Loss of One Allele of ARF Rescues Mdm2 Haploinsufficiency Effects On

Loss of one allele of ARF rescues Mdm2 haploinsufﬁciency effects on apoptosis and lymphoma development

Christine M Eischen*,1,2, Jodi R Alt1,2 and Peng Wang1

1Eppley Institute for Research in Cancer, University of Nebraska Medical Center, Omaha, NE 68198, USA; 2Department of Pathology and Microbiology, University of Nebraska Medical Center, Omaha, NE 68198, USA

The tumor suppressor p19ARF inhibits Mdm2, which ARF or p53is inactivated in over half of these tumors restricts the activity of p53. Complicated feedback and (Eischen et al., 1999). In addition, in lymphomas that control mechanisms regulate ARF, Mdm2, and p53 emerge in ARF þ /À or p53 þ /ÀEm-myc transgenics, the interactions. Here we report that ARF haploinsufficiency second allele of ARF or p53 is deleted in 77 or 100% of completely rescued the p53-dependent effects of Mdm2 these tumors, respectively (Eischen et al., 1999; Schmitt haploinsufficiency on B-cell development, survival, and et al., 1999). Therefore, ARF and p53guard against transformation. In contrast to Mdm2 þ /À B cells, Mdm2 þ /À oncogene-initiated tumorigenesis by activating apoptosis B cells deficient in ARF were similar to wild-type B cells in and consequently, are frequently targeted for inactiva- their rates of growth and apoptosis and activation of p53. tion in lymphomas that overexpress oncogenes. Consequently, the profoundly reduced numbers of B cells Mdm2 is a key intermediary in the ARF–p53tumor in Mdm2 þ /ÀEl-myc transgenic mice were restored to suppressor pathway. Mdm2 functions as an E3ubiqui- normal levels in ARF þ /ÀMdm2 þ /ÀEl-myc transgenics. tin ligase, and inactivates p53by ubiquitylating and Additionally, ARF þ /ÀMdm2 þ /ÀEl-myc transgenics devel- targeting it for degradation by the proteasome (Honda oped lymphomas at rates analogous to those observed for et al., 1997; Freedman and Levine, 1998; Roth et al., wild-type El-myc transgenics, demonstrating that loss of 1998). Mdm2 can also bind to p53and block its one allele of ARF rescued the protracted lymphoma transactivation functions (Momand et al., 1992). ARF is latency in Mdm2 þ /ÀEl-myc transgenics. Importantly, in a nucleolar protein and regulates p53by binding to ARF þ /ÀMdm2 þ /ÀEl-myc transgenic lymphomas, p53 was Mdm2 and blocking Mdm2’s ability to ubiquitylate p53 inactivated at the frequency observed in lymphomas of and to inhibit p53transcription functions (Kamijo et al., wild-type El-myc transgenics. Collectively, these results 1998; Pomerantz et al., 1998; Stott et al., 1998; Zhang support a model whereby the stoichiometry of Mdm2 and et al., 1998; Honda and Yasuda, 1999). Whereas ARF ARF controls apoptosis and tumor development, which or p53is commonly inactivated in cancers, Mdm2 should have significant implications in the treatment of is frequently overexpressed in murine and human malignancies that have inactivated ARF. malignancies, including lymphomas (Watanabe et al., Oncogene (2004) 23, 8931–8940. doi:10.1038/sj.onc.1208052 1996; Momand et al., 1998; Eischen et al., 1999). Thus, Published online 27 September 2004 ARF, Mdm2, and p53function together to suppress tumor development. Keywords: Mdm2; ARF; lymphoma; apoptosis; Myc; In vitro and in vivo experiments have shown that p53 Mdm2 is necessary to regulate p53activity during development and under stressful conditions, such as oncogene overexpression and following DNA damage. Firstly, early embryonic lethality of Mdm2-null embryos is rescued by loss of p53 (Jones et al., 1995; Montes de Introduction Oca Luna et al., 1995). Secondly, deleting p53 blocks Mdm2 haploinsufficient B cells from undergoing spon- The tumor suppressors p19ARF and p53are essential in taneous and Myc-induced apoptosis (Alt et al., 2003). blocking lymphoma development initiated by oncogenes. Thirdly, mice engineered to express low levels of Mdm2 ARF and p53mediate Myc-induced apoptosis in B cells, (hypomorphic) are more sensitive to apoptosis and and consequently loss of ARF or p53 accelerates Myc- death induced by gamma radiation, which is rescued by induced B-cell lymphomagenesis (Eischen et al., 1999; loss of p53 (Mendrysa et al., 2003). Finally, lymphomas Schmitt et al., 1999). Furthermore, in the lymphomas that arise in Mdm2 þ /ÀEm-myc transgenic mice preferen- that arise in Em-myc transgenic mice, which overexpress tially harbor mutations that inactivate p53 (Alt et al., Myc in the B-cell compartment (Adams et al., 1985), 2003). Therefore, a certain threshold of Mdm2 is required to harness p53activity under stressful condi- *Correspondence: CM Eischen; E-mail: [email protected] tions, and failure to do so results in apoptosis. Received 7 April 2004; revised 23June 2004; accepted 24 July 2004; Although at face value the p53tumor suppressor 10.1038/sj.onc.1208052; published online 27 September 2004 pathway appears linear, complicated feedback control Stoichiometry of ARF and Mdm2 regulates lymphomagenesis CM Eischen et al 8932 mechanisms are also operational in the pathway (Wu et al., 1993; Stott et al., 1998). For example, Mdm2 is a direct transcription target upregulated by p53(Barak et al., 1993; Juven et al., 1993; Perry et al., 1993; Wu et al., 1993), whereas ARF expression is elevated in cells lacking p53 or in cells harboring p53 mutations, and this appears to be Mdm2-independent (Kamijo et al., 1998; Zindy et al., 1998). However, Mdm2 expression can also influence ARF function in certain situations. For example, ARF overexpression fails to induce a cell cycle arrest in p53-null cells (Kamijo et al., 1997), yet is capable of doing so in cells lacking both p53 and Mdm2 (Weber et al., 2000). Furthermore, in fibroblasts and B cells, the induction of p53by oncogenes requires ARF (Robertson and Jones, 1998; Stott et al., 1998; Zindy et al., 1998; Eischen et al., 1999), yet nucleolar sequestration of Mdm2 by ARF is not necessarily required for p53activation (Korgaonkar et al., 2002). Collectively, these findings underscore the complexity of the ARF/Mdm2/p53network and, in particular, the consequences of ARF : Mdm2 interactions on cell survival and transformation are not resolved. Here we report that loss of only one allele of ARF fully compensates for the effects of Mdm2 haploinsufficiency in B-cell development, survival, and transformation in Em-myc transgenic mice. The results support the model whereby the stoichiometry of the ARF–Mdm2 complex is a critical arbiter of cell survival and transformation. Figure 1 Loss of ARF confers resistance to apoptosis to Mdm2 heterozygous pre-B cells. Prior to any detectable disease, bone marrow from two ARFÀ/À mice (triangles), two ARFÀ/ÀMdm2 þ /À mice (circles), two wild-type mice (squares), and two Mdm2 þ /À mice Results (crosses) was placed into IL-7-containing medium on day 0. Cells were counted at the indicated intervals, and net population doublings of the pre-B cells was calculated. Trypan blue dye þ /À ARF loss rescues the spontaneous apoptosis of Mdm2 exclusion was used to determine viability and propidium iodide pre-B cells (PI) staining followed by flow cytometry verified cell death was apoptosis Bone marrow-derived primary Mdm2 þ /À pre-B cells undergo spontaneous apoptosis in tissue culture and therefore do not grow (Alt et al., 2003), whereas pre-B rescued by loss of ARF,asARFÀ/ÀMdm2 þ /À bone cells lacking ARF are resistant to spontaneous apoptosis marrow cultures consistently had low apoptotic indices and can proliferate indefinitely (Eischen et al., 1999; (>86% viable at day 18) similar to those of wild-type Randle et al., 2001). p53mediates the spontaneous cultures (>89% viable at day 18). Therefore, ARF loss apoptosis of Mdm2 þ /À pre-B cells (Alt et al., 2003), yet fully compensates for the apoptosis sensitivity conferred the influence ARF had on this process was unclear. We by Mdm2 haploinsufficiency in pre-B cells. Furthermore, therefore evaluated whether loss of ARF would influence as the rates of growth of wild-type and ARFÀ/ÀMdm2 þ /À the survival and growth of Mdm2 þ /À pre-B cells in vitro. pre-B cells were similar, these results also suggest that Bone marrow from ARFÀ/À, Mdm2 þ /À, ARFÀ/ÀMdm2 þ /À, Mdm2 haploinsufficiency restores normal proliferative and wild-type littermate mice was placed into IL-7- rates even in cells lacking ARF. containing medium, which selects for pre-B cells that emerge within 14 days (Eischen et al., 1999). As Loss of one allele of ARF inhibits the increased p53 previously reported, ARF-null pre-B cells grew more activity in Mdm2 þ /À lymphocytes rapidly than wild-type pre-B cells (Figure 1; Eischen et al., 1999). In contrast to Mdm2 þ /À pre-B cells, ARFÀ/À Previously, we established that p53mediated the Mdm2 þ /À pre-B cells emerged from bone marrow and spontaneous apoptosis of Mdm2 haploinsufficient B proliferated as well as wild-type pre-B cells (Figure 1). To cells (Alt et al., 2003). In addition, thymocytes from determine whether the differences in the rates of mice with an Mdm2 hypomorphic allele showed an proliferation were attributable to differences in apopto- increase in p53transcriptional activity (Mendrysa et al., sis, we measured cell viability and DNA fragmentation in 2003). To determine whether a deficiency in ARF rescues the bone marrow cultures. As anticipated, the high rates apoptosis of Mdm2 þ /À B cells by inhibiting p53, we of spontaneous apoptosis of Mdm2 þ /À bone marrow cells evaluated p53induction and activity following gamma (o29% viable at day 18; Alt et al., 2003) were fully irradiation of mice with both alleles or only one allele of

