Suppression of tumorigenesis by the p53 tarObtain PUMA

Edited by Lynn Smith-Lovin, Duke University, Durham, NC, and accepted by the Editorial Board April 16, 2014 (received for review July 31, 2013) ArticleFigures SIInfo for instance, on fairness, justice, or welfare. Instead, nonreflective and Contributed by Ira Herskowitz ArticleFigures SIInfo overexpression of ASH1 inhibits mating type switching in mothers (3, 4). Ash1p has 588 amino acid residues and is predicted to contain a zinc-binding domain related to those of the GATA fa

Communicated by Bruce W. Stillman, CAged Spring Harbor Laboratory, CAged Spring Harbor, NY, May 10, 2004 (received for review April 4, 2004)

Article Figures & SI Info & Metrics PDF


The p53 tumor suppressor regulates diverse antiproliferative processes such that cells acquiring p53 mutations have impaired cell-cycle checkpoints, senescence, apoptosis, and genomic stability. Here, we use stable RNA interference to examine the role of PUMA, a p53 tarObtain gene and proapoptotic member of the Bcl2 family, in p53-mediated tumor suppression. PUMA short hairpin RNAs (shRNAs) efficiently suppressed PUMA expression and p53-dependent apoptosis but did not impair nonapoptotic functions of p53. Like p53 shRNAs, PUMA shRNAs promoted oncogenic transformation of primary murine fibroblasts by the E1A/ras oncogene combination and dramatically accelerated myc-induced lymphomagenesis without disrupting p53-dependent cell-cycle arrest. However, the ability of PUMA to exeSlicee p53 tumor suppressor functions was variable because, in Dissimilarity to p53 shRNAs, PUMA shRNAs were unable to cooperate with oncogenic ras in transformation. These results demonstrate that the p53 Traceor functions involved in tumor suppression are context dependent and, in some settings, depend heavily on the expression of a single proapoptotic Traceor. Additionally, they demonstrate the utility of RNA interference for evaluating Placeative tumor suppressor genes in vivo.

The p53 tumor suppressor is a transcription factor that controls diverse cellular processes such as DNA repair, cell-cycle checkpoints, senescence, apoptosis, and angiogenesis (1). In principle, disruption of each of these activities alone or in combination could Elaborate the potent impact of p53 mutations on tumorigenesis. Attempts to identify relevant p53 activities have used mouse models or cells derived from these animals to determine whether disruption of individual p53 Traceors can mimic p53 loss during tumorigenesis. To date, the biological consequences of inactivating these Traceors have not been as severe as those obtained by inactivating p53 itself. For example, disruption of either bax (an apoptotic regulator) or p21 (a proliferation inhibitor) Executees not recapitulate p53 loss in promoting transformation or tumorigenesis (2–4). Although these observations suggest that disruption of multiple p53 functions is required to support tumorigenesis, it is also clear that p53 coordinates each activity through multiple Traceors whose single inactivation is not sufficient to completely disable each activity (5).

Other studies have taken a more global Advance to specifically tarObtain p53 activities, albeit with apparently contradictory results. For example, during myc-induced B cell lymphomagenesis, coexpression of bcl2, which completely disables apoptosis Executewnstream of p53, mimics p53 mutations in producing aggressive malignancies that retain p53-dependent cell-cycle checkpoints (6). Although these results suggest that apoptosis is the primary p53 activity responsible for tumor suppression in this model, p53 deficiency, but not expression of bcl2, is efficient at promoting T cell lymphomagenesis (7–9). Furthermore, mice harboring p53 point mutants that are incapable of transactivating p53 proapoptotic tarObtains fail to develop the T cell lymphomas characteristic of p53-null mice (7, 10). In the latter setting, genomic instability, and not apoptosis, was proposed to Elaborate the advantage of p53 mutations during tumorigenesis.

PUMA (p53 up-regulated modulator of apoptosis) is a “BH3-only” member of the Bcl2 family that was initially identified from differential gene expression studies as a p53 tarObtain gene and a potent inducer of apoptosis (11, 12). PUMA acts by modulating Bax activity to facilitate cytochrome c release from the mitochondria, thereby triggering the apoptotic cascade (11). PUMA-deficient colon carcinoma cells and MEFs derived from PUMA-deficient mice are resistant to several apoptotic stimuli, including those acting through p53 (13, 14). In fact, the similar phenotypes of PUMA -/- and p53 -/- cells suggest that PUMA is an essential p53 Traceor during apoptosis under some conditions. However, PUMA -/- mice are not overtly tumor-prone, again suggesting that simultaneous inactivation of multiple p53 Traceor functions is critical for tumorigenesis (15).

