sign in sign up
<<
Home , P14arf

(2007) 26, 4627–4634 & 2007 Nature Publishing Group All rights reserved 0950-9232/07 $30.00 www.nature.com/onc ORIGINAL ARTICLE ARF promotes accumulation of retinoblastoma through inhibition of

DLF Chang1, W Qiu1, H Ying2, Y Zhang3, C-Y Chen4 and Z-XJ Xiao1,2

1Graduate Program in Molecular Medicine, Department of Medicine, Boston University School of Medicine, Boston, MA, USA; 2Department of Biochemistry, Boston University School of Medicine, Boston, MA, USA; 3The Pulmonary Center, Boston University School of Medicine, Boston, MA, USA and 4Department of Pathology, Boston University School of Medicine, Boston, MA, USA

The INK4a/ARF locus,encoding two tumor suppressor binds to and inhibits D-associated kinases , INK4a and p14ARF (ARF),plays key roles in resulting in hypophosphorylated many cellular processes including proliferation, (Rb), a biologically active Rb species in tumor suppres- ,cellular senescence and differentiation. Inacti- sion (Weinberg, 1995). It was later discovered that the vation of INK4a/ARF is one of the most frequent events same locus encodes a second protein through the usage during human development. Although p16INK4a is of a separate with a distinct first (exon a critical component in retinoblastoma protein (Rb)- 1b), which results in translation from an alternate mediated growth regulatory pathway,p14 ARF plays a (ARF) encoding a basic, nucleolar pivotal role in the activation of upon oncogenic stress protein named p14ARF in human (p19ARF in mouse). signals. A body of evidence indicates that ARF also ARF shares no homology with p16INK4a possesses growth suppression functions independent of protein and is activated by oncogenic stress signals, such p53,the mechanism of which is not well understood. We as Ras, c-, E1A and (Serrano, 2000; Sherr, have recently shown that MDM2 interacts with Rb and 2001). promotes -dependent Rb degradation. In this Mice lacking both INK4a and ARF alleles (INK4aÀ/À/ study,we show that ARF disrupts MDM2–Rb interaction ARFÀ/À) are prone to spontaneous and carcinogen- resulting in Rb accumulation. Wild-type ARF,but not induced tumor formation (Serrano et al., 1996). ARF mutant defective in MDM2 interaction,stabilizes Selective disruption of ARF alone recapitulates the Rb and inhibits colony foci formation independent of p53. phenotypic consequences seen in the INK4aÀ/À/ARFÀ/À In addition,ablation of Rb impairs ARF function in double-null mice (Kamijo et al., 1997; Sharpless et al., growth suppression. Thus,this study demonstrates that 2004). The ARFÀ/À cells are highly susceptible to Ras ARF plays a direct role in regulation of Rb and suggests transformation and are resistant to both Ras- and that inactivation of ARF may lead to defects in both p53 culture-induced senescence (Groth et al., 2000). In and Rb pathways in human cancer development. addition, exon 1b is frequently targeted in human Oncogene (2007) 26, 4627–4634; doi:10.1038/sj.onc.1210254; (Kumar et al., 1998; Iida et al., 2000; Randerson- published online 5 February 2007 Moor et al., 2001; Rizos et al., 2001; Viswanathan et al., 2001). The N-terminal portion (amino acids (aa) 1–64) Keywords: ARF; retinoblastoma protein; MDM2; cell of p14ARF is necessary and sufficient for growth arrest cycle; cancer (Quelle et al., 1997; Zhang et al., 1998). ARF functions to activate p53 through inhibition of MDM2 (Pomerantz et al., 1998; Zhang et al., 1998). ARF interacts with MDM2 in the central acidic domain (CAD) (Weber et al., 2000b; Bothner et al., 2001) and Introduction antagonizes MDM2 E3 ligase function, resulting in p53 stabilization (Kamijo et al., 1998; Pomerantz et al., The INK4a/ARF locus, located on chromosome 9p21, is 1998; Stott et al., 1998; Zhang et al., 1998). In fact, ARF one of the most frequently targeted loci for inactivation inhibits several known functions of MDM2 toward p53, in human cancer (Hall and Peters, 1996; Hainaut et al., including MDM2’s ability to block p53 1997). The INK4a/ARF encodes the p16INK4a and (Kamijo et al., 1998; Pomerantz et al., 1998; Stott et al., p14ARF tumor suppressor proteins. p16INK4a specifically 1998), and to ubiquitinate p53 (Honda and Yasuda, 1999; Midgley et al., 2000; Llanos et al., 2001). In addition, ARF sequesters MDM2 in the to Correspondence: Dr Z-XJ Xiao, Department of Biochemistry, Boston prevent nucleo-cytoplasmic shuttling (Tao and Levine, University School of Medicine, 715 Albany St., K-423, Boston, MA 1999; Weber et al., 1999, 2000b; Zhang and Xiong, 02118, USA. 1999). E-mail: [email protected] Received 5 June 2006; revised 30 November 2006; accepted 4 December A growing body of evidence suggests that ARF 2006; published online 5 February 2007 possesses growth suppressive properties independent of ARF upregulates Rb through inhibition of MDM2 DLF Chang et al 4628 p53. ARFÀ/À mice develop carcinomas and neurogenic (Figure 1a). Transfection efficiency was comparable as tumors rarely seen in p53À/À mice. In addition, c-Myc- shown by a comparable