Ras induces elevation of E2F-1 mRNA levels.

Both E2F-1 and Ras play pivotal roles in regulation of cell proliferation and, in some biological settings, they collaborate in cell transformation. We show here that activated Ras induces an increase in E2F-1 mRNA and protein levels. This Ras-induced increase in E2F-1 levels is dependent on both MEK and PKB and it is RB-independent. The effect of Ras on up-regulation of E2F-1 mRNA is at the level of mRNA stability. Our data describe a novel functional link between Ras and the RB/E2F pathway. Furthermore, we suggest that one of the molecular mechanisms underlying the collaboration between Ras and E2F-1 involves a Ras-induced elevation of transcriptionaly active E2F-1 levels . cells they were selected with 2 µ g/ml puromycin for 24hr. Then Actinomycin D (10 µ g/ml, Sigma A-9415) was added for the indicated times. For cyclohexamide treatment: 24hr after transfection of 293 cells, 10 µ g/ml of cyclohexamide (Sigma C0934) were added to the plates for the indicated times and then cells were harvested. For MG-132 treatment: 24hr after transfection of 293 cells, 50 µ M of MG-132 (Calbiochem 474790) were added to the plates for 2hr. and then cells were harvested.


Introduction
The E2F transcription factors control cell cycle-dependent expression of genes that are essential for cell proliferation (for review see (1,2)). The DNA binding complex named E2F is a heterodimeric complex consisting of an E2F component and a dimerization partner, DP.
To date, six E2F genes and two DP genes have been cloned (2). E2F-1, -2 and -3 represent a subgroup of the E2F family, and they are specifically regulated by RB and not by the RB related proteins, p107 and p130.
The Ras gene family encodes small GTP-binding proteins that play a critical role in cell growth control as pivotal mediators of mitogenic signals from tyrosine kinase receptors (Reviewed by (8)). Mutations in Ras genes which result in constitutively activated Ras proteins are frequent in human tumors (reviewed by (9)).
Co-expression of activated Ras together with E2F-1 and its heterodimeric partner DP-1 leads to the formation of morphologically transformed foci in primary rat embryo fibroblasts and these cells induce tumor formation in nude mice (10). Furthermore, double transgenic animals over-expressing E2F-1 and activated Ras in their epidermis develop skin tumors (11).
The molecular mechanisms underlying the co-operation between Ras and E2F-1 in cell transformation are currently not fully understood. When expressed alone, either deregulated E2F-1 or constitutively active Ras transform immortal rodent cells (12,13), but not primary cells. In fact, expression of either E2F-1 or activated Ras in primary cells leads to cell cycle arrest resembling premature senescence (14)(15)(16). In both cases induction of the senescencelike phenotype involves up-regulation of the expression of p19 ARF which neutralizes MDM2 and thereby stabilizes p53 (6,17).
E2F activity is tightly regulated by a number of mechanisms during cell cycle progression.
E2F/DP heterodimer formation facilitates binding to and negative regulation by the product of the retinoblastoma gene, RB and its related proteins p107 and p130, collectively referred to as the pocket proteins. Indeed, complexes of unphosphorylated RB and E2F/DP act as transcriptional repressors, which contribute to RB-dependent cell cycle arrest in G1. Complex formation is cell cycle regulated via phosphorylation of the pocket proteins by Cdk4/cyclin D by guest on March 20, 2020 http://www.jbc.org/ Downloaded from and Cdk2/cyclin E heterodimers. These phosphorylations lead to dissociation of E2F/pocket protein complexes, resulting in free, transcriptionaly active E2F/DP heterodimers. The combination of cessation of repression of some E2F-regulated genes and activation of others by the now activated transcription factor(s) is a major step in promoting G1 exit. Additional controls of E2F-1 activity include up-regulation of its DNA binding activity by acetylation (18) and down-regulation of this activity via phosphorylation of DP-1 by Cdk2/cyclin A (19)(20)(21). E2F-1 mRNA and protein levels are also tightly regulated. It is subjected to cell cycle dependent transcriptional control and its mRNA level peak in late G1. In addition, E2F-1 is a short-lived protein and is degraded by the proteasome pathway (22)(23)(24).
In view of the ability of both Ras and E2F-1 to play a role in cellular trasnformation on the one hand and to induce premature cell senescence on the other hand, we studied possible functional relationships between Ras and E2F-1. We report here that activated Ras upregulates both E2F-1 mRNA and protein levels. This constitutes a novel mechanism of regulating E2F-1 levels.

