DNA-BINDING SPECIFICITY OF THE E2F GENE FAMILY,
AND CLONING AND CHARACTERIZATION
OF A NOVEL FAMILY MEMBER
by
Brian A. Fairchild
B.S., Cellular and Molecular Biology
University of Michigan, 1992
Submitted to the Department of Biology
in partial fulfillment of the Requirements for the Degree of
Doctor of Philosophy in Biology
at the
ASSACHUSETTS INSTITUTE
OFTECHNO 0
Massachusetts Institute of Technology
~R
February 1999
1 999
.1
© 1999 Massachusetts Institute of Technology
All rights reserved
Signature
of A uthor
Certified
by
Accepted
by
...... .
....
..................................................
Brian A. Fairchild
Department of Biology
January 7, 1999
....................................
............
Jacqueline A. Lees
Assistant Professor of Biology
Thesis Supervisor
.
. .
.... ..
Orr-Weaver
J ....... .............................................Tenry
Professor of Biology
Co-Chair, Department Committee on Graduate Students
For Susan, on whose love and support I rely daily.
DNA-BINDING SPECIFICITY OF THE E2F GENE FAMILY,
AND CLONING AND CHARACTERIZATION
OF A NOVEL FAMILY MEMBER
by
Brian A. Fairchild
Submitted to the Department of Biology
on January 7, 1999 in partial fulfillment of the
requirements for the degree of Doctor of Philosophy in
Biology
ABSTRACT
The E2F transcription factors have been implicated in the regulation of cell cycle
progression and apoptosis. The E2F proteins are differentially regulated, and have distinct
functions in addition to any shared ones. This work describes the identification of a novel E2F
protein, E2F-6. Unlike the previously characterized E2F proteins, E2F-6 lacks a transcriptional
activation domain and does not associate with the tumor suppressor pRB or its homologs p107 or
p130. E2F-6 must therefore function in ways different from those of the previously identified E2F
proteins. The identification of FIP- 1, a novel zinc-finger protein which specifically interacts with
E2F-6, is also described.
The distinct functions of the E2F proteins are presumed to result from the differential
regulation of E2F-responsive target genes by individual E2F proteins. This work explores the
hypothesis that this specificity of target gene regulation could be mediated by differences in the
preferred DNA binding sequences. A PCR-based selection to determine the preferred DNA
binding sequences for different E2F complexes identified substantially similar sequences for each
E2F complex tested. Furthermore, these sequences were able to bind to each complex with similar
affinities. The sequences also showed very little differential specificity when used in
transcriptional activation assays with different E2F proteins. Specificity in target gene activation
by particular E2F proteins is therefore more likely to be mediated by differences in promoter
context than by differences in the sequences of the E2F sites.
Thesis Supervisor: Jackie Lees
Title: Assistant Professor of Biology
TABLE OF CONTENTS
....................
ract............................................................3
Abstract
Chapter I: Introduction
......... .. 7
O verview .........................................................................
B. Adenovirus: a tool to identify mechanisms of cellular proliferation control ........... 8
9
B l. El A deregulates cellular proliferation ............................................
A.
B2. p300 and CBP are bound, and possibly inactivated, by ElA................12
C. ElA inactivates the tumor and apoptosis suppressor pRB.............................15
15
Cl. RB is frequently inactivated in human tumors .................................
16
RB participates in restriction point control ....................................
C3. Cyclin-dependent kinases hyperphosphorylate pRB at the restriction point... 17
C2.
RB suppresses apoptosis.........................................................23
C5. RB inactivation is insufficient to promote transformation ...................
C6. ElA also binds the pRB homologs p107 and p130 ...........................
C4.
D.
25
C7. pRB suppresses growth through its "large pocket".................... 27
E2F function is antagonized by pRB, p107, and p130..................................29
30
DI. The E2F and DP gene families.................................................
D2.
E.
24
pRB-mediated
repression.........................................................31
D3. p107 and p130 also bind E2F and can repress transcription ................
E2F promotes proliferation and apoptosis ...............................................
33
35
E2.
E2F promotes progression through GI........................................35
37
E2F promotes transformation and apoptosis ..................................
E3.
E2F as an oncogene, and as a tumor suppressor ..............................
39
F. E2F genes: specificity in function and in regulation ....................................
41
El.
Individual E2F genes have specific functions.................................41
F2. The E2F genes are regulated in overlapping but distinct ways..............43
47
E2F genes: transcriptional targets........................................................
F1.
G.
G1. Some E2F targets are regulated by repression, some by activation ........
49
G2. Individual E2F genes may regulate specific subsets
of transcriptional targets .......................................................
H.
I.
50
S u m m ary ......................................................................................
51
R eferences ........................................................................................
52
Chapter II: The individual E2F family members bind a common
DNA recognition sequence
. .. 69
A . A b stract .....................................................................................
70
Introduction ....................................................................................
B.
..... 74
...
C. Results .........................................................................
74
C1. Production of recombinant proteins............................................
C2. E2F site selection..................................................................77
C3. Relative affinities of E2F complexes for selected sites.......................
C4. Sequence variations do not confer E2F complex specificity
in transactivation................................................................85
D.
E.
F.
G.
D iscu ssion .....................................................................................
Materials and Methods......................................................................
Acknowledgements ..........................................................................
References .....................................................................................
85
90
94
97
98
Chapter III: E2F-6, a member of the E2F family that can behave as a
transcriptional repressor
2
A . A b stract.........................................................................................10
10 3
B . Introdu ction ....................................................................................
10 6
C . R esu lts ..........................................................................................
C1. Isolation of cDNAs encoding an E2F family member..........................106
E2F-6 is widely expressed in vivo................................................107
C3. E2F-6 displays low-affinity E2F DNA-binding activity ....................... 113
C4. E2F-6 is unable to activate the transcription of E2F-responsive genes.......116
120
D iscu ssio n ......................................................................................
123
Materials and Methods ........................................................................
126
Acknowledgements ...........................................................................
127
R eferences .....................................................................................
C2.
D.
E.
F.
G.
Chapter IV: FIP-1, a zinc-finger protein, interacts specifically with E2F-6
A . Ab stract.........................................................................................13
B . Introduction ....................................................................................
C.
0
13 1
13 5
R e su lts ..........................................................................................
C1. Isolation of FIP-l...................................................................135
140
C2. Expression pattern of FIP-1 .......................................................
C3. Identification of E2F-6 regions interacting with FIP-1.........................143
D.
E.
C4. Specific interaction of FIP-1 with E2F-6 ........................................
D iscu ssio n ......................................................................................
Materials and Methods ........................................................................
146
14 9
152
F.
Acknowledgements ...........................................................................
154
G.
R eferences .....................................................................................
15 5
Chapter V: Conclusion
A.
161
Cloning and characterization of novel E2F genes .........................................
Al. What are the biological functions of E2F-6?.....................................161
A2.
Additional E2F genes must exist..................................................164
B. Toward an understanding of the mechanisms of E2F target specificity.................166
C. Broadening our understanding of E2F: tumor genetics and genetic screens ........... 170
E2F and cancer genetics............................................................170
C2. Genetic screens could uncover addditional E2F functions or regulators ..... 173
C3. Genetics, molecular biology and biochemistry in the study of E2F .......... 174
Cl.
D.
R eferen ces......................................................................................17
Acknowledgements .....................................................................................
6
178
CHAPTER I
Introduction
A.
Overview
The transcription factor E2F confers a GI/S-specific pattern of transcription on a variety of
cellular genes. This is accomplished in part through the association of E2F with members of the
pRB family of transcriptional repressors during early in the G1 phase of the cell cycle. This
association may be critical to the well-documented tumor-suppressive and apoptosis-suppressive
activities of pRB. Similarly, the adenoviral protein ElA dissociates E2F from pRB family
members, and this function of the ElA protein contributes to its transforming and pro-apoptotic
activities. This chapter discusses the biological and biochemical activities of E A, pRB, and E2F,
with special attention to the genetic and biochemical heterogeneity of the E2F transcription factor.
7
B.
Adenovirus: a tool to identify mechanisms of cellular proliferation control
The regulation of cell division and of programmed cell death is critical to the growth,
development and reproduction of multicellular organisms. Deregulated cell growth and division is
the hallmark and perhaps the definition of cancer. Cancer cells also demonstrate distinct changes in
their susceptibility to programmed cell death. The proper regulation of these processes is therefore
a study of enormous medical interest. Some viruses induce cells to enter the DNA synthesis phase
(S phase) of the cell cycle, which promotes replication of the viral DNA. In addition to promoting
re-entry of the cell cycle and, hence, cellular proliferation, these viruses also manipulate the
regulation of apoptotic control pathways to prevent the premature death of an infected cell. Viruses
have therefore been used extensively by cancer biologists as tools to identify key components of
the cellular regulatory pathways controlling proliferation and apoptosis. One such virus is the
adenovirus.
The adenoviruses are a family of double-stranded DNA viruses that infect humans and
other animals. They have been the subjects of intensive study, in part because of their role in
human disease: of the more than forty human-specific adenoviruses described so far, some are
associated with respiratory infections, others with ocular disease and still others with
gastrointestinal infections (see Lucher, 1995 for a review). Of particular interest to cancer
biologists, however, is the observation that adenoviral infection can induce cellular transformation.
They are also known to evade mechanisms inducing programmed cell death (apoptosis) (reviewed
in White, 1995). The mechanisms through which adenoviruses override the normal regulation of
cellular growth, proliferation, and death are being carefully investigated, primarily as a tool to
improve our understanding of those normal regulatory mechanisms and how tumor cells might
escape them.
Adenoviruses can infect either dividing or quiescent cells. Upon infection by adenovirus,
quiescent cells enter the DNA synthesis phase (S phase) of the cell cycle, with a concomitant
induction of the expression of dihydrofolate reductase (DHFR), thymidine kinase, and other genes
associated with cellular DNA replication. This induction of reentry of quiescent cells into the cell
8
cycle is believed to be mediated by the products of the adenoviral gene first expressed upon
infection, the ElA gene.
B1.
EtA deregulates cellular proliferation
The adenoviral El A gene is expressed as at least five different alternatively spliced
messenger RNAs (mRNAs), of which the two most abundant are the two largest, known as 12S
and 13S (reviewed in Zantema and van der Eb, 1995). The 12S and 13S mRNAs of adenovirus
type 5 (Ad5), one of the most studied species of adenovirus, encode proteins of 243 and 289
amino acids, respectively. The ElA proteins from the various species of adenovirus are
homologous to each other. This homology is concentrated in three conserved domains, referred to
as CR1, CR2, and CR3. The 12S mRNA encodes both CR1 and CR2 as well as intervening and
flanking sequences; the 13S mRNA encodes a nearly identical gene product, differing only by the
additional presence of CR3.
ElA drives proliferationand apoptosis
Adenoviral strains bearing small deletions in their genome have been generated to study the
roles of particular genes in the life cycle of the virus. The E2A and E2B genes, for example, are
required for the replication of the viral genome, whereas the EIB gene products are required to
prevent apoptosis of the host cell during infection (reviewed in Philipson, 1995). Viruses that lack
the El A region are deficient in multiple biological activities. The El A gene is the first viral gene to
be expressed after infection; expression of the 13S mRNA is essential for efficient transcription of
the remaining viral genes and, therefore, for the execution of the remainder of the viral life cycle.
Internal deletions within the ElA gene have revealed that this function requires CR3, consistent
with the ability of the 13S but not the 12S splice forms to mediate this transactivation (reviewed in
Jones, 1995). As mentioned above, adenoviruses that lack a functional EIB region induce
apoptosis upon infection; this apoptosis is induced by the ElA gene products (White, 1995, and
references therein).
9
The El A gene products are also involved in the stimulation of cell cycle progression of the
host cell. Whereas normal viruses will rapidly induce the host cell to enter S phase, viruses that
lack the EIA region are unable to do so (Howe et al., 1990). This does not seem to result from the
requirement for this region in the transactivation of other viral genes. CR3, which is necessary and
sufficient to induce the expression of the viral genes, is not required for S phase induction: the 12S
mRNA species, which lacks CR3, is competent to promote the reentry of the cell into the cell cycle
(Lillie et al., 1987; Zerler et al., 1987; Bellett et al., 1989; Howe et al., 1990), and expression of
EIA by plasmid microinjection is sufficient to induce S phase in the absence of viral infection
(Lillie et al., 1987).
In addition to promoting re-entry of the host cell into the cell cycle, ElA can also promote
the transformation of primary cells. Adenoviral infection of cells that do not support the complete
viral life cycle can lead to cellular transformation (reviewed in Lucher, 1995); transformation is
rarely observed in other cell types, presumably as a result of cellular lysis at the end of the infection
cycle. A combination of deletion studies and transfection experiments have determined that the
ElA gene is capable of mediating cellular transformation (reviewed in Williams et al., 1995).
Expression of EIA in the absence of viral infection causes focus formation in primary baby rat
kidney (BRK) cells, although the foci degenerate as a result of ElA-induced apoptosis (White,
1995, and references therein). Coexpression of ElA with another gene that prevents ElA-induced
apoptosis, such as E IB, bcl-2, or activated ras, is sufficient to induce complete and stable
transformation of primary cells (Whyte et al., 1989; White, 1995).
10
The effects of EJA are mediated by specific domains of the protein
Two distinct regions of the El A protein contribute to its ability to promote cell
proliferation. The first is the amino-terminal region of the protein, including CR1 and a nonconserved region at the extreme amino-terminus, and the second is CR2. Mutation of either of
these regions impairs the ability of ElA to induce DNA synthesis in most assays, and mutation of
both regions completely abrogates S phase induction by ElA (Lillie et al., 1987; Zerler et al.,
1987; Bellett et al., 1989; Howe et al., 1990). The same two regions are required for the
transforming activity of ElA (Lillie et al., 1986; Moran et al., 1986; Lillie et al., 1987; Whyte et
al., 1988b). The ability of ElA to induce S phase and its ability to induce cellular transformation
are therefore genetically inseparable. Interestingly, the same regions are also required for ElAinduced apoptosis (Samuelson and Lowe, 1997).
The EIA proteins do not seem to possess an intrinsic DNA-binding activity. It has
therefore long been assumed that the cellular effects of El A expression in the absence of viral
infection are mediated by protein-protein interactions. Consequently, E lA-associated proteins
have been intensively studied in an attempt to determine the mechanisms through which ElA
promotes cellular transformation and apoptosis. Immunoprecipitation of ElA from an adenovirustransformed cell line revealed that a number of proteins can be found in a stable complex with EIA;
the most abundant of these have molecular weights of approximately 105, 107 and 300 kD (Yee
and Branton, 1985; Harlow et al., 1986). These proteins are cellular in origin: they are also
present in a variety of cells that have not been exposed to adenoviral infection (Harlow et al.,
1986). The 300 kD protein, known as p300, requires both CR1 and the amino-terminus of ElA
for efficient binding to that protein (Whyte et al., 1989; Eckner et al., 1994). Efficient binding of
ElA to p107 requires the presence of CR2; maximal binding to p105 also requires part of CR1
(Whyte et al., 1989). Given that these same regions of ElA are known to mediate its effects on
cellular proliferation and apoptosis, much study has been focused on p300, p105 and p107 as
potential effectors of ElA function.
1 1
B2.
p300 and CBP are bound, and possibly inactivated, by ElA
The p300 gene was identified by screening an expression library with an antiserum to p300
protein purified through its association through ElA (Eckner et al., 1994). p300 binds to the
amino-terminus of E l A; mutations in CR1 or in the extreme amino-terminus of E lA are capable of
disrupting the interaction between ElA and p300, which would be consistent with a role for p300
in ElA-induced cell proliferation or apoptosis. The p300 protein is homologous to the CREBbinding protein CBP (Eckner et al., 1994; Lundblad et al., 1995). This homology includes the
region used by p300 to bind to ElA, and in vitro studies have found that CBP and p300 bind to
ElA with similar affinities (Lundblad et al., 1995). Thus, both p300 and CBP are bound by ElA
and one or both may participate in the deregulation of cellular proliferation by ElA.
Like CBP, p300 also binds to many other transcription factors in the cell, suggesting a role
as a transcriptional adaptor. Both have been reported to bind a wide variety of transcripton factors
and to have histone acetyltransferase (HAT) activity (Giles et al., 1998), indicating that these
proteins probably also have a role in chromatin remodeling at promoters that are being actively
transcribed. Studies on ElA-mediated repression of the SV40 enhancer suggest that p300 and/or
CBP may have a required role in the activation of some promoters and that ElA may sequester or
otherwise inactivate p300 and CBP, preventing this activation. In addition, they are consistent
with p300 and/or CBP mediation of at least some of the in vivo effects of ElA.
p300, CBP, and cellularproliferation
The regions of EIA required for binding to p300 and CBP also mediate EIA's promotion
of cellular proliferation and apoptosis. In fact, a point mutation abrogating the ability of ElA to
bind to p300 in a co-immunoprecipitation assay, but not its ability to bind p105 or p107, also
abrogates the transforming activity of E1A(Whyte et al., 1988b; Whyte et al., 1989).
Furthermore, no El A mutation has been described in which the transforming activity is retained
but p300-binding is not, and all tested mutations in ElA that retain p300-binding activity also retain
12
the ability to promote S phase entry and apoptosis. Nevertheless, no direct evidence currently
indicates a role for either p300 or CBP in either the promotion of S phase or of apoptosis.
Studies of the effects of EIA on p300-mediated transcriptional modulation suggest that
E l A neutralizes p300 function, at least in some contexts, perhaps through sequestration of the
protein (see above). If inactivation of p300 or CBP by ElA does indeed promote cellular
transformation, their inactivation might also be expected at some frequency in naturally occuring
human tumors. To date, there has also been one report of somatic mutation in p300 in a colorectal
carcinoma and a gastric carcinoma; these are associated with loss of heterozygosity at the locus,
suggesting that the mutations may be more likely to be hypomorphic than neomorphic (Muraoka et
al., 1996). Patients with Rubinstein-Taybi syndrome, a disease caused by inheritance of an
inactivated allele of the CBP gene, have been reported to bear a mildly increased predisposition to
certain tumor types, particularly to certain tumors of the nervous system (Miller and Rubinstein,
1995). CBP may, therefore, act as a tumor suppressor in some cellular contexts, although no
studies have yet addressed the fate of the second CBP allele in Rubinstein-Taybi tumors.
In addition, both p300 and CBP have been found in translocations with MLL, a homolog
of the DrosophilaTrithoraxgene, in leukemia patients (reviewed in Giles et al., 1998). CBP has
also been found in translocations with MOZ, a putative acetyltransferase, in patients with acute
myelocytic leukemia (Borrow et al., 1996). The fusion products of these translocations are
believed to be expressed, and are believed to be tumorigenic as a result of a gain-of-function, in
contrast to the hypomorphic mutations observed in Rubenstein-Taybi and the presumably
hypomorphic mutations in the two carcinomas mentioned above, suggesting that the mechanism
for their tumorigenicity may be unrelated to the mechanism(s) in the loss-of-function scenarios.
In summary, the sequestration of p300 or CBP is required for some of the transcriptional
effects of E lA. Sequestration or other modification of their activity might also play a role in the
deregulation of cell proliferation or apoptosis by ElA, although no direct evidence supports this
hypothesis. The alterations in p300 and CBP in some human tumors does suggest that these
proteins may participate in the regulation of cellular proliferation in some contexts. As
13
transcriptional adaptors, however, both p300 and CBP have been reported to bind a wide variety
of transcription factors. p300 and CBP are therefore potentially implicated in an immense number
of signaling pathways. Perhaps only the discovery of a tumorigenic mutation specifically
affecting only a limited subset of these pathways will permit cancer biologists to identify which
functions of p300 and CBP are relevant to proliferation control and most urgently deserving of
further study.
14
C.
ElA inactivates the tumor and apoptosis suppressor pRB
Like p300, the 105 kD ElA-associated protein, p105, binds to regions of ElA that mediate
its effects on cellular proliferation and apoptosis; in the case of p105, optimal binding to ElA
requires both CR1 and CR2. A combination of immunological methods and protease digestion
experiments revealed that p105 is identical to pRB, the product of the retinoblastoma-susceptibility
gene, RB (Whyte et al., 1988a).
C1.
RB is frequently inactivated in human tumors
The RB gene was originally identified based on its frequent mutation, often homozygous
deletion, in human retinoblastomas (Friend et al., 1986; Friend et al., 1987; Lee et al., 1987).
Predisposition to retinoblastomas is inherited as an autosomal dominant trait, and usually results in
in bilateral retinoblastomas, whereas unilateral retinoblastomas are more often the result of sporadic
mutations (Knudson, 1971). Statistical analyses on patients with unilateral or bilateral
retinoblastomas suggested a requirement for two mutational events with similar mutation rates in
the genesis of retinoblastomas (Knudson, 1971). The frequent loss of heterozygosity observed in
retinoblastomas on the long arm of chromosome 13 in the vicinity of the retinoblastoma
susceptibility locus is consistent with a tumor-suppressive function for this locus (Cavenee et al.,
1983; Dryja et al., 1984). Although gross genomic deletions of RB, the retinoblastomasusceptibility gene, are found in only about 30% of retinoblastoma cell lines studied (Friend et al.,
1986; Friend et al., 1987; Lee et al., 1987), gross alterations in the transcript are seen in almost all
retinoblastoma cells (Friend et al., 1987; Lee et al., 1987) . In one study the protein was
undetectable in all eighteen retinoblastoma cultures tested, including thirteen short-term cultures
(Horowitz et al., 1990). This locus is believed to be inactivated in all human retinoblastomas, and
this inactivation is rate-limiting for tumor formation.
Humans who inherit a one defective (and one functional) copy of the RB gene develop
retinoblastomas with approximately 95% penetrance, and most of these retinoblastomas are
bilateral. Patients who survive retinoblastoma often develop other tumors, including sarcomas
15
such as osteosarcomas. Indeed, mutations of the RB gene have been detected in primary tumors
and/or tumor-derived cell lines from osteosarcomas, synovial sarcomas, small cell lung
carcinomas, breast carcinomas, prostate carcinomas, cervical carcinomas and other tumors in
addition to retinoblastomas (Harbour et al., 1988; Lee et al., 1988; T'Ang et al., 1988; Yokota et
al., 1988; Bookstein et al., 1989; Varley et al., 1989; Horowitz et al., 1990). These mutations are
often biallelic deletions, although missense and nonsense mutations have also been described.
Thus, the RB gene clearly plays an important role in tumor suppression in a wide variety of human
cell types.
A role for RB in the prevention of tumorigenesis is also supported by mouse studies. Mice
that inherit an inactivated allele of RB develop pituitary tumors, and less frequently cancer of the
thyroid (Jacks et al., 1992; Williams et al., 1994). The cause(s) of the differences in the tissue
specificity of tumorigenesis between mice and humans in the case of RB mutation is (are) not
known; nevertheless, the RB gene does play an essential role in the prevention of tumorigenesis in
both species.
C2.
RB participates in restriction point control
The commitment of a mammalian cell to replicate or to remain quiescent is usually
determined during the GI phase of the cell cycle. During early- to mid-GI the growth of
mammalian cells in tissue culture is dependent on the continuous presence of growth factors,
usually provided by supplementation of the growth medium with serum; removal of the serum
during this period of GI induces the cells to enter a quiescent state (reviewed in Pardee, 1989).
Later, during late G1, S or G2, progression through the cell cycle becomes serum-independent.
The point at which the cell switches from serum-dependence to serum-independence is termed the
restriction point (Pardee, 1989). The period of serum-dependence is also characterized by a
marked sensitivity to the protein synthesis inhibitor cycloheximide: fibroblasts treated with
cycloheximide will arrest in early G 1, but once they have passed the restriction point they become
much less sensitive to cycloheximide levels (reviewed in Pardee, 1989).
16
The RB gene contributes to control over passage through the restriction point. Mouse
embryonic fibroblasts (MEFs) in which both alleles of the RB gene have been mutated lose some,
but not all, aspects of restriction point control. RB-deficient MEFs are smaller than wild-type
MEFs and spend less time in GI (Herrera et al., 1996; Lukas et al., 1995). RB-deficient MEFs
lack the extreme sensitivity of wild-type MEFs to cycloheximide during early- and mid-GI
(Herrera et al., 1996). Serum-dependence is also markedly reduced in RB-deficient MEFs (Lukas
et al., 1995; Brugarolas et al., 1998), although not eliminated: complete serum withdrawal before
late GI will still arrest RB-deficient MEFs (Herrera et al., 1996). Given that RB-deficient MEFs
retain some degree of serum-dependence, loss of RB does not abrogate all forms of restriction
point control, but RB is nevertheless required for complete enforcement of the checkpoint.
RB mutation also leads to a relaxation of control over the GUS transition in some tissues of
the developing mouse. Specifically, aberrant S phase entry is observed in RB-deficient mice in the
lens and both the central and peripheral nervous systems (Morgenbesser et al., 1994; Macleod et
al., 1996). Inactivation of the RB gene is therefore sufficient to permit inappropriate entry into the
cell cycle of some cell types that are normally quiescent. As noted above, the regions used by El A
to bind the RB-encoded protein, pRB, are involved in the promotion of S phase by ElA. Given
that ElA expression and RB inactivation share this phenotype, binding of ElA to pRB may
inactivate a pRB function and thus prevent (full) enforcement of normal controls on the GUS
transition.
C3.
Cyclin-dependent kinases hyperphosphorylate pRB at the restriction point
Concomitant with passage of the cell through the restriction point, the majority of the RB
protein, pRB, becomes hyperphosporylated (DeCaprio et al., 1989). This hyperphosphorylation
is mediated by the cyclin-dependent kinases (CDKs). CDKs have been demonstrated to play a
central role in promoting progression through the cell cycle in a wide variety of organisms,
including fusion and fission yeast, fruit flies, frogs, and mammals, among others. CDK activity
17
requires association of the CDK subunit with a cyclin subunit; modulation of cyclin protein levels
during the cell cycle seems to be one mechanism regulating CDK activity during the cell cycle.
Cyclins D and E promote passage through GJ
The first CDK complexes known to be active in the GI phase of the mammalian cell cycle
are CDK4 and CDK6 in association with one of the "D-type" cyclins, DI, D2, and D3. Treatment
of mammalian cells in tissue culture with mitogens induces expression of D-type cyclins at the
RNA and protein levels, but both disappear rapidly upon withdrawal of the mitogens (Matsushime
et al., 1991). Overexpression of D-type cyclins leads to a shortening of GI and smaller cell size
(Jiang et al., 1993; Quelle et al., 1993; Resnitzky et al., 1994) Rodent fibroblast cells
overexpressing D-type cyclins also show a reduced (but not eliminated) dependency on serum for
cell cycle progression (Quelle et al., 1993; Resnitzky et al., 1994). Microinjection of a monoclonal
antibody to cyclin DI during early GI prevents S phase entry, whereas no effect is observed closer
to the GI/S transition (Baldin et al., 1993; Quelle et al., 1993). Similar results are observed for
antibodies to cyclin D2 (Lukas et al., 1995b) or cyclin D3 (Bartkova et al., 1998) in cells
expressing these cyclins. Thus, the D-type cyclins, presumably in combination with their partners
CDK4 and CDK6, play an essential role in promoting passage through GI, and this role is
arguably preceding or concomitant with passage through the restriction point.
Like the D-type cyclins in conjunction with CDK4 and CDK6, cyclin E/ CDK2 plays an
essential role in the GI/S transition. Ectopic expression of cyclin E shortens the GI phase of the
cell cycle; coexpression of cyclin Dl with cyclin E shortens GI even further (Resnitzky et al.,
1994; Resnitzky and Reed, 1995). Microinjection of an antibody to cyclin E during early GI
prevents passage through the GI/S transition, whereas microinjection later has no effect (Ohtsubo
et al., 1995). Similarly, overexpression of dominant negative CDK2 arrests cells in GI (Hofmann
and Livingston, 1996).
1 8
Cyclins D and E participatein the hyperphosphorylationof pRB
Multiple lines of evidence indicate that pRB is phosphorylated by CDK complexes with Dtype cyclin subunits. The D-type cyclins can bind directly to pRB (Dowdy et al., 1993; Ewen et
al., 1993) and, in combination with CDK4 or CDK6, will phosphorylate pRB in vitro (Ewen et
al., 1993; Kato and Sherr, 1993) on sites that are also phosphorylated in vivo (Kato and Sherr,
1993). Cyclin D/CDK4 and cyclin D/CDK6 complexes also phosphorylate pRB more efficiently
than they phosphorylate histone HI; they are the only CDKs known to have this subtrate
specificity. Co-expression of cyclin DI with CDK4 induces hyperphosphorylation of pRB in
transient transfections (Lundberg and Weinberg, 1998). Furthermore, pRB is maintained in a
primarily hypophosphorylated state in cells overexpressing p16, a specific inhibitor of CDK4 and
CDK6, indicating that these kinase activities are required for full phosphorylation of the protein
(Lundberg and Weinberg, 1998).
pRB is also phosphorylated by cyclin E/ CDK2 complexes. Overexpression of cyclin E is
sufficient to induce hyperphosporylation of pRB (Hinds et al., 1992; Lundberg and Weinberg,
1998), and cyclin E/ CDK2 complexes can phosphorylate pRB directly in vitro (Sherr, 1994).
Overexpression of a dominant negative form of CDK2 prevents pRB from reaching a fully
phosphorylated state (Lundberg and Weinberg, 1998). Cyclin E protein levels and cyclin Eassociated kinase activity both peak in late GI (Ohtsubo et al., 1995), at the time when pRB first
becomes hyperphosphorylated. The phosphorylation of specific residues of pRB is significantly
restricted in irradiated wild-type MEFs, in which the CDK inhibitor p21 is induced and cyclin Eassociated kinase activity is inhibited, but not in p21-deficient MEFs, suggesting that specific pRB
phosphorylation events require active cyclin E/ CDK2 complexes (Moberg, 1998). The
phosphorylation of pRB on serine 780, in contrast, is mediated specifically by cyclin D (Kitagawa
et al., 1996).
19
Inactivationof pRB, in part by D-type cyclins, is requiredfor passage through GJ
Both D-type cyclins and cyclin E, together with their cognate CDK subunits, can therefore
phosphorylate pRB in vitro, are required for full pRB hyperphosphorylation in vivo, and promote
passage through the GUS transition. In the case of the D-type cyclins, the inactivation of pRB
seems to be the only function required for cell cycle progression. Microinjection of an antibody to
cyclin DI can only induce a GI arrest in cells with wild-type pRB: cells without functional pRB are
unaffected, although sensitivity can be restored by ectopic expression of wild-type pRB in these
cells (Lukas et al., 1994; Lukas et al., 1995a). Similarly, only cells with wild-type pRB can be
arrested in GI by ectopic expression of the CDK4/ CDK6 inhibitor p16 (Lukas et al., 1995).
Because overexpression of D-type cyclins and mutations inactivating pRB both promote Sphase progression and because inactivation of pRB relieves the requirement for D-type cyclinassociated kinase activity, it seems clear that 1) pRB is a critical target for this kinase activity in
vivo and 2) that this activity is required for the hyperphosphorylation and inactivation of pRB and
for the subsequent passage through the GUS transition. This must, furthermore, be the only
requisite target for this activity in cells grown in tissue culture, given that the activity is dispensable
in cells that lack wild-type pRB. Cyclin E, in contrast, must have additional critical targets in vivo:
microinjection of an antibody to cyclin E inhibits cell cycle progression even in cells that lack
functional pRB (Ohtsubo et al., 1995).
HyperphosphorylatedpRBno longer interactswith a viral oncoprotein
It is interesting to note that ElA is not the only viral oncoprotein that is known to bind
pRB. SV40 large T antigen and the E7 protein of human papilloma virus (HPV) are among the
best known transforming viral proteins (other than ElA) that are known to interact with pRB
through regions required for their transforming functions (DeCaprio et al., 1988; Dyson et al.,
1989; Dyson et al., 1992). At least SV40 large T seems to interact specifically with the
underphosphorylated form of pRB (DeCaprio et al., 1988), and the same is suspected to be true of
the other viral oncoproteins (Weinberg, 1995). Hyperphosphorylated pRB has therefore lost at
20
least two functions: inhibition of S phase progression and binding to viral oncoproteins. This also
indicates that any functions of pRB that must be abrogated by the oncoproteins are absent in the
hyperphosphorylated protein. Oncoprotein binding and CDK-mediated hyperphosphorylation
interfere with the same function of pRB, and this function is required for progression to S phase.
Genes regulatingthe hyperphosphorylationof pRB are also mutated in human tumors
This requirement for the inactivation of pRB prior to passage through the GI/S transition is
likely to be the reason RB is often mutated in human tumors. If this is the case, extragenic
mutations that also have the effect of inactivating pRB should also be observed in some tumors.
Indeed, mutations that cause elevated levels of D-type cyclin-associated kinase activity are
frequently observed in human tumors. Cyclin DI was originally identified through its
rearrangement in some B-cell lymphomas (Motokura et al., 1991), and cyclin DI levels have also
been shown to be elevated in a number of other tumor types, in some cases as a result of
amplification of the locus (Herwig and Strauss, 1997). Cyclin D2 has also been shown to be a
proto-oncogene in the mouse (Hanna et al., 1993). Amplification of CDK4 has also been
observed in some gliomas and glioblastomas (He et al., 1994; Schmidt et al., 1994). Similarly,
mutation of p16 is observed in familial melanoma (Hussussian et al., 1994; Kamb et al., 1994),
and deletion of p 16, often in combination with its homolog p 15, another specific inhibitor of
CDK4 and CDK6, is observed in a wide variety of tumor types (Weinberg, 1995). Although the
interpretation of the deletion studies is complicated by the recent identification of another tumor
suppressor, p19-ARF, present at the same locus (Quelle et al., 1995; Kamijo et al., 1997), the
identification of mutations affecting only p16 prove that p16 is clearly a tumor suppressor in its
own right (Quelle et al., 1997). Furthermore, the observation that small cell lung carcinomas
almost invariably demonstrate mutations affecting either pRB expression or p16 expression, but
not both, argues strongly that pRB and p16 are components of the same oft-targeted pathway in the
tumorigenesis of these tissues (Otterson et al., 1994). Thus, the inactivation of pRB by CDK
complexes promotes passage through the restriction point, and the mutagenesis of either pRB,
21
Figure 1.
p16
cyclin D/
CDK 4
pRB
(tumor suppressor)
(oncogenes)
(tumor suppressor)
E2F
Figure 1. Inactivation of pRB is associated with tumorigenesis.
Inhibitory interactions are depicted by inverted "T" symbols.
CDK complexes, or CDK inhibitors can lead to a loss of proliferative control and to tumorigenesis
(Figure 1).
C4.
RB suppresses apoptosis
Some of the tissues in which unscheduled S phase entry is observed in RB-deficient mice
are also prone to apoptosis (Morgenbesser et al., 1994; Macleod et al., 1996). The abnormal
levels of apoptosis observed in the RB-deficient lens or central nervous system (CNS) are
dependent on the presence of wild-type p53, a tumor suppressor and an inducer of apoptosis: mice
that lack both pRB and p53 do not exhibit elevated levels of apoptosis in either the lens or CNS,
although aberrant S phase entry is still observed (Morgenbesser et al., 1994; Macleod et al., 1996).
Aberrant apoptotic figures are also observed in the peripheral nervous systems (PNS) of RBdeficient mice, but this apoptosis is apparently p53-independent (Macleod et al., 1996). RBdeficient MEFs are also sensitized to apoptosis-inducing drugs such as methotrexate (Almasan et
al., 1995).
Overexpression studies also suggest pRB can act to inhibit apoptosis. HeLa cells are a
human cervical carcinoma-derived cell line expressing low levels of HPV oncoproteins,
inactivating their endogenous pRB and p53; overexpression of pRB can protect HeLa cells from
apoptosis induced by overexpressed p53 (Haupt et al., 1995). Expression of wild-type pRB in
SAOS-2 cells, a tumor cell line lacking functional pRB or p53, promotes cell survival following
treatment with ionizing radiation (Haas-Kogan et al., 1995). Wild-type pRB can therefore inhibit
either p53-dependent or -independent apoptosis in certain contexts.
The anti-apoptotic effects of pRB are consistent with the observed pro-apoptotic effects of
ElA proteins that can bind and inactivate pRB. They are also consistent with the known proapoptotic effects of cyclin DI. Overexpression of cyclin Dl can induce apoptosis in multiple cell
lines (Kranenburg et al., 1996; Sofer-Levi and Resnitzky, 1996). Elevated levels of cyclin DI
have been observed in neuronal cells undergoing apoptosis in reponse to growth factor
withdrawal, suggesting that cyclin DI-mediated apoptosis may indeed have in vivo relevance.
23
Overexpression of the CDK4/6 inhibitor p16 has been shown to inhibit apoptosis during myocyte
differentiation (Wang and Walsh, 1996). Thus, either the loss of pRB function through
mutagenesis or its inactivation through hyperphosphorylation can promote apoptosis in a variety of
tissues.
C5.
