Differential Regulation of E2F1, DP1, and the E2F1/DP1 Complex by ARF

Abstract

The tumor suppressor protein ARF inhibits MDM2 to activate and stabilize p53. Recent studies provided evidence for p53-independent tumor suppression functions of ARF. For example, it has been shown that ARF induces proteolysis of certain E2F species, including E2F1. In addition, ARF relocalizes E2F1 from the nucleoplasm to nucleolus and inhibits E2F1-activated transcription. Because DP1 is a functional partner of the E2F family of factors, we investigated whether DP1 is also regulated by ARF. Here we show that DP1 associates with ARF. Coexpression of ARF relocalizes DP1 from the cytoplasm to the nucleolus, suggesting that DP1 is also a target of the ARF regulatory pathways. Surprisingly, however, the E2F1/DP1 complex is refractory to ARF regulation. Coexpression of E2F1 and DP1 blocks ARF-induced relocalization of either subunit to the nucleolus. The E2F1/DP1 complex localizes in the nucleoplasm, whereas ARF is detected in the nucleolus, suggesting that ARF does not interact with the E2F1/DP1 complex. Moreover, we show that E2F1 is more stable in the presence of ARF when coexpressed with DP1. These results suggest that ARF differentially regulates the free and heterodimeric forms of E2F1 and DP1. DP1 is a constitutively expressed protein, whereas E2F1 is mainly expressed at the G₁/S boundary of the cell cycle. Therefore, the E2F1/DP1 complex is abundant only between late G₁ and early S phase. Our results on the differential regulation E2F1, DP1, and the E2F1/DP1 complex suggest the possibility that ARF regulates the function of these cell cycle factors by altering the dynamics of their heterodimerization during progression from G₁ to S phase.

The ARF/INK4A locus encodes two tumor suppressor proteins, p16^Ink4A and a 19-kDa ARF protein in mice or 14-kDa ARF protein in humans (32, 37, 44). p16^Ink4A is an inhibitor of the CDK4/6 kinases, and it causes accumulation of the underphosphorylated forms of the retinoblastoma protein. Therefore, the tumor suppressor function of p16^Ink4A is tightly linked to the retinoblastoma protein pathway of tumor suppression (reviewed in references 37 and 44).

Genetic studies in mice have suggested a potent tumor suppression function of the ARF protein. For example, mice lacking only the ARF function developed tumors at a frequency that is similar to loss of the ARF/p16^INK4A locus, suggesting that, at least in mice, the tumor suppression pathway of ARF is significant (14-16). The function of ARF has been biochemically linked to p53. It has been shown that ARF can stabilize and stimulate the activities of p53 (22, 23, 27, 31, 42), and it does so by regulating MDM2, which is an inhibitor of p53. MDM2 has been shown to possess a p53-specific E3 ubiquitin ligase activity (8, 11, 19, 30, 36). It binds to p53 and causes ubiquitination followed by proteolysis of p53. The ARF protein associates with MDM2 and inhibits its ability to ubiquitinate p53 (11).

ARF also possesses a p53-independent tumor suppression function. The first indication came from studies on transgenic mice expressing the Myc gene (5). Premalignant B cells expressing transgenic Myc and lacking both ARF and p53 proliferated at a much faster rate than those expressing Myc and lacking p53 or ARF alone. These observations suggested additional p53-independent functions of ARF. Further evidence for a p53-independent tumor suppression function of ARF came from studies that compared tumor frequencies in mice lacking ARF, MDM2, and p53 with those in mice lacking p53 and MDM2 or p53 alone (43). Mice nullizygous for ARF, p53, and MDM2 developed tumors at a frequency greater than that observed in mice lacking both p53 and MDM2 or p53 alone. Moreover, reintroduction of ARF in fibroblasts lacking ARF, p53, and MDM2 caused a G₁ arrest, suggesting that ARF can interact with targets other than p53 and MDM2 to inhibit cell proliferation. Studies on the simian virus 40 T antigen also provided evidence for the p53-independent tumor suppression function of ARF (4). Finally, a recent study, with mutants of ARF, indicated that the sequences involved in MDM2 regulation are not sufficient for the growth suppression function of ARF (17).

