BMS-387032

E2F4 deficiency promotes drug-induced apoptosis

Yihong Ma, Scott N. Freeman & W. Douglas Cress

To cite this article: Yihong Ma, Scott N. Freeman & W. Douglas Cress (2004) E2F4 deficiency promotes drug-induced apoptosis, Cancer Biology & Therapy, 3:12, 1262-1269, DOI: 10.4161/ cbt.3.12.1239
To link to this article: http://dx.doi.org/10.4161/cbt.3.12.1239

Published online: 09 Sep 2004.

Submit your article to this journal

Article views: 43

View related articles

Citing articles: 9 View citing articles

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=kcbt20

[Cancer Biology & Therapy 3:12, 1262-1269, December 2004]; ©2004 Landes Bioscience

Research Paper
E2F4 Deficiency Promotes Drug-Induced Apoptosis

Yihong Ma
Scott N. Freeman
W. Douglas Cress*
Molecular Oncology Program; H. Lee Moffitt Cancer Center and Research Institute; University of South Florida College of Medicine; Tampa, Florida USA
*Correspondence to: Doug Cress; Molecular Oncology Program; The H. Lee Moffitt Cancer Center and Research Institute; 12902 Magnolia Drive; Tampa, Florida 33612-9497 USA; Tel.: 813.979.6703; Fax: 813.979.6700; Email: cressd@mof-
fitt.usf.edu
Received 09/03/04; Accepted 09/14/04
Previously published online as a Cancer Biology & Therapy E-publication: http://www.landesbioscience.com/journals/cbt/abstract.php?id=1239

KEY WORDS
E2F, chemotherapy, apoptosis, BMS-387032, cyclin dependent kinase

ACKNOWLEDGEMENTS
The authors would like to thank Drs. Michele Glozak, Eric Haura, Gerold Bepler and anonymous reviewers at Bristol-Meyer Squibb for scientific input and for comments on the manuscript.
This work was supported by funds from the National Cancer Institute (CA90489-01, W.D.C.), the Department of Defense (National Functional Genomics Pilot Project, 12-12990-01-01, W.D.C.) and by the Molecular Biology, Flow Cytometry, Analytical Microscopy, and Molecular Imaging Core Facilities of the Moffitt Research Institute.

ABSTRACT
E2F1 and E2F4 are known to have opposing roles in cell cycle control. In the present work, we examine the role of both E2F1 and E2F4 in apoptosis induced by three cyclin-dependent kinase inhibitors (roscovitine, BMS-387032, and flavopiridol) as well as by three established chemotherapeutic drugs (VP16, cisplatin and paclitaxel). We find that E2F4 levels are diminished following treatment with cyclin dependent kinase inhibitors (flavopiridol, roscovitine and BMS-387032) or with DNA damaging drugs (cisplatin and VP16). In contrast, each of these drugs induced E2F1. We find that mouse fibroblasts nullizygous for the E2F4 gene are more sensitive to apoptosis induced by roscovitine, flavopiridol, cisplatin, and VP16, whereas E2F1-deficient fibroblasts are less sensitive. Likewise, we find that RNAi-mediated reductions in E2F4 in human cancer cells results in increased drug sensitivity. Taken together, these results support a model in which E2F1 and E2F4 play opposing roles during drug-induced apoptosis.

INTRODUCTION
The term E2F refers to a family of transcription factors that have been implicated in the control of cellular proliferation, differentiation and survival. Knockout studies in mice have made it clear that each member of the E2F family has unique biological roles.1-8
It was first demonstrated over a decade ago that E2F1 expression could drive otherwise quiescent cells into S phase.9 Subsequent studies have compared the activities of various members of the E2F family and have defined three functional classes.10,11 The first class, comprising E2F1, 2 and 3A, induce S phase with high efficiency. The second class, including E2F4 and 5, induces S phase with low efficiency and serves primarily in growth arrest together with members of the Rb family. The most recently discovered members of the E2F family, E2Fs 6 and 7, also appear to be primarily involved in growth restraint and differentiation.8,12-17 The final class, represented by E2F1 alone, induces apoptosis. 10,11,18-20 More recent work, using different model systems, has challenged some details of the initial classification of E2F family members.21 For example, in one model system the over expression of E2F2 or 3A induces apoptosis as efficiently as E2F1, although E2F4 expression does not.21 This observation demonstrates that E2F1 is not the only E2F family member with the potential to induce apoptosis; E2F2 and -3 may share this activity. However, in all studies E2F4 appears to lack the ability to induce apoptosis.10,18-23
While E2F2 and 3A can induce apoptosis under certain circumstances, it appears likely that E2F1 is the most important physiological inducer of apoptosis. For example, E2F1-nullizygous mice develop tumors late in life revealing the important tumor suppressor activity of E2F1 that is most likely related to its ability to induce apoptosis.1,2,24 Furthermore, it has been known for some time that treatment of cells with various DNA damaging agents results in induction of E2F1 levels,25-29 but not of other members of the E2F family. A recent study demonstrates that E2F1 is a physiological target of the ATM (ataxia-telangiectasia-mutated) kinase and that treatment with DNA damaging chemotherapeutic agents induce E2F1 levels via the ATM/ATR pathway.25 E2F2 and 3 are not induced by damaging drugs and are not ATM targets. Although other members of the E2F family do not induce apoptosis under most physiological conditions, it is not clear whether other members of the family might oppose the apoptosis-inducing activity of E2F1.
We have previously shown that E2F1 is induced in response to treatment with flavopiridol, a cyclin dependent kinase inhibitor currently in clinical trials against a number of malignancies.30-36 In the current work, we expand our analysis to include two other cdk inhibitors roscovitine37-39 and BMS-38703240-43 and find that both drugs induce E2F1. During the course of this study we observed that E2F4 levels were consistently