Experimental Procedures
Cell Culture. Rat-1a-MT-wtE2F-1 cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum (FCS) and G418 (500µg/ml, GibcoBRL). 293 and 293T cells were grown in DMEM supplemented with 10% fetal calf serum (FCS). H1299 cells were grown in RPMI supplemented with 10% FCS. Swiss 3T3, NIH 3T3 and 3T3 fibroblasts derived from RB -/mice were grown in DMEM supplemented with 10% bovine calf serum (BCS). All cells were maintained at 37°C in a humidified 8% CO 2 -containing atmosphere. For Actinomycin D treatment: After infection of Rat-1a-MT-wtE2F-1 cells they were selected with 2µg/ml puromycin for 24hr. Then Actinomycin D (10µg/ml, Sigma A-9415) was added for the indicated times. For cyclohexamide treatment: 24hr after transfection of 293 cells, 10µg/ml of cyclohexamide (Sigma C0934) were added to the plates for the indicated times and then cells were harvested. For MG-132 treatment: 24hr after transfection of 293 cells, 50µM of MG-132 (Calbiochem 474790) were added to the plates for 2hr. and then cells were harvested.

Retroviral infection, transfections and reporter assay.
Cells of the packaging cell line 293T (2x10 6 cells) were co-transfected with 10µg of ψ ecotropic packaging plasmid, pSV-ψ-E-MLV, providing packaging helper function and 10µg of the relevant plasmid using the calcium phosphate method. Chloroquin (25µM final concentration, Sigma C6628) was added to the transfection medium. After 8hr the transfection medium was replaced with fresh DMEM supplemented with 10% FCS, and 5ml of medium containing retroviruses were collected at 6hr intervals. Five collections were pooled together and frozen in aliquots. For infection of Swiss 3T3 or Rat1a, cells were incubated for 5hr at 37°C in 3 ml of retroviral supernatant, supplemented with 8µg/ml of polybrene (Sigma H9268). Then, 7 ml of DMEM, containing 10% FCS were added, and after 24hr the medium was replaced with fresh medium containing 10% serum and 2µg/ml of puromycin (Sigma P7130). Transfections -H1299 and 293 cells were transfected by a calcium phosphate method. NIH 3T3 cells and 3T3 fibroblasts derived from RB -/mice were transfected using lipofectAMINE reagent (GIBCO BRL). Cells were harvested 24-40hr following transfection. Cell lysis, β-Gal, and luciferase assays were performed essentially as described (27).

Western blot analysis.
For western blot analysis 30µg of protein from each lysate, as determined by Bradford assay, or 20µl from the luciferase assay extracts were loaded, resolved by electrophoresis on an SDS 10% polyacrylamide gel and transferred to filters (Protran, BA 85, S&S). Filters were incubated with the indicated antibodies over night in PBS with 0.05% Tween and 5% dry milk. Binding of the primary antibodies was detected using an enhanced chemiluminescence kit (ECL, Amersham).

