INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Murine double minute 2 (mdm2), which was first described as one of the genes amplified on the double minute chromosomes present in the spontaneously transformed BALB/c 3T3 murine cell line 3T3 DM (6), is an evolutionarily conserved gene whose protein product, MDM2, is endowed with growth-promoting activity. The overexpression of MDM2 can immortalize rodent primary fibroblasts and transform a variety of cell lines (14), and the human homologue of mdm2, hdm2, is frequently overexpressed in tumors (44). MDM2 can antagonize transcriptional activation by p53 (39) and p53-dependent growth arrest (8), which is mediated through the ability of the N-terminal region of MDM2 to bind to the N-terminal trans activation domain of p53 (7, 32, 39). Furthermore, mice deficient for mdm2 are not viable unless they lack the p53 gene (23, 33), suggesting that the regulation of p53 is an important physiological function of MDM2. In addition to its ability to antagonize p53-dependent transcription, MDM2 can promote the degradation of p53, apparently through a ubiquitin-dependent proteasome pathway (18, 28), and recent studies have shown that MDM2 undergoes nucleocytoplasmic export and continuously shuttles between the nucleus and cytoplasm, and a nuclear export signal (NES) in MDM2 may be involved in a similar export mechanism similar to that of the human immunodeficiency virus rev protein (10, 46). Some evidence has suggested that the inhibition of MDM2 export interferes with the ability of MDM2 to destabilize p53 (46).

Also, MDM2 has been shown to functionally interact with a number of other proteins, including the retinoblastoma tumor suppressor protein pRb (52), E2F (38), the ribosomal protein L5 (37), and the human TATA box binding protein (32), and the C-terminal RING finger in MDM2 may interact with RNA (13). Overall, it appears that although the regulation of p53 is an important target in mediating the oncogenic effects of MDM2, it is likely that additional proteins can be subject to control by MDM2.

The E2F family of transcription factors plays an important role in orchestrating early cell cycle progression (12, 30). Hypophosphorylated pRb, which is active in negative growth control, binds to E2F and thereafter becomes progressively phosphorylated during the cell cycle through the sequential activity of cdk's, with the subsequent release of E2F allowing the transcriptional activation of target genes required for the early cell cycle. In mammalian cells E2F is composed of one of six E2F family members bound to a DP family member (12, 29), in which the transcriptional activity of E2F is mediated through a C-terminal activation domain that is under the control of pRb or its relatives p107 and p130, proteins that physically interact with E2F in a heterodimer and in a temporally specific fashion (12). Many target genes that are subject to control by E2F have been identified. In general, these genes encode proteins that are required for cell cycle progression and, for example, include enzymes for DNA synthesis, such as dihyduofolate reductase, or regulatory proteins, such as cyclins A and E (50).

It has been shown that MDM2 can bind to pRb through a domain located in the C-terminal region of pRb which overlaps with a domain required for the interaction of pRb with E2F and for growth suppression (52). Furthermore, a direct interaction with both the E2F and DP subunits of the E2F heterodimer may be important in augmenting the response of E2F-regulated genes (38), and in p53^/ tumor cells, MDM2 can override pRb- or p107-induced growth arrest (11). In p53^/ Rb^/ cells MDM2 can mediate anchorage-independent growth, an effect that requires a domain in MDM2 suggested to be involved in E2F regulation (11). Therefore, considerable evidence supports a functional connection between MDM2 and E2F.

An overwhelming body of evidence has established a role for E2F in promoting proliferation, although certain E2F family members may also function as negative regulators of cell growth (12). For example, overexpression of E2F-1 promotes cell cycle progression and is sufficient to drive quiescent cells into S phase (22), and both E2F-1 and DP-1 can cooperate with an activated ras oncogene to transform cells (22). However, E2F-1^/ knockout mice suffer a significant increase in the incidence of tumors, arguing that E2F-1 is also endowed with growth-inhibitory and tumor suppressor activities (15, 52, 53). Indeed, this apparent dichotomous function of E2F-1 may be partially attributable to its apoptotic function, as the overexpression of E2F-1 can not only cause quiescent cells to enter S phase but also promote apoptosis (25, 45, 48, 51).