Recent studies also provided biochemical evidence for the p53-independent function of ARF. It has been demonstrated (25) that ARF could associate with certain members of the E2F family of transcription factors (E2Fs) and induce their degradation through the 26S proteasome pathway. The degradation is associated with a relocalization. For example, ARF expression caused relocalization of E2F1 from mainly nucleoplasmic to mainly nucleolar. The ARF effect is not true for all E2Fs; for example, E2F6 was not affected by ARF. Experiments with mouse embryo fibroblasts lacking both copies of p53 indicated that E2F1 could associate with ARF in a p53-independent manner. Moreover, expression of ARF in cells containing functionally defective p53 reduced the steady-state levels of E2F1. In addition, expression of a dominant negative p53 had no effect on the ARF-induced destabilization of E2F1 and E2F3 in U2OS cells. It was also shown that E2F1 could partially overcome ARF-induced growth arrest of p53-defective cells (25). Taken together, these observations suggest that ARF can regulate certain members of the E2Fs in a p53-independent manner.

These observations are significant with regard to the tumor suppressor functions of ARF because E2Fs stimulate expression of a number of genes that are critical for DNA replication and mitosis (13, 29). Another study (6) investigated the effect of human ARF on the transcriptional activity of E2F1. This group also demonstrated an interaction between ARF and E2F1. Moreover, they showed that ARF inhibited E2F1-activated transcription in a manner independent of p53. Furthermore, exon 1b of ARF was sufficient to bind E2F1 and inhibit E2F1-activated transcription. It is very possible that the inhibition of the E2F1-activated transcription is a result of E2F1 proteolysis induced by ARF. However, additional roles of ARF as a transcriptional repressor of the E2F1-regulated genes cannot be ruled out.

The E2F family of transcription factors bind DNA as heterodimers in conjunction with the DP family of factors (1, 7, 10). DP1 is the best-studied member of the DP family of proteins. E2Fs and DP1 contain hydrophobic heptad repeats, which are believed to be involved in heterodimer formation through coil-coil interactions. DP1 also possesses a DNA-binding domain that contributes to the sequence-specific DNA binding activity of the E2Fs. DP1 by itself has very little transcriptional activity; however, it cooperates with the E2Fs to activate transcription of the E2F target genes (1, 10). DP1 is also a critical component of the E2F-retinoblastoma protein repressor complex. It has been shown that heterodimerization with DP1 is important for a stable interaction between E2F1 and retinoblastoma protein (10).

The activity of DP1 is believed to be regulated through phosphorylation by cyclin A-cdk2. It has been shown that cyclin A-cdk2 binds to an N-terminal region in E2F1 and phosphorylates DP1, which leads to a loss of DNA binding by the E2F1/DP1 complex (18). Thus, DP1 is both a functional and a regulatory partner of E2F1. In this study, we show that DP1 is also a target of ARF. Moreover, the E2F1/DP1 complex is refractory to the regulatory effects of ARF.

MATERIALS AND METHODS

Cell cultures.

U2OS and SAOS2 cells were obtained from the American Type Culture Collection. The cells were maintained in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum.

Expression plasmids.

An expression plasmid expressing hemagglutinin (HA)-tagged p19^ARF driven by a cytomegalovirus promoter was constructed by subcloning a fragment containing the p19^ARF cDNA along with HA sequences into the BamHI and XhoI sites of pCDNA3. The expression plasmids expressing T7 epitope-tagged proteins were constructed with PCR in which the upstream primers contained a sequence encoding the T7 epitope in frame with the first ATG of the respective cDNA. Appropriate restriction enzyme sites were also engineered into the primers to facilitate cloning,

For T7-E2F1, the upstream and downstream primers used were ACGGTACCCACCATGGCTAGCATGACCGGCGGACAGCAGATGGGCATGGCCTTGGCCGGG and ACTCTAGATCAGAAATCCAGGGG, respectively, and the amplified fragment was cloned as a KpnI (5′)-XbaI (3′) fragment into pCDNA3. For T7-DP1 the upstream and downstream primers used were CGGGATCCCGCCATGGCTAGCATGACCGGCGGACAGCAGATGGGCATGGCAAAAGATGCCGGTCT and GCTCTAGAGCTCAGTCGTCCTCGTCATTCT, respectively, and the amplified fragment was cloned as a BamHI (5′)-XbaI (3′) fragment into pCDNA3. For T7-p19^ARF, the upstream and downstream primers were GCGGTACCCACCATGGCTAGCATGACCGGCGGACAGCAGATGGGCATGGGTCGCAGGTTC and AGTCTCGAGCTATGCCCGTCGGTC, respectively, and the amplified fragment was cloned as a KpnI (5′)-XhoI (3′) fragment into pCDNA3.