Oncogene Stoichiometry of ARF and Mdm2 regulates lymphomagenesis CM Eischen et al 8933 Mdm2 and/or ARF. As expected, the levels of p53and the p53transcriptional target p21 in unirradiated splenocytes from wild-type and Mdm2 þ /À mice were below the level of detection (Figure 2a). Additionally, steady-state levels of Mdm2 were reduced in Mdm2 þ /À splenocytes when compared to splenocytes from wild- type mice (Figure 2a). Following gamma irradiation, p53protein is stabilized and activated resulting in increased levels of the p53transcriptional targets Mdm2 and p21 in all genotypes analyzed (Figures 2a and b). Specifically, Mdm2 protein was upregulated by gamma irradiation in splenocytes from Mdm2 þ /À and þ /À þ /À ARF Mdm2 mice, but not to the same extent as in þ /À þ /À Figure 2 The increased p53transcriptional activity in Mdm2 irradiated wild-type and ARF splenocytes (Figures 2a lymphocytes is inhibited by loss of a single allele of ARF.(a, b) and b). In contrast, the levels of p21 were higher in Wild-type (WT), Mdm2 þ /À (M þ /À), ARF þ /À (A þ /À), and ARF þ /À splenocytes from irradiated Mdm2 þ /À mice when com- Mdm2 þ /À (A þ /ÀM þ /À) mice were gamma irradiated with 10 Gy pared to p21 levels in splenocytes from irradiated wild- (IR þ in panel a and all samples in panel b) or left unirradiated type and ARF þ /À mice (Figures 2a and b). Importantly, (IRÀ in panel a). At 4 h following irradiation, spleens were collected þ /À from all mice. Protein lysates of splenocytes were Western blotted the increased levels of p21 in irradiated Mdm2 with antibodies specific for p53, p21, and Mdm2. Asterisks in (a) splenocytes were consistently reduced in mice haplo- indicate the location of nonspecific bands. Three separate experi- insufficient for both Mdm2 and ARF (Figure 2b). ments with different litters of mice are denoted by numbers in (b) The increased p53activity in Mdm2 þ /À splenocytes was not accompanied by an increase in p53protein levels, as all genotypes had similar levels of p53 following irradiation (Figures 2a and b). This result is consistent with experiments on thymocytes that have an Mdm2 hypomorphic allele (Mendrysa et al., 2003). Therefore, loss of one allele of ARF inhibits p53 transcriptional activity in Mdm2 þ /À lymphocytes without altering the levels of p53.