expression of co-transfected induced lymphomas are more aggressive in ARFÀ/À/ GFP. In addition, expression of p19ARF by retroviral p53À/À mice compared to either p53À/À or ARFÀ/À mice infection in H1299 cells also led to Rb accumulation. alone (Eischen et al., 1999). Triple knock mice Importantly, ARF led to a significant increase in nullizygous for ARF, p53 and Mdm2 develop multiple hypophosphorylated Rb (Figure 1b). tumors at higher frequency than that of mice lacking both p53 and Mdm2 or p53 alone (Weber et al., 2000a). Recently, it has been shown that tumors emerged p14ARF upregulates Rb in an MDM2-dependent manner significantly earlier in mice lacking both ARF and p53 Rb consists of the N-terminus, the central portion (A–B than in the mice lacking p53 alone (Christophorou et al., domain) and the C-pocket (Figure 2a). The A–B 2006). Notably, ARF can upregulate hypophosphory- domain, referred to as the Rb small pocket (RbSP), lated Rb and inhibits progression in p21À/À binds to several viral oncoproteins and a subset of cells (Modestou et al., 2001). These observations cellular proteins. The Rb large pocket (RbLP), contain- indicate that ARF growth suppression function can be ing the RbSP and C-pocket, is required for Rb growth achieved independent of the p53– pathway. suppressive function. As the Rb C-pocket binds to We have shown previously that MDM2 interacts with MDM2 (Xiao et al., 1995; Sdek et al., 2004), we Rb and inhibits its growth suppressive function (Xiao examined the role of Rb C-pocket in ARF-mediated Rb et al., 1995). MDM2 binds to Rb C-pocket, blocks Rb- upregulation. H1299 cells were co-transfected with E2F interaction and promotes cell cycle progression RbLP or RbSP and p14ARF expression plasmids. (Sdek et al., 2004). Recently, we demonstrated that Western-blot analysis showed that co-expression of MDM2 facilitates Rb interaction with the 20S protea- ARF led to a significant increase in RbLP, but not some resulting in accelerated Rb degradation (Sdek RbSP proteins (Figure 2b), suggesting that the Rb C- et al., 2005). pocket is critical for ARF effect on Rb. We next In this study, we investigated the role of ARF tumor examined whether ARF could directly inhibit MDM2- suppressor protein in the regulation of Rb. Our results mediated Rb degradation. As shown in Figure 2c, show that ARF inhibits MDM2-mediated Rb degrada- expression of ARF-stabilized MDM2, consistent with tion by disruption of MDM2–Rb interaction. In previous reports (Stott et al., 1998; Honda and Yasuda, addition, we show that lack of ARF led to reduced Rb 1999; Llanos et al., 2001). Importantly, whereas MDM2 protein levels independent of p53, and that ARF growth destabilized Rb, as expected, co-expression of ARF suppression function is impaired in the absence of Rb. led to concomitant MDM2 and Rb stabilization (Figure 2c), suggesting that ARF upregulates Rb through inhibition of MDM2. We then examined the effect of ARF on Rb in p53À/À murine embryonic Results fibroblast (MEF) and p53À/À/MDM2À/À cells. As shown in Figure 2d, ARF markedly induced Rb in p53À/À MEF ARF Expression of p14 increases Rb levels independent cells but failed to do so in p53À/À/MDM2À/ÀÀ/À cells. of p53 Taken together, these data suggest that ARF inhibits MDM2 has been shown to facilitate Rb degradation MDM2 and upregulates Rb. (Sdek et al., 2005; Uchida et al., 2005). Given the critical role of ARF in regulation of MDM2, we investigated whether ARF can modulate MDM2-mediated Rb ARF inhibits colony foci formation independent of p53 ARF degradation. Co-expression of p14 and Rb in p53- The N-terminus of ARF (ARF-N; aa 1–64) is critical for null H1299 cells led to a marked increase in Rb levels ARF-mediated growth suppressive function. Notably, both p14ARF and p19ARF N-terminal domains contain MDM2-binding modules (Zhang et al., 1998; Weber et al., 2000b). Thus, we examined the effects of wild type, ARF-N and ARF-C (aa 64–132) on the expression of MDM2 and Rb. Our results showed that whereas wild-type ARF and ARF-N promoted MDM2 expres- sion, keeping with its reported ability to block MDM2 E3 ligase activity (Figure 3a), they also led to the accumulation of Rb protein (Figure 3b). In contrast, ARF-C was unable to affect either MDM2 or Rb (Figure 3a and b). We next examined the effect of ARF Figure 1 ARF increases Rb protein levels independent of p53. and ARF mutants on colony foci formation. As shown (a) H1299 cells were co-transfected with p14ARF, myc-tagged Rb in Figure 3c, wild-type ARF as well as ARF-N (aa 1–64) (myc-Rb) and GFP expression plasmids. Cell lysates were analysed significantly suppressed colony foci formation of H1299 for expression of myc-tagged Rb, ARF, GFP and actin by Western-blot analyses. (b) H1299 cells were infected by retrovirus cells. In contrast, ARF-C was less efficient in suppres- encoding p19ARF or vector. Western-blot analyses were performed sing colony foci formation. These data, therefore, as indicated. indicate that ARF can inhibit colony foci formation