The DP1 d205-277 mutant was created by PCR in two steps. The region between amino acids 1 and 204 was amplified with the upstream primer DP1mut1 (ACGGATCCATGGCAAAAGAT) and downstream primer DP1mut2 (AAGGTACCGTTCTGACATTC). Similarly, the region between amino acids 278 and 410 was amplified with the upstream primer DP1mut3 (AAGGTACCAATGACAAATTT) and downstream primer DP1mut4 (GCTCTAGAGCTCAGTCGTCCTCGTCATTCT). An XhoI site was engineered into the primers designated DP1mut 2 and DP1mut3. Following PCR amplification and digestion with XhoI, the fragments were ligated, and a heat-inactivated aliquot of the ligation mix was used to do PCR with the DP1 upstream and downstream primers (DP1mut1 and DP1mut4) as described above. The amplified product was cloned as a BamHI (5′)-XbaI (3′) fragment into pCDNA3. The T7-tagged version of the DP1 mutant was made in a similar fashion except that the sequence for the T7 tag was engineered into the primer designated DP1mut1.

DNA transfections.

All DNA transfections, unless and until mentioned, were done by the calcium phosphate precipitate method as described previously (9).

CAT assays.

Chloramphenicol acetyltransferase (CAT) assays were performed by the xylene extraction method of Seed and Sheen as described earlier (35).

Immunoprecipitation and Western blotting.

The transfected cells were harvested 48 h after transfection. The cells were washed twice with phosphate-buffered saline (PBS) and suspended in NETT250 buffer [20 mM Tris-HCl (pH 8.0), 0.1 mM EDTA, 250 mM NaCl, 0.5% Triton X-100] for 1 h at 4°C. After incubation, the lysates were centrifuged at 13,000 × g for 10 min, and the supernatants were used for immunoprecipitation. p19^ARF antibody (R562; GeneTex, San Antonio, Tex.) was used for immunoprecipitation. Cell lysates (1.2 mg) were incubated with the antibody for 2 h at 4°C. Protein A-Sepharose was then added, and the tubes were rocked for 1 h at 4°C. The beads were then collected by centrifugation. Precipitates were washed three times with 400 μl of the NETT250 buffer. The bound proteins were subjected to Western blot analysis.

Western blot analysis was performed with anti-rabbit or anti-mouse Fab fragments conjugated to horseradish peroxidase (Amersham) and the ECL Western blot detection reagent (Amersham) according to the manufacturer's instructions. The T7 epitope tag-horseradish peroxidase conjugated antibody was obtained from Novagen. The E2F1 antibody (KH95) was from Santa Cruz Biotechnology. The polyclonal antibody against p19^ARF (R562) was obtained from GeneTex. The DP1 antibody (WTH16) was from NeoMarkers.

Decay rate analysis.

U2OS cells or SAOS2 cells were transfected with the indicated expression plasmids by the calcium phosphate method. The transfections were done in duplicate for each combination of plasmids indicated. Sixteen hours after transfection, the cells were trypsinized, pooled, and replated. Twenty-four hours after transfection, cycloheximide (ICN Pharmaceuticals) was added to the cells at a final concentration of 20 μg/ml, the cells were harvested at the indicated times, and extracts were prepared with lysis buffer [20 mM HEPES (pH 7.9), 400 mM NaCl, 1 mM EDTA, 0.1% NP-40, 1 mM phenylmethylsulfonyl fluoride, and 10% glycerol]. The extracts were run on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gels, and Western blot analysis was performed as described above.

Immunostaining and confocal microscopy.

U20S cells were grown in 24-well plates containing coverslips. For DP1 localization, cells were transfected with either T7-DP1 (0.2 μg) alone or a combination of plasmids expressing T7-DP1 or T7-DP1 d205-277 and p19^ARF (0.6 μg). Cells were also transected with plasmids expressing T7-Cul4B (0.2 μg) alone or in combination with p19^ARF to serve as a control.

For E2F1 localization, cells were either transfected with T7-E2F1 (0.2 μg) alone or cotransfected with T7-E2F1 and p19^ARF (0.6 μg) or with T7-E2F1, p19^ARF, and either DP1 or DP1 d205-277 (0.2 μg each) expression plasmids. Transfections were done with Lipofectamine 2000 reagent (Invitrogen). Twenty-four hours after transfection, the cells were fixed with methanol, blocked with 5% goat serum in PBS, and probed with T7 tag monoclonal antibody (1:250 dilution) and polyclonal p19^ARF antibody (1:250 dilution). The proteins were detected with tetramethyl rhodamine isocyanate (TRITC)-conjugated anti-mouse and fluorescein isothiocyanate (FITC)-conjugated anti-rabbit immunoglobulin secondary antibodies (both at 1:200 dilution). Finally, the coverslips were washed and mounted on glass slides with Vectashield mounting medium (Vector Laboratories). The immunofluorescence was detected and images were taken with a CLSM 510 microscope (Zeiss) and a 63× Acrophlan water immersion objective.