ARF haploinsufﬁciency impairs accelerated rates of Myc- induced apoptosis in Mdm2 þ /À B cells Resistance to p53- and ARF-mediated apoptosis is a rate-limiting step in Myc-induced lymphomagenesis (Eischen et al., 1999; Schmitt et al., 1999), and loss of one allele of ARF reduces p53activity in Mdm2 þ /À lymphocytes. To address the effects of ARF loss on Myc-induced apoptosis in the context of Mdm2 heterozygosity, we isolated bone marrow from Em-myc transgenic mice heterozygous for Mdm2 and/or ARF prior to lymphoma onset. Bone marrow from ARF þ /À Em-myc transgenics explanted into tissue culture (day 0) grew as well as bone marrow from wild-type Em-myc transgenic mice (Figure 3). As previously reported, Mdm2 þ /ÀEm-myc transgenic bone marrow was unable to grow ex vivo (Figure 3), due to very high rates of spontaneous apoptosis (Alt et al., 2003). In contrast, ARF þ /ÀMdm2 þ /ÀEm-myc transgenic bone marrow did survive explantation and proliferate, with pre-B cells emerging in 9 days (Figure 3). However, the ARF þ /À Mdm2 þ /ÀEm-myc pre-B cells never grew as well as wild- Figure 3 The sensitivity of Mdm2 heterozygous bone marrow to type or ARF þ /À Em-myc transgenic pre-B cells, due to Myc-induced apoptosis ex vivo is partially rescued by loss of one allele of ARF. Prior to any detectable lymphoma, bone marrow the higher rates of spontaneous apoptosis in these from one ARF þ /ÀEm-myc transgenic mouse (crosses), two wild-type cultures, which were consistently o75% viable. Due to Em-myc transgenics (squares), two ARF þ /ÀMdm2 þ /ÀEm-myc trans- the rapid development of lymphoma in ARFÀ/ÀMdm2 þ /À genics (circles), and two Mdm2 þ /ÀEm-myc transgenic mice (trian- Em-myc transgenics (see below), transformed precursor gles) was placed into IL-7-containing medium on day 0. At the indicated intervals, cells were counted, and pre-B-cell growth was B cells were always present in the bone marrow cultures, calculated as net population doublings. Trypan blue dye exclusion which excluded the mice from these and other analyses. was used to determine viability and apoptosis was measured by PI It should also be noted that although the wild-type, staining followed by ﬂow cytometry

Oncogene Stoichiometry of ARF and Mdm2 regulates lymphomagenesis CM Eischen et al 8934 ARF þ /À, and ARF þ /ÀMdm2 þ /ÀEm-myc transgenic bone marrow explants proliferated in the first 15 days, all of the cells died by day 25 irrespective of genotype due to Myc transgene expression, as previously reported for wild-type Em-myc transgenic bone marrow cells (Eischen et al., 1999). Therefore, Mdm2 haploinsufficient B cells are particularly prone to undergo apoptosis ex vivo especially when Myc is overexpressed, and loss of a single allele of ARF only partially rescues the combined effects of Myc overexpression and stress from tissue culture conditions. To determine whether a deficiency in ARF rescues Mdm2 þ /À B cells from the apoptotic effects of Myc overexpression without the added stresses of tissue culture, apoptosis in vivo was measured. Specifically, splenocytes from Em-myc transgenics doubly heterozygous for Mdm2 and ARF and from control mice were analyzed for apoptosis prior to any detectable disease. The apoptotic index of splenocytes from nontransgenic Mdm2 þ /À mice was slightly higher than that of wild-type, ARF þ /À, and ARF þ /ÀMdm2 þ /À mice, which had comparable percentages of apoptotic cells (Figure 4). A similar but more dramatic difference was observed when splenocytes from Em-myc transgenics were evaluated. The percentage of apoptotic splenocytes in ARF þ /ÀMdm2 þ /ÀEm-myc transgenics was analogous to the percentage in wild-type Em-myc and ARF þ /ÀEm-myc transgenics (Figure 4). Thus, loss of one allele of ARF conferred resistance to Mdm2 þ /À B cells to Myc-induced apoptosis in vivo and restored to wild type the apoptotic index of Mdm2 þ /ÀEm-myc transgenics. To assess whether the increased resistance to Figure 4 Loss of one allele of ARF confers resistance to Mdm2 þ /À apoptosis from ARF haploinsufficiency resulted in B cells to Myc-induced apoptosis in vivo. PI-stained splenocytes þ /À the restoration of B-cell populations in Mdm2 Em- from Em-myc transgenic negative (TgÀ) and Em-myc transgenic myc transgenics, splenic B cells from Em-myc trans- positive (Tg þ ) of the indicated genotypes prior to any detectable genics haploinsufficient for Mdm2 and/or ARF prior disease were analyzed by flow cytometry. DNA with less than 2N content (sub-G1) was quantitated from two (wild-type TgÀ, to any detectable disease were analyzed by flow þ /À þ /À þ /À þ /À þ þ Mdm2 TgÀ, Mdm2 Tg þ ) and three (ARF TgÀ, ARF cytometry. Normally, mature B cells (IgM ,CD19 ) Mdm2 þ /À TgÀ, wild-type Tg þ , ARF þ /À Tg þ , ARF þ /ÀMdm2 þ /À comprise 40–50% of the cells in a mouse spleen, and Tg þ ) separate mice and averaged. Error bars represent one B-cell numbers from wild-type, ARF þ /À,andMdm2 þ /À standard deviation spleens reflected this percentage (Figure 5). There were slightly reduced numbers (32–35%) of mature B cells and a population of B-cell precursors (9–10%, IgMÀ, þ þ /À ARF haploinsufficiency reverses the deleterious effects of CD19 ) in the spleens of the wild-type and ARF Mdm2 haploinsufficiency on Myc-induced Em-myc transgenics (Figure 5). A small percentage of lymphomagenesis precursor B cells in the periphery of Em-myc transgenics is typical of these mice (Adams et al., Em-myc transgenic mice lacking ARF and/or p53 1985; Langdon et al., 1986). Notably, spleens in develop lymphomas at highly accelerated rates (Eischen ARF þ /ÀMdm2 þ /ÀEm-myc transgenics had normal num- et al., 1999; Schmitt et al., 1999), whereas Mdm2 þ /ÀEm- bers of mature (41%) B cells and also lacked myc transgenics have a protracted rate of lymphoma precursor B cells (Figure 5), indicating a complete development (Alt et al., 2003). Given the effects of loss rescue of the B-cell developmental defects observed in of a single ARF allele in rescuing B-cell development and Mdm2 þ /ÀEm-myc transgenic mice. In addition, there survival in Mdm2 þ /ÀEm-myc transgenics, we tested were similar numbers of peripheral blood lymphocytes whether ARF haploinsufficiency would also restore the in ARF þ /ÀMdm2 þ /ÀEm-myc transgenics as compared normal course of lymphoma development in Mdm2 þ /À to wild-type Em-myc transgenic mice (data not Em-myc transgenics. As expected, Mdm2 þ /ÀEm-myc shown). Therefore, an ARF deficiency rescues the transgenics had an extended lifespan and ARF þ /ÀEm- negative effects of Myc overexpression and Mdm2 myc transgenic mice a shortened lifespan, when com- haploinsufficiency on B-cell development and survival pared to wild-type Em-myc transgenic littermates in vivo. (Figure 6; Eischen et al., 1999; Alt et al., 2003). Notably,