We have explored the regulation of apoptosis by E2F in different types of p53^/ cells and specifically addressed the influence of hDM2 on apoptosis. We find that hDM2 can prevent cells from entering E2F-dependent apoptosis and further that this outcome is critically dependent on the DP component of the E2F heterodimer. Cells that have been rescued from apoptosis by hDM2 contain lower levels of E2F, which we show is dependent on the presence of the DP subunit and on the ability of MDM2 to promote E2F degradation. Moreover, the regulation of E2F correlates with an hDM2-dependent influence on the nuclear accumulation of DP-1, suggesting that an altered intracellular distribution is important in mediating the physiological effects of hDM2. Most importantly, in rescued cells hDM2 and E2F-DP cooperate in promoting DNA synthesis and cell viability. These data define a new level of interplay between the pathways of control mediated by hDM2 and E2F that occurs independently of p53. Furthermore, they support a model whereby in tumor cells hDM2 can antagonize the apoptotic properties of E2F and thereby maintain E2F in a permanent state of growth stimulation.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Cell culture and transfection. SAOS2 cells and early-passage p53^/ mouse embryo fibroblasts (MEFs) (35) were maintained in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% fetal calf serum. For transient transfections, immunostaining, and terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end labeling (TUNEL) and bromodeoxyuridine (BrdU) assays, about 3 × 10⁵ SAOS2 cells or 2 × 10⁵ p53^/ MEFs were seeded onto glass coverslips in 30-mm-diameter dishes and transfected 12 h later using the calcium phosphate procedure described previously (54), in which the total amount of transfected DNA was made equivalent with empty vector. The cells were transfected for 6 h, washed, and grown under low-serum conditions (0.2% fetal calf serum) for a further 18 h prior to being fixed for immunostaining, BrdU staining, or TUNEL assay. Transfection efficiency usually was between 18 and 24%.

Expression vectors. The plasmids pCMVHA-HA-E2F-5 and pHA-E2F-5TAD were described previously (1). The plasmid pCMVE2F-4, which encodes human E2F-4, was described previously (4). The plasmid pRcCMVHA-E2F-1wt, encoding human E2F-1, was a kind gift from W. Krek (26). Plasmids pG4-DP-1 and the DP-1 derivatives pG4DP3, pG4DP3E, pG4DP3DEF, and pCMV gal have been previously described (3, 9, 31). Plasmid pC53-SN3, expressing wild-type p53 (2), and the hDM2 expression vectors (7, 8) have been previously described.

Apoptosis TUNEL assay and immunostaining. Cells on coverslips were transfected as indicated in the figures and, at the appropriate time, fixed for 30 min in 4% paraformaldehyde in phosphate-buffered saline (PBS) at room temperature. The cells were permeabilized at 4°C for 2 min in 0.1% Triton X-100-0.1% sodium citrate and treated for an additional 10 min with 10% fetal calf serum in PBS. TUNEL staining (Boehringer Mannheim) for apoptotic cells was carried out according to the manufacturer's instructions. For immunostaining, the coverslips were incubated with the primary antibody for 30 min at room temperature and washed in PBS containing 1% bovine serum albumin. Rabbit polyclonal antibodies were used at a 1:200 dilution, and mouse monoclonal antibodies were used at a 1:1,000 dilution unless otherwise stated. Secondary antibodies, either anti-mouse fluorescein isothiocyanate- or tetramethyl rhodamine isothiocyanate-conjugated or anti-rabbit fluorescein isothiocyanate- or tetramethyl rhodamine isothiocyanate-conjugated antibodies (Southern Biotechnology Associates Inc.), were incubated for '30 min at room temperature and washed in PBS containing 1% bovine serum albumin with 1 µg of DAPI (4',6'-diamidino-2-phenylindole)/ml to visualize nuclei. Transfected cells were identified by immunostaining them with an appropriate antibody, and the level of apoptosis is expressed relative to the number of cells transfected (antihemagglutinin [HA] was used to detect E2F-1, E2F-4, and E2F-5; anti-DP-1 was used to detect DP-1; and Ab-1 [Calbiochem] was used to detect hDM2); the data shown were derived from at least three independent experiments.