Figure 1. Cdk inhibitors induce E2F1 levels and repress E2F4 levels prior to entry into apoptosis. Logarithmically growing H1299 cells were treated with varying doses of (A) roscovitine, (B) BMS-387032 or (C) paclitaxel, as indicated. Induction of cell death was confirmed by measuring sub-G1 DNA content after 48 hr of drug treatment. E2F1, E2F4 and beta-actin levels were determined by Western blot. An “*” indicates that the apoptosis assay was not performed for that particular dose of drug.

decreased following treatment with flavopiridol, roscovitine and BMS-387032 and with the DNA damaging drugs cisplatin and VP16. Upon examination of multiple cell lines and with multiple chemotherapeutic agents, we find that E2F4 levels are diminished by drug treatment, in sharp contrast to E2F1. In further contrast to E2F1, deficiency of E2F4 predisposes cells to drug-induced apoptosis. During the preparation of our work for publication, researchers studying cell death induced by ionizing radiation demonstrated that E2F4 levels and its association with p130 were increased following irradiation.44 Although this effect was not universal in all of the cell lines studied by the group, this finding supports the idea that E2F4 is serving a regulatory role in cellular survival. Furthermore, this work44 demonstrated that an E2F4 interfering RNA increased the sensitivity of cells to ionizing radiation. We believe these results with ionizing radiation44 and our results with chemotherapeutic drugs support a model in which E2F1 promotes apoptosis, whereas E2F4 promotes cell survival. Furthermore, both studies support the idea that agents that diminish E2F4 activity may enhance cancer therapy.

METHODS AND MATERIALS
Cell Lines and Cell Culture. The H1299 nonsmall cell lung cancer (which lacks expression of p53 protein) line was grown in Dulbecco’s modified eagles medium (DMEM) supplemented with 2 mM L-glutamine, 5% fetal bovine serum (FBS), and 1% penicillin/streptomycin (PS). MDA-MB-231 and MCF7 breast cancer cell lines were grown in DMEM supplemented with 10% FBS. 3T3 immortalized mouse embryo fibroblasts (MEFS) derived from littermate wild type and E2F knock out mice were a gift from Drs. Rachel Rempel and Joseph Nevins (Duke University) and were grown in DMEM with 15% FBS. 293 cells were from the American Type Culture Collection and were grown in DMEM with 5% FBS. Aventis Pharmaceuticals provided flavopiridol. VP16, cisplatin, roscovitine and pacilitaxel were purchased from Sigma and concentrated stock solutions were prepared in dimethysulfoxide. Dr. Jack Hunt (Bristol-Myers Squib) provided BMS-387032, which was dissolved in water.
Flow Cytometry. Levels of apoptosis were determined in triplicate (unless otherwise indicated by absence of error bars) by flow cytometry using a PharMingen APO-BrdU kit without modification. After treatments,