Oncogene Stoichiometry of ARF and Mdm2 regulates lymphomagenesis CM Eischen et al 8935

Figure 5 B-cell development in Mdm2 þ /ÀEm-myc transgenics is rescued by loss of one allele of ARF. Splenic cells were isolated from mice of the indicated genotypes prior to any detectable lymphoma. Cells were stained with fluorescent B-cell-specific antibodies (IgM and CD19) and subjected to flow cytometry analysis. Fluorescently labeled isotype antibody controls were used to determine the location of the quadrant axes. The percentage of cells in the top two quadrants is denoted in their respective quadrant; the percentage of cells in the bottom left quadrant is indicated in the bottom right quadrant. Data shown are representative of four separate experiments

transgenic mice (21 weeks) (Figure 6). The lymphomas that arose in the ARF þ /ÀMdm2 þ /ÀEm-myc transgenics were typical pre-B/B-cell lymphomas and developed in the peripheral lymph nodes and spleen, as is character- istic of lymphomas arising in wild type and ARF þ /ÀEm- myc transgenics (Adams et al., 1985; Eischen et al., 1999). Therefore, since ARF heterozygosity rescues the effects of Mdm2 haploinsufficiency, and a decrease in Mdm2 compensates for a deficiency in ARF, these results suggest that the stoichiometry of both ARF and Mdm2 is a critical determinant of the rate of lymphomagenesis. Figure 6 The rate of Myc-induced lymphoma development in ARF þ /ÀMdm2 þ /ÀEm-myc transgenics is similar to wild-type Em-myc transgenic mice. Kaplan–Meier survival curves of the indicated Inactivation of both p53 and ARF is selected for in genotype of Em-myc transgenic mice are plotted. The average lymphomas arising in ARF þ /ÀMdm2 þ /ÀEm-myc survivals in weeks are 43.5 (Mdm2 þ /ÀEm-myc transgenic), 8.2 (ARF þ /ÀEm-myc transgenic), 24.1 (ARF þ /ÀMdm2 þ /ÀEm-myc trans- transgenics genic), and 21.3(wild-type E m-myc transgenic). Log-rank test Mdm2 Po0.001. The number of mice in each group is denoted by the ‘n’ Previously we reported that haploinsufficiency values, and vertical lines indicate ages of surviving mice. mediates its inhibitory effects on Myc-induced lympho- Lymphoma was documented in all of the mice. The right-censored magenesis through activation of p53(Alt et al., 2003). nature of the Kaplan–Meier curves is due to the study being Moreover, loss of functional p53is the dominant terminated before all of the mice were killed alteration observed in lymphomas that arise in Em-myc transgenics (Eischen et al., 1999; Schmitt et al., 1999), the inhibition of lymphomagenesis in Mdm2 þ /ÀEm-myc and this selection becomes over-riding in the context of transgenics (44 weeks average survival) was completely an Mdm2 haploinsufficiency (Alt et al., 2003). By blocked in mice heterozygous for both Mdm2 and ARF contrast, in ARF þ /ÀEm-myc transgenics, nearly all of (24 weeks average survival) (Figure 6). Additionally, the the lymphomas suffered loss of the second allele of ARF, accelerated rate of lymphomagenesis in ARF þ /ÀEm-myc and none of these tumors had mutations or deletions in transgenics (8.2 weeks average survival) was inhibited by p53 (Eischen et al., 1999). We therefore addressed the loss of a single allele of Mdm2 (Figure 6). The average status of ARF and p53 in the B-cell lymphomas that survival of ARF þ /ÀMdm2 þ /ÀEm-myc transgenics was arose in ARF þ /ÀMdm2 þ /ÀEm-myc transgenics. Notably, analogous to the average survival of wild-type Em-myc p53 mutations were detected in 22% of the lymphomas

Oncogene Stoichiometry of ARF and Mdm2 regulates lymphomagenesis CM Eischen et al 8936 Table 1 p53, ARF, and Mdm2 status in lymphomas from ARF+/À Mdm2+/ÀEm-myc transgenics Genetic alteration Expression of (frequency), mouse/ tumor analyzeda p53b ARFc Mdm2d

p53 mutation (21.7%) 18 Mutant Overexpressed Moderately overexpressed 36 Mutant Overexpressed Moderately overexpressed 52 Mutant Overexpressed Basal 68 Mutant Overexpressed Basal 80 Mutant Overexpressed Overexpressed

