[CANCER RESEARCH 63, 7410 –7422, November 1, 2003]
Flavopiridol-Induced Apoptosis during S Phase Requires E2F-1 and Inhibition of
Cyclin A-Dependent Kinase Activity1
Jingrui Jiang,2 Christian B. Matranga,2, 3 Dongpo Cai, Vaughan M. Latham, Jr., Xinxin Zhang, April M. Lowell,
Fabio Martelli,4 and Geoffrey I. Shapiro5
Department of Medical Oncology and Lowe Center for Thoracic Oncology [J. J., C. B. M., D. C., V. M. L. Jr., X. Z., A. M. L., G. I. S.] and Department of Cancer Biology [F. M.],
Dana-Farber Cancer Institute, and Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School [G. I. S.], Boston, Massachusetts 02115
ABSTRACT
Transformed cells are selectively sensitized to apoptosis induced by the
cyclin-dependent kinase inhibitor flavopiridol after their recruitment to S
phase. During S phase, cyclin A-dependent kinase activity neutralizes
E2F-1 allowing orderly S phase progression. Inhibition of cyclin A-dependent kinase by flavopiridol could cause inappropriately persistent E2F-1
activity during S phase traversal and exit. Transformed cells, with high
baseline levels of E2F-1 activity, may be particularly sensitive to cyclin
A-dependent kinase inhibition, as the residual level of E2F-1 activity that
persists may be sufficient to induce apoptosis. Here, we demonstrate that
flavopiridol treatment during S phase traversal results in persistent expression of E2F-1. The phosphorylation of E2F-1 is markedly diminished,
whereas that of the retinoblastoma protein is minimally affected, so that
E2F-1/DP-1 heterodimers remain bound to DNA. In addition, manipulation of E2F-1 levels leads to predictable outcomes when cells are exposed
to flavopiridol during S phase. Tumor cells expressing high levels of
ectopic E2F-1 are more sensitive to flavopiridol-induced apoptosis during
S phase compared with parental counterparts, and high levels of ectopic
E2F-1 expression are sufficient to sensitize nontransformed cells to flavopiridol. Furthermore, E2F-1 activity is required for flavopiridolinduced apoptosis during S phase, which is severely compromised in cells
homozygous for a nonfunctional E2F-1 allele. Finally, the response to
flavopiridol during S phase is blunted in cells expressing a nonphosphorylatable E2F-1 mutant incapable of binding cyclin A, suggesting that the
modulation of E2F-1 activity produced by flavopiridol-mediated cyclindependent kinase inhibition is critical for the apoptotic response of S
phase cells.
INTRODUCTION
The Rb6 tumor suppressor pathway is universally disrupted in
human cancer. Rb is the prototype member of a family of proteins that
govern cell cycle progression, and prevents S phase entry in its active,
hypophosphorylated form. The G1-S transition requires phosphorylation of Rb by cdks (1), which are regulated by the state of their own
phosphorylation, as well as their association with activating cyclins
and endogenous cdk inhibitors (2, 3). Whereas some tumors lack Rb
itself, most overexpress cyclins or lose expression of cdk inhibitors,
resulting in deregulated cdk activity and functional inactivation of Rb
(4, 5). Inhibition of cdk activity in exponentially growing transformed
Received 3/19/03; revised 8/13/03; accepted 8/29/03.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance with
18 U.S.C. Section 1734 solely to indicate this fact.
1
Supported by NIH Grant R01 CA90687.
2
J. J. and C. B. M. contributed equally to this work.
3
Present address: Graduate School of Biological Sciences, University of Massachusetts, Worcester, MA 01655.
4
Present address: Laboratorio Patologia Vascolare, Isituto Dermopatico
dell’Immacolata-IRCCS, Via dei Monti di Creta 104, 00167 Rome, Italy.
5
To whom requests for reprints should be addressed, at Department of Medical
Oncology, Dana-Farber Cancer Institute, Dana 810A, 44 Binney Street, Boston, MA
02115. Phone: (617) 632-4942; Fax: (617) 632-1977; E-mail: geoffrey_shapiro@
dfci.harvard.edu.
6
The abbreviations used are: Rb, retinoblastoma; cdk, cyclin-dependent kinase; MEF,
mouse embryo fibroblast; PARP, poly (ADP-ribose) polymerase; TUNEL, terminal
deoxynucleotidyltransferase-mediated nick end labeling; BCS, bovine calf serum; TdT,
terminal deoxytransferase.
cells frequently results in cell cycle arrest, with cytostatic effects on
tumor growth (reviewed in Ref. 6).
Rb prevents cell cycle progression by binding to E2F transcription
factor family members; Rb-E2F complexes are potent transcriptional
repressors (7). The sequential phosphorylation of Rb during G1 by
cyclin D- and E-dependent kinases results in derepression and eventual release of free, transcriptionally active E2F, which, along with a
heterodimeric DP family member partner, directs the expression of
genes required for S phase entry and advance (8). However, following
S phase entry, E2F is only activated transiently. Orderly S phase
progression requires the timely inactivation of E2F, in part accomplished by cyclin A-dependent kinase activity. Cyclin A-cdk2 stably
associates with E2F-1, and directs phosphorylation of the E2F/DP
heterodimer, neutralizing its DNA binding capacity (9 –12). Cyclin
A-cdk1 holoenzymes also contribute to E2F-1 phosphorylation (13).
Inappropriately persistent E2F-1 activity during S phase traversal
and exit has been shown to result in apoptosis. For example, the
ectopic expression of an E2F-1 mutant incapable of binding to cyclin
A or of a nonphosphorylatable DP-1 mutant causes S phase delay and
apoptosis (14). Both p53-dependent and p53-independent apoptotic
pathways may be activated in response to E2F-1 (15).
Disruption of Rb or its upstream regulators in tumor cells would be
expected to result in high levels of free E2F-1 compared with levels
found in their normal counterparts. Transformed cells may be particularly dependent on cyclin A-dependent kinase activity during S
phase to limit E2F-1 activity and prevent apoptosis. A reduction in
cyclin A-dependent kinase activity may result in the persistence of
residual E2F activity that may cause the death of cells with high
enough E2F-1 activity to surpass the threshold required to induce
apoptosis and may therefore be lethal to transformed cells yet tolerable to normal cells. This concept was tested recently using short
peptides that block the interaction of cyclin A-cdk2 with substrates
such as E2F-1. Introduction of these peptides induced S phase arrest
and apoptosis; cell death occurred selectively in transformed
cells (16).
Flavopiridol is a small molecule cdk inhibitor currently in clinical
development (17–19). As flavopiridol inhibits multiple cyclin-cdk
holoenzymes (including cyclin D-cdk4, cyclin A/E-cdk2, and cyclin
B-cdk1), it induces G1 and G2 arrest in a wide variety of exponentially
growing cells (18). In many solid tumor cell lines, death of exponentially growing cells only occurs after prolonged exposures to concentrations of drug greater than the range necessary to inhibit cdks,
suggesting that cytotoxicity is related to the effects of high flavopiridol concentrations on other cellular targets (20).
However, we have recently shown that cells can be sensitized to
flavopiridol if they are first recruited to S phase. Instead of an initial
cytostatic cell cycle arrest, treatment of cells during S phase results in
inhibition of S phase progression and apoptosis at concentrations of
drug that correspond to cdk inhibition (21). The result is similar
whether cells are recruited to S phase by synchronization or by prior
exposure to noncytotoxic concentrations of chemotherapy drugs that
cause S phase accumulation, including cisplatin and gemcitabine. In
7410
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2003 American Association for Cancer
Research.
FLAVOPIRIDOL-INDUCED APOPTOSIS IS E2F-1-DEPENDENT
were plated in the absence or presence of 100 ␮M ZnCl2. Twenty-four h after
plating, cells were treated with DMSO or flavopiridol either in the absence or
presence of ZnCl2.
3T3 Cells. E2F-1 ⫹/⫹, E2F-1 ⫹/⫺ and E2F-1 ⫺/⫺ MEFs were harvested
from embryos isolated at 13.5 days after the mating of E2F-1 ⫹/⫺ mice (26).
Primary MEFs at passage 0 were cultured in DMEM containing 10% FCS.
Spontaneous immortalization of primary MEFs was accomplished according to
the 3T3 protocol described by Todaro and Green (27). Cells of each genotype
were counted and seeded at a density of 3 ⫻ 105/60-mm dish every 3 days with
a medium change the subsequent day, for 24 passages, at which time protein
was isolated and cells were characterized for levels of E2F-1, p16INK4A, and
Rb. Experiments were performed with cells between passages 25 and 30.
Restoration of E2F-1 expression in E2F-1 ⫺/⫺ cells was accomplished by
retroviral infection. Viral supernatants from Bing packaging cells transfected
with pBABE-puro or pBABE-puro-E2F-1 were used to infect exponentially
growing E2F-1 ⫺/⫺ cells. After 48 h of infection, cells were replated and
subjected to sequential treatment with hydroxyurea and flavopiridol.
Drug Treatment. Stock solutions of flavopiridol (provided by the Drug
Synthesis & Chemistry Branch, Developmental Therapeutics Program, Division of Cancer Treatment and Diagnosis, National Cancer Institute), hydroxyurea (Sigma-Aldrich Co., St. Louis, MO), and gemcitabine (Eli Lilly and Co.,
Indianapolis, IN) were prepared as described previously (21, 28). Flavopiridol
was used at concentrations ranging from 150 to 300 nM. Cells were treated with
1 mM hydroxyurea to achieve synchronization at the G1-S boundary. Gemcitabine concentrations ranging from 10 nM to 10 ␮M were used. The concentration
and protocol for gemcitabine treatment varied with each cell line and were
designed to allow accumulation of cells in early-mid S phase or at the G1-S
boundary before flavopiridol treatment. Cells were plated 24 –30 h before drug
treatment and were at subconfluent density at the time of drug exposure. For
sequential drug treatments, cells were washed in PBS before application of the
second drug.
Fluorescence-Activated Cell-Sorting Analysis. Cell cycle analysis was
performed as described previously (21, 28). Nonadherent and adherent cells
were combined. After fixation and treatment in 500 ␮g/ml RNase A, cells were
resuspended in 1 ml 69 ␮M propidium iodide in 30 mM sodium citrate. Cells
were analyzed for DNA content by flow cytometry using the ModFit program
(Verity Software House, Topsham, ME).
Detection of Apoptosis by Flow Cytometry. A fluorescein apoptosis
detection kit was used (Promega, Madison, WI), as described previously (21,
28). Nonadherent and adherent cells were combined. After formaldehyde and
ethanol fixation, cells were incubated with fluorescein-12 dUTP in the absence
or presence of TdT. After washes prescribed in the manufacturer’s instructions,
cells were resuspended in PBS containing 5 ␮g/ml propidium iodide and 500
␮g/ml RNase A. Cells were analyzed for DNA content and apoptosis using
two-color flow cytometry. Apoptosis was quantitated as the percentage of cells
MATERIALS AND METHODS
shifting to fluorescein-positivity in the presence of TdT.