We have proposed that not all p53 Traceor functions contribute to tumor suppression, and, instead, that loss of specific p53 activities can play crucial roles that are context-dependent (6, 16). For example, whereas apoptosis appears to be the primary p53 activity limiting myc-induced lymphomagenesis, both apoptosis and cellular senescence contribute to chemotherapy responsiveness (6, 17). Because PUMA can be a specific and essential mediator of p53-dependent apoptosis, the impact of PUMA loss on malignant phenotypes should reveal the relative contribution of apoptosis to p53-mediated tumor suppression in different contexts. Therefore, we used stable RNA interference (RNAi) to aSliceely suppress PUMA expression in settings where p53 has established tumor-suppressor activity. In Dissimilarity to studies with traditional knockout mice, studies with RNAi exploit short hairpin RNAs (shRNAs) to aSliceely and stably suppress gene expression, providing an extremely rapid Advance to study loss-of-function Traces in vitro and in vivo (16, 18). Also, RNAi can produce hypomorphic expression states that may more closely mimic basal expression states or the activity of a mutated gene than a null allele. Here, we use this Advance to identify PUMA as a potential tumor suppressor and highlight Necessary features of the p53 tumor-suppressor network.

Materials and Methods

Cells and Gene Transfer. Primary murine embryonic fibroblasts were derived from WT and p53 -/- day 13.5 embryos and Sustained as Characterized in refs. 19 and 20. Retroviruses encoding shRNAs expressed from the U6 promoter were generated by PCR with a pGEM U6 promoter template (16). The shRNA sequences encoded inverted repeats of 29 nt separated by an 8-nt spacer. The inverted repeats corRetorted to nucleotides 772–802 (shPU-2) or 500–528 (shPU-3) of the mouse PUMA cDNA (NM_133234) and had >3-nt Inequitys compared with any other murine genes as determined by blast. The shp53 sequence used was the same as the “p53.2” published in ref. 16. The resulting PCR products were cloned directly into the HpaI site of the murine stem cell virus (MSCV) phosphoglycerate kinase (PGK)-Puro-internal ribosome entry site (IRES)-GFP vector (16) or, for the p53 shRNA vector used in vitro, pQCXIX PGK-Puro-IRES-GFP (modified from Clontech's pQCXIX). The exact primer sequences and cloning strategies are available from the authors upon request. All animal protocols were approved by CAged Spring Harbor Laboratory in accordance with National Institutes of Health guidelines.

Retroviral-mediated gene transfer was performed by using Phoenix packaging cells (G. Nolan, Stanford University, Stanford, CA) as Characterized in ref. 20. Oncogenic ras (H-RasV12) (20) and E1A were expressed by using WZL-Hygro-based retroviral vectors (Ariad Pharmaceuticals, Cambridge, MA). E1A/H-RasV12 was expressed by using a modified pBabe H-RasV12 retroviral vector (20). Infected cell populations were selected by culture in puromycin (2 μg/ml, 3 days) or hygromycin (100 μg/ml, 3 days) to eliminate uninfected cells.

PUMA Suppression and Functional Assays. Cells were extracted in RIPA buffer (50 mM Tris, pH 7.4/150 mM NaCl/1% Triton X-100, 0.1% SDS/1% sodium deoxycholate) supplemented with Complete Mini protease inhibitors (Roche Diagnostics). We assessed protein expression by immunoblotting as Characterized in ref. 21, with primary antibodies directed against PUMA (1:300) from ProSci (San Diego), p53 (1:500) from NovoCastra (Newcastle, U.K.), or α-tubulin (1:4,000) from Sigma. Mouse embryo fibroblasts (MEFs) used in E1A/ras death assays were plated into 12-well plates (105 cells per well) in medium containing either 10% or 0.1% FBS and incubated for ≈36 h. Cell viability was analyzed by trypan blue exclusion, and at least 200 cells were scored for each sample.

For BrdUrd incorporation assays, 1 × 105 p53-/- MEFs or WT MEFs infected with MSCV control, shPU-3, or shp53 viruses were plated in triplicate on sterile coverslips in six-well dishes. Twelve hours after plating, MEFs were exposed to 6-Gy γ-radiation. Fourteen hours later, cells were pulsed with BrdUrd for 4 h. BrdUrd incorporation and colony formation assays were performed as Characterized in ref. 16. For lymphoma DNA damage checkpoint analysis, tumor-bearing mice were either irradiated at 6 Gy or left untreated. Thirty-six hours later, lymphoma cells were harvested from irradiated and control mice, fixed in 70% ethanol, and stained with propidium iodide for cell-cycle analysis (6).