BrdU staining. Following transfection and overnight incubation in 0.2% serum, cells were incubated with BrdU (Boehringer Mannheim) for between 5 and 10 min. The cells were fixed in 70% ethanol-2 mM glycine (pH 2.0) for 20 min at 20°C, and BrdU incorporation was detected according to the manufacturer's instructions. Transfected cells were identified by immunostaining them with an appropriate antibody, and the level of BrdU incorporation was expressed relative to the total number of cells transfected.

Immunoblotting. For immunoblotting, about 5 × 10⁵ SAOS2 cells or 3.5 × 10⁵ p53^/ MEFs were seeded in 60-mm-diameter dishes and transfected with plasmid as described above. The cells were harvested, snap frozen, and subsequently lysed in sodium dodecyl sulfate sample buffer before separation on a 10% polyacrylamide gel. Protein loading was standardized by -galactosidase activity (derived from transfected pCMV gal) and carried out as described previously (9). The results shown are representative of three independent experiments.

Colony assay. For the colony assay, about 5 × 10⁵ SAOS2 cells in 60-mm-diameter dishes were transfected with 12 µg of each plasmid. Equal amounts of neoresistant plasmid DNA were added to each transfection. The cells were transfected for 6 h, split, and grown in 10% DMEM with 0.5 mg of G418 (Sigma)/ml added on the third day to select for resistant colonies. Alternatively, following transfection, the cells were incubated for 18 h in low serum prior to being split and subjected to selection. The data shown were derived from three independent experiments.

Antibodies. Rabbit polyclonal antibodies to DP-1 and DP-3 have been described previously (9). Other primary antibodies used were the anti-E2F-1 mouse monoclonal antibody KH95 (Santa Cruz), anti-p53 mouse monoclonal antibody DO1 (Santa Cruz), the anti-MDM2 mouse monoclonal antibody Ab-1, and the anti-HA mouse polyclonal antibody (BabCo).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Apoptotic properties of E2F family proteins. It is known that E2F-1 can induce apoptosis (45, 48, 51), but the properties of the other family members, such as E2F-4 and -5, have not been elucidated. To explore their apoptotic activities, we compared the properties of exogenous E2F-1 to those of E2F-4 and -5 in SAOS2 osteosarcoma cells into which E2F expression vectors were transfected under low-serum growth conditions (31). SAOS2 cells, which are p53^/ Rb^/, undergo apoptosis upon the introduction of exogenous E2F-1 (45). By TUNEL assay, and consistent with previous reports, E2F-1 could promote apoptosis, whereas under the same conditions, E2F-4 and E2F-5 had negligible activity (Fig. 1a, compare bar 2 to 3 and 6).