Western Blot Analysis. Whole cell lysates were prepared in NP40 lysis
Tumor Cell Lines. NCI-H661 and A549 non-small cell lung cancer cell
lines as well as U2OS osteosarcoma cells were obtained from the American buffer [50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 1% NP40, and 1 mM
Type Culture Collection (Rockville, MD) and maintained in DMEM supple- phenylmethylsulfonyl fluoride], and clarified by centrifugation. Nuclear extracts were prepared using NE-PER Nuclear Extraction Reagents (Pierce,
mented with 10% BCS.
Generation of Cell Lines Expressing Ectopic E2F-1. Stable cell lines Rockford, IL), according to the manufacturer’s instructions. The nuclear lywere generated by retroviral infection. Fugene 6 reagent was used to transfect sates were prepared in buffer supplemented with a protease inhibitor mixture
Bing packaging cells (22) with pBabe-puro or pBabe-puro into which a cDNA (Calbiochem-Novabiochem, San Diego, CA) and 1 mM phenylmethylsulfonyl
fluoride. Protein concentrations were determined using the Bradford assay
encoding wild-type E2F-1 with a hemagluttinin tag had been cloned (pBABEpuro-E2F-1-WT-HA; Refs. 14, 23). Viral supernatants were used to infect (Bio-Rad, Richmond, CA), and equivalent amounts (10 –100 ␮g) were subNCI-H661, A549, and U2OS cells, which were then selected and maintained jected to SDS-PAGE. Western blotting was performed as described previously
in DMEM with 10% BCS containing 2 ␮g/ml puromycin. Mass populations of (28). The following antibodies were used: 1:500 dilution of anti E2F-1 (C-20;
cells infected with the pBABE-puro virus were isolated, as were mass popu- Santa Cruz Biotechnology, Santa Cruz, CA); 1:1000 dilution of anti-E2F-1
lations and individual clones arising from the pBABE-puro-E2F-1-WT-HA clone KH95 (Santa Cruz Biotechnology); 1 ␮g/ml anti-HA, clone 12CA5
infections. In transient experiments, NCI-H661 cells were plated at 5 ⫻ 105 (Roche Molecular Biochemicals, Indianapolis, IN); 1:1000 dilution of anticells/well in six-well plates and transfected with 1 ␮g pRcCMV, or pRcCMV- p16INK4A (M-156; Santa Cruz Biotechnology); 1:1000 dilution of anti-cleaved
containing hemagluttinin-tagged wild-type E2F-1 or [E2F-1 (⌬24)] (9), en- PARP (Cell Signaling Technology, Beverly, MA); 1:500 dilution of anti-DP-2
coding an E2F-1 species incapable of cyclin A binding. In these experiments, (C-20; Santa Cruz Biotechnology); 1:1000 dilution of anti-p21Waf1/Cip1 and
cells were cotransfected with 4 ␮g of a construct encoding DP-2 (pCMV- anti-p27Kip1 (Signal Transduction Laboratories, Lexington, KY); 1:1000 dilution of anti-phospho-p27Kip1 (pT187; Zymed Laboratories, South San FranDP-2) to permit nuclear localization of the E2F-1 mutant species (24).
Rat1a Fibroblasts. Rat1a fibroblasts with stably integrated p1093 or cisco, CA); and 1:5000 dilution of antitubulin, clone DM 1A (Sigma-Aldrich
p1093-E2F-1 (in which E2F-1 cDNA is under the control of the metallothio- Co.). For analysis of Rb, the following antibodies were used at 1:1000 dilution:
nein promoter) were maintained in DMEM supplemented with 10% BCS and anti-Rb, clone G3–245 (PharMingen, San Diego, CA), anti-COOH-terminal
500 ␮g/ml neomycin (25). Subconfluent Rat 1a-p1093 or p1093-E2F-1 cells Rb, anti-Rb (pS795), anti-Rb (pS780), anti-Rb (pS807/811), and anti-Rb
7411
addition, flavopiridol-induced apoptosis during S phase occurs selectively in transformed cells (21).
The mechanism of S phase sensitization to flavopiridol may be
related to the known consequences of cdk inhibition during S phase.
Flavopiridol-mediated inhibition of cyclin A-dependent kinase activity during S phase would be expected to result in decreased phosphorylation of the E2F/DP heterodimer, permitting inappropriately persistent E2F-1 activity that induces an apoptotic response. In addition,
cyclin A-dependent kinase inhibition may also account for the selective killing of transformed cells when they are exposed to flavopiridol
during S phase (16).
In this report, we have tested several predictions of this model.
First, we show that following synchronization at the G1-S boundary,
flavopiridol treatment results in persistent expression of E2F-1 during
S phase traversal. Consistent with the inhibition of cyclin A-cdk2 by
flavopiridol, the phosphorylation of E2F-1 is markedly diminished, so
that it remains capable of binding to DNA. The phosphorylation of Rb
is only minimally affected, and increased complex formation between
Rb and E2F-1 is not detected during S phase in the presence of
flavopiridol. Moreover, we show that manipulation of the cellular
level of E2F-1 activity affects the apoptotic response induced by
flavopiridol during S phase. For example, ectopic expression of high
levels of E2F-1 renders transformed cells more sensitive to flavopiridol during S phase than parental controls expressing lower levels. In
addition, nontransformed cells engineered to express high levels of
E2F-1 are similarly sensitized, suggesting that it is indeed the high
baseline levels of E2F-1 in transformed cells that accounts for their
selective sensitivity to flavopiridol during S phase. Furthermore,
disruption of E2F-1 activity prevents the apoptotic response to flavopiridol during S phase. We have investigated this question using
immortalized cells homozygous for a nonfunctional E2F-1 allele.
Finally, we show that expression of an E2F-1 mutant that cannot be
regulated by cyclin A-dependent kinase activity diminishes the degree
of sensitization to flavopiridol during S phase, indicating that the
modulation of E2F-1 activity resulting from flavopiridol-mediated
cdk inhibition contributes to induction of the apoptotic response.
These data explain the sensitivity of transformed cells to flavopiridol
during S phase, as well as the synergism observed when flavopiridol
follows chemotherapy agents that retard S phase progression.
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2003 American Association for Cancer
Research.
FLAVOPIRIDOL-INDUCED APOPTOSIS IS E2F-1-DEPENDENT
(pT821; all from Cell Signaling Technology, Beverly, MA or Biosource
International, Camarillo, CA).
Immunoprecipitation. For the detection of Rb/E2F-1 complexes, 1 mg of
nuclear lysate was precleared with 2 ␮g of mouse or rabbit IgG and then used
for immunoprecipitation with 2 ␮g of anti-E2F-1 antibody (C20; Santa Cruz
Biotechnology) or anti-Rb antibody, clone G3–245 (PharMingen). Immunoprecipitates were collected on protein A/G agarose beads (Santa Cruz Biotechnology), washed in NP40 buffer [50 mM Tris (pH 8.0), 150 mM NaCl, and 1%
NP40], boiled in sample buffer and subjected to SDS-PAGE, followed by
Western blotting using the opposite antibody to detect complex formation.
Recovery of E2F-1 and Rb was monitored by Western blotting with anti-E2F-1
(KH95; Santa Cruz Biotechnology) or anti-Rb (Cell Signaling Technology)
antibodies, respectively.
Assessment of E2F-1 Phosphorylation. After release from a hydroxyureainduced G1-S block, U2OS cells were treated for 8 h with either DMSO or 300
nM flavopiridol in phosphate-free DMEM containing 10% dialyzed BCS,
supplemented with 1 mCi of [32P]-Pi. Labeled cells were washed in PBS,
released by trypsinization, and used for the preparation of nuclear lysates,
which were subjected to immunoprecipitation with an anti-E2F-1 antibody
(C-20; Santa Cruz Biotechnology), used at a 4 ␮g/ml, a saturating concentration. Immunoprecipitates were collected using protein A-Sepharose beads,
washed with lysis buffer, and boiled for 5 min in 50 ␮l of sample buffer.
Radiolabeled bands were analyzed by 10% SDS-PAGE followed by autoradiography. After boiling in sample buffer, 5 ␮l was electrophoresed separately
and subjected to Western blotting for E2F-1.
Analysis of E2F-1/DP-1 DNA Binding Activity. NCI-H661 or U2OS
cells were treated with 1 mM hydroxyurea for 24 h, followed by release into
DMSO or 300 nM flavopiridol. Nuclear lysates were collected at 2, 5, and 8 h
after release from the block at the G1-S boundary, and applied to Mercury
TransFactor kits (Clontech Laboratories, Palo Alto, CA), according to the
manufacturer’s instructions. Kits contain a 96-well format in which wells are
coated with an oligonucleotide containing the E2F-1/DP-1 consensus binding
sequence. After 1 h, the nuclear lysates were removed from the well, which
was washed and treated with anti-E2F-1 (1:50) or anti-DP-1 primary antibody
(1:100), followed by additional washing and incubation with horseradish
peroxidase-anti-mouse IgG to detect the bound primary antibody. After addition of substrate, the enzymatic product was measured with a microtiter plate
reader at absorbance 655 nM (29).
Statistical Analysis. Statistical evaluation of data were performed with the
two-tailed, unpaired Student’s t test. P ⬍ 0.05 was considered statistically
significant.
RESULTS
Flavopiridol Treatment During S Phase Inhibits Cell Cycle
Progression and Results in Persistent Expression of E2F-1. In
many exponentially growing solid tumor cell lines, treatment with
flavopiridol causes arrest at the G1 and G2 boundaries. G1 arrest is
associated with direct inhibition of G1 cdk activity and transcriptional
repression of D-cyclins, leading to decreased E2F-1 expression and a
diminution in cellular S phase DNA content. To examine the effects
of flavopiridol treatment during S phase, NCI H661, A549, and U2OS
cells were synchronized at the G1-S boundary with hydroxyurea, and
released into DMSO or 300 nM flavopiridol (Figs. 1 and 2). Fig. 1
demonstrates that S phase progression is slowed after flavopiridol
Fig. 1. Flavopiridol treatment during S phase slows cell cycle
progression and causes persistent expression of E2F-1 in NCI-H661
cells. A, cells were treated with 1 mM hydroxyurea for 24 h,
achieving synchronization at the G1-S boundary. Cells were then
released into DMSO or 300 nM flavopiridol for 2, 4, 6, 8, or 10 h.
Adherent and nonadherent cells were pooled, fixed, and analyzed by
flow cytometry. B, nuclear lysates were extracted directly after
treatment with hydroxyurea (1H), or at the indicated times after
release into DMSO or flavopiridol, and subjected to Western blotting. Results demonstrate the persistent expression of E2F-1 as cells
traverse S/G2 in the presence of flavopiridol. Levels of cyclin A and
cdk2 are stable, whereas levels of p27Kip1 fall in flavopiridol. Rb
was analyzed with an antibody recognizing phosphorylated and
unphosphorylated forms (G3-245, designated Rb), a COOH-terminal antibody, as well as with phosphospecific antibodies. Minimal
dephosphorylation of Rb occurs over this time course, as indicated
by results with phosphospecific Rb antibodies. C, cells were treated
with 1 mM hydroxyurea alone or followed by release into DMSO for
2 h (D2) or flavopiridol for 4 or 6 h (F4 and F6). One mg of nuclear
lysate was subjected to immunoprecipitation with anti-E2F-1 (C-20)
or control rabbit IgG antibodies, followed by Western blotting with
anti-Rb (G3-245) or anti-E2F-1 (KH95) antibodies (top blot, first
four lanes, and bottom blot). In addition, 50 ␮g of each extract (D2,
F4, and F6) was subjected to Western blotting with the anti-Rb
(G3-245) antibody (top blot, last 3 lanes). Exposure of the top blot
was ⬃1 h; exposure of the bottom blot was ⬃1 min. The data
demonstrate that only a small amount of E2F-1 is in complex with
Rb after release from the hydroxyurea-induced block and that E2F-1
does not accumulate in a complex with Rb as cells traverse S phase
in the presence of flavopiridol.