Oncogene ARF upregulates Rb through inhibition of MDM2 DLF Chang et al 4629

Figure 2 p14ARF upregulates Rb in an MDM2-dependent manner. (a) Schematic representation of Rb pocket domains in binding to MDM2 (Sdek et al., 2004). RbLP, Rb large pocket; RbSP, Rb small pocket. (b) H1299 cells were co-transfected with pCMV-RbLP or pCMV-RbSP plasmid in the presence of GFP expression plasmid and p14ARF or vector DNA. Equal amounts of total protein were separated on SDS–PAGE and subjected to Western-blot analysis. (c) C33A cells were co-transfected with MDM2, Rb, GFP and an increasing amount of myc-p14ARF expression plasmids (0.5, 1 and 2 mg), as indicated. Equal amounts of total protein were separated on SDS–PAGE and subjected to Western-blot analysis. (d) p53À/À or p53À/À/MDM2À/À MEF cells were infected with recombinant adenovirus encoding p14ARF or GFP for 24 h. Cell lysates were subjected to Western-blot analysis. independent of p53, which is associated with its ability protein in wild type and ARFÀ/À MEF cells. As shown in to induce Rb. Figure 4c, ARFÀ/À MEF cells had reduced Rb protein levels compared to wild-type MEF cells. Taken together, these data suggest that ARF stabilizes Rb protein ARF inhibits MDM2–Rb interaction through inhibition of MDM2–Rb association. ARF binds to MDM2 in the CAD and inhibits MDM2- mediated p53 ubiquitination and degradation (Weber et al., 2000b; Bothner et al., 2001). Notably, ARF can Rb is important in ARF-mediated growth inhibition modulate MDM2 interaction with its binding partners, To examine the role of Rb in ARF growth suppression including KAP1 (Wang et al., 2005) and YY1 (Sui et al., function, we infected U2-OS cells with retrovirus 2004). Since Rb interacts with MDM2 in the CAD encoding short hairpin RNAi for Rb or vector (Sdek et al., 2004), we examined whether ARF affects control (pMSCV). Knockdown of Rb was effective MDM2–Rb interaction. As shown in Figure 4a, (Figure 5a) and led to an increase in MDM2–Rb interaction was readily detected by co- (Figure 5b). Retroviral infection of p19ARF led to immunoprecipitation (IP) experiments (lane 3). Expres- significant growth inhibition (60%) in wild-type U2- sion of wild-type p14ARF led to a stable p14ARF–MDM2 OS cells. In contrast, p19ARF-induced cell growth complex formation and, importantly, a significant inhibition was significantly attenuated (23%) in cells reduction of MDM2–Rb interaction (lane 6). In expressing shRb (Figure 5b), indicating that Rb is contrast, ARF-C was unable to bind to MDM2 and essential for ARF-induced growth suppression. In had little effect on MDM2–Rb interaction (lane 9). In a addition, p19ARF significantly inhibited cell growth reverse IP-Western experiment, expression of ARF also in p53À/À MEF cells but not in p53À/Àmdm2À/À cells led to reduced MDM2–Rb interaction (Figure 4b). In (Figure 5c), demonstrating that MDM2 is critical for addition, we examined the expression levels of Rb ARF growth suppression independent of p53.

Oncogene ARF upregulates Rb through inhibition of MDM2 DLF Chang et al 4630

Figure 4 p14ARF inhibits MDM2–Rb interaction. H1299 cells were co-transfected with MDM2 and Rb in the presence of myc-p14ARF or myc-p14ARF-C expressing plasmids for 24 h and treated with Figure 3 p14ARF inhibits colony foci formation independent of MG132 for 14 h. Cell lysates were subjected to immunoprecipita- p53. H1299 cells were co-transfected with MDM2 (a) or myc- tion using a control antibody CD19, SMP14 (an antibody specific tagged Rb (b) in the presence of p14ARF, p14ARF-N or p14ARF-C for MDM2) (a), or XZ77 (an antibody specific for Rb) (b). The expression plasmids. Equal amounts of total protein were subjected immunoprecipitated protein complexes were subjected to Western- to Western-blot analyses, as indicated. (c) H1299 cells expressing blot analyses for Rb, MDM2 or myc-p14ARF, as indicated. Total p14ARF, p14ARF-N or p14ARF-C were grown for 21 days in DMEM protein (60 mg) from transfected H1299 cells was directly loaded on supplemented with 2% FBS and G418. Colonies were visualized by the gel as control (input) for Rb, MDM2 or p14ARF, respectively. staining with crystal violet. Total cell numbers were counted and (c) Wild-type or ARF-null MEFs were examined for Rb expres- presented as percentage of controls. Two independent experiments sion. Equal amounts of total protein (120 mg for Rb, 40 mg for in duplicates were performed. p19ARF and actin) were separated by SDS-PAGE and subjected to Western-blot analysis.