View larger version (45K): [in this window] [in a new window]   FIG. 1.   Properties of E2F responsible for apoptosis. (a) SAOS2 cells were transfected as indicated with 6 µg of E2F-1 (bar 2), 6 µg of E2F-4 (bar 3), or 6 µg of E2F-5 (bar 6) alone or with (+) 6 µg of DP3- (bars 4 and 7) or 6 µg of DP3-E (bars 5 and 8). After 6 h, the cells were washed and further grown overnight in 0.2% fetal calf serum in DMEM. Apoptosis was assayed ~18 h after transfection, and apoptotic cells were quantitated as described in the text. The total amount of transfected DNA was made up with pcDNA3. The percentage of apoptosis is expressed relative to the total number of transfected cells, and the data shown represent the average values (+ standard deviations) from three independent experiments. (b) The intracellular distribution of the indicated E2F proteins was determined under apoptotic conditions after SAOS2 cells were transfected with 6 µg of E2F-1 (i), 6 µg of E2F-4 (ii), 6 µg of both E2F-4 and DP-3 (iii), 6 µg of both E2F-4 and DP-3E (iv), 6 µg of E2F-5 (v), or 6 µg of both E2F-5 and DP-3 (vi) and treated as described. Anti-HA was used to detect E2F-1 (i), -4 (ii, iii, and iv), and -5 (v and vi). (c) SAOS2 cells were transfected with 6 µg of E2F-5 (bars 2 and 3) or 6 µg of E2F-5TAD (bars 4 and 5) together with 6 µg of DP-3 (bars 2 and 4) or 6 µg of DP-3DEF (bars 3 and 5) and treated as described for panel (a). (d) The intracellular distribution of the indicated proteins was determined under apoptotic conditions after SAOS2 cells were transfected with 6 µg of E2F-5 together with 6 µg of DP-3 (i) or DP-3DEF (ii). The intracellular distribution of E2F-5 is shown.

hDM2 overcomes E2F-mediated apoptosis in SAOS2 p53^/ cells. The fact that MDM2 activity is integrated with the pRb-E2F pathway of growth control (38, 52) prompted us to examine the possibility that hDM2 could alter the apoptotic activity of E2F; experiments were performed with p53^/ cells to rule out any influence of p53. For these studies, we capitalized on the information described above and in the first instance explored the influence of hDM2 on E2F-dependent apoptosis in SAOS2 cells. Therefore, hDM2 was expressed in cells undergoing apoptosis as a result of E2F-1 expression, and the level of apoptotic cells was compared to that of transfected cells expressing both components of the physiological heterodimer, namely E2F-1 and DP-1. We found that the apoptotic activity of E2F-1, when expressed alone, was not affected by hDM2, as the number of apoptosing cells remained at a level very similar to that caused by E2F-1 alone (Fig. 2a, compare bars 3 and 4), a result consistent with earlier studies, which established that hDM2 fails to overcome E2F-1-dependent apoptosis (21). Remarkably, however, although the level of apoptosis was only marginally greater when DP-1 and E2F-1 were coexpressed (Fig. 2a, compare bars 3 and 5), there was a highly significant decline in apoptosis upon the expression of hDM2 (Fig. 2a, compare bars 5 and 6), similar results being obtained with DP-3 (data not shown). Thus, hDM2 can down-regulate E2F-dependent apoptosis, but it does so in a fashion that reflects the presence of the DP subunit.

View larger version (33K): [in this window] [in a new window]   FIG. 2.   hDM2 overcomes E2F-dependent apoptosis in a DP subunit-dependent fashion in SAOS2 cells. (a) SAOS2 cells were transfected as indicated with 6 µg of E2F-1 (bars 3 to 6), 6 µg of DP-1 (bars 5 and 6), or 6 µg of hDM2 (bars 2, 4, and 6). After 6 h, the cells were washed and grown overnight in 0.2% fetal calf serum in DMEM. Apoptosis was assayed ~18 h after transfection, and apoptotic cells were quantitated as described in the text. The percentage of apoptosis is expressed relative to the total number of transfected cells, and the data shown represent the average values (+ standard deviations) from three independent experiments. +, present; , absent. (b) SAOS2 cells were transfected as indicated with 6 µg of E2F-4 (bars 3, 4, and 5) or E2F-5 (bars 6, 7, and 8) together with 6 µg of DP-3 (bars 4, 5, 7, and 8) and 6 µg of hDM2 (bars 2, 5, and 8) and treated as described for panel a. (c) The intracellular distribution of the indicated proteins was determined under apoptotic conditions after SAOS2 cells were transfected with the appropriate expression vectors encoding the indicated proteins as described in for panel a. Anti-HA was used to detect E2F-1, anti-DP-1 was used to detect DP-1, and anti-MDM2 was used to detect hDM2 (data not shown). The transfected expression vectors are indicated, and the immunostained protein is shown by white lettering. Note that iii and v, and iv and vi, show the same field stained with either anti-HA (iii and iv) or anti-DP-1 (v and vi) and that the images shown result from the same time of exposure.