RESULTS

ARF associates with DP1 and relocalizes DP1 to the nucleolus.

Both the human and mouse ARF gene products have been shown to associate with the E2F family of factors (6, 25). Because DP1 is an essential DNA-binding partner of the E2F family of factors, we decided to investigate whether DP1 could associate with ARF. U2OS cells were transfected with a plasmid expressing T7 epitope-tagged DP1 or a mutant DP1 which lacks the sequences required for heterodimerization with E2F1 along with a mouse ARF expression plasmid. Transfections were carried out in the absence of any E2F expression plasmid. An unrelated protein with the T7 tag (Cul-4B) was used as a control. Extracts from the transfected cells were subjected to immunoprecipitation with ARF antibody, and the immunoprecipitates were assayed with T7 antibody for the presence of DP1 and Cul-4B.

We could easily detect coimmunoprecipitation of the wild-type DP1 protein with ARF (Fig. 1). In the same experiment, Cul-4B or the mutant DP1 was not coimmunoprecipitated with ARF. These results provide evidence for an association of DP1 with the ARF tumor suppressor protein. Moreover, the interaction is dependent upon the sequences in DP1 required for binding to E2F1. It is possible that the E2F1-binding motif in DP1 is also required for binding to ARF.

FIG. 1.

p19^ARF associates with DP1. U2OS cells were transfected with plasmids that express T7 epitope-tagged Cul4B (10 μg), DP1 (10 μg), or DP1 d205-277 (10 μg) along with a p19^ARF (10 μg) expression plasmid. Forty-eight hours after transfection, the cells were harvested, and total cell extract was prepared as described in Materials and Methods. (Right) One milligram of the total cell extract was subjected to immunoprecipitation (IP) with R562 antibody (for p19^ARF), and the immunoprecipitates were subjected to Western blot (WB) analysis. The blot was probed with horseradish peroxidase-linked T7 antibody (Novagen) to detect coimmunoprecipitating Cul4B, DP1, and DP1 d205-277. (Left) Extracts were also tested for the expression of T7-Cul4B, T7-DP1, T7-DP1 d205-277, and p19^ARF by probing the blot with T7 and p19^ARF antibodies.

It was shown that ARF is able to relocalize its associated proteins to the nucleolus. For example, coexpression of ARF causes a relocalization of E2F1 and MDM2 to the nucleolus (3, 25, 33, 39, 43, 44). Therefore, we investigated whether the subcellular localization of DP1 changes when coexpressed with ARF.

For this purpose, U2OS cells were transfected with a plasmid expressing T7 epitope-tagged DP1 or the mutant DP1 in the presence and absence of an ARF expression plasmid. T7 epitope-tagged Cul-4B was also used as a control. The transfections were carried out with cells grown on coverslips. Following transfection, the coverslips were subjected to immunostaining with T7 antibody for DP1 and ARF antibody to detect the localization of ARF. Fluorescently conjugated secondary antibodies (TRITC for DP1 and FITC for ARF) were used for visualization. The subcellular localizations of DP1 and ARF were visualized with a confocal microscope.

Consistent with previous observations (24), we found that DP1 is mainly localized to the cytoplasm when expressed alone (Fig. 2A). DP1 lacks a functional nuclear localization signal, and its nuclear localization is dependent upon the DNA-binding partners, the E2F family of proteins (24). It was shown that binding to E2F1 is critical for nuclear localization of DP1 (24). Interestingly, coexpression of ARF resulted in a relocalization of DP1 from mainly cytosolic to mainly nucleolar (Fig. 2A and Table 1). The colocalization of DP1 with ARF is also consistent with their interaction. The mutant DP1 and Cul-4B failed to colocalize with ARF in the nucleolus. The nucleolar localization of DP1 in the presence of ARF was confirmed by staining for nucleolin (Fig. 2B). Taken together, these results suggest that ARF associates with DP1 and relocalizes it to the nucleolus.

FIG.2.

p19^ARF relocalizes DP1 to the nucleolus. (A) U2OS cells grown on coverslips and transfected with plasmids expressing either T7-DP1 (0.2 μg) or T7-Cul4B (0.2 μg) alone or a combination of plasmids expressing T7-DP1 and p19^ARF (0.5 μg), T7-DP1 d205-277 (0.2 μg), and p19^ARF or T7-Cul4B and p19^ARF. At 24 h after transfection, cells were fixed and probed with T7 epitope tag monoclonal antibody (1:500 dilution) and R562 antibody (p19^ARF; 1:200 dilution). TRITC-labeled anti-mouse and FITC-labeled anti-rabbit immunoglobulin antibodies (both 1:200 dilution) were used as the secondary antibodies. The immunofluorescence was detected by confocal microscopy as described in Materials and Methods. (B) U2OS cells grown on coverslips were cotransfected with plasmids expressing T7-DP1 and p19^ARF. At 24 h after transfection, the cells were fixed and probed with T7 epitope tag monoclonal antibody (1:500 dilution) and goat polyclonal C23 antibody (nucleolin; 1:200 dilution). TRITC-labeled anti-mouse and FITC-labeled anti-goat immunoglobulin antibodies (both 1:200 dilution) were used as secondary antibodies. The immunofluorescence was detected by confocal microscopy as described in Materials and Methods.