floating and attached cells (released by trypsinization) were pooled, washed and resuspended in PBS containing 1% paraformaldehyde as a cross-linker. Cross-linked cells were then washed twice in PBS and fixed with ice-cold 70% ethanol for at least four hours. Fixed cells were again washed in PBS and resuspended with reaction buffer, TdT enzyme, and bromo-dUTP for 1 hr at 37˚C. Cells were subsequently rinsed with 1.0 ml of PBS and resus- pended with fluorescein-labeled anti-BrdU in the dark for 30 min at room temperature. Propidium iodide (50-g/ml) and RNase (0.25-mg/ml) were added, and the cells were incubated for 30 min. At least 1 x 104 cells per experimental condition were analyzed for fluorescence on a Becton- Dickinson FACScan using Cell Quest software. For cells transfected with GFP and E2F4RNAi, at least ten thousand GFP-positive cells were acquired for sub-G1 analysis. In several experiments relative levels of apoptosis were determined by PI staining (estimating sub-G1 DNA content) as a less expensive alternative to the APO-BrdU stain. This approach is justified since most of the drugs used in these experiments are well known to induce apoptosis coincident with inducing sub-G1 DNA content.
Western Blots. Cell lysates were normalized for total protein content (50 g) and subjected to SDS-PAGE. Primary antibodies used in these studies consisted of E2F4 (Santa Cruz Biotechnology; SC-1082), E2F1 (Santa Cruz Biotechnology; SC-251) and -actin (Sigma; A5441). Detection of proteins was accomplished using horseradish-peroxidase-conjugated secondary anti- bodies and enhanced chemiluminescence (ECL) purchased from Amersham.
The pBS/U6-E2F4 RNAi Vector. The RNAi vector, pBS/U6, which is presently marketed as pSilencer by Ambion, was a kind gift of Dr. Yang Shi (Harvard University). Plasmid pBS/U6-E2F4-RNAi was derived from pBS/U6 by ligation of a double-stranded oligonucleotide into pBS/U6 cleaved with Apa I and Eco RI. The sequences of the two single-stranded oligonucleotides were 5’-AATTAAAAAA CCCTTCTACC TCCTTTGAG T CTCTTGAACT CAAAGGAGGT AGAAGGG GGC C-3’ and 5’-CCCTTCTACC TCCTTTGAG T TCAAGAGACT CAAAGGAGGT
AGAAGGG TTT TTT-3’. The oligonucleotides were purchased from Integrated DNA Technologies and were gel purified prior to annealing and ligation.
Transient Transfection of E2F4-RNAi. Plasmids pBS/U6 (empty vector) and pBS/U6-E2F4-RNAi were transfected into 293 cells using calcium phosphate together with the AdTrack plasmid in a 4:1 ratio. The AdTrack vector, which expresses the GFP protein, was used to mark transfected cells

Figure 2. E2F1 deficiency reduces apoptosis induced by cdk inhibitors. (A) Logarithmically growing H1299 cells (white histograms) derived with an empty pBS/U6 vector and their E2F1-deficient counterparts (H1299-E2F1RNAi, shaded histograms) were treated with single doses of 0.5 M BMS-387032 (BMS), 10 M cisplatin (CP), 0.2 M flavopiridol (FP), 20 M roscovitine (RO), 0.2 M paclitaxel (TA) or 10 M VP-16 (VP), for 48 hrs, as indicated. Levels of apoptosis determined using an Apo-BrdU assay are indicated. In this experiment, all drugs gave a statistically significant difference (P value < 0.05) between E2F1 wt and E2F1-deficient cells. (B) The indicated H1299-derived cell lines were treated with either 0.2-M flavopiridol (FP) or 10-M VP16 for the indicated periods of time (in hours). E2F1, E2F4 and beta-actin levels in treated cells were determined by Western blot.

and was a gift from Dr. Tim Kowalik (University of Massachusetts). Approximately 16–18 h after transfection cells were washed twice with PBS and fresh medium containing drugs was applied. Cells were harvested for flow cytometry 56 hr after drug treatment.

RESULTS
In previous work, we demonstrated that E2F1 levels are induced in response to treatment with flavopiridol, a cdk inhibitor. To determine if E2F1 induction is common to cdk inhibitors, we treated H1299 cells (a p53-deficient human lung cancer cell line) with the cdk inhibitors rosco- tivine and BMS-387032. As a control, cells were also treated with paclitaxel, which induces apoptosis via the inhibition microtubule formation. Figure 1 demonstrates that roscotivine, BMS-387032, and paclitaxel all induce apoptosis in a dose-dependent manner. Roscotivine and BMS-387032 induced E2F1, while paclitaxel did not.
In previous work, we derived an H1299 cell line that lacks E2F1 due to stable expression of a small hairpin E2F1 inhibitory RNA (shRNAi).45 To determine if E2F1 deficiency would provide protection from roscotivine and BMS-387032, we treated H1299-U6-E2F1RNAi cells and control cells (H1299-U6) derived with empty vector with these drugs and compared apoptosis in the two lines. Figure 2A reveals that E2F1-deficiency signifi- cantly protects H1299 cells from apoptosis induced by both cdk inhibitors. In addition, we find that E2F1-deficiency also protects the cells from treatment with several DNA damaging drugs (known to induce E2F1 via activation of ATM) as well as the microtubule inhibitor paclitaxel (which does not induce E2F1) (Fig. 1). These data suggest that E2F1 may play a very general role in cellular predisposition toward apoptosis, even in cases such as paclitaxel in which E2F1 levels are not specifically induced by the drug. Figure 2B