RB inactivation is insufficient to promote transformation
pRB is a plausible target for E lA-mediated S phase entry and apoptosis and mutations that
inactivate pRB promote tumorigenesis. In addition, inactivation of pRB may eventually prove to
be required for cellular transformation in vitro. For example, mutants of SV40 large T antigen that
have lost the ability to bind to pRB have also lost the ability to transform wild-type MEFs
(Christensen and Imperiale, 1995; Zalvide and DeCaprio, 1995). There is, however, no evidence
that the loss of pRB is sufficient to promote cellular transformation . pRB-deficient MEFs are not
transformed; like wild-type MEFs, pRB-deficient MEFs require functional oncogenes to induce
transformation. The large T mutants that cannot bind pRB are also unable to transform RBdeficient MEFs (Christensen and Imperiale, 1995; Zalvide and DeCaprio, 1995). Similar results
were obtained with studies using the HPV E7 protein (Zalvide and DeCaprio, 1995). Thus, the
inactivation of pRB does not, on its own, significantly increase the efficiency of cellular
transformation. Furthermore, the pRB binding domain of the oncoprotein is required even in cells
lacking pRB; inactivation of this region must also compromise additional functions required for
transformation. In fact, the regions of ElA required for binding to pRB are also involved in
binding to the ElA-associated proteins p107 and p130 (Whyte et al., 1989; Dyson et al., 1992),
raising the possibility that they, alone or in combination with pRB, may mediate the effects of ElA
on cellular transformation.
Indeed, although RB loss promotes tumorigenesis, S phase entry, and apoptosis, mouse
embryos without a functional RB gene survive until embryonic day 13.5; death is most likely the
result of a hematopoietic defect (Jacks et al., 1992). Although some tissues exhibit abnormally
high levels of S phase entry and apoptosis, most embryonic tissues appear entirely normal, having
24
undergone many normal cell cycles and appropriate differentiation. In chimeric mouse embryos
carrying some percentage of RB-deficient cells, those cells can contribute to most, if not all, tissues
(Williams et al., 1994). In conclusion, despite the clear importance of pRB in preventing
tumorigenesis and its contribution to the prevention of inappropriate S phase entry and apoptosis,
there are clearly additional safeguards in most cells working in concert with the pRB pathway to
prevent the deregulation of these processes.
C6. ElA also binds the pRB homologs p107 and p130
The EA-associated proteins p107 and p130 have been cloned and are homologs of pRB
(Ewen et al., 1991; Hannon et al., 1993; Li et al., 1993; Mayol et al., 1993). p107 and p130 are
more homologous to each other than either is to pRB; nevertheless, they share noticeable
homology to pRB throughout their length, particularly in the carboxy-terminal half of the protein.
They also share an insertion between two domains highly conserved among pRB, p107, and p130.
These two domains are referred to collectively as the "small pocket"; the insertion is referred to as
the "spacer", and contains a motif with homology to the CDK inhibitors p21, p27, and p57.
Indeed, both p107 and p130 are able to stably interact with cyclin A/ CDK2 or cyclin E/ CDK2 and
to inhibit their kinase activity (Woo et al., 1997; Castano et al., 1998). Like the overexpression of
p21 or p27, which arrests cells in GI (Xiong et al., 1993; Toyoshima and Hunter, 1994),
overexpression of p107 in some cells induces a GI arrest, and this arrest is dependent on the
integrity of the cyclin/ CDK binding regions of p107.
Neitherp107 norp130 isfrequently mutated in human tumors
In contrast to the frequent mutations of RB in human tumors, the evidence potentially
implicating p107 or p130 in tumorigenesis is sparse. Chromosomal alterations in the p107 gene
have been described in one cell line derived from a T-cell leukemia, one line derived from a B-cell
lymphoma, and one primary tumor from an acute lymphoblastic leukemia (Takimoto et al., 1998).
Full-length p107 protein was no longer detectable in either cell line. One point mutation at the
25
p130 locus has been described in a cell line derived from a small cell lung cancer (Helin et al.,
1997). The possibility remains that the cell line mutations could have arisen during tissue culture;
furthermore, the lack of tumor-free controls from the affected patients allow the possibility that the
mutations may be unrelated to the genesis of the cancer. It is also possible that the inactivation of
p107 or p130 may indeed contribute to tumorigenesis, if infrequently. Mutation of p107 in the
mouse has variable consequences depending on the genetic background (Lee et al., 1996; LeCouter
et al., 1998a). In no background has the loss of p107 been reported to promote tumorigenesis,
although loss of p107 does induce hyperplasia in certain myeloid tissues of one mouse strain
(LeCouter et al., 1998a). Chimeric mice with cells deficient in both pRB and p107 develop
retinoblastoma at high frequency, suggesting that p107 may share a tumor-suppressive function
with pRB, at least in this system (Lee et al., 1996). As seen for p107 mutations, mutations
inactivating p130 have dramatically differing consequences in different genetic backgrounds, but
these consequences do not seem to include an increased frequency of tumorigenesis (Cobrinik et
al., 1996; LeCouter et al., 1998b).
p107 and p130 in apoptosis, transformation,and cell cycle progression
p107 and p130 associate with the same domains of El A required for interaction of the latter
with pRB (Dyson et al., 1992), raising the possibility that they might participate in the regulation
of apoptosis, transformation, or the GUS transition. Although mice lacking p107 are healthy and
fertile, mice lacking both p107 and pRB demonstrate increased levels of apoptosis in the liver and
CNS than mice only lacking pRB, indicating that they may share an apoptosis-suppressing
function (Lee et al., 1996). Some mice in which the p130 gene has been mutated show extensive
apoptosis in a variety of tissues, but this apoptosis is only observed in certain genetic backgrounds
(LeCouter et al., 1998b).
In contrast, no direct evidence implicates either protein in the regulation of cellular
transformation. p10 7 and p130, like pRB, do associate with other transforming viral oncoproteins
such as SV40 large T antigen and HPV E7, and the inability of a T antigen mutant that has lost the
26
ability to bind pRB, p107 or p130 to transform wild-type or pRB-deficient MEFs may point to a
transformation-suppressive function for p107 and/or p130. Interestingly, an additional domain of
T antigen that promotes the degradation of p130 and changes in the phosphorylation state of p107
and p130 is also required to induce MEFs to grow to high density or in low serum; this domain is
dispensable in MEFs deficient in both p107 and p130 (Zalvide et al., 1998).
The nature of the role played by these proteins in the regulation of the GUS transition
remains somewhat controversial. p130 is expressed primarily in GO cells, although it is still
detectable in GI and S phase (Cobrinik et al., 1993). The expression of p107, in contrast, is
strongly induced upon reentry of cells into the cell cycle (Moberg et al., 1996). Like pRB, p107
and p130 can be phosphorylated by GI-specific CDK complexes (Li et al., 1993; Beijersbergen et
al., 1995). As mentioned above, overexpression of p107 can induce a GI arrest. The effects of
p107 or p130 loss on cellular proliferation in mice are strongly dependent on the genetic strain of
the mice. In one genetic background, both MEFs and peripheral lymphocytes from these mice
retain restriction point controls, although lymphocytes from mice lacking both p107 and p130 are
hypersensitive to stimulation by concanavalin A and MEFs deficient in both proteins have been
reported to enter S phase slightly earlier than wild-type cells following release from serum
starvation (Cobrinik et al., 1996; Mulligan et al., 1998). In another background, loss of p107
leads to a doubling of the proliferation rate of MEF cultures, and loss of p130 in the mouse leads to
extensive aberrant S phase entry in neural tissues. There exist, therefore, biological contexts in
which p107 and p130 can play critical roles in proliferation control, although the genetic modifer or
modifiers that must also contribute to proper cellular regulation remain unknown.
C7.
pRB suppresses growth through its "large pocket"
pRB, pl0 7 , and p130 bind to viral oncoproteins through their conserved "small pocket"
(Hu et al., 1990; Huang et al., 1990; Kaelin et al., 1990; Ewen et al., 1991; Hannon et al., 1993).
Mutations in this region of pRB that prevent its interaction with viral oncoproteins also inhibit the
ability of ectopically expressed pRB to suppress the growth of RB-deficient tumor-derived cell
27
lines (Qian et al., 1992; Qin et al., 1992; Hiebert, 1993). Similarly, although overexpression of
wild-type pRB can inhibit apoptosis in SAOS-2 cells, expression of a mutant pRB that is unable to
bind E l A is also unable to inhibit apoptosis in these cells (Haas-Kogan et al., 1995). This
suggests a model in which ElA interferes with a function of pRB through direct binding to the
domain of the protein required for that function.
While the small pocket is necessary for pRB-mediated growth suppression, it is not
sufficient (Qian et al., 1992; Qin et al., 1992; Hiebert, 1993). The minimal region of pRB required
to suppress growth includes additional sequences immediately to the carboxy-terminus of the small
pocket (Qin et al., 1992; Hiebert, 1993); together with the small pocket, this region is referred to as
the "large pocket", and is somewhat conserved with p107 and p130. All known tumor-derived
mutations of RB disrupt the large pocket through deletion, truncation or point mutation (Sellers and
Kaelin, 1996). This region, then, appears central to the function of pRB both in vitro and in vivo.
Like ElA, pRB is believed to carry out its biological activities through interactions with
other proteins. Phosphorylation of the protein by CDK complexes, for example, seems to regulate
the inactivation of the protein as the cell prepares to traverse the GI/S transition. While this reveals
much about the proper regulation of the protein and the importance of the timing of its biological
activity, it does not reveal how pRB exerts its activity on the cell. Although many proteins have
been reported to bind to pRB in vitro, one particular cellular factor has been shown to stably
interact with pRB both in vivo and in vitro in a manner specifically dependent on an intact,
underphosphorylated large pocket: the transcription factor E2F.
28
D.
E2F function is antagonized by pRB, p107, and p130
pRB binds to E2F and inhibits cell proliferation
The E2F transcription factor was originally identified as an El A-responsive cellular
transcription factor with a sequence-specific DNA binding activity (Kovesdi et al., 1986b; Kovesdi
et al., 1986a; Kovesdi et al., 1987; Yee et al., 1987). In the context of adenoviral infection, E2F
binds to two sites in the promoter of the viral E2 gene; these sites are essential for efficient
expression of the gene (Kovesdi et al., 1986b). Not surprisingly, E2F sites have also been
identified in a number of endogenous cellular genes, and have been shown to mediate the
upregulation of these genes at the GUS transition. Most of these genes are regulatory genes
required in GI or S, such as RB (Shan et al., 1994) or B-myb (Lam and Watson, 1993; Lam et
al., 1994; Lam et al., 1995), or genes required for the machinery of DNA synthesis, such as
dihydrofolate reductase (Means et al., 1992; Wade et al., 1992; Slansky et al., 1993; Wade et al.,
1995) or DNA polymerase alpha (Pearson et al., 1991).
In cycling cells, a significant percentage of E2F is associated with pRB (Bandara and La
Thangue, 1991; Chellappan et al., 1991). This interaction is disrupted by wild-type ElA, but not
by mutant forms that have lost the ability to bind pRB. Furthermore, all known tumor-derived
mutations of pRB disrupt its ability to bind E2F (Sellers and Kaelin, 1996). The inactivation of
pRB function, either in tumor-derived mutations or in association with viral oncogenes, frees E2F
from its interactions with pRB.
In normal cells, the association between pRB and E2F is regulated during the cell cycle.
Hypophosphorylated pRB, which predominates during cell cycle arrest, can be found in
association with E2F. pRB that has been hyperphosphorylated by cyclin/ CDK complexes
following passage through the restriction point (DeCaprio et al., 1989), in contrast, has lost its
affinity for E2F (Chellappan et al., 1991; Knudsen and Wang, 1997) Mutations of the
phosphorylation sites in pRB enhance the abilility of the protein to induce a cell cycle arrest when
overexpressed, but this arrest can be overcome by the overexpression of an E2F subunit (Qin et
29
al., 1995). Thus, hypophosphorylated pRB binds E2F and inhibits cell proliferation, whereas the
hyperphosphorylation of pRB leads to the dissociation of E2F and passage through the GUS
transition.
D1.
The E2F and DP gene families
Although E2F was originally described as a DNA-binding activity, it is in fact rather
heterogeneous, resulting from the proteins encoded by at least seven distinct genes in mammals
(Helin et al., 1992; Kaelin et al., 1992; Shan et al., 1992; Girling et al., 1993; Ivey-Hoyle et al.,
1993; Lees et al., 1993; Beijersbergen et al., 1994; Ginsberg et al., 1994; Hijmans et al., 1995;
Ormondroyd et al., 1995; Sardet et al., 1995; Wu et al., 1995; Zhang and Chellappan, 1995;
Rogers et al., 1996). These are grouped into two families of genes, referred to as E2F (E2F-1
through -5) and DP (DP-1 and -2). Either of the DP polypeptides are capable of heterodimerizing
with any of the E2F polypeptides; each of the possible ten resulting pairs are observed in cell
extracts. Heterodimerization is essential for efficient binding to DNA and transcriptional activation
by the complex (Bandara et al., 1993; Helin et al., 1993; Krek et al., 1993).
Whereas the sequence conservation between the two DP genes is high throughout the
length of their coding sequences, the sequence homology among the E2F genes is concentrated in a
few key domains. The domain encoding the DNA-binding function is the most highly conserved,
and does not resemble DNA binding domains from other transcription factors. This domain is
followed immediately by a domain required for efficient heterodimerization with the DP proteins.
The transcriptional activation function of the complex is encoded by a conserved region in the 3'
end of the E2F genes. Based on sequence conservation and gene structure, E2F-4 and -5 form a
subfamily of E2F proteins distinct from the subfamily including E2F-1, -2, and -3. This latter
group contains an amino-terminal region that includes a conserved nuclear localization signal
(Verona et al., 1997); this region is absent in E2F-4 and -5.
Domains required for DNA binding and dimerization are also found in the DP proteins, and
these regions show limited homology to the corresponding regions of the E2F proteins. No
30
domain in the DP proteins contributing to transcriptional activation by the complex (other than the
dimerization and DNA-binding domains) has been identified; rather, presence of the DP moiety in
the E2F/DP complex is essential for DNA binding and the potentiation of the transcriptional
activation function of E2F. The DP polypeptide also promotes the interaction of the complex with
pRB, although this binding occurs primarily through the transcriptional activation domain of the
E2F subunit.
D2.
pRB-mediated repression
The association of pRB with E2F is regulated during the cell cycle. This regulation is
believed to occur both through the specific phosphorylation of pRB at the GUS transition by CDK
complexes and through transcriptional regulation of the RB gene. pRB/ E2F complexes cannot be
detected in quiescent cells, but can be detected in late G 1 as these cells reenter the cell cycle and the
transcription of pRB is activated (Ikeda et al., 1996; Moberg et al., 1996). They can also be
detected in S phase. This is believed to result from newly synthesized pRB, not yet
hyperphosphorylated by CDK complexes, binding to E2F. Importantly, when cells are induced to
arrest and differentiate, pRB is no longer present in a primarily hyperphosphorylated form, and a
much greater percentage of E2F complexes associate with pRB (Ikeda et al., 1996).
When bound to E2F, pRB inhibits E2F-mediated transcription. This was first indicated by
studies of the E2 promoter: overexpression of pRB inhibits the expression of the promoter in a
manner dependent on the presence of the E2F sites, and this effect is only observed with alleles of
pRB that are competent to bind E2F (Hiebert et al., 1992). The repression of the E2 promoter can
be alleviated by co-expression of EtA or HPV E7 (Hiebert et al., 1992). This derepression can be
explained by the in vitro and in vivo observations that EIA induces the dissociation of pRB from
E2F complexes (Bagchi et al., 1990). pRB does not prevent DNA binding by the complex;
endogenous pRB-E2F complexes are easily detectable in gel shift assays (Chellappan et al., 1991;
Moberg et al., 1996). The inhibitory activity of pRB is instead specifically a repression of
transcription by the intact DNA-bound complex.
3 1
pRB represses basaltranscription
The presence of pRB, however, does not merely mask the transcriptional activation domain
of the E2F moiety, but also represses the basal level of transcription when present at a promoter.
This was first suggested by the observation that the overexpression of pRB inhibits the E2
promoter more than the simple deletion of the E2F sites does, and this additional inhibition is
dependent on the integrity of the E2F sites (Hiebert et al., 1992). Similarly, it has been shown that
the insertion of E2F sites can repress artificial promoters in cells with wild-type pRB, whereas the
same insertion activates the expression of those promoters in cells without functional pRB, or in
cells expressing wild-type ElA (Weintraub et al., 1992). That this repression is the direct result of
the action of pRB on the promoter has been demonstrated with fusion proteins between pRB and
the DNA-binding domain of the yeast GAL4 protein: this fusion can repress the expression of
artificial promoters in a manner dependent on the presence of either a GAL4 site (to which GAL4pRB could directly bind) or an E2F site (to which GAL4-pRB could bind through an E2F
intermediary) (Adnane et al., 1995; Bremner et al., 1995; Weintraub et al., 1995). Thus, although
"free E2F", E2F/DP complexes not associated with pRB, can activate transcription, pRB-E2F
complexes repress basal levels of transcription.
The mechanisms of pRB-mediated repression are under active investigation. Repression
requires only the "small pocket" of pRB, and can be alleviated by overexpression of viral
oncoproteins or of cyclins A or E. Repression may be effected by interactions between pRB and
other promoter elements (Dyson, 1998) and/or by recruitment of histone deacetylase activity to the
promoter by pRB (Brehm et al., 1998; Luo et al., 1998; Magnaghi-Jaulin et al., 1998).
32
D3.
p107 and p130 also bind E2F and can repress transcription
pRB, p107, andp130 bind E2F at distinct times during the cell cycle
Although "free E2F" and pRB-associated E2F account for a significant percentage of the
E2F in the cell, other E2F complexes exist as well. Like the pRB association, these complexes can
be disrupted by ElA (Bagchi et al., 1990). The cloning of the ElA-associated proteins p107 and
p130 and their significant homology to pRB led rapidly to their identification as components of
these "other" E2F complexes (Ewen et al., 1991; Cao et al., 1992; Shirodkar et al., 1992; Cobrinik
et al., 1993).
E2F complexes containing p107 and those containing p130 show distinctly different
patterns of appearance and disappearance during the cell cycle. p130/ E2F complexes are present
at their highest levels in quiescent cells, and become increasingly less abundant as cells reenter the
cell cycle, although they can still be detected as cells enter S phase (Cobrinik et al., 1993; Ikeda et
al., 1996; Moberg et al., 1996). p107/ E2F complexes, in contrast, are not detectable in quiescent
cells, first appearing in late GI and reaching their highest levels in S phase (Moberg et al., 1996).
Binding of p107 or p130 to CDK complexes and E2F is not mutually exclusive; the five
polypeptide complex p107 or p130/ cyclin A or E/ CDK 2/ E2F/ DP complex can be observed
bound to DNA in a gel shift assay, although the potential biological function of a presumably
inactive kinase in this complex remains unknown. Like pRB, p107 and p130 can repress E2F
activity (Bandara et al., 1993; Zamanian and La Thangue, 1993; Bremner et al., 1995; Ferreira et
al., 1998; Pierce et al., 1998b) and may be associated with a histone deacetylase (Ferreira et al.,
1998). p13 0 is therefore proposed to mediate the repression of E2F target genes in quiescent cells;
the role p107/ E2F complexes might be playing, at times in the cell cycle when known E2F target
genes are actively expressed, remains unclear.
33
pRB, p107, and p130 bind to distinctforms of E2F
It has been proposed that the failure to detect specific roles for p107 or p130 in apoptosis or
growth control in some contexts could result from extensive redundancy of function among pRB,
p107, and p130. In this model, the three proteins would share a common suppressive function,
such as transcriptional repression at E2F DNA-binding sites, and only the inactivation of all three
genes could reveal their importance to the biology of the cell. Alternatively, the differences in the
biology of pRB, p107, and p130 could result from their associations with biochemically distinct
forms of E2F. Indeed, pRB can be found in association with E2F-1, -2, -3, or -4 (Moberg et al.,
1996), whereas p107 is associated only with E2F-4 (Moberg et al., 1996) and p130 with both
E2F-4 and E2F-5 (Hijmans et al., 1995; Sardet et al., 1995); although the DP subunit is always
present in these complexes, the identity of the subunit, DP- 1 or DP-2, does not seem to affect to
the specificity of the interaction.
A different subset of E2F transcriptional targets is deregulated in MEFs lacking p107 and
p130 than the subset deregulated in pRB-deficient MEFs. Cyclin E expression, for example, is
upregulated in pRB-deficient MEFs, but not in p107/ p130-deficient cells (Herrera et al., 1996;
Hurford et al., 1997). B-myb, in contrast, is derepressed in p107/p130-deficient MEFs, but
shows a wild-type pattern of expression in pRB-deficient cells (Hurford et al., 1997). These
proteins therefore regulate different biochemical pathways in these cells. Given that pRB, p107
and p130 associate with different E2F proteins, this specific deregulation could result from
differential promoter selection mediated either by the E2F subunit or by interactions between the
pRB homolog and other promoter elements .
34
E.
E2F promotes proliferation and apoptosis
El.
E2F promotes progression through G1
Like ElA, ectopic expression of the E2F genes promotes entry into S phase. E2F-1, -2,
-3, -4, and -5 all have this ability in some experiments, although neither E2F-4 nor -5 efficiently
promotes S phase in others (Johnson et al., 1993; Beijersbergen et al., 1994; Lukas et al., 1996).
Overexpression of E2F-I can override or prevent GI arrest induced by a variety of stimuli,
including transforming growth factor-beta, gamma-irradiation, and the overexpression of CDK
inhibitors p16, p21, p27 or dominant-negative CDK2 (DeGregori et al., 1995; Johnson et al.,
1993; Lukas et al., 1996; Mann and Jones, 1996; Schwarz et al., 1995). Overexpression of
E2F- 1 in mouse megakaryocytes also induces their hyperproliferation and interferes with their
terminal differentiation (Guy et al., 1996). S phase promotion by E2F-1 requires both its DNAbinding and transactivation functions (Johnson et al., 1993; Qin et al., 1994), whereas mutations
in E2F- 1 with reduced affinity for pRB enhance the ability of E2F- 1 to promote S phase entry
(Shan et al., 1996).
In addition to their ability to induce inappropriate entry into S phase when misexpressed,
transcriptional activation via the E2F genes may also be essential for the G1/S transition.
Overexpression of dominant-negative forms of DP- 1 or DP-2 can promote the accumulation of
SAOS-2 cells in GI (Wu et al., 1996; Fan and Bertino, 1997). Similarly, microinjection of a
dominant-negative form of E2F- 1 can prevent the stimulation of S phase entry by El A
(Dobrowolski et al., 1994), and inhibition of E2F DNA-binding activity inhibits S phase induction
following serum stimulation (Ishizaki et al., 1996). Although all five E2F genes share the ability
to induce S phase entry when overexpressed, microinjection of an antibody to E2F-3 inhibits S
phase progression in a rat embryo fibroblast cell line (REF52), whereas an antibody to E2F-I was
without effect, suggesting an specific role for E2F-3 in G/S progression (Leone et al., 1998).
E2F-3-deficient MEFs proliferate more slowly than wild-type MEFs; however, they do proliferate,
35
and mice that lack E2F-3 are fully viable, arguing that E2F-3 is rate-limiting for progression but
not irreplaceable (P. Humbert, K. Rogers, S. Dandapani, and J. Lees, unpublished data).
E2F complexes promote cell cycle progressionin Drosophila
Homologs of E2F, DP, and pRB have also been identified in D. melanogaster(Dynlacht et
al., 1994a; Ohtani and Nevins, 1994; Du et al., 1996a; Sawado et al., 1998), where they also seem
to be involved in the regulation of cell proliferation and S phase progression. Loss-of-function
mutations in an E2F homolog, dE2F, or a DP homolog, dDP, are lethal, usually at the late larval or
pupal stages of development, and are associated with a slower progression through S phase
(Royzman et al., 1997; Duronio et al., 1998). Elimination of dE2F function in a compartment of
the developing wing has also been shown to slow cell cycle progression, as has overexpression of
the RB homolog RBF (Neufeld et al., 1998). In contrast, the ectopic expression of dE2F induces
an unscheduled entry into S phase in a variety of Drosophilatissues (Asano et al., 1996; Duronio
et al., 1996), and in the wing promotes rapid cellular proliferation; this effect is further enhanced
by coexpression of dDP (Neufeld et al., 1998). The promotion of cellular proliferation by E2F
and DP proteins is therefore a function highly conserved across metazoan evolution.
Thus, Drosophilaembryos and larvae do not require zygotic dE2F or dDP expression for
progression, albeit slow, through S phase. If the maternal stores of E2F-encoding mRNA have
been exhausted, this suggests that some components of E2F activity are partially dispensable for
cell cycle progression in this organism during its early development. A second DrosophilaE2F
homolog, termed dE2F2, has recently been described (Sawado et al., 1998), and the potential
existence of additional E2F or DP homologs cannot yet be dismissed; the mutations in dE2F and
dDP are therefore not necessarily eliminating all E2F complexes in these cells. Furthermore,
additional arguments suggest that the expression of transcriptional targets of E2F may indeed be
required for S phase progression in Drosophila. First, some Drosophilatarget genes known to be
regulated by dE2F, such as cyclin E and PCNA, are also known to be required for S phase
(Henderson et al., 1994; Knoblich et al., 1994). Second, some E2F transcriptional targets are
36
derepressed by loss-of-function mutations in dDP (Royzman et al., 1997; Duronio et al., 1998),
raising the possibility that, although E2F-mediated transcriptional activation is clearly requisite for
efficient progression through S phase, derepression of E2F target genes may suffice for passage
through S. Thus, although dE2F and dDP are required for efficient passage through S phase, the
question of a general requirement for the induction of E2F transcriptional targets in DrosophilaS
phase progression remains unresolved.
E2.
E2F promotes transformation and apoptosis
E2F genes also have some capacity to promote cellular transformation in vitro.
Overexpression of E2F- 1 has been reported to cooperate in a transformation assay with a mutant
form of HPV E7 that has lost its ability to bind to the pRB family, suggesting that activation of the
E2F pathway may be the critical role of this domain in transforming viral oncoproteins (Melillo et
al., 1994). Overexpression of the DP proteins, either alone or in combination with E2F-1 or E2F4, has been shown to cooperate with activated ras in transformation assays, (Beijersbergen et al.,
1994; Johnson et al., 1994a; Jooss et al., 1995), although they seem to cooperate less well with
ras than c-myc or ElA does in these assays. Overexpression of E2F-1, -2, or -3 in NIH 3T3 cells
also promotes transformation (Xu et al., 1995). Mutational studies of E2F-1 indicate that its
transforming activity requires an intact DNA-binding domain (Xu et al., 1995), although a
heterologous transcriptional activation domain can substitute for the endogenous one (Johnson et
al., 1994a).
E2F-1 promotes the induction of apoptosis
The activation of E2F-I also promotes apoptosis. Overexpression of E2F-1 induces
apoptosis in a variety of systems (Qin et al., 1994; Shan and Lee, 1994; Wu and Levine, 1994;
Kowalik et al., 1995). Ectopic expression of the Drosophila protein dE2F also induces apoptosis
in multiple cell types (Asano et al., 1996; Du et al., 1996b; Neufeld et al., 1998), indicating that,
like the promotion of cellular proliferation, this function has been well conserved during evolution.
37
Much of the apoptosis induced by the overexpression of E2F- 1 in mammalian systems depends on
the presence of functional p53 protein (Qin et al., 1994; Wu and Levine, 1994; Kowalik et al.,
1995). The observation that ectopic expression of E2F-1 induces expression of p19-ARF
(DeGregori et al., 1997; Bates et al., 1998), a protein that activates p53-mediated transcription and
stabilizes p53 protein levels (Pomerantz et al., 1998; Zhang et al., 1998) suggests that at least some
of the effects of E2F-1 on apoptosis may be mediated by this pathway. E2F-1 overexpression has
also been shown to promote apoptosis in some systems lacking functional p53 (Hsieh et al., 1997;
Phillips et al., 1997). The effects of E2F-1 on p53-mediated apoptosis depend on the DNAbinding and transcriptional activation functions of E2F-1, although p53-independent apoptosis can
be induced in SAOS-2 cells by a form of E2F- 1 lacking a transcriptional activation domain.
Ectopic expression of E2F-1 induces p53-dependent apoptosis in keratinocytes of transgenic mice
(Pierce et al., 1998a), and p53-independent apoptosis in the testes (Holmberg et al., 1998).
The induction of apoptosis is not a general consequence of E2F activation. Rather, the
ability to promote apoptosis is limited to particular E2F complexes. Ectopic expression of E2F- 1
in REF52 cells induces apoptosis, whereas expression of E2F-2, -3, -4, or -5 does not (DeGregori
et al., 1997). Similarly, apoptosis is induced in an astrocytoma cell line by ectopic expression of
either E2F-I or E2F-2, but not by the other E2F genes (Dirks et al., 1998). In both of these cases
E2F overexpression, sometimes in combination with DP- 1, was observed to promote cell cycle
progression regardless of which E2F was used in the experiment; only the ability to induce
apoptosis appears specific in these assays.
Experiments in mice indicate that E2F-1 may regulate apoptosis in a variety of tissues.
Mice lacking functional E2F- 1 develop thymic hypertrophy, and thymocytes from these mice are
partially resistant to apoptosis (Field et al., 1996). Homozygous deletion of the E2F-1 gene also
rescues most of the aberrant apoptosis and S phase entry observed in the lens and CNS of mice
lacking pRB, although aberrant apoptosis in the PNS is still observed (Tsai et al., 1998).
Similarly, loss of E2F- 1 rescues apoptosis in transgenic mice expressing in the choroid plexus a
fragment of SV40 large T antigen (T-121) retaining its ability to bind pRB-family members (Pan et
38
al., 1998). Thus, E2F-1 activity promotes the induction of apoptosis in vivo as well as in vitro.
Once again, this suggests some specificity of E2F function.
E3.
E2F as an oncogene, and as a tumor suppressor
Although the pRB pathway is routinely mutated in human cancers and although the
expression of cellular or viral oncogenes can lead to an activation of E2F transcriptional activity,
the E2F genes themselves do not seem to be mutational targets in human tumors. Alterations in the
length of a polyserine repeat in E2F-4 have been reported, although this was believed to result
from the genomic stability associated with tumorigenesis and not believed to have been a causative
event in tumorigenesis (Yoshitaka et al., 1996; Souza et al., 1997; Walsh et al., 1998). A variety
of hypotheses may explain the lack of tumor-promoting E2F mutations in tumors. The
transcriptional targets of multiple E2F genes may promote tumorigenesis, in which case mutations
in regulators of E2F, such as pRB, may be strongly favored statistically, because these mutations
simultaneously deregulate the products of multiple E2F genes. Another possibility is that
mutations in E2F genes that abolish binding to pRB-family members without reducing
transcriptional activity may be rare or even impossible (R. Verona and J. Lees, unpublished data).
Experiments with E2F-I deficient mice, however, do indicate a role for this gene in
tumorigenesis. E2F- 1 deficiency prevents thyroid tumorigenesis in mice heterozygous for an RB
null mutation, and reduces the incidence of pituitary tumorigenesis (Yamasaki et al., 1998). E2F-1
deficiency also inhibits cellular proliferation in T- 121 induced tumors of the choroid plexus (Pan et
al., 1998). A function supplied by E2F-1 is therefore essential for efficient tumorigenesis in some
contexts, indicating that the deregulation of E2F by mutagenesis of the pRB/ cyclin D/ CDK4/ p16
pathway is highly relevant to tumorigenesis affected by this pathway (Figure 1). Mice deficient for
E2F- 1 develop late onset tumors of the reproductive tract more often than wild-type mice,
indicating that E2F-1 can also act as a tumor suppressor (Yamasaki et al., 1996). It has not yet
been determined if this suppressive role for E2F-I is cell autonomous. If it is, it could indicate that
removal of a pRB-E2F-1 complex derepresses an oncogenic target gene; alternatively,
39
transcriptional activation by E2F- 1 might be necessary for the expression of a tumor suppressor in
these tissues, perhaps related to the pro-apoptotic function of E2F-1.
40
F.
Fl.
E2F genes: specificity in function and in regulation
Individual E2F genes have specific functions
The studies of El A have led to a transcriptional activity, E2F, which participates in the
regulation of cellular proliferation and apoptosis both in normal tissues and in tumors. To
understand the mechanisms through which E2F regulates these processes requires the identification
of the specific E2F genes involved in the regulation of each process; it also requires an
understanding of the ways specific E2F proteins are regulated.
The existence of some functional overlap among E2F genes seems very likely; for example,
all five E2F genes have been reported to promote cell cycle progression when overexpressed,
although in some cases this effect requires the coexpression of a DP subunit. There is,
nevertheless, strong evidence that individual E2F genes also play distinct roles in the regulation of
cell biology. As discussed previously, E2F-I is involved in the regulation of apoptosis, a role not
indicated for the other E2F genes with the possible exception of E2F-2. E2F- 1 may also play a
specific role in the deregulation of the cell cycle caused by loss of pRB. E2F-3 may also have a
specific role in Gl/S progression, as indicated by the proliferative defect in E2F-3-deficient MEFs
and the ability of an E2F-3-specific antibody to inhibit S phase entry. Unlike mice deficient for
E2F-l or -3, mice without E2F-4 display hematopoietic defects (P. Humbert, K. Rogers, and J.
Lees, unpublished data). Mice without E2F-5 die at about six weeks of age with hydrocephalus,
apparently the result of excessive cerebrospinal fluid production in the choroid plexus (Lindeman et
al., 1998). The defects observed in mice lacking E2F-4 or E2F-5 may suggest a role for E2F
complexes in differentiation, or may be consequences of cell cycle deregulation in certain tissue
types. Regardless of the mechanisms involved, specific E2F genes must supply a variety of nonredundant functions important at the cellular and organismal levels. This is consistent with the
argument that pRB, p107, and p130 differ in their biological activities because they regulate
different subsets of E2F proteins (Figure 2). pRB may suppress tumorigenesis because the E2F
proteins regulated by pRB may have specific functions relevant to tumorigenesis, whereas neither
41
Figure 2.
p107
pRB
frequently mutated
in tumors
N OT frequently
mutated in
tu mors
p130
NOT frequently
mutated in
tumors
E2F-1
E2F-2
E2F-3
E2F-4
E2F-5
promotes
promotes
S phase
romotes
phase
entry,
proliferation
required in
4ematopoesis
loss
apoptosis,
proliferation
entry?
Figure 2. The pRB family members inhibit different E2F proteins with distinct functions.
Inhibitory interactions are depicted by the lines connecting pRB and E2F family members.
Functions of each protein are listed below the name of the protein.
disregulates
CSF
production
p107 nor p130 is frequently lost in cancer because the E2F proteins they regulate may have
functions less likely to promote tumorigenesis. While this hypothesis remains speculative, its
underlying premise that the E2F proteins differ in some of their functions is apparently true.
F2.
The E2F genes are regulated in overlapping but distinct ways
Given that the E2F proteins have non-overlapping functions in addition to any shared
functions, it is perhaps unsurprising that there are differences in their methods of regulation.
Obviously, the best characterized method of regulation is the association of E2F complexes with
pRB-family members: E2F-5 associating only with p130, E2F-1, -2, and -3 associating only with
pRB, and E2F-4 associating with pRB, p107, and p130. The observation that the three pRBfamily members have both distinct and overlapping functions is therefore consistent with the
distinct and (probably) overlapping functions observed for the E2F proteins they regulate.
Mutation of the pRB-family members deregulates specific subsets of E2F target genes; these
subsets, then, presumably mediate the specific biological effects of E2F and pRB family members
on cell biology.
E2F mRNA levels are regulatedduring development
Individual E2F family members are also regulated by control of their expression,
localization, and DNA binding activities. E2F-2 and E2F-4, for example, are widely expressed in
the developing mouse (Dagnino et al., 1997a). All five E2F transcripts can be detected by in situ
hybridization in the kidney, liver, lung, and heart of the developing mouse (Dagnino et al., 1997a).
E2F-5, however, appears absent from the developing thymus and from epithelial tissues that are
not undergoing terminal differentiation (Dagnino et al., 1997a; Dagnino et al., 1997b). E2F-3 is
also undetectable in the developing thymus, and neither E2F-1 nor E2F-3 can be detected in the
developing choroid plexus or in chondrocytes (Dagnino et al., 1997a). In skeletal muscle, only
E2F-2 can be readily detected by in situ hybridization, although E2F-3, -4, and -5, but not E2F-1,
can also be detected (Dagnino et al., 1997a). Whether this control over E2F mRNA levels during
43
development results from transcriptional control or from regulation of mRNA stability remains
unresolved. Regardless of the mechanism(s) involved, the restricted expression of particular
subsets of E2F genes in some tissues is intriguing and would be consistent with the inactivation of
certain E2F-dependent regulatory pathways in a tissue-specific manner.
E2F levels and localizationare regulatedduring the cell cycle
Much more is known about the regulation of E2F proteins from studies of cells grown in
vitro. The primary E2F complexes in quiescent cells are E2F-4/ DP and E2F-5/ DP, each of which
are associated with p130 and therefore assumed to be repressing the expression of E2F target
genes; the E2F-4-containing complexes outnumber the E2F-5 complexes. As cells enter the cell
cycle and approach S phase, E2F-4 and E2F-5 become increasingly excluded from the nucleus,
such that by S phase the vast majority of E2F-4 and -5 complexes are cytoplasmic (Lindeman et
al., 1997; Muller et al., 1997; Verona et al., 1997). This relocation of E2F-4 may promote the
derepression of E2F target genes at the approach of the GI/S transition, as E2F-4 is primarily
associated with p130 as cells are first reentering the cell cycle and with pRB and p107 in late GI
and S (Moberg et al., 1996). The relocation of both proteins may also prevent the premature
activation of E2F target genes in early- to mid- GI when free E2F-4 complexes first appear. E2F4 and -5 are the only E2F proteins known to undergo changes in subcellular localization; E2F-1,
-2, and -3 all seem to be constitutively nuclear (Verona et al., 1997).