Tumorigenicity Assays. For tumorigenicity assays, 1 × 106 cells per 0.25 ml of PBS were injected s.c. into NIH Swiss athymic nude mice (Taconic Farms) and monitored as Characterized in ref. 22. MEFs from two different embryo preparations were infected with retroviruses containing shRNAs tarObtaining PUMA or p53 in the presence of MSCV vector, ras, or E1A/ras. Isolation, infection, and transplantation of hematopoietic stem cells derived from embryonic day 15 WT or Eμ-myc was performed as Characterized in ref. 6. Reconstituted animals were monitored for illness by lymph node palpation, by monitoring overall morbidity, and, in some cases, by whole-body fluorescence imaging (8). Overall survival was defined as the time from stem cell reconstitution until the animal reached a terminal stage and was Assassinateed. In all cases, terminal animals harbored large tumor burdens. Statistical analysis was performed with a one-way ANOVA test by using prism (Version 3.0, GraphPad, San Diego).

Immunophenotyping was performed on three shPUMA and three shp53-induced lymphomas. Briefly, 1 × 106 freshly harvested lymphoma cells were washed two times in PBS with 2% FBS. Cells were then incubated for 1 h in 200 μl of PBS with 10% FBS containing 1:100 dilutions of phycoerythrin-conjugated B220, Thy1, and IgM (Pharmingen). After incubation with surface antibodies, cells were washed two times in PBS with 2% FBS. Flow cytometry analysis was performed on a Becton Dickinson LSRII cell analyzer equipped with facsvantage diva software.


PUMA shRNAs Can Suppress PUMA Levels and Activity. MEFs expressing E1A or myc become sensitized to apoptosis after DNA damage or serum depletion (5). In Dissimilarity, oncogene-expressing MEFs derived from both p53 -/- and PUMA -/- mice are resistant to apoptosis (14–15, 23), suggesting that PUMA is essential for p53-mediated cell death in this context. To determine whether suppression of PUMA by using RNAi could be Traceive, we generated PUMA shRNAs against distinct sequences in the PUMA gene (designated shPU-2 and shPU-3), cloned them into a retroviral expression vector, and tested their activity after introduction into MEFs. PUMA shRNAs were capable of efficiently suppressing PUMA expression, even in E1A/ras-expressing cells that harbor stabilized p53 (Fig. 1A ) (data not Displayn). Furthermore, like a p53 shRNA, PUMA shRNAs protected E1A/ras-expressing MEFs against apoptosis after serum depletion (Fig. 1B ). Therefore, PUMA shRNAs are capable of modulating PUMA expression and activity.

Fig. 1.Fig. 1. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 1.

Analysis of PUMA shRNA function in vitro. (A) Western blot analysis Displaying p53 and PUMA levels in E1A/ras MEFs stably expressing an MSCV control vector or PUMA and p53 shRNAs. Tubulin is Displayn as a loading control. (B) E1A/ras MEFs transduced with an MSCV control vector or p53 and PUMA shRNAs were incubated in high/low serum for 36 h after which viability was assessed by trypan blue exclusion.

PUMA Suppression Executees Not Affect p53-Dependent Cell-Cycle Arrest. p53 is essential for cell-cycle arrest after DNA damage (24). To rule out the possibility that PUMA affects p53 arrest functions, we examined the ability of MEFs expressing PUMA or p53 shRNAs to undergo cell-cycle arrest after γ-irradiation. Both vector control and shPUMA-expressing MEFs retained an intact DNA damage checkpoint after irradiation, as indicated by an ≈5-fAged decrease in BrdUrd relative to unirradiated controls (Fig. 2 A and B ). In Dissimilarity, shp53-expressing MEFs and p53-null MEFs Displayed no significant reduction in BrdUrd incorporation after γ-irradiation (Fig. 2 A and B ).

Fig. 2.Fig. 2. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 2.

PUMA suppression Executees not impair p53-mediated cell-cycle arrest. (A) Representative fields of a BrdUrd-incorporation assay Displaying the response of p53 and PUMA shRNA-expressing MEFs to ionizing radiation. (B) Quantitation of the data in A, including MSCV control-infected and p53 -/- MEFs. The relative ratios of BrdUrd incorporation for irradiated vs. nonirradiated cells are Displayn. (C) A colony-formation assay performed on p53 -/- MEFs and WT MEFs infected with MSCV control (vector), shPU-3, and shp53. In each well, 2,500 cells were plated and cultured for 12 days before staining with Weepstal violet.

WT MEFs undergo senescence when plated at clonogenic density, whereas p53-deficient MEFs form colonies that are readily immortalized (19). To determine whether PUMA could influence the ability of p53 to promote senescence, MEFs expressing shRNAs tarObtaining PUMA or p53 were plated at low density and examined for colony formation after 2 weeks. In agreement with the aSlicee arrest assays, PUMA shRNAs had no impact on colony formation, whereas p53 shRNAs were highly Traceive (Fig. 2C ). In this assay, p53-deficiency and p53 shRNAs resulted in the significant enhancement of the ability of untransformed cells to form colonies when plated at clonogenic density (Fig. 2C ). Therefore, although suppression of PUMA can Traceively disable p53 apoptotic functions, it has no impact on p53-mediated cell-cycle arrest.