demonstrates that flavopiridol, a cdk inhibitor, and VP16, a DNA damaging drug, both induce E2F1 levels in a time dependent manner and that this induction does not occur in the H1299 cell line expressing the E2F1 small hairpin RNAi. Similar results were obtained with the other cdk inhibitors and DNA damaging drugs, but not with paclitaxel (data not shown). It may be noteworthy that the reductions in E2F4 are slightly slower and less pro- nounced in the E2F1-deficient cells (see Discussion).
We noted in Figures 1 and 2 that E2F4 protein levels generally declined in cells treated with apoptosis-inducing drugs; suggesting that its down regulation might play a role in apoptosis. To test this, we treated two breast cancer cell lines (MDA-MD-231 and MCF-7) with several drugs to determine if E2F4 was consistently down regulated by chemotherapeutic drugs in other human cancer cell lines. Figure 3 reveals that in every case, except paclitaxel, E2F4 was moderately to significantly down regulated following treatment with drugs, coincident with apoptosis. Taken together, these results suggest that certain cdk inhibitors and DNA-damaging drugs may activate signaling pathways that lead to down regulation of E2F4.
To determine whether up regulation of E2F1 has a causative role in apoptosis induced by the drugs under study or merely change coincidentally, we compared an established E2F1 nullizygous mouse embryo fibroblast (MEFs) line with a wild-type (WT) control derived in parallel from sibling mice. If E2F1 drives drug-induced apoptosis, it would be predicted that E2F1-nullizygous MEFs would be less sensitive to drug treatments than cells derived from WT littermates. Figure 4A demonstrates that E2F1-null MEFs are significantly less sensitive than their WT counterparts to most of the drugs tested. Figure 4B confirms that all of the drugs tested lead to a decline in E2F4. The observation that E2F1-null MEFs have more E2F4 after drug treatment than their wild-type counterparts may suggest that E2F1 negatively regulates E2F4 and this is considered in the Discussion.

Figure 3. Cdk inhibitors and DNA damaging agents repress E2F4 in breast cancer lines. (A) Logarithmically growing MDA-MB-231 breast cancer cells were treated for 48 hrs with single doses of 0.5 M BMS-387032 (BMS), 10 M cisplatin (CP), 0.2 M flavopiridol (FP), 20 M roscovitine (RO), 0.2 M paclitaxel (TA) or 10 M VP-16 (VP), as indicated. Induction of apoptosis was confirmed by measuring sub-G1 DNA content in the treated cells as indicated.
(B) Same as (A), except MCF-7 cells were used. (C) Logarithmically growing MDA-MB-231 breast cancer cells were treated as indicated and E2F1, E2F4 and beta-actin levels were determined by Western blot. In the left panel cells were treated for 48 hrs. In the right panel cells were treated with single doses of 0.5 M BMS-387032 (BMS), 20 M roscovitine (RO), or 0.2 M paclitaxel (TA). (D) Same as (C), except MCF-7 cells were used.

To explore the idea that downregulation of E2F4 has a causative role in apoptosis we compared an established E2F4 nullizygous mouse embryo fibroblast (MEFs) line with a wild-type (WT) control line derived in par- allel from sibling animals. If E2F4 down regulation contributes to apopto- sis, it would be predicted that E2F4-null MEFS would be more sensitive to various chemotherapeutic drugs than WT controls. Figure 5A demonstrates that the E2F4 knockout line is significantly more sensitive than its WT counterpart to most of the drugs tested. Figure 5B confirms that all of the drugs tested lead to a decline in E2F4 in the MEFs that have E2F4, as occurs in human cancer cell lines. Having observed that E2F1-null MEFs are less sensitive to drugs and that E2F4-null MEFs are more sensitive, we expanded genetic analysis to
human cells. Toward this end we generated a series of potential E2F4 shRNAi expression vectors and tested them for efficacy using an E2F4-driven luciferase assay and Western blots as previously described for E2F1.45 An

effective shRNAi was identified (see Material and Methods for details) and was cotransfected transiently together with a GFP reporter into HEK 293 cells, which were chosen because of their relatively high transfection efficiency (estimated at 20% based on GFP expression). Since we had not previously characterized HEK 293 cells, it was important to establish that HEK 293s behave in a similar manner to the other lines tested following treatment with chemotherapeutic drugs. Figure 6A reveals that the 293 cells are indeed sensitive to drug concentrations used. Figure 6B demonstrates that endogenous E2F4 levels are reduced in the HEK 293 cells upon treatment with the drugs, just as they are in MEFs, MDA-MB-231s, MCF-7s and H1299s. We wanted to verify that the E2F4 RNAi was functional in the HEK 293 cells; however, since we could only transfect approximately 20% of the HEK 293 cells it was not easy to see reductions in E2F4 levels upon transfection of the RNAi. To circumvent this problem, a Flag-Tagged version of E2F4 was cotransfected into the HEK 293s either with empty RNAi

Figure 4. E2F1-null MEFs are more resistant to drug-induced apoptosis than WT counterparts. (A) Logarithmically growing MEFs derived from littermate WT and E2F1-/- mice cells were treated for approximately 48 hrs with the indicated chemotherapeutic drugs and apoptosis was determined by flow cytometry using an Apo-BrdU kit (BD Pharmingen). Since these fibroblasts are in general less sensitive to drug induced apoptosis the drug concentrations used here were twice those used in the treatment of human cancer lines. Specifically, these experiments used a single dose of 1 M BMS-387032 (BMS), 20 M cisplatin (CP), 0.4 M flavopiridol (FP), 40 M roscovitine (RO), 0.4 M paclitaxel (TA) or 20 M VP-16 (VP), as indicated. In this experiment, all treatments gave a statistically significant difference (P value < 0.05) between E2F1 WT and E2F1-deficient cells, except roscotivine (RO). (B) Logarithmically growing MEFs derived from littermate WT and E2F1-/- mice cells were treated for 24 or 48 hrs with the indicated chemotherapeutic drugs and E2F4 and beta-actin levels were determined by Western.