Discussion of human sarcomas (Momand et al., 1998). It is well documented that MDM2 functions in p53-dependent Several studies have suggested a functional interplay and -independent pathways. Recently, we demonstrated between ARF and Rb. MEF cells devoid of all three Rb that MDM2 binds to Rb at the C-terminus to promote family members are resistant to senescence and ARF- the proteasome-dependent Rb degradation (Sdek et al., induced cell growth arrest (Dannenberg et al., 2000; 2005). In this study, we demonstrated that ARF inhibits Sage et al., 2000). Furthermore, loss of ARF accelerates MDM2 resulting in stabilization of Rb. First, the Rb pituitary tumor development in Rb heterozygous mice C-pocket, a binding site for MDM2, is critical for ARF- (Rb þ /À/ARFÀ/À) and the effects are greater than those mediated Rb induction. Second, wild-type ARF and seen in Rb þ /À/p53À/À mice. Significantly, alteration of the ARF-N terminus capable of MDM2 interaction the ARF locus was also seen in tumors from Rb þ /À/ can stabilize Rb and inhibit cell growth. In contrast, ARF þ /À mice (Tsai et al., 2002), indicating a selective the ARF-C mutant unable to bind to MDM2 cannot advantage of ARF inactivation. Additionally, induction stabilize Rb and/or inhibit cell growth. Third, ARF fails of growth arrest by ARF in immortalized cells is to upregulate Rb in cells lacking MDM2. Fourth, ARF inhibited by simultaneous inactivation of both p53 and but not ARF-C inhibits MDM2–Rb interaction. More- Rb but not by inactivation of p53 alone (Carnero et al., over, loss of ARF leads to downregulation of Rb 2000), suggesting that ARF has p53-independent protein levels in MEF cells. Importantly, depletion of activity by acting, in part, through the Rb pathway. Rb attenuates ARF growth suppression function. Interestingly, ARF can inhibit E7-mediated Rb degra- Notably, although our data show that ablation of Rb dation (Pan et al., 2003). These observations collectively substantially impairs ARF growth suppression function, imply a functional link between ARF and Rb. it has been reported that inhibition of Rb by E7 in triple Overexpression of MDM2 is associated with a variety knockout MEF (p53À/À,Mdm2À/À and ARFÀ/À) cells of human tumors/cancers, including approximately 30% only marginally affects ARF-induced growth arrest

Oncogene ARF upregulates Rb through inhibition of MDM2 DLF Chang et al 4631 (Weber et al., 2000a). Although the reasons for this growth suppression function for ARF independent of discrepancy are unclear, it is possible that E7-mediated p53. Interestingly, since ARF is a downstream target of inhibition of Rb is affected by ARF, which has been E2F, activation of Rb would lead to inhibition of E2F reported by others (Pan et al., 2003). In addition, it activity resulting in downregulation of ARF expression. should be noted that Rb knockdown does not com- This feedback system would likely contribute to the pletely impairs ARF-induced cell growth arrest, indicat- tight control of ARF expression. ing that ARF can function independent of p53 and Rb c-Abl has been shown to phosphorylate MDM2 at pathways. tyrosine 394 and inhibit MDM2-mediated ubiquitina- Taken together, these data indicate that ARF induces tion and nuclear export of p53 (Sionov et al., 1999, 2001; Rb through the inhibition of MDM2, suggesting a novel Goldberg et al., 2002). Recently, c-Abl is reported to phosphorylate MDM2 at tyrosine 276 and promote ARF–MDM2 interaction and sequestration of MDM2 in the nucleolus (Dias et al., 2006). Notably, this site (Y276) is in proximity to the Rb- binding domain (aa 254–264) of MDM2. Thus, it is plausible that c-Abl might affect Rb levels by enhancing ARF–MDM2 interaction and MDM2 nucleolar locali- zation. Recently, it was reported that ARF promotes Rb accumulation through inhibition of Tip60-dependent (Leduc et al., 2006). Interestingly, acetyla- tion of Rb C-pocket at 873 and lysine 874 appears to enhance MDM2–Rb interaction (Chan et al., 2001; Nguyen et al., 2004). Therefore, it is plausible that acetylation of Rb C-pocket may enhance or stabilize MDM2–Rb interaction resulting in accelerated Rb degradation. Although our data favor the possibility that ARF enhances Rb protein stability by direct inhibition of MDM2, it is possible that ARF can also modulate MDM2–Rb interaction through inhibition of Rb acetylation. Further experiments are needed to address these possibilities. It is clear that Rb is regulated at multiple levels. In addition to gene , such as deletion and point mutations that disrupt Rb structure and function, modulation of Rb protein phosphorylation remains a key regulatory mechanism for Rb function. Recent studies indicate that regulation of Rb protein stability appears to be an important neoplastic strategy (Ying and Xiao, 2006), as exemplified by human papilloma viral oncoprotein E7 (Boyer et al., 1996; Berezutskaya et al., 1997; Wang et al., 2001), cellular oncoproteins gankyrin (Higashitsuji et al., 2000) and MDM2 (Sdek et al., 2005; Uchida et al., 2005). This study further