7412
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2003 American Association for Cancer
Research.
FLAVOPIRIDOL-INDUCED APOPTOSIS IS E2F-1-DEPENDENT
Fig. 2. Flavopiridol treatment during S phase causes persistent expression of E2F-1 in
A549 and U2OS cells without evidence of Rb dephosphorylation. Cells were treated with
1 mM hydroxyurea for 24 h, followed by release into DMSO or 300 nM flavopiridol for
2, 4, 6, 8, or 10 h. In both of these cell lines, E2F-1 expression persists in the presence of
flavopiridol, without detectable effects on the phosphorylation of Rb. During S phase
traversal in flavopiridol, levels of cyclin A and cdk2 are stable, whereas levels of
p21Waf1/Cip1 and p27Kip1 fall.
exposure. In addition, whereas E2F-1 levels are diminished in DMSOtreated cells as they approach the G2 boundary, they remain persistently high in the presence of flavopiridol. In contrast, levels of cyclin
A and cdk2 are not affected and remain constant over the first 10 h of
drug treatment, whereas levels of Cip/Kip proteins are decreased by
flavopiridol during S phase traversal. Our previous studies demonstrated that in NCI-H661 cells treated with flavopiridol in this manner,
a small amount of apoptosis is detected at 6 h, but becomes more
substantial after 12 h of drug exposure (21). In A549 and U2OS cells,
the slowing of S phase progression is not as marked, and the onset of
apoptosis of the S/G2 population is more delayed, occurring between
18 and 24 h. Nonetheless, in all of these cell line lines, persistent
expression of E2F-1 in the presence of flavopiridol precedes the onset
of apoptosis as cells traverse S and G2.
These experiments were performed after release from synchronization at the G1-S boundary, past the peak of cyclin E-cdk2 activity, a
time when Rb is phosphorylated and incapable of E2F-1 binding.
Nonetheless, it is possible that flavopiridol-mediated cdk inhibition
could results in Rb dephosphorylation during S phase, resulting in
E2F-1 inactivation. Therefore, we examined Rb phosphorylation in
response to flavopiridol after release from the hydroxyurea-induced
G1-S block. In A549 and U2OS cells, dephosphorylation of Rb during
S phase traversal in the presence of flavopiridol is not detected, either
with an antibody recognizing phosphorylated and unphosphorylated
forms or with phosphospecific antibodies (Fig. 2). In NCI-H661 cells,
a slightly faster migrating form of Rb is present at 8 and 10 h after
exposure to flavopiridol during S phase. This band is present, but not
as apparent when an antibody recognizing the COOH terminus of Rb
is used. Therefore, a low level of either dephosphorylation or caspase
cleavage (30), or both could generate the faster migrating band. If Rb
dephosphorylation occurs, it is only to a small degree, as the phosphorylation at many of the cdk4- and cdk2-specific sites is largely
preserved (Refs. 31–34; Fig. 1B), with no change in the phosphorylation of Rb at S780 and S795, and only a slight decrease at S807/811,
T821, and S249/252. In addition, as cells pass through S phase, an
increase in complex formation between E2F-1 and Rb is not observed
(Fig. 1C). After release from hydroxyurea into DMSO for a short
period (2 h) or flavopiridol for longer periods (4 or 6 h), nuclear
lysates were subjected to immunoprecipitation with an anti-E2F-1
antibody, followed by Western blotting with either anti-Rb or antiE2F-1 antibodies. Complexes containing E2F-1 and Rb are only
detected after prolonged exposures of the blots, indicating that E2F-1
remains largely free of Rb during S phase in the presence of flavopiridol. Similar results were obtained after release into flavopiridol for 8 h
and with reciprocal immunoprecipitations (data not shown).
E2F-1 Phosphorylation Is Inhibited by Flavopiridol during S
Phase, Facilitating Persistent DNA Binding of the E2F-1/DP-1
Heterodimer. Following activation of E2F-1 during early S phase,
cyclin A-cdk2 phosphorylates both E2F-1 and DP-1, neutralizing the
DNA binding capacity of the heterodimer (9 –12). Appropriately
timed deactivation of E2F-1 by cyclin A-cdk2 is critical for proper S
phase progression. Cyclin A-cdk2 inhibition by flavopiridol should
result in decreased phosphorylation of E2F-1, which would result in
persistent DNA binding of the transcription factor. Fig. 3 demonstrates that phosphorylation of E2F-1 is reduced over the course of S
phase traversal in the presence of flavopiridol in U2OS cells. Similar
data were obtained in NCI-H661 cells (data not shown). Consequently, both E2F-1 and DP-1 remain capable of binding to their
consensus sequence throughout the course of S phase (Fig. 4), demonstrated by an ELISA assay designed to detect DNA binding by
either E2F-1 or DP-1.
High Levels of E2F-1 Additionally Sensitize Transformed Cells
to Flavopiridol-Induced Apoptosis during S Phase. To determine
whether the amount of E2F-1 expressed can affect sensitivity to
flavopiridol-mediated apoptosis during S phase, NCI-H661, A549,
and U2OS cells were engineered to express high levels of ectopic
E2F-1 by retroviral transfer. Fig. 5A demonstrates that E2F-1 expression could be increased several fold with this strategy. Examination of
nuclear and cytoplasmic lysates indicated that in pBABE-E2F-1infected cells, ⬃80% of the ectopic E2F-1 is nuclear (data not shown).
Under conditions of normal exponential growth in 10% serum, these
cell lines were able to tolerate increased expression of E2F-1 without
a significant change in cell cycle pattern (Fig. 5B). Control and E2F-1
overexpressing cells were then treated with hydroxyurea to achieve a
7413
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2003 American Association for Cancer
Research.
FLAVOPIRIDOL-INDUCED APOPTOSIS IS E2F-1-DEPENDENT
activity could account for the differential sensitivity of transformed
cells to flavopiridol-induced apoptosis during S phase, we used a rat
fibroblast cell line stably transfected with an E2F-1 cDNA under
control of the metallothionein promoter (16, 25). After induction with
zinc, Rat 1a cells transfected with p1093-E2F-1 express high levels of
E2F-1, whereas cells transfected with the parent vector maintain
expression of baseline levels of E2F-1 (Fig. 7A). Under conditions of
exponential growth in 10% serum, induction of E2F-1 did not significantly affect the cell cycle pattern of transfected cells (Fig. 7B, left
panels). Treatment of Rat1a fibroblasts with either 250 or 300 nM
flavopiridol for 24 h slows cell cycle progression but does not result
in tight cell cycle arrest. Because these cells do not undergo pro-
Fig. 3. Flavopiridol treatment during S phase results in decreased phosphorylation of
E2F-1. Top, U2OS cells were synchronized at the G1-S boundary with hydroxyurea for
24 h. After a PBS rinse, cells were incubated in phosphate-free medium containing
dialyzed FBS and 1 mCi Pi in the presence of DMSO or 300 nM flavopiridol for 8 h.
Nuclear lysates were extracted and immunoprecipitation was performed with an antiE2F-1 antibody (C-20). The antibody was used at saturating concentration such that no
detectable E2F-1 was found in the supernatant after immunoprecipitation (data not
shown). Immunoprecipitates were collected on protein A-agarose beads, boiled in sample
buffer, and subjected to electrophoresis on 10% SDS-PAGE. After boiling, 10% of the
sample was electrophoresed separately and subjected to Western blotting with anti-E2F-1.
The results demonstrate the persistent expression of E2F-1 with reduced phosphorylation
in the presence of flavopiridol and are representative of those seen in three separate
experiments. Bottom, the cell cycle protein of cells released from hydroxyurea into DMSO
or flavopiridol for 6 h in phosphate-free medium is shown, demonstrating that S phase
progression is similar to that seen in full medium.
block at the G1-S boundary, followed by release into flavopiridol.
When 300 nM flavopiridol is used, this treatment has been shown to
kill most transformed cell types. Therefore, in these experiments, the
concentration of flavopiridol was lowered to 150 nM so that fewer
cells in the control population underwent apoptosis. In separate experiments, we confirmed that this concentration of flavopiridol causes
an ⬃50% inhibition of cdk2 activity in these cell lines (data not
shown). Fig. 5B demonstrates that high levels of ectopic E2F-1
additionally sensitize U2OS cells to flavopiridol treatment after release into S phase, so that following synchronization at the G1-S
boundary, differences in the degree of cell death between control and
E2F-1 overexpressing cells are evident after 24 h of flavopiridol
exposure. Because the sub-G1 peak does not represent the full degree
of DNA fragmentation occurring as cells die from S/G2, TUNEL
analysis was performed; after synchronization and 32 h of flavopiridol
exposure, a large proportion of E2F-1 overexpressing cells die compared with control cells. Similar results were obtained for NCI-H661
and A549 cells (Fig. 5C).
Levels of E2F-1 also affect the sensitivity of cells to flavopiridol
following recruitment to early S phase by a noncytotoxic concentration of gemcitabine (Fig. 6). In both A549 and U2OS cells, ectopic
expression of E2F-1 did not sensitize cells to gemcitabine treatment
alone at the concentrations used in these experiments, but did predispose to a greater degree of cell death to the sequential combination of
gemcitabine and flavopiridol. In the case of U2OS cells, 10 nM
gemcitabine and 300 nM flavopiridol results in significant apoptosis
after 36 – 48 h of flavopiridol exposure. When a high level of ectopic
E2F-1 is expressed, apoptosis is observed earlier (at 24 h) and with a
lower flavopiridol concentration (Fig. 6B).
Overexpression of E2F-1 Sensitizes Nontransformed Cells to
Flavopiridol-Mediated Apoptosis. In our previous work, recruitment to S phase selectively sensitized transformed cell populations to
flavopiridol (21). Because of Rb pathway disruption, transformed
cells are expected to have higher levels of free E2F-1 activity than
their normal counterparts. To examine whether high levels of E2F-1
Fig. 4. E2F-1 and DP-1 remain capable of DNA binding during S phase traversal in the
presence of flavopiridol. A, NCI-H661 cells were treated with 1 mM hydroxyurea for 24 h,
followed by release into DMSO or 300 nM flavopiridol. At the indicated times, nuclear
lysates were extracted and applied to wells coated with the E2F-1 consensus sequence, and
subjected to an ELISA assay with an anti-E2F-1 antibody. Detection of E2F-1 bound to
its consensus sequence was accomplished with a secondary antibody coupled to horseradish peroxidase. The Y-axis indicates absorbance at 655 nM after the addition of
substrate. Graphs represent the data from 5 experiments performed at 2 and 5 h after
release from hydroxyurea, and 7 experiments performed at 8 h. At 8 h, the difference in
the amount of E2F-1 capable of DNA binding in DMSO and flavopiridol treated cells was
statistically significant (P ⫽ 0.00005). In flavopiridol-treated cells, the amount of E2F-1
capable of DNA binding was statistically greater at 8 h compared with 2 h (P ⫽ 0.02).