vector or with the E2F4RNAi vector. Since the un-transfected cells do not express Flag-E2F4 the signals shown in Figure 6C correspond solely to the transfected cells and reveals that the E2F4 shRNAi is highly effective against Flag-E2F4 in the HEK 293 cells in transient assays.
Having established that HEK 293 cells behaved like other cells, they were transfected with a GFP expressing vector to mark transfected cells together with either an empty shRNAi expression vector or with a vector that expresses an E2F4 shRNAi. After 16 hrs, transfected cells were treated with various chemotherapeutic drugs to induce cell death. Cell death in GFP positive cells was then measured 56 hrs following drug treatment. Figure 6D reveals that transient E2F4 deficiency clearly sensitizes these cells to drug-induced apoptosis consistent with the results using E2F4-null MEFs.

DISCUSSION
Previous work has shown that E2F1 and E2F4 possess opposing activities in cell cycle control.16 Our present work suggests that a similar opposition also exists regarding the ability of E2F1 and E2F4 to influence cell survival in the face of chemotherapeutic drug treatment. Specifically, we find that treatment with several apoptosis-inducing drugs (including three cdk inhibitors and two DNA damaging agents) caused significant increases in E2F1 levels and declines in the levels of E2F4 protein in a manner that parallels induction of apoptosis. Furthermore, these changes in E2F1 and E2F4 do not appear to be merely coincidental with apoptosis since we demonstrate that E2F1 deficiency decreases sensitivity, while E2F4 deficiency increases sensitivity to drug-induced apoptosis.

One drug, paclitaxel, did not appear to specifically activate the expression of E2F1 or repress the expression of E2F4. In spite of this, E2F deficiency reduced sensitivity to paclitaxel, whereas E2F4 deficiency increased sensitivity. These results suggest that both E2F1 and E2F4 modulate predisposition to undergo drug-induced apoptosis in general. We propose that E2F1 and E2F4 oppose one another in the control of expression of proteins that prepare a cell to undergo apoptosis. For example, it is known that E2F1 activates the expression levels of a number the pro-caspase proteins that mediate cell death.46 This activation of pro-caspase expression is not by itself sufficient to induce apoptosis; however, in the presence of an exogenous apoptotic signal, such as paclitaxel treatment, the abundant pro-caspases are cleaved into active forms and then rapidly mediate the demise of the cell.
A second important point of this paper is that certain chemother- apeutic drugs, such as cdk inhibitors and DNA damaging agents, clearly activate signaling pathways that specifically target E2F1 and E2F4. Based on the literature, we would expect that the DNA damaging drugs are stabilizing the E2F1 protein via activation of the ATM/ATR pathway,25-29 while cdk inhibitors would be expected to stabilize the E2F1 protein by inhibition of cyclin A/cdk complexes that normally down regulate E2Fs1-3A in S phase.47,48 However, in preliminary experiments (not shown) we fail to see an increase in the half-life of E2F1 following treatment with either class of drugs. Instead, we find that all of these drugs induce E2F1 transcriptionally. There has been virtually no published work describing the mechanisms

Figure 5. E2F4-null MEFs are more sensitive to drug-induced apoptosis than WT counterparts. (A) Logarithmically growing MEFs derived from littermate WT and E2F4-/- mice cells were treated for approximately 48 hrs with the indicated chemotherapeutic drugs and apoptosis was determined by flow cytometry using an Apo-BrdU kit (BD Pharmingen) as described in the legend of Figure 4A. In this experiment, all treatments gave a statistically significant difference (P value < 0.05) between E2F4 WT and E2F4-deficient cells, except paclitaxel (TA). The experiments presented in Figures 4 and 5 are not directly comparable since the paired cell lines were derived and the experiments were performed at different times. (B) Logarithmically growing MEFs derived from littermate WT and E2F1-/- mice cells were treated for 24 or 48 hrs with the indicated chemotherapeutic drugs and E2F4 and beta-actin levels were determined by Western.