Reentry of cells into the cell cycle is also associated with the de novo synthesis of E2F
proteins. Whereas E2F-4 levels are only mildly elevated upon exit of cells from a quiescent state,
levels of E2F-1, -2, and -3 are strongly induced (Hsiao et al., 1994; Johnson et al., 1994b;
Neuman et al., 1994; Moberg et al., 1996). The induction of E2F-1 and -2 appears to occur at the
transcriptional level and are mediated, at least in part, by the presence of E2F binding sites in their
promoters (Hsiao et al., 1994; Johnson et al., 1994b; Neuman et al., 1994; Sears et al., 1997);
E2F-3 is also upregulated at the transcriptional level, although this upregulation is much less
dramatic than the induction of its protein levels or DNA binding activity (Moberg et al., 1996;
Leone et al., 1998). This discrepancy may be explained by the observation that the E2F-3 protein
44
in NIH 3T3 cells that have been blocked in S phase is 40-fold more stable than in asynchronous
cells (Flores et al., 1998). Although E2F-1 and E2F-2 expression and DNA binding activities are
strongly induced upon reentry of cells into the cell cycle, accumulation of E2F complexes at the
GUS transition in continuously cycling cells has only been described for E2F-3.
Protein stability may also regulate other members of the E2F family. At least E2F- 1 and
E2F-4, when overexpressed, are targeted for degradation by the ubiquitin-proteasome pathway
(Hateboer et al., 1996; Hofmann et al., 1996). Coexpression of pRB-family members competent
to bind the E2F in question results in the stabilization of the E2F complex; this is postulated to
result from the masking of a carboxy-terminal epitope on the E2F moiety required for its targeting
and/or degradation. If endogenous E2F complexes are regulated in the same manner, this would
suggest that the levels of free E2F complexes are reduced by their specific degradation, whereas
complexed E2F is relatively stable in comparison. Whether or not the degradation of free E2F
complexes is regulated in a cell cycle-dependent manner, beyond the cell cycle-dependent pRBfamily associations, is unknown.
Some E2F complexes are inactivated by specific phosphorylation
The DNA binding activity of some E2F complexes is regulated by phosphorylation. E2F-1
can interact with cyclin A through the same region of E2F- 1 required for its nuclear localization
(Dynlacht et al., 1994b; Krek et al., 1994; Xu et al., 1994; Kitagawa et al., 1995; Krek et al.,
1995; Dynlacht et al., 1997). This kinase is active, and phosphorylates in vitro both subunits of an
E2F-1/ DP-1 complex on sites that are also phosphorylated in vivo. Cyclin A-mediated
phosphorylation of the DP- 1 subunit in this complex leads to a loss of DNA binding activity by the
complex (Dynlacht et al., 1994b; Krek et al., 1994; Xu et al., 1994; Kitagawa et al., 1995). E2F2 and E2F-3 are also associated with a kinase activity in vivo (Krek et al., 1994), and incubation of
E2F-3/DP-1 complexes with cyclin A/ CDK2 inhibits their DNA binding activity (Dynlacht et al.,
1997), suggesting that this downregulation may be shared amongst these three E2F complexes.
E2F-4/DP-1 and E2F-5/DP-1 complexes, in contrast, are unaffected by treatment with cyclin A/
45
CDK2. Downregulation of E2F-1/DP-1 complexes by cyclin A/ CDK2 requires the cyclin A
binding domain of E2F-1; transplantation of this domain to the amino-terminus of E2F-4 confers to
E2F-4/DP-1 complexes susceptibility to cyclin A/ CDK2-mediated inhibition (Dynlacht et al.,
1997). Interestingly, although CDK complexes containing cyclin B are capable of phosphorylating
E2F-1/DP-1 complexes on the same residues as cyclin A complexes do, cyclin B cannot bind
stably to E2F-1 and cannot inhibit DNA binding by E2F-1/DP-I complexes (Dynlacht et al.,
1997), indicating that phosphorylation of DP-1, although necessary, is not sufficient to inhibit
DNA binding activity.
The activation of E2F target gene expression at the approach of the GUS transition, then, is
mediated by phosphorylation and release of the pRB-family members and de novo synthesis of
E2F- 1, -2, and -3. The inactivation of E2F target gene expression in S phase is effected by the
relocalization of E2F-4 and -5 to the cytoplasm and the inhibition of E2F-1, -2, and -3 by cyclin A/
CDK2. The importance of this inactivation is underlined by the observation that overexpression of
a form of E2F- 1 lacking the cyclin A-binding domain or a form of DP- 1 which cannot be
phosphorylated delays S phase progression (Krek et al., 1995). Only alleles of E2F-1 that retain
DNA binding activity can induce this delay; alleles also retaining the transcriptional activation
domain induce apoptosis following the S phase delay, inducing apoptosis much more efficiently
than wild-type E2F-1 does. Inactivation of E2F activity may, therefore, be required for efficient
progression through S phase and completion of the cell cycle.
Thus, the E2F genes are not redundant: they have individual functions in addition to any
shared functions, and are regulated in overlapping but distinct ways. This underscores the
importance of characterizing the functions of each member of the E2F gene family. In addition,
antibody experiments indicate that E2F- 1, -2, -3, -4, and -5 are insufficient to account for all of the
E2F DNA binding activity in proliferating cells. The identification of the remaining members of
the E2F family is obviously critical to any complete understanding of the functions of E2F in cell
biology.
46
G.
E2F genes: transcriptional targets
Given that E2F activity is rate-limiting for S phase entry under a variety of conditions, it is
of great interest to identify the transcriptional targets of E2F promoting cellular proliferation. E2F
has been proposed to regulate many genes expressed at the Gl/S transition, such as pRB, p107,
E2F- 1, E2F-2, cyclin E, cyclin A, B-myb, DHFR, DNA polymerase alpha, HsOrc 1, and cdc6.
Both cyclin E and cyclin A are required for the GUS transition, and cdc6 is required for DNA
synthesis, as are, presumably, DHFR, DNA polymerase alpha, and perhaps HsOrc 1. If all of
these genes, as well as others which have not yet been identified, are legitimate and direct E2F
targets, it may be very difficult to establish a target gene or genes tha can substitute for E2F activity
in the promotion of DNA synthesis.
Cyclin E
Cyclin E has, nevertheless, been intensively studied as a candidate for such a role. Cyclin
E levels are induced in pRB-deficient MEFs (Herrera et al., 1996; Hurford et al., 1997) and in
cells overexpressing E2F-1, -2, -3, or -4 (DeGregori et al., 1997), and cyclin E is rate-limiting for
S phase progression in a wide variety of cells. Overexpression of cyclin B can overcome a GI
arrest imposed by overexpression of p1 6, by a form of pRB lacking phosphorylation sites, or by
dominant-negative DP-1, and does so without a significant induction of E2F activity. Although the
lack of E2F activity would probably lead to the eventual depletion of nucleotide pools for DNA
synthesis and thereby prevent further cellular proliferation, overexpression of cyclin E in this
system is apparently sufficient to permit at least one transition past the restriction point (Lukas et
al., 1997). In contrast, studies on irradiated cells, in which elevated levels of CDK inhibitors lead
to the inactivation of cyclin E-associated kinase activity and G 1 arrest, have shown that
overexpression of E2F- 1 can permit entry into S phase without (full) rescue of cyclin E-associated
kinase activity (DeGregori et al., 1995). In this system, then, hyperactivation of other E2F targets
can substitute for activation of cyclin E. While these two findings are difficult to reconcile, it has
been suggested that normally cycling cells may require both cyclin E and another E2F target for
47
timely progression through the restriction point, and that hyperactivation (perhaps unphysiological)
of either can be sufficient for S phase entry in some circumstances (Lukas et al., 1997; Dyson,
1998). These results are reminiscent of data regarding cyclin E and cyclin DI: both are required in
normally cycling cells in tissue culture, overexpression of either can accelerate S phase entry, and
overexpression of both causes a further acceleration.
The relative contributions of E2F activity and cyclin E in the promotion of S phase entry
have also been studied in D. melanogaster. As indicated above, mutations of dE2F or dDP retard
S phase progression, whereas null mutations of cyclin E prevent almost all DNA synthesis
following the first sixteen cell cycles in early Drosophilaembryogenesis (Knoblich et al., 1994).
Cyclin E mRNA levels are highly dependent on dE2F and dDP in the endocycling cells of the
larval gut (Duronio et al., 1995; Royzman et al., 1997; Duronio et al., 1998). In contrast, the
mRNA levels of E2F target genes in the developing Drosophilacentral nervous system depend on
the expression of cyclin E (Duronio et al., 1995). Thus, as in mammals, cyclin E appears to be
both an activator or E2F, perhaps through phosphorylation of the DrosophilaRB homolog RBF,
and a transcriptional target of E2F. In quiescent epidermal cells of the early Drosophilaembryo,
the induction of S phase by ectopic expression of dE2F and dDP is not observed in embryos
lacking functional cyclin E, although the mRNA levels of E2F target genes are efficiently induced,
indicating that the overexpression of dE2F with dDP cannot bypass the S phase requirement for
cyclin E in this tissue (Duronio et al., 1996). Ectopic expression of cyclin E induces S phase in
this tissue even in embryos with a mutation in dE2F (Duronio et al., 1995). This induction,
however, is not reported to occur at a stage in embryogenesis at which the epidermal cells have
been quiescent for a prolonged period of time, suggesting that dE2F transcriptional targets other
than cyclin E may also be important in the promotion of S phase. Thus, in both Drosophilaand
mammals, cyclin E is a very important transcriptional target for E2F-mediated promotion of cell
cycle progression, but does not seem to be the sole important target.
48
cdc6
Another target proposed to participate in the promotion of S phase entry by E2F is cdc6. In
yeast, cdc6 is directly involved in the initiation of DNA synthesis, and the same is believed to be
true in humans. The human cdc6 mRNA is most abundant at the GUS transition, at least in part as
a result of cell cycle-dependent transcription (Leone et al., 1998). The variations in cdc6
transcription depend on two E2F binding sites in the cdc6 promoter (Hateboer et al., 1998; Ohtani
et al., 1998; Yan et al., 1998). Both sites are occupied throughout the cell cycle, and their
mutation causes constitutive, low levels of expression from the promoter, preventing both
repression of the promoter in arrested cells and activation of the promoter in S phase cells.
Microinjection of an antibody to cdc6 prevents entry into S phase, whereas the overexpression of
cdc6 in cycling cells increases the percentage of cells in S phase, as does overexpression of cyclin
E; coexpression of both increases this percentage even further. A more interesting question is
whether or not cdc6 can induce S phase entry in quiescent cells, either alone or in cooperation with
cyclin E, although there are some indications that cdc6 may be unstable in quiescent cells
(Hateboer, 1998).
G1.
Some E2F targets are regulated by repression, some by activation
The existence of non-overlapping functions for individual E2F genes and their differential
regulation by transcription, localization, phosphorylation, stability and associated proteins raises
the possibility that not all E2F target genes will be affected by E2F in the same way. Indeed, some
evidence suggests this may be the case. For example, deletion of the E2F site in the mouse DHFR
promoter abolishes activation of the promoter in S phase without affecting expression levels in
early GI (Means et al., 1992). In contrast, the E2F site in the B-myb promoter is required for
repression in early GI, but not for peak expression in S phase (Lam and Watson, 1993) and the
E2F sites in the cdc6 promoter may be required both for repression and for activation (Hateboer et
al., 1998). The B-myb results are consistent with in vivo footprinting data indicating that its E2F
site is occupied in quiescent fibroblasts, but that this occupation is no longer detectable once the
49
cells enter S phase (Zwicker et al., 1996), and with the derepression of B-myb observed in MEFs
deficient for both p107 and p130. Thus, E2F sites in specific promoters do not all confer the same
regulation to those promoters, although they do seem to share conferral of maximal promoter
activity at the GUS transition.
G2.
Individual E2F genes may regulate specific subsets of transcriptional targets
Overexpression studies are also consistent with individual E2F complexes preferentially
regulating certain promoters. For example, ectopic expression of E2F-1 activates the endogenous
p19-ARF gene in REF52 cells, whereas E2F-2 activates it more weakly and the other E2F genes
do not activate it (DeGregori et al., 1997). In contrast, E2F-2 overexpression is the most efficient
at activating DHFR, E2F-3 is the most efficient activator of CDK2, and E2F- 1, -2, -3, and -4
share the ability to activate cyclin E expression. In many overexpression assays, specificity of
function can be lost; for example, E2F-2 will bind to p107 when overexpressed, although this
interaction is not believed to occur in vivo (Moberg, 1998). That individual E2F genes could have
demonstrably different effects on particular transcriptional targets even when overexpressed
strongly suggests that a given E2F may preferentially regulate only a subset of potential E2F target
genes in vivo. This is also consistent with the differential effects of mutation of pRB or p107 and
p130 on E2F target gene expression, and with the different biological effects of individual E2F
genes. Unraveling which target genes are regulated in which ways (repression, activation, or
both) by which E2F genes, and how this specificity is conferred, remains the subject of intensive
study.
50
H.
Summary
The E2F transcription factors participate in the regulation of a variety of cellular processes,
including progression through GI and apoptosis. Although there are similarities among the E2F
proteins, they are differentially regulated during development and the cell cycle. They also have
distinct functions, as determined by both overexpression studies and mutagenesis studies as well
as other data. As a family of transcription factors, E2F proteins are believed to effect these distinct
functions through the regulation of specific target genes. Indeed, mutagenesis of the E2F
inhibitors pRB or p107 and p130 derepresses distinct sets of E2F target genes, and ectopic
expression of individual E2F proteins has differential effects on specific E2F target genes. The
mechanism permitting this specifity of target gene activation by individual E2F proteins remains
unknown. One hypothesis, that E2F proteins might vary in their DNA-binding specificity, is
explored in CHAPTER II.
The five E2F proteins described in this chapter are known to be insufficient to account for
all of the cellular E2F DNA-binding activity. CHAPTER III describes the identification of a novel
member of the E2F family, E2F-6, and its initial characterization. CHAPTER IV details the
characterization of a zinc-finger protein interacting specifically with E2F-6. Finally, CHAPTER V
discusses the impact of these findings on the larger questions of E2F function and regulation.
5 1
I.
References
Adnane, J., Shao, Z., and Robbins, P. D. (1995). The retinoblastoma susceptibility gene product
represses transcription when directly bound to the promoter. J Biol Chem 270, 8837-43.
Almasan, A., Yin, Y., Kelly, R. E., Lee, E. Y., Bradley, A., Li, W., Bertino, J. R., and Wahl,
G. M. (1995). Deficiency of retinoblastoma protein leads to inappropriate S-phase entry,
activation of E2F-responsive genes, and apoptosis. Proc Natl Acad Sci U S A 92, 5436-40.
Asano, M., Nevins, J. R., and Wharton, R. P. (1996). Ectopic E2F expression induces S phase
and apoptosis in Drosophila imaginal discs. Genes Dev 10, 1422-32.
Bagchi, S., Raychaudhuri, P., and Nevins, J. R. (1990). Adenovirus ElA proteins can
dissociate heteromeric complexes involving the E2F transcription factor: a novel mechanism for
ElA trans- activation. Cell 62, 659-69.
Baldin, V., Lukas, J., Marcote, M. J., Pagano, M., and Draetta, G. (1993). Cyclin DI is a
nuclear protein required for cell cycle progression in G1. Genes Dev 7, 812-21.
Bandara, L. R., Buck, V. M., Zamanian, M., Johnston, L. H., and La Thangue, N. B. (1993).
Functional synergy between DP- 1 and E2F- 1 in the cell cycle-regulating transcription factor
DRTF1/E2F. Embo J 12, 4317-24.
Bandara, L. R., and La Thangue, N. B. (1991). Adenovirus Ela prevents the retinoblastoma
gene product from complexing with a cellular transcription factor. Nature 351, 494-7.
Bartkova, J., Lukas, J., Strauss, M., and Bartek, J. (1998). Cyclin D3: requirement for GI/S
transition and high abundance in quiescent tissues suggest a dual role in proliferation and
differentiation. Oncogene 17, 1027-37.
Bates, S., Phillips, A. C., Clark, P. A., Stott, F., Peters, G., Ludwig, R. L., and Vousden, K.
H. (1998). p14ARF links the tumour suppressors RB and p53 [letter]. Nature 395, 124-5.
Beijersbergen, R. L., Carlee, L., Kerkhoven, R. M., and Bernards, R. (1995). Regulation of the
retinoblastoma protein-related p107 by GI cyclin complexes. Genes Dev 9, 1340-53.
Beijersbergen, R. L., Kerkhoven, R. M., Zhu, L., Carlee, L., Voorhoeve, P. M., and Bernards,
R. (1994). E2F-4, a new member of the E2F gene family, has oncogenic activity and associates
with p107 in vivo. Genes Dev 8, 2680-90.
Bellett, A. J., Jackson, P., David, E. T., Bennett, E. J., and Cronin, B. (1989). Functions of the
two adenovirus early El A proteins and their conserved domains in cell cycle alteration, actin
reorganization, and gene activation in rat cells. J Virol 63, 303-10.
Bookstein, R., Lee, E. Y., Peccei, A., and Lee, W. H. (1989). Human retinoblastoma gene:
long-range mapping and analysis of its deletion in a breast cancer cell line. Mol Cell Biol 9,
1628-34.
Borrow, J., Stanton, V. P., Jr., Andresen, J. M., Becher, R., Behm, F. G., Chaganti, R. S.,
Civin, C. I., Disteche, C., Dube, I., Frischauf, A. M., Horsman, D., Mitelman, F., Volinia, S.,
Watmore, A. E., and Housman, D. E. (1996). The translocation t(8;16)(p Il;pl13) of acute
myeloid leukaemia fuses a putative acetyltransferase to the CREB-binding protein. Nat Genet 14,
33-41.
52
Brehm, A., Miska, E. A., McCance, D. J., Reid, J. L., Bannister, A. J., and Kouzarides, T.
(1998). Retinoblastoma protein recruits histone deacetylase to repress transcription. Nature 391,
597-601.
Bremner, R., Cohen, B. L., Sopta, M., Hamel, P. A., Ingles, C. J., Gallie, B. L., and Phillips,
R. A. (1995). Direct transcriptional repression by pRB and its reversal by specific cyclins. Mol
Cell Biol 15, 3256-65.
Brugarolas, J., Bronson, R. T., and Jacks, T. (1998). p2 1 is a critical CDK2 regulator essential
for proliferation control in Rb-deficient cells. J Cell Biol 141, 503-14.
Cao, L., Faha, B., Dembski, M., Tsai, L. H., Harlow, E., and Dyson, N. (1992). Independent
binding of the retinoblastoma protein and p107 to the transcription factor E2F. Nature 355, 1769.
Castano, E., Kleyner, Y., and Dynlacht, B. D. (1998). Dual cyclin-binding domains are required
for p107 to function as a kinase inhibitor. Mol Cell Biol 18, 5380-91.
Cavenee, W. K., Dryja, T. P., Phillips, R. A., Benedict, W. F., Godbout, R., Gallie, B. L.,
Murphree, A. L., Strong, L. C., and White, R. L. (1983). Expression of recessive alleles by
chromosomal mechanisms in retinoblastoma. Nature 305, 779-84.
Chellappan, S. P., Hiebert, S., Mudryj, M., Horowitz, J. M., and Nevins, J. R. (1991). The
E2F transcription factor is a cellular target for the RB protein. Cell 65, 1053-61.
Christensen, J. B., and Imperiale, M. J. (1995). Inactivation of the retinoblastoma susceptibility
protein is not sufficient for the transforming function of the conserved region 2-like domain of
simian virus 40 large T antigen. J Virol 69, 3945-8.
Cobrinik, D., Lee, M. H., Hannon, G., Mulligan, G., Bronson, R. T., Dyson, N., Harlow, E.,
Beach, D., Weinberg, R. A., and Jacks, T. (1996). Shared role of the pRB-related p130 and
p107 proteins in limb development. Genes Dev 10, 1633-44.
Cobrinik, D., Whyte, P., Peeper, D. S., Jacks, T., and Weinberg, R. A. (1993). Cell cyclespecific association of E2F with the p130 E lA-binding protein. Genes Dev 7, 2392-404.
Dagnino, L., Fry, C. J., Bartley, S. M., Farnham, P., Gallie, B. L., and Phillips, R. A. (1997a).
Expression patterns of the E2F family of transcription factors during mouse nervous system
development. Mech Dev 66, 13-25.
Dagnino, L., Fry, C. J., Bartley, S. M., Farnham, P., Gallie, B. L., and Phillips, R. A.
(1997b). Expression patterns of the E2F family of transcription factors during murine epithelial
development. Cell Growth Differ 8, 553-63.
DeCaprio, J. A., Ludlow, J. W., Figge, J., Shew, J. Y., Huang, C. M., Lee, W. H., Marsilio,
E., Paucha, E., and Livingston, D. M. (1988). SV40 large tumor antigen forms a specific
complex with the product of the retinoblastoma susceptibility gene. Cell 54, 275-83.
DeCaprio, J. A., Ludlow, J. W., Lynch, D., Furukawa, Y., Griffin, J., Piwnica-Worms, H.,
Huang, C. M., and Livingston, D. M. (1989). The product of the retinoblastoma susceptibility
gene has properties of a cell cycle regulatory element. Cell 58, 1085-95.
53
DeGregori, J., Kowalik, T., and Nevins, J. R. (1995). Cellular targets for activation by the
E2Fl transcription factor include DNA synthesis- and GI/S-regulatory genes [published erratum
appears in Mol Cell Biol 1995 Oct;15(10):5846-7]. Mol Cell Biol 15, 4215-24.
DeGregori, J., Leone, G., Miron, A., Jakoi, L., and Nevins, J. R. (1997). Distinct roles for
E2F proteins in cell growth control and apoptosis. Proc Natl Acad Sci U S A 94, 7245-50.
DeGregori, J., Leone, G., Ohtani, K., Miron, A., and Nevins, J. R. (1995). E2F-1
accumulation bypasses a GI arrest resulting from the inhibition of GI cyclin-dependent kinase
activity. Genes Dev 9, 2873-87.
Dirks, P. B., Rutka, J. T., Hubbard, S. L., Mondal, S., and Hamel, P. A. (1998). The E2Ffamily proteins induce distinct cell cycle regulatory factors in p16-arrested, U343 astrocytoma
cells. Oncogene 17, 867-76.
Dobrowolski, S. F., Stacey, D. W., Harter, M. L., Stine, J. T., and Hiebert, S. W. (1994). An
E2F dominant negative mutant blocks EIA induced cell cycle progression. Oncogene 9, 260512.
Dowdy, S. F., Hinds, P. W., Louie, K., Reed, S. I., Arnold, A., and Weinberg, R. A. (1993).
Physical interaction of the retinoblastoma protein with human D cyclins. Cell 73, 499-511.
Dryja, T. P., Cavenee, W., White, R., Rapaport, J. M., Petersen, R., Albert, D. M., and Bruns,
G. A. (1984). Homozygosity of chromosome 13 in retinoblastoma. N Engl J Med 310, 550-3.
Du, W., Vidal, M., Xie, J. E., and Dyson, N. (1996a). RBF, a novel RB-related gene that
regulates E2F activity and interacts with cyclin E in Drosophila. Genes Dev 10, 1206-18.
Du, W., Xie, J. E., and Dyson, N. (1996b). Ectopic expression of dE2F and dDP induces cell
proliferation and death in the Drosophila eye. Embo J 15, 3684-92.
Duronio, R. J., Bonnette, P. C., and O'Farrell, P. H. (1998). Mutations of the Drosophila dDP,
dE2F, and cyclin E genes reveal distinct roles for the E2F-DP transcription factor and cyclin E
during the G1-S transition [published erratum appears in Mol Cell Biol 1998 May;18(5):3 101].
Mol Cell Biol 18, 141-51.
Duronio, R. J., Brook, A., Dyson, N., and O'Farrell, P. H. (1996). E2F-induced S phase
requires cyclin E. Genes Dev 10, 2505-13.
Duronio, R. J., O'Farrell, P. H., Xie, J. E., Brook, A., and Dyson, N. (1995). The
transcription factor E2F is required for S phase during Drosophila embryogenesis. Genes Dev 9,
1445-55.
Dynlacht, B. D., Brook, A., Dembski, M., Yenush, L., and Dyson, N. (1994a). DNA-binding
and trans-activation properties of Drosophila E2F and DP proteins. Proc Natl Acad Sci U S A 91,
6359-63.
Dynlacht, B. D., Flores, 0., Lees, J. A., and Harlow, E. (1994b). Differential regulation of E2F
transactivation by cyclin/cdk2 complexes. Genes Dev 8, 1772-86.
Dynlacht, B. D., Moberg, K., Lees, J. A., Harlow, E., and Zhu, L. (1997). Specific regulation
of E2F family members by cyclin-dependent kinases. Mol Cell Biol 17, 3867-75.
Dyson, N. (1998). The regulation of E2F by pRB-family proteins. Genes Dev 12, 2245-62.
54
Dyson, N., Guida, P., Munger, K., and Harlow, E. (1992). Homologous sequences in
adenovirus E lA and human papillomavirus E7 proteins mediate interaction with the same set of
cellular proteins. J Virol 66, 6893-902.
Dyson, N., Howley, P. M., Munger, K., and Harlow, E. (1989). The human papilloma virus16 E7 oncoprotein is able to bind to the retinoblastoma gene product. Science 243, 934-7.
Eckner, R., Ewen, M. E., Newsome, D., Gerdes, M., DeCaprio, J. A., Lawrence, J. B., and
Livingston, D. M. (1994). Molecular cloning and functional analysis of the adenovirus ElAassociated 300-kD protein (p300) reveals a protein with properties of a transcriptional adaptor.
Genes Dev 8, 869-84.
Ewen, M. E., Sluss, H. K., Sherr, C. J., Matsushime, H., Kato, J., and Livingston, D. M.
(1993). Functional interactions of the retinoblastoma protein with mammalian D- type cyclins.
Cell 73, 487-97.
Ewen, M. E., Xing, Y. G., Lawrence, J. B., and Livingston, D. M. (1991). Molecular cloning,
chromosomal mapping, and expression of the cDNA for p107, a retinoblastoma gene productrelated protein. Cell 66, 1155-64.
Fan, J., and Bertino, J. R. (1997). Functional roles of E2F in cell cycle regulation. Oncogene
14, 1191-200.
Ferreira, R., Magnaghi-Jaulin, L., Robin, P., Harel-Bellan, A., and Trouche, D. (1998). The
three members of the pocket proteins family share the ability to repress E2F activity through
recruitment of a histone deacetylase. Proc Natl Acad Sci U S A 95, 10493-8.
Field, S. J., Tsai, F. Y., Kuo, F., Zubiaga, A. M., Kaelin, W. G., Jr., Livingston, D. M.,
Orkin, S. H., and Greenberg, M. E. (1996). E2F-1 functions in mice to promote apoptosis and
suppress proliferation. Cell 85, 549-61.
Flores, A. M., Kassatly, R. F., and Cress, W. D. (1998). E2F-3 accumulation is regulated by
polypeptide stability. Oncogene 16, 1289-98.
Friend, S. H., Bernards, R., Rogelj, S., Weinberg, R. A., Rapaport, J. M., Albert, D. M., and
Dryja, T. P. (1986). A human DNA segment with properties of the gene that predisposes to
retinoblastoma and osteosarcoma. Nature 323, 643-6.
Friend, S. H., Horowitz, J. M., Gerber, M. R., Wang, X. F., Bogenmann, E., Li, F. P., and
Weinberg, R. A. (1987). Deletions of a DNA sequence in retinoblastomas and mesenchymal
tumors: organization of the sequence and its encoded protein [published erratum appears in Proc
Natl Acad Sci U S A 1988 Apr;85(7):2234]. Proc Natl Acad Sci U S A 84, 9059-63.
Giles, R. H., Peters, D. J., and Breuning, M. H. (1998). Conjunction dysfunction: CBP/p300 in
human disease. Trends Genet 14, 178-83.
Ginsberg, D., Vairo, G., Chittenden, T., Xiao, Z. X., Xu, G., Wydner, K. L., DeCaprio, J. A.,
Lawrence, J. B., and Livingston, D. M. (1994). E2F-4, a new member of the E2F transcription
factor family, interacts with p107. Genes Dev 8, 2665-79.
Girling, R., Partridge, J. F., Bandara, L. R., Burden, N., Totty, N. F., Hsuan, J. J., and La
Thangue, N. B. (1993). A new component of the transcription factor DRTF1/E2F. Nature 365,
468.
55
Guy, C. T., Zhou, W., Kaufman, S., and Robinson, M. 0. (1996). E2F-1 blocks terminal
differentiation and causes proliferation in transgenic megakaryocytes. Mol Cell Biol 16, 685-93.
Haas-Kogan, D. A., Kogan, S. C., Levi, D., Dazin, P., T'Ang, A., Fung, Y. K., and Israel, M.
A. (1995). Inhibition of apoptosis by the retinoblastoma gene product. Embo J 14, 461-72.
Hanna, Z., Jankowski, M., Tremblay, P., Jiang, X., Milatovich, A., Francke, U., and Jolicoeur,
P. (1993). The Vin-1 gene, identified by provirus insertional mutagenesis, is the cyclin D2.
Oncogene 8, 1661-6.
Hannon, G. J., Demetrick, D., and Beach, D. (1993). Isolation of the Rb-related p130 through
its interaction with CDK2 and cyclins. Genes Dev 7, 2378-91.
Harbour, J. W., Lai, S. L., Whang-Peng, J., Gazdar, A. F., Minna, J. D., and Kaye, F. J.
(1988). Abnormalities in structure and expression of the human retinoblastoma gene in SCLC.
Science 241, 353-7.
Harlow, E., Whyte, P., Franza, B. R., Jr., and Schley, C. (1986). Association of adenovirus
early-region IA proteins with cellular polypeptides. Mol Cell Biol 6, 1579-89.
Hateboer, G., Kerkhoven, R. M., Shvarts, A., Bernards, R., and Beijersbergen, R. L. (1996).
Degradation of E2F by the ubiquitin-proteasome pathway: regulation by retinoblastoma family
proteins and adenovirus transforming proteins. Genes Dev 10, 2960-70.
Hateboer, G., Wobst, A., Petersen, B. 0., Le Cam, L., Vigo, E., Sardet, C., and Helin, K.
(1998). Cell cycle-regulated expression of mammalian CDC6 is dependent on E2F. Mol Cell Biol
18, 6679-97.
Haupt, Y., Rowan, S., and Oren, M. (1995). p53-mediated apoptosis in HeLa cells can be
overcome by excess pRB. Oncogene 10, 1563-71.
He, J., Allen, J. R., Collins, V. P., Allalunis-Turner, M. J., Godbout, R., Day, R. S. r., and
James, C. D. (1994). CDK4 amplification is an alternative mechanism to p16 gene homozygous
deletion in glioma cell lines. Cancer Res 54, 5804-7.
Helin, K., Holm, K., Niebuhr, A., Eiberg, H., Tommerup, N., Hougaard, S., Poulsen, H. S.,
Spang-Thomsen, M., and Norgaard, P. (1997). Loss of the retinoblastoma protein-related p130
protein in small cell lung carcinoma. Proc Natl Acad Sci U S A 94, 6933-8.
Helin, K., Lees, J. A., Vidal, M., Dyson, N., Harlow, E., and Fattaey, A. (1992). A cDNA
encoding a pRB-binding protein with properties of the transcription factor E2F. Cell 70, 337-50.
Helin, K., Wu, C. L., Fattaey, A. R., Lees, J. A., Dynlacht, B. D., Ngwu, C., and Harlow, E.
(1993). Heterodimerization of the transcription factors E2F- 1 and DP- 1 leads to cooperative transactivation. Genes Dev 7, 1850-61.
Henderson, D. S., Banga, S. S., Grigliatti, T. A., and Boyd, J. B. (1994). Mutagen sensitivity
and suppression of position-effect variegation result from mutations in mus209, the Drosophila
gene encoding PCNA. Embo J 13, 1450-9.
Herrera, R. E., Sah, V. P., Williams, B. 0., Makela, T. P., Weinberg, R. A., and Jacks, T.
(1996). Altered cell cycle kinetics, gene expression, and GI restriction point regulation in Rbdeficient fibroblasts. Mol Cell Biol 16, 2402-7.
56
Herwig, S., and Strauss, M. (1997). The retinoblastoma protein: a master regulator of cell cycle,
differentiation and apoptosis. Eur J Biochem 246, 581-601.
Hiebert, S. W. (1993). Regions of the retinoblastoma gene product required for its interaction
with the E2F transcription factor are necessary for E2 promoter repression and pRb-mediated
growth suppression. Mol Cell Biol 13, 3384-91.
Hiebert, S. W., Chellappan, S. P., Horowitz, J. M., and Nevins, J. R. (1992). The interaction
of RB with E2F coincides with an inhibition of the transcriptional activity of E2F. Genes Dev 6,
177-85.
Hijmans, E. M., Voorhoeve, P. M., Beijersbergen, R. L., van 't Veer, L. J., and Bernards, R.
(1995). E2F-5, a new E2F family member that interacts with p130 in vivo. Mol Cell Biol 15,
3082-9.
Hinds, P. W., Mittnacht, S., Dulic, V., Arnold, A., Reed, S. I., and Weinberg, R. A. (1992).
Regulation of retinoblastoma protein functions by ectopic expression of human cyclins. Cell 70,
993-1006.
Hofmann, F., and Livingston, D. M. (1996). Differential effects of cdk2 and cdk3 on the control
of pRb and E2F function during G 1 exit. Genes Dev 10, 851-61.
Hofmann, F., Martelli, F., Livingston, D. M., and Wang, Z. (1996). The retinoblastoma gene
product protects E2F- 1 from degradation by the ubiquitin-proteasome pathway. Genes Dev 10,
2949-59.
Holmberg, C., Helin, K., Sehested, M., and Karlstrom, 0. (1998). E2F-1-induced p53independent apoptosis in transgenic mice. Oncogene 17, 143-55.
Horowitz, J. M., Park, S. H., Bogenmann, E., Cheng, J. C., Yandell, D. W., Kaye, F. J.,
Minna, J. D., Dryja, T. P., and Weinberg, R. A. (1990). Frequent inactivation of the
retinoblastoma anti-oncogene is restricted to a subset of human tumor cells. Proc Natl Acad Sci U
S A 87, 2775-9.
Howe, J. A., Mymryk, J. S., Egan, C., Branton, P. E., and Bayley, S. T. (1990).
Retinoblastoma growth suppressor and a 300-kDa protein appear to regulate cellular DNA
synthesis. Proc Natl Acad Sci U S A 87, 5883-7.
Hsiao, K. M., McMahon, S. L., and Farnham, P. J. (1994). Multiple DNA elements are
required for the growth regulation of the mouse E2F1 promoter. Genes Dev 8, 1526-37.
Hsieh, J. K., Fredersdorf, S., Kouzarides, T., Martin, K., and Lu, X. (1997). E2F1-induced
apoptosis requires DNA binding but not transactivation and is inhibited by the retinoblastoma
protein through direct interaction. Genes Dev 11, 1840-52.
Hu, Q. J., Dyson, N., and Harlow, E. (1990). The regions of the retinoblastoma protein needed
for binding to adenovirus ElA or SV40 large T antigen are common sites for mutations. Embo J
9, 1147-55.
Huang, S., Wang, N. P., Tseng, B. Y., Lee, W. H., and Lee, E. H. (1990). Two distinct and
frequently mutated regions of retinoblastoma protein are required for binding to SV40 T antigen.
Embo J 9, 1815-22.
57
Hurford, R. K., Jr., Cobrinik, D., Lee, M. H., and Dyson, N. (1997). pRB and p107/p130 are
required for the regulated expression of different sets of E2F responsive genes. Genes Dev 11,
1447-63.
Hussussian, C. J., Struewing, J. P., Goldstein, A. M., Higgins, P. A., Ally, D. S., Sheahan, M.
D., Clark, W. H., Jr., Tucker, M. A., and Dracopoli, N. C. (1994). Germline p16 mutations in
familial melanoma. Nat Genet 8, 15-21.
Ikeda, M. A., Jakoi, L., and Nevins, J. R. (1996). A unique role for the Rb protein in
controlling E2F accumulation during cell growth and differentiation. Proc Natl Acad Sci U S A
93, 3215-20.
Ishizaki, J., Nevins, J. R., and Sullenger, B. A. (1996). Inhibition of cell proliferation by an
RNA ligand that selectively blocks E2F function. Nat Med 2, 1386-9.
Ivey-Hoyle, M., Conroy, R., Huber, H. E., Goodhart, P. J., Oliff, A., and Heimbrook, D. C.
(1993). Cloning and characterization of E2F-2, a novel protein with the biochemical properties of
transcription factor E2F. Mol Cell Biol 13, 7802-12.
Jacks, T., Fazeli, A., Schmitt, E. M., Bronson, R. T., Goodell, M. A., and Weinberg, R. A.