Bars, ⫾SD. B, the same analysis was performed with an anti-DP-1 antibody. For each
condition, a minimum of three and a maximum of five determinations were used to
generate the graphs. At 8 h, the difference in the amount of DP-1 capable of DNA binding
in DMSO and flavopiridol-treated cells reached statistical significance (P ⫽ 0.02). In
flavopiridol-treated cells, the amount of DP-1 capable of DNA binding was statistically
greater at 8 h compared with 2 h (P ⫽ 0.04). The overall results demonstrate that the
E2F-1/DP-1 heterodimer remains capable of binding to DNA over the course of S phase
traversal in flavopiridol. Bars, ⫾SD.
7414
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2003 American Association for Cancer
Research.
FLAVOPIRIDOL-INDUCED APOPTOSIS IS E2F-1-DEPENDENT
optosis, evidenced by the large sub-G1 peak in p1093-E2F-1 cells
maintained in zinc and flavopiridol. This was confirmed when cell
death was measured by TUNEL assay or by PARP cleavage (Fig. 7,
C and D). Viable cell number over 48 h of flavopiridol treatment is
shown in Fig. 7E. Cells expressing parent vector grow sluggishly in
the presence or absence of zinc, consistent with flavopiridol-induced
slowing of cell cycle progression. In the subclone expressing the
E2F-1 inducible vector, viable cell count is relatively stable over the
48 h of flavopiridol treatment in the absence of zinc. However, after
induction of E2F-1 in the presence of zinc, viable cell number decreases by 24 h, and the cell population is almost entirely dead by
48 h. Therefore, high levels of E2F-1 can confer flavopiridol sensitivity on a nontransformed cell line otherwise resistant to flavopiridolinduced apoptosis.
Confirmatory experiments were performed in these cells after S
phase recruitment with a noncytoxic concentration of gemcitabine
(Fig. 8A). The sequential gemcitabine/flavopiridol combination resulted in the greatest amount of apoptosis after induction of ectopic
E2F-1 in the presence of zinc (Fig. 8, B and C).
Disruption of E2F-1 Activity Inhibits the Apoptotic Response to
Flavopiridol during S Phase. To confirm the critical role of E2F-1
in flavopiridol-induced apoptosis during S phase, we examined 3T3
cells derived from E2F-1 ⫹/⫹, ⫹/⫺, and ⫺/⫺ MEFs. Fig. 9A
demonstrates that E2F-1 was only detectable in nuclear extracts from
Fig. 5. High levels of ectopic E2F-1 additionally sensitize tumor cell lines to flavopiridol during S phase. NCI-H661, U2OS, and A549 cells were infected with control (Babe)
and wild-type E2F-1 (WT) -expressing retroviruses, and stable cell lines were generated
after selection in puromycin. Data for mass populations are shown in the case of U2OS
and A549 infections; data from a representative clone are shown for the NCI-H661
infection. A, Western blot analysis of protein extracted from control and wild-type E2F-1
expressing cells, demonstrating high levels of ectopic E2F-1 expression in cells infected
with the WT virus; ectopic E2F-1 carries a hemagluttinin (HA) tag and can also be
identified with an anti-HA antibody. Blots were stripped and reprobed with an anti-tubulin
antibody to demonstrate equal loading. B, U2OS cells infected with control (Babe) or
wild-type E2F-1-expressing retroviruses were treated with hydroxyurea to achieve synchronization at the G1-S boundary and then released into either DMSO or 150 nM
flavopiridol for 24 h. Adherent and nonadherent cells were pooled, fixed, and stained with
propidium iodide for flow cytometry. Analysis of untreated cells demonstrates that a high
level of ectopic E2F-1 expression does not alter the cell cycle distribution. After release
into S phase, cell death is detected only in E2F-1 overexpressing cells. C, cells were
treated with 1 mM hydroxyurea for 24 h. After a PBS rinse, cells were released into
medium containing 150 nM flavopiridol for an additional 24 h (NCI-H661 and A549) or
32 h (U2OS). At the conclusion of treatment, adherent and nonadherent cells were pooled,
fixed, and subjected to the flown cytometry-based TUNEL assay in which apoptosis was
quantitated. In the case of NCI-H661 cells, data from 7 experiments were combined; for
U2OS and A549 cells, data represent a compilation of 10 experiments. In all of the cases,
differences in the amount of apoptosis in pBABE- and pBABE-E2F-1 (WT) -infected cells
were statistically significant (H661, P ⫽ 0.0000043; U2OS, P ⫽ 0.0000064; A549,
P ⫽ 0.003). Bars, ⫾SD.
nounced G1 arrest after flavopiridol exposure, recruitment to S phase
first was not necessary, and we were able to examine the response of
control and E2F-1 overexpressing cells to flavopiridol alone. As
shown by the fluorescence-activated cell sorter analyses in Fig. 7B,
induction of E2F-1 sensitized these cells to flavopiridol-induced ap-
Fig. 6. High levels of ectopic E2F-1 additionally sensitize tumor cell lines to flavopiridol following S phase recruitment by noncytotoxic concentrations of gemcitabine. A,
control A549 cells (Babe) and those expressing ectopic wild-type E2F-1 (WT) were
treated with 30 ␮M gemcitabine for 1.5 h, followed by a 24-h washout to achieve
synchronization in early S phase (Gem). Cells were subsequently treated for 24 h with
either DMSO or 150 nM flavopiridol. Previous experiments have demonstrated that S
phase delay by flavopiridol is less marked in A549 cells than in other cell types, so that
exposure to flavopiridol after S phase synchronization results in cell death from the entire
S/G2 population (21). B, control U2OS cells (Babe) and those expressing ectopic wildtype E2F-1 (WT) were treated with 10 nM gemcitabine for 24 h, which results in S phase
accumulation in these cells. After a PBS rinse, cells were then treated with DMSO, 150
nM, or 300 nM flavopiridol for an additional 24 h. Adherent and nonadherent cells were
pooled, fixed, and subjected to the flow cytometry-based TUNEL assay, and apoptosis
was quantitated. In the presence of high levels of ectopic E2F-1, apoptosis after the
sequential gemcitabine/flavopiridol combination occurs at lower concentrations of flavopiridol and at earlier times. Data represent a compilation of three experiments. Differences in the degree of apoptosis between pBABE- and pBABE-E2F-1 (WT) -infected cells
after sequential gemcitabine and flavopiridol approached statistical significance at 150 nM
flavopiridol (P ⫽ 0.069) and reached statistical significance at 300 nM flavopiridol
(P ⫽ 0.0017). Bars, ⫾SD.
7415
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2003 American Association for Cancer
Research.
FLAVOPIRIDOL-INDUCED APOPTOSIS IS E2F-1-DEPENDENT
Fig. 7. High levels of ectopic E2F-1 are sufficient to sensitize nontransformed cells to flavopiridol. A, characterization of Rat1a fibroblasts engineered to express control vector
(p1093) or ectopic E2F-1 under the control of the metallothionein promoter (p1093-E2F-1). Subconfluent cells were left untreated or treated with 100 ␮M ZnCl2 for 24 h, after which
protein was extracted and subjected to Western blotting with an anti-E2F-1 antibody (clone KH95). High levels of E2F-1 are induced in p1093-E2F-1 cells in the presence of zinc.
B, cell cycle profiles of control (p1093) and p1093-E2F-1 cells. Cells were fixed and stained with propidium iodide for flow cytometry. Left panels demonstrate that after growth in
100 ␮M ZnCl2 for 24 h, a high level of ectopic E2F-1 expression does not significantly alter the cell cycle distribution of these cells. Treatment with 250 nM flavopiridol for 24 or 48 h
results in cell death primarily in E2F-1 overexpressing cells. C, p1093 or p1093-E2F-1 cells were grown in the absence or presence of ZnCl2 for 24 h, and then exposed to DMSO,
250 nM, or 300 nM flavopiridol in the absence or presence of ZnCl2 for an additional 48 h. At the end of 48 h, adherent and nonadherent cells were pooled, fixed, and subjected to
the flow cytometry-based TUNEL assay in which apoptosis was quantitated. Only Rat1a fibroblasts expressing high levels of ectopic E2F-1 undergo significant apoptosis after 48 h
of flavopiridol exposure. Data represent a compilation of a minimum of three and a maximum of six experiments for each condition. In p1093-E2F-1 cells, differences in the amount
of apoptosis in the absence and presence of zinc were statistically significant at both 250 nM flavopiridol (P ⫽ 0.048) and 300 nM flavopiridol (P ⫽ 0.008). Bars, ⫾SD. D, cells were
grown in the absence or presence of ZnCl2 for 24 h and then treated with DMSO or 300 nM flavopiridol in the absence or presence of ZnCl2 for an additional 48 h. Nonadherent and
adherent cells were pooled, and whole cell lysates were subjected to Western blotting using antibodies detecting cleaved PARP and tubulin. E, left, subconfluent p1093 cells were grown
in the absence or presence of zinc for 24 h, and then treated with 300 nM flavopiridol for an additional 48 h in the absence (F) or presence (E) of zinc. Time 0 represents the initiation
of flavopiridol exposure. At the indicated times, adherent and nonadherent cells were collected, and viable cell count determined by trypan blue exclusion. Right, the same experiment
performed with p1093-E2F-1 cells.
⫹/⫹ and ⫹/⫺ cells. Several sets of cell lines were characterized, and
growth curves and cell cycle distribution patterns were comparable
(Fig. 9B). During immortalization, these cell lines retained p16INK4A
expression and displayed markedly diminished expression of the Rb
protein (Fig. 9A). After release from a hydroxyurea-induced block
into flavopiridol, cells expressing E2F-1 underwent cell death. In
contrast, apoptosis was diminished in cells lacking E2F-1 (Fig. 9, C
and D). When E2F-1 expression was restored to E2F-1 ⫺/⫺ cells,
7416
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2003 American Association for Cancer
Research.
FLAVOPIRIDOL-INDUCED APOPTOSIS IS E2F-1-DEPENDENT
resulted in cell death in cell lines expressing E2F-1 and was markedly
compromised in cells in which E2F-1 expression was genetically
disrupted, when measured by either TUNEL assay (Fig. 10B) or
PARP cleavage (data not shown).