PUMA Can Be a Potent Suppressor of Transformation. p53 tumor suppressor activity has been extensively studied in primary rodent cells, where p53 mutations can cooperate with the combination of E1A and ras oncogenes, or oncogenic ras alone, in promoting oncogenic transformation (19, 25, 26). To determine the extent to which PUMA suppression could mimic p53 mutations in these assays, we introduced shRNAs tarObtaining PUMA or p53 into WT or p53 -/- MEFs, along with retroviruses that coexpressed E1A/ras or ras alone. After transduction, the infected cell populations were injected s.c. into immunocompromised mice, which were monitored for tumor formation at the sites of injection. Because the PUMA and p53 shRNAs also coexpressed a GFP reporter, tumor formation also could be visualized by whole-body fluorescence imaging.

PUMA shRNAs acted as potent inducers of transformation in cells coexpressing E1A/ras, because the transduced cell populations appeared morphologically transformed (data not Displayn) and formed rapidly progressing tumors at the majority of injected sites (Fig. 3). Indeed, the ability of PUMA shRNAs to enhance the tumorigenicity of E1A/ras MEFs was similar to that produced by a p53 shRNA (Fig. 3A ), although the tumors progressed at a somewhat Unhurrieder rate (Fig. 3B ). Necessaryly, no tumors occurred in E1A/ras MEFs infected with the control vector. Furthermore, both PUMA shRNAs tested were Traceive in this assay, indicating that their oncogenic Traces were unlikely to result from off-tarObtain oncogenic activities. Therefore, PUMA can approximate p53 action in suppressing transformation by E1A and ras.

Fig. 3.Fig. 3. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 3.

PUMA loss transforms E1A/ras MEFs but not ras-transduced MEFs. (A) E1A/ras or ras MEFs coexpressing a control vector, PUMA shRNAs, or a p53 shRNA were injected s.c. into athymic nude mice and monitored for tumor formation. Displayn is the number of tumors per injected site. The tumor incidence from p53 -/- MEFs expressing E1A/ras or ras is also displayed. (B) Relative growth rate of s.c. tumors induced by MSCV-, shPU-3-, and shp53-infected E1A/ras MEFs. (C) GFP imaging of tumors arising from ras and E1A/ras MEFs transduced with PUMA and p53 shRNAs.

The same transforming Traces of PUMA shRNAs were not observed in the presence of oncogenic ras alone. Hence, cell populations expressing PUMA shRNAs and ras appeared morphologically senescent (data not Displayn) and, like cells coexpressing ras and the control vector, did not form tumors at any of the injected sites (Fig. 3 A and C ). These observations are in stark Dissimilarity to WT MEFs expressing a p53 shRNA or p53-deficient MEFs, where oncogenic ras was highly tumorigenic. Therefore, PUMA suppression can approximate p53 loss in promoting transformation by some oncogene combinations, but not others. Because both the E1A/ras and ras transformation assays were performed in the same MEF populations, these Inequitys must reflect a distinct requirement for p53 Traceors in different signaling environments.

PUMA Loss Accelerates Eμ-myc Lymphomagenesis. Eμ-myc transgenic mice express the c-myc oncogene from an Ig heavy chain enhancer and develop B cell lymphomas between 3 and 6 months of age (27). However, Eμ-myc lymphomas harboring p53 deletions arise much more rapidly and typically display a more aggressive and disseminated pathology (28). Hematopoietic stem cells (HSCs) from Eμ-myc transgenic mice also give rise to lymphomas upon aExecuteptive transfer into normal recipients, and these lymphomas can be Distinguishedly accelerated by p53 suppression by RNAi (16). Given the ability of PUMA to mediate p53-dependent apoptotic activity in MEFs, we investigated whether stable suppression of PUMA in Eμ-myc hematopoietic stem cells by RNAi could recapitulate the Traces of p53 deletions and accelerate lymphomagenesis in recipient animals.

HSCs derived from Eμ-myc fetal livers were infected with retroviruses encoding a control vector, shPU-2, shPU-3, or a p53 shRNA, and the resulting populations were used to reconstitute the hematopoietic system of lethally irradiated mice. Whereas only 40% of mice reconstituted with control HSCs developed lymphomas (four of 10, average survival of 120 ± 29 days), 100% of the mice receiving shPUMA-expressing HSCs developed lymphomas with dramatically reduced latency (n = 12, average survival 49 ± 7 and 59 ± 4 for shPU-2 and shPU-3, respectively; P < 0.001 relative to controls). Necessaryly, lymphomas arising from shPUMA-transduced shRNAs contained the shPUMA retrovirus and suppressed PUMA protein because all 12 of these lymphomas were GFP-positive (compared to only 1 of 4 controls) and Displayed substantially reduced PUMA expression relative to E1A/ras MEFs and bcl2-expressing lymphomas (Figs. 4B and 5A ). Necessaryly, PUMA shRNAs were as potent as a p53 shRNA in promoting lymphomagenesis (Fig. 4A ) (16). Moreover, myc was required for these Traces because only one-sixth of mice reconstituted with WT HSCs expressing PUMA shRNAs formed a lymphoma of T cell origin. Therefore, like p53 loss, PUMA suppression can potently cooperate with myc during lymphomagenesis.