p53 deletion (0%) Figure 7 Analysis of p53, ARF, and Mdm2 status in lymphomas None from ARF þ /ÀMdm2 þ /ÀEm-myc transgenic mice. (a) Whole-cell extracts of 22 separate lymphomas (designated by the numbers ARF deletion (43.5%) on the top of blot) arising in ARF þ /ÀMdm2 þ /ÀEm-myc transgenic 34 Low Undetected Overexpressed mice were Western blotted for p53(top panel), p19 ARF (second 50 Low Undetected Moderately panel), Mdm2 (third panel), and as a loading control b-actin overexpressed (bottom panel). Lysate from a lymphoma from a wild-type Em-myc 65 Low Undetected Overexpressed transgenic mouse that harbors mutant p53, overexpresses ARF, 72 Low Undetected Basal and overexpresses the three isoforms of Mdm2 protein (three 225 Low Undetected Basal arrows) was run as a control for p53, ARF, and Mdm2 protein 236 Low Undetected Basal expression (first lane, labeled C). (b) Southern blot analysis for 366 Low Undetected Moderately ARF was performed on genomic DNA isolated from lymphomas overexpressed from ARF þ /ÀMdm2 þ /ÀEm-myc transgenic mice. Genomic DNA 372 Low Undetected Overexpressed from an ARF þ /À and ARFÀ/À mouse was run as controls for the 383 Low Undetected Basal location of the wild-type and mutant alleles of ARF. The locations 401 Low Undetected Overexpressed of the wild-type (WT) and knockout (MT) alleles of ARF are indicated. Lack of a band denotes deletion of the allele ARF overexpression, p53 wild type (8.7%) 66 Low Overexpressed Moderately overexpressed 423Low Overexpressed Basal þ /À þ /À from ARF Mdm2 Em-myc transgenics (tumors 68, Mdm2 overexpression only 80, 52, 36, 18 in Figure 7a, Table 1), a frequency (13.0%) analogous to that observed in lymphomas from wild- 67 Basal Undetected Overexpressed type Em-myc transgenics (24%) (Eischen et al., 1999). 86 Basal Undetected Overexpressed Mutant p53protein was overexpressed in these lym- 227 Basal Undetected Overexpressed phomas and all bore missense mutations in the p53 No detectable alteration DNA-binding domain, the hot spot region for muta- (13.0%) tions. For example, p53 mutations at codons 270 and 05 Basal Undetected Basal 242 were identified in lymphomas from ARF þ /ÀMdm2 þ /À 75 Basal Undetected Basal Em-myc transgenic mice, and these two codons are those 245 Basal Undetected Basal most frequently mutated in p53 in human cancer aA total of 23lymphomas were analysed by Western and Southern blot (codons 273and 245, respectively) (Hainaut et al., and are denoted by numbers. bProtein expression, mutation verified by 1998). Similar to wild-type Em-myc transgenics (Eischen sequencing cDNA, p53 gene present in all tumors. cProtein expression, et al., 1999), we failed to detect alterations in ARF, Southern blot verified that ARF gene was deleted in 10 lymphomas. d Mdm2, or p53in 13%of the lymphomas emerging in Overexpression was determined by Western blot (see Figure 7a) ARF þ /ÀMdm2 þ /ÀEm-myc transgenics (Figures 7a and b, Table 1). ARF protein is normally not detectable in unstressed tissues and cells (Kamijo et al., 1997; Zindy and is associated with a corresponding decrease in et al., 1998); therefore, all lymphomas that did not lymphomas displaying loss of heterozygosity of ARF. overexpress ARF were evaluated for ARF deletions. p53can also be inactivated by Mdm2 overexpression Importantly, in lymphomas arising in ARF þ /ÀMdm2 þ /À (reviewed in Momand et al., 2000). Mdm2 protein is Em-myc transgenics, 43.5% had deleted the remaining overexpressed in approximately half of the lymphomas wild-type ARF allele (Figure 7b, Table 1), an exceed- that arise in wild-type Em-myc transgenics (Eischen et al., ingly rare event in Mdm2 þ /ÀEm-myc transgenic lympho- 1999) and is also overexpressed in lymphomas having mas (4%) (Alt et al., 2003). Not a single p53 deletion inactivated ARF or p53(Eischen et al., 1999, 2001; Alt was detected in any of the lymphomas from ARF þ /À et al., 2003). Analysis of all the lymphomas from ARF þ /À Mdm2 þ /ÀEm-myc transgenic mice (Table 1). Thus, loss Mdm2 þ /ÀEm-myc transgenics revealed that 57% (13of of one allele of Mdm2 drives the selection of p53 23) of these lymphomas overexpressed Mdm2 (Figure 7a, mutations in ARF þ /ÀEm-myc transgenic lymphomas, Table 1), strengthening the notion that Mdm2 over-

Oncogene Stoichiometry of ARF and Mdm2 regulates lymphomagenesis CM Eischen et al 8937 expression is selected for during Myc-induced lymphomagenesis. Of the lymphomas that had deleted ARF or mutated p53, Mdm2 was also overexpressed in over half of these lymphomas (Figure 7, Table 1). Thus, despite the fact that ARF þ /ÀMdm2 þ /ÀEm-myc transgenic mice are haploinsufﬁcient for Mdm2, Mdm2 overexpression was selected for as frequently in ARF þ /ÀMdm2 þ /ÀEm- myc transgenics as in wild-type Em-myc transgenics during Myc-induced lymphomagenesis. Therefore, the levels of both ARF and Mdm2 together regulate p53and dictate the spectrum of genetic alterations selected for during Myc-induced lymphomagenesis.

Discussion Figure 8 Stoichiometry of ARF and Mdm2 regulates Myc- ARF, Mdm2, and p53function in a pathway that induced apoptosis and lymphoma development. Schematics of inhibits lymphoma development by prompting apopto- the ARF–Mdm2–p53tumor suppressor pathway in E m-myc transgenic mice of the indicated genotype. The pathway illustrated sis when oncogenes are overexpressed (Zindy et al., for wild-type Em-myc transgenics (furthest left) is altered for each 1998; Eischen et al., 1999; Schmitt et al., 1999). B cells of the genotypes to the right to reflect the changes that occur in the lacking ARF and/or p53 are resistant to Myc-induced pathway as a consequence of ARF and/or Mdm2 haploinsuffi- apoptosis (Eischen et al., 1999; Schmitt et al., 1999), and ciency. Smaller symbols of ARF and Mdm2 indicate heterozygosity, whereas the different sizes of the p53symbols represent the accordingly B-cell lymphoma development is accelerated differences in activation of p53. In ARF þ /ÀEm-myc transgenic mice in Em-myc transgenics that lack ARF and/or p53 (second from left), there is less ARF to regulate Mdm2, and this (Eischen et al., 1999; Schmitt et al., 1999). However, leads to decreased p53activation and apoptosis and accelerated when p53’s negative regulator Mdm2 is limiting, as in lymphoma development (Zindy et al., 1998; Eischen et al., 1999). þ /À Mdm2 þ /À mice, B cells are rendered exquisitely sensitive The levels of Mdm2 in Mdm2 Em-myc transgenics (third from the left) are too low to regulate p53effectively, resulting in increased to p53-dependent apoptosis induced by Myc. As a net apoptosis and inhibition of lymphoma development (Alt et al., result, B-cell lymphoma development is markedly 2003). In ARF þ /ÀMdm2 þ /ÀEm-myc transgenics (furthest right), the impaired in Mdm2 þ /ÀEm-myc transgenic mice, and the decreased levels of ARF allow for a sufficient level of Mdm2 to lymphomas that do arise in these mice have an increased regulate p53and re-establish the equilibrium in the pathway; consequently, B-cell apoptosis and lymphoma development are frequency of p53mutations (Alt et al., 2003). Moreover, restored to wild-type Em-myc transgenic rates (Figures 4 and 6) the rates of B-cell apoptosis and lymphomagenesis are analogous in p53 þ /ÀEm-myc transgenic and Mdm2 þ /À p53 þ /ÀEm-myc transgenic mice (Alt et al., 2003), indicating that p53mediates the haploinsufficiency regulation, apoptosis, cancer development, and survival effects of Mdm2 in Em-myc transgenics. The results (Figure 8). presented here reveal that regulation of p53-dependent apoptosis by Mdm2 is more complex than previously realized and underscore the importance of ARF as a ARF regulates Mdm2 under stressful conditions stoichiometric regulator of Mdm2. p53expression must be strictly regulated in cells to Strikingly, our unpredicted results demonstrate that a inhibit its negative consequences on cell growth and loss of a single allele of ARF completely rescues the survival. For example, a deficiency or inactivation of effects of Mdm2 haploinsufficiency on B-cell develop- p53results in the development of cancer, whereas ment, survival, and transformation in Em-myc transgenic uncontrolled p53results in apoptosis and can lead to mice. Rates of B-cell apoptosis and development of B- the death of the organism. As a guardian of p53, Mdm2 cell lymphoma in ARF þ /ÀMdm2 þ /ÀEm-myc transgenics is itself also tightly regulated, and ARF is a key were equivalent to those of wild-type Em-myc trans- regulator in this respect (reviewed in Sherr and Weber, genics. Moreover, the increased transcriptional activity 2000). Recent studies have suggested that control of p53 of p53in Mdm2 þ /À lymphocytes was restored to wild- by ARF through Mdm2 occurs under specific stressful type levels with loss of one allele of ARF. Therefore, the conditions and possibly in certain cell types. Specifically, threshold level of Mdm2 necessary to regulate p53in ARF is required to regulate Mdm2 in cells placed into Em-myc transgenics is re-established when a single allele tissue culture and in lymphocytes and mouse embryo of ARF is lost (Figure 8). Similarly, it would follow that fibroblasts (MEFs) that overexpress Myc (Zindy et al., the levels of ARF required to harness Mdm2 activity in 1998; Eischen et al., 1999; Schmitt et al., 1999; Sherr and Em-myc transgenic mice are restored when there is a DePinho, 2000). In contrast, other stimuli, including haploinsufficiency in Mdm2 (Figure 8). Taken together, lethal doses of gamma irradiation, overexpression of the these findings are consistent with the concept that there E2f1 oncogene, and the development of hematopoietic is a delicate balance of ARF and Mdm2, and pushing and nonhematopoietic cells, do not appear to rely on this equilibrium in either direction drastically alters p53 Mdm2 regulation by ARF (Kamijo et al., 1997, 1999;