Figure 5 ARF-mediated cell growth inhibition is impaired in the absence of Rb. (a) U2-OS cells were infected with retroviral pMSCV vector or pMSCV-RbshRNA. Cell lysates were examined for Rb protein levels by Western-blot analysis. (b) Stable U2-OS cells expressing shRb or pMSCV vector were superinfected with concentrated pBabe-p19ARF retrovirus or pBabe vector alone. At 24 h after infection, 25 000 cells were reseeded on p35 plates and allowed to grow for 3 days. Total numbers of cells were counted (b) and also presented as percentage of growth inhibition after normalization to control cells infected by pBabe vector alone (b, lower panel). Expression of ARF was examined by Western-blot analysis (inset). (c) p53À/À and p53À/À mdm2À/À MEF cells were infected with pBabe-p19ARF retrovirus or pBabe vector alone. At 24 h after infection, 25 000 cells were reseeded on p35 plates and allowed to grow for 3 days. Total numbers of cells were counted using a hemacytometer. Two independent experiments in dupli- cates were performed. Expression of p19ARF was analysed by Western-blot analysis (inset).

Oncogene ARF upregulates Rb through inhibition of MDM2 DLF Chang et al 4632 supports the notion that regulation of Rb protein Transfection and adenovirus/retrovirus infection stability is important in . It is conceivable Transient transfection was carried out using Lipofecta- that inactivation of ARF leads to disruption of p53 and mine2000 transfection reagent (Invitrogen) or calcium phos- Rb pathways both of which play a critical role in human phate precipitation (Clontech Laboratories, Inc., Palo Alto, cancer development. CA, USA). Equal amounts of total DNA plasmids were obtained by supplementing empty expression vector DNA. Cells were harvested 24 or 48 h after transfection. Adenoviral-p14ARF infection was performed as described Materials and methods previously (Sdek et al., 2005) in 2% FBS/DMEM for 2 h in a 371C incubator with 5% CO2. Infected cells were then washed Cell culture and plasmids with PBS and replaced with fresh 12% FBS/DMEM. H1299 and C33A cells were maintained in Dulbecco’s modified For high-efficiency retroviral infection, concentrated retro- Eagles medium (DMEM), supplemented with 10% fetal viruses (viral titers of 108–109 cfu/ml) were prepared as bovine serum (FBS) and 1% penicillin G/streptomycin sulfate described (Burns et al., 1993). Briefly, 293FT cells were (Invitrogen, Carlsbad, CA, USA) at 371C in a humidified 5% transiently transfected in triplicate using retroviral DNA and À/À CO2 incubator. Wild type and ARF (provided by Dr accessory plasmids by Lipofectamine2000. At 12 h after Charles Sherr), p53À/À and p53À/À/MDM2À/À (provided by Dr transfection, the media were changed; 48 h later, the super- Guillermina Lozano) MEF cells were maintained in DMEM natant was collected and centrifuged at 3000 r.p.m. for 10 min supplemented with 12% FBS, 1% penicillin G/streptomycin to remove debris and filtered through a 0.45 mM filter. The sulfate, 2 mML-glutamine (Invitrogen) and 100 mM MEM non- retroviral particles were then concentrated by ultra-centrifuga- essential amino acids (Invitrogen). tion (27 000 r.p.m., 1 h 45 min at 41C), resuspended in fresh pcDNA-p14ARF-N and pcDNA-p14ARF-C plasmids were pro- media supplemented with polybrene (10 mg/ml). Cells were vided by Dr Yanping Zhang. pBABE-puro-p19ARF and selected in growth media supplemented with puromycin (2 mg/ pMSCV-puro-RbshRNA were provided by Dr Ronald A ml) 24 h after infection. DePinho and Dr Scott Lowe, respectively.