regulating E2F4 protein levels. Although we have little evidence for any particular mechanism at this point, we do note that E2F4 levels are elevated in E2F1-null cells both before and after drug treatment and that increases in E2F1 tend to occur before decreases in E2F4. This observations suggests that E2F4 may be transcriptionally repressed by E2F1, as is Mcl-1, another E2F1-repressed pro-survival gene.45,49,50
A recent report describes an increase in E2F4 protein and its associ- ation with p130 within the first twenty-four hours following treatment of prostate carcinoma cells with -irradiation.44 Although this effect was not observed in similarly treated breast cancer cell lines and the study did not look at E2F4 levels at times beyond 24 hours, this work does indicate that E2F4 may respond in different ways to different insults in different cell lines. We also observe modest
increases in E2F4 levels at early times following drug treatment; however, over time E2F4 levels are reduced as cells enter apoptosis. A potential explanation for these results would suggests that the initial increase in E2F4 protein does not drive apoptosis, as does E2F1, but is rather a futile attempt to survive irradiation. We conclude this since both studies demonstrate that depleting E2F4 using RNAi made cells more sensitive to radiation or to drugs.
We do not know specifically how E2F4 deficiency contributes to apoptosis. We note that E2F4-deficient MEFs grow normally in culture suggesting that E2F4’s contribution to survival is only relevant during conditions when potent apoptotic signals have been received. Over-expression of E2F4 alone does not significantly protect cells from drug or E2F1-induced apoptosis (data not shown), and thus,

we do not think that E2F4 acts as a dominant negative molecule to block E2F1. Rather, the most likely mechanism is that E2F4-containing complexes, such as the E2F4/DP1/p130 complex,44 block the apop- tosis inducing action of E2F1 by stable occupancy of potential E2F1 binding sites in E2F1-regulated promoters. In fact, we find that the E2F4 partner p130 is also reduced following drug treatment (not shown). In the absence of E2F4, we propose that this repressive complex cannot form. Presently, it is not clear how other E2F family members contribute to drug-induced apoptosis. A recent paper studying E2F3-deficient animals suggests that E2F3B, another growth restraining member of the E2F family, may also serve to promote survival specifically by blocking the action of E2F1.51
Although important questions remain, one issue that is clear is that targeting E2F4 expression with shRNAi increases the sensitivity of cancer cells to drug or radiation-induced apoptosis. This observation could be of clinical relevance. For example, a potential use of this information may lie in predicting response to chemotherapeutic drugs. Based on our experiments in cell culture it might be predicted that tumors with high E2F1 and low E2F4 might be more sensitive to chemotherapeutic treatments. Alternatively, as RNAi technology develops it may be possible to transiently reduce E2F4 levels in cancer tissue using E2F4 RNAi, such as the one described in this work. Transient reductions in E2F4 might enhance the efficacy of standard chemotherapeutic drugs.

Figure 6. An E2F4 small hairpin inhibitory RNA sensitizes cells to drug-induced apoptosis. (A) Logarithmically growing 293 cells were treated with the indicated drugs and apoptosis levels were determined. (B) 293 cells were treated with drugs as indicated and E2F4 levels were determined by Western blot. (C) 293 cells were transfected with a Flag-E2F4 expression vector (indicated by a “+”) with or without an E2F4 shRNAi expression vector. Transfection of empty pcDNA3 or pBS/U6 are indicated with a “-“. Expression levels of the Flag-E2F4 reporter at 24 and 48 hrs after transfection were determined by Western blot. (D) 293 cells were transfected with the empty RNAi vector pBS/U6 or with an E2F4 shRNAi expression vector (shaded) together with a GFP expression vector to identify transfected cells. Following transfection, cells were treated with the indicated drugs and then fixed for analysis by flow cytometry. Apoptosis in GFP-positive cells was determined based on sub-G1 DNA content. In this experiment, all treatments gave a statistically significant difference (P value < 0.05) between E2F4 WT and E2F4-deficient cells, except flavopiridol (FP).

References
1. Yamasaki L, Jacks T, Bronson R, Goillot E, Harlow E, Dyson NJ. Tumor induction and tissue atrophy in mice lacking E2F-1. Cell 1996; 85:537-48.
2. Field SJ, Tsai FY, Kuo F, Zubiaga AM, Kaelin Jr WG, Livingston DM, et al. E2F-1 func- tions in mice to promote apoptosis and suppress proliferation. Cell 1996; 85:549-61.
3. Murga M, Fernandez-Capetillo O, Field SJ, Moreno B, Borlado LR, Fujiwara Y, et al. Mutation of E2F2 in mice causes enhanced T lymphocyte proliferation, leading to the development of autoimmunity. Immunity 2001; 15:959-70.
4. Humbert PO, Rogers C, Ganiatsas S, Landsberg RL, Trimarchi JM, Dandapani S, et al. E2F4 is essential for normal erythrocyte maturation and neonatal viability. Mol Cell 2000; 6:281-91.
5. Rempel RE, Saenz-Robles MT, Storms R, Morham S, Ishida S, Engel A, et al. Loss of E2F4 activity leads to abnormal development of multiple cellular lineages. Mol Cell 2000; 6:293-306.
6. Gaubatz S, Lindeman GJ, Ishida S, Jakoi L, Nevins JR, Livingston DM, et al. E2F4 and E2F5 play an essential role in pocket protein-mediated G1 control. Mol Cell 2000; 6:729-35.
7. Lindeman GJ, Dagnino L, Gaubatz S, Xu Y, Bronson RT, Warren HB, et al. A specific, nonproliferative role for E2F-5 in choroid plexus function revealed by gene targeting. Genes Dev 1998; 12:1092-8.
8. Storre J, Elsasser HP, Fuchs M, Ullmann D, Livingston DM, Gaubatz S. Homeotic trans- formations of the axial skeleton that accompany a targeted deletion of E2f6. EMBO Rep 2002; 3:695-700.
9. Johnson DG, Schwarz JK, Cress WD, Nevins JR. Expression of transcription factor E2F1 induces quiescent cells to enter S phase. Nature 1993; 365:349-52.
10. DeGregori J, Leone G, Miron A, Jakoi L, Nevins JR. Distinct roles for E2F proteins in cell growth control and apoptosis. Proc Natl Acad Sci USA 1997; 94:7245-50.
11. Kowalik TF, De Gregori J, Schwarz JK, Nevins JR. E2F1 overexpression in quiescent fibroblasts leads to induction of cellular DNA synthesis and apoptosis. J Virol 1995; 69:2491-500.
12. Di Stefano L, Jensen MR, Helin K. E2F7, a novel E2F featuring DP-independent repres- sion of a subset of E2F-regulated genes. Embo J 2003; 22:6289-98.
13. de Bruin A, Maiti B, Jakoi L, Timmers C, Buerki R, Leone G. Identification and charac- terization of E2F7, a novel mammalian E2F family member capable of blocking cellular proliferation. J Biol Chem 2003; 278:42041-9.