hDM2 down-regulates the levels of E2F and DP proteins. Since the immunostaining results (Fig. 2) suggested that hDM2 can down-regulate the levels of the E2F and DP subunits, we tested this idea by directly measuring E2F and DP levels under conditions of E2F-dependent apoptosis. Firstly, we studied the effects of hDM2 on E2F-1, either alone or together with DP-1. Although the presence of hDM2 failed to affect the level of E2F-1 (Fig. 3a, compare lanes 4 and 5), a clear reduction in E2F-1 was apparent in cells that coexpressed DP-1 (Fig. 3a, compare lanes 6 and 7), thus suggesting that a DP partner is necessary to allow hDM2 to modulate the levels of E2F-1. In contrast to E2F-1, the level of DP-1 was reduced either in the presence or absence of E2F-1 upon coexpressing hDM2 (Fig. 3b, compare lane 6 with 7 and lane 8 with 9), consistent with the idea that the DP subunit is the component in the E2F heterodimer that is targeted by hDM2.

View larger version (22K): [in this window] [in a new window]   FIG. 3.   hDM2 regulates the levels of E2F and DP subunits. (a and b) SAOS2 cells were transfected as indicated with expression vectors encoding E2F-1 (12 µg; lanes 4, 5, 6, and 7) or DP-1 (12 µg; lanes 6, 7, 8, and 9) together with hDM2 (12 µg) as indicated. After 6 h, the cells were washed and further grown for ~18 h in 0.2% fetal calf serum in DMEM. The cells were harvested and assayed for immunoblotting with anti-E2F-1 (a) or anti-HA for DP-1 (b) as described in the text. The band apparent in panel b, lane 4, results from a nonspecific reaction of the anti-HA antibody. (c) SAOS2 cells were transfected as indicated with expression vectors encoding E2F-5 (12 µg; lanes 3, 4, 5, and 6) and DP-3 (12 µg; lanes 5 and 6) together with hDM2 (12 µg; lanes 2, 4, and 6) treated as described above and were immunoblotted with anti-HA for E2F-5. (d) SAOS2 cells were transfected as indicated with expression vectors encoding p53 (12 µg; lanes 2 and 3) or hDM2 (12 µg; lane 3) treated as described above and were immunoblotted with anti-p53. For each of the experiments shown, the amount of extract loaded was determined by reference to the level of  galactosidase activity derived from pCMV gal (see Materials and Methods). +, present; , absent.

hDM2 modulates E2F activity in p53^/ MEFs. As SAOS2 cells are human tumor cells which have sustained mutations in genes other than p53, such as Rb, it was important to establish that similar effects were observed in nontumorigenic p53^/ cells. To this end, we assessed the effect of hDM2 in early-passage MEFs in which the p53 gene had been knocked out through homologous recombination (35).