Inhibition of Cyclin A-Dependent Kinase Activity by Flavopiridol Likely Contributes to the Induction of Apoptosis during
S Phase. To validate the importance of inhibition of cyclin Adependent kinase activity by flavopiridol in the induction of apoptosis
during S phase, an E2F-1 mutant incapable of binding to cyclin A
[E2F-1 (⌬24)] was examined. Because this mutant is not phosphorylatable by cyclin A-cdk2, its expression is expected to result in
constitutive deregulated E2F-1 activity, which itself can induce apoptosis (14, 23). Effects of flavopiridol in the presence of this mutant
are expected to be diminished, because inhibition of cyclin A-cdk2
will not affect the activity of this E2F-1 species.
Fig. 11A demonstrates the expression of wild-type E2F-1 and
[E2F-1 (⌬24)] after transient transfection in NCI-H661 cells. Because
cyclin A-binding mutants of E2F-1 also lack nuclear localization
signals, these constructs were introduced along with DP-2, permitting
similar levels of nuclear expression of wild-type and mutant E2F-1
(24). After transfection, cells were synchronized and released into
DMSO or flavopiridol, again at reduced concentration (150 nM) to
avoid a large amount of cell death in control populations. Fig. 11B
demonstrates that expression of wild-type E2F-1 sensitized NCI-H661
cells to flavopiridol-induced apoptosis during S phase by 8-fold in the
context of these experimental conditions. As expected, nuclear expression of the nonphosphorylatable [E2F-1 (⌬24)] resulted in baseline apoptosis in S phase cells exposed only to DMSO. However,
when the cyclin A-binding mutant is expressed, treatment with flavopiridol modulated apoptosis only by 1.5–2-fold. These data suggest
that inhibition of cyclin A-dependent kinase-mediated phosphorylation of E2F-1/DP-1 plays a major role in the induction of apoptosis by
flavopiridol during S phase.
DISCUSSION
Fig. 8. Overexpression of E2F-1 sensitizes nontransformed cells to flavopiridol after
recruitment to S phase by noncytotoxic concentrations of gemcitabine. A, p1093-E2F-1
cells were treated in the absence of ZnCl2 with 60 nM gemcitabine for 1.5 h. After a 24 h
washout in the absence of ZnCl2, cells were recruited to S phase. B, after gemcitabine
treatment in the absence of ZnCl2 (as in A), cells were treated with 300 nM flavopiridol in
the absence or presence of ZnCl2 for an additional 24 h. Nonadherent and adherent cells
were pooled, fixed, and subjected to the flow cytometry-based TUNEL assay in the
presence or absence of TdT. Numbers represent the percentage of fluorescein-positive
cells. C, control (p1093) or p1093-E2F-1 cells were treated with gemcitabine in the
absence of ZnCl2, as in A, to achieve S phase recruitment. Cells were then treated with
DMSO or 300 nM flavopiridol in the absence or presence of ZnCl2 for an additional 24 h.
Nonadherent and adherent cells were pooled, fixed, and subjected to TUNEL assay. Data
from three experiments were used to generate the graphs. In p1093-E2F-1 cells, differences in the amount of apoptosis in the absence and presence of zinc after sequential
treatment with gemcitabine and flavopiridol reached statistical significance (P ⫽ 0.028).
Bars, ⫾SD.
apoptosis after release from a hydroxyurea-induced block into flavopiridol occurred at a degree comparable with that seen in E2F-1
⫹/⫹ cells (Fig. 10A). Similarly, following S phase recruitment with
noncytotoxic concentrations of gemcitabine, release into flavopiridol
Flavopiridol is the first potent inhibitor of cdks to enter clinical
trials. The response of many solid tumors, both in vitro and in vivo, is
cell cycle arrest, with reversible, cytostatic effects on tumor growth.
We have shown previously that recruitment of cells to S phase,
accomplished by cell synchronization or by pretreatment with chemotherapy agents that impose S phase delay, markedly sensitizes them
to flavopiridol (21). The mechanism of S phase sensitivity and selectivity may be related to inhibition of cdk2 activity by flavopiridol
during S phase, preventing the properly timed deactivation of E2F
(9 –11, 16), resulting in S phase delay and apoptosis.
Here, we have examined the effect of flavopiridol treatment as cells
pass through S phase on the expression of several cell cycle proteins.
In flavopiridol-treated cells, S phase progression is slowed and E2F-1
levels persist as cells move toward the G2-M boundary. Consistent
with flavopiridol-mediated inhibition of cyclin A-dependent kinase
activity, the phosphorylation of E2F-1 during S phase is markedly
diminished in the presence of drug, so that the E2F-1/DP-1 heterodimer remains capable of DNA-binding. These events either coincide with or precede the onset of apoptosis in the cell lines examined
(21).
Importantly, in the presence of flavopiridol, E2F-1 likely remains
bound to DNA free of Rb. This is because Rb dephosphorylation
occurs at most to a small degree in response to flavopiridol over the
course of S phase traversal. In these experiments, the cell cycle period
during which cells are exposed to flavopiridol is after cyclin D-cdk4/6
and cyclin E-cdk2 holoenzymes have already phosphorylated Rb.
After the G1-S boundary, Rb typically remains phosphorylated to
7417
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2003 American Association for Cancer
Research.
FLAVOPIRIDOL-INDUCED APOPTOSIS IS E2F-1-DEPENDENT
Fig. 9. Flavopiridol-induced apoptosis following recruitment to S phase is compromised in cells lacking functional E2F-1 alleles. A, characterization of 3T3 cells with
E2F-1 ⫹/⫹, ⫹/⫺, and ⫺/⫺ genotypes. Western blot analysis for two sets of 3T3 cell
lines. Each set contains lines of the designated genotype, derived from MEFs, which, in
turn, were derived from embryos generated from a cross of E2F ⫹/⫺ mice. To identify
E2F-1 in these cells, 100 ␮g of nuclear protein was subjected to Western blotting with a
polyclonal anti-E2F-1 antibody (C20). Whole cell lysates were used for Western blotting
to identify Rb and p16INK4A, respectively. The positive control for Rb expression was a
lysate from p19ARF ⫺/⫺ MEFs. B, representative cell cycle profiles from E2F-1 ⫹/⫹ and
E2F-1 ⫺/⫺ 3T3 cells. Exponentially growing cells were collected, fixed, and stained with
propidium iodide for flow cytometry. The absence of E2F-1 did not alter the cell cycle
profile. C, E2F-1 ⫹/⫹ and ⫺/⫺ 3T3 cells were treated with 1 mM hydroxyurea for 24 h.
After synchronization at the G1-S boundary, cells were rinsed with PBS and released into
300 nM flavopiridol for an additional 24 h. Adherent and nonadherent cells were pooled,
fixed, and subjected to the flow cytometry-based TUNEL assay in the absence or presence
of TdT. The primary data demonstrate that cell death is significantly diminished in E2F-1
⫺/⫺ cells; numbers indicate the percentage of fluorescein-positive cells. D, E2F-1 ⫹/⫹
or ⫺/⫺ 3T3 cells were treated for 24 h with DMSO, 150 nM, or 300 nM flavopiridol for
24 h. Alternatively, after synchronization with 1 mM hydroxyurea for 24 h, cells were
released into DMSO, 150 nM, or 300 nM flavopiridol, as in B. Adherent and nonadherent
cells were pooled, fixed, and subjected to TUNEL assay to quantitate apoptosis. After S
phase recruitment, flavopiridol-induced apoptosis in E2F-1 ⫺/⫺ cells is severely compromised. Data are shown for ⫹/⫹ and ⫺/⫺ 3T3 cells from one set. A minimum of 3 and
a maximum of 12 experiments were used to generate the graphs. Differences in the
amount of apoptosis in E2F-1 ⫹/⫹ and ⫺/⫺ cells treated with hydroxyurea followed by
flavopiridol were statistically significant at 150 nM flavopiridol (P ⫽ 0.00027) and 300 nM
flavopiridol (P ⫽ 0.0015). Bars, ⫾SD.
Fig. 10. A, E2F-1 re-expression in E2F-1 ⫺/⫺ cells restores their ability to undergo
flavopiridol-mediated apoptosis following release from a hydroxyurea-induced block at
the G1-S boundary. E2F-1 ⫺/⫺ cells were infected with control (C, pBABE) or E2F-1
(pBABE-E2F-1) -expressing retroviruses. The Western blot demonstrates the restoration
of E2F-1 expression in pBABE-E2F-1-infected cells. Cells were treated with 1 mM
hydroxyurea for 24 h, rinsed with PBS, and then exposed to DMSO (1H/D) or 300 nM
flavopiridol (1H/F) for an additional 24 h. Adherent and nonadherent cells were pooled,
fixed, and subject to TUNEL assay to quantitate apoptosis. Data from one (1H/D) or two
(1H/F) experiments were used to generate the graphs. In flavopiridol-treated cells,
differences in the amount of apoptosis between pBABE and pBABE-E2F-1 infected cells
were statistically significant (P ⫽ 0.01). Bars, ⫾SD. B, flavopiridol-induced apoptosis
following recruitment to S phase by gemcitabine is compromised in cells lacking functional E2F-1 alleles. E2F-1 ⫹/⫹, ⫹/⫺, or ⫺/⫺ 3T3 cells were treated with 1 or 10 ␮M
(1 or 10 G) gemcitabine for 24 h, concentrations that result in accumulation in S phase or
at the G1-S boundary in these cells. After a PBS rinse, cells were treated with DMSO or
300 nM flavopiridol for an additional 24 h. Adherent and nonadherent cells were pooled,
fixed, and subjected to TUNEL assay to quantitate apoptosis. Apoptosis induced by the
sequential gemcitabine/flavopiridol combination is inhibited in the absence of E2F-1. Data
are shown for 1 set of ⫹/⫹, ⫹/⫺, and ⫺/⫺ cells; a minimum of 3 and maximum of 10
experiments were used to generate the graphs for each condition. The 1 ␮M gemcitabine/
DMSO sequential treatment was performed only once. Ps demonstrated statistically
significant differences between ⫹/⫹ or ⫹/⫺ and ⫺/⫺ cells (For 10G/F: ⫹/⫹ versus
⫺/⫺, P ⫽ 0.000001; ⫹/⫺ versus ⫺/⫺, P ⫽ 0.000095; for 1G/F: ⫹/⫹ versus ⫺/⫺,
P ⫽ 0.0034; ⫹/⫺ versus ⫺/⫺, P ⫽ 0.006). Bars, ⫾SD.
permit S phase completion; maintenance of Rb phosphorylation most
likely does not require continued cdk activity, which is typically
diminished during S phase by expression of endogenous inhibitors,
such as p19INK4D (35), as well as reduction of levels of cyclin E,
7418
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2003 American Association for Cancer
Research.
FLAVOPIRIDOL-INDUCED APOPTOSIS IS E2F-1-DEPENDENT
Fig. 11. The effect of flavopiridol during S phase is blunted in NCI-H661 cells
expressing [E2F-1 (⌬24)], a cyclin A-binding defective mutant. A, subconfluent cells were
cotransfected with vector alone or constructs encoding either hemagglutinin-tagged wildtype E2F-1 or [E2F-1 (⌬24)], along with a construct encoding DP-2 (pCMV-DP-2).
Nuclear lysates extracted 48 h after transfection were subjected to Western blotting with
anti-E2F-1, anti-HA, or anti-DP-2 antibodies. B, transfections were carried out as in A.