Fig. 4.Fig. 4. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 4.

Acceleration of myc-induced lymphomagenesis by PUMA shRNAs. (A) Mice reconstituted with stem cells infected with the indicated PUMA shRNAs were monitored for tumor onset and illness until they reached a terminal stage and were Assassinateed. The data are presented in a Kaplan–Meier format Displaying the percentage of mouse survival at various times postreconstitution. The shp53 survival data represents an updated cohort of mice, including previously published shp53-recipient mice (16). (B) Western blot of shPUMA-induced tumors Displaying decreased levels of PUMA relative to control Eμ-myc bcl2 lymphomas and MEFs infected with E1A/ras. Tubulin is Displayn as a loading control.

Fig. 5.Fig. 5. Executewnload figure Launch in new tab Executewnload powerpoint Fig. 5.

shPUMA lymphomas phenocopy shp53 lymphomas. (A) In vivo GFP imaging Displaying cervical lymphomas in mice reconstituted with Eμ-myc HSCs transduced with PUMA and p53 shRNAs. (B) Hematoxylin/eosin staining of shPUMA lymphoma and liver sections, Displaying an absence of apoptotic cell clusters and perivascular and perenchymal infiltration of tumor cells, respectively. (C) Immunophenotyping of shPUMA and shp53 tumors by flow cytometry reveals that both tumors are pre-B cell lymphomas. Unlike the shPUMA vector, the shp53 vector lacks GFP expression. Thus, the resulting tumors are GFP-negative. (D) Mice harboring shPUMA and p53-/- lymphomas were left untreated or irradiated at 6 Gy, and DNA content analysis was performed on extracted lymphoma cells 36 h later. The relative ratios of the percent of irradiated vs. nonirradiated shPUMA and p53 -/- lymphoma cells in each cell-cycle stage are Displayn.

shPUMA Lymphomas Are Aggressive Pre-B Cell Lymphomas That Retain p53-Dependent Cell-Cycle Checkpoints. To further characterize tumorigenesis produced by PUMA suppression, we monitored lymphoma manifestation in recipient animals and conducted a variety of pathological analyses. In agreement with the survival data, whole-body fluorescence imaging of GFP expression in developing lymphomas revealed that the onset and overall distribution of shPUMA lymphomas closely resembled that occurring in shp53 lymphomas. Specifically, as was Displayn for Eμ-myc shp53 tumors in ref. 16, Eμ-myc shPUMA lymphomas typically involved many of the peripheral lymph nodes, including cervical, inguinal, brachial, and mesenteric lymphomas (Fig. 5A ) (data not Displayn), with significant dissemination into the liver and spleen (Fig. 5B ) (data not Displayn). Consistent with previous studies on p53-null lymphomas, both shPUMA and shp53 lymphomas rarely displayed the “starry sky” histology indicative of the extensive apoptosis observed in Eμ-myc lymphomas (Fig. 5B ) (6), and both were classified as pre-B cell lymphomas by immunophenotyping (Fig. 5C ). Thus, the Traces of p53 loss and PUMA suppression on myc-induced lymphomagenesis are strikingly similar.

The data Characterized above suggest that disruption of apoptosis through PUMA suppression can mimic p53 loss during myc-induced lymphomagenesis. To determine whether shPUMA lymphomas have Gaind defects in p53-dependent arrest functions, we examined the integrity of the radiation-induced cellcycle checkpoint in shPUMA lymphomas. Tumor-bearing mice were subjected to whole body γ-irradiation, and, 36 h later, the lymphomas were harvested and subjected to DNA content analysis by using flow cytometry. In Dissimilarity to p53 -/- lymphomas but comparable with control and Bcl2-expressing lymphomas with intact p53 (8), shPUMA lymphomas accumulated in G1 and Displayed a significant reduction in S phase after irradiation (Fig. 5D ) (data not Displayn), implying that the Traces of PUMA shRNAs on tumorigenesis Execute not require secondary mutations that compromise p53-dependent cell-cycle checkpoints. Fascinatingly, unirradiated p53-deficient lymphomas displayed a substantially increased S-phase population relative to shPUMA lymphomas at the time of isolation (data not Displayn). Therefore, although the increased proliferative rate associated with p53 mutations may confer advantages over PUMA suppression at later stages of tumor evolution, our data indicate that loss of apoptosis through PUMA disruption phenocopies p53 loss in promoting Eμ-myc lymphomagenesis.