Oncogenic Ras elicits an increase in E2F-1 levels
The effect of Ras on E2F-1 was studied initially using co-transfection experiments. Coexpression of E2F-1 and oncogenic Ras (H-RasV12) in 293 cells led to a significant increase of E2F-1 protein level. This increase in the levels of E2F-1 did not depend on co-expression of its heterodimeric partner, DP-1, since it was detected both in the absence and presence of DP-1 ( Fig 1A). Similar results were obtained using the human lung carcinoma cell line H1299 (data not shown). Levels of E2F-2 and E2F-3 were similarly elevated upon coexpression of oncogenic Ras (Fig. 1B). However, this Ras-induced increased of E2F levels was not shared by all E2Fs and, as previously reported (31), activated Ras did not cause a significant change in the levels of another E2F family member, E2F-4. This suggests that the effect of H-RasV12 on E2F level is specific to the E2F-1, -2, -3 sub-family. Additional studies focused on the effect of oncogenic Ras on E2F-1 levels.
Therefore, we tested whether the differences we detect in E2F-1 protein level are due to   1C). Moreover, endogenous Ras activity was required for the cell cycle-dependent induction of endogenous E2F-1 since dominant negative H-RasN17 interfered with this induction (Fig. 1C). Similar results were obtained using Swiss 3T3 murine fibroblasts (data not shown). These results corroborate our data with exogenous E2F-1 and further implicate Ras in the physiological regulation of E2F-1 levels.
E2F-1 is a short-lived protein and is degraded through the ubiquitin-proteasome pathway (22,24). Therefore, a plausible mechanism explaining the Ras-induced increase in E2F-1 protein levels involves interfering with E2F-1 degradation. However, this is not the case since analysis of E2F-1 protein stability by measurement of its half-life time indicated that the E2F1 protein stability was not significantly altered in the presence of activated Ras (Fig 2A,   B). Inhibition of proteasome activity further supported the notion that activated Ras does not have a notable effect on E2F-1 degradation. As expected, treating cells with a proteasome inhibitor, MG132, resulted in an increase in E2F-1 protein levels. In cells co-expressing E2F-1 and Ras, E2F-1 protein level was significantly higher prior to MG132 addition but it was further increased, to a similar extent, upon MG132 addition ( Fig 2C). Thus, the Ras-induced increase in E2F-1 protein levels is most probably not due to inhibition of E2F-1 protein degradation.
Next, we studied the effect of Ras on E2F-1 levels using a Rat-1a-derived cell line containing a stably integrated Zinc-inducible E2F-1 (Rat-1a-MT-wtE2F-1) (4). Infection of these cells with a retrovirus harboring H-RasV12 resulted in a significant increase in E2F-1 protein after addition of Zinc (Fig. 3A). This increase was accompanied by an elevation in E2F-1 mRNA (Fig. 3B). A similar elevation was detected in endogenous E2F-1 mRNA levels upon infection of Swiss 3T3 cells with a retrovirus harboring H-RasV12. Hence, Ras elevates E2F-1 mRNA levels. Ras-dependent transcriptional regulation of the E2F-1 gene is probably not the underlying mechanism since the increase in E2F-1 was detected also when E2F-1 expression was driven by heterologous promoters (Figures 1A, 3A). Therefore, we next studied the effect of Ras on the half-life of E2F-1 mRNA. As can be seen in figure 3C exogenous E2F-1 mRNA was easily detectable in Rat-1a-MT-wtE2F-1 cells, however, its levels were reduced below detection level 3 hours after addition of actinomycin D. In the presence of activated Ras E2F-1 mRNA levels were not significantly changed after 3 hours and only slightly reduced after 5 hours of the same treatment. These data indicate that activated Ras leads to a significant increase in E2F-1 mRNA stability.
The Ras-induced increase in E2F-1 mRNA levels was detected when oncogenic Ras was coexpressed with an E2F-1 expression plasmid containing the full E2F-1 coding region and no increase was detected when using an E2F expression plasmid containing bases 1-1089 of the coding region (encoding amino acids 1-363). In contrast, co-expression of oncogenic Ras did not elevate the mRNA levels of a truncated E2F-1 containing only bases 381-1311 of its coding region (Fig 3D). These data indicate that the effect of oncogenic Ras on E2F-1 mRNA stability is mediated by an element within bases 1-381 of the E2F-1 coding sequence.