(1992). Effects of an Rb mutation in the mouse. Nature 359, 295-300.
Jiang, W., Kahn, S. M., Zhou, P., Zhang, Y. J., Cacace, A. M., Infante, A. S., Doi, S.,
Santella, R. M., and Weinstein, I. B. (1993). Overexpression of cyclin DI in rat fibroblasts
causes abnormalities in growth control, cell cycle progression and gene expression. Oncogene 8,
3447-57.
Johnson, D. G., Cress, W. D., Jakoi, L., and Nevins, J. R. (1994a). Oncogenic capacity of the
E2F1 gene. Proc Natl Acad Sci U S A 91, 12823-7.
Johnson, D. G., Ohtani, K., and Nevins, J. R. (1994b). Autoregulatory control of E2F1
expression in response to positive and negative regulators of cell cycle progression. Genes Dev 8,
1514-25.
Johnson, D. G., Schwarz, J. K., Cress, W. D., and Nevins, J. R. (1993). Expression of
transcription factor E2F 1 induces quiescent cells to enter S phase. Nature 365, 349-52.
Jones, N. (1995). Transcriptional modulation by the adenovirus ElA gene. Curr Top Microbiol
Immunol 199, 59-80.
Jooss, K., Lam, E. W., Bybee, A., Girling, R., Muller, R., and La Thangue, N. B. (1995).
Proto-oncogenic properties of the DP family of proteins. Oncogene 10, 1529-36.
Kaelin, W. G., Jr., Ewen, M. E., and Livingston, D. M. (1990). Definition of the minimal
simian virus 40 large T antigen- and adenovirus ElA-binding domain in the retinoblastoma gene
product. Mol Cell Biol 10, 3761-9.
Kaelin, W. G., Jr., Krek, W., Sellers, W. R., DeCaprio, J. A., Ajchenbaum, F., Fuchs, C. S.,
Chittenden, T., Li, Y., Farnham, P. J., Blanar, M. A., and et al. (1992). Expression cloning of a
cDNA encoding a retinoblastoma-binding protein with E2F-like properties. Cell 70, 351-64.
Kamb, A., Gruis, N. A., Weaver-Feldhaus, J., Liu, Q., Harshman, K., Tavtigian, S. V.,
Stockert, E., Day, R. S. r., Johnson, B. E., and Skolnick, M. H. (1994). A cell cycle regulator
potentially involved in genesis of many tumor types. Science 264, 436-40.
58
Kamijo, T., Zindy, F., Roussel, M. F., Quelle, D. E., Downing, J. R., Ashmun, R. A.,
Grosveld, G., and Sherr, C. J. (1997). Tumor suppression at the mouse INK4a locus mediated
by the alternative reading frame product p19ARF. Cell 91, 649-59.
Kato, J. Y., and Sherr, C. J. (1993). Inhibition of granulocyte differentiation by GI cyclins D2
and D3 but not D1. Proc Natl Acad Sci U S A 90, 11513-7.
Kitagawa, M., Higashi, H., Jung, H. K., Suzuki-Takahashi, I., Ikeda, M., Tamai, K., Kato, J.,
Segawa, K., Yoshida, E., Nishimura, S., and Taya, Y. (1996). The consensus motif for
phosphorylation by cyclin D1-Cdk4 is different from that for phosphorylation by cyclin A/E-Cdk2.
Embo J 15, 7060-9.
Kitagawa, M., Higashi, H., Suzuki-Takahashi, I., Segawa, K., Hanks, S. K., Taya, Y.,
Nishimura, S., and Okuyama, A. (1995). Phosphorylation of E2F-1 by cyclin A-cdk2.
Oncogene 10, 229-36.
Knoblich, J. A., Sauer, K., Jones, L., Richardson, H., Saint, R., and Lehner, C. F. (1994).
Cyclin E controls S phase progression and its down-regulation during Drosophila embryogenesis
is required for the arrest of cell proliferation. Cell 77, 107-20.
Knudsen, E. S., and Wang, J. Y. (1997). Dual mechanisms for the inhibition of E2F binding to
RB by cyclin- dependent kinase-mediated RB phosphorylation. Mol Cell Biol 17, 5771-83.
Knudson, A. G., Jr. (1971). Mutation and cancer: statistical study of retinoblastoma. Proc Natl
Acad Sci U S A 68, 820-3.
Kovesdi, I., Reichel, R., and Nevins, J. R. (1986a). ElA transcription induction: enhanced
binding of a factor to upstream promoter sequences. Science 231, 719-22.
Kovesdi, I., Reichel, R., and Nevins, J. R. (1986b). Identification of a cellular transcription
factor involved in ElA trans- activation. Cell 45, 219-28.
Kovesdi, I., Reichel, R., and Nevins, J. R. (1987). Role of an adenovirus E2 promoter binding
factor in E lA-mediated coordinate gene control. Proc Natl Acad Sci U S A 84, 2180-4.
Kowalik, T. F., DeGregori, J., Schwarz, J. K., and Nevins, J. R. (1995). E2F1 overexpression
in quiescent fibroblasts leads to induction of cellular DNA synthesis and apoptosis. J Virol 69,
2491-500.
Kranenburg, 0., van der Eb, A. J., and Zantema, A. (1996). Cyclin DI is an essential mediator
of apoptotic neuronal cell death. Embo J 15, 46-54.
Krek, W., Ewen, M. E., Shirodkar, S., Arany, Z., Kaelin, W. G., Jr., and Livingston, D. M.
(1994). Negative regulation of the growth-promoting transcription factor E2F-1 by a stably bound
cyclin A-dependent protein kinase. Cell 78, 161-72.
Krek, W., Livingston, D. M., and Shirodkar, S. (1993). Binding to DNA and the retinoblastoma
gene product promoted by complex formation of different E2F family members. Science 262,
1557-60.
Krek, W., Xu, G., and Livingston, D. M. (1995). Cyclin A-kinase regulation of E2F-1 DNA
binding function underlies suppression of an S phase checkpoint. Cell 83, 1149-58.
59
Lam, E. W., Bennett, J. D., and Watson, R. J. (1995). Cell-cycle regulation of human B-myb
transcription. Gene 160, 277-81.
Lam, E. W., Morris, J. D., Davies, R., Crook, T., Watson, R. J., and Vousden, K. H. (1994).
HPV16 E7 oncoprotein deregulates B-myb expression: correlation with targeting of p107/E2F
complexes. Embo J 13, 871-8.
Lam, E. W., and Watson, R. J. (1993). An E2F-binding site mediates cell-cycle regulated
repression of mouse B- myb transcription. Embo J 12, 2705-13.
LeCouter, J. E., Kablar, B., Hardy, W. R., Ying, C., Megeney, L. A., May, L. L., and
Rudnicki, M. A. (1998a). Strain-dependent myeloid hyperplasia, growth deficiency, and
accelerated cell cycle in mice lacking the Rb-related p107 gene. Mol Cell Biol 18, 7455-65.
LeCouter, J. E., Kablar, B., Whyte, P. F., Ying, C., and Rudnicki, M. A. (1998b). Straindependent embryonic lethality in mice lacking the retinoblastoma- related p130 gene. Development
125, 4669-4679.
Lee, E. Y., To, H., Shew, J. Y., Bookstein, R., Scully, P., and Lee, W. H. (1988).
Inactivation of the retinoblastoma susceptibility gene in human breast cancers. Science 241, 21821.
Lee, M. H., Williams, B. 0., Mulligan, G., Mukai, S., Bronson, R. T., Dyson, N., Harlow, E.,
and Jacks, T. (1996). Targeted disruption of p107: functional overlap between p107 and Rb.
Genes Dev 10, 1621-32.
Lee, W. H., Shew, J. Y., Hong, F. D., Sery, T. W., Donoso, L. A., Young, L. J., Bookstein,
R., and Lee, E. Y. (1987). The retinoblastoma susceptibility gene encodes a nuclear
phosphoprotein associated with DNA binding activity. Nature 329, 642-5.
Lees, J. A., Saito, M., Vidal, M., Valentine, M., Look, T., Harlow, E., Dyson, N., and Helin,
K. (1993). The retinoblastoma protein binds to a family of E2F transcription factors. Mol Cell
Biol 13, 7813-25.
Leone, G., DeGregori, J., Yan, Z., Jakoi, L., Ishida, S., Williams, R. S., and Nevins, J. R.
(1998). E2F3 activity is regulated during the cell cycle and is required for the induction of S
phase. Genes Dev 12, 2120-30.
Li, Y., Graham, C., Lacy, S., Duncan, A. M., and Whyte, P. (1993). The adenovirus ElAassociated 130-kD protein is encoded by a member of the retinoblastoma gene family and
physically interacts with cyclins A and E. Genes Dev 7, 2366-77.
Lillie, J. W., Green, M., and Green, M. R. (1986). An adenovirus Ela protein region required
for transformation and transcriptional repression. Cell 46, 1043-51.
Lillie, J. W., Loewenstein, P. M., Green, M. R., and Green, M. (1987). Functional domains of
adenovirus type 5 E la proteins. Cell 50, 1091-100.
Lindeman, G. J., Dagnino, L., Gaubatz, S., Xu, Y., Bronson, R. T., Warren, H. B., and
Livingston, D. M. (1998). A specific, nonproliferative role for E2F-5 in choroid plexus function
revealed by gene targeting. Genes Dev 12, 1092-8.
Lindeman, G. J., Gaubatz, S., Livingston, D. M., and Ginsberg, D. (1997). The subcellular
localization of E2F-4 is cell-cycle dependent. Proc Natl Acad Sci U S A 94, 5095-100.
60
Lucher, L. A. (1995). Abortive adenovirus infection and host range determinants. Curr Top
Microbiol Immunol 199, 119-52.
Lukas, J., Bartkova, J., Rohde, M., Strauss, M., and Bartek, J. (1995a). Cyclin DI is
dispensable for GI control in retinoblastoma gene- deficient cells independently of cdk4 activity.
Mol Cell Biol 15, 2600-11.
Lukas, J., Bartkova, J., Welcker, M., Petersen, 0. W., Peters, G., Strauss, M., and Bartek, J.
(1995b). Cyclin D2 is a moderately oscillating nucleoprotein required for GI phase progression in
specific cell types. Oncogene 10, 2125-34.
Lukas, J., Herzinger, T., Hansen, K., Moroni, M. C., Resnitzky, D., Helin, K., Reed, S. I., and
Bartek, J. (1997). Cyclin E-induced S phase without activation of the pRb/E2F pathway. Genes
Dev 11, 1479-92.
Lukas, J., Pagano, M., Staskova, Z., Draetta, G., and Bartek, J. (1994). Cyclin DI protein
oscillates and is essential for cell cycle progression in human tumour cell lines. Oncogene 9, 70718.
Lukas, J., Parry, D., Aagaard, L., Mann, D. J., Bartkova, J., Strauss, M., Peters, G., and
Bartek, J. (1995). Retinoblastoma-protein-dependent cell-cycle inhibition by the tumour
suppressor p16. Nature 375, 503-6.
Lukas, J., Petersen, B. 0., Holm, K., Bartek, J., and Helin, K. (1996). Deregulated expression
of E2F family members induces S-phase entry and overcomes p16INK4A-mediated growth
suppression. Mol Cell Biol 16, 1047-57.
Lundberg, A. S., and Weinberg, R. A. (1998). Functional inactivation of the retinoblastoma
protein requires sequential modification by at least two distinct cyclin-cdk complexes. Mol Cell
Biol 18, 753-61.
Lundblad, J. R., Kwok, R. P., Laurance, M. E., Harter, M. L., and Goodman, R. H. (1995).
Adenoviral E lA-associated protein p300 as a functional homologue of the transcriptional coactivator CBP. Nature 374, 85-8.
Luo, R. X., Postigo, A. A., and Dean, D. C. (1998). Rb interacts with histone deacetylase to
repress transcription. Cell 92, 463-73.
Macleod, K. F., Hu, Y., and Jacks, T. (1996). Loss of Rb activates both p53-dependent and
independent cell death pathways in the developing mouse nervous system. Embo J 15, 6178-88.
Magnaghi-Jaulin, L., Groisman, R., Naguibneva, I., Robin, P., Lorain, S., Le Villain, J. P.,
Troalen, F., Trouche, D., and Harel-Bellan, A. (1998). Retinoblastoma protein represses
transcription by recruiting a histone deacetylase. Nature 391, 601-5.
Mann, D. J., and Jones, N. C. (1996). E2F-l but not E2F-4 can overcome p16-induced GI cellcycle arrest. Curr Biol 6, 474-83.
Matsushime, H., Roussel, M. F., Ashmun, R. A., and Sherr, C. J. (1991). Colony-stimulating
factor 1 regulates novel cyclins during the GI phase of the cell cycle. Cell 65, 701-13.
61
Mayol, X., Grana, X., Baldi, A., Sang, N., Hu, Q., and Giordano, A. (1993). Cloning of a
new member of the retinoblastoma gene family (pRb2) which binds to the ElA transforming
domain. Oncogene 8, 2561-6.
Means, A. L., Slansky, J. E., McMahon, S. L., Knuth, M. W., and Farnham, P. J. (1992). The
HIPI binding site is required for growth regulation of the dihydrofolate reductase gene promoter.
Mol Cell Biol 12, 1054-63.
Melillo, R. M., Helin, K., Lowy, D. R., and Schiller, J. T. (1994). Positive and negative
regulation of cell proliferation by E2F- 1: influence of protein level and human papillomavirus
oncoproteins. Mol Cell Biol 14, 8241-9.
Miller, R. W., and Rubinstein, J. H. (1995). Tumors in Rubinstein-Taybi syndrome. Am J Med
Genet 56, 112-5.
Moberg, K. Molecular Analysis of the Role of E2F Proteins in the pRB Pathway.
Massachusetts Institute of Technology, 1998.
Moberg, K., Starz, M. A., and Lees, J. A. (1996). E2F-4 switches from p130 to p107 and pRB
in response to cell cycle reentry. Mol Cell Biol 16, 1436-49.
Moran, E., Zerler, B., Harrison, T. M., and Mathews, M. B. (1986). Identification of separate
domains in the adenovirus ElA gene for immortalization activity and the activation of virus early
genes. Mol Cell Biol 6, 3470-80.
Morgenbesser, S. D., Williams, B. 0., Jacks, T., and DePinho, R. A. (1994). p53-dependent
apoptosis produced by Rb-deficiency in the developing mouse lens. Nature 371, 72-4.
Motokura, T., Bloom, T., Kim, H. G., Juppner, H., Ruderman, J. V., Kronenberg, H. M., and
Arnold, A. (1991). A novel cyclin encoded by a bcll-linked candidate oncogene. Nature 350,
512-5.
Muller, H., Moroni, M. C., Vigo, E., Petersen, B. 0., Bartek, J., and Helin, K. (1997).
Induction of S-phase entry by E2F transcription factors depends on their nuclear localization. Mol
Cell Biol 17, 5508-20.
Mulligan, G. J., Wong, J., and Jacks, T. (1998). p 130 is dispensable in peripheral T
lymphocytes: evidence for functional compensation by p107 and pRB. Mol Cell Biol 18, 20620.
Muraoka, M., Konishi, M., Kikuchi-Yanoshita, R., Tanaka, K., Shitara, N., Chong, J. M.,
Iwama, T., and Miyaki, M. (1996). p3 0 0 gene alterations in colorectal and gastric carcinomas.
Oncogene 12, 1565-9.
Neufeld, T. P., de la Cruz, A. F., Johnston, L. A., and Edgar, B. A. (1998). Coordination of
growth and cell division in the Drosophila wing. Cell 93, 1183-93.
Neuman, E., Flemington, E. K., Sellers, W. R., and Kaelin, W. G., Jr. (1994). Transcription
of the E2F- 1 gene is rendered cell cycle dependent by E2F DNA-binding sites within its promoter
[published erratum appears in Mol Cell Biol 1995 Aug;15(8):4660]. Mol Cell Biol 14, 6607-15.
Ohtani, K., and Nevins, J. R. (1994). Functional properties of a Drosophila homolog of the
E2F1 gene. Mol Cell Biol 14, 1603-12.
62
Ohtani, K., Tsujimoto, A., Ikeda, M., and Nakamura, M. (1998). Regulation of cell growthdependent expression of mammalian CDC6 gene by the cell cycle transcription factor E2F.
Oncogene 17, 1777-85.
Ohtsubo, M., Theodoras, A. M., Schumacher, J., Roberts, J. M., and Pagano, M. (1995).
Human cyclin E, a nuclear protein essential for the G1-to-S phase transition. Mol Cell Biol 15,
2612-24.
Ormondroyd, E., de la Luna, S., and La Thangue, N. B. (1995). A new member of the DP
family, DP-3, with distinct protein products suggests a regulatory role for alternative splicing in the
cell cycle transcription factor DRTF1/E2F. Oncogene 11, 1437-46.
Otterson, G. A., Kratzke, R. A., Coxon, A., Kim, Y. W., and Kaye, F. J. (1994). Absence of
p1 61NK4 protein is restricted to the subset of lung cancer lines that retains wildtype RB.
Oncogene 9, 3375-8.
Pan, H., Yin, C., Dyson, N. J., Harlow, E., Yamasaki, L., and Dyke, T. V. (1998). Key roles
for E2F1 in signaling p53-dependent apoptosis and in cell division within developing tumors. Mol
Cell 2, 283-92.
Pardee, A. B. (1989). GI events and regulation of cell proliferation. Science 246, 603-8.
Pearson, B. E., Nasheuer, H. P., and Wang, T. S. (1991). Human DNA polymerase alpha gene:
sequences controlling expression in cycling and serum-stimulated cells. Mol Cell Biol 11, 208195.
Philipson, L. (1995). Adenovirus--an eternal archetype. Curr Top Microbiol Immunol 199, 124.
Phillips, A. C., Bates, S., Ryan, K. M., Helin, K., and Vousden, K. H. (1997). Induction of
DNA synthesis and apoptosis are separable functions of E2F- 1. Genes Dev 11, 1853-63.
Pierce, A. M., Gimenez-Conti, I. B., Schneider-Broussard, R., Martinez, L. A., Conti, C. J.,
and Johnson, D. G. (1998a). Increased E2F1 activity induces skin tumors in mice heterozygous
and nullizygous for p53. Proc Natl Acad Sci U S A 95, 8858-63.
Pierce, A. M., Schneider-Broussard, R., Philhower, J. L., and Johnson, D. G. (1998b).
Differential activities of E2F family members: unique functions in regulating transcription. Mol
Carcinog 22, 190-8.
Pomerantz, J., Schreiber-Agus, N., Liegeois, N. J., Silverman, A., Alland, L., Chin, L., Potes,
J., Chen, K., Orlow, I., Lee, H. W., Cordon-Cardo, C., and DePinho, R. A. (1998). The Ink4a
tumor suppressor gene product, pl9Arf, interacts with MDM2 and neutralizes MDM2's inhibition
of p53. Cell 92, 713-23.
Qian, Y., Luckey, C., Horton, L., Esser, M., and Templeton, D. J. (1992). Biological function
of the retinoblastoma protein requires distinct domains for hyperphosphorylation and transcription
factor binding. Mol Cell Biol 12, 5363-72.
Qin, X. Q., Chittenden, T., Livingston, D. M., and Kaelin, W. G., Jr. (1992). Identification of
a growth suppression domain within the retinoblastoma gene product. Genes Dev 6, 953-64.
63
Qin, X. Q., Livingston, D. M., Ewen, M., Sellers, W. R., Arany, Z., and Kaelin, W. G., Jr.
(1995). The transcription factor E2F-1 is a downstream target of RB action. Mol Cell Biol 15,
742-55.
Qin, X. Q., Livingston, D. M., Kaelin, W. G., Jr., and Adams, P. D. (1994). Deregulated
transcription factor E2F- 1 expression leads to S-phase entry and p53-mediated apoptosis. Proc
Natl Acad Sci U S A 91, 10918-22.
Quelle, D. E., Ashmun, R. A., Shurtleff, S. A., Kato, J. Y., Bar-Sagi, D., Roussel, M. F., and
Sherr, C. J. (1993). Overexpression of mouse D-type cyclins accelerates GI phase in rodent
fibroblasts. Genes Dev 7, 1559-71.
Quelle, D. E., Cheng, M., Ashmun, R. A., and Sherr, C. J. (1997). Cancer-associated
mutations at the INK4a locus cancel cell cycle arrest by p16INK4a but not by the alternative
reading frame protein p19ARF. Proc Natl Acad Sci U S A 94, 669-73.
Quelle, D. E., Zindy, F., Ashmun, R. A., and Sherr, C. J. (1995). Alternative reading frames of
the INK4a tumor suppressor gene encode two unrelated proteins capable of inducing cell cycle
arrest. Cell 83, 993-1000.
Resnitzky, D., Gossen, M., Bujard, H., and Reed, S. I. (1994). Acceleration of the GUS phase
transition by expression of cyclins D1 and E with an inducible system. Mol Cell Biol 14, 166979.
Resnitzky, D., and Reed, S. I. (1995). Different roles for cyclins DI and E in regulation of the
G1-to-S transition. Mol Cell Biol 15, 3463-9.
Rogers, K. T., Higgins, P. D., Milla, M. M., Phillips, R. S., and Horowitz, J. M. (1996). DP2, a heterodimeric partner of E2F: identification and characterization of DP-2 proteins expressed in
vivo. Proc Natl Acad Sci U S A 93, 7594-9.
Royzman, I., Whittaker, A. J., and Orr-Weaver, T. L. (1997). Mutations in Drosophila DP and
E2F distinguish GI-S progression from an associated transcriptional program. Genes Dev 11,
1999-2011.
Samuelson, A. V., and Lowe, S. W. (1997). Selective induction of p53 and chemosensitivity in
RB-deficient cells by ElA mutants unable to bind the RB-related proteins. Proc Natl Acad Sci U S
A 94, 12094-9.
Sardet, C., Vidal, M., Cobrinik, D., Geng, Y., Onufryk, C., Chen, A., and Weinberg, R. A.
(1995). E2F-4 and E2F-5, two members of the E2F family, are expressed in the early phases of
the cell cycle. Proc Natl Acad Sci U S A 92, 2403-7.
Sawado, T., Yamaguchi, M., Nishimoto, Y., Ohno, K., Sakaguchi, K., and Matsukage, A.
(1998). dE2F2, a novel E2F-family transcription factor in Drosophila melanogaster. Biochem
Biophys Res Commun 251, 409-15.
Schmidt, E. E., Ichimura, K., Reifenberger, G., and Collins, V. P. (1994). CDKN2
(p16/MTS 1) gene deletion or CDK4 amplification occurs in the majority of glioblastomas. Cancer
Res 54, 6321-4.
Schwarz, J. K., Bassing, C. H., Kovesdi, I., Datto, M. B., Blazing, M., George, S., Wang, X.
F., and Nevins, J. R. (1995). Expression of the E2F1 transcription factor overcomes type beta
transforming growth factor-mediated growth suppression. Proc Natl Acad Sci U S A 92, 483-7.
64
Sears, R., Ohtani, K., and Nevins, J. R. (1997). Identification of positively and negatively
acting elements regulating expression of the E2F2 gene in response to cell growth signals. Mol
Cell Biol 17, 5227-35.
Sellers, W. R., and Kaelin, W. G. (1996). RB [corrected] as a modulator of transcription
[published erratum appears in Biochim Biophys Acta 1996 Dec 9;1288(3):E-1]. Biochim Biophys
Acta 1288, Mi-5.
Shan, B., Chang, C. Y., Jones, D., and Lee, W. H. (1994). The transcription factor E2F-1
mediates the autoregulation of RB gene expression. Mol Cell Biol 14, 299-309.
Shan, B., Durfee, T., and Lee, W. H. (1996). Disruption of RB/E2F-1 interaction by single
point mutations in E2F- 1 enhances S-phase entry and apoptosis. Proc Natl Acad Sci U S A 93,
679-84.
Shan, B., and Lee, W. H. (1994). Deregulated expression of E2F-1 induces S-phase entry and
leads to apoptosis. Mol Cell Biol 14, 8166-73.
Shan, B., Zhu, X., Chen, P. L., Durfee, T., Yang, Y., Sharp, D., and Lee, W. H. (1992).
Molecular cloning of cellular genes encoding retinoblastoma-associated proteins: identification of a
gene with properties of the transcription factor E2F. Mol Cell Biol 12, 5620-31.
Sherr, C. J. (1994). GI phase progression: cycling on cue. Cell 79, 551-5.
Shirodkar, S., Ewen, M., DeCaprio, J. A., Morgan, J., Livingston, D. M., and Chittenden, T.
(1992). The transcription factor E2F interacts with the retinoblastoma product and a p107-cyclin A
complex in a cell cycle-regulated manner. Cell 68, 157-66.
Slansky, J. E., Li, Y., Kaelin, W. G., and Farnham, P. J. (1993). A protein synthesisdependent increase in E2F1 mRNA correlates with growth regulation of the dihydrofolate
reductase promoter [published erratum appears in Mol Cell Biol 1993 Nov;13(11):7201]. Mol Cell
Biol 13, 1610-8.
Sofer-Levi, Y., and Resnitzky, D. (1996). Apoptosis induced by ectopic expression of cyclin DI
but not cyclin E. Oncogene 13, 2431-7.
Souza, R. F., Yin, J., Smolinski, K. N., Zou, T. T., Wang, S., Shi, Y. Q., Rhyu, M. G.,
Cottrell, J., Abraham, J. M., Biden, K., Simms, L., Leggett, B., Bova, G. S., Frank, T.,
Powell, S. M., Sugimura, H., Young, J., Harpaz, N., Shimizu, K., Matsubara, N., and Meltzer,
S. J. (1997). Frequent mutation of the E2F-4 cell cycle gene in primary human gastrointestinal
tumors. Cancer Res 57, 2350-3.
T'Ang, A., Varley, J. M., Chakraborty, S., Murphree, A. L., and Fung, Y. K. (1988).
Structural rearrangement of the retinoblastoma gene in human breast carcinoma. Science 242,
263-6.
Takimoto, H., Tsukuda, K., Ichimura, K., Hanafusa, H., Nakamura, A., Oda, M., Harada, M.,
and Shimizu, K. (1998). Genetic alterations in the retinoblastoma protein-related p107 gene in
human hematologic malignancies. Biochem Biophys Res Commun 251, 264-8.
Toyoshima, H., and Hunter, T. (1994). p27, a novel inhibitor of GI cyclin-Cdk protein kinase
activity, is related to p21. Cell 78, 67-74.
65
Tsai, K. Y., Hu, Y., Macleod, K. F., Crowley, D., Yamasaki, L., and Jacks, T. (1998).
Mutation of E2f- 1 suppresses apoptosis and inappropriate S phase entry and extends survival of
Rb-deficient mouse embryos. Mol Cell 2, 293-304.
Varley, J. M., Armour, J., Swallow, J. E., Jeffreys, A. J., Ponder, B. A., T'Ang, A., Fung, Y.
K., Brammar, W. J., and Walker, R. A. (1989). The retinoblastoma gene is frequently altered
leading to loss of expression in primary breast tumours [published erratum appears in Oncogene
1990 Feb;5(2):245]. Oncogene 4, 725-9.
Verona, R., Moberg, K., Estes, S., Starz, M., Vernon, J. P., and Lees, J. A. (1997). E2F
activity is regulated by cell cycle-dependent changes in subcellular localization. Mol Cell Biol 17,
7268-82.
Wade, M., Blake, M. C., Jambou, R. C., Helin, K., Harlow, E., and Azizkhan, J. C. (1995).
An inverted repeat motif stabilizes binding of E2F and enhances transcription of the dihydrofolate
reductase gene. J Biol Chem 270, 9783-91.
Wade, M., Kowalik, T. F., Mudryj, M., Huang, E. S., and Azizkhan, J. C. (1992). E2F
mediates dihydrofolate reductase promoter activation and multiprotein complex formation in human
cytomegalovirus infection. Mol Cell Biol 12, 4364-74.
Walsh, T., Chappell, S. A., Shaw, J. A., and Walker, R. A. (1998). Microsatellite instability in
ductal carcinoma in situ of the breast. J Pathol 185, 18-24.
Wang, J., and Walsh, K. (1996). Resistance to apoptosis conferred by Cdk inhibitors during
myocyte differentiation. Science 273, 359-61.
Weinberg, R. A. (1995). The retinoblastoma protein and cell cycle control. Cell 81, 323-30.
Weintraub, S. J., Chow, K. N., Luo, R. X., Zhang, S. H., He, S., and Dean, D. C. (1995).
Mechanism of active transcriptional repression by the retinoblastoma protein. Nature 375, 812-5.
Weintraub, S. J., Prater, C. A., and Dean, D. C. (1992). Retinoblastoma protein switches the
E2F site from positive to negative element. Nature 358, 259-61.
White, E. (1995). Regulation of p53-dependent apoptosis by ElA and EiB. Curr Top Microbiol
Immunol 199, 34-58.
Whyte, P., Buchkovich, K. J., Horowitz, J. M., Friend, S. H., Raybuck, M., Weinberg, R. A.,
and Harlow, E. (1988a). Association between an oncogene and an anti-oncogene: the adenovirus
ElA proteins bind to the retinoblastoma gene product. Nature 334, 124-9.
Whyte, P., Ruley, H. E., and Harlow, E. (1988b). Two regions of the adenovirus early region
IA proteins are required for transformation. J Virol 62, 257-65.
Whyte, P., Williamson, N. M., and Harlow, E. (1989). Cellular targets for transformation by the
adenovirus ElA proteins. Cell 56, 67-75.
Williams, B. 0., Schmitt, E. M., Remington, L., Bronson, R. T., Albert, D. M., Weinberg, R.
A., and Jacks, T. (1994). Extensive contribution of Rb-deficient cells to adult chimeric mice with
limited histopathological consequences. Embo J 13, 4251-9.
66
Williams, J., Williams, M., Liu, C., and Telling, G. (1995). Assessing the role of ElA in the
differential oncogenicity of group A and group C human adenoviruses. Curr Top Microbiol
Immunol 199, 149-75.
Woo, M. S., Sanchez, I., and Dynlacht, B. D. (1997). p130 and p107 use a conserved domain
to inhibit cellular cyclin- dependent kinase activity. Mol Cell Biol 17, 3566-79.
Wu, C. L., Classon, M., Dyson, N., and Harlow, E. (1996). Expression of dominant-negative
mutant DP-1 blocks cell cycle progression in G1. Mol Cell Biol 16, 3698-706.
Wu, C. L., Zukerberg, L. R., Ngwu, C., Harlow, E., and Lees, J. A. (1995). In vivo
association of E2F and DP family proteins. Mol Cell Biol 15, 2536-46.
Wu, X., and Levine, A. J. (1994). p53 and E2F-1 cooperate to mediate apoptosis. Proc Natl
Acad Sci U S A 91, 3602-6.
21
Xiong, Y., Hannon, G. J., Zhang, H., Casso, D., Kobayashi, R., and Beach, D. (1993). p is
a universal inhibitor of cyclin kinases. Nature 366, 701-4.
Xu, G., Livingston, D. M., and Krek, W. (1995). Multiple members of the E2F transcription
factor family are the products of oncogenes. Proc Natl Acad Sci U S A 92, 1357-61.
Xu, M., Sheppard, K. A., Peng, C. Y., Yee, A. S., and Piwnica-Worms, H. (1994). 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, 8420-31.
Yamasaki, L., Bronson, R., Williams, B. 0., Dyson, N. J., Harlow, E., and Jacks, T. (1998).
Loss of E2F-1 reduces tumorigenesis and extends the lifespan of Rbl(+/- )mice. Nat Genet 18,
360-4.
Yamasaki, L., Jacks, T., Bronson, R., Goillot, E., Harlow, E., and Dyson, N. J. (1996).
Tumor induction and tissue atrophy in mice lacking E2F-1. Cell 85, 537-48.
Yan, Z., DeGregori, J., Shohet, R., Leone, G., Stillman, B., Nevins, J. R., and Williams, R. S.
(1998). Cdc6 is regulated by E2F and is essential for DNA replication in mammalian cells. Proc
Natl Acad Sci U S A 95, 3603-8.
Yee, A. S., Reichel, R., Kovesdi, I., and Nevins, J. R. (1987). Promoter interaction of the ElAinducible factor E2F and its potential role in the formation of a multi-component complex. Embo J
6, 2061-8.
Yee, S. P., and Branton, P. E. (1985). Detection of cellular proteins associated with human
adenovirus type 5 early region IA polypeptides. Virology 147, 142-53.
Yokota, J., Akiyama, T., Fung, Y. K., Benedict, W. F., Namba, Y., Hanaoka, M., Wada, M.,
Terasaki, T., Shimosato, Y., Sugimura, T., and et al. (1988). Altered expression of the
retinoblastoma (RB) gene in small-cell carcinoma of the lung. Oncogene 3, 471-5.
Yoshitaka, T., Matsubara, N., Ikeda, M., Tanino, M., Hanafusa, H., Tanaka, N., and Shimizu,
K. (1996). Mutations of E2F-4 trinucleotide repeats in colorectal cancer with microsatellite
instability. Biochem Biophys Res Commun 227, 553-7.
Zalvide, J., and DeCaprio, J. A. (1995). Role of pRb-related proteins in simian virus 40 large-Tantigen- mediated transformation. Mol Cell Biol 15, 5800-10.
67
Zalvide, J., Stubdal, H., and DeCaprio, J. A. (1998). The J domain of simian virus 40 large T
antigen is required to functionally inactivate RB family proteins. Mol Cell Biol 18, 1408-15.
Zamanian, M., and La Thangue, N. B. (1993). Transcriptional repression by the Rb-related
protein p107. Mol Biol Cell 4, 389-96.
Zantema, A., and van der Eb, A. J. (1995). Modulation of gene expression by adenovirus
transformation. Curr Top Microbiol Immunol 199, 1-23.
Zerler, B., Roberts, R. J., Mathews, M. B., and Moran, E. (1987). Different functional domains
of the adenovirus ElA gene are involved in regulation of host cell cycle products. Mol Cell Biol
7, 821-9.
Zhang, Y., and Chellappan, S. P. (1995). Cloning and characterization of human DP2, a novel
dimerization partner of E2F. Oncogene 10, 2085-93.
Zhang, Y., Xiong, Y., and Yarbrough, W. G. (1998). ARF promotes MDM2 degradation and
stabilizes p53: ARF-INK4a locus deletion impairs both the Rb and p53 tumor suppression
pathways. Cell 92, 725-34.
Zwicker, J., Liu, N., Engeland, K., Lucibello, F. C., and Muller, R. (1996). Cell cycle
regulation of E2F site occupation in vivo. Science 271, 1595-7.
68
CHAPTER I
The individual E2F family members bind a common DNA recognition
sequence
B Fairchild, S Estes and JA Lees
(Author's contribution: all data, except some of the sequencing
associated with the binding site selections.)
A.
Abstract
The E2F transcription factors have been implicated in the regulation of cellular proliferation
and of apoptosis. They are inhibited by the product of the retinoblastoma-susceptibility gene,
pRB, and by its homologs p107 and p130. E2F transcription factors are heterodimers composed
of an E2F subunit and a DP subunit. In humans, six E2F proteins and two DP proteins have been
identified; the pRB family members bind to specific subsets of the E2F proteins. Both
overexpression studies and genetic studies indicate that the E2F proteins have a variety of distinct
functions in vivo. Mutation of pRB family members derepresses distinct subsets of E2F target
genes, as does overexpression of individual E2F proteins. These observations strongly suggest
that the E2F proteins regulate different subsets of E2F-responsive genes, and raises the possibility
that E2F proteins could differ in their DNA binding specificities. We have determined that the
preferred DNA binding sequences for complexes containing DP-I with E2F-1, -2, -3, or -4 are
indistinguishable. Each sequence selected for binding to one of the complexes also binds the
others with comparable affinities. More importantly, each sequence in the context of a minimal
promoter can serve as a site for transcriptional activation by any of the complexes. Sequence
differences within E2F sites are therefore unlikely to mediate target specificity by particular E2F
proteins in vivo.
69
B.
Introduction
E2F is a transcriptional regulator that plays a pivotal role in the regulation of cellular
proliferation (reviewed in Beijersbergen and Bernards, 1996; Dyson, 1998). Many E2Fresponsive genes have been identified and their products are components of either the DNA
replication (for example DNA polymerase u, HsOrc 1, and cdc6) or the cell cycle control (for
example cyclin E, cyclin A and cdc2) machinery. In each case, E2F acts to restrict the transcription
of these genes to the stage of the cell cycle in which their products are known to function.
E2F is regulated by the retinoblastoma protein (Bandara and La Thangue, 1991; Chellappan
et al., 1991), a tumor suppressor that is functionally inactivated in almost all human tumors
(reviewed in Dyson, 1998). Consistent with its anti-proliferative role, the binding of the
retinoblastoma protein (pRB) blocks the transcriptional activity of E2F (Hiebert et al., 1992). In
addition, overexpression studies indicate that the resultant pRB-E2F complex can act as a
transcriptional repressor in which E2F provides the sequence-specific DNA binding and pRB
inhibits transcription through the recruitment of a histone deacetylase (reviewed in DePinho,
1998). The interaction between pRB and E2F is regulated by the cell cycle-dependent
phosphorylation of the retinoblastoma protein by one or more of the cell cycle-dependent kinases
(reviewed in Dyson, 1998). Phosphorylation of the retinoblastoma protein releases it from E2F,
switching E2F-responsive genes from a fully repressed to a fully induced state. Consistent with
this model, the timing of E2F-responsive gene transcription correlates closely with the timing of
pRB phosphorylation at the GI/S transition.