Twenty-four h after transfection, cells were treated with 1 mM hydroxyurea for 24 h. After
a PBS rinse, cells were treated with either DMSO or 150 nM flavopiridol for an additional
24 h. Nonadherent and adherent cells were pooled and subjected to TUNEL assay. Graphs
represent compilation of data from three experiments in the case of cells transfected with
vector alone or the construct encoding wild-type E2F-1 and six experiments for cells
transfected with the construct encoding (E2F-1 ⌬24). In the case of vector-expressing
cells, differences in the amount of apoptosis after treatment with hydroxyurea/DMSO or
hydroxyurea/flavopiridol were not statistically significant (P ⫽ 0.308). These treatments
did induce significant differences in apoptosis in cells expressing wild-type E2F-1
(P ⫽ 0.0029). Expression of (E2F-1 ⌬24) produce significantly more baseline apoptosis
compared with expression of wild-type E2F-1 (P ⫽ 0.026). The degree of apoptosis
induced by flavopiridol was blunted in mutant expressing cells compared with wild-type
(P ⫽ 0.018), indicating that the ability of flavopiridol to modulate E2F-1 via inhibition of
cyclin A-dependent kinase activity is required for a maximal apoptotic response. However, in (E2F-1 ⌬24)-expressing cells, flavopiridol induced an increase in apoptosis over
baseline (i.e., DMSO) that was statistically significant (P ⫽ 0.0006), suggesting that other
flavopiridol targets in addition to cyclin A-dependent kinase may also contribute to
E2F-1-dependent apoptosis.
events preserved in transformed cell types (36). Therefore, any additional reduction in cdk activity by flavopiridol specifically during S
phase is unlikely to have major impact on the degree of Rb phosphorylation already accomplished during cell cycle progression.
Flavopiridol treatment during S phase also resulted in diminished
expression of both p21Waf1/Cip1 and p27Kip1. The effect on p27Kip1
expression was surprising, as decreased phosphorylation of this protein (as shown in Fig. 2) after cdk2 inhibition should result in its
stabilization (37, 38). Decreased levels of Cip/Kip proteins during S
phase in the presence of flavopiridol may possibly be explained by
decreased transcription mediated by inhibition of cdk9-pTEFb (39 –
43) or by cleavage of these proteins by caspases early after the
induction of apoptosis (44 – 46).
A model in which the apoptotic response of cells to flavopiridol
during S phase is dependent on modulation of E2F-1 activity by cyclin
A-dependent kinase inhibition leads to several testable predictions.
First, cells with high baseline levels of E2F-1 activity should be more
sensitive to cyclin A-dependent kinase inhibition than cells with lower
levels, because the higher levels of E2F-1 that persist are more likely
to achieve the threshold activity necessary to generate an apoptotic
response during S phase traversal and exit. Here, we have tested this
hypothesis in several tumor cell lines engineered to express high
levels of ectopic E2F-1. The high levels of expression were tolerable
to these cells, all grown in high serum, and presumably ectopic protein
was subject to the same phosphorylation and ubiquitination events
that regulate endogenous E2F-1 activity. Following recruitment to S
phase by synchronization or with noncytotoxic doses of gemcitabine,
cells expressing high levels of E2F-1 were more sensitive to flavopiridol than their parental counterparts. In general, these differences were
most notable at a low flavopiridol concentration, i.e., 150 nM; at the
300 nM concentration, parental cells are also sensitive to hydroxyurea/
flavopiridol and gemcitabine/flavopiridol treatments, so that differences were more difficult to appreciate. The exception was U2OS
cells, in which S phase retardation and apoptosis are more separable
in time; less of a delay in apoptosis occurred in cells expressing high
levels of ectopic E2F-1 (Fig. 6B).
In previous work, we demonstrated that transformed cell populations were selectively sensitized to flavopiridol treatment during S
phase (21). The universal alteration in expression of cell cycle proteins in human tumors is expected to produce higher levels of free
E2F-1 activity in tumor cells compared with their normal counterparts. In this report, we have demonstrated that overexpression of
E2F-1 in nontransformed Rat1a fibroblasts is sufficient to sensitize
these cells to flavopiridol. These data suggest that higher levels of
E2F-1 activity could indeed account for the selective killing of transformed cells when flavopiridol treatment occurs during S phase.
The mechanism for the sensitivity of S phase cells also predicts that
disruption of E2F activity will inhibit the apoptotic response to
flavopiridol. We have tested this using 3T3 cells derived from E2F-1
⫹/⫹, ⫹/⫺, and ⫺/⫺ MEFs. The ability of E2F-1 ⫺/⫺ cells to
undergo flavopiridol-induced apoptosis during S phase is markedly
compromised, indicating that E2F-1 activity is essential for a maximal
apoptotic response. In addition, E2F-1 ⫹/⫹ cells underwent apoptosis
to a greater degree than E2F-1 ⫹/⫺ cells, suggesting a gene dosage
effect.
While our data are in agreement with a previous analysis of flavopiridol in E2F-1 ⫺/⫺ cells (47), it is noteworthy that our experimental conditions were very different. In the former experiments,
nonimmortalized MEFs were used, and exponentially growing cells
were exposed to flavopiridol. A difference in the degree of cell death
at 72 h (⬃14 versus 4%) was reported. Presumably, this is the cell
death occurring late after cell cycle arrest that occurs in both transformed and nontransformed cells. It is of interest that E2F-1 expression may affect the degree of late cell death as well, in addition to the
earlier cell death that occurs when cells are treated during S phase.
Our data are also consistent with results in exponentially growing
NCI-H1299 cells, in which the depletion of E2F-1 by RNA interference (RNAi) was reported to compromise flavopiridol-induced apoptosis (47). Our attempts to deplete E2F-1 activity in the cell lines
examined here, via RNAi or expression of dominant-negative E2F-1
mutants, has resulted in derivatives with higher G1 and diminished S
phase content compared with parental cells. Among transformed
tumor cell lines, cycling cells are more sensitive to flavopiridol than
arrested cells (28, 48), so this approach has not allowed us to definitively attribute decreased sensitivity of these cells to the absence of
E2F-1. In the cell lines used in this study, the expression of ectopic
E2F-1 or the absence of endogenous E2F-1 did not affect cell cycle
patterns in exponentially growing cells. Therefore, it is likely that
E2F-1 is playing a primary role in flavopiridol-induced apoptosis
during S phase, and that the effects seen are not dependent on
E2F-1-mediated cell cycle progression or secondary to altered cell
cycle distribution in engineered cells.
It will next be important to determine the mechanism by which
E2F-1 induces an apoptotic response after flavopiridol-mediated cdk
7419
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2003 American Association for Cancer
Research.
FLAVOPIRIDOL-INDUCED APOPTOSIS IS E2F-1-DEPENDENT
inhibition during S phase. E2F-1 has been shown to induce apoptosis
in cells by both p53-dependent and p53-independent pathways (25,
49 –54). For example, when induced in growth factor-deprived
quiescent cells, E2F-1 promotes S phase entry and p53-dependent
apoptosis, in part via induction of p14ARF (55, 56). However, the
induction of apoptosis resulting from uninterrupted E2F-1 activity
beyond G1-S occurs in cells cultured in serum-containing medium and
is independent of p53 (14, 23). Whereas A549 and U2OS cells express
wild-type p53, NCI-H661 cells are hemizygous for p53 (57) and do
not have normal p53 function in response to DNA damage, making it
likely that the effects we are observing are independent of p53.
Mechanisms of p53-independent E2F-1-induced apoptosis include
pathways that are both dependent and independent of E2F-1 transcriptional transactivation activity. It will be of interest to determine
whether E2F-1 transcriptional targets such as p73 (58 – 60), caspase 3,
or Apaf-1 (61, 62) are essential for flavopiridol-mediated cytotoxicity
during S phase. However, flavopiridol-mediated inhibition of cyclin
T/cdk9 (pTEFb) results in a global decrease in efficient transcriptional
elongation (39), suggesting that increased E2F-1 transcriptional activity may be difficult to detect in flavopiridol-treated cells. Furthermore, an E2F-1 mutant lacking transcriptional transactivation activity,
E2F-1 (1–374), can induce apoptosis as efficiently as wild-type E2F-1
(51, 52), which may occur via intersection with nuclear factor ␬Bmediated survival pathways (63– 65). Finally, it has also been proposed that E2F-1 may also mediate down-regulation of Mcl-1 in
flavopiridol-treated cells, although down-regulation of Mcl-1 occurs
in E2F-1-deficient cells as well, presumably via pTEFb inhibition
(47). Full analysis of these issues will require direct measurement of
E2F-1-mediated transcriptional activity in the presence of flavopiridol
during S phase as well as an analysis of the effect of flavopiridol on
E2F-1 transcriptional targets. In addition, expression of E2F-1 mutants lacking either transactivation or DNA-binding activity, or the
ability to bind to nuclear factor ␬B during S phase may be instructive.
Our results in gemcitabine-treated cells contrast with those reported
in gastrointestinal cancer cell lines treated with sequential gemcitabine and flavopiridol (66). In these cells, both reduced levels of E2F-1
and phosphorylated Rb were noted, contributing to reduced E2F-1
activity and transcription of ribonucleotide reductase, which could
increase sensitivity to gemcitabine. Differences in the point of cell
cycle arrest induced by gemcitabine (G1-S versus mid-S) and differences in the times after combination treatment that proteins were
analyzed may account for some of the discrepancies. Furthermore,
because E2F-1-induced apoptosis in flavopiridol treated cells may not
require transcription, it is possible that decreased transcriptional activity mediated by cdk9 inhibition could result in decreased ribonucleotide reductase even whereas there is still sufficient E2F-1 activity
to induce apoptosis by other mechanisms.
To assess the contribution of cyclin A-dependent kinase inhibition
to the apoptotic response during S phase, we expressed an E2F-1
mutant containing a deletion that abrogates cyclin A-binding. When
expressed alone in NCI-H661 cells, this nonphosphorylatable mutant
provides a constitutive deregulated E2F-1 activity that itself induces
apoptosis. This is detected continuously during our experimental
manipulation, including synchronization with hydroxyurea, followed
by release into DMSO. Compared with cells engineered to express
wild-type E2F-1, in mutant-expressing cells the effect of flavopiridolmediated inhibition of cyclin A-dependent kinase activity on the
apoptotic response was diminished by several-fold.
Cyclin A-cdk2 has been proposed to be responsible for the majority
of E2F-1 phosphorylation in vivo, and complexes of E2F-1 with cyclin
A-cdk2 have been described both in vitro and in vivo (11). However,
E2F-1 is also phosphorylated by cyclin A-cdk1 at S375, which may
promote the formation of Rb-E2F-1 complexes, contributing to the
turn-off of E2F-1 activity late in the cell cycle (13). Consequently,
inhibition of cdk1 by flavopiridol may permit the persistence of E2F-1
free of Rb. Therefore, whereas our data indicate that the ability of
flavopiridol to modulate E2F-1 activity during S phase contributes to
apoptosis and strongly suggest that inhibition of cyclin A-dependent
kinase activity is involved, we have not yet definitively distinguished
whether cyclin A-cdk2, cyclin A-cdk1, or both are the critical flavopiridol targets implicated in E2F-1-induced apoptosis.