Western-blot analyses and immunoprecipitation experiments Colony foci formation assay and cell growth analysis For Western-blot analyses, cells were washed with cold PBS For colony foci formation, H1299 cells were transiently transfected with p14ARF expression plasmid or vector using and lysed in EBC250 lysis buffer (50 mM Tris-HCl, pH 8.0, Lipofectamine2000. At 24 h after transfection, 10 000 cells 250 mM NaCl, 0.5% Nonidet P-40, 1 mM PMSF, 2 mg/ml leupeptin and 2 mg/ml aprotinin). Equal amounts of protein were reseeded in duplicate on 60 mm plates and grown in 2% (20–100 mg) were separated on sodium dodecyl sulfate–poly- FBS/DMEM supplemented with G418 (800 ng/ml) (Invitro- acrylamide gel electrophoresis (SDS-PAGE) gels, transferred gen). The G418-containing media was changed every 3 days. to polyvinylidene difluoride membrane and hybridized to an About 18–21 days after selection, cells were rinsed with  1 appropriate primary antibody and horseradish peroxidase- PBS and then stained with 0.5% crystal violet (C-3886, Sigma- conjugated secondary antibody for subsequent detection by Aldrich, St Louis, MO, USA) solution in 70% ethanol. enhanced chemiluminescence. Another set of plates was trypsinized and cell numbers were For IP-Western analyses, H1299 cells were transfected with counted using a hemocytometer. plasmids using Lipofectamine2000 (Invitrogen). About 24 h For cell growth analysis, U2-OS or MEF cells were infected with concentrated pBABE-puro or pBABE-p19ARF retro- after transfection, cells were treated with 20 mM MG132 for 14 h and lysed in EBC150 buffer (same as EBC250 except virus. At 24 h after infection, 25 000 infected cells were reseeded onto 35 mm dishes and grown for 3 days in 10% 150 mM NaCl). Whole-cell lysates were precleared with 75 mlof protein A/G slurry (Santa Cruz Biotechnology, Santa Cruz, FBS/DMEM supplemented with puromycin (0.5 mg/ml). Cell CA, USA). Total protein (1.5 mg) was incubated with 4 mg numbers were quantified using a hemocytometer. At least two of antibody specific for MDM2 (SMP14), myc-tag (9E10) independent experiments in duplicates were performed. or control antibody against CD19 at 41C overnight. Immuno- protein complexes were resolved by SDS-PAGE followed by Western-blot analyses. Acknowledgements Specific antibodies were used to detect MDM2 (SMP14, Santa Cruz Biotechnology), Rb (G3-245, BD PharMingen), We thank Drs Ronald A DePinho, Scott Lowe and Yanping underphosphorylated Rb (G99-529, BD PharMingen), RbLP Zhang for plasmids and viruses. We are grateful to Drs and RbSP proteins (a mixture of monoclonal antibodies XZ91 Charles Sherr and Guillermina Lozano for MEF cells. We and XZ77, gift of Dr N Dyson), p19ARF (ab80, Abcam), thank Yian Xiao for critical reading of the manuscript. This p14ARF (C-18, Santa Cruz Biotechnology), GFP (FL, Santa work was supported by the Karin Grunebaum Cancer Cruz Biotechnology), b-galactosidase (Z378A, Promega, Research Fellowship and the Boston University Graduate Madison, WI, USA), actin (C-11, Santa Cruz Biotechnology) Student Research Fellowship (to DLC) and NIH grants (to and myc (9E10, Santa Cruz Biotechnology). ZX Xiao).

References

Berezutskaya E, Yu B, Morozov A, Raychaudhuri P, Bothner B, Lewis WS, DiGiammarino EL, Weber JD, Bothner SJ, Bagchi S. (1997). Differential regulation of the pocket Kriwacki RW. (2001). Defining the molecular basis of Arf domains of the retinoblastoma family proteins by and Hdm2 interactions. J Mol Biol 314: 263–277. the HPV16 E7 oncoprotein. Cell Growth Differ 8: Boyer SN, Wazer DE, Band V. (1996). E7 protein of human 1277–1286. papilloma virus-16 induces degradation of retinoblastoma