E2F is also regulated by two additional proteins, p107 and p130 (Cao et al., 1992; Devoto
et al., 1992; Shirodkar et al., 1992; Cobrinik et al., 1993). These two proteins share significant
sequence similarity with the retinoblastoma protein (Ewen et al., 1991; Hannon et al., 1993; Li et
al., 1993; Mayol et al., 1993) and they repress E2F in a similar manner to pRB in overexpression
assays (Zhu et al., 1993; Starostik et al., 1996). Despite these similarites, the loss of these "pocket
proteins" has dramatically different biological effects in mice. Mice heterozygous for a loss-offunction mutation at the retinoblastoma locus develop tumors at 100% penetrance, whereas mice
70
lacking only p107 or p130 do not demonstrate an increased tumor predisposition (Jacks et al.,
1992; Cobrinik et al., 1996; Lee et al., 1996). Homozygous mutations at these three loci also have
different developmental effects in mice (Jacks et al., 1992; Cobrinik et al., 1996; Lee et al., 1996;
LeCouter et al., 1998a; LeCouter et al., 1998b). Thus, despite the similarities among pRB, p107,
and p1 30, they have distinct biological functions in vivo.
The pocket proteins also appear to regulate E2F in distinct ways. First, pRB, p107 and
p130 bind to E2F at distinct stages of the cell cycle (Mudryj et al., 1991; Cao et al., 1992; Lees et
al., 1992; Shirodkar et al., 1992; Chittenden et al., 1993; Cobrinik et al., 1993; Moberg et al.,
1996). Second, the loss of pRB, p107 and/or p130 in otherwise isogenic cell lines results in the
specific deregulation of different subsets of E2F-responsive genes (reviewed in Dyson, 1998).
These findings have led a number of labs to suggest that the pRB-E2F, p107-E2F and p130-E2F
complexes will have distinct target specificities in vivo.
To date, at least eight human genes have been identified that encode components of E2F
DNA-binding activity (reviewed in Dyson, 1998). These can be divided into two distinct groups,
termed E2F (1 through 6) and DP (1 and 2), whose products heterodimerize to generate functional
E2F complexes (Bandara et al., 1993; Helin et al., 1993; Krek et al., 1993; Wu et al., 1995). In
vivo studies indicate that the endogenous E2F activity arises from the concerted action of multiple
E2F/DP complexes (Wu et al., 1995; Moberg et al., 1996). Although these complexes do share,
to some extent, the ability to promote cell cycle progression when ectopically expressed, both
overexpression studies and genetic studies indicate that these complexes also have distinct
functions. Ectopic expression of E2F-1, for example, can induce apoptosis in a rat fibroblast cell
line (REF52), whereas the overexpression of E2F-2, -3, -4, or -5 does not (DeGregori et al.,
1997). Similarly, loss of E2F-1 in the mouse yields a defect in thymocyte apoptosis (Field et al.,
1996), whereas loss of E2F-5 affects choroid plexus function (Lindeman et al., 1998). A specific
role for E2F-3 has been proposed in the GI/S transition in cycling cells (Leone et al., 1998). E2F
complexes therefore have individual functions in addition to any shared ones.
71
pRB, p107 and p130 each target a subset of the endogenous E2F/DP complexes (Dyson et
al., 1993; Lees et al., 1993; Beijersbergen et al., 1994; Ginsberg et al., 1994; Hijmans et al.,
1995; Sardet et al., 1995; Vairo et al., 1995). Although the DP subunit is essential for pocket
protein binding, the specificity of this interaction is determined by the E2F moiety. E2F-1, -2, or
-3 complexes bind specifically to pRB. In contrast, p107 and p130 only associate with E2F-4 and
-5 complexes, and E2F-6 does not associate with any of the pocket proteins (Cartwright et al.,
1998; Gaubatz et al., 1998; Trimarchi et al., 1998).
Sequence comparisons also clearly
differentiate the pRB-specific E2F proteins (E2F-1, -2, and -3) from the others. Strikingly, in the
DNA-binding domain there are several residues that are absolutely conserved among the pRBspecific E2F proteins, whereas a different amino acid at that position is conserved between the
p107/p 130-binding E2F proteins. These observations have led several groups to propose that the
E2F proteins will regulate different E2F-responsive genes, and that this would account for the
different biological properties of the pocket proteins and of the E2F genes. Indeed, ectopic
expression of individual E2F proteins has been demonstrated to differentially activate E2Fresponsive genes in multiple cell lines (DeGregori et al., 1997; Dirks et al., 1998). For example,
expression of E2F- 1 efficiently activates the expression of cyclin A but not of cdk2 in REF52 cells,
whereas the opposite is true for E2F-3, and E2F-2 is the most efficient at activating DHFR
expression (DeGregori et al., 1997). These data argue very strongly that individual E2F genes
regulate specific E2F-responsive promoters in vivo.
E2F DNA-binding activity was originally defined as the ability to recognize and bind to one
of the E2F sites (TTTCGCGC) derived from the adenoviral E2 promoter. However, E2Fresponsive elements have been identified within many cellular promoters and there is considerable
variation between these sequences and the canonical E2 site. Moreover, a single nucleotide
difference between two known E2F-response elements has been proposed to alter the spectrum of
associated E2F complexes (Wells et al., 1996; Wells et al., 1997). This suggests that differences
in the target specificity of the two E2F subgroups could arise, at least in part, from differences in
their DNA binding specificity. In this study, we have used an in vitro DNA binding site selection
72
assay to identify the optimal DNA binding sequences of four different E2F/DP- 1 heterodimers.
The high affinity DNA binding sites selected for these complexes were indistinguishable from one
another. All four E2F/DP-1 complexes bound to these sites with similar, high affinities, and
transactivated a reporter bearing these sites with similar efficiencies. We therefore conclude that
variations in the primary sequences of high-affinity E2F sites are unlikely to specify which E2F
proteins can regulate a given promoter.
73
C.
C1.
Results
Production of recombinant proteins
To better understand the DNA binding specificity of the individual E2F/ DP complexes, we
sought to identify their optimal recognition sequences. To establish a convenient source of each
complex, we used baculoviral vectors to express the proteins in insect cells. A baculovirus-based
system provided several advantages over other expression systems (Dynlacht et al., 1994). First,
it allows the generation of insect cell extracts containing high levels of the desired protein. Second,
it confers the ability to form a functional E2F complex by tolerating the infection of both E2F- and
DP-expressing viruses. Finally, the E2F complexes expressed in this system are likely to contain
any post-translational modifications necessary for their normal function. To take advantage of this
system we constructed a series of recombinant baculoviruses expressing hemagluttinin (HA)-
tagged E2F-2, -3, or -4 (BV-HAE2F-X) or DP-I (BV-HADP-1). A recombinant baculovirus
expressing untagged E2F-1 (BV-E2F-1) was also used (Dynlacht et al., 1994). Single infections
of insect cells with these viruses resulted in high level expression of the expected E2F or DP- 1 as
judged by western blotting of whole cell extracts from infected cells (Figure 1A). BV-HAE2F-4
infection resulted in the expression of multiple species of E2F-4, consistent with the multiple
phosphorylated forms of the protein found in mammalian cells (Figure IA and data not shown).
We next asked whether the resulting E2F proteins were functional by testing their ability to
dimerize with DP-I and to bind to DNA. Insect cells were dually infected with BV-HADP-I and
individual E2F viruses and the resulting cell extracts were tested for DNA binding activity in gel
shift assays using an oligonucleotide containing the canonical E2 site as a probe. At the levels of
input protein used, we were unable to detect any E2F DNA-binding activity in extracts of cells that
had been mock- or singly-infected (Figure 1 B, lanes 1, 6, 11, 16 and 21, and data not shown). As
expected, extracts of dually infected cells showed considerable levels of E2F DNA binding
activity. The presence of the expected E2F subunit and DP- 1 in each infection was confirmed by
74
Figure 1A.
0
0
0
a E2F-3
a E2F-2
a E2F-1
0
0
a E2F-4
a HA
Figure lB.
no antibody
1
k
2
.~t.
3
4
-.- j--a
5
-i.*i: i i
7
6
:,j
a E2-2a
E2F-1
8
9
10 11
12 13
E2F-3
14 15 16 17 18
a E2F-4
19 20 21 22 23 24 25
Figure 1. (preceding page) Overexpression of E2F subunits by infection of Hi5 insect cells with
BV-E2F-1, BV-HAE2F-2-4 and/ or BV-HADP-1. (A) Three days post-infection cell extracts were
prepared from singly- or mock-infected cells and analyzed by Western blotting with E2F- or HAspecific monoclonal antibodies. The amounts of protein loaded were as follows: E2F-1, 20 ng;
E2F-2, 200 ng; E2F-3, 5[tg; E2F-4, 100 ng; DP-1, 500 ng. In each case, an equal amount of
mock-infected extract was run in parallel. (B) Generation of high levels of E2F DNA-binding
activity by dual infection of Hi5 cells with BV-HADP-l and a BV-E2F. Extracts were prepared
three days post-infection and used in gel shift assays with the canonical E2 site. 50 ng of total
protein from each infection were used per binding reaction, except for the BV-HAE2F-3/ BVHADP- 1 infection in which 400 ng were utilized.
76
the ability of specific monoclonal antibodies to supershift the relevant complexes (Figure lB lanes
7, 13, 19, and 25, and data not shown).
C2.
E2F site selection
To identify DNA binding sites, the E2F complexes were incubated with a pool of
32 p-
labeled double-stranded oligonucleotides containing an internal region of 25 random positions.
Free and bound oligonucleotides were separated by electrophoresis on 4% native gels (Figure 2).
To insure that only relatively high affinity E2F binding sites would be captured, a titration of
protein concentration was performed and only the lowest input of protein yielding a detectable band
in the gel shift was used for the selection. The oligonucleotides bound each E2F complex were
isolated from the gel, amplified by PCR and subjected to five additional rounds of selection with a
protein titration performed at each round. After the sixth round, the isolated oligonucleotides were
cloned and 28-36 independent clones were sequenced for each E2F/ DP-1 complex.
Figure 3 shows an alignment of the oligonucleotides sequenced after six rounds of
selection with the E2F-1/DP-1 extract. We found that several selected oligonucleotides contained
perfect matches with known E2F-responsive elements, validating the efficiency of the selection
protocol. These include E2F-response elements from the E2F-1 (1.6), CDC2 (1.22), and DHFR
(1.24, 1.28, and 1.31) promoters. The remaining selected oligonucleotides each contained a
sequence that was highly homologous to known E2F binding sites. These selected E2F- 1 binding
sites had two common features. First, these sequences contained an internal G/C-rich core that
was usually five or six nucleotides long. Most notable was the presence of a highly conserved
CGC trinucleotide at positions 3-5 of the core. Second, the G/C-rich core was frequently flanked
on either side by adenine or thymine residues. A comparison of these sequences yielded the
consensus sequence WTTSSCGCSWWWW where W represents A or T and S represents G or C
(Figure 4).
The oligonucleotides derived from the other E2F/DP- 1 selections showed considerable
similarity to those derived from the E2F-I/DP-I selection (Figures 3 and 4). Each of the
77
Figure 2.
E2F
DP-1
GGCTGAGTCTGAACGGATCCNNNNNNNNNNNNNNNNNNNNNNNNNCCTCGAGACTGAGCGTCG
CCGACTCAGACTTGCCTAGGNNNNNNNNNNNNNNNNNNNNNNNNNGGAGCTCTGACTCGCAGC
Amplify by PCR
Repeat EMSA
Amplify by PCR
Repeat EMSA
1
W
e~~
-- MM
Figure 2. (preceding page) Summary of the site selection scheme. Extracts of infected cells were
incubated with a pool of randomized oligonucleotides and separated by gel shift assay. After
autoradiography, DNA from the shifted complexes was excised from the gel, eluted, and amplified
by PCR. A total of six rounds of selection were performed.
Figure 3. (following page) Oligonucleotides derived from six rounds of E2F-1/ DP-1 selection.
Sequences are aligned to the highly conserved CGC (boldface). Randomized positions within the
oligonucleotides are capitalized, whereas invariant positions are depicted in lower case.
Figure 4. (second following page) Comparison of the consensi derived from the selections with
DP-1 and E2F-1, -2, -3, or -4. For each position, the frequency at which a given nucleotide
appears is indicated.
79
0)
r-4
40
(2'
01
W2
1 U0 01U
0
C) -(2(
E,
As0 4 U.
(20
002200
U m
0)
44400
1
(H
(0
4OPH4U22
4
000
HH
HO
V-
2
0
(N
)
4 (22(2
(qcicl
O00(0000
i
N
H
8<8888
H
HHHH
~
00
((
((NN(
(NN((N4
4J-(0
(0
lioo
00
HH
(2
(2
00
0
P
(~
00p
u0
OH
0(20
0l
(N
0 U
E
u
(
(N
(Ni
NNNmmN(I
Z3
0
O
(N
H
ooos
(NNNN
(2 u u r
m
0(2
uu
0 (200
U0(2(2
HO
88888888
H(2 ()C uuu 0 H(2 0
OH2( u2 4 L u(2HH(
u
N
U-
HH2
HH 00(2H(2(
oooooou
~~
(N
000
0) r.)
0000uu(m2
U
EU (MH
UU681"
U00H H
HH
88
(20M (
HUO
H U 000(u2
H
H HH
N
(20
p(00
~ <u
HH
HH
(Ni
m
01M(
)01 0
0 0
(2m
UUm (Ol 0U)V
M 0 0((
M dU
(2001
0 ) uuu r
0
(2u
0
0022((
HOOOO(
0
0
0(2
H(2
HH
00
000
E-H(((002HE'(220(2
HH(O2HHH
0000000
U4
HH
0
(2UOHuU H 4000(2(2(2(2 u 0 (2 UOOH H
OOH(2(
H(2HH000(200000 OO8 O"(2H8(200(
HHHHH2(2 (
(20000
Uooooooooo
0
E-.~~~ (2
0l
Wr w (2
N((NNN
(n
E- 0000000
0
U0( 0 00(
(2(i((0
U4M
EU u
(00 000
4(0(0 M
(2(200
(00(0000
mmmEm 000(2 RrdO
U (2
08 )0(
00 0
0
0
0) 08%
HE0 68 0 00 0 (2 (2 0)
V
0(20l4JM
0(24
0~
N m
u
0(
M4OVUE
u
(2
(2
U(2
4J (Z
0000
00
0i
00() m 000
0 )
(2
(2
(20
(0H 0(20l
HH g(d2OHE4(
OH
0
HO)
U0u2n(2UH
4
u0((
(N (N CA
H
((0
g
0H00OH
'El
OH 00
u 00 H 0 HH 000
OO
0u0 H HH H OH H4 00u (
48a08888888
HHH
HHO
0(
01
(2
0
0
44
NN
) -v( (n
(N
HHHr-wHHH H HJ H Hi H H-
00
H
W rH
H H
H H
H
wHHHHHHH
00
0
u
N(
'0
(2
W2
(2
(NN (N(NV(N(N(N-(M(M(N(N(NM(*(n(N(N
H H H H n mHH H H, H H
0(
MUOOHH(U
H OH
HO ()O
04U000440
M O O HH (2(20(2(2
0
U Q EOH
H(H22
0-2 (2 O(a0H ( 20(
4 H( (2HO
H - OO
HM 2
0000(
(2
0(022
HOH00
H (20
(200(20
HOO 0
O
000(0
0((0
OH OH (2
0 n-2(2(02(2(20 02 H
0 (2(02(2
O
(2 HO 01zOOH
0 a O(l
00(2(202
0)-2
000(2000(200(H
OOOOH
HOO
00000(
UUUO8888UUOUUUOUUUUUOUU8OOUUUU8UUUUOU
O000000000H
0(
000
00
000000000000000
2~O(~0000(0
000
00~~~~ 000
40
m U
H~~
H H Hn
0
HO
0
(2
U
0
0
0
(20
H
0
H(2
E-
(20
0
0
U ()
~0 00
U
0
40
u
(2
0
OHU
UH
(2..((2
0) 000
ooo
H4
0(2
H
U(2
00
(2
l
0
4j01
0
m0
0
(2) 0
0
I(2OH(2mO(U2
m
(2l
M
lr
(a~2 m
(2
00
6
O(2J
g
0(200
g p
E0
(2(2H OH (2
000 0 4E 0
0(20
I0
OHH1
uuuu
E
M M-
0~ 00
0
0
0(0
u 0 0(2
(n * n w r2 w.m
m
HO (2 (2OOH 1u ) 0 H
U2 HM2
00(2 (0H
HHU4 1 (a(2HOO(2 (0
( (HOHOO(20
(200
(2H
(20
0 (2
0uu~uuu~uuu
IME'UOO M~
U
011H
01
(2
M
0(2
HH 4H (
UO0
4-2
p
oo
4O O
2
""HH2
Egggggggggggg
0
(2
01
HH
00
0
0
0
ME
0(22(0H
0 OHU)(O8888888
uU o3Ho~
HHOOOOO
(2
U 0
000
00(
OH
gUr
00
0Htl.
0 <0
g02 (0(
0 U U
H4J
0 u
-j
0
H
0
0U20
0(2(200F4 (2
(
00
0
0~ (
0
b)24J
00
0 (M
mJ
0
0
0
(20
0
010
-0
(2
(2
0
0
(2
(2208
(
HrVPHPOH
Ho
(000(2EUH
0 00
l
U0 00 HO OO 2(
UO O
oo
000
H
(2
0
01
0
UU
0
(2
(2
0 8 0
02 0 (2H
(20
MH 00
41(
0
00
(2(0
OHuU
H2
(2IM
Mpu
00)1H
((2
) H
(0
(2
O
O
0
U2
000 00
20U00
00(2
1 .,E4
0
U
(2(2
H(
0
gggggg
400(00
0
U
S(2
, 000
(2(0
-0(20
U2
002
O(240
(2(2
0
UoE~uU
888888888888888888800
8o888o8o88o8uo8888488ouu8888
4
a
0y
0 00
(2 F0
)0b
(24
(2mO00iH"
Figure 4.
E2F-1/DP-1 Consensus:
lUU-
-
80-
G
A
G_
.
G
T
G G
G
C
A
C
C
G
60
A
C
T
-
C
_
C
A
T
ACG
TTCCCGC
T
4020
-
A
A
GAAATTT
T
A
E2F-2/DP- 1 Consensus:
100-
G
80-
T
-_c
c
c
AGG
-6
G
-
60-
C
GcGG
C
G
T
A
T A
-
AAA
A
T
--
-
TTCCCGC
C
T
40-
- A A A T
CAATTC
20
E2F-3/DP-1 Consensus:
80-
GG
C
G
A
C
G
T6-T-TT T
T TC
C C
60
G
T
G
GG
G
-C
AG
C
AA
A
G
A
CGC-T
CC
-T
40-
A -AA
C
A
CATTTTA
A
20-
T
-I
0-
-
I-
-
-
E2F-4/DP- 1 Consensus:
0
T
80
G
C
-
T
G
G
-
G
AC
A 60-
T
C
G
T
c
T TC
TTCCCGC-_
A
-C
40-
AA
0 -
A
A
-
AT
A
T
sequences selected for binding to E2F-2, -3, or -4 containing complexes contained the CGC
trinucleotide. This was typically present in a five or six nucleotide G/C core flanked by A/T-rich
sequences. Occasionally, the G/C run extended to seven or more residues in length. However,
this extended G/C-rich sequence was detected at similar frequency in each of the individual E2F
selections. For all four E2F proteins, there was a marked preference for thymine in the two
positions directly 5' of the G/C core. The nucleotides following the G/C run are enriched in
adenine and thymine, and at a significant frequency this A/T-rich stretch is composed principally or
exclusively of either adenine or thymine. We therefore conclude that heterodimers of E2F- 1, -2,
-3, or -4 with DP- 1 selectively bind to the same high affinity sequence, WTTSSCGCSWWWW,
in vitro.
The trinucleotide CGC was invariant in the selections involving E2F-2, -3, or -4. In
contrast, 3 of the 36 E2F-1-derived sequences contained a CGG at these positions. This raised the
possibility that E2F- 1 complexes may bind with high affinity to sequences containing a CGG. If
so, while promoter sequences with a CGC might be regulated by any of the E2F proteins,
sequences with a CGG in its place might be regulated exclusively by E2F- 1. To test this
possibility, we compared the affinities of E2F-I and other E2F proteins for CGG-bearing
sequences (Figure 5). When used in binding competition experiments, oligonucleotides with a
CGG sequence bound less well to each of the four E2F complexes than the canonical, CGCcontaining E2F sequence. Thus, the CGG trinucleotide promotes only low affinity binding.
Furthermore, the oligonucleotides with the CGG sequence were no more efficient at competing for
E2F-1 binding than at competing for binding with the other E2F complexes, arguing that E2F-1
complexes are no more tolerant of this sequence variation than other E2F complexes. We therefore
conclude that replacement of the CGC trinucleotide with CGG gives rise to a low affinity E2F
binding site that does not show any preference for binding to E2F- 1 relative to the other E2F
family members.
82
Figure 5.
2000
.0
1000-
LU
U no competitor
U + E2-CGGG
S+
Cl)
E2F-1/DP-1
E2F-2/DP-1
Transfected protein
E2F-4/DP-1
E2-CGGC
Figure 5. (preceding page) E2F complexes share similar, low tolerances for CGG sequences.
C33-A cells were transfected with constructs expressing DP-1 and the indicated E2F protein. 1 gg
of transfected cell extract was subjected to gel shift analysis. in the presence or absence of 100 ng
of the indicated competitor DNA sequences. The amount of binding to the E2-derived probe in the
absence or presence of 100 ng of the indicated competitor DNA sequences was determined by
quantitation on a STORM 860 (Molecular Dynamics).
84
C3.
Relative affinities of E2F complexes for selected sites
To confirm that any E2F heterodimer could indeed bind to these selected sites, eight
sequences, varying in the length and composition of the internal G/C core as well as the
arrangement of flanking sequences, were used in gel shift assays. In these experiments the E2F
complexes all bound to any given probe with qualitatively similar efficiencies (data not shown).
Furthermore, complexes containing DP-2 (instead of DP-1) also bound all eight sequences with
relatively high affinity (data not shown), demonstrating that the DNA sequences recognized by DP1 can also be recognized by DP-2.
In order to define more precisely the affinity of the different E2F complexes for these
selected sites, we selected one each of the E2F- 1-, E2F-2- and E2F-4-selected sequences to be
used as competitors for the binding of the E2F/ DP-1 dimers to the canonical E2F site (Figure 6).
Although there was as much as a ten-fold difference in the affinity of the E2F heterodimers for
different selected sites (Figure 6, compare 3P with 4E), the relative affinity of the E2F complexes
for any given site did not vary more than two- to four-fold. This confirms that, in vitro, subtle
sequence differences within or outside an E2F site do not selectively dictate which family members
can bind to that site.
C4.
Sequence variations do not confer E2F complex specificity in transactivation
Based on the in vitro binding studies, we hypothesized that variations in the sequences of
E2F sites are unlikely to determine which E2F proteins are capable of promoting transcription at a
given promoter. To test this notion, we constructed a series of reporter constructs that contained a
tandem repeat of the three E2F DNA-binding sites, 1.30, 3P and 4E, upstream of a common
minimal promoter. Since the transcriptional activity of the individual E2F proteins is known to
vary considerably (Verona et al., 1997), we compared the ability of each E2F complex to activate
the test reporters versus one containing the canonical E2F site from the E2 promoter. We saw a
significant difference in the activity of these test reporters (Figure 7). The reporter containing the
low affinity E2F binding site, 3P, gave the lowest levels of activation in the reporter assays. In
85
Figure 6.
4
3-
E2F-1/DP-1
0U
4)
U
2
E2F-2/DP-1
U E2F-3/DP-1
[9 E2F-4/DP-1
7T)
0
1.30
3P
Competitor
4E
Figure 6. (preceding page) Competition for recombinant E2F DNA-binding with selected sites.
50 ng (E2F-1 and 2), 200 ng (E2F-3) or 100 ng (E2F-4) of extract from dually infected cells were
incubated with a 32 P-labelled double-stranded wild-type E2 oligonucleotide in the presence or
absence of increasing amounts of the indicated unlabelled competitor (2, 10, 50, or 300 ng).
Figure 7. (following page) E2F-mediated transactivation through selected sites. C33-A cells
were transfected with an artificial reporter construct containing two copies of a wild-type or mutant
E2F site, or two copies of selected sequences 1.30, 3P, or 4E, upstream of a CAT reporter; CMVbased constructs expressing DP- 1 and the indicated E2F were also transfected. For E2F-1, 100 ng
of CMV-E2F-1 and 100 ng of CMV-DP1 were transfected; for E2F-2 and -3, 300 ng of the CMVE2F construct and 300 ng of CMV-DP-1; for E2F-4, 1000 ng of CMV-E2F-4 and 1000 ng CMVDP- 1; the total amount of CMV promoter present in each transfection was maintained at 2 g by
cotransfection of an appropriate amount of CMV plasmid. Depicted in the graph is the ratio of the
activity of each E2F/DP complex on the reporter in question to its activity on the wild-type
reporter. Transfection efficiency was internally controlled by the cotransfection of 2gg of RSVluciferase in each experiment. The error bars depict the standard error of the mean.
87
Figure 7.
2
1.4
.
1.0
.
N
0.8
C.6E2F-3/
DP-1
0 E2F-4/ DP-1
> 0.40.2
0.0
_____
E2
E2F-1/ DP-1
E2F-2/ DP-1
E2m
1.30
reporter
3P
4E
contrast, the presence of the higher affinity E2F binding sites, 1.30 or 4P, resulted in significantly
higher reporter activity. Most importantly, when compared to their ability to activate the canonical
E2 reporter, we did not detect any significant difference in the ability of the individual E2F
complexes to activate a given test reporter (Figure 7). This was observed over a range of input
plasmid concentrations (data not shown), proving that the reporter system is not at saturation. We
therefore conclude that the sequence of the E2F site can significantly alter the efficiency of E2F
binding and thereby the level of transactivation but it is unable to specify the identity of the E2F
complexes which can bind and regulate the target gene.
89
D.
Discussion
The E2F genes vary both in their biological functions and in their ability to activate the
expression of endogenous target genes when overexpressed. In this paper, we have investigated
the possibility that these differences in E2F function could result from differences in their DNA
binding specificities. Through an in vitro site selection assay we have identified high-affinity
binding sites for heterodimers containing human DP- 1 and E2F- 1, -2, -3, or -4. The high-affinity
sites for each of these complexes showed considerable homology not only to each other but also to
known E2F binding sites from E2F-responsive promoters. By comparing the oligonucleotides
selected in these assays, we propose the following consensus E2F DNA-binding site:
WTTSSCGCSWWWW, where W represents A or T and S represents G or C. Beyond these
thirteen positions, we detected no obvious nucleotide preference at any position, suggesting that
flanking sequences do not contribute to binding site specificity in these in vitro assays. Based on
DNA binding site competition studies, the internal CGC trinucleotide appears to be critical to highaffinity DNA binding. This motif is not, however, sufficient for binding: it is always found in the
context of a longer series of guanine and cytosine residues. High affinity E2F binding seemed to
tolerate some flexibility in the length of this G/C run. Although G/C stretches of five, six, or
seven residues were most abundant in the selected sites, examples with eight or nine residues were
occasionally observed. Regardless of the length of the G/C core, each of the E2F complexes tested
showed a marked preference for adenine or thymine at the positions directly flanking the G/C-rich
sequences. This was most obvious in the first two positions prior to the core which were strongly
biased toward thymine.
The similarity of the sequences derived from the E2F-1, -2, -3, and -4 selections suggested
that there may be little or no difference among high-affinity DNA binding sites recognized by the
various E2F complexes. Indeed, binding competition studies with sequences derived from the
E2F-1, -3, and -4 selections suggest that complexes with DP-1 and E2F-1, -2, -3, or -4 can all
bind each of these sequences with comparable affinities. This is consistent with studies comparing
90
the affinities of various E2F complexes for a sequence from the murine B-myb promoter
(TCTCCCGCCAAGT) to their affinities for the canonical E2 sequence. Complexes containing
E2F-1, -2, -3, -4, or -5 have been reported to bind those two sequences with similar affinities, in
the presence of murine DP-1, -2, or -3 and in the absence or presence of pRB or p107 (Zwicker et
al., 1996). In both that study and in the present one, the largest differences in the affinities of
individual E2F complexes for a given DNA sequence were two- to four-fold. Thus, sequences
that bind to one E2F complex with high affinity can also bind to other E2F complexes with similar
affinities. Furthermore, the DNA sequences from the binding site selection did not differentiate
between E2F- 1, -2, -3, or -4 in a transcriptional activation assay. We therefore conclude that the
differences in the effects of specific E2F proteins at individual promoters are unlikely to be
mediated by differences in the sequences of the E2F DNA binding sites.
Tao et al. have previously reported that the consensus binding sites for E2F-4/ DP- 1 and
for pRB/ E2F-1I/ DP-1 complexes differ from those of E2F-1I/ DP-1, E2F-1I/ DP-2, and E2F-4/
DP-2 (Tao et al., 1997). This conclusion is clearly at odds with our own. We suggest that this
discrepancy is due to the manner in which they derived their consensus binding sites. First, their
analysis included multiple copies of a relatively small number of identical oligonucleotides,
suggesting that they resulted from repeated PCR amplification of a small number of selection
events. Although each of our selections yielded twenty-eight to thirty-six inserts per experiment,
every sequence was different suggesting it arose from an independent binding event. Second, Tao
et al. included partially palindromic sites in both possible orientations. In each case, we believe
that the inclusion of these repeat sequences is likely to have introduced significant bias into the
consensus. In both the E2F-4/ DP-1 and the pRB/ E2F-1/ DP-1 selections there were high
percentages of identical sequences, most of which were palindromic. Indeed, if the identical
sequences from their E2F-4/ DP- 1 selection are excluded, and the palindromic sequences are only
counted once, the resulting consensus strongly resembles their E2F- 1/ DP- 1 consensus, and our
consensus for both complexes. This reaffirms our observations that different E2F proteins share
the same preferred DNA recognition sequences.
91
There are also other important similarities between our data and theirs. For example, we
both find a tolerance for either cytosine or guanine at a variety of positions in the C/G core (Tao et
al., 1997). Also, just as we find that affinity differences among E2F complexes for our selected
sequences do not vary by more than two- to four-fold, they found that the ratio of the half-life of
E2F- 1/ DP- 1 to the half-life of E2F-4/ DP- 1 complexes on individual sequences does not vary by
more than two- to four-fold. Indeed, they found a four-fold difference between two sequences that
do not vary in their E2F binding site (TTTTCGCGCCAAAGGA and
TTTTCGCGCCAAAAGGG), suggesting that this may represent the degree of variation within
the assays. They also found that their consensus sequences varied in how efficiently they confer
transcriptional repression when substituted for the E2F site in the B-myb promoter. Indeed, most
of the differences they observe in these reporter assays could be explained by differences in the
efficiency of the sites in mediating the transcriptional effects of E2F, as we observed in our assays.
Other factors may also contribute to the differences they observe: the introduction of consensus
sequences that are extended palindromes might affect the promoter architecture, or unrelated
complexes could fortuitously recognize their sequences.
E2F complexes have different biological roles in the organism. E2F-5, for example, has a
specific, required function in the choroid plexus (Lindeman et al., 1998). E2F-1 promotes
apoptosis (reviewed in Dyson, 1998), whereas E2F-3 may have a specific function in S phase
entry in cycling cells (Leone et al., 1998). E2F complexes also seem to differentially regulate E2Fresponsive genes (DeGregori et al., 1997; Dirks et al., 1998). This study has focused on
sequences selected for their high-affinity binding to E2F complexes, and therefore has not
addressed the possibility that certain low-affinity E2F sites may preferentially bind particular E2F
complexes. Variations in the sequences of high-affinity E2F DNA binding sites, however, do not
appear to confer specificity of E2F complex binding or of transactivation.
We propose that this specificity may instead result from differences in the promoter context
of the E2F sites. This is consistent with observations that, although the B-myb E2F site
specifically mediates transcriptional repression in Go/ G1 and the DHFR site specifically mediates
92
transcriptional activation at the GI/ S transition, these effects do not depend on the sequence of the
E2F site, but on the promoter (Fry et al., 1997). Some promoters may contain transcription
factors that can selectively bind and stabilize particular E2F proteins, conferring target specificity.
The transcriptional activation domains of individual E2F proteins may differ in their ability to
promote transcription, depending on the identity of other factors at the promoter. Interactions of
other transcription factors with DP or pRB family members may also prove important. Indeed, all
of these factors may participate in the specificity of E2F transcriptional regulation in vivo.
93
E.
Materials and Methods
Generationof recombinantE2F
The baculovirus transfer vector pVL 1392 (Pharmingen) was modified by adding the coding
sequence of the hemagluttinin (HA) tag as an EcoRI/ BamHI restriction fragment to create the
plasmid pVL1392-HA. The coding sequences of E2F-2, -3, -4 and DP-1 were subsequently
cloned as BamHI restriction fragments into this vector. Recombinant baculoviruses were
generated in Sf9 cells grown in Grace's media supplemented with 10% foetal bovine serum
according to the manufacturer's instructions. A baculovirus expressing E2F-1 was generously
provided by Brian Dynlacht (Dynlacht et al., 1994). After three rounds of amplification, titers
were determined and Hi5 cells were singly or dually (E2F/DP- 1) with an MOI of 5. Three days
post-infection, cells were harvested and whole cell extracts prepared as previously described
(Moberg et al., 1996), except that the lysates were cleared by centrifugation at 14,000 RPM for
one hour.
Western blotting
Extracts from singly infected Hi5 cells were separated by SDS-PAGE (10%) and
electroblotted to a nylon membrane. Total protein loaded varied from 20 ng (BV-E2F- 1 infection)
to 5 Rg (BV-HAE2F-3 infection) with an equal amount of uninfected Hi5 extract run in parallel as
a control for each infection. The following monoclonal antibodies were used for detection: KH20
(E2F-l) (Helin et al., 1992), LLF2-4/ LLF2-7 (E2F-2), LLF3-1 (E2F-3) (Moberg et al., 1996),
LLF4-2 (E2F-4) and 12CA5 (DP-1). Antibodies LLF2-4, LLF2-7, and LLF4-2 were generated as
previously described (Moberg et al., 1996). Antigen/ antibody complexes were detected by
enhanced chemiluminescence (Amersham).
94
Site selections
A pool of radioactive, double-stranded oligonucleotides 63 residues in length with 25
randomized positions flanked by constant regions was incubated with insect cell extracts containing
heterodimers of DP-1 with E2F-1, -2, -3, or -4. DNA that was bound to proteins was separated
from unbound DNA by gel shift as described below. This protein-bound DNA was eluted from
the gel, amplified by PCR in the presence of X- 32 P-dCTP, and gel purified as described elsewhere
(Pomerantz and Sharp, 1994). This PCR product was then incubated with E2F/ DP- 1 complexes
and the entire procedure repeated another five times. At all stages, only the minimum amount of
insect cell extract necessary to produce a gel shift was included. For E2F-1/ DP-1 complexes, the
amounts of cell lysate used in the selections were 50 ng of total protein, 50 ng, 50 ng, 2.5 ng, 0.5
ng, and 0.5 ng for rounds 1 through 6, respectively; for E2F-2/DP-1, 50 ng, 50 ng, 50 ng, 2.5
ng, 2.5 ng, and 0.5 ng; for E2F-3/ DP-1, 200 ng, 200 ng, 200 ng, 10 ng, 10 ng, and 2.5 ng; for
E2F-4/DP-1, 200 ng, 50 ng, 10 ng, lOng, 10 ng, and 10 ng. Selected sites were amplified by
PCR, cloned into pBluescriptIISK+ (Stratagene) and characterized by a double-stranded dideoxynucleotide sequencing protocol (USB). Sequences were aligned with assistance from the
CLUSTAL program found at http://dot.imgen.bcm.tmc.edu:933 1/multialign/Options/clustalw.html.
Gel shift assays and affinity studies
Gel shifts were performed as described (Moberg et al., 1996) with extracts from cells
infected with baculoviruses expressing DP and E2F subunits. For supershifting E2F/DNA
complexes the following monoclonal antibodies specific for the indicated E2F were included in the
binding reaction: KH20 (E2F-1) (Helin et al., 1992), LLF2-5 (E2F-2), LLF3-1 (E2F-3) (Moberg
et al., 1996), and LLF4-1 (E2F-4) (Verona et al., 1997). Antibody LLF2-5 was generated exactly
as previously described (Moberg et al., 1996). The sequences of the double-stranded
oligonucleotides used in the binding site competition experiments are as follows: wild-type E2:
ATTTAAGTTTCGCGCCCTTTCCAA; mutant E2: ATTTAAGTTTCGATCCCTTTCCAA; E295
CGGC: TGACTAGTTTCCCGGCTTTCTCAAA; E2-CGGG: TGACTAGTTTCCCGGGTTTCTCAAA; 1.30: TCGAGGGTGGCCTTCGCTTTCCCGCCAAATGGATC; 3P: TCGAGGCCCAACCGTTTCCCGCGTGTCACGGATC; 4E: TCGAGGTGAAAAGTATTCCCGCGCTTGTGGATC. Binding quantitation was performed on a STORM 860 using ImageQuant vi.2
(Molecular Dynamics).