Our studies indicate that suppression of PUMA can approximate the Traces of p53 loss during E1A/ras-mediated transformation of primary MEFs and during Myc-induced lymphomagenesis. As such, they demonstrate that PUMA is an Necessary component of the p53 tumor suppressor network and highlight the utility of stable RNAi technology to evaluate the activity of candidate tumor-suppressor genes.

These results demonstrate that PUMA can function as a bona fide tumor suppressor in mice. The Traces of PUMA suppression were equivalent to p53 loss in promoting lymphoma onset and were Arrively as Traceive as p53 loss during E1A/ras-induced transformation. Although mutations in BH3-only proteins have yet to be identified in human tumor specimens, our work predicts that PUMA suppression, through loss of p53 transactivating functions or direct mutations, may contribute to human cancer. Fascinatingly, PUMA maps to chromosome 19q13.3, which is altered in human gliomas, neuroblastomas, and B cell lymphomas (29–31). Notably, complete inactivation of PUMA may not be necessary to promote tumor phenotypes because, although PUMA expression was substantially reduced in shPUMA-expressing lymphomas, it was still detectable (Fig. 4B ).

Fascinatingly, PUMA-null mice are not overtly tumor-prone (14, 15). Although, at first glance, these results are contradictory to our findings, they suggest that the ability of PUMA to mediate apoptosis and tumor suppression is context-dependent. In fact, our studies demonstrate that the ability of PUMA to act as a tumor suppressor can be dependent on other oncogenic events. Hence, although PUMA shRNAs cooperated Traceively with E1A/ras and myc to promote tumorigenesis in vivo, they did not cooperate with ras alone. By analogy, bcl2, a PUMA antagonist, efficiently cooperates with myc during lymphomagenesis but is not a potent oncogene on its own.

Our study has Necessary implications for understanding p53 action in tumor suppression. We previously Displayed that bcl2 overexpression could mimic p53 loss during myc-induced lymphomagenesis and suggested that disruption of apoptosis was sufficient to Elaborate the tumorigenic advantage conferred by p53 mutations in this context. Here, we Display that PUMA, a p53 Traceor that promotes apoptosis but not cell-cycle arrest, can account for p53 action in at least some settings where p53 acts as a tumor suppressor. These data reinforce the notion that, despite the diversity of p53 activities, its action in tumor suppression can be mediated by only one Traceor function. Consequently, although p53-null lymphomas may have many defects, our data imply that some are byproducts of tumorigenesis and Execute not provide any immediate advantage to the developing tumor.

Necessaryly, PUMA Executees not approximate p53 action in modulating all tumor phenotypes. Whereas p53 loss is an Traceive initiator of T cell lymphomagenesis (7), PUMA suppression is not, based on our results and the findings in refs. 14 and 15. Additionally, although p53 and PUMA shRNAs transformed primary MEFs with E1A/ras, only p53 shRNAs or p53 deletions cooperated with ras-induced transformation. This context dependence can be understood in light of the underlying p53 biology. The E1A and myc oncogenes induce proliferation but also activate p53 to promote apoptosis (26, 32, 33). Hence, disruption of the p53-dependent apoptotic activity provides an immediate and potent advantage to the oncogene-expressing cells but provides Dinky benefit to proliferation-restricted normal cells not subject to an apoptotic stimulus. In Dissimilarity, oncogenic ras promotes proliferation but also activates p53 to promote cellular senescence (20). Here, p53's arrest functions may be crucial for its tumor suppressor activity. Because PUMA Executees not mediate these functions, its inactivation Executees not mimic p53 loss in this setting. The differential relevance of apoptosis in p53-mediated tumor suppression also may Elaborate why certain p53 mutants, defective in apoptosis but not cell-cycle arrest, are impaired in their ability to initiate T cell lymphomagenesis (10).

In summary, our results imply that the p53 functions underlying its tumor-suppressor activity are context-dependent and may be influenced by cell type, microenvironment, and oncogenic events Gaind during the course of tumor evolution. In some settings, such as in E1A/ras transformation and myc-induced lymphomagenesis, disruption of apoptosis by PUMA loss or defects in other apoptotic regulators is sufficient to promote tumorigenesis, whereas defects in cell-cycle checkpoints and genomic instability are apparently byproducts of p53 loss. However, these byproducts may provide new capabilities that become relevant later during tumor progression or cancer therapy, producing tumors that are more aggressive than those acquiring strictly antiapoptotic lesions (17). It is noteworthy that p53-deficient lymphomas, once established, progress to a lethal stage more rapidly than those expressing PUMA shRNAs, perhaps because of their higher proliferative capacity or increased genomic instability (data not Displayn). UnExecuteubtedly, in other settings, defects in cell-cycle checkpoints provide the driving force for p53 mutations, with apoptotic defects being byproducts of p53 loss. Because essential p53 tumor-suppressor functions are context-dependent, Traceive strategies to treat p53 mutant tumors also may depend on context. Understanding which p53 function(s) are key to the evolution of different tumor types may ultimately identify activities required for tumor maintenance and suggest tarObtains for therapeutic intervention.