The Ras induced increase in E2F-1 levels is mediated by MEK and PKB and it is RB independent
Since Ras activates a number of signal transduction pathways, we tested which of these pathways plays a role in controlling E2F-1 protein levels. Co-expression of either activated PKB or activated MEK together with E2F-1 resulted in elevated E2F-1 protein levels ( Figure 4A,B), indicating that both the PI3K/PKB pathway and the MAPK/MEK pathway can contribute to the Ras-dependent increase in E2F-1 levels. The Ras-induced increase in E2F-1 protein levels was not blocked by either the PI3K inhibitor Wortmanin alone or by the MEK inhibitor PD-098059 alone (Fig. 4C). However, it was diminished upon simultaneous treatment of cells with both inhibitors (Fig. 4C), further supporting the notion that both PKB and MEK play a role in the Ras-induced increase of E2F-1 levels.
A well-established effect of activated Ras on the RB/E2F pathway involves Ras-induced accumulation of cyclin D1. This accumulated cyclin D1 complexes with Cdk4, resulting in phosphorylation of RB and release of active E2Fs. Indeed, the presence of functional RB was shown to be essential for some of the effects of Ras on cell proliferation (26,37).
However, this effect of Ras on the cyclin D/Cdk4/RB/E2F pathway does not lead to an increase in E2F-1 levels. To test whether RB is required for the increase that we observe, we studied the effect of H-RasV12 on the levels of an E2F-1 mutant lacking the RB binding domain, E2F-1∆18. As can be seen in figure 5A levels of both wtE2F-1 and E2F-1∆18 were similarly increased upon co-expression of activated Ras suggesting that the effect of Ras is RB-independent. Furthermore, co-expression of H-RasV12 and E2F-1 in 3T3 fibroblasts derived from RB -/mice led to an increase in E2F-1 protein levels similar to that detected in wt 3T3 cells ( Figure 5B). Thus, RB is dispensable for this increase.
The Ras-induced increase in E2F-1 levels is accompanied by an increase in the promoter activity of an E2F-regulated gene suggesting that the Ras-induced E2F-1 is transcriptionaly active ( fig. 6).
Overall, the data presented here demonstrate that activated Ras leads to an increase in E2F-1 mRNA and protein levels that results in enhanced E2F transcriptional activity.

Discussion
Both Ras and E2F-1 play pivotal roles in the control of cell proliferation and they collaborate in cell transformation both in tissue culture experiments and in transgenic mice.
However, the molecular mechanism(s) by which these two regulators of cell growth cooperate are not fully understood. This collaboration may involve initiation of distinct cascades of events that synergies downstream to E2F-1 and Ras, though an effect of one of these two proteins on the activity of the other may also contribute to such a collaboration.
We demonstrate here that oncogenic Ras induces an increase in the levels of E2F-1 mRNA and protein levels. Moreover, the dominant negative RasN17 abolishes the induction of E2F-1 in quiescent cells upon serum stimulation, indicating that endogenous Ras plays a role in the regulation of endogenous E2F-1 levels.
High levels of transcriptionaly active E2F-1 were shown to induce S-phase entry in quiescent immortal cells (3,4,38). Ras brings about the accumulation of transcriptionaly active E2F-1 (Fig. 6) and therefore, this ability of Ras to up-regulate E2F-1 levels may be an important factor in the biological activities of E2F-1. Furthermore, it may play an important role in the collaboration between Ras and E2F-1 in controlling cell growth, although additional molecular mechanisms are, most probably, involved. For example, Ras and E2F-1 might collaborate on the activation of the phosphatase Cdc25A that is a known E2F target gene (7) and is directly phosphorylated and activated by the Ras/Raf pathway  (Fig 2). Instead we observed a Ras-induced increase in the levels of E2F-1 mRNA (Fig. 3B and 3D). This increase in E2F-1 mRNA levels was seen when E2F-1 expression was driven by different promoters, making a specific Ras dependent trasncriptional effect unlikely. The mechanism underlying this Ras-induced increase in E2F-1 mRNA level is enhancement of mRNA stability. This is evident from the Rasinduced increase in E2F-1 mRNA levels in the presence of Actinomycin D (Fig. 3C). The first 381 bases of E2F-1 coding sequence mediate the response to Ras-induced signals since the mRNA levels of E2F-1 lacking these 381 are not affected by co-expression of oncogenic Ras (Fig 3D).  (37). The effect of Ras on cyclin D results in an RB dependent effect on E2F activity, and it leads to increased E2F activity without significant changes in E2F levels. In contrast, the Ras-induced increase in E2F-1 levels described here is RB independent and it occurs in the absence of RB and also in the absence of the RB binding domain in the E2F-1 molecule.
The Ras-induced increase in E2F-1 levels and activity, reported here, constitutes a novel functional link between Ras and E2F-1 and it may be part of the explanation for their ability to collaborate in affecting cell growth.