Oncogene ARF upregulates Rb through inhibition of MDM2 DLF Chang et al 4633 protein through the –proteasome pathway. Cancer Kumar R, Sauroja I, Punnonen K, Jansen C, Hemminki K. Res 56: 4620–4624. (1998). Selective deletion of exon 1 beta of the p19ARF gene Burns JC, Friedmann T, Driever W, Burrascano M, Yee JK. in metastatic melanoma cell lines. Chromosomes (1993). Vesicular stomatitis virus G glycoprotein pseudo- Cancer 23: 273–277. typed retroviral vectors: concentration to very high titer and Leduc C, Claverie P, Eymin B, Col E, Khochbin S, Brambilla E efficient gene transfer into mammalian and nonmammalian et al. (2006). p14(ARF) promotes RB accumulation through cells. Proc Natl Acad Sci USA 90: 8033–8037. inhibition of its Tip60-dependent acetylation. Oncogene 25: Carnero A, Hudson JD, Price CM, Beach DH. (2000). 4147–4154. p16INK4A and p19ARF act in overlapping pathways in Llanos S, Clark PA, Rowe J, Peters G. (2001). Stabilization of cellular immortalization. Nat Cell Biol 2: 148–155. p53 by p14ARF without relocation of MDM2 to the Chan HM, Krstic-Demonacos M, Smith L, Demonacos C, nucleolus. Nat Cell Biol 3: 445–452. La Thangue NB. (2001). Acetylation control of the Midgley CA, Desterro JM, Saville MK, Howard S, Sparks A, retinoblastoma tumour-suppressor protein. Nat Cell Biol 3: Hay RT et al. (2000). An N-terminal p14ARF peptide 667–674. blocks Mdm2-dependent ubiquitination in vitro and can Christophorou MA, Ringshausen I, Finch AJ, Swigart LB, activate p53 in vivo. Oncogene 19: 2312–2323. Evan GI. (2006). The pathological response to DNA damage Modestou M, Puig-Antich V, Korgaonkar C, Eapen A, does not contribute to p53-mediated tumour suppression. Quelle DE. (2001). The alternative reading frame tumor Nature 443: 214–217. suppressor inhibits growth through p21-dependent and p21- Dannenberg JH, van Rossum A, Schuijff L, te Riele H. (2000). independent pathways. Cancer Res 61: 3145–3150. Ablation of the retinoblastoma gene family deregulates Momand J, Jung D, Wilczynski S, Niland J. (1998). The G(1) control causing immortalization and increased cell MDM2 gene amplification database. Nucleic Acids Res 26: turnover under growth-restricting conditions. Genes Dev 14: 3453–3459. 3051–3064. Nguyen DX, Baglia LA, Huang SM, Baker CM, McCance DJ. Dias SS, Milne DM, Meek DW. (2006). c-Abl phosphorylates (2004). Acetylation regulates the differentiation-specific Hdm2 at tyrosine 276 in response to DNA damage and functions of the retinoblastoma protein. EMBO J 23: regulates interaction with ARF. Oncogene 25: 6666–6671. 1609–1618. Eischen CM, Weber JD, Roussel MF, Sherr CJ, Cleveland JL. Pan W, Datta A, Adami GR, Raychaudhuri P, Bagchi S. (1999). Disruption of the ARF-Mdm2-p53 tumor suppressor (2003). P19ARF inhibits the functions of the HPV16 E7 pathway in Myc-induced lymphomagenesis. Genes Dev 13: oncoprotein. Oncogene 22: 5496–5503. 2658–2669. Pomerantz J, Schreiber-Agus N, Liegeois NJ, Silverman A, Goldberg Z, Vogt Sionov R, Berger M, Zwang Y, Perets R, Alland L, Chin L et al. (1998). The Ink4a tumor suppressor Van Etten RA et al. (2002). Tyrosine phosphorylation of gene product, p19Arf, interacts with MDM2 and neutralizes Mdm2 by c-Abl: implications for p53 regulation. EMBO J MDM2’s inhibition of p53. Cell 92: 713–723. 21: 3715–3727. Quelle DE, Cheng M, Ashmun RA, Sherr CJ. (1997). Groth A, Weber JD, Willumsen BM, Sherr CJ, Roussel MF. Cancer-associated mutations at the INK4a locus cancel (2000). Oncogenic Ras induces p19ARF and growth arrest cell cycle arrest by p16INK4a but not by the alternative in mouse embryo fibroblasts lacking p21Cip1 and p27Kip1 reading frame protein p19ARF. Proc Natl Acad Sci USA 94: without activating -dependent kinases. J Biol Chem 669–673. 275: 27473–27480. Randerson-Moor JA, Harland M, Williams S, Cuthbert- Hainaut P, Soussi T, Shomer B, Hollstein M, Greenblatt M, Heavens D, Sheridan E, Aveyard J et al. (2001). A germline Hovig E et al. (1997). Database of p53 gene somatic deletion of p14(ARF) but not CDKN2A in a melanoma– mutations in human tumors and cell lines: updated neural system tumour syndrome family. Hum Mol Genet 10: compilation and future prospects. Nucleic Acids Res 25: 55–62. 151–157. Rizos H, Puig S, Badenas C, Malvehy J, Darmanian AP, Hall M, Peters G. (1996). Genetic alterations of , cyclin- Jimenez L et al. (2001). A melanoma-associated germline dependent kinases, and Cdk inhibitors in human cancer. Adv in exon 1beta inactivates p14ARF. Oncogene 20: Cancer Res 68: 67–108. 5543–5547. Higashitsuji H, Itoh K, Nagao T, Dawson S, Nonoguchi K, Sage J, Mulligan GJ, Attardi LD, Miller A, Chen S, Williams B Kido T et al. (2000). Reduced stability of retinoblastoma et al. (2000). Targeted disruption of the three Rb-related protein by gankyrin, an oncogenic ankyrin-repeat protein genes leads to loss of G(1) control and immortalization. overexpressed in hepatomas. Nat Med 6: 96–99. Genes Dev 14: 3037–3050. Honda R, Yasuda H. (1999). Association of p19(ARF) with Sdek P, Ying H, Chang DL, Qiu W, Zheng H, Touitou R et al. Mdm2 inhibits activity of Mdm2 for tumor (2005). MDM2 promotes proteasome-dependent ubiquitin- suppressor p53. EMBO J 18: 22–27. independent degradation of retinoblastoma protein. Mol Iida S, Akiyama Y, Nakajima T, Ichikawa W, Nihei Z, Cell 20: 699–708. Sugihara K et al. (2000). Alterations and hypermethylation Sdek P, Ying H, Zheng H, Margulis A, Tang X, Tian K et al. of the p14(ARF) gene in gastric cancer. Int J Cancer 87: (2004). The central acidic domain of MDM2 is critical in 654–658. inhibition of Rb-mediated suppression of E2F and cell Kamijo T, Weber JD, Zambetti G, Zindy F, Roussel MF, growth. J Biol Chem 279: 53317–53322. Sherr CJ. (1998). Functional and physical interactions of the Serrano M. (2000). The INK4a/ARF locus in murine ARF tumor suppressor with p53 and Mdm2. Proc Natl Acad tumorigenesis. Carcinogenesis 21: 865–869. Sci USA 95: 8292–8297. Serrano M, Lee H, Chin L, Cordon-Cardo C, Beach D, Kamijo T, Zindy F, Roussel MF, Quelle DE, Downing JR, DePinho RA. (1996). Role of the INK4a locus in tumor Ashmun RA et al. (1997). Tumor suppression at the mouse suppression and cell mortality. Cell 85: 27–37. INK4a locus mediated by the alternative reading frame Sharpless NE, Ramsey MR, Balasubramanian P, Castrillon DH, product p19ARF. Cell 91: 649–659. DePinho RA. (2004). The differential impact of p16(INK4a)