TABLE 1.

Subcellular localization of E2F1, DP1, and the E2F1/DP1 complex in the presence of ARF

Transfected gene(s)	% of cells with E2F1 or DP1 in the:
	Nucleoplasm or cytoplasm	Nucleolus
T7-E2F1	100 (nucleoplasm)	0
T7-DP1	100 (cytoplasm)	0
E2F1 + T7-DP1	100 (nucleoplasm)	0
T7-E2F1 + ARF	8 (nucleoplasm)	92
T7-DP1 + ARF	12 (cytoplasm)	88
T7-E2F1 + DP1 + ARF	76 (nucleoplasm)	24
T7-DP1 + E2F1 + ARF	80 (nucleoplasm)	20

Coexpression of E2F1 and DP1 inhibits ARF-induced relocalization to the nucleolus.

ARF relocalizes E2F1 to the nucleolus (25). Therefore, we sought to investigate whether ARF would relocalize a complex of E2F1 and DP1 by coexpressing E2F1 and DP1 in the presence of ARF. For this purpose, U2OS cells were transfected with plasmids expressing T7 epitope-tagged E2F1 and untagged DP1 or T7-tagged DP1 and untagged E2F1 in the presence and absence of an ARF expression plasmid. The transfections were carried out with cells grown on coverslips. Following transfection, the coverslips were subjected to immunostaining with T7 antibody for E2F1 or DP1 and ARF antibody to detect the localization of ARF.

As expected, E2F1 was mainly found in the nucleoplasm when expressed in the absence of ARF (Fig. 3 and Table 1). In the presence of ARF expression, there was a dramatic change; the majority of E2F1 colocalized with the ARF protein to nuclear structures (Fig. 3) previously shown to be nucleoli (25). Interestingly, expression of DP1 reversed or blocked the effect of ARF. In the presence of DP1 expression, E2F1 was localized mainly in the nucleoplasm even in the presence of the ARF protein. The mutant DP1 failed to retain E2F1 in the nucleoplasm.

FIG. 3.

Coexpression of E2F1 and DP1 blocks p19^ARF-mediated nucleolar relocalization. U2OS cells grown on coverslips were transfected with plasmid expressing T7 epitope-tagged E2F1 alone (0.2 μg) or a combination of plasmids expressing T7-E2F1 and p19^ARF (0.5 μg), T7-E2F1, p19^ARF, and DP1 (0.3 μg), T7-E2F1, p19^ARF, and DP1 (d205-277) (0.3 μg), or T7-DP1 (0.2 μg), E2F1 (0.3 μg), and p19^ARF. At 24 h after transfection, cells were fixed and probed with T7 epitope tag monoclonal antibody (1:500 dilution) and R562 antibody (p19^ARF; 1:200 dilution). TRITC-labeled anti-goat and FITC-labeled anti-rabbit immunoglobulin antibodies (both 1:200 dilution) were used as the secondary antibodies. The immunofluorescence was detected by confocal microscopy as described in Materials and Methods. The scale bars in the T7-E2F1 alone panel correspond to 5 μm; in all other panels they correspond to 10 μm.

To see the localization of DP1 under similar conditions, cells were transfected with T7-tagged DP1 and untagged E2F1 in the presence and absence of ARF. As expected from previous studies (24), in the presence of E2F1 expression, DP1 localizes in the nucleus. Moreover, coexpression of E2F1 blocked the ARF-mediated localization of DP1 to the nucleolus. Results of the subcellular localization studies are also summarized in Table 1. The localization of T7-E2F1 and T7-DP1 in 50 to 60 transfected cells positive for all transgenes was examined, and the percentages of cells exhibiting nucleolar, nucleoplasmic, and cytoplasmic staining for E2F1 or DP1 are shown in Table 1. These results clearly demonstrate that, when expressed together, the E2F1 and DP1 proteins fail to colocalize with ARF in the nucleolus and remain in the nucleoplasm. The differential localization of ARF and the E2F1/DP1 complex would be consistent with the notion that the E2F1/DP1 complex does not associate with ARF inside the cell.