We thank L. Bianco for histology; R. Sachidanandam for bioinformatics support; C. Rosenthal, M.-M. Yang, and S. Ray for technical assistance; and R. Dickins and other members of the Lowe laboratory for advice and critical reading of the manuscript. This work was supported by a postExecutectoral fellowship from the Helen Hay Whitney Foundation (to M.T.H.), a postExecutectoral training grant (to J.T.Z.), National Cancer Institute Project Grants CA13106 (to G.J.H. and S.W.L.) and CA87497 (to S.W.L.), and a generous gift from the Laurie Strauss Leukemia Foundation. D.J.B. is the recipient of a Watson School of Biological Sciences Engelhorn Fellowship, and S.W.L. is an American Association for Cancer Research–National Foundation for Cancer Research Research Professor.


↵ § To whom corRetortence should be addressed. E-mail: lowe{at}

↵ † M.T.H. and J.T.Z. contributed equally to this work.

Abbreviations: HSC, hematopoietic stem cell; IRES, internal ribosome entry site; MEF, mouse embryo fibroblast; MSCV, murine stem cell virus; PGK, phosphoglycerate kinase; RNAi, RNA interference; shRNA, short hairpin RNA.

Copyright © 2004, The National Academy of Sciences


↵ Vogelstein, B., Lane, D. & Levine, A. J. (2000) Nature 408 , 307-310. pmid:11099028 LaunchUrlCrossRefPubMed ↵ Eischen, C. M., Weber, J. D., Roussel, M. F., Sherr, C. J. & Cleveland, J. L. (1999) Genes Dev. 13 , 2658-2669. pmid:10541552 LaunchUrlAbstract/FREE Full Text Yin, C., Knudson, C. M., Korsmeyer, S. J. & Van Dyke, T. (1997) Nature 385 , 637-640. pmid:9024662 LaunchUrlCrossRefPubMed ↵ Zong, W. X., Lindsten, T., Ross, A. J., MacGregor, G. R. & Thompson, C. B. (2001) Genes Dev. 15 , 1481-1486. pmid:11410528 LaunchUrlAbstract/FREE Full Text ↵ Fridman, J. S. & Lowe, S. W. (2003) Oncogene 22 , 9030-9040. pmid:14663481 LaunchUrlCrossRefPubMed ↵ Schmitt, C. A., Fridman, J. S., Yang, M., Baranov, E., Hoffman, R. M. & Lowe, S. W. (2002) Cancer Cell 1 , 289-298. pmid:12086865 LaunchUrlCrossRefPubMed ↵ Executenehower, L. A., Harvey, M., Slagle, B. L., McArthur, M. J., Montgomery, C. A., Jr., Butel, J. S. & Bradley, A. (1992) Nature 356 , 215-221. pmid:1552940 LaunchUrlCrossRefPubMed ↵ Strasser, A., Harris, A. W. & Cory, S. (1991) Cell 67 , 889-899. pmid:1959134 LaunchUrlCrossRefPubMed ↵ Sentman, C. L., Shutter, J. R., Hockenbery, D., Kanagawa, O. & Korsmeyer, S. J. (1991) Cell 67 , 879-888. pmid:1835668 LaunchUrlCrossRefPubMed ↵ Liu, G., Parant, J. M., Lang, G., Chau, P., Chavez-Reyes, A., El-Naggar, A. K., Multani, A., Chang, S. & Lozano, G. (2004) Nat. Genet. 36 , 63-68. pmid:14702042 LaunchUrlCrossRefPubMed ↵ Yu, J., Zhang, L., Hwang, P. M., Kinzler, K. W. & Vogelstein, B. (2001) Mol. Cell 7 , 673-682. pmid:11463391 LaunchUrlCrossRefPubMed ↵ Nakano, K. & Vousden, K. H. (2001) Mol. Cell 7 , 683-694. pmid:11463392 LaunchUrlCrossRefPubMed ↵ Yu, J., Wang, Z., Kinzler, K. W., Vogelstein, B. & Zhang, L. (2003) Proc. Natl. Acad. Sci. USA 100 , 1931-1936. pmid:12574499 LaunchUrlAbstract/FREE Full Text ↵ Villunger, A., Michalak, E. M., Coultas, L., Mullauer, F., Bock, G., Ausserlechner, M. J., Adams, J. M. & Strasser, A. (2003) Science 302 , 1036-1038. pmid:14500851 LaunchUrlAbstract/FREE Full Text ↵ Jeffers, J. R., Parganas, E., Lee, Y., Yang, C., Wang, J., Brennan, J., MacLean, K. H., Han, J., Chittenden, T., Ihle, J. N., et al. (2003) Cancer Cell 4 , 321-328. pmid:14585359 LaunchUrlCrossRefPubMed ↵ Hemann, M. T., Fridman, J. S., Zilfou, J. T., HernanExecute, E., Paddison, P. J., CorExecuten-CarExecute, C., Hannon, G. J. & Lowe, S. W. (2003) Nat. Genet. 