/

/

View larger version (23K): [in this window] [in a new window]   FIG. 4.   hDM2 overcomes E2F-dependent apoptosis in p53/ MEFs. (a) p53/ MEFs were transfected with expression vectors encoding hDM2 (6 µg; bars 3, 4, 6, 8, and 10), E2F-1 (6 µg; bars 3, 4, 5, and 6), DP-1 (6 µg; bars 7 and 8), or p53 (6 µg; bars 9 and 10) as indicated. After 6 h, the cells were washed and grown overnight in 0.2% fetal calf serum in DMEM. Apoptosis was assayed ~18 h after transfection, and the apoptotic cells were counted. The percentage of apoptosis is expressed relative to the total number of cells counted, and the data shown (+ standard deviations) were derived from two separate experiments. (b) p53/ MEFs were transfected as indicated with expression vectors encoding hDM2 (12 µg; lanes 1, 3, and 5), E2F-1 (12 µg; lanes 2, 3, 4, and 5), or DP-1 (12 µg; lanes 4 and 5). The cells were treated as described in the legend to Fig. 3a and immunoblotted with anti-HA for E2F-1 and DP-1; E2F-1 and DP-1 are indicated. +, present; , absent.

/

/

hDM2 cooperates with E2F in DNA synthesis and colony-forming activity. We reasoned that in overcoming E2F-dependent apoptosis, it was likely that hDM2 promoted cell cycle progression and enhanced cell viability. Therefore, we monitored the level of DNA synthesis through BrdU incorporation in cells grown under low-serum conditions. When expressed alone, either E2F-1 or DP-1 stimulated BrdU incorporation (Fig. 5c, compare bars 2 and 6), and coexpressing E2F-1 with DP-1 resulted in a level of BrdU incorporation similar to that with E2F-1 or DP-1 alone (Fig. 5c, compare bars 2, 4, and 6). However, although hDM2 had a small effect on BrdU incorporation in the presence of E2F-1, a marked induction was apparent when hDM2 was coexpressed with E2F-1 and DP-1 (Fig. 5c, bar 5), suggesting that in the presence of both subunits, hDM2 can stimulate a greater level of DNA synthesis than in the presence of either subunit alone.

View larger version (15K): [in this window] [in a new window]   FIG. 5.   hDM2 regulation of E2F promotes DNA synthesis and confers a survival advantage. (a) SAOS2 cells were transfected as indicated with expression vectors encoding E2F-1 (6 µg; bars 2, 3, 4, and 5) or DP-1 (6 µg; bars 4, 5, and 6) together with hDM2 (6 µg; bars 1, 3, and 5). After 6 h, the cells were washed and grown for ~18 h in 0.2% fetal calf serum in DMEM. BrdU incorporation was assayed at ~18 h as described in the text, and immunostaining was performed in parallel to determine transfection efficiency (data not shown). The BrdU incorporation was determined relative to the transfection efficiency, and the data shown represent the average values (+ standard deviations) from three independent experiments where the BrdU incorporation is expressed relative to the control treatment of pcDNA3-transfected cells. (b) SAOS2 cells were transfected as indicated with expression vectors encoding E2F-1 (12 µg; bars 2 and 3), DP-1 (12 µg; bars 3 and 4), and hDM2 (12 µg; bars 1, 2, 3, and 4). After 6 h, the cells were washed and grown in either 0.2% () or 10% () fetal calf serum for ~18 h, after which the cells were placed under G418 selection under normal culture conditions; each bar represents the average data from a single experiment involving duplicate treatments. The average number of colonies over three distinct experiments obtained after transfection with pcDNA3 was 588. The colonies were stained with crystal violet, and colony-forming activity is presented as the increase in colony numbers in the presence of hDM2 relative to the numbers in the absence of hDM2 in cells treated with the indicated expression vectors. +, present; , absent.

Domains in hDM2 required for the regulation of E2F activity. We continued to analyze the mechanism through which hDM2 modulates the activity of the E2F heterodimer by examining the regions in hDM2 that are necessary for the observed effects on E2F-dependent apoptosis, colony formation, and altered protein level. We employed a number of mutant proteins derived from hDM2 (Fig. 6a) that have previously been examined for their regulatory effects on p53 activity (7). Thus, mutant hDM2^222-437 lacks most of the C-terminal half but retains the p53 binding domain, NLS, NES, and RING finger and has been reported to bind to p53, but it fails to affect the level of p53 protein (8, 28). hDM2^1-440 has a truncation in the C-terminal region and has wild-type activity for p53 regulation (8), and hDM2^59-89 possesses a deletion in the p53 interaction domain and fails to regulate p53 activity (7).