Coexpression of DP1 blocks ARF-induced proteolysis of E2F1.

Consistent with the previous result (25), we observed that ARF expression induces proteolysis of E2F1. These experiments were performed by transfecting U2OS cells with plasmids expressing E2F1 or DP1 in the presence or absence of the ARF expression plasmid. For each set of experiments, two plates of cells were transfected with the expression plasmids. To equalize for the transfection efficiencies, 16 h after transfection (after addition of the DNA precipitates), cells from the same set were trypsinized and pooled, and approximately equal numbers of cells were divided among five plates. Twenty-four hours after replating, cells were treated with cycloheximide at a final concentration of 20 μg/ml. At the indicated times, cells were harvested, and the extracts were subjected to Western blot assays and probed with a monoclonal antibody against E2F1 or the T7 epitope for DP1.

As expected, E2F1 decayed at faster rates in the presence of ARF than in its absence (Fig. 4). The half-life of E2F1 was reduced from greater than 4 h to less than 2 h. We observed that the half-life of DP1 was also reduced in the presence of ARF. However, the effect of ARF on the stability of DP1 was moderate compared to that of E2F1 (Fig. 4).

FIG. 4.

p19^ARF reduces the half-life of E2F1 and DP1. Two plates (10-cm dish) of U2OS cells were transfected with plasmids that express E2F1 (2 μg) or T7 epitope-tagged DP1 (5 μg) in the presence or absence of a plasmid expressing p19^ARF (10 μg). Sixteen hours after transfection, the cells were trypsinized, and the cells from the same set were pooled and replated into five plates. After 24 h, cycloheximide was added to the medium at a final concentration of 20 μg/ml, and the cells were harvested at the indicated time points. Total cell extract was prepared as described in Materials and Methods. One hundred and fifty micrograms of the transfected cell extracts was analyzed by Western blot assay. The blot was probed with either KH95 antibody (for E2F1) or T7 epitope tag monoclonal antibody (for T7-DP1).

Because ARF exhibited a dramatic effect on the decay rate of E2F1, we investigated the effect of DP1 on the ARF-induced decay of E2F1. Since DP1 blocked the ARF-mediated nucleolar localization, we predicted that DP1 would block the proteolysis of E2F1 induced by ARF. Our preliminary experiments on the steady-state levels of E2F1 indicated that expression of ARF reduced the level of E2F1, and the effect was reversed by the coexpression of DP1 (not shown).

To further investigate the effect of DP1, we looked at the decay rate of E2F1 in the presence of DP1 expression. U2OS (Fig. 5, upper panel) or SAOS2 (Fig. 5, lower panel) cells were transfected with plasmids expressing E2F1 and ARF in the presence of a DP1 expression plasmid. As in the previous experiment, 16 h following transfection, cells from the same set were pooled and equally divided among five plates. Twenty-four hours after replating, cells were treated with cycloheximide and harvested at the indicated time points following cycloheximide treatment. Analysis of the levels of E2F1 clearly indicated a stabilizing effect of DP1. Expression of DP1 blocked the ARF-induced decay of E2F1 (Fig. 5). The mutant DP1, which does not form heterodimers with E2F1 and is unable to retain E2F1 in the nucleoplasm, was also unable to stabilize E2F1. The results reinforce the notion that the E2F1/DP1 complex is refractory to the regulatory effects of ARF.

FIG. 5.

DP1 blocks p19^ARF-induced proteolysis of E2F1. Two plates (10-cm dish) of U2OS cells or SAOS2 cells were transfected with a plasmid expressing E2F1 (2 μg) in the presence or absence of a plasmid expressing p19^ARF (10 μg). Wherever indicated, the cells were also transfected with plasmids expressing either DP1 or DP1 d205-277 (5 μg). Sixteen hours after transfection, the cells were trypsinized, and the cells from the same set were pooled and replated into five plates. After 24 h of transfection, cycloheximide was added to the medium at a final concentration of 20 μg/ml, and the cells were harvested at the indicated time points. Total cell extracts were prepared as described in Materials and Methods. One hundred and fifty micrograms of the transfected cell extracts was analyzed by Western blot assay. The blot was probed with monoclonal antibody KH95 (for E2F1).

Coexpression of DP1 reverses ARF inhibition of E2F1-activated transcription.

The human ARF protein was shown to be an inhibitor of E2F1-activated transcription (6). Consistent with that, we observed that the mouse ARF protein is also a potent inhibitor of E2F1-activated transcription (Fig. 6). Moreover, viral oncogenes (E1A, E7, and T antigen) which are known to stimulate E2F1-activated transcription failed to reverse the inhibitory effect of p19^ARF (not shown). Because DP1 expression blocked the ARF-induced nucleolar localization and proteolysis of E2F1, we predicted that DP1 would reverse the ARF inhibition of the E2F1-activated transcription.