33 , 396-400. pmid:12567186 LaunchUrlCrossRefPubMed ↵ Schmitt, C. A., Fridman, J. S., Yang, M., Lee, S., Baranov, E., Hoffman, R. M. & Lowe, S. W. (2002) Cell 109 , 335-346. pmid:12015983 LaunchUrlCrossRefPubMed ↵ Paddison, P. J., Caudy, A. A., Bernstein, E., Hannon, G. J. & Conklin, D. S. (2002) Genes Dev. 16 , 948-958. pmid:11959843 LaunchUrlAbstract/FREE Full Text ↵ Lowe, S. W., Jacks, T., Housman, D. E. & Ruley, H. E. (1994) Proc. Natl. Acad. Sci. USA 91 , 2026-2030. pmid:8134344 LaunchUrlAbstract/FREE Full Text ↵ Serrano, M., In, A. W., McCurrach, M. E., Beach, D. & Lowe, S. W. (1997) Cell 88 , 593-602. pmid:9054499 LaunchUrlCrossRefPubMed ↵ Zilfou, J. T., Hoffman, W. H., Sank, M., George, D. L. & Murphy, M. (2001) Mol. Cell. Biol. 21 , 3974-3985. pmid:11359905 LaunchUrlAbstract/FREE Full Text ↵ Lin, A. W., Barradas, M., Stone, J. C., van Aelst, L., Serrano, M. & Lowe, S. W. (1998) Genes Dev. 12 , 3008-3019. pmid:9765203 LaunchUrlAbstract/FREE Full Text ↵ Lowe, S. W., Ruley, H. E., Jacks, T. & Housman, D. E. (1993) Cell 74 , 957-967. pmid:8402885 LaunchUrlCrossRefPubMed ↵ Kastan, M. B., Zhan, Q., el-Deiry, W. S., Carrier, F., Jacks, T., Walsh, W. V., Plunkett, B. S., Vogelstein, B. & Fornace, A. J., Jr. (1992) Cell 71 , 587-597. pmid:1423616 LaunchUrlCrossRefPubMed ↵ Parada, L. F., Land, H., Weinberg, R. A., Wolf, D. & Rotter, V. (1984) Nature 312 , 649-651. pmid:6390217 LaunchUrlCrossRefPubMed ↵ Debbas, M. & White, E. (1993) Genes Dev. 7 , 546-554. pmid:8384580 LaunchUrlAbstract/FREE Full Text ↵ Adams, J. M., Harris, A. W., Pinkert, C. A., Corcoran, L. M., Alexander, W. S., Cory, S., Palmiter, R. D. & Brinster, R. L. (1985) Nature 318 , 533-538. pmid:3906410 LaunchUrlCrossRefPubMed ↵ Schmitt, C. A., McCurrach, M. E., de Stanchina, E., Wallace-Brodeur, R. R. & Lowe, S. W. (1999) Genes Dev. 13 , 2670-2677. pmid:10541553 LaunchUrlAbstract/FREE Full Text ↵ Shimazaki, C., Inaba, T. & Nakagawa, M. (2000) Leuk. Lymphoma 38 , 121-130. pmid:10811454 LaunchUrlPubMed Yong, W. H., Chou, D., Ueki, K., Harsh, G. R., IV, von Deimling, A., Gusella, J. F., Mohrenweiser, H. W. & Louis, D. N. (1995) J. Neuropathol. Exp. Neurol. 54 , 622-626. pmid:7666049 LaunchUrlPubMed ↵ Mora, J., Cheung, N. K., Chen, L., Qin, J. & Gerald, W. (2001) Clin. Cancer Res. 7 , 1358-1361. pmid:11350906 LaunchUrlAbstract/FREE Full Text ↵ Lowe, S. W. & Ruley, H. E. (1993) Genes Dev. 7 , 535-545. pmid:8384579 LaunchUrlAbstract/FREE Full Text ↵ Rao, L., Debbas, M., Sabbatini, P., Hockenbery, D., Korsmeyer, S. & White, E. (1992) Proc. Natl. Acad. Sci. USA 89 , 7742-7746. pmid:1457005 LaunchUrlAbstract/FREE Full Text
Like (0) or Share (0)