Oncogene Stoichiometry of ARF and Mdm2 regulates lymphomagenesis CM Eischen et al 8938 Stott et al., 1998; Russell et al., 2002; Tolbert et al., denominator of some tumor types, and this may be 2002; O’Leary et al., 2004). For example, death caused independent of their effects on p53function. by lethal doses of gamma irradiation of Mdm2 þ /À and Mdm2 hypomorphic mice, which express lower levels of Mdm2 protein, is not affected by deletion of ARF Mdm2 as a therapeutic target in tumors lacking ARF (O’Leary et al., 2004). Loss of ARF also fails to rescue Previous studies have shown that ARF overexpression the lethality of Mdm2-null embryos (CM Eischen, blocks Mdm2’s ability to inhibit p53(reviewed by Sherr unpublished data), and the reduced body weights and and Weber, 2000). We show here that the primary role decreased lymphocyte numbers in Mdm2 hypomorphic of Myc-induced ARF is to regulate Mdm2. However, mice are unaltered by deletion of ARF (O’Leary et al., our findings also reveal that Mdm2 levels regulate the 2004). Consistent with the former reports, we show here effects of ARF expression. Specifically, ARF-null pre-B- / that in Mdm2 þ ÀEm-myc transgenic mice, loss of only cell growth rates were reduced to those of wild-type pre- one allele of ARF restored to wild-type B-cell develop- B cells by decreasing Mdm2 levels. Additionally, B cells ment, susceptibility to apoptosis, lymphoma latency, lacking ARF are normally resistant to the apoptotic and the spectrum of genetic alterations selected for effects of Myc (Eischen et al., 1999; Schmitt et al., 1999), during lymphomagenesis. Additionally, deletion of ARF yet loss of one allele of Mdm2 conferred sensitivity to rescued the high levels of spontaneous apoptosis that are Myc-induced apoptosis and inhibited Myc-induced / a hallmark of Mdm2 þ À pre-B cells in tissue culture. lymphomagenesis in Em-myc transgenics that were also Thus, our genetic data support and extend earlier in haploinsufficient for ARF. These results suggest that vitro findings (Zindy et al., 1998; Eischen et al., 1999; modulating Mdm2 levels or function would have Schmitt et al., 1999; Sherr and DePinho, 2000) that significant impact on tumors bearing alterations in ARF regulates Mdm2 in B cells when Myc is over- ARF expression. In accordance with this idea, Mdm2 expressed and in pre-B cells in tissue culture. However, antisense oligonucleotides, which can induce apoptosis our data also show that the increased activation of p53 of a variety of human cancer cell lines, have recently / in Mdm2 þ À mice from gamma irradiation is reduced to been used to treat several kinds of cancers in xenograft wild-type activity with loss of one allele of ARF. models (Wang et al., 2001). Other approaches being Collectively, these results indicate that ARF is required investigated primarily by the pharmaceutical industry to regulate Mdm2 in certain scenarios, but not in others, are to identify small molecules that inhibit Mdm2 and and thus, this pathway must be dependent on other p53interactions (Chene, 2003;Vassilev et al., 2004) or factors and/or signaling pathways. Mdm2’s E3ubiquitin ligase activity (Sun, 2003).There- fore, such approaches should be particularly beneficial ARF and Mdm2 cooperate in tumorigenesis to treat tumors that have inactivated ARF and still express wild-type p53. Several studies implicate the cooperation of ARF loss and Mdm2 overexpression on cell cycle and tumorigenesis (Lundgren et al., 1997; Eischen et al., 1999; Foster and Lozano, 2002; Alt et al., 2003; Moore et al., 2003), Materials and methods and that this can occur in a p53-independent manner. Specifically, defects in mammary gland development are Transgenic and knockout mice accentuated in ARF-null mice overexpressing Mdm2 in The congenic C57Bl/6 Em-myc transgenic mouse strain breast epithelial cells (Foster and Lozano, 2002). These (Adams et al., 1985) was generously provided by Drs Alan Mdm2-mediated mammary gland defects also occurred Harris (Walter and Eliza Hall Institute, Melbourne, Australia) in p53-null mice (Lundgren et al., 1997). Additionally, in and Charles Sidman (University of Cincinnati, Cincinnati, lymphomas that arise in ARF-null Em-myc transgenics, OH, USA). The Mdm2 þ /À (C57Bl/6 Â 129/Sv) (Montes de Oca À/À or in lymphomas from wild-type Em-myc transgenics Luna et al., 1995) and the ARF (congenic C57Bl/6) mice bearing deletions of ARF, Mdm2 is overexpressed in at (Kamijo et al., 1997) were kindly provided by Dr Guillermina Lozano (MD Anderson Cancer Center, Houston, TX, USA) least half of these tumors (Eischen et al., 1999, 2001). and Drs Charles Sherr and Martine Roussel (St Jude Moreover, Mdm2 is frequently overexpressed in lym- Children’s Research Hospital, Memphis, TN, USA), respec- phomas that have mutated p53(Eischen et al., 1999; Alt tively. The Mdm2 þ /À mice were crossed to Em-myc transgenic et al., 2003). These results would suggest that ARF mice to generate F1’s. The F1’s were then crossed to ARFÀ/À deletion or p53mutation and Mdm2 overexpression are and ARF þ /À mice to generate F2 ARF þ / þ Mdm2 þ /ÀEm-myc, two separate selected events during tumor development ARF þ /ÀMdm2 þ / þ Em-myc, ARF þ /ÀMdm2 þ /ÀEm-myc, and wild- that independently contribute to lymphomagenesis. type Em-myc transgenics. Nontransgenics were also generated Indeed, a deficiency in ARF and overexpression of from the F1 crosses and used as littermate controls where Mdm2 has been shown by others to be capable of indicated. All mice were monitored daily for disease, and all accelerating tumor development (Moore et al., 2003). research with these mice complied with federal, state, and institutional guidelines. A Kaplan–Meier survival analysis was This decreased tumor latency in ARF-null/Mdm2-over- performed for each genotype. Since the Kaplan–Meier plot expressing mice is purported to be independent of p53, (Figure 6) includes mice that were still alive at the time the plot as loss of both p53 and ARF did not alter tumor latency was generated, these cases are right censored because the study (Weber et al., 2000; Moore et al., 2003). Therefore, loss was terminated before death occurred. To determine the of ARF and overexpression of Mdm2 may be a common statistical significance of the survival between the different