Transient transfectionsand transcriptionalreporterassays
C33-A cells were maintained in DMEM supplemented with 10% fetal bovine serum;
transient transfections were performed as described (Trimarchi et al., 1998). For reporter assays,
chloramphenicol acetyltransferase (CAT) reporter constructs were generated by replacing the XhoI/
Sal region of Elb-CAT with two tandem copies of the following sequences: for 2xE2-CAT,
GAAGTTTCGCGCCCTTTG; for 2xE2mut-CAT, GAAGTTTCGATCCCTTTG; for 2x1.30CAT, GAACTTTCCCGCCAAATG; for 2x3P-CAT, GAAGTTTCCCGCGTGTCG; for 2x4ECAT, GAATATTCCCGCGCTTTG. 4 pg of one of these reporters was cotransfected with 2 gg
of RSV-luc, 19 gg of pBluescriptIIKS+ and the indicated CMV-based E2F and DP- I expression
constructs; the total amount of CMV promoter in each transfection for reporter assays was
maintained at 2 gg by the addition of CMV plasmid as appropriate. CAT and luciferase activities
were assayed as described (Lees et al., 1993).
96
F.
Acknowledgements
We gratefully acknowledge Brian Dynlacht for providing the E2F-1-expressing
baculovirus, many members of the Lees lab, past and present, for helpful advice and discussions,
Steve Bell and Patrick McCaw for advice about quantitation, Joel Pomerantz and Ernest Fraenkel
for critical help in setting up the site selection experiments, and Raluca Verona and Jeff Trimarchi
for critical reading of sections of this manuscript.
97
G.
References
Bandara, L. R., Buck, V. M., Zamanian, M., Johnston, L. H., and La Thangue, N. B. (1993).
Functional synergy between DP-1 and E2F-1 in the cell cycle-regulating transcription factor
DRTF1/E2F. Embo J 12, 4317-24.
Bandara, L. R., and La Thangue, N. B. (1991). Adenovirus Ela prevents the retinoblastoma
gene product from complexing with a cellular transcription factor. Nature 351, 494-7.
Beijersbergen, R. L., and Bernards, R. (1996). Cell cycle regulation by the retinoblastoma family
of growth inhibitory proteins. Biochim Biophys Acta 1287, 103-20.
Beijersbergen, R. L., Kerkhoven, R. M., Zhu, L., Carlee, L., Voorhoeve, P. M., and Bernards,
R. (1994). E2F-4, a new member of the E2F gene family, has oncogenic activity and associates
with p107 in vivo. Genes Dev 8, 2680-90.
Cao, L., Faha, B., Dembski, M., Tsai, L. H., Harlow, E., and Dyson, N. (1992). Independent
binding of the retinoblastoma protein and p107 to the transcription factor E2F. Nature 355, 1769.
Cartwright, P., Muller, H., Wagener, C., Holm, K., and Helin, K. (1998). E2F-6: a novel
member of the E2F family is an inhibitor of E2F- dependent transcription. Oncogene 17, 611-23.
Chellappan, S. P., Hiebert, S., Mudryj, M., Horowitz, J. M., and Nevins, J. R. (1991). The
E2F transcription factor is a cellular target for the RB protein. Cell 65, 1053-61.
Chittenden, T., Livingston, D. M., and DeCaprio, J. A. (1993). Cell cycle analysis of E2F in
primary human T cells reveals novel E2F complexes and biochemically distinct forms of free E2F.
Mol Cell Biol 13, 3975-83.
Cobrinik, D., Lee, M. H., Hannon, G., Mulligan, G., Bronson, R. T., Dyson, N., Harlow, E.,
Beach, D., Weinberg, R. A., and Jacks, T. (1996). Shared role of the pRB-related p130 and
p107 proteins in limb development. Genes Dev 10, 1633-44.
Cobrinik, D., Whyte, P., Peeper, D. S., Jacks, T., and Weinberg, R. A. (1993). Cell cyclespecific association of E2F with the p130 ElA-binding protein. Genes Dev 7, 2392-404.
DeGregori, J., Leone, G., Miron, A., Jakoi, L., and Nevins, J. R. (1997). Distinct roles for
E2F proteins in cell growth control and apoptosis. Proc Natl Acad Sci U S A 94, 7245-50.
DePinho, R. A. (1998). Transcriptional repression. The cancer-chromatin connection [news;
comment]. Nature 391, 533, 535-6.
Devoto, S. H., Mudryj, M., Pines, J., Hunter, T., and Nevins, J. R. (1992). A cyclin A-protein
kinase complex possesses sequence-specific DNA binding activity: p33cdk2 is a component of the
E2F-cyclin A complex. Cell 68, 167-76.
Dirks, P. B., Rutka, J. T., Hubbard, S. L., Mondal, S., and Hamel, P. A. (1998). The E2Ffamily proteins induce distinct cell cycle regulatory factors in p1 6-arrested, U343 astrocytoma cells
[In Process Citation]. Oncogene 17, 867-76.
Dynlacht, B. D., Flores, 0., Lees, J. A., and Harlow, E. (1994). Differential regulation of E2F
transactivation by cyclin/cdk2 complexes. Genes Dev 8, 1772-86.
98
Dyson, N. (1998). The regulation of E2F by pRB-family proteins. Genes Dev 12, 2245-62.
Dyson, N., Dembski, M., Fattaey, A., Ngwu, C., Ewen, M., and Helin, K. (1993). Analysis of
p107-associated proteins: p107 associates with a form of E2F that differs from pRB-associated
E2F-1. J Virol 67, 7641-7.
Ewen, M. E., Xing, Y. G., Lawrence, J. B., and Livingston, D. M. (1991). Molecular cloning,
chromosomal mapping, and expression of the cDNA for p107, a retinoblastoma gene productrelated protein. Cell 66, 1155-64.
Field, S. J., Tsai, F. Y., Kuo, F., Zubiaga, A. M., Kaelin, W. G., Jr., Livingston, D. M.,
Orkin, S. H., and Greenberg, M. E. (1996). E2F-1 functions in mice to promote apoptosis and
suppress proliferation. Cell 85, 549-61.
Fry, C. J., Slansky, J. E., and Farnham, P. J. (1997). Position-dependent transcriptional
regulation of the murine dihydrofolate reductase promoter by the E2F transactivation domain. Mol
Cell Biol 17, 1966-76.
Gaubatz, S., Wood, J. G., and Livingston, D. M. (1998). Unusual proliferation arrest and
transcriptional control properties of a newly discovered E2F family member, E2F-6. Proc Natl
Acad Sci U S A 95, 9190-5.
Ginsberg, D., Vairo, G., Chittenden, T., Xiao, Z. X., Xu, G., Wydner, K. L., DeCaprio, J. A.,
Lawrence, J. B., and Livingston, D. M. (1994). E2F-4, a new member of the E2F transcription
factor family, interacts with p107. Genes Dev 8, 2665-79.
Hannon, G. J., Demetrick, D., and Beach, D. (1993). Isolation of the Rb-related p130 through
its interaction with CDK2 and cyclins. Genes Dev 7, 2378-91.
Helin, K., Lees, J. A., Vidal, M., Dyson, N., Harlow, E., and Fattaey, A. (1992). A cDNA
encoding a pRB-binding protein with properties of the transcription factor E2F. Cell 70, 337-50.
Helin, K., Wu, C. L., Fattaey, A. R., Lees, J. A., Dynlacht, B. D., Ngwu, C., and Harlow, E.
(1993). Heterodimerization of the transcription factors E2F- 1 and DP- 1 leads to cooperative transactivation. Genes Dev 7, 1850-61.
Hiebert, S. W., Chellappan, S. P., Horowitz, J. M., and Nevins, J. R. (1992). The interaction
of RB with E2F coincides with an inhibition of the transcriptional activity of E2F. Genes Dev 6,
177-85.
Hijmans, E. M., Voorhoeve, P. M., Beijersbergen, R. L., van 't Veer, L. J., and Bernards, R.
(1995). E2F-5, a new E2F family member that interacts with p130 in vivo. Mol Cell Biol 15,
3082-9.
Jacks, T., Fazeli, A., Schmitt, E. M., Bronson, R. T., Goodell, M. A., and Weinberg, R. A.
(1992). Effects of an Rb mutation in the mouse [see comments]. Nature 359, 295-300.
Krek, W., Livingston, D. M., and Shirodkar, S. (1993). Binding to DNA and the retinoblastoma
gene product promoted by complex formation of different E2F family members. Science 262,
1557-60.
LeCouter, J. E., Kablar, B., Hardy, W. R., Ying, C., Megeney, L. A., May, L. L., and
Rudnicki, M. A. (1998a). Strain-dependent myeloid hyperplasia, growth deficiency, and
99
accelerated cell cycle in mice lacking the Rb-related p107 gene [In Process Citation]. Mol Cell Biol
18, 7455-65.
LeCouter, J. E., Kablar, B., Whyte, P. F., Ying, C., and Rudnicki, M. A. (1998b). Straindependent embryonic lethality in mice lacking the retinoblastoma- related p130 gene. Development
125, 4669-4679.
Lee, M. H., Williams, B. 0., Mulligan, G., Mukai, S., Bronson, R. T., Dyson, N., Harlow, E.,
and Jacks, T. (1996). Targeted disruption of p107: functional overlap between p107 and Rb.
Genes Dev 10, 1621-32.
Lees, E., Faha, B., Dulic, V., Reed, S. I., and Harlow, E. (1992). Cyclin E/cdk2 and cyclin
A/cdk2 kinases associate with p107 and E2F in a temporally distinct manner. Genes Dev 6,
1874-85.
Lees, J. A., Saito, M., Vidal, M., Valentine, M., Look, T., Harlow, E., Dyson, N., and Helin,
K. (1993). The retinoblastoma protein binds to a family of E2F transcription factors. Mol Cell
Biol 13, 7813-25.
Leone, G., DeGregori, J., Yan, Z., Jakoi, L., Ishida, S., Williams, R. S., and Nevins, J. R.
(1998). E2F3 activity is regulated during the cell cycle and is required for the induction of S
phase. Genes Dev 12, 2120-30.
Li, Y., Graham, C., Lacy, S., Duncan, A. M., and Whyte, P. (1993). The adenovirus E lAassociated 130-kD protein is encoded by a member of the retinoblastoma gene family and
physically interacts with cyclins A and E. Genes Dev 7, 2366-77.
Lindeman, G. J., Dagnino, L., Gaubatz, S., Xu, Y., Bronson, R. T., Warren, H. B., and
Livingston, D. M. (1998). A specific, nonproliferative role for E2F-5 in choroid plexus function
revealed by gene targeting. Genes Dev 12, 1092-8.
Mayol, X., Grana, X., Baldi, A., Sang, N., Hu, Q., and Giordano, A. (1993). Cloning of a
new member of the retinoblastoma gene family (pRb2) which binds to the ElA transforming
domain. Oncogene 8, 2561-6.
Moberg, K., Starz, M. A., and Lees, J. A. (1996). E2F-4 switches from p130 to p107 and pRB
in response to cell cycle reentry. Mol Cell Biol 16, 1436-49.
Mudryj, M., Devoto, S. H., Hiebert, S. W., Hunter, T., Pines, J., and Nevins, J. R. (1991).
Cell cycle regulation of the E2F transcription factor involves an interaction with cyclin A. Cell 65,
1243-53.
Pomerantz, J. L., and Sharp, P. A. (1994). Homeodomain determinants of major groove
recognition. Biochemistry 33, 10851-8.
Sardet, C., Vidal, M., Cobrinik, D., Geng, Y., Onufryk, C., Chen, A., and Weinberg, R. A.
(1995). E2F-4 and E2F-5, two members of the E2F family, are expressed in the early phases of
the cell cycle. Proc Natl Acad Sci U S A 92, 2403-7.
Shirodkar, S., Ewen, M., DeCaprio, J. A., Morgan, J., Livingston, D. M., and Chittenden, T.
(1992). The transcription factor E2F interacts with the retinoblastoma product and a p107-cyclin A
complex in a cell cycle-regulated manner. Cell 68, 157-66.
100
Starostik, P., Chow, K. N., and Dean, D. C. (1996). Transcriptional repression and growth
suppression by the p107 pocket protein. Mol Cell Biol 16, 3606-14.
Tao, Y., Kassatly, R. F., Cress, W. D., and Horowitz, J. M. (1997). Subunit composition
determines E2F DNA-binding site specificity. Mol Cell Biol 17, 6994-7007.
Trimarchi, J. M., Fairchild, B., Verona, R., Moberg, K., Andon, N., and Lees, J. A. (1998).
E2F-6, a member of the E2F family that can behave as a transcriptional repressor. Proc Natl Acad
Sci U S A 95, 2850-5.
Vairo, G., Livingston, D. M., and Ginsberg, D. (1995). Functional interaction between E2F-4
and p 1 30: evidence for distinct mechanisms underlying growth suppression by different
retinoblastoma protein family members. Genes Dev 9, 869-81.
Verona, R., Moberg, K., Estes, S., Starz, M., Vernon, J. P., and Lees, J. A. (1997). E2F
activity is regulated by cell cycle-dependent changes in subcellular localization. Mol Cell Biol 17,
7268-82.
Wells, J., Held, P., Illenye, S., and Heintz, N. H. (1996). Protein-DNA interactions at the major
and minor promoters of the divergently transcribed dhfr and rep3 genes during the Chinese
hamster ovary cell cycle. Mol Cell Biol 16, 634-47.
Wells, J. M., Illenye, S., Magae, J., Wu, C. L., and Heintz, N. H. (1997). Accumulation of
E2F-4.DP- 1 DNA binding complexes correlates with induction of dhfr gene expression during the
GI to S phase transition. J Biol Chem 272, 4483-92.
Wu, C. L., Zukerberg, L. R., Ngwu, C., Harlow, E., and Lees, J. A. (1995).
association of E2F and DP family proteins. Mol Cell Biol 15, 2536-46.
In vivo
Zhu, L., van den Heuvel, S., Helin, K., Fattaey, A., Ewen, M., Livingston, D., Dyson, N., and
Harlow, E. (1993). Inhibition of cell proliferation by p107, a relative of the retinoblastoma
protein. Genes Dev 7, 1111-25.
Zwicker, J., Liu, N., Engeland, K., Lucibello, F. C., and Muller, R. (1996). Cell cycle
regulation of E2F site occupation in vivo. Science 271, 1595-7.
101
CHAPTER III
E2F-6, a member of the E2F family that can behave as a
transcriptional repressor
JM Trimarchi, B Fairchild, R Verona, K Moberg, N Andon,
and JA Lees
PNAS, 1998, Vol. 95: 2850-5.
Copyright (1998) National Academy of Sciences, U. S. A.
(Author's contribution: identification of E2F-6)
A.
Abstract
The E2F family of proteins is required to establish the correct cell-cycle-dependent
transcription of genes that direct the process of cell division. All previously identified E2F proteins
can act in a similar manner; depending on whether or not they are associated with the cell cycle
inhibitors the retinoblastoma protein (pRB), p107, or p1 30, they can either repress or activate the
transcription of E2F responsive genes. We now report the cloning and characterization of another
E2F family member, E2F-6, whose structure is reminiscent of the dominant inhibitors of other
transcription factor families. The dimerization and DNA-binding properties of E2F-6 are similar to
those of the other E2F family members. However, it is not regulated by pRB, p107, or p130, and
it is unable to activate transcription. Instead, it can act to repress the transcription of E2F
responsive genes by countering the activity of the other E2F complexes via a pRB-, p107-, or
p130-independent mechanism.
102
B.
Introduction
The retinoblastoma gene (RB-1) is one of the best studied of the tumor suppressor genes.
Although it was originally identified by virtue of its absence in retinoblastomas, subsequent studies
have shown that it is absent or mutated in at least one-third of all human tumors (Weinberg, 1992).
The product of this gene [the retinoblastoma protein (pRB)] is also an essential target of the
transforming proteins of the small DNA tumor viruses (Dyson and Harlow, 1992). In
untransformed cells, the growth-suppressive properties of pRB are regulated by its cell-cycledependent phosphorylation (Bartek et al., 1996). This phosphorylation is catalyzed by the cellcycle-dependent kinase cyclin D-CDK4/ 6 (Bartek et al., 1996). Consistent with this hypothesis,
many human tumors contain activating mutations within either the cyclin DI or CDK4 genes or
have lost the CDK4/ 6- specific inhibitor p16 (Pollock et al., 1996).
The growth-suppressive properties of pRB are largely dependent upon its ability to regulate
a cellular transcription factor, E2F (Nevins, 1992; Bartek et al., 1996). Many E2F-responsive
genes have been identified, and their products are required for entry into, or passage through, the
cell cycle. Consistent with its antiproliferative role, pRB inhibits the transcriptional activity of E2F
(Hiebert et al., 1992). Moreover, the resultant pRB-E2F complex actively represses the
transcription of E2F-responsive genes by blocking the activity of adjacent transcription factors
(Weintraub et al., 1992; Bremner et al., 1995; Weintraub et al., 1995). Phosphorylation of pRB
causes it to dissociate from E2F, thereby switching E2F-responsive genes from the repressed to
the induced state.
E2F is regulated by two additional proteins called p107 and p130 (Beijersbergen and
Bernards, 1996). These proteins share significant sequence similarity with pRB (Dyson and
Harlow, 1992; Beijersbergen and Bernards, 1996), and in overexpression assays they repress E2F
in a similar manner to pRB (Zhu et al., 1993; Starostik et al., 1996). However, genetic studies
indicate that pRB, p107, and p130 have distinct properties in vivo. Whereas pRB is mutated in
30% of all human tumors, neither p107 nor p130 has been proven to suppress tumorigenesis
103
(Weinberg, 1996). Moreover, analysis of mutant mouse strains indicates pRB is essential for
development, whereas loss of p107 or p130 in a similar genetic background does not alter viability
or tumor incidence (Clarke et al., 1992; Jacks et al., 1992; Lee et al., 1992; Cobrinik et al., 1996;
Lee et al., 1996). It is unclear exactly how the different biological properties of these proteins
relate to the manner in which they regulate E2F, but pRB, p107, and p130 are known to bind E2F
at defined but distinct stages of the cell cycle (Beijersbergen and Bernards, 1996).
To date, at least seven human genes have been identified that encode components of the
E2F transcriptional activity (Beijersbergen and Bernards, 1996). These can be divided into two
distinct groups, named E2F (E2F-1 through -5) and DP (DP-1 and -2). The products of these
groups heterodimerize to give rise to high-affinity DNA-binding activity and transcriptional
activation (Bandara et al., 1993; Helin et al., 1993; Krek et al., 1993; Wu et al., 1995). In vivo
studies confirm that the endogenous E2F activity is generated from the combined properties of
multiple E2F-DP complexes (Wu et al., 1995; Moberg et al., 1996). The individual E2F-DP
species have different pRB-, p107-, and p130-binding properties. Although the DP subunit is
essential, the E2F moiety mediates the specificity of this interaction. Complexes containing E2F- 1,
-2, or -3 associate with pRB but not p107 or p130 in vivo (Dyson et al., 1993; Lees et al., 1993).
In contrast, E2F-4 and -5 complexes are capable of binding p107 and p130 (Beijersbergen et al.,
1994; Ginsberg et al., 1994; Hijmans et al., 1995). Consistent with these findings, sequence
comparisons suggest that the family of E2F proteins can be subdivided into two distinct
subgroups. The pRB-specific E2Fs (E2F-1 through -3) have an extended N-terminal domain that
is absent in both E2F-4 and -5. There is also considerable variation in the sequence of DNAbinding, dimerization, and transactivation domains between members of the two E2F subgroups
(E2F-1 through -3 versus E2F-4 and -5). These observations have led to the hypothesis that these
two subgroups will play distinct roles in vivo that will at least partially account for the different
biological consequences of loss of pRB, p107, or p130.
To examine the biochemical and functional properties of the endogenous E2Fs, we had
previously developed specific antisera for each of the components of the E2F family (Moberg et
104
al., 1996). With these reagents, we have been able to demonstrate that the known E2F proteins are
unable to account for all of the endogenous E2F-DP DNA-binding activity (Moberg et al., 1996).
Specifically, gel shift analysis revealed that an activity exists in gel shift analysis with the same
DNA binding specificity as other E2F complexes, but that is not recognized by antibodies to the
known E2F proteins. The activity is, however, recognized by anti-DP antibodies. Thus, there
exists at least one novel protein that, like known E2F proteins, binds to DP and to an E2F DNA
binding sequence. We have therefore sought to identify additional members of the E2F family. In
this study, we describe the cloning and characterization of an additional E2F family member, E2F6. The DNA-binding and dimerization domains of E2F-6 are highly related to the corresponding
domains of the previously identified family members, but this protein lacks the sequences
necessary for either transactivation or pRB-, p107-, or p130-binding. We conclude that the E2F
family contains a third subgroup of proteins whose structure is highly reminiscent of the dominant
inhibitors of other transcription factor families.
105
C. Results
C1.
Isolation of cDNAs encoding an E2F family member
At the start of this study, five genes had been identified that encode members of the E2F
family of proteins. We have shown previously that these proteins account for a significant
proportion of the endogenous E2F-DP complexes, but there must be at least one additional E2F
(Moberg et al., 1996). The greatest homology between the known E2F family members maps to
the DNA-binding domain. In an attempt to identify additional E2F family members, we searched
the EST database for cDNA clones with homology to E2F DNA-binding domains. As expected,
we were able to identify ESTs derived from E2F-1, -2, -3, -4, or -5. In addition, this search
identified one mouse (GenBank accession no. AA041604/ AA050073) and one human (GenBank
accession no. AA127210) EST that did not correspond to the known E2Fs. These clones were
highly related to one another at the nucleotide level, suggesting that they were mouse and human
homologs of the same gene. The core of previously identified E2F DNA-binding domains
includes an invariant stretch of fifteen amino acids (QKRRIYDITNVLEGI); this motif was highly,
but not completely, conserved in one of the predicted translation products and almost all of the
variation between these two clones was in the third base position of this ORF. These data strongly
suggested that this was the correct ORF and that these cDNAs were excellent candidates to encode
another E2F.
With both the EST database and standard library screening techniques, we identified
multiple overlapping cDNAs that encompassed 2,027 bp of the human gene. This sequence starts
within the ORF and extends to the poly(A)+ tail. None of the human cDNAs diverged from this
assembled sequence. We also identified two noncontiguous mouse cDNAs. The first is a 638-bp
clone that shares significant homology with the 3' untranslated region of the human gene. The
second includes 216 bp of 5' untranslated region followed by the initiating ATG (as judged by the
presence of a good Kozak consensus and an upstream in-frame termination codon) and 536 bp of
ORF. Sequence comparison suggested that the human cDNA lacks only a short region (18 bp) of
5' coding sequence (data not shown). The overlapping regions of the human and mouse sequence
106
are highly conserved at both the nucleotide (83%) and predicted amino acid (92%) level (Fig. lA).
We therefore concluded that we had identified the mouse and human orthologs of another gene,
hereafter designated E2F-6, that share significant homology with the known E2Fs.
The previously identified E2F proteins have been divided into two distinct subgroups
(E2F- 1 through -3 versus E2F-4 and -5) on the basis of differences in their amino acid sequence.
The central portion of the E2F-6 protein shared considerable homology with the domains of E2F- 1
through -5 that are known to mediate their dimerization (both the leucine zipper domain and region
of homology known as the marked box) and DNA-binding properties. Within these domains,
many of the residues that are absolutely conserved between E2F- 1 through -5 are also maintained
in E2F-6. However, at a significant proportion of these conserved residues, E2F-6 had alternative
codon usage from that found in the other E2F genes (data not shown). Moreover, when we
examined the amino acid positions that are known to distinguish E2F- 1 through -3 from E2F-4 and
-5, the corresponding residue in E2F-6 rarely fit into either subgroup (see Fig. 1B). The
distinction between E2F-6 and the other E2F family members was further underscored by the
degree of sequence variation outside of the DNA-binding and dimerization domains. The Nterminal domain of E2F-6 was of intermediate length, relative to those of the two E2F subgroups.
More importantly, the E2F-6 protein terminated just 42 amino acids beyond the marked box motif
(the C-terminal portion of the dimerization domain). As a result this protein lacks the sequences
that are known to mediate either the transcriptional activation of pRB-, p107-, or p130-binding
properties of the other E2F proteins. Thus, these findings suggest that E2F-6 represents a third
subclass of the E2F protein family that is likely to display distinct properties from the previously
identified members.
C2.
E2F-6 is widely expressed in vivo.
E2F-1 through -5 are all expressed in a wide variety of tissues (Helin et al., 1992; Kaelin et
al., 1992; Ivey-Hoyle et al., 1993; Lees et al., 1993; Beijersbergen et al., 1994; Ginsberg et al.,
1994; Hijmans et al., 1995). We were therefore interested to establish the expression pattern of
107
Figure 1.
A
hE2F6
HEKLPSLLLDPTEETVRRRCRDPINVEGLLPSKIRINLEDNVQYVSMRKALKVKRPRFDVSLVYLTRKFMDLVRSAPGGILDLNKVAT
mE2F6
MSQQRTARRQPSLLVDPTAQTVRRRCRDPINVEGNLPSKIRINLEENVQYVSMRKALKVKRPRFDVSLVYLTRKFMDLVRSAPACILDLNKVAT
hE2F6
KLGVRKRRVYDITNVLDGIDLVEKKSKNHIRWIGSDLSNFGAVPQQKKLQEELSDLSAMEDALDELIKDCAQQLFELTDDKENERLAYVTYQDI
mE2F6
KLGVRKRRVYDITNVLDGIELVEKKSKNHIRWIGSDLNNFGAAPQQKKLQAELSDLSAMEDALDELIKDCAQQLLELTDDKENE
hE2F6
HSIQAFHEQIVIAVKAPAETRLDVPAPREDSVTVHIRSTNEPIDVYLCEVEQGQTSNKRSEGVGTSSSESTHPEGPEEEENPQQSEELLEVSN*
B
E2F2
E2F3
E2F1
E2F6
MLQGPR . ALASAAGQTPKVVPAMSPTELWPSGLSSPQLCPATATYYTPLYPQTAPPAAAPG ..... TCLDATPHGPEGQV .........
MRKGIQPALEQYLVTAGGGEGAAVVAAAAAASMDKRALLASPGFAAAAAAAAAPGAYIQILTTNTSTTSCSSSLQSGAVAA. . GPLLPSAPGAEQTAG ..... SLLYTTPHGPSSRAGLLQQPPALG
MALAGAPAGGPCAPALEALLGAGALRLLDSSQIVIISAAQDASAPPAPTGPAAPAAGPCDPDLLLFATPQAPRPTPS .........
HEKLPS.....LLLDPTEETVRRRC..........
E2F2
E2F3
E2F1
E2F6
E2F4
E2F5
QKRRIYDITNVLEGIQLIRKKAKNNIQWVGRGM
n
.VRCLPAGRLPAKRKLDLEGIGRPVVPE. FPTPKGKC. IRVDGLPSPKTPKSPGEKTRYDTSLGLLTKKFI YEnSESEGID4NAAEVLDV.
RGGSGGGGGPPAKRRLELGESGHQYLSDGLKTPKGKGRAALRSPDSPKTPKSPSEKTRYDTSLGLLTKKFIQLLSESEDGVLDLNWAAEVLDV.QKRRIYDITNVLEGIQLIRKKAKNNIQWMGCSL
HPGKGVKSPGEKSRYETSLNLTTKRFLELLSHSADGVVDLNWAAEVLKV.QKRRIYDITNVLEGIQLIAKKSKNHIQWLGSH.
. APRPALGRPPVKRRLDLET . DHQYLAESSGPARGRGR ......
RDPINVEGLLPSKIRINLED ...................... NVQYVSMRKALK . .VKRPRFDVSLVYLTRKFMDLVRSAPGGILDLNKVATKLGVR.KRRVYDITNVLDGIDLVEKKSKNHIRWIGSDL
GTPSRHEKSLGLLTTKFVSLLQEAKDGVLDLKLAADTLAVRQKRRIYDITNVLEGIGLIEKKSKNSIQWKGVGP
MAEAGPQAPPPP ......
MAAAE PASSGQQAPAGQGQGQRPPPQPPQAQAPQPPPPPQLGGAGGGSSRHEKSLQLLTTKVS LLEAKDGVDKA__DTLAVRKRRIYDITNVLEGIDLEKKSKNS OWKGVGA
E2F2
E2F3
E2F1
E2F6
E2F4
E2F5
VQEPDS
TEDNLQIYLKSTQGPIEVYLCPEE ......
T. ETHS
E. SLQIHLASTQGPIEVYLCPEE ...
TVGGIS
. . TTVGVGGRLEGLTQDLRQLQESEQQLDHLMNICTTQLRLLSEDTDSQRLAYVTCQDLRSIADPAEQMVMVIKAPPETQLQAVDS ..... SE.NFQISLKSKQGPIDVFLCPEE ......
QGQTSN
......
REDSVTVHIRSTNEPIDVYLCEVE
.....
. SM. FGAVPQQKKLQEELSDLSAMEDALDELIKDCAQQLFELTDDKENERLAYVTYQDIHSIQAFHEQIVIAVKAPAETRLDVPAP
GCNTREIADKLIELKAEIEELQQREQELDQHKVWVQQSIRNVTEDVQNSCLAYVTHEDI. . CRCFAGDTLLAIRAPSGTSLEVPIPE .GLNGQKKYQIHLKSVSGPIEVLLVNKEAWSSPPVAVPVP
GCNTKEVIDRLRYLKAEIEDLELKERELDQQKLWLQQSIKNVMDDSINNRFSYVTHEDI. .CNCFNGDTLLAIQAPSGTQLEVPIPEMGQNGQKKYQINLKSHSGPIHVLLINKESSSSKPVVFPVP
E2F2
E2F3
E2F1
E2F6
E2F4
E2F5
PSEEPLPSTSTLCPSPD.......SAQPSSSTDPSIMEPTASSVPAPAPTPQQAPPPPSLVPLEATDSLLELPHPLLQQTEDQFLSPTLACSS..................................
PMKTNNQDHNGNIPKPA........SKD..........LASTNSG..........HSDCSVSMGNL.... SPLASPANLLQQTEDQIPSNL ... EG. ..................................
PGKTPSQEVTSEEENRATDSATIVSPPPSSPPSSLTTDPSQSLLSLEQEPLLSRMGSLRAPVDEDRLSPLVAADSLLEHVREDFSGLL ... PE ..................................
KRSEGVGTSSSESTHPE ....... .GPEEEENPQQSEELLEVSN*
PPEDFVQSPSAVSTPPPLPKPALAQSQEASRPNSPQLTPTAVPGSAEVQGMAGPAAEITVSGGPGTDSKDSGELSSLPLGPTTLDTRPLQSSALLDSSSSSSSSSSSSSNSNSSSSSGPNPSTSFEP
.RSQALQQTSATDISSGS ........................
PPDDLTQPSSQSLTPVTPQKSSMA ......................................... TQNLPEQHVSE .............
DNA binding domain
Marked box
Leucine zipper
. FEDPTRPGKQQQLGQELKELMNTEQALDQLIQSCSLSFKHLTEDKANKRLAYVTYQDIRAVGNFKEQTVIAVKAPPQTRLEVPDR .....
. SEDGGMLAQCQGLSKEVTELSQEEKKLDELIQSCTLDLKLLTEDSDNQRLAYVTYQDIRKISGLKDQTVIVVKAPPETRLEVPDS ..
*
0
00
Pocket protein binding
PLISFSPSLDQDDYLWGLEAGEGISDLFDSYDLGDLLIN*
.........................................
E2F2
......................................... PFVNLLPPLLQEDYLLSLGEEEGISDLFDAYDLEKLPLVEDFMCS*
E2F3 ......
......................................... EFISLSPPHEALDYHFGLEEGEGIRDLFD.CDFGDLTPLDF*
E2F1
E2F6
IKADPTGVLELPKELSEIFDPTRECMSSELLEELMSSEVFAPLLRLSPPPGDHDYIYNLDESEGVCDLFDLPVLNL*
E2F4
E2F5 ............................. ISGDIIDELMSSDVF. PLLRLSPTPAD DYNFNLDDNEGVCDLFDVQILNY*
Figure 1. (preceding page) E2F-6 is another member of the E2F family. (A) Amino acid
sequence comparison of human versus murine E2F-6. (B) Amino acid sequence comparison of
the human E2F proteins. Domains responsible for DNA binding, dimerization (leucine zipper,
marked box), and pocket protein binding are indicated. Conserved residues are denoted in
boldface type.
109
E2F-6. Initially, we isolated poly(A)+ RNA from the indicated human cell lines and screened them
for the presence of the E2F-6 mRNA by Northern blotting using a probe derived from the coding
sequence (Fig. 2B). In these and every other cell line examined, this probe hybridized with similar
stringency to two distinct messages. To date, each of the E2F-6 cDNAs that we have identified
corresponds to a single common transcript. It was therefore unclear why the E2F-6 probe detected
two different mRNAs. To address this issue, we rescreened these Northern blots with a probe
derived from the 3' untranslated region. This second probe also hybridized to the same two
transcripts (data not shown), indicating that they must both contain at least some of the sequences
from both the coding and 3' untranslated region of our E2F-6 cDNA. We also examined the
expression of E2F-6 in human tissues. For these experiments, we screened a human tissue blot
(CLONTECH) with the probe corresponding to the E2F-6 coding region (Fig. 2A). As with
cultured cells, we detected two E2F-6 transcripts in every tissue examined. We therefore conclude
that E2F-6 will be expressed in most, if not all, cell and tissue types. At this time, we are unable to
explain the structural difference(s) between the two mRNAs.
Given the existence of the two E2F-6 transcripts, we could not rule out the possibility that
alternative splicing could give rise to protein products with two distinct C-termini, one that
corresponds to that encoded by our identified cDNA clones and one that more closely resembles
the domain structure of the previously identified E2F proteins. To address this issue, we raised a
polyclonal antiserum against the predicted human E2F-6 protein. Control experiments confirmed
that the antiserum recognized epitopes throughout the E2F-6 protein but did not cross-react to any
of the other known E2Fs (data not shown). When tested on Western blots, the antiserum detected
a single 35-kDa protein in all tested cell lines that migrated slightly faster than an HA-tagged
version of the E2F-6 protein produced by transient transfection (Fig. 2B and data not shown). We
therefore conclude that the E2F-6 protein is expressed in vivo and it exists predominantly in the
form predicted by the cDNA clones.
110
Figure 2.
+A4
<<0
.A,2*B2F
-g
Sa
ww"
Figure 2. (preceding page) Expression patterns of E2F-6. (A) Northern blot analysis of
poly(A)+ RNA from the indicated cell lines and tissues (human tissue blot from CLONTECH),
screened with a probe derived from the full-length E2F-6 coding sequence. (B) Western blot
analysis of extracts derived from C33-A cells transiently transfected with pCMV-HA-E2F-6 (200
ng) and from untransfected U2OS, C33-A, ML- 1, and 293 cells (50 jg). The positions of HAE2F-6 and endogenous E2F-6 proteins are indicated by arrows. The band running beneath HAE2F-6 is a degradation product of this transfected protein (data not shown). The asterisk denotes a
nonspecific band.
112
C3.
E2F-6 displays low-affinity E2F DNA-binding activity.
Given the unusual structure of E2F-6, we wished to establish whether this protein retained
any of the properties of the known E2F family members. All previously identified E2Fs can bind
to either of the human DPs and this heterodimerization is known to be a prerequisite for the highaffinity DNA binding. We therefore initiated our analysis by comparing the ability of E2F-1, E2F4, or an HA-tagged version of E2F-6 in the presence or absence of the DP (either DP- 1 or DP-2)
or pocket (pRB, p107, or p130) proteins. The transfectants were labelled with [35 S]-methionine
and then subjected to immunoprecipitation with the indicated antibodies (Fig. 3A). Although there
was considerable variation in the efficiency of the individual transfections, these experiments
allowed us to assess the DP- and pocket protein-binding properties of the individual E2F proteins.
A monoclonal antibody specific for the HA tag was able to recover either DP- 1 or DP-2 in
approximately stoichiometric amounts with HA-E2F-6 (Fig. 3A, lanes 18 and 21). Moreover,
HA-E2F-6 seemed to bind to these proteins as well as either E2F-1 or E2F-4 (Fig. 3A, compare
lanes 18 and 21 with lanes 3 and 11 and lanes 6 and 14). This confirmed that E2F-6 is able to
heterodimerize with the DP proteins despite the sequence variation within its presumed
dimerization domain. Consistent with previous studies, the E2F-1-DP-l (Fig. 3A, lanes 4 and 5)
and E2F-4-DP-l (Fig. 3A, lanes 12 and 13) complexes were able to associate with pRB and p107
when overexpressed. In contrast, we did not detect any interaction between the HA-E2F-6-DP
complexes any of the three pocket proteins (Fig. 3A, lanes 19 and 20 and data not shown). This
finding is consistent with the absence of a pocket protein-binding motif within the E2F-6 protein
sequence.