In addition, flavopiridol treatment during S phase does result in cell
death above that which is achieved by expression of the mutant that is
not phosphorylated by cyclin A-dependent kinase activity (Fig. 11). In
this regard, other cyclin-cdk holoenzymes have been reported to
contribute to E2F-1 phosphorylation during S phase. For example,
E2F-1 also interacts with the general transcription factor TFIIH multisubunit protein complex. The kinase activity associated with this
complex, including cyclin H, cdk7, and MAT-1, phosphorylates
E2F-1 at S408 and T433 (67). Mutation of these sites to alanine
greatly enhances the stability of E2F-1, preventing recognition by
proteins involved in the targeting of E2F-1 for degradation. Therefore,
the inhibition of cdk2, cdk1, and cdk7 by flavopiridol could all
contribute to the inappropriately persistent E2F-1 activity critical for
the apoptotic response during S phase. While the site of cdk2 phosphorylation has not been definitively mapped, E2F-1 has the consensus amino acid sequence (Ser-Pro-X-Lys/Arg) at S307 (Ser-Pro-GlyLys) for phosphorylation by cdk2 (12), a site distinct from those
phosphorylated by cdk1 and cdk7. Phosphospecific antibodies may
ultimately permit determination of which cdks are the critical targets
for the modulation of E2F-1 activity by flavopiridol.
The involvement of multiple cdks in the phosphorylation of E2F-1
may explain recent results in which siRNAs used to down-regulate
only cdk2 have not induced effects on cell cycle progression or
apoptosis (68). Other experiments have used ectopic expression of
p27Kip1 (69, 70), inhibitory peptides (16), or dominant-negative cdk2
mutants (71). These are all designed to inhibit cdk2, but also inhibit
other cdks either directly or indirectly, and may be more appropriate
models for the action of flavopiridol during S phase. As predicted,
these approaches have been associated with profound effects on S/G2
progression as well as with apoptosis in many cell types. Small
molecules that inhibit cdk2 and cdk1 (but not cdk4/6) are at various
stages of preclinical and clinical development (72). These drugs may
be E2F-1-metic even without prior S phase recruitment. Our data
support the assertion that such compounds may induce apoptosis in
transformed cells via E2F-1-dependent pathways.
ACKNOWLEDGMENTS
We thank Barrett Rollins and William Kaelin for critical reading of the
manuscript. Constructs and cell lines used in this work were provided by Peter
Adams, Meredith Irwin, and William Kaelin (Dana-Farber Cancer Institute);
Christiane Wirbelauer and Wilhelm Krek (Friedrich Miescher-Institut, Basel,
Switzerland); and Jacqueline Lees (Massachusetts Institute of Technology,
Cambridge, MA). We also thank Joyce LaVecchio, Maris Handley, and Kim
Lowe of the Dana-Farber flow cytometry facility for technical help.
REFERENCES
1. Weinberg, R. A. The retinoblastoma protein and cell cycle control. Cell, 81: 323–330,
1995.
2. Sherr, C. J. Mammalian G1 cyclins. Cell, 73: 1059 –1065, 1993.
3. Sherr, C. J., and Roberts, J. M. CDK inhibitors: positive and negative regulators of
G1-phase progression. Genes Dev., 13: 1501–1512, 1999.
4. Hall, M., and Peters, G. Genetic alterations of cyclins, cyclin-dependent kinases, and
cdk inhibitors in human cancer. Adv. Cancer Res., 68: 67–108, 1996.
5. Sherr, C. J. Cancer cell cycles. Science 274: 1672–1677, 1996.
6. Shapiro, G. I., and Harper, J. W. Anticancer drug targets: cell cycle and checkpoint
control. J. Clin. Investig., 104: 1645–1653, 1999.
7420
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2003 American Association for Cancer
Research.
FLAVOPIRIDOL-INDUCED APOPTOSIS IS E2F-1-DEPENDENT
7. Harbour, J. W., and Dean, D. C. The Rb/E2F pathway: expanding roles and emerging
paradigms. Genes Dev., 14: 2393–2409, 2000.
8. Dyson, N. The regulation of E2F by pRB-family proteins. Genes Dev., 12: 2245–
2262, 1998.
9. Krek, W., Ewen, M. E., Shirodkar, S., Arany, Z., Kaelin, W. G., Jr., and Livingston,
D. M. Negative regulation of the growth-promoting transcription factor E2F-1 by a
stably bound cyclin A-dependent protein kinase. Cell, 78: 161–172, 1994.
10. Dynlacht, B. D., Flores, O., Lees, J. A., and Harlow, E. Differential regulation of E2F
transactivation by cyclin-cdk2 complexes. Genes Dev., 8: 1772–1786, 1994.
11. Xu, M., Sheppard, K. A., Peng, C. Y., Yee, A. S., and Piwinica-Worms, H. Cyclin
A/Cdk2 binds directly to E2F-1 and inhibits the DNA-binding activity of E2F-1/DP-1
by phosphorylation. Mol. Cell. Biol., 14: 8420 – 8431, 1994.
12. Kitagawa, M., Higashi, H., Suzuki-Takahashi, I., Segawa, K., Hanks, S. K., Taya, Y.,
Nishimura, S., and Okuyama, A. Phosphorylation of E2F-1 by cyclin A-cdk2.
Oncogene, 10: 229 –236, 1995.
13. Peeper, D. S., Keblusek, P., Helin, K., Toebes, M., van der Eb, A. J., and Zantema,
A. Phosphorylation of a specific cdk site in E2F-1 affects its electrophoretic mobility
and promotes pRB-binding in vitro. Oncogene, 10: 39 – 48, 1995.
14. Krek, W., Xu, G., and Livingston, D. M. Cyclin A-kinase regulation of E2F-1 DNA
binding function underlies suppression of an S phase checkpoint. Cell, 83: 1149 –
1158, 1995.
15. Phillips, A. C., and Vousden, K. H. E2F-1 induced apoptosis. Apoptosis, 6: 173–182,
2001.
16. Chen, Y. N. P., Sharma, S. K., Ramsey, T. M., Jiang, L., Martin, M. S., Baker, K.,
Adams, P. D., Bair, K. W., and Kaelin, W. G. Selective killing of transformed cells
by cyclin/cyclin-dependent kinase 2 antagonists. Proc. Natl. Acad. Sci. USA, 96:
4325– 4329, 1999.
17. Senderowicz, A. M., and Sausville, E. A. Preclinical and clinical development of
cyclin-dependent kinase modulators. J. Natl. Cancer Inst. (Bethesda), 92: 376 –387,
2000.
18. Sedlacek, H. H., Czech, J., Naik, R., Kaur, G., Worland, P., Losiewicz, M., Parker,
B., Carlson, B., Smith, A., Senderowicz, A., and Sausville, E. A. Flavopiridol (L86
8275; NSC 649890), a new kinase inhibitor for tumor therapy. Int. J. Oncol., 9:
1143–1168, 1996.
19. Kelland, L. R. Flavopiridol, the first cyclin-dependent kinase inhibitor to enter the
clinic: current status. Exp. Opin. Investig. Drugs, 9: 2903–2911, 2000.
20. Sedlacek, H. H. Mechanisms of action of flavopiridol. Crit. Rev. Oncol. Hematol., 38:
139 –170, 2001.
21. Matranga, C. B., and Shapiro, G. I. Selective sensitization of transformed cells to
flavopiridol-induced apoptosis following recruitment to S phase. Cancer Res., 62:
1707–1717, 2002.
22. Pear, W. S., Nolan, G. P., Scott, M. L., and Baltimore, D. Production of high-titer
helper-free retroviruses by transient transfection. Proc. Natl. Acad. Sci. USA, 90:
8392– 8396, 1993.
23. Pruschy, M., Wirbelauer, C., Glanzmann, C., Bodis, S., and Krek, W. E2F-1 has
properties of a radiosensitizer and its regulation by cyclin A kinase is required for
cell survival of fibrosarcoma cells lacking p53. Cell Growth Differ., 10: 141–146,
1999.
24. Verona, R., Moberg, K., Estes, S., Starz, M., Vernon, J. P., and Lees, J. A. E2F
activity is regulated by cell cycle-dependent changes in subcellular localization. Mol.
Cell. Biol., 17: 7268 –7282, 1997.
25. Qin, X. Q., Livingston, D. M., Kaelin, W. G., and Adams, P. D. Deregulated
transcription factor E2F-1 expression leads to S phase entry and p53-mediated
apoptosis. Proc. Natl. Acad. Sci. USA, 91: 10918 –10922, 1994.
26. Field, S. J., Tsai, F. Y., Kuo, F., Zubiaga, A. M., Kaelin, J., W. G., Livingston, D. M.,
Orkin, S. H., and Greenberg, M. E. E2F-1 functions in mice to promote apoptosis and
suppress proliferation. Cell, 85: 549 –561, 1996.
27. Todaro, G. J., and Green, H. Quantitative studies of the growth of mouse embryo cells
in culture and their development into established lines. J. Cell Biol., 17: 299 –313,
1963.
28. Shapiro, G. I., Koestner, D. A., Matranga, C. B., and Rollins, B. J. Flavopiridol
induces cell cycle arrest and p53-independent apoptosis in non-small cell lung cancer
cell lines. Clin. Cancer Res., 5: 2925–2938, 1999.
29. Benotmane, A. M., Hoylaerts, M. F., Collen, D., and Belayew, A. Nonisotopic
quantitative analysis of protein-DNA interactions at equilibrium. Anal. Biochem.,
250: 181–185, 1997.
30. Janicke, R. U., Walker, P. A., Lin, X. Y., and Porter, A. G. Specific cleavage of the
retinoblastoma protein by an ICE-like protease in apoptosis. EMBO J., 15: 6969 –
6978, 1996.
31. Knudsen, E. S., and Wang, J. Y. Differential regulation of retinoblastoma protein
function by specific cdk phosphorylation sites. J. Biol. Chem., 271: 8313– 8320,
1996.
32. Kitagawa, M., Higashi, H., Jung, H. K., Suzuki-Takahashi, I., Ikeda, M., Tamai, K.,
Kato, J. Y., Segawa, K., Yoshida, E., Nishimura, S., and Taya, Y. The consensus
motif for phosphorylation by cyclin D1-cdk4 is different from that for phosphorylation by cyclin A/E-cdk2. EMBO J., 15: 7060 –7069, 1996.
33. Zarkowska, T., and Mittnacht, S. Differential phosphorylation of the retinoblastoma protein by G1/S cyclin-dependent kinases. J. Biol. Chem., 272: 12738 –
12746, 1997.
34. Connell-Crowley, L., Harper, J. W., and Goodrich, D. W. Cyclin D1/Cdk4 regulates
retinoblastoma protein-mediated cell cycle arrest by site-specific phosphorylation.
Mol. Biol. Cell, 8: 287–301, 1997.
35. Hirai, H., Roussel, M. F., Kato, J. Y., Ashmun, R. A., and Sherr, C. J. Novel INK4
proteins, p19 and p18, are specific inhibitors of cyclin D-dependent kinases CDK4
and CDK6. Mol. Cell. Biol., 15: 2672–2681, 1995.