Oncogene Stoichiometry of ARF and Mdm2 regulates lymphomagenesis CM Eischen et al 8939 genotypes of Em-myc transgenic mice, log-rank tests were were sacrificed and spleens collected and frozen for later performed. analysis. Four separate litters of mice were used for these analyses. Isolation and culture of primary pre-B cells Bone marrow of 9- to 12-week-old wild-type, Mdm2 þ /À, ARFÀ/À, Western blotting ARFÀ/ÀMdm2 þ /À, Mdm2 þ /ÀEm-myc transgenic, ARF þ /ÀEm-myc þ /À þ /À Frozen mouse spleens of various genotypes and pre-B/B-cell transgenic, ARF Mdm2 Em-myc transgenic, and wild-type lymphomas (3–5 mm3 chunk) were lysed and protein was Em-myc transgenic mice was placed into tissue culture to select extracted as previously described (Zindy et al., 1998; Eischen for the outgrowth of pure (>98%) populations of pre-B cells þ þ À À et al., 1999). Equal amounts of protein were separated by (CD19 , B220 , CD43 , IgM ), as previously described SDS–PAGE, transferred to nitrocellulose membranes (Pro- (Eischen et al., 1999). Phenotype analysis was performed tran, Schleicher and Schuell, Dassel, Germany), and blotted 7–14 days postexplant with fluorescent B-cell-specific anti- with antibodies specific for p19ARF (GeneTex, San Antonio, bodies from Southern Biotechnology (Birmingham, AL, USA) TX, USA), p53(Ab-7, Calbiochem, La Jolla, CA, USA), and PharMingen (San Diego, CA, USA), and flow cytometry Mdm2 (C-18, Santa Cruz Inc., Santa Cruz, CA, USA), p21 (FACSCalibur; BD Immunocytometry Systems, San Jose, CA, (F5, Santa Cruz), and b-actin (Sigma, St Louis, MO, USA). USA) verified that the cells in culture were pre-B cells. HRP-linked secondary antibodies (Amersham Pharmacia Biotech, Piscataway, NJ, USA) and Pierce Supersignal (Rock- Viability and apoptosis assays ford, IL, USA) were used to detect bound immunocomplexes. Bone marrow was explanted into IL-7-containing medium on day 0 (see above), and cell viability was determined at specific Southern blotting and sequencing intervals by Trypan blue dye exclusion. For apoptosis measurements, pre-B cells or disaggregated splenocytes were Genomic DNA from each of the lymphomas that emerged in stained with PI without or with, respectively, ethanol fixation. ARF þ /ÀMdm2 þ /ÀEm-myc transgenic mice was digested with The FACSCaliber was used to analyse the PI-stained samples, AflII or BamHI. Following electrophoretic separation and and CellQuest software (BD Immunocytometry Systems, San transfer to Nytran (Scheicher and Schuell, Dassel, Germany) Jose, CA, USA) was used to quantitate fragmented (sub-G1) membranes, the digested DNAs were probed with radioactive DNA. cDNAs coding ARF (exon 1b) (for AflII-digested DNA) and p53 (exons 2–10) (for BamHI-digested DNA). Genomic DNA þ /À À/À À/À Phenotype analysis from ARF , ARF , and p53 mice was used for controls. p53 was sequenced from RT–PCR products from RNA Whole spleens from wild-type, Mdm2 þ /À, ARF þ /À, wild-type isolated from lymphomas arising in ARF þ /ÀMdm2 þ /ÀEm-myc Em-myc transgenic, Mdm2 þ /ÀEm-myc transgenic, ARF þ /ÀEm-myc transgenic, as previously described (Eischen et al., 1999). transgenic, and ARF þ /ÀMdm2 þ /ÀEm-myc transgenic mice were minced and strained through a 100 mm nylon filter and red blood cells were lysed. Splenocytes were incubated with one to Acknowledgements three fluorescently labeled antibodies (CD19-PE, IgM-FITC, We thank Drs John Cleveland and Hua Xiao for critically B220-CyChrome, and/or fluorescent isotype controls) per reviewing the manuscript, Drs Guillermina Lozano, Martine sample and evaluated on a FACSCalibur instrument. Data Roussel, Charles Sherr, Alan Harris, and Charles Sidman for were analysed with CellQuest software. kindly providing breeder mice that were essential for these studies, Dr Jane Meza for the Kaplan–Meier analysis, and the personnel in the Eppley Institute’s animal facility. This work Gamma irradiation was supported by NCI grant CA098139, the Wanda Rizzo Wild-type, Mdm2 þ /À, ARF þ /À, and ARF þ /ÀMdm2 þ /À litter- memorial fund, the Eppley Institute for Research in Cancer, mates were unirradiated or subjected to whole-body gamma and NIH training grant T32 CA09476 (JRA). CME is a irradiation (10 Gy) from a 137cesium irradiator. After 4 h, mice Leukemia and Lymphoma Society Scholar.