After establishing that E2F-6 can associate with the DP proteins, we next examined the
DNA-binding activity of these heterodimers. Transient transfection was used to generate the
relevant E2F complexes, exactly as described above, and the resultant cell extracts were measured
for total protein content. These were then tested in gel retardation assays with the consensus E2F
site and equivalent levels of total protein (Fig. 3B). Consistent with previous studies, E2F-I and
E2F-4 were unable to bind DNA in the absence of a co-transfected DP protein (Fig. 3B, lanes 3
113
Figure 3.
*A44
44
Lot..............
+ -,,+,
* 4..............
......................................... 4.
inu
ml4
. 8t
Bx
44
4
44wE
oo*,
rn
4*~4w
aI2P4
r,7!-
Figure 3. (preceding page) Dimerization and DNA binding properties of E2F-6. (A) C33-A
cells were transiently transfected with expression vectors encoding E2F-1, E2F-4, or HA-E2F-6 in
the presence or absence of DP (DP-1 or DP-2) or pocket proteins (pRB, p107) and
immunoprecipitated with the indicated antibodies. (B) C33-A cells were transiently transfected
with the identical combinations of expression vectors as in A. Gel shift assays were carried out on
the indicated unlabelled cell extracts (1.5 gg of total protein per lane) by using the consensus E2F
site from the adenoviral E2 promoter (TTTCGCGCCCTTT). Western blot assays were carried out
on the same unlabelled cell extracts (300 ng of total protein per lane) by using monoclonal
antibodies against E2F-1 (KH95), E2F-4 (LLF4-1), or the HA tag (12CA5). Gel retardation
assays have also been conducted by using the additional E2F sites TTTCCCGCCTTT,
TTTCCCGCCAAA, TTTCCCGCGTGT, or ATTCCCGCGCTTT with similar differences in
affinity.
1 15
and 8). In a similar manner, we were unable to detect any increase in the levels of E2F DNA
binding activity in cells that were transfected with HA-E2F-6 alone (Fig. 3B, lane 13). In the same
assay, E2F-1, -4, and -6 were able to bind to DNA when associated with either DP-1 (Fig. 3B,
lanes 4, 9, and 14) or DP-2 (Fig. 3B, lanes 7, 12, and 17). However, we consistently detected
less E2F DNA binding activity in extracts derived from cells transfected with HA-E2F-6 rather
than E2F-1 or E2F-4 (Fig. 3B, compare lanes 14-17 with lanes 4-7 and 9-12). Given this finding,
we also assessed the expression levels of these E2Fs in Western blots (Fig. 3B). Because the
blots were probed with different antibodies, we cannot make definitive conclusions about the
relative levels of these proteins. However, these antibodies have similar avidities for their
respective antigens (unpublished observations), suggesting that E2F-1 and E2F-4 are not
expressed at significantly higher levels than HA-E2F-6. This suggests that HA-E2F-6 containing
complexes have a lower affinity for the consensus E2F site relative to complexes containing either
E2F-1 or -4. This low DNA-binding activity was observed with multiple independent
transfections and was not improved by removal of the HA tag (data not shown). Given these
observations, we also tested the E2Fs for their ability to bind to other E2F binding sites (data not
shown). The DNA-binding activity of E2F-6-DP did increase when we used the probes
TTTCCCGCC(A/T)(A/T)(A/T).
However, this site was previously identified by site selection
assays as the preferred recognition sequence of complexes containing E2F-1, -2, -3, or -4, and
these species also have a higher affinity for this sequence than for any other E2F site (see
CHAPTER II). In fact, relative to the other E2F species, E2F-6-DP complexes bound less well to
each of the five probes tested. We therefore conclude that E2F-6 has a lower affinity for DNA, at
least when associated with either of the known DP proteins.
C4.
E2F-6 is unable to activate the transcription of E2F-responsive genes.
E2F-6 lacks the C-terminal sequences that are known to mediate the transcriptional activity
of E2F-1 through -5. However, its shortened C-terminal domain is highly charged and contains a
significant proportion of acidic residues (22%), suggesting that it might activate transcription via
116
this alternative motif. To test this hypothesis, we transiently transfected C33-A cells with a
chimeric reporter construct, E2F 4 -CAT, in which the expression of the CAT gene is controlled by
a minimal promoter containing four consensus E2F sites upstream of the IB TATA box (Helin et
al., 1992) and increasing amounts of the eukaryotic expression vectors encoding E2F- 1, E2F-4, or
HA-E2F-6 (Fig. 4A). These transfections were conducted in either the absence or presence of
CMV-DP-2 as indicated. E2F-1 and E2F-4 both substantially increased the activation of this
reporter. In contrast, E2F-6 did not bring about any increase in the level of E2F transcriptional
activity, either in the absence or presence of cotransfected DP proteins (Fig. 4A). Instead, the
increasing input levels of E2F-6 steadily inhibited the ability of the endogenous E2F complexes to
activate this reporter (Fig. 4B). We did not see any inhibition of reporter activity when the E2F
sites were deleted, indicating that this repression was specific for E2F-responsive reporters (data
not shown). This inhibition was observed in either the absence or presence of an excess of
exogenously expressed DP protein (Fig. 4B and data not shown).
After finding that E2F-6 can block the activity of the endogenous E2F complexes, we
wanted to establish whether it could also inhibit a cotransfected E2F. For this experiment, C33-A
cells were transiently transfected with the E2F 4-CAT reporter in either the absence or presence of
CMV-E2F-1 (50 ng) and increasing amounts of CMV-E2F-6 (Fig. 4C). At this input level, E2F-1
increased the activation of this reporter by approximately six-fold. This was effectively inhibited
by increasing input levels of CMV-E2F-6. Exactly as described above, this repression was
observed in either the absence or presence of cotransfected DP (data not shown). We therefore
conclude that E2F-6 is unable to activate transcription and, at least when overexpressed, can inhibit
the transcriptional activity of the other E2F species.
117
Figure 4.
A
12000-
30002500-
10000-
- CMV-DP
2000
-
1500
-
+ CMV-DP-2
(3 pg)
8000-
U
C~)
6000-
H
1000
-
4000-
500
-
2000 -
0
oil
-- l
I
0
3 0 ,0. 10.3 1 3 1 0 ,0.1 0.3 1 3 ,1
(pg)
CMV-F4 (jg) CMV-F6 (pg)
CMV-Fl
010.10.3 1
U uu.1u..,
U.1 0.3
- CMV-DP
U
250-
U
C.)
H
100
50-011
CMV-F6 (ng)
1
3
U
.
-
+
0
0 1
+
+
+
+
+
+
+
3 10 30 100 3001000
--
0 0.10.3 1 3
CMV-F4 (jg) CMV-F6 (jig)
2500-
-c
1500-
150-
0-
IU.IU.
2000-
200-
Reporter
- , U
.g
3000-
350 300-
1
CMV-F1 (pig)
C
B
----
---
1000500
0
Reporter
+
CMV-Fl (50 ng) CMV-F6 (pg)
-
+ + +
+ + +
- .05 .5
+ +
+ +
5 25
MV-DP
Figure 4. (preceding page) Effects of E2F-6 on transactivation of an E2F 4 -CAT reporter C33-A
cells were transiently transfected in duplicate with 4 [tg of E2F 4 -CAT, 2 gg of Rous sarcoma
virus-luciferase (as an internal control), 14 gg of carrier DNA (pBIIKS+ and pCMV-neo-Bam) in
the absence or presence of 3 jg of pCMV-HA-DP2 plus pCMV-E2F expression vectors as
indicated. CAT and luciferase activity were determined 24 h after transfection. The values shown
are the average of duplicate transfectants for representative experiments.
119
D. Discussion
E2F plays a pivotal role in the regulation of cellular proliferation by controlling the
expression of genes that are essential for either entry into, or passage through, the cell cycle. This
activity is extremely complex, arising from the combined action of multiple E2F-DP heterodimers.
Although there are differences in their cell cycle regulation, the previously identified E2F-DP
species can all display the same transcriptional properties. In the presence of an associated pocket
protein, they each behave as transcriptional repressors. Once released from the pocket proteins,
these same E2F-DP species can activate transcription. Although the DP moiety is required for
activity, the transcriptional activation and pocket protein-binding properties of the E2F-DP
complexes are entirely dependent upon sequences within the E2F protein.
Herein we describe the cloning and characterization of the human and murine orthologs of a
gene that contains the most conserved domains of the known E2F family members, including the
core sequences required for DNA binding and dimerization. Functional studies confirm that the
human protein can dimerize with either DP- 1 or DP-2 and the resultant complexes can bind to DNA
in a sequence-specific manner. These findings confirm that this protein is a genuine member of the
E2F family, and we have, therefore, named it E2F-6. Despite this designation, our data indicate
that this protein functions in a different manner than the other E2F family members. (i) E2F-6containing complexes bind to DNA significantly less well than the other E2F-DP species.
Although we have yet to calculate dissociation constants for these interactions, our preliminary
studies suggest that E2F-6-containing complexes recognize the same target sequences as the other
E2F species but they bind to these sites with at least a five-fold lower affinity. (ii) E2F-6 lacks the
domains that are known to mediate the transactivation and pocket protein-binding properties of the
other E2Fs and it is unable to perform either of these functions. (iii) Overexpressed E2F-6 is able
to block the transcriptional activity of either cotransfected E2F- 1 or the endogenous E2F
complexes. Although the mechanism of this repression is unclear, our data indicate that
transcriptional inhibition is not due solely to the sequestration of DP from the other E2F family
members by abnormally high levels of E2F-6.
120
Many other transcription factor families include one or more members that behave as
transcriptional inhibitors but there is considerable variation in the mechanism by which repression
is mediated. The Id proteins of the myogenic basic-helix-loop-helix family, the I-POU protein of
the POU domain family, and the CHOP protein of the c/EBP-like family lack the sequences
necessary for high-affinity DNA-binding and, therefore, function in a DNA-independent manner
(Benezra et al., 1990; Treacy et al., 1991; Ron and Habener, 1992). Instead, they are able to
inhibit transcription by binding to the essential heterodimeric partners of the positively-acting
family members and forming nonfunctional complexes. There are also several examples of
inhibitory proteins that repress transcription in a DNA-binding-dependent manner. In this case,
inhibition can be mediated by either exclusion of other family members from the DNA or through
the recruitment of cellular factors that actively inhibit transcription. One example of the latter case
is the Mad-Max complex that has recently been shown to mediate repression in a sequence-specific
manner through its ability to recruit histone deacetylase activity to its target genes (Pazin and
Kadonaga, 1997).
Because the E2F proteins function as part of a heterodimeric complex, it is possible to
envisage that the repressive effects of E2F-6 could be mediated by any of the repressive effects of
E2F-6 could be mediated by any of the mechanisms described above. Although we cannot rule out
any of these models, two observations suggest that E2F-6 is unlikely to act solely through the
sequestration of the DP proteins. (i) Unlike the other transcriptional repressors that act by this
mechanism, E2F-6 retains the ability to bind to DNA. (ii) At least as observed in our
overexpression assays, addition of excess DP protein does not prevent E2F-6 from inhibiting the
activity of the other E2F complexes. This strongly suggests that the observed repression is
dependent upon an additional function(s) of the E2F-6 protein. DP-binding might still be required
for this repression as an essential component of an actively repressive complex.
Clearly, the experiments described above do not allow us to address the mechanism of
action of the endogenous E2F-6 species. In fact, it is still unclear whether or not E2F-6 functions
as a repressor in vivo. It is possible to envisage alternate models for E2F-6 function. For
121
example, the DP proteins are known to be present in excess relative to the E2Fs in vivo and E2F-6
could exist to chaperone this pool of "unbound" DP. If correct, the "chaperone" model yields two
clear predictions about the endogenous E2F-6 protein. (i) It should exist at sufficiently high levels
in vivo to be able to bind to any DP protein that is not associated with E2F-1 through -5. (ii) E2F6 should have a lower affinity/ avidity for the DPs than any of the other E2F family members to
ensure the efficient transfer of the DP from E2F-6 to E2F-1 through -5. Clearly, the "repressor"
model of E2F-6 action also yields testable predictions. Regardless of whether or not its repressive
properties are dependent or independent of its ability to bind to DNA, E2F-6 should have a similar
or greater affinity for the DP proteins than the other E2Fs. Depending on its mechanism of action,
it could also be present at either greater (if it acts through the sequestration of DP) or similar/ lower
(if it regulates transcription in a sequence-specific manner) levels relative to the other E2Fs in vivo.
A direct comparison of the relative affinity/ avidity of the individual E2F family members for DP
should help us to distinguish between these "chaperone" and "repressor" models, as will
information about the relative abundance and DNA binding activity of the endogenous E2F-6
protein. Although we have yet to establish the true phyiological role of E2F-6, its broad
expression pattern suggests that it makes an important contribution to the regulation of E2F activity
in vivo.
122
E. Materials and Methods
cDNA identificationand characterization
GenBank, EMBL, and DDBJ databases were searched by using the TBLASTN program at
the National Center for Biotechnology Information with a consensus E2F DNA-binding domain
sequence derived from the primary amino acid sequence of the DNA-binding domains of E2F- 1,
-2, -3, -4, and -5. The identified E2F-6 human and mouse expressed sequence tags (ESTs) were
obtained from Research Genetics (Huntsville, AL). A human fetal brain cDNA library (Stratagene)
was screened with a 1.6 kbp EcoRI fragment of a human EST labelled with [X- 32 P]dCTP by
random priming. Hybridization was performed at 42'C in 5x SSC/ 5x Denhardt's solution/ 30%
formamide/ 0.5% SDS/ dextran sulfate (50 gg/ mL)/ salmon sperm DNA (150 gg/ mL). Filters
were washed at 55'C for three twenty minute periods in lx SSC/ 0.1% SDS. Positive clones were
identified by autoradiography. Exonuclease III (New England Biolabs) digestion was used to
generate nested deletions of both EST and cDNA clones, which were then sequenced with
Sequenase 2.0 (United States Biochemical).
Plasmidconstruction
The human E2F-6 ORF was amplified by PCR with the primers 6.6
(GTTAGGATCCATGCGGCACGAGAAGTTACCCAG) and 6.10
(CTCAGGATCCATCAGTTGCTTACTTCAAG) and Vent polymerase. The PCR product was
digested with BamHI and subcloned into pHACMV-neo-Bam to generate pCMV-HA-E2F-6. The
plasmids pE2F4 -CAT, pRSV-luciferase, pCMV-E2F-1, pCMV-E2F-4, pCMV-DP-1, pCMV-HADP-1, pCMV-HA-DP2, pCMV-RB, and pCMV-p 107 have been described (Helin et al., 1993; Wu
et al., 1995; Moberg et al., 1996). The 6x His-tagged E2F-6 vector was constructed by
amplifying the E2F-6 ORF with primers 6.6 and 6.12
(CACTAAGCTTATCAGTTGCTTACTTCAAGCA). The PCR product was digested with
BamHI and EcoRI and subcloned into pQE30 (Qiagen, Chatsworth, CA).
123
Northern blot analysis
The cell line Northern blot [containing poly(A)+ RNA from 293 (adenocarcinoma), HeLa
(cervical carcinoma), ML- 1 (myeloid leukemia), T98G (neuroblastoma), MCF7 (breast cancer), or
C33-A (cervical carcinoma)] or a human tissue blot [containing 2 gg (per lane) of poly(A)+ RNA
isolated from the indicated tissues (CLONTECH)] was screened with probes corresponding to the
full-length E2F-6 ORF or a 1,052-bp XmnI-SacI fragment from the 3' untranslated region. These
fragments were labelled with [k-
32 P]dCTP
by using the Prime-It II kit (Stratagene). The blots
were hybridized for 18 h at 65'C in 0.5 M sodium phosphate, pH 7.5/ 1 mM EDTA, pH 8.0/ 5%
SDS/ 1% BSA and washed three times with Ix SSC/ 0.1% SDS at 65'C before autoradiography.
Polyclonalantibody productionand Western blotting
The full-length 6x His-tagged human E2F-6 protein (amino acids 1-275) was expressed in
bacteria, purified over a Ni 2 +-nitrilotriacetic acid-agarose resin (Qiagen), and used to immunize
mice. Western blotting was performed as described by Moberg et al.
Transient transfection and in vitro assays
Cells were grown under standard conditions in DMEM supplemented with 10% fetal calf
serum. Transient transfections were performed as described (Moberg et al., 1996). For the
immunoprecipitation and gel retardation assays, transfections were conducted with 10 ptg of each
of the indicated plasmids plus pCMV-neo-Bam to give a total of 30 pg. Gel shift assays were
carried out as described (Moberg et al., 1996) with unlabelled cell extracts normalized for total
protein concentration. For immunoprecipitations, the cells were labelled with 250 gCi of [35S]_
methionine Express labelling mix (NEN) in methionine-free medium (GIBCO/ BRL) for 3.5 h.
Immunoprecipitations were performed as described (Lees et al., 1993) with the following
antibodies: 12CA5 [anti-haemagluttinin (HA) tag], KH20 [anti-E2F-1 (Helin et al., 1992)], LLF41 [anti-E2F-4 (Moberg et al., 1996)], sc-610x [anti-DP-1 (Santa Cruz Biotechnology)], sc-829x
124
[anti-DP-2 (Santa Cruz Biotechnology)]. Precipitates were resolved on 10% SDS polyacrylamide
gels and detected by fluorography. For transactivation assays, C33-A cells were transfected in
duplicate with 4 gg of E2F 4 -CAT, 2 jig of RSV-luciferase (as an internal control for transfection
efficiency), 14 gg of carrier DNA (pBIIKS+), and the indicated amounts of the pCMV-E2F
expression vectors. These transfections were performed in the presence or absence of 3 jg of
pCMV-HA-DP2. Within each experiment, the total concentration of CMV expression vector was
kept constant by the addition of pCMV-neo-Bam. Chloramphenicol acetyltransferase (CAT) and
luciferase assays were conducted as described in Lees et al (Lees et al., 1993).
125
F.
Acknowledgements
We are grateful to members of the Lees laboratory for helpful discussion throughout the
course of this study and to Patrick Humbert and Steve Bell for critical reading of the manuscript.
This work was supported by a grant from the National Institutes of Health. J. A. L. is a recipient
of the Professor Irwin Sizer and Helen Sizer Career Development Professorship.
126
G.
References
Bandara, L. R., Buck, V. M., Zamanian, M., Johnston, L. H., and La Thangue, N. B. (1993).
Functional synergy between DP- 1 and E2F- 1 in the cell cycle-regulating transcription factor
DRTF1/E2F. Embo J 12, 4317-24.
Bartek, J., Bartkova, J., and Lukas, J. (1996). The retinoblastoma protein pathway and the
restriction point. Cuff Opin Cell Biol 8, 805-14.
Beijersbergen, R. L., and Bernards, R. (1996). Cell cycle regulation by the retinoblastoma family
of growth inhibitory proteins. Biochim Biophys Acta 1287, 103-20.
Beijersbergen, R. L., Kerkhoven, R. M., Zhu, L., Carlee, L., Voorhoeve, P. M., and Bernards,
R. (1994). E2F-4, a new member of the E2F gene family, has oncogenic activity and associates
with p107 in vivo. Genes Dev 8, 2680-90.
Benezra, R., Davis, R. L., Lockshon, D., Turner, D. L., and Weintraub, H. (1990). The protein
Id: a negative regulator of helix-loop-helix DNA binding proteins. Cell 61, 49-59.
Bremner, R., Cohen, B. L., Sopta, M., Hamel, P. A., Ingles, C. J., Gallie, B. L., and Phillips,
R. A. (1995). Direct transcriptional repression by pRB and its reversal by specific cyclins. Mol
Cell Biol 15, 3256-65.
Clarke, A. R., Maandag, E. R., van Roon, M., van der Lugt, N. M., van der Valk, M., Hooper,
M. L., Berns, A., and te Riele, H. (1992). Requirement for a functional Rb-I gene in murine
development [see comments]. Nature 359, 328-30.
Cobrinik, D., Lee, M. H., Hannon, G., Mulligan, G., Bronson, R. T., Dyson, N., Harlow, E.,
Beach, D., Weinberg, R. A., and Jacks, T. (1996). Shared role of the pRB-related p130 and
p107 proteins in limb development. Genes Dev 10, 1633-44.
Dyson, N., Dembski, M., Fattaey, A., Ngwu, C., Ewen, M., and Helin, K. (1993). Analysis of
p107-associated proteins: p107 associates with a form of E2F that differs from pRB-associated
E2F-1. J Virol 67, 7641-7.
Dyson, N., and Harlow, E. (1992). Adenovirus ElA targets key regulators of cell proliferation.
Cancer Surv 12, 161-95.
Ginsberg, D., Vairo, G., Chittenden, T., Xiao, Z. X., Xu, G., Wydner, K. L., DeCaprio, J. A.,
Lawrence, J. B., and Livingston, D. M. (1994). E2F-4, a new member of the E2F transcription
factor family, interacts with p107. Genes Dev 8, 2665-79.
Helin, K., Lees, J. A., Vidal, M., Dyson, N., Harlow, E., and Fattaey, A. (1992). A cDNA
encoding a pRB-binding protein with properties of the transcription factor E2F. Cell 70, 337-50.
Helin, K., Wu, C. L., Fattaey, A. R., Lees, J. A., Dynlacht, B. D., Ngwu, C., and Harlow, E.
(1993). Heterodimerization of the transcription factors E2F- 1 and DP- 1 leads to cooperative transactivation. Genes Dev 7, 1850-61.
Hiebert, S. W., Chellappan, S. P., Horowitz, J. M., and Nevins, J. R. (1992). The interaction
of RB with E2F coincides with an inhibition of the transcriptional activity of E2F. Genes Dev 6,
177-85.
127
Hijmans, E. M., Voorhoeve, P. M., Beijersbergen, R. L., van 't Veer, L. J., and Bernards, R.
(1995). E2F-5, a new E2F family member that interacts with p130 in vivo. Mol Cell Biol 15,
3082-9.
Ivey-Hoyle, M., Conroy, R., Huber, H. E., Goodhart, P. J., Oliff, A., and Heimbrook, D. C.
(1993). Cloning and characterization of E2F-2, a novel protein with the biochemical properties of
transcription factor E2F. Mol Cell Biol 13, 7802-12.
Jacks, T., Fazeli, A., Schmitt, E. M., Bronson, R. T., Goodell, M. A., and Weinberg, R. A.
(1992). Effects of an Rb mutation in the mouse [see comments]. Nature 359, 295-300.
Kaelin, W. G., Jr., Krek, W., Sellers, W. R., DeCaprio, J. A., Ajchenbaum, F., Fuchs, C. S.,
Chittenden, T., Li, Y., Farnham, P. J., Blanar, M. A., and et al. (1992). Expression cloning of a
cDNA encoding a retinoblastoma-binding protein with E2F-like properties. Cell 70, 351-64.
Krek, W., Livingston, D. M., and Shirodkar, S. (1993). Binding to DNA and the retinoblastoma
gene product promoted by complex formation of different E2F family members. Science 262,
1557-60.
Lee, E. Y., Chang, C. Y., Hu, N., Wang, Y. C., Lai, C. C., Herrup, K., Lee, W. H., and
Bradley, A. (1992). Mice deficient for Rb are nonviable and show defects in neurogenesis and
haematopoiesis [see comments]. Nature 359, 288-94.
Lee, M. H., Williams, B. 0., Mulligan, G., Mukai, S., Bronson, R. T., Dyson, N., Harlow, E.,
and Jacks, T. (1996). Targeted disruption of p107: functional overlap between p107 and Rb.
Genes Dev 10, 1621-32.
Lees, J. A., Saito, M., Vidal, M., Valentine, M., Look, T., Harlow, E., Dyson, N., and Helin,
K. (1993). The retinoblastoma protein binds to a family of E2F transcription factors. Mol Cell
Biol 13, 7813-25.
Moberg, K., Starz, M. A., and Lees, J. A. (1996). E2F-4 switches from p130 to p107 and pRB
in response to cell cycle reentry. Mol Cell Biol 16, 1436-49.
Nevins, J. R. (1992). E2F: a link between the Rb tumor suppressor protein and viral
oncoproteins. Science 258, 424-9.
Pazin, M. J., and Kadonaga, J. T. (1997). What's up and down with histone deacetylation and
transcription? Cell 89, 325-8.
Pollock, P. M., Pearson, J. V., and Hayward, N. K. (1996). Compilation of somatic mutations
of the CDKN2 gene in human cancers: non-random distribution of base substitutions. Genes
Chromosomes Cancer 15, 77-88.
Ron, D., and Habener, J. F. (1992). CHOP, a novel developmentally regulated nuclear protein
that dimerizes with transcription factors C/EBP and LAP and functions as a dominant- negative
inhibitor of gene transcription. Genes Dev 6, 439-53.
Starostik, P., Chow, K. N., and Dean, D. C. (1996). Transcriptional repression and growth
suppression by the p107 pocket protein. Mol Cell Biol 16, 3606-14.
Treacy, M. N., He, X., and Rosenfeld, M. G. (1991). I-POU: a POU-domain protein that
inhibits neuron-specific gene activation [see comments]. Nature 350, 577-84.
128
Weinberg, R. A. (1992). The retinoblastoma gene and gene product. Cancer Surv 12, 43-57.
Weinberg, R. A. (1996). E2F and cell proliferation: a world turned upside down. Cell 85, 4579.
Weintraub, S. J., Chow, K. N., Luo, R. X., Zhang, S. H., He, S., and Dean, D. C. (1995).
Mechanism of active transcriptional repression by the retinoblastoma protein. Nature 375, 812-5.
Weintraub, S. J., Prater, C. A., and Dean, D. C. (1992). Retinoblastoma protein switches the
E2F site from positive to negative element. Nature 358, 259-61.
Wu, C. L., Zukerberg, L. R., Ngwu, C., Harlow, E., and Lees, J. A. (1995). In vivo
association of E2F and DP family proteins. Mol Cell Biol 15, 2536-46.
Zhu, L., van den Heuvel, S., Helin, K., Fattaey, A., Ewen, M., Livingston, D., Dyson, N., and
Harlow, E. (1993). Inhibition of cell proliferation by p107, a relative of the retinoblastoma
protein. Genes Dev 7, 1111-25.
129
CHAPTER IV
FIP-1, a zinc-finger protein, interacts specifically
with E2F-6.
B Fairchild, JM Trimarchi, and JA Lees
(Author's contribution: Identification of FIP- 1
and mapping of the E2F-6 domain that binds FIP- 1.)
A.
Abstract
E2F transcription factors have been implicated in the control of GI/S-specific transcription
and in the regulation of cellular proliferation. We recently reported the discovery of a novel E2F
protein, E2F-6, that differs from other known E2F species in its lack of a transcriptional activation
domain; E2F-6 is also the only E2F protein not known to be regulated by the pRB-family of
transcriptional repressors. We now report the isolation of a novel, ubiquitously expressed protein
interacting specifically with E2F-6 in yeast two-hybrid and co-immunoprecipitation assays. This
novel protein, which we have named FIP- 1, contains a putative C2 -C2 zinc finger with homology
to the DNA-binding zinc fingers of certain nuclear pore proteins. FIP-1 bears strong homology
throughout its length to YAF-2, a YYI-binding protein of unknown function. FIP-1 binds
specifically to the dimerization domain of E2F-6, and can associate with E2F-6 but not E2F-1, -2,
-3, or -4 as determined by co-immunoprecipitation. Thus, FIP-1 may participate in E2F-6-specific
functions or in their regulation.
130
B.
Introduction
The E2F family of transcription factors bind DNA in a sequence-specific manner and
regulate the induction of expression of a variety of genes at the GUS transition. So far, eight
genes have been identified that may contribute to this DNA-binding activity (Helin et al., 1992;
Kaelin et al., 1992; Shan et al., 1992; Girling et al., 1993; Ivey-Hoyle et al., 1993; Lees et al.,
1993; Beijersbergen et al., 1994; Ginsberg et al., 1994; Buck et al., 1995; Hijmans et al., 1995;
Ormondroyd et al., 1995; Sardet et al., 1995; Wu et al., 1995a; Zhang and Chellappan, 1995;
Rogers et al., 1996; Morkel et al., 1997; Cartwright et al., 1998; Gaubatz et al., 1998; Trimarchi et
al., 1998). Their gene products can be divided into two related subgroups based on sequence
homology. One subgroup, referred to as the DP proteins (DP-I and -2) can heterodimerize with
the members of the other subgroup, called the E2F proteins (E2F-1, -2, -3, -4, -5, and -6); this
heterodimerization is required for efficient DNA binding by the complex (Bandara et al., 1993;
Helin et al., 1993; Krek et al., 1993; Wu et al., 1995a; Trimarchi et al., 1998).
Although E2F activity is required in a general way for the regulation of GI/ S-specific
transcription, the E2F proteins also have specific roles in the regulation of cellular proliferation and
apoptosis. E2F- 1, for example, is required for efficient apoptosis of mouse thymocytes (Field et
al., 1996). E2F-3 is rate-limiting for proliferation of mouse embryonic fibroblasts in culture (P.
Humbert, S. Dandapani, K. Rogers, and J. Lees, unpublished observations), and an antibody to
E2F-3 inhibits S phase induction in a rat embryo fibroblast cell line (Leone et al., 1998). Both
E2F-4 and E2F-5 have also been shown to have specific, required functions in mouse development
(Lindeman et al., 1998), and P. Humbert, K. Rogers, S. Dandapani, and J. Lees, unpublished
observations). In contrast, E2F-1, -2, -3, -4, and -5 can all promote the induction of S phase
when overexpressed (Johnson et al., 1993; Beijersbergen et al., 1994; Qin et al., 1994; Shan and
Lee, 1994; DeGregori et al., 1995; Schwarz et al., 1995; Lukas et al., 1996; Shan et al., 1996;
DeGregori et al., 1997; Dirks et al., 1998), although E2F-5 may require the coexpression of a DP
protein for this activity.
131
The E2F proteins are regulated at multiple levels, including transcription, localization, and
protein stability. The best characterized mechanism regulating E2F activity is the association of
E2F/ DP complexes with members of the pRB family. pRB is the product of the retinoblastomasusceptibility gene RB, which is mutated in 30% of human cancers (Weinberg, 1996). pRB and
its homologs p107 and p130 are transcriptional repressors, and their binding to the transcriptional
activation domain of E2F proteins not only inhibits activation but also represses basal transcription
at E2F-regulated promoters (Weintraub et al., 1992; Bandara et al., 1993; Zamanian and La
Thangue, 1993; Adnane et al., 1995; Bremner et al., 1995; Weintraub et al., 1995; Ferreira et al.,
1998; Pierce et al., 1998). pRB, p107, and p130 bind preferentially to different E2F complexes:
pRB can be found with E2F-1, -2, -3, or -4 (Moberg et al., 1996), while p107 associates only
with E2F-4 (Moberg et al., 1996) and p130 binds to either E2F-4 or E2F-5 (Hijmans et al., 1995;
Sardet et al., 1995). This association is believed to participate in the cell cycle-specific regulation
of E2F target genes. At the approach of the GI/S transition, pRB becomes hyperphosphorylated
by cyclin/ CDK complexes(DeCaprio et al., 1989; Hinds et al., 1992; Ewen et al., 1993; Kato et
al., 1993; Sherr, 1994; Lundberg and Weinberg, 1998), disrupting its association with E2F
(Chellappan et al., 1991; Knudsen and Wang, 1997). pRB/ E2F complexes are therefore believed
to repress E2F transcriptional targets in early GI, whereas "free" E2F complexes activate
transcription in late GI and S phase. Some E2F complexes can be inactivated by cyclin Amediated phosphorylation, which may contribute to the return of E2F activity to basal levels later in
the cell cycle (Dynlacht et al., 1994; Krek et al., 1994; Xu et al., 1994; Kitagawa et al., 1995;
Krek et al., 1995; Dynlacht et al., 1997).
The proper regulation of E2F activity is essential to the cell and the organism. Mice born
with only one wild-type copy of the RB gene uniformly develop tumors, and mice completely
deficient for pRB die in utero with defects in the regulation of S phase entry and apoptosis (Jacks
et al., 1992); the increased tumorigenesis and inappropriate S phase entry and apoptosis are largely
E2F dependent (Tsai et al., 1998; Yamasaki et al., 1998). Mice born without p107 and p130 die
neonatally (Cobrinik et al., 1996). Overexpression of a form of E2F-1 bearing a deletion of the
132
cyclin A-binding domain in E2F- 1 leads to a delay in S phase progression and an enhanced
induction of apoptosis (Krek et al., 1995). Thus, inappropriate E2F activity can have a variety of
deleterious effects. Furthermore, overexpression of dominant-negative forms of E2F-1, DP-1, or
DP-2 can interfere with S phase entry (Dobrowolski et al., 1994; Wu et al., 1996; Fan and
Bertino, 1997). Levels of E2F activity that are both sufficient and not excessive are therefore
maintained by a variety of cellular mechanisms.
It is therefore of interest that, although E2F-6 contains DNA-binding and DP-dimerization
domains with homology to the other E2F proteins, no transcriptional activation domain has been
identified in E2F-6, nor does it interact with the pRB-family members. In fact, when
overexpressed, E2F-6 represses the transcription of an artificial promoter bearing E2F sites
(Morkel et al., 1997; Cartwright et al., 1998; Gaubatz et al., 1998; Trimarchi et al., 1998). While
it has not been established that endogenous E2F-6 complexes repress E2F target genes, the lack of
a domain mediating transcriptional activation or association with pRB-family members clearly
suggests that, at a minimum, E2F-6 is likely to function in ways that are distinct from those of
other E2F family members. The deregulation of E2F activity is associated with a deregulation of
cellular proliferation and apoptosis; an understanding of the role(s) played by E2F-6 in these or
other processes requires a knowledge of the in vivo activity of E2F-6, of the mechanisms
regulating its activity and of the genes it regulates.
To date, analysis of E2F-6 DNA-binding activity and of the effects of E2F-6 on
transcription have relied exclusively on overexpression of the protein. Although overexpression
studies may suggest functions for a given gene, they are subject to a variety of limitations. One
such limitation lies in the assumption that all proteins required for the function of that gene have
been provided in the system. DP- 1, for example, activates an E2F-responsive promoter very
poorly when overexpressed by itself, but very efficiently when coexpressed with E2F- 1 (Bandara
et al., 1993; Helin et al., 1993; Krek et al., 1993). If E2F-6 complexes include another protein in
vivo that has not been provided in the overexpression assays, these assays may not accurately
recapitulate the normal functions of E2F-6. We have therefore executed a yeast two-hybrid screen
133
to isolate proteins that can associate with E2F-6. We have identified a novel zinc-finger protein
that we have named FIP-1 (E2F-6 Interacting Protein-1), and demonstrated that this protein
interacts specifially with E2F-6.
134
C.
Results
C1.
Isolation of FIP-1
By yeast two-hybrid, we screened an activated human T cell library for proteins that could
interact with an E2F-6 fusion protein. The E2F-6 fusion protein included most of the aminoterminus of this protein, as well as its putative DNA-binding and dimerization domains, fused in
frame to the GAL4 DNA-binding domain; the T cell library transformants are fusions with the
GAL4 transcriptional activation domain. We screeened approximately 4 x 105 transformants, and
identified four independent clones with the ability to interact specifically with the E2F-6 fusion
protein but not with pRB or DP-1 fusion proteins. The strength of the protein-protein interactions
in this two-hybrid system can be estimated by the relative abilities of the yeast clones to proliferate
in the presence of increasing levels of an inhibitory drug (3-aminotriazole). By this criterion, the
four E2F-6-interacting clones could be further subdivided into two clones that grew moderately
well, indicating a moderate interaction, and two clones that grew very well, indicating a strong
interaction. This distinction was also observed when the library plasmids were isolated from these
clones and retransformed into the host yeast strain expressing the E2F-6 fusion protein, indicating
that the differential growth properties of the yeast were indeed conferred by the library plasmid.
Both of the library plasmids conferring moderate growth in 3-aminotriazole encode
fragments of DP-2. Both DP-1 and DP-2 have been shown to interact when overexpressed with
E2F-6 as assayed by co-immunoprecipitation, and to cooperate with E2F-6 in a DNA-binding
assay. The identification of DP-2 in the two-hybrid was therefore not an unexpected result. In
contrast, no clones encoding DP-1 fragments were obtained. This suggests that the two-hybrid
screen was not saturating. Indeed, E2F-6 does interact efficiently with DP- 1 when the two are coexpressed in this two-hybrid system (data not shown), and DP-I is known to be expressed in
activated T cells (Moberg et al., 1996). Screening of additional transformants would therefore be
expected to identify some plasmids encoding fragments of DP- 1, and might reveal additional E2F6-interacting proteins as well.
135
The library plasmids conferring robust growth in combination with E2F-6 in the twohybrid encode fragments of the same novel protein, that we have named FIP- 1. The library
plasmid inserts include more than a kilobase of sequence, with a potential open reading frame of at
least 228 amino acids (Figure 1). One of the FIP-1 clones begins at the twelfth amino acid in this
potential open reading frame: (PKRQAKPAA ... ) . The second clone begins more than one
hundred nucleotides upstream. Two ATG codons in the correct frame, very close to each other,
are present in this upstream region; both have similarity to the Kozak consensus sequence for
initiating methionines , although it remains possible that we may not have isolated the entire open
reading frame. It is perhaps worth noting that the region upstream of the potential initiating
methionine(s) is extremely GC-rich, which would be consistent with proximity to a site of
transcriptional initiation.