14. Kherrouche Z, Begue A, Stehelin D, Monte D. Molecular cloning and characterization of the mouse E2F6 gene. Biochem Biophys Res Commun 2001; 288:22-33.
15. Trimarchi JM, Fairchild B, Wen J, Lees JA. The E2F6 transcription factor is a component of the mammalian Bmi1-containing polycomb complex. Proc Natl Acad Sci USA 2001; 98:1519-24.
16. Trimarchi JM, Lees JA. Sibling rivalry in the E2F family. Nat Rev Mol Cell Biol 2002; 3:11-20.
17. Dahme T, Wood J, Livingston DM, Gaubatz S. Two different E2F6 proteins generated by alternative splicing and internal translation initiation. Eur J Biochem 2002; 269:5030-6.
18. DeGregori J, Kowalik T, Nevins JR. Cellular targets for activation by the E2F1 transcrip- tion factor include DNA synthesis- and G1/S-regulatory genes. Mol Cell Biol 1995; 15:4215-24.
19. Mann DJ, Jones NC. E2F-1 but not E2F-4 can overcome p16-induced G1 cell-cycle arrest. Curr Biol 1996; 6:474-83.
20. Lukas J, Petersen BO, Holm K, Bartek J, Helin K. Deregulated expression of E2F family members induces S-phase entry and overcomes p16INK4A-mediated growth suppression. Mol Cell Biol 1996; 16:1047-57.
21. Vigo E, Muller H, Prosperini E, Hateboer G, Cartwright P, Moroni MC, et al. CDC25A phosphatase is a target of E2F and is required for efficient E2F-induced S phase. Mol Cell Biol 1999; 19:6379-95.
22. DeGregori J, Leone G, Ohtani K, Miron A, Nevins JR. E2F-1 accumulation bypasses a G1 arrest resulting from the inhibition of G1 cyclin-dependent kinase activity. Genes Dev 1995; 9:2873-87.
23. Wang D, Russell JL, Johnson DG. E2F4 and E2F1 have similar proliferative properties but different apoptotic and oncogenic properties in vivo. Mol Cell Biol 2000; 20:3417-24.
24. Weinberg RA. E2F and cell proliferation: A world turned upside down. Cell 1996; 85:457-9.
25. Lin WC, Lin FT, Nevins JR. Selective induction of E2F1 in response to DNA damage, mediated by ATM-dependent phosphorylation. Genes Dev 2001; 15:1833-44.
26. Huang Y, Ishiko T, Nakada S, Utsugisawa T, Kato T, Yuan ZM. Role for E2F in DNA dam- age-induced entry of cells into S phase. Cancer Res 1997; 57:3640-3.
27. Blattner C, Sparks A, Lane D. Transcription factor E2F-1 is upregulated in response to DNA damage in a manner analogous to that of p53. Mol Cell Biol 1999; 19:3704-13.
28. Hofferer M, Wirbelauer C, Humar B, Krek W. Increased levels of E2F-1-dependent DNA binding activity after UV- or gamma-irradiation. Nucleic Acids Res 1999; 27:491-5.

