C.A., Desterro, J.M., Saville, M.K., Howard, S., Sparks, A., Hay, R.T. and [21] Yang, H.Y., Wen, Y.Y., Lin, Y.I., Pham, L., Su, C.H., Yang, H., Chen, J. and Lee, M.H. Lane, D.P. (2000) An N-terminal p14ARF peptide blocks Mdm2-dependent (2007) Roles for negative cell regulator 14-3-3sigma in control of MDM2 ubiquitination in vitro and can activate p53 in vivo. Oncogene 19, activities. Oncogene 26, 7355–7362. 2312–2323. [22] Geraghty, K.M. et al. (2007) Regulation of multisite phosphorylation and 14-3- [28] Weber, J.D., Taylor, L.J., Roussel, M.F., Sherr, C.J. and Bar-Sagi, D. (1999) 3 binding of AS160 in response to IGF-1, EGF, PMA and AICAR. Biochem. J. 407, Nucleolar Arf sequesters Mdm2 and activates p53. Nat. Cell Biol. 1, 20–26. 231–241. [29] Stott, F.J. et al. (1998) The alternative product from the human CDKN2A locus, [23] Lopez-Pajares, V., Kim, M.M. and Yuan, Z.M. (2008) Phosphorylation of MDMX p14(ARF), participates in a regulatory feedback loop with p53 and MDM2. mediated by Akt leads to stabilization and induces 14-3-3 binding. J. Biol. Embo J. 17, 5001–5014. Chem. 283, 13707–13713.