References

Adams JM, Harris AW, Pinkert CA, Corcoran LM, Honda R, Tanaka H and Yasuda H. (1997). FEBS Lett., 420, Alexander WS, Cory S, Palmiter RD and Brinster RL. 25–27. (1985). Nature, 318, 533–538. Honda R and Yasuda H. (1999). EMBO J., 18, 22–27. Alt JR, Greiner TC, Cleveland JL and Eischen CM. (2003). Jones SN, Roe AE, Donehower LA and Bradley A. (1995). EMBO J., 22, 1442–1450. Nature, 378, 206–208. Barak Y, Juven T, Haffner R and Oren M. (1993). EMBO J., Juven T, Barak Y, Zauberman A, George DL and Oren M. 12, 461–468. (1993). Oncogene, 8, 3411–3416. Chene P. (2003). Nat. Rev. Cancer, 3, 102–109. Kamijo T, van de Kamp E, Chong MJ, Zindy F, Diehl JA, Eischen CM, Roussel MF, Korsmeyer SJ and Cleveland JL. Sherr CJ and McKinnon PJ. (1999). Cancer Res., 59, (2001). Mol. Cell. Biol., 21, 7653–7662. 2464–2469. Eischen CM, Weber JD, Roussel MF, Sherr CJ and Cleveland Kamijo T, Weber JD, Zambetti G, Zindy F, Roussel MF JL. (1999). Genes Dev., 13, 2658–2669. and Sherr CJ. (1998). Proc. Natl. Acad. Sci. USA, 95, Foster CJ and Lozano G. (2002). Oncogene, 21, 3525–3531. 8292–8297. Freedman DA and Levine AJ. (1998). Mol. Cell. Biol., 18, Kamijo T, Zindy F, Roussel MF, Quelle DE, Downing JR, 7288–7293. Ashmun RA, Grosveld G and Sherr CJ. (1997). Cell, 91, Hainaut P, Hernandez T, Robinson A, Rodriguez-Tome P, 649–659. Flores T, Hollstein M, Harris CC and Montesano R. (1998). Korgaonkar C, Zhao L, Modestou M and Quelle DE. (2002). Nucleic Acids Res., 26, 205–213. Mol. Cell. Biol., 22, 196–206.

Oncogene Stoichiometry of ARF and Mdm2 regulates lymphomagenesis CM Eischen et al 8940 Langdon WY, Harris AW, Cory S and Adams JM. (1986). Russell JL, Powers JT, Rounbehler RJ, Rogers PM, Cell, 47, 11–18. Conti CJ and Johnson DG. (2002). Mol. Cell. Biol., 22, Lundgren K, Montes de Oca Luna R, McNeill YB, Emerick 1360–1368. EP, Spencer B, Barﬁeld CR, Lozano G, Rosenberg MP and Schmitt CA, McCurrach ME, de Stanchina E, Wallace- Finlay CA. (1997). Genes Dev., 11, 714–725. Brodeur RR and Lowe SW. (1999). Genes Dev., 13, Mendrysa SM, McElwee MK, Michalowski J, O’Leary KA, 2670–2677. Young KM and Perry ME. (2003). Mol. Cell. Biol., 23, Sherr CJ and DePinho RA. (2000). Cell, 102, 407–410. 462–472. Sherr CJ and Weber JD. (2000). Curr. Opin. Genet. Dev., 10, Momand J, Jung D, Wilczynski S and Niland J. (1998). 94–99. Nucleic Acids Res., 26, 3453–3459. Stott FJ, Bates S, James MC, McConnell BB, Starborg M, Momand J, Wu HH and Dasgupta G. (2000). Gene, 242, 15–29. Brookes S, Palmero I, Ryan K, Hara E, Vousden KH and Momand J, Zambetti GP, Olson DC, George D and Peters G. (1998). EMBO J., 17, 5001–5014. Levine AJ. (1992). Cell, 69, 1237–1245. Sun Y. (2003). Cancer Biol. Ther., 2, 623–629. Montes de Oca Luna R, Wagner DS and Lozano G. (1995). Tolbert D, Lu X, Yin C, Tantama M and Van Dyke T. (2002). Nature, 378, 203–206. Mol. Cell. Biol., 22, 370–377. Moore L, Venkatachalam S, Vogel H, Watt JC, Wu CL, Vassilev LT, Vu BT, Graves B, Carvajal D, Podlaski F, Steinman H, Jones SN and Donehower LA. (2003). Filipovic Z, Kong N, Kammlott U, Lukacs C, Klein C, Oncogene, 22, 7831–7837. Fotouhi N and Liu EA. (2004). Science, 303, 844–848. O’Leary KA, Mendrysa SM, Vaccaro A and Perry ME. Wang H, Prasad G, Buolamwini JK and Zhang R. (2001). (2004). Mol. Cell. Biol., 24, 186–191. Curr. Cancer Drug Targets, 1, 177–196. Perry ME, Piette J, Zawadzki JA, Harvey D and Levine AJ. Watanabe T, Ichikawa A, Saito H and Hotta T. (1996). Leuk. (1993). Proc. Natl. Acad. Sci. USA, 90, 11623–11627. Lymphoma, 21, 391–397. Pomerantz J, Schreiber-Agus N, Liegeois NJ, Silverman A, Weber JD, Jeffers JR, Rehg JE, Randle DH, Lozano G, Alland L, Chin L, Potes J, Chen K, Orlow I, Lee HW, Roussel MF, Sherr CJ and Zambetti GP. (2000). Genes Dev., Cordon-Cardo C and DePinho RA. (1998). Cell, 92, 14, 2358–2365. 713–723. Wu X, Bayle JH, Olson D and Levine AJ. (1993). Genes Dev., Randle DH, Zindy F, Sherr CJ and Roussel MF. (2001). Proc. 7, 1126–1132. Natl. Acad. Sci. USA, 98, 9654–9659. Zhang Y, Xiong Y and Yarbrough WG. (1998). Cell, 92, Robertson KD and Jones PA. (1998). Mol. Cell. Biol., 18, 725–734. 6457–6473. Zindy F, Eischen CM, Randle DH, Kamijo T, Cleveland JL, Roth J, Dobbelstein M, Freedman DA, Shenk T and Sherr CJ and Roussel MF. (1998). Genes Dev., 12, Levine AJ. (1998). EMBO J., 17, 554–564. 2424–2433.

Oncogene