FIG. 6.

DP1 overcomes p19^ARF inhibition of E2F1-activated transcription. (A) SAOS2 cells were transfected with the indicated CAT reporter gene. Where indicated, the transfection mix also contained an expression plasmid for E2F1 or the following combination of expression plasmids: E2F1 and p19^ARF; E2F1, p19^ARF, and DP1; or E2F1, p19^ARF, and DP1 d205-277. The transfections were carried out by the calcium phosphate precipitation method as described in Materials and Methods. A plasmid expressing β-galactosidase was included to control for transfection efficiencies. Average activation of the CAT gene activity in three independent experiments is shown. (B) U2OS cells were transfected with the indicated CAT reporter gene. The cells were also transfected with the indicated combination of plasmids expressing E2F1, DP1, and p19^ARF. The transfections were carried out by the calcium phosphate precipitation method as described in Materials and Methods. A plasmid expressing β-galactosidase was included to control for transfection efficiencies. Average activation of CAT gene activity in three independent experiments is shown.

Transcription experiments were carried out in the presence of DP1 expression with both SAOS2 cells (Fig. 6A) and U2OS cells (Fig. 6B). A CAT reporter gene with an E2F1-reponsive promoter (E2-CAT) was used in this study. E2F1 stimulated transcription from this promoter. Expression of mouse ARF inhibited the E2F1-activated transcription. More interestingly, coexpression of DP1 could overcome the ARF inhibition in a dose-dependent manner (Fig. 6). Coexpression of DP1 increases the level of the E2F1/DP1 complex, which is resistant to the inhibitory effects of ARF. A mutant DP1 that does not form heterodimers with E2F1 also failed to overcome ARF inhibition of the E1F1-activated transcription (Fig. 6A). Moreover, in the presence of DP1 expression, a much higher amount of ARF was needed to inhibit transcription. These results are also consistent with the notion that the E2F1/DP1 heterodimer is resistant to the regulatory functions of ARF.

DISCUSSION

The E2F family of transcription factors are centrally important in mammalian cell division. E2Fs are also targets of a variety of oncogenes and tumor suppressor genes (40). Therefore, it is likely that the E2F-inhibitory activity of ARF is linked to its tumor suppression function. Analysis of ARF mutants indicates a link between the E2F-inhibitory activity and the G₁ arrest function of ARF (A. Datta, C. Korgaonkar, D. Quelle, and P. Raychaudhuri, unpublished observations). Therefore, an understanding of the mechanism by which ARF inhibits E2F will be important.

Towards this end, we show that the DP1 subunit of the E2F family of transcription factors also associates with ARF. Moreover, binding of ARF leads to a relocalization of DP1 from the cytoplasm to the nucleolus. In addition, there was a modest increase in the decay rate of DP1 in the presence of ARF. These results suggest that DP1, which is an essential DNA-binding partner of the E2Fs, is also a target of the ARF regulatory pathways. Interestingly, we observed that ARF regulates E2F1 and DP1 effectively when these proteins are expressed individually. Coexpression of E2F1 and DP1 blocks the regulatory effects of ARF. When coexpressed, E2F1 and DP1 exist mainly in the form of E2F1/DP1 complex. Our results suggest that the E2F1/DP1 complex, unlike E2F1 and DP1, is refractory to the inhibitory effects of ARF.

We observed that a mutant DP1 which lacks the region important for heterodimerization with E2F1 also failed to associate with ARF. While it is possible that DP1 associates with ARF through E2F1, we favor the model in which DP1 associates with ARF independently of the E2Fs through its heterodimerization domain, because ARF could quantitatively relocalize DP1 from cytoplasm to the nucleolus in the absence of exogenous E2F (Fig. 2). Independent binding through the heterodimerization domain would raise the possibility that E2F1 and ARF compete for binding to DP1. We observed evidence for such a competition. When all three proteins were coexpressed, E2F1 outcompeted ARF for binding to DP1. ARF was found mainly in the nucleolus, while the E2F1 and DP1 proteins were in the nucleoplasm (Fig. 3). The lack of colocalization of the E2F1/DP1 complex with ARF would be consistent with the notion that ARF does not interact with the heterodimeric complex of E2F1 and DP1. This would also explain why we failed to detect a DNA-bound complex of E2F (which involves both E2F1 and DP1) that contains ARF (data not shown).