View larger version (28K): [in this window] [in a new window]   FIG. 6.   Domains in hDM2 required to regulate E2F. (a) Diagrammatic summary of wild-type hDM2 (indicating the relevant domains) and the hDM2 mutant derivatives. (b) SAOS2 cells were transfected under low-serum conditions as indicated with expression vectors encoding E2F-1 (6 µg) and DP-1 (6 µg) together with the hDM2 derivative hDM21-440, hDM2222-437, or hDM259-89 (6 µg; bars 2, 3, and 4). Apoptotic cells were assayed as described in the legend to Fig. 2. The values shown indicate the percent change in apoptotic activity relative to the control treatment (cells transfected with pcDNA3) and represent the average values (+ standard deviations) from three independent experiments. The control-treated cells usually showed less than 5% apoptosis. (c) SAOS2 cells were transfected as indicated with expression vectors encoding E2F-1 (12 µg) and DP-1 (12 µg) together with the indicated hDM2 derivatives (12 µg; bars 2, 3, and 4) and treated as described (grown in 0.2% fetal calf serum) in the legend to Fig. 5b. The percent change in colony numbers was calculated as the colony activity in the presence of hDM2 relative to that in the absence of hDM2, and the data represent the average values (+ standard deviations) derived from three independent experiments. (d and e) SAOS2 cells were transfected as indicated with expression vectors encoding E2F-1 (12 µg; lanes 1 to 4) and DP-1 (12 µg; lanes 3 and 4) together with wild-type hDM2 (d) or hDM2222-437 (e) (12 µg; lanes 2 and 4). The cells were harvested and assayed by immunoblotting with anti-E2F-1 or anti-HA for DP-1 as indicated. (f) SAOS2 cells were transfected as indicated with expression vectors encoding p53 (12 µg; lanes 1 and 2) together with hDM2222-437 (12 µg; lane 2). The cells were harvested and assayed by immunoblotting with anti-p53 or anti-MDM2 as indicated. The amount of extract loaded was determined by reference to the level of  galactosidase activity derived from pCMV gal. +, present; , absent.

222-437

59-89

222-437

59-89

222-437

222-437

222-437

222-437

222-437

222-437

222-437

59-89

59-89

hDM2 modulates the intracellular location of DP-1. The interaction between DP-1 and hDM2 (38) and the role of MDM2 in nucleocytoplasmic shuttling (46), combined with the data presented here, led us to consider the possibility that hDM2 can influence the intracellular location of DP-1. In the first instance, we performed the experiments with apoptotic cells coexpressing wild-type hDM2 and DP-1. As expected (9, 46), hDM2 was present predominantly in nuclei (Fig. 7a), whereas DP-1 was located mostly in the cytosol (Fig. 2 and 7a). In contrast, expressing hDM2 with DP-1 caused DP-1 to undergo a dramatic alteration in intracellular distribution: thus, although hDM2 possessed a nuclear location, a significant nuclear presence was observed for DP-1 (Fig. 7, compare b to d), suggesting that hDM2 can modulate the intracellular distribution of DP-1.

View larger version (39K): [in this window] [in a new window]   FIG. 7.   Influence of hDM2 on the intracellular distribution of DP-1. The intracellular distribution of the indicated proteins was determined under apoptotic conditions after SAOS2 cells were transfected with expression vectors encoding hDM2 (6 µg) (a), DP-1 (6 µg) (b), or both hDM2 and DP-1 (6 µg each) (c, d, e, f, g, and h). The cells were treated as described in the text with anti-DP-1 (b, d, f, g, and h) and anti-MDM2 (a, c, e, g, and i). Panels a and b show different cells, whereas panels c and d, e and f, g and h, and i and j show the same cells stained with anti-MDM2 and anti DP-1, respectively; note that the images result from different exposure times. The hDM2 mutant derivatives are indicated. Similar results were obtained with U20S cells (data not shown).