Oncogene ARF upregulates Rb through inhibition of MDM2 DLF Chang et al 4634 or p19(ARF) deficiency on cell growth and tumorigenesis. Wang C, Ivanov A, Chen L, Fredericks WJ, Seto E, Oncogene 23: 379–385. Rauscher III FJ et al. (2005). MDM2 interaction with Sherr CJ. (2001). The INK4a/ARF network in tumour nuclear corepressor KAP1 contributes to p53 inactivation. suppression. Nat Rev Mol Cell Biol 2: 731–737. EMBO J 24: 3279–3290. Sionov RV, Coen S, Goldberg Z, Berger M, Bercovich B, Wang J, Sampath A, Raychaudhuri P, Bagchi S. (2001). Both Ben-Neriah Y et al. (2001). c-Abl regulates p53 levels under Rb and E7 are regulated by the ubiquitin proteasome normal and stress conditions by preventing its nuclear pathway in HPV-containing cervical tumor cells. Oncogene export and ubiquitination. Mol Cell Biol 21: 5869–5878. 20: 4740–4749. Sionov RV, Moallem E, Berger M, Kazaz A, Gerlitz O, Weber JD, Jeffers JR, Rehg JE, Randle DH, Lozano G, Ben-Neriah Y et al. (1999). c-Abl neutralizes the inhibitory Roussel MF et al. (2000a). p53-independent functions of the effect of Mdm2 on p53. J Biol Chem 274: 8371–8374. p19(ARF) tumor suppressor. Genes Dev 14: 2358–2365. Stott FJ, Bates S, James MC, McConnell BB, Starborg M, Weber JD, Kuo ML, Bothner B, DiGiammarino EL, Brookes S et al. (1998). The alternative product from the Kriwacki RW, Roussel MF et al. (2000b). Cooperative human CDKN2A locus, p14(ARF), participates in a signals governing ARF–mdm2 interaction and nucleolar regulatory feedback loop with p53 and MDM2. EMBO J localization of the complex. Mol Cell Biol 20: 2517–2528. 17: 5001–5014. Weber JD, Taylor LJ, Roussel MF, Sherr CJ, Bar-Sagi D. Sui G, Affar el B, Shi Y, Brignone C, Wall NR, Yin P et al. (1999). Nucleolar Arf sequesters Mdm2 and activates p53. (2004). Yin Yang 1 is a negative regulator of p53. Cell 117: Nat Cell Biol 1: 20–26. 859–872. Weinberg RA. (1995). The retinoblastoma protein and cell Tao W, Levine AJ. (1999). P19(ARF) stabilizes p53 by cycle control. Cell 81: 323–330. blocking nucleo-cytoplasmic shuttling of Mdm2. Proc Natl Xiao ZX, Chen J, Levine AJ, Modjtahedi N, Xing J, Acad Sci USA 96: 6937–6941. Sellers WR et al. (1995). Interaction between the retino- Tsai KY, MacPherson D, Rubinson DA, Nikitin AY, Bronson R, blastoma protein and the oncoprotein MDM2. Nature 375: Mercer KL et al. (2002). ARF mutation accelerates pituitary 694–698. tumor development in Rb+/À mice. Proc Natl Acad Sci Ying H, Xiao ZX. (2006). Targeting retinoblastoma protein USA 99: 16865–16870. for degradation by . Cell Cycle 5: 506–508. Uchida C, Miwa S, Kitagawa K, Hattori T, Isobe T, Otani S Zhang Y, Xiong Y. (1999). Mutations in human ARF exon 2 et al. (2005). Enhanced Mdm2 activity inhibits pRB disrupt its nucleolar localization and impair its ability to function via ubiquitin-dependent degradation. EMBO J 24: block nuclear export of MDM2 and p53. Mol Cell 3: 160–169. 579–591. Viswanathan M, Tsuchida N, Shanmugam G. (2001). Selective Zhang Y, Xiong Y, Yarbrough WG. (1998). ARF promotes deletion of p14(ARF) exon 1beta of the INK4a locus in MDM2 degradation and stabilizes p53: ARF-INK4a locus oral squamous cell carcinomas of Indians. Oral Oncol 37: deletion impairs both the Rb and p53 tumor suppression 341–344. pathways. Cell 92: 725–734.

Oncogene