36. Grimison, B., Langan, T. A., and Sclafani, R. A. p16INK4A tumor suppressor
function in lung cancer cells involves cyclin-dependent kinase 2 inhibition by Cip/Kip
protein redistribution. Cell Growth Differ., 11: 507–515, 2000.
37. Sheaff, R. J., Groudine, M., Gordon, M., Roberts, J. M., and Clurman, B. E. Cyclin
E-cdk2 is a regulator of p27Kip1. Genes Dev., 11: 1464 –1478, 1997.
38. Vlach, J., Hennecke, S., and Amati, B. Phosphorylation-dependent degradation of the
cyclin-dependent kinase inhibitor p27. EMBO J., 16: 5334 –5344, 1997.
39. Chao, S. H., and Price, D. H. Flavopiridol inactivates p-TEFb and blocks most RNA
polymerase II transcription in vivo. J. Biol. Chem., 276: 31793–31799, 2001.
40. Lam, L. T., Pickeral, O. K., Peng, A. C., Rosenwald, A., Hurt, E. M., Giltnane,
J. M., Averett, L. M., Zhao, H., Davis, R. E., Sathyamoorthy, M., Wahl, L. M.,
Harris, E. D., Mikovits, J. A., Monks, A. P., Hollingshead, M. G., Sausville, E. A.,
and Staudt, L. M. Genomic-scale measurement of mRNA turnover and the
mechanisms of action of the anti-cancer drug flavopiridol. Genome Biol., 2:
research0041. Epub, 2001.
41. Medema, R. H., Kops, G. J. P. L., Bos, J. L., and Burgering, B. M. T. AFX-like
forkhead transcription factors mediate cell-cycle regulation by Ras and PKB through
p27Kip1. Nature (Lond.), 404: 782–787, 2000.
42. Cartee, L., Wang, Z., Decker, R. H., Chellappan, S. P., Fusaro, G., Hirsch, K. G.,
Sankala, H. M., Dent, P., and Grant, S. The cyclin dependent kinase inhibitor (CDKI)
flavopiridol disrupts phorbol 12-myristate 13-acetate-induced differentiation and
CDKI expression while enhancing apoptosis in human myeloid leukemia cells.
Cancer Res., 61: 2583–2591, 2001.
43. Motwani, M., Jung, C., Sirotnak, F. M., She, Y., Shah, M. A., Gonen, M., and
Schwartz, G. K. Augmentation of apoptosis and tumor regression by flavopiridol in
the presence of CPT-11 in HCT116 colon cancer monolayers and xenografts. Clin.
Cancer Res., 7: 4209 – 4219, 2001.
44. Gervais, J. L. M., Seth, P., and Zhang, H. Cleavage of cdk inhibitor p21Cip1/Waf1 by
caspases is an early event during DNA damage-induced apoptosis. J. Biol. Chem.,
273: 19207–19212, 1998.
45. Levkau, B., Koyama, H., Raines, E. W., Clurman, B. E., Herren, B., Orth, K.,
Roberts, J. M., and Ross, R. Cleavage of p21Cip1/Waf1 and p27Kip1 mediates apoptosis
in endothelial cells through activation of cdk2: role of a caspase cascade. Mol. Cell,
1: 553–563, 1998.
46. Zhang, Y., Fujita, N., and Tsuruo, T. Caspase-mediated cleavage of p21Waf1/Cip1
converts cancer cells from growth arrest to undergoing apoptosis. Oncogene, 18:
1131–1138, 1999.
47. Ma, Y., Cress, D., and Haura, E. B. Flavopiridol-induced apoptosis is mediated
through up-regulation of E2F-1 and repression of Mcl-1. Mol. Cancer Ther., 2:
73– 81, 2003.
48. Brüsselbach, S., Nettelbeck, D. M., Sedlacek, H. H., and Muller, R. Cell cycleindependent induction of apoptosis by the anti-tumor drug flavopiridol in endothelial
cells. Int. J. Cancer, 77: 146 –152, 1998.
49. Wu, X., and Levine, A. J. p53 and E2F-1 cooperate to mediate apoptosis. Proc. Natl.
Acad. Sci. USA, 91: 3602–3606, 1994.
50. Shan, B., and Lee, W. H. Deregulated expression of E2F-1 induces S phase entry and
leads to apoptosis. Mol. Cell. Biol., 14: 8166 – 8173, 1994.
51. Phillips, A. C., Bates, S., Ryan, K. M., Helin, K., and Vousden, K. H. Induction of
DNA synthesis and apoptosis are separable functions of E2F-1. Genes Dev., 11:
846 –1666, 1997.
52. Hseih, J. K., Frdersdorf, S., Kouzarides, T., Martin, K., and Lu, X. E2F-1-induced
apoptosis requires DNA binding but not transactivation and is inhibited by the
retinoblastoma protein through direct interaction. Genes Dev., 11: 1840 –1852,
1997.
53. Hunt, K. K., Deng, J., Liu, T. J., Wilson-Heiner, M., Swisher, S. G., Clayman, G., and
Hung, M. C. Adenovirus-mediated overexpression of the transcription factor E2F-1
induces apoptosis in human breast and ovarian carcinoma cell lines and does not
require p53. Cancer Res., 57: 4722– 4726, 1997.
54. Holmberg, C., Helin, K., Sehested, M., and Karlstrom, O. E2F-1-induced p53independent apoptosis in transgenic mice. Oncogene, 17: 143–155, 1998.
55. Bates, S., Phillips, A. C., Clark, P. A., Stott, F., Peters, G., Ludwig, R. L., and
Vousden, K. H. p14 ARF links the tumour suppressors RB and p53. Nature (Lond.),
395: 124 –125, 1998.
56. Sherr, C. Tumor surveillance via the ARF-p53 pathway. Genes Dev., 12: 2984 –2991,
1998.
57. Mitsudomi, T., Steinberg, S. M., Nau, M. M., Carbone, D., D’Amico, D., Bodner, S.,
Oie, H. K., Linnoila, R. I., Mulshine, J. L., Minna, J. D., and Gazdar, A. F. p53 gene
mutations in non-small cell lung cancer cell lines and their correlation with the
presence of ras mutations and clinical features. Oncogene, 7: 171–180, 1992.
58. Lissy, N. A., Davis, P. K., Irwin, M., Kaelin, W. G., and Dowdy, S. F. A common
E2F-1 and p73 pathway mediates cell death induced by TCR activation. Nature
(Lond.), 407: 642– 645, 2000.
59. Irwin, M., Marin, M. C., Phillips, A. C., Seelan, R. S., Smith, D. I., Liu, W., Flores,
E. R., Tsai, K. Y., Jacks, T., Vousden, K., and Kaelin, W. G. Role for the p53
homologue p73 in E2F-1 induced apoptosis. Nature (Lond.), 407: 645– 648, 2000.
60. Stiew, T., and Putzer, B. M. Role of the p53-homologue p73 in E2F-1-induced
apoptosis. Nat. Genet., 26: 464 – 469, 2000.
61. Moroni, M. C., Hickman, E. S., Denchi, E. L., Caprara, G., Colli, E., Cecconi, F.,
Muller, H., and Helin, K. Apaf-1 is a transcriptional target for E2F and p53. Nat. Cell
Biol., 3: 552–558, 2001.
62. Muller, H., Bracken, A. P., Vernell, R., Moroni, M. C., Christians, F., Grassilli, E.,
Prosperini, E., Vigo, E., Oliner, J. D., and Helin, K. E2Fs regulate the expression of
genes involved in differentiation, development, proliferation and apoptosis. Genes
Dev., 15: 267–285, 2001.
63. Phillips, A., Ernst, M., Bates, S., Rice, N., and Vousden, K. E2F-1 potentiates cell
death by blocking antiapoptotic signaling pathways. Mol. Cell, 4: 771–781, 1999.
7421
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2003 American Association for Cancer
Research.
FLAVOPIRIDOL-INDUCED APOPTOSIS IS E2F-1-DEPENDENT
64. Tanaka, H., Matsumura, I., Ezoe, S., Satoh, Y., Sakamaki, T., Albanese, C., Machii,
T., Pestell, R. G., and Kanakura, Y. E2F1 and c-Myc potentiate apoptosis through
inhibition of NF-␬B activity that facilitates MnSOD-mediated ROS elimination. Mol.
Cell, 9: 1017–1029, 2002.
65. Kim, D. M., Koo, S. Y., Jeon, K., Kim, M. H., Lee, J., Hong, C. Y., and Jeong, S.
Rapid induction of apoptosis by combination of flavopiridol and tumor necrosis factor
(TNF)- or TNF-related apoptosis-inducing ligand in human cancer cell lines. Cancer
Res., 63: 549 –736, 2003.
66. Jung, C. P., Motwani, M. V., and Schwartz, G. K. Flavopiridol increases sensitization
to gemcitabine in human gastrointestinal cancer cell lines and correlates with downregulation of the ribonucleotide reductase M2 subunit. Clin. Cancer Res., 7: 2527–
2536, 2001.
67. Vandel, L., and Kouzarides, T. Residues phosphorylated by TFIIH are required for
E2F-1 degradation during S-phase. EMBO J., 18: 4280 – 4291, 1999.
68. Tetsu, O., and McCormick, F. Proliferation of cancer cells despite cdk2 inhibition.
Cancer Cell, 3: 233–245, 2003.
69. Katayose, Y., Kim, M., Rakkar, A. N. S., Li, Z., Cowan, K. H., and Seth, P.
Promoting apoptosis: a novel activity associated with the cyclin-dependent kinase
inhibitor p27. Cancer Res., 57: 5441–5445, 1997.
70. Wang, X., Gorospe, M., Huang, Y., and Holbrook, N. J. p27Kip1 overexpression
causes apoptotic death of mammalian cells. Oncogene, 1997.
71. Hu, B., Mitra, J., van den Heuvel, S., and Enders, G. H. S and G2 roles for cdk2
revealed by inducible expression of a dominant-negative mutant in human cells. Mol.
Cell. Biol., 21: 2755–2766, 2001.
72. Shapiro, G. I. Small molecule inhibitors of cyclin-dependent kinases. In: M.
Blagosklonny (ed.), Cell Cycle Checkpoints and Cancer, pp. 208 –234. Georgetown,
TX: Landes Bioscience, 2001.
7422
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2003 American Association for Cancer
Research.
Flavopiridol-Induced Apoptosis during S Phase Requires
E2F-1 and Inhibition of Cyclin A-Dependent Kinase Activity
Jingrui Jiang, Christian B. Matranga, Dongpo Cai, et al.
Cancer Res 2003;63:7410-7422.
Updated version
Cited articles
Citing articles
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/63/21/7410
This article cites 68 articles, 37 of which you can access for free at:
http://cancerres.aacrjournals.org/content/63/21/7410.full#ref-list-1
This article has been cited by 13 HighWire-hosted articles. Access the articles at:
http://cancerres.aacrjournals.org/content/63/21/7410.full#related-urls
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 15, 2017. © 2003 American Association for Cancer
Research.