It is likely that DP1 has a much higher affinity for E2F1 than for ARF because we could almost quantitatively block the ARF-induced nucleolar localization of DP1 by E2F1 with a low level of the E2F1 expression plasmid (Fig. 3 and Table 1). E2F1, DP1, and ARF were expressed with a common expression vector (pcDNA3; Invitrogen). Therefore, it is also noteworthy that a relatively small amount of DP1 expression plasmid (50 ng) was able reverse the inhibition of the E2F1-activated transcription by 2 μg of ARF expression plasmid (Fig. 6A). Thus, under our experimental conditions, the E2F1/DP1 heterodimer is a stronger complex than the ARF/E2F1 or the ARF/DP1 complex. This is also consistent with the observation that a much higher level of ARF was needed to inhibit transcription when both E2F1 and DP1 were expressed (Fig. 6).

How does ARF inhibit E2F-activated transcription? Although E2F1 and DP1 are functional partners, expression of these two proteins is regulated differentially. For example, DP1 is expressed constitutively, whereas several members of the E2F family of proteins (E2F1, E2F2, and E2F3) are expressed mainly at the G₁/S boundary of the cell cycle (see reference 29 and references therein). These E2F family members, after their expression, associate with the available pool of DP1 to participate in the transcription of a variety of DNA replication genes and genes involved in mitosis (13). Thus, in early G₁ phase, there is a pool of unbound DP1, which might be a target for ARF regulation (Fig. 7). ARF would sequester away all the available unbound DP1, which in turn would favor a dissociation of the E2F1/DP1 complex present at low levels in G₁ phase. The dissociated E2F1 and DP1 molecules will become further targets of ARF. This would be consistent with the G₁ arrest function of ARF.

FIG. 7.

Schematic diagram showing how ARF regulates the free DP1 and E2F1 subunits. DP1 is expressed constitutively, whereas E2F1 is expressed between the mid- and late G₁ phase of the cell cycle. The E2F1/DP1 complex is abundant only at the G₁/S boundary. In our model, formation of the E2F1/DP1 complex at the G₁/S boundary is the target of ARF inhibition. ARF depletes DP1 in early G₁ phase, and the newly synthesized E2F1, later in G₁ phase, becomes a further target of ARF inhibition.

It is also possible that ARF targets these proteins when there is a disproportionate increase in the level of one of the subunits. Oncogenic stimulation increases levels of the E2Fs. It is possible that ARF, which is an E2F1-induced gene (2), acts to trim down the levels of the E2Fs. This would be consistent with the notion that ARF functions as a failsafe mechanism that protects cells from tumorigenic consequences of oncogene activation (37)

The E2F1/DP1 complex is abundant only at the boundary of G₁/S phases (29), at a time when the cell has committed to progression through S phase. The E2F1/DP1 complex stimulates expression of genes necessary for progression through the S and G₂/M phases (13). Since ARF is one of the E2F-induced genes, one would expect an increased expression of ARF at the G₁/S boundary. Therefore, the resistance of the E2F1/DP1 complex to ARF regulation would ensure progression through S phase. It is possible that ARF curtails any excessive activity of E2F1, which would otherwise lead to apoptosis. E2F1 was shown to induce cellular senescence in human primary fibroblasts through increased expression of ARF (4a). This study ectopically expressed E2F1 alone and relied upon endogenous DP1, a condition under which ARF is dominant over E2F1. It would be interesting to determine whether coexpression of E2F1/DP1 induces senescence or apoptosis.

It is unclear whether the ARF-induced nucleolar localization is essential for the proteolysis of E2F1 and DP1. Nucleolar localization of the ARF-MDM2 complex has been studied in detail (41). It has been shown that the ARF-mediated relocalization is dependent upon a cryptic nucleolar localization signal, the R/KR/KXR/K motif, in MDM2. It was suggested that binding of ARF to MDM2 causes a conformational change in MDM2, unmasking the cryptic nucleolar signal in MDM2 (41). However, it is unclear how nucleolar localization signals direct localization of a protein or a complex of proteins.

Interestingly, both E2F1 and DP1 contain a potential nucleolar signal, between residues 181 and 185 for E2F1 and between residues 105 and 108 for DP1. These signals map within the DNA-binding domain and very close to the heterodimerization domain of both proteins. We speculate that, in the absence of heterodimerization, as in the case of MDM2, the cryptic nucleolar signal in DP1 and E2F1 becomes active after binding to ARF and plays a role in the nucleolar localization of the ARF-DP1 or ARF-E2F1 complex. Following heterodimerization of E2F1 and DP1, however, the cryptic nucleolar signals in E2F1 and DP1 may not be available for relocalization to the nucleolus. Clearly, further studies will be necessary to test this model.