1-440

222-437

59-89

59-89

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

hDM2 overcomes E2F-dependent apoptosis. A significant conclusion from the study reported here relates to the ability of hDM2 to overcome E2F-dependent apoptosis and the important role played by the DP subunit of the E2F heterodimer in facilitating the antiapoptotic effect of hDM2. This effect is exerted independently of the established role of hDM2 in regulating the p53 response (44), as most importantly, the ability of hDM2 to control E2F-dependent apoptosis was apparent in p53^/ cells. Furthermore, the data suggest that once cells have been rescued from apoptosis by hDM2, hDM2 and the E2F-DP heterodimer cooperate to enhance cell viability and cell cycle progression; thus, the data establish the fact that an antiapoptotic and growth-promoting activity of hDM2 may be channelled through a single cellular target. In a physiological setting, the results imply that hDM2 can alter the cellular consequences of deregulating the E2F pathway from the induction of apoptosis to that of growth stimulation, a process that may be critically important during tumorigenesis.

/

/

/

Does hDM2 regulate an S-phase checkpoint? A variety of studies have ascribed both oncogenic and apoptotic activity to E2F (12). In this respect, given the important role that E2F plays in the control of proliferation, it makes considerable sense that cellular mechanisms should exist that monitor E2F and, in circumstances of aberrant control, stimulate apoptosis. Indeed, it is likely that E2F-dependent apoptosis represents a physiologically relevant process, as E2F-1^/ mice have defects in apoptosis during normal thymocyte ontogeny (15). Furthermore, the apoptosis seen in Rb^/ mice shows a clear tissue dependence on p53, as apoptosis is dependent upon p53 in the eye lens but is p53 independent in the peripheral nervous system (35, 40).

View larger version (25K): [in this window] [in a new window]   FIG. 8.   Model for regulation of E2F by hDM2. (a and b) It is envisaged that inappropriately high levels of E2F-DP activity cause the activation of a checkpoint pathway of control which thereafter leads to apoptosis (a). Since hDM2 can influence apoptosis by causing a reduction in E2F and DP subunit levels (b), it is suggested that this process may allow hDM2 to modulate checkpoint activity, limit apoptosis, and thereby facilitate cell cycle progression. The data suggest that the DP subunit is instrumental in enabling hDM2 to regulate E2F-dependent apoptosis. (c) Summary of the effects of hDM2 and mutant derivatives on E2F-DP regulation. (1), data not shown.

hDM2 influences E2F-DP levels and augments E2F-dependent growth. The ability of hDM2 to regulate the levels of E2F and DP correlated with the rescue of cells from apoptosis and induction of DNA synthesis. While it is necessary to clarify the mechanism involved in the down-regulation of E2F and DP subunits, it is noteworthy that the regulation of p53 by hDM2 involves a stimulation of ubiquitin-dependent degradation mediated by the E3 ligase-like activity in hDM2 (18, 28), and further, that the regulation of E2F and DP protein levels involves ubiquitin-dependent degradation (17, 19). It is consistent with this idea that the N-terminal region of hDM2, which is required for p53 regulation (39, 41), plays an important role in the regulation of E2F-DP activity, as the hDM2 derivative hDM2^59-89 has reduced capacity to control E2F-DP activity. It is possible that the p53 binding domain of hDM2 plays a wider role in mediating the effects of hDM2. However, despite these parallels between the regulation of p53 and that of E2F-DP, it is likely that each process is mechanistically distinct. This was most clearly suggested from studying hDM2^222-437, which cannot influence p53 levels (28) but possesses properties very similar to those of wild-type hDM2 in the control of the E2F-DP heterodimer.

/