29. Meng RD, Phillips P, El-Deiry WS. p53-independent increase in E2F-1 expression enhances the cytotoxic effects of etoposide and of adriamycin. Int J Oncol 1999; 14:5-14.
30. Tan AR, Headlee D, Messmann R, Sausville EA, Arbuck SG, Murgo AJ, et al. Phase I clin- ical and pharmacokinetic study of flavopiridol administered as a daily 1-hour infusion in patients with advanced neoplasms. J Clin Oncol 2002; 20:4074-82.
31. Thomas JP, Tutsch KD, Cleary JF, Bailey HH, Arzoomanian R, Alberti D, et al. Phase I clinical and pharmacokinetic trial of the cyclin-dependent kinase inhibitor flavopiridol. Cancer Chemother Pharmacol 2002; 50:465-72.
32. Tan AR, Swain SM. Review of flavopiridol, a cyclin-dependent kinase inhibitor, as breast cancer therapy. Semin Oncol 2002; 29:77-85.
33. Schwartz GK, O’Reilly E, Ilson D, Saltz L, Sharma S, Tong W, et al. Phase I study of the cyclin-dependent kinase inhibitor flavopiridol in combination with paclitaxel in patients with advanced solid tumors. J Clin Oncol 2002; 20:2157-70.
34. Wang HK. Flavopiridol. National cancer institute. Curr Opin Investig Drugs 2001; 2:1149-55.
35. Shapiro GI, Supko JG, Patterson A, Lynch C, Lucca J, Zacarola PF, et al. A phase II trial of the cyclin-dependent kinase inhibitor flavopiridol in patients with previously untreated stage IV nonsmall cell lung cancer. Clin Cancer Res 2001; 7:1590-9.
36. Schwartz GK, Ilson D, Saltz L, O’Reilly E, Tong W, Maslak P, et al. Phase II study of the cyclin-dependent kinase inhibitor flavopiridol administered to patients with advanced gas- tric carcinoma. J Clin Oncol 2001; 19:1985-92.
37. Meijer L, Borgne A, Mulner O, Chong JP, Blow JJ, Inagaki N, et al. Biochemical and cel- lular effects of roscovitine, a potent and selective inhibitor of the cyclin-dependent kinases cdc2, cdk2 and cdk5. Eur J Biochem 1997; 243:527-36.
38. Gray N, Detivaud L, Doerig C, Meijer L. ATP-site directed inhibitors of cyclin-dependent kinases. Curr Med Chem 1999; 6:859-75.
39. McClue SJ, Blake D, Clarke R, Cowan A, Cummings L, Fischer PM, et al. In vitro and in vivo antitumor properties of the cyclin dependent kinase inhibitor CYC202 (R-roscovi- tine). Int J Cancer 2002; 102:463-8.
40. Dai Y, Grant S. Small molecule inhibitors targeting cyclin-dependent kinases as anticancer agents. Curr Oncol Rep 2004; 6:123-30.
41. Misra RN, Xiao HY, Kim KS, Lu S, Han WC, Barbosa SA, et al. N-(cycloalkylamino)acyl-2-aminothiazole inhibitors of cyclin-dependent kinase 2. N-[5-[[[5-(1,1-dimethylethyl)-2-oxazolyl]methyl]thio]-2-thiazolyl]-4- piperidinecarbox- amide (BMS-387032), a highly efficacious and selective antitumor agent. J Med Chem 2004; 47:1719-28.
42. Senderowicz AM. Novel small molecule cyclin-dependent kinases modulators in human clinical trials. Cancer Biol Ther 2003; 2:S84-95.
43. Senderowicz AM. Small-molecule cyclin-dependent kinase modulators. Oncogene 2003; 22:6609-20.
44. DuPree EL, Mazumder S, Almasan A. Genotoxic stress induces expression of E2F4, lead- ing to its association with p130 in prostate carcinoma cells. Cancer Res 2004; 64:4390-3.
45. Ma Y, Cress WD, Haura EB. Flavopiridol-induced apoptosis is mediated through up-reg- ulation of E2F1 and repression of Mcl-1. Mol Cancer Ther 2003; 2:73-81.
46. Nahle Z, Polakoff J, Davuluri RV, McCurrach ME, Jacobson MD, Narita M, et al. Direct coupling of the cell cycle and cell death machinery by E2F. Nat Cell Biol 2002; 4:859-64.
47. Jiang J, Matranga CB, Cai D, Latham Jr VM, Zhang X, Lowell AM, et al. Flavopiridol-induced apoptosis during S phase requires E2F-1 and inhibition of cyclin A-dependent kinase activity. Cancer Res 2003; 63:7410-22.
48. Matranga CB, Shapiro GI. Selective sensitization of transformed cells to flavopiridol-induced apoptosis following recruitment to S-phase. Cancer Res 2002; 62:1707-17.
49. Croxton R, Ma Y, Song L, Haura EB, Cress WD. Direct repression of the Mcl-1 promot- er by E2F1. Oncogene 2002; 21:1359-69.
50. Croxton R, Ma Y, Cress WD. Differences in DNA binding properties between E2F1 and E2F4 specify repression of the Mcl-1 promoter. Oncogene 2002; 21:1563-70.
51. Aslanian A, Iaquinta PJ, Verona R, Lees JA. Repression of the Arf tumor suppressor by E2F3 is required for normal cell cycle kinetics. Genes Dev 2004; 18:1413-22.