Although both library plasmid inserts encoding FIP-1 share a common 3' end apparently
resulting from the priming of an oligo-dT primer to a complementary tract of adenosine residues in
the FIP-1 message, comparison of this sequence to known expressed sequence tags (ESTs) reveals
that the FIP- 1 message extends at least another kilobase in this direction (Figure 1). Multiple
sequences derived from this 3' region of the gene have been mapped to the short arm of
chromosome 3 (Unigene entry Hs.56936; http://www.ncbi.nlm.nih.gov/UniGene/Hs.Home.html)
. These sequences are believed to reside near chromosomal marker D3S 1261, which is located at
3pl3-pl4.1 (Hughes et al., 1994), and references therein). This region of chromosome 3 is the
target of frequent deletion in a variety of human carcinomas (Yokoyama et al., 1992; Buchhagen et
al., 1996; Corless et al., 1996; Hu et al., 1996; Iqbal et al., 1996; Leonard et al., 1996; Chu et al.,
1998; Halbert et al., 1998; Rader et al., 1998; Venugopalan et al., 1998).
The amino acid sequence of FIP-1 has multiple noteworthy features. FIP-1 has significant
homology throughout its length to YAF-2, a protein identified in a two-hybrid screen with the
transcription factor YY1 (Figure 2) (Kalenik et al., 1997). YAF-2 can bind directly to the zinc
finger domain of YY1 in vitro and can counteract YY 1-mediated repression in overexpression
assays (Kalenik et al., 1997). The high degree of conservation between the amino-terminus of
136
Figure 1.
1
71
ACTCCCCGGACGGCGTTTCTCCTCCGAGCGGCGCCGGTTTCGGCTTGGGGGGGGCGGGGTACAGCCATCC
ATGACCATGGGCGACAAGAAGAGCCCGACCAGGCCAAAAAGACAAGCGAAACCTGCCGCAGACGAAGGGT
M T M G D K K S P T R P K R Q A K P A A D E G F
141
TTTGGGATTGTAGCGTCTGCACCTTCAGAAACAGTGCTGAAGCCTTTAAATGCAGCATCTGCGATGTGAG
I
C
D
V
R
E A
F
K
C
S
T
F
R
N
S
A
W
D
C
S
V
C
211
GAAAGGCACCTCCACCAGAAAACCTCGGATCAATTCTCAGCTGGTGGCACAACAAGTGGCACAACAGTAT
P
R
I
N
S Q L V A Q Q V A Q Q Y
K
G
T
S
T
R
K
281
GCCACCCCACCACCCCCTAAAAAGGAGAAGAAGGAGAAAGTTGAAAAGCAGGACAAAGAGAAACCTGAGA
K
D _._.K _ E ....
_K
P.._ E _-_
V _._
E .._._K _-_-_Q _._
P
P
P
P
K... K _._
E _-_-K _ K .._.E _._._K
A
T
351
AAGACAAGGAAATTAGTCCTAGTGTTACCAAGAAAAATACCAACAAGAAAACCAAACCAAAGTCTGACAT
S P S V T K K N T N K K T K P K S D I
I
_K-- E
..... D ....
421
TCTGAAAGATCCTCCTAGTGAAGCAAACAGCATACAGTCTGCAAATGCTACAACAAAGACCAGCGAAACA
N
A
T
T
K
T
S
E T
D
P
P
S
E A
N
S
I
Q
S
A
L
K
491
AATCACACCTCAAGGCCCCGGCTGAAAAACGTGGACAGGAGCACTGCACAGCAGTTGGCAGTAACTGTGG
N
H
T
S R P R L K N V D R S T A Q Q L A V T V G
561
GCAACGTCACCGTCATTATCACAGACTTTAAGGAAAAGACTCGCTCCTCATCGACATCCTCATCCACAGT
N
631
V
T
I
I
V
T
D
F
K
E
K
T
R
S
S
S
T
S
S
S
T
V
GACCTCCAGTGCAGGGTCAGAACAGCAGAACCAGAGCAGCTCGGGGTCAGAGAGCACAGACAAGGGCTCC
T
S
S
A
G
S
E
Q
Q
N
Q
S
S
S
G
S
E
S
T
D
K
G
S
701
TCCCGTTCCTCCACGCCAAAGGGCGACATGTCAGCAGTCAATGATGAATCTTTCTGAAATTGCACATGGA
S R S S T P K G D M S A V N D E S F *
771
841
911
981
1051
1121
1191
1261
1331
1401
1471
1541
1611
1681
ATTGTGAAAACTATGAATCAGGGTATGAAATTCAAAACCTCCACCTGCCCATGCTGCTTGCATCCCTGGA
GAATCTTCTGTGGACATCGACCTCTTAGTGATGCTGCCAGGATAATTTCTGCTTGCCATGGGCATCTGGC
CACCAAGGAATTTCGCACCCTGACGATTACTCTTGACACTTTTATGTATTCCATTGTTTTATATGATTTT
CCTAACAATCATTTATAATTGGATGTGCTCCTGAATCTACTTTTTATAAAAAAAAAAAAAAtctgctgtg
cacaattttccatgtacattacaactggttttttgtttttgttttgttgggggagggaacttttatttat
tgtgttcacaaactccatcctttcagcatatccttttaagtttagttctttcttccagttatactatgta
ctatcagttttgatataactatatatatataaatataaaattatatataaagggttatttgaaaccaatc
catggcacgctggtgcttgatacactgtgaagtgaatacaacattgaacagttacagatctgggacagtc
ccttctatgaaagtgctgaaatttaattaaaatcagtcttacatgaagtatgttccaatccatgtgggaa
cttgactctctcatctgtctaaagagtactggacgatataaaaatatatattttttaaacaatgtgatct
caaatttaaagactgctccagatagcctgcatttgcaatggaataactgacaaatcacaagtggtttagt
tgggcagggctttgatcattcaaaagtaactaaagtagctccagaatgccaagtattcgtgtaaattacg
gttacatgttatcatttgctgttcttacataagcactcatgaaaatatggtattctgtaacttgaattcc
atccattttccagacgtctactcatgtctgaggtaaatctagaaattgtcttagttttaggattgaaaca
1751
gtctataaactgtatttttggtccatccaggaagctagtcccttgtttctcctttctacatgacattgca
1821
1891
1961
2031
gtggtggtttctgtaattaaaaatttgtttgcctcatgtccctttgtctgataaaccttcactctaccga
ttcagttgtgagtattctttttttccttctcaaaacctactatgatttgttttactgaacaaaggttatc
aaccacacatccagtcctgacatggagcttttcagtgtttggagacatttctcaatcccctgctgtggta
ggaactccagtggtgaacgcgttgcgcgcctgcagccagagttgcagggaaagctcgtacttactgcgag
2101
cagcatgtaatcttttttcttcctggacataaagatagcttgagtaaactgttctatttcattctcttca
2171
ctctttttactgtcttgcaaaaaaaa
Figure 1. (preceding page) Nucleotide sequence and predicted translation product of FIP-1.
Nucleotides derived from the plasmids isolated in the two-hybrid screen are capitalized, whereas
nucleotides that have only been inferred based on ESTs present in Genbank are in lower case. We
have not confirmed the accuracy of these sequences; however, the sequences have been reported in
at least four independent ESTs at each nucleotide position. The predicted translation product is
depicted underneath the corresponding mRNA codons; the potential initiating methionines are
depicted in bold-face. A solid line underlines the potential zinc finger, and a dotted line underlines
the lysine/ glutamic acid rich domain.
Figure 2. (following page) FIP- 1 shares extensive homology with the YY 1-associated factor
YAF-2. The amino acid sequences of FIP- 1 and YAF-2 were compared using the BESTFIT
program of the Genetics Computer Group software package. Identities are depicted by a vertical
line ( I), and gaps are depicted with periods (.). The conserved putative zinc finger is depicted in
bold-face.
138
Figure 2.
FIP-1
MTMGDKKSPTRPKRQAKPAADEGFWDCSVCTFRNSAEAFKCSICDVRKGTST
YAF-2
. . MGDKKSPTRPKRHAKPSSDEGYWDCSVCTFRNSAEAFKCMMCDVRKGTST
FIP-1
RKPRINSQLVAQQVAQQYATPPPPKKEKKEKVEKQDKEKPEKDKEISPSVTK
YAF-2
RKPRPVSQLVAQQVTQQFVPPTQSKKEKKDKVE ...KEKSEKE ..... .TTSK
FIP-1
KNTNKKTKPKSDILKDPPSEANSIQSANATTKTSETNHTSRPRLKNVDRSTA
YAF-2
KNSHKKT ................................. RPRLKNVDRSSA
FIP-1
QQLAVTVGNVTVIITDFKEKTRSSSTSSSTVTSSAGSEQQNQSSSGSESTDK
YAF-2
QHLEVTVGDLTVIITDFKEKTKSPPAS ..... .SAASADQHSQSGSSSDNTER
FIP-1
GSSRSSTPKGDMSAVNDESF
YAF-2
GMSRSSSPRGEASSLNG.. .
||
I| |lii
I I
111111
llIIIIIIIIIIII|
I Al li
I I
I
I
I
II|
| |
|
111111
YAF-2 and the proposed amino-terminus of FIP-I is consistent with the possibility that one of the
methionines highlighted in Figure 1 may indeed be the legitimate site of translational initiation.
FIP-I shares with YAF-2 a C-X 2 -C-X 10 -C-X 2 -C zinc finger motif in the amino-terminal region of
the proteins (Figures 1 and 2). The zinc fingers of FIP-1 and YAF-2 are most highly homologous
to nupl53 and nup358 (Sukegawa and Blobel, 1993; McMorrow et al., 1994; Wu et al., 1995b;
Yokoyama et al., 1995), both of which are believed to be associated with the nuclear pore. FIP-1
and YAF-2 also share a domain rich in lysine and glutamic acid. There are also differences
between FIP- 1 and YAF-2. FIP- 1 bears a thirty-three amino acid insertion in the middle of the
protein, absent in YAF-2. The carboxy-terminal region of FIP- 1 is less well conserved to YAF-2
than most of the rest of the protein is (<40% identity vs. -80%), and is rich in serine and threonine
residues and poor in hydrophobic residues (5 of the last 52 residues). Interestingly, there are
ESTs homologous to the FIP-I/ YAF-2 family from a variety of multicellular organisms including
fruit flies and nematodes; the homology is particularly striking in the amino terminus (40/60
residues identical from D. melanogasterFIP-1I/YAF-2 (GenBank entry AA26427 1, for example) to
human FIP-1). Unfortuneately, the functions of this gene in these organisms remain unknown.
C2.
Expression pattern of FIP-1
We have investigated the pattern of expression of the endogenous FIP- 1 gene by Northern
blot analysis. The levels of FIP- 1 mRNA expression vary among the human tissues assayed;
levels were highest in samples derived from placenta, lung, and kidney, and were present at more
moderate levels in muscle, liver, brain, pancreas, and heart (Figure 3). Despite the variations in
expression levels, a 5 kb FIP- 1 mRNA is expressed in all tissues examined; a second, slightly
larger message is also detectable in the placental sample. A similar 5 kb message is also expressed
in all cell lines we have examined to date (data not shown). E2F-6 mRNA has also been detected
in these tissues and cell lines, consistent with the possibility that the two proteins may interact in
vivo.
140
40
Figure 3. (preceding page) FIP- 1 is widely expressed in vivo. A PCR product corresponding to
the entire putative open reading frame was radiolabelled and used to probe a human tissue northern
blot (Clontech) as previously described. This probe does not cross to YAF-2 (data not shown); the
YAF-2 message is also considerably smaller than the FIP-1 message (-2 kb vs. ~ 5 kb).
142
The existence of a ubiquitous 5 kb message indicates that we have not yet isolated a fulllength FIP-1 cDNA. The 3 kb that we have not located could potentially result from additional
sequences upstream and/or downstream of the 2 kb we have already identified. Theoretically, the
additional sequences could also be interior, in whole or in part, to our current sequence as a result
of alternative splicing; however, the lack of smaller detectable FIP- I mRNA splice forms argues
against this possibility.
C3.
Identification of E2F-6 regions interacting with FIP-1
As a first step toward characterizing the interaction of FIP- 1 with E2F-6, we have identified
the domains of E2F-6 that mediate this interaction. We were interested, for example, to know if
FIP-1 binds to a region of E2F-6 that is largely conserved in the other E2F proteins, such as the
DNA binding domain, or to a more divergent region, such as the amino-terminus. We have
therefore expressed, in the two-hybrid system, various domains of E2F-6 to determine which are
necessary and which are sufficient for interaction with FIP- 1 (Figure 4). Deletions from the
carboxy-terminus of E2F-6 revealed that the domain believed to be required for interaction with the
DP proteins, the "dimerization domain", including the "leucine zipper" and "marked box" regions,
is also required for the interaction of E2F-6 with FIP- 1. Deletions from the other direction indicate
that neither the amino-terminal region nor the DNA-binding domain of E2F-6 is required for its
interaction with FIP-1; indeed, the dimerization domain of E2F-6 is both necessary and sufficient
for the interaction. This raises a number of specific questions. First, the similarity in the
requirements of DP proteins and FIP- 1 for binding to E2F-6 raises the possibility that their binding
to E2F-6 may be mutually exclusive; this possibility is being actively investigated. Second, the
dimerization domain of E2F-6 is somewhat conserved with other E2F proteins. We have therefore
attempted to determine if FIP- 1 can also interact with other members of the E2F family.
143
Figure 4.
FULL
LENGTH
E2F-6
DNA
BINDING
DOMAIN
DIMERIZATION
DOMAIN
RE
INTERACTION
WITH
FIP-1
E2F-6
CONSTRUCT
na...
..
.....
I
I
1
'NE
Figure 4. (preceding page) The dimerization domain of E2F-6 mediates its interaction with FIP1. Fusion proteins expressing the indicated regions of E2F-6 fused to the GAL4 DNA-binding
domain were expressed in yeast and their ability to interact with FIP-1 was assessed by yeast twohybrid. The strength of the observed interaction, as inferred by resistance to 3-aminotriazole, is
depicted in the right-hand column. "+++" indicates strong resistance to 3-aminotriazole, whereas
"-" indicates susceptibility to the drug.
145
C4.
Specific interaction of FIP-1 with E2F-6
We have subcloned the putative open reading frame of FIP-I into a eukaryotic expression
vector to assess its ability to interact with E2F proteins in mammalian cells; a haemagluttinin (HA)
epitope tag was included at the amino-terminus of FIP-1. This construct expresses efficiently in
transiently transfected C33-A cells, a human cervical carcinoma cell line, as determined by Western
blotting and probing with an anti-HA antibody (data not shown).
The ability of FIP-I to interact with E2F proteins was assayed by co-immunoprecipitation
assays from
35 S-labeled
C33-A cells (Figure 5). Using an anti-HA antibody, FIP-l is efficiently
immunoprecipitated from extracts transfected with the construct expressing this protein; similarly,
E2F-1, -2, -3, -4 , and -6 can be specifically immunoprecipitated with the appropriate antibodies
when overexpressed (Figure 5 and data not shown). E2F-6 can also be immunoprecipitated from
cells co-expressing E2F-6 and FIP- 1 and treated with an antibody to the HA tag on FIP- 1,
indicating that E2F-6 and FIP-I are interacting in this assay and that E2F-6 is coimmunoprecipitated by virtue of its association with FIP-I (Figure 5). Similarly, an antibody to
E2F-6 immunoprecipitates both FIP- 1 and E2F-6 when the two are coexpressed, although
inefficiently: whereas nearly comparable amounts of E2F-6 are immunoprecipitated through an
E2F-6 antibody or when co-immunoprecipitated with FIP- 1, FIP- 1 is clearly immunoprecipitated
more efficiently through an anti-HA antibody than it is through the anti-E2F-6 antibody in the
presence of co-expressed E2F-6. The reasons for this difference are currently under investigation.
Although FIP- 1 interacts with E2F-6 in this assay, no interaction between FIP- 1 and E2F1, E2F-2, or E2F-4 can be detected. In cells overexpressing both E2F- 1 and FIP-1, for example,
treatment with an antibody to E2F- 1 efficiently immunoprecipitated E2F- 1, but not FIP- 1, whereas
an anti-HA antibody immmunoprecipitated FIP- 1, but not E2F- 1. Similar results were observed
for E2F-2, -3, and -4 (Figure 5 and data not shown). Thus, the capacity to interact with FIP-I is
not a general property of E2F proteins, but is specific to E2F-6.
146
A
L
X
X
A
~X
A
%~
A
A
x S
x
C
%I-
I
If
VH
9-d33
VH
VH
VH
VH
VH
vH
I VH
0
I~
-V
0
-n
Figure 5. (preceding page) FIP- 1 associates specifically with E2F-6. Constructs expressing the
proteins indicated at the top of the figure were transfected into C33A cells, which were
subsequently labelled with
35 S-methionine,
lysed, and immunoprecipitated with an antibody to the
epitope indicated at the bottom of the figure. The resulting proteins were resolved by SDS-PAGE.
The lower arrow on each side of the figure indicates the position of FIP- 1 protein, whereas the
upper arrow on the left side indicates the approximate positions of E2F-1, -2, and -4, and the
upper arrow on the right indicates the position of E2F-6.
148
D.
Discussion
E2F activity participates in the regulation of cellular proliferation and apoptosis. The
ubiquitous expression of E2F-6 suggests that it may play an important role in cellular regulation.
Although E2F-6 has been shown, when overexpressed, to bind DP proteins and DNA and to
repress E2F-mediated transcription, efforts to confirm these activities for endogenous E2F-6
protein have proven difficult. For example, endogenous E2F-6/ DP proteins have not yet been
reported, nor has E2F-6 yet been detected as a component of endogenous E2F DNA binding
activity as assayed by standard E2F electrophoretic mobility shift assays (J. Trimarchi and J. Lees,
unpublished data), perhaps consistent with its relatively low affinity for E2F binding sites in these
assays when overexpressed.
The E2F literature is replete with examples of E2F-associated proteins that mediate or
regulate the activity of these complexes. These range from the DP proteins, required for high
affinity DNA binding and transcriptional activation by E2F-1 through -5, to the repressors pRB,
p107, and p130, to cyclin A, which phosphorylates some E2F complexes, inhibiting their
transcriptional activity. We have therefore executed an unbiased screen for E2F-6-associated
proteins in an attempt to clarify the functions of endogenous E2F-6 and the mechanisms of the
regulation of its activities. We have isolated a novel protein that, like E2F-6, is expressed in all
tissues and cell lines examined, and is therefore a plausible candidate for a regulator or mediator of
E2F-6 function. Our screen was not saturating; further screening has the potential to uncover
additional E2F-6-interacting proteins.
This novel protein, FIP-1, bears signifiant homology to the YY 1-associated protein YAF2. Beyond its ability to bind YY1, there are few data to indicate the in vivo function(s) of YAF-2.
Its message is upregulated upon differentiation of skeletal myotubes or cardiac myoctes (Kalenik et
al., 1997). When overexpressed, YAF-2 can activate the expression of the YY1-repressible
skeletal alpha-actin promoter (Kalenik et al., 1997). YAF-2 could therefore be hypothesized to
counteract the effects of YY 1 in some contexts. YAF-2 is not believed to encode a transcriptional
149
activation domain; rather, it has been suggested that YAF-2 may promote the specific degradation
of YY1 upon muscle cell differentiation, although this remains highly speculative.
The putative zinc finger in FIP- 1 (and YAF-2) is homologous to zinc fingers found in the
nuclear pore proteins nup153 and nup358. These proteins have at least four zinc fingers each, all
of the type W-x-C-x-x-C-x(3)-N-x(6)-C-x-x-C: the same pattern observed in the single zinc finger
in FIP-1 and YAF-2 as well as in mdm2 and other proteins. Although mdm2 is known to interact
with p53 through the amino-terminus of the former protein (Chen et al., 1993; Oliner et al., 1993)
and has been reported to interact with E2F-1/DP-1 through that same region (Martin et al., 1995),
no function has yet been ascribed to this mdm2 zinc finger, which is located near the middle of the
protein. Both nup153 and nup358 have been reported to bind DNA in a zinc-dependent manner
(Sukegawa and Blobel, 1993; Yokoyama et al., 1995); the zinc finger domain of nupl53 is
sufficient for this interaction. YAF-2 has been reported to have a moderate affinity for DNA
cellulose (Kalenik et al., 1997). Given the low affinity of E2F-6/DP complexes for DNA in
overexpression assays, it is tempting to speculate that FIP- 1 may promote the interaction of E2F-6
complexes with DNA. Nevertheless, no data yet support that conclusion, and the zinc finger of
FIP- 1 may instead play a role in protein-protein interactions with E2F-6 or other proteins, or
simply be required to maintain the overall structure of the protein. Importantly, although FIP-1
and YAF-2 share homology to nup153 and nup358 in their zinc finger, neither FIP-1 nor YAF-2
bears the FG repeats that are the hallmark of nuclear pore proteins.
Although the identification of FIP- 1 as a specific E2F-6 binding protein is intriguing, it has
not yet clarified in any way the in vivo functions of E2F-6. Many questions remain to be
addressed. With what proteins is endogenous E2F-6 associated, and is FIP- 1 among them? In
what proportion? If endogenous E2F-6 is associated with FIP-1, many more questions arise.
Does the presence of FIP- 1 in the complex preclude the presence of DP proteins, or does the
trimolecular E2F-6/ DP/ FIP-I complex exist? Does FIP-1 promote DNA-binding by the complex?
Does FIP-I recruit additional proteins? Does it affect the stability of the complex, or its
localization? Given that E2F-6 is the only known E2F that is not believed to be regulated by the
150
pRB family of proteins, it may have functions distinctly different from other E2F proteins, and will
likely be regulated by different mechanisms. The identification of E2F-6-interacting proteins such
as FIP- 1, coupled with studies of endogenous E2F-6, should help to elucidate the particular roles
of E2F-6 in the regulation of transcription and of cell biology.
151
E. Materials and Methods
Yeast two-hybrid
A fragment of E2F-6 whose amino-terminus was that reported in (ref) and carboxyterminus was the end of the marked box (... PIDVYLCEVE) was cloned as a SacI/ SalI fragment
into pc97 (Sardet et al., 1995) to generate an in-frame fusion with the DNA-binding domain of
GAL4. This plasmid, called pc97-F6, was transformed into S. cerevisiae strain MaV 103 (Sardet
et al., 1995) using standard protocols. After establishing that the fusion protein neither interacts
with the GAL4 activation domain nor transactivates on its own, this strain expressing pc97-F6 was
transformed with a library derived from activated human T cells. The transformations were plated
on SC -Leu -Trp plates. After two days, the transformations were replica-plated to SC -Leu -Trp
-His plates supplemented with 10 mM 3-aminotriazole, a drug that selects for protein-protein
interaction in this system (Sardet et al., 1995). After an additional two days, the transformations
were again replica plated to SC -Leu -Trp -His plates, this time supplemented with 30 mM 3aminotriazole. After two days, colonies were restreaked on SC -Leu -Trp -His +30 mM 3aminotriazole plates. The library plasmid was isolated from the yeast and transformed into bacteria
using standard protocols. Plasmids were re-transformed into MaV103 yeast expressing pc97-F6
to verify that 3-aminotriazole resistance was conferred by the library plasmid. Some sequencing
was done manually using standard protocols, the remainder was done with an automated
sequencer.
For mapping the FIP- 1 binding region in E2F-6, all fragments were cloned as SacI/ SalI
fragments into pc97 and transformed into MaV103 yeast expressing the library plasmids. The
amino acid boundaries of the fragments are as follows: amino-terminus, HEKLPS ...
NVQYSMR; DNA-binding domain, KALKVK ... KNHIRW; dimerization domain, IGSDLS..
. YLCEVE. Strength of interactions was inferred from degree of resistance to 3-aminotriazole in
the two-hybrid assay.
152
Northern blot analysis
Northerns were performed exactly as previously described (Trimarchi et al., 1998), except
that the probe was derived from a PCR product corresponding to the predicted open reading frame
of FIP-1 (MTMGDKK . . . VNDESF*). The blots were hybridized and washed exactly as
previously described (Trimarchi et al., 1998).
Transient transfectionsand immunoprecipitations
Calcium phosphate were performed as previously described (Trimarchi et al., 1998). The
eukaryotic expression vector for FIP- 1, pCMV-HA-FIP- 1, was generating by cloning a PCR
product containing the predicted open reading frame of FIP- 1 as a BamHI fragment into
pHACMV-neo-Bam. Similarly, pCMV-E2F-6 was generated by cloning the BamHI fragment
from pCMV-HA-E2F-6 into pCMV-neo-Bam. The plasmids pCMV-E2F-1, pCMV-E2F-2,
pCMV-E2F-3, pCMV-E2F-4, and pCMV-DPl have all been previously described (Helin et al.,
1993; Moberg et al., 1996).
Immunoprecipitations were performed as previously described with the following
antibodies: 12CA5 [anti-haemagglutinin (HA) epitope tag], KH20 [anti-E2F-1 (Helin et al.,
1992)], LLF2-5 [anti-E2F-2], LLF3-3 and LLF3-5 [anti-E2F-3], LLF4-1 [anti-E2F-4 (Moberg et
al., 1996)], and LLF6-1 [anti-E2F-6]. The monoclonal antibodies LLF2-5, LLF3-3, and LLF3-5
were generated exactly as previously described (Moberg et al., 1996). For LLF6- 1, mice were
immunized with His-tagged E2F-6 protein as previously described (Trimarchi et al., 1998);
splenocytes were isolated and monoclonal antibody-producing cell lines were generated as
described (Moberg et al., 1996).
153
F.
Acknowledgements
We are grateful to members of the Lees laboratory for helpful discussion throughout the
course of this study and to Patrick Humbert and Steve Bell for critical reading of the manuscript.
This work was supported by a grant from the National Institutes of Health. J. A. L. is a recipient
of the Professor Irwin Sizer and Helen Sizer Career Development Professorship.
154
G.
References
Adnane, J., Shao, Z., and Robbins, P. D. (1995). The retinoblastoma susceptibility gene product
represses transcription when directly bound to the promoter. J Biol Chem 270, 8837-43.
Bandara, L. R., Buck, V. M., Zamanian, M., Johnston, L. H., and La Thangue, N. B. (1993).
Functional synergy between DP-1 and E2F-1 in the cell cycle-regulating transcription factor
DRTF1/E2F. Embo J 12, 4317-24.
Beijersbergen, R. L., Kerkhoven, R. M., Zhu, L., Carlee, L., Voorhoeve, P. M., and Bernards,
R. (1994). E2F-4, a new member of the E2F gene family, has oncogenic activity and associates
with p107 in vivo. Genes Dev 8, 2680-90.
Bremner, R., Cohen, B. L., Sopta, M., Hamel, P. A., Ingles, C. J., Gallie, B. L., and Phillips,
R. A. (1995). Direct transcriptional repression by pRB and its reversal by specific cyclins. Mol
Cell Biol 15, 3256-65.
Buchhagen, D. L., Worsham, M. J., Dyke, D. L., and Carey, T. E. (1996). Two regions of
homozygosity on chromosome 3p in squamous cell carcinoma of the head and neck: comparison
with cytogenetic analysis. Head Neck 18, 529-37.
Buck, V., Allen, K. E., Sorensen, T., Bybee, A., Hijmans, E. M., Voorhoeve, P. M., Bemards,
R., and La Thangue, N. B. (1995). Molecular and functional characterisation of E2F-5, a new
member of the E2F family. Oncogene 11, 31-8.
Cartwright, P., Muller, H., Wagener, C., Holm, K., and Helin, K. (1998). E2F-6: a novel
member of the E2F family is an inhibitor of E2F- dependent transcription. Oncogene 17, 611-23.
Chellappan, S. P., Hiebert, S., Mudryj, M., Horowitz, J. M., and Nevins, J. R. (1991). The
E2F transcription factor is a cellular target for the RB protein. Cell 65, 1053-61.
Chen, J., Marechal, V., and Levine, A. J. (1993). Mapping of the p53 and mdm-2 interaction
domains. Mol Cell Biol 13, 4107-14.
Chu, T. Y., Shen, C. Y., Chiou, Y. S., Lu, J. J., Perng, C. L., Yu, M. S., and Liu, H. S.
(1998). HPV-associated cervical cancers show frequent allelic loss at 3p14 but no apparent
aberration of FHIT mRNA. Int J Cancer 75, 199-204.
Cobrinik, D., Lee, M. H., Hannon, G., Mulligan, G., Bronson, R. T., Dyson, N., Harlow, E.,
Beach, D., Weinberg, R. A., and Jacks, T. (1996). Shared role of the pRB-related p130 and
p107 proteins in limb development. Genes Dev 10, 1633-44.
Corless, C. L., Aburatani, H., Fletcher, J. A., Housman, D. E., Amin, M. B., and Weinberg, D.
S. (1996). Papillary renal cell carcinoma: quantitation of chromosomes 7 and 17 by FISH,
analysis of chromosome 3p for LOH, and DNA ploidy. Diagn Mol Pathol 5, 53-64.
DeCaprio, J. A., Ludlow, J. W., Lynch, D., Furukawa, Y., Griffin, J., Piwnica-Worms, H.,
Huang, C. M., and Livingston, D. M. (1989). The product of the retinoblastoma susceptibility
gene has properties of a cell cycle regulatory element. Cell 58, 1085-95.
DeGregori, J., Leone, G., Miron, A., Jakoi, L., and Nevins, J. R. (1997). Distinct roles for
E2F proteins in cell growth control and apoptosis. Proc Natl Acad Sci U S A 94, 7245-50.
155
DeGregori, J., Leone, G., Ohtani, K., Miron, A., and Nevins, J. R. (1995). E2F-1
accumulation bypasses a GI arrest resulting from the inhibition of GI cyclin-dependent kinase
activity. Genes Dev 9, 2873-87.
Dirks, P. B., Rutka, J. T., Hubbard, S. L., Mondal, S., and Hamel, P. A. (1998). The E2Ffamily proteins induce distinct cell cycle regulatory factors in p16-arrested, U343 astrocytoma cells
[In Process Citation]. Oncogene 17, 867-76.
Dobrowolski, S. F., Stacey, D. W., Harter, M. L., Stine, J. T., and Hiebert, S. W. (1994). An
E2F dominant negative mutant blocks ElA induced cell cycle progression. Oncogene 9, 260512.
Dynlacht, B. D., Flores, 0., Lees, J. A., and Harlow, E. (1994). Differential regulation of E2F
transactivation by cyclin/cdk2 complexes. Genes Dev 8, 1772-86.
Dynlacht, B. D., Moberg, K., Lees, J. A., Harlow, E., and Zhu, L. (1997). Specific regulation
of E2F family members by cyclin-dependent kinases. Mol Cell Biol 17, 3867-75.
Ewen, M. E., Sluss, H. K., Sherr, C. J., Matsushime, H., Kato, J., and Livingston, D. M.
(1993). Functional interactions of the retinoblastoma protein with mammalian D- type cyclins.
Cell 73, 487-97.
Fan, J., and Bertino, J. R. (1997). Functional roles of E2F in cell cycle regulation. Oncogene
14, 1191-200.
Ferreira, R., Magnaghi-Jaulin, L., Robin, P., Harel-Bellan, A., and Trouche, D. (1998). The
three members of the pocket proteins family share the ability to repress E2F activity through
recruitment of a histone deacetylase. Proc Natl Acad Sci U S A 95, 10493-8.
Field, S. J., Tsai, F. Y., Kuo, F., Zubiaga, A. M., Kaelin, W. G., Jr., Livingston, D. M.,
Orkin, S. H., and Greenberg, M. E. (1996). E2F-1 functions in mice to promote apoptosis and
suppress proliferation. Cell 85, 549-61.
Gaubatz, S., Wood, J. G., and Livingston, D. M. (1998). Unusual proliferation arrest and
transcriptional control properties of a newly discovered E2F family member, E2F-6. Proc Natl
Acad Sci U S A 95, 9190-5.
Ginsberg, D., Vairo, G., Chittenden, T., Xiao, Z. X., Xu, G., Wydner, K. L., DeCaprio, J. A.,
Lawrence, J. B., and Livingston, D. M. (1994). E2F-4, a new member of the E2F transcription
factor family, interacts with p107. Genes Dev 8, 2665-79.
Girling, R., Partridge, J. F., Bandara, L. R., Burden, N., Totty, N. F., Hsuan, J. J., and La
Thangue, N. B. (1993). A new component of the transcription factor DRTF1/E2F. Nature 365,
468.
Halbert, A. R., Harrison, W. R., Hicks, M. J., Davino, N., and Cooley, L. D. (1998).
Cytogenetic analysis of a scapular chondromyxoid fibroma. Cancer Genet Cytogenet 104, 52-6.
Helin, K., Lees, J. A., Vidal, M., Dyson, N., Harlow, E., and Fattaey, A. (1992). A cDNA
encoding a pRB-binding protein with properties of the transcription factor E2F. Cell 70, 337-50.
Helin, K., Wu, C. L., Fattaey, A. R., Lees, J. A., Dynlacht, B. D., Ngwu, C., and Harlow, E.
(1993). Heterodimerization of the transcription factors E2F- 1 and DP- 1 leads to cooperative transactivation. Genes Dev 7, 1850-61.
156
Hijmans, E. M., Voorhoeve, P. M., Beijersbergen, R. L., van 't Veer, L. J., and Bernards, R.
(1995). E2F-5, a new E2F family member that interacts with p130 in vivo. Mol Cell Biol 15,
3082-9.
Hinds, P. W., Mittnacht, S., Dulic, V., Arnold, A., Reed, S. I., and Weinberg, R. A. (1992).
Regulation of retinoblastoma protein functions by ectopic expression of human cyclins. Cell 70,
993-1006.
Hu, L. F., Eiriksdottir, G., Lebedeva, T., Kholodniouk, I., Alimov, A., Chen, F., Luo, Y.,
Zabarovsky, E. R., Ingvarsson, S., Klein, G., and Emberg, I. (1996). Loss of heterozygosity
on chromosome arm 3p in nasopharyngeal carcinoma. Genes Chromosomes Cancer 17, 118-26.
Hughes, A. E., Newton, V. E., Liu, X. Z., and Read, A. P. (1994). A gene for Waardenburg
syndrome type 2 maps close to the human homologue of the microphthalmia gene at chromosome
3pI2-p14.1. Nat Genet 7, 509-12.
Iqbal, M. A., Akhtar, M., and Ali, M. A. (1996). Cytogenetic findings in renal cell carcinoma.
Hum Pathol 27, 949-54.
Ivey-Hoyle, M., Conroy, R., Huber, H. E., Goodhart, P. J., Oliff, A., and Heimbrook, D. C.
(1993). Cloning and characterization of E2F-2, a novel protein with the biochemical properties of
transcription factor E2F. Mol Cell Biol 13, 7802-12.
Jacks, T., Fazeli, A., Schmitt, E. M., Bronson, R. T., Goodell, M. A., and Weinberg, R. A.
(1992). Effects of an Rb mutation in the mouse [see comments]. Nature 359, 295-300.
Johnson, D. G., Schwarz, J. K., Cress, W. D., and Nevins, J. R. (1993). Expression of
transcription factor E2F1 induces quiescent cells to enter S phase. Nature 365, 349-52.
Kaelin, W. G., Jr., Krek, W., Sellers, W. R., DeCaprio, J. A., Ajchenbaum, F., Fuchs, C. S.,
Chittenden, T., Li, Y., Farnham, P. J., Blanar, M. A., and et al. (1992). Expression cloning of a
cDNA encoding a retinoblastoma-binding protein with E2F-like properties. Cell 70, 351-64.
Kalenik, J. L., Chen, D., Bradley, M. E., Chen, S. J., and Lee, T. C. (1997). Yeast two-hybrid
cloning of a novel zinc finger protein that interacts with the multifunctional transcription factor
YY 1. Nucleic Acids Res 25, 843-9.
Kato, J., Matsushime, H., Hiebert, S. W., Ewen, M. E., and Sherr, C. J. (1993). Direct
binding of cyclin D to the retinoblastoma gene product (pRb) and pRb phosphorylation by the
cyclin D-dependent kinase CDK4. Genes Dev 7, 331-42.
Kitagawa, M., Higashi, H., Suzuki-Takahashi, I., Segawa, K., Hanks, S. K., Taya, Y.,
Nishimura, S., and Okuyama, A. (1995). Phosphorylation of E2F-1 by cyclin A-cdk2.
Oncogene 10, 229-36.
Knudsen, E. S., and Wang, J. Y. (1997). Dual mechanisms for the inhibition of E2F binding to
RB by cyclin- dependent kinase-mediated RB phosphorylation. Mol Cell Biol 17, 5771-83.
Krek, W., Ewen, M. E., Shirodkar, S., Arany, Z., Kaelin, W. G., Jr., and Livingston, D. M.
(1994). Negative regulation of the growth-promoting transcription factor E2F-1 by a stably bound
cyclin A-dependent protein kinase. Cell 78, 161-72.
157
Krek, W., Livingston, D. M., and Shirodkar, S. (1993). Binding to DNA and the retinoblastoma
gene product promoted by complex formation of different E2F family members. Science 262,
1557-60.
Krek, W., Xu, G., and Livingston, D. M. (1995). Cyclin A-kinase regulation of E2F-1 DNA
binding function underlies suppression of an S phase checkpoint. Cell 83, 1149-58.
Lees, J. A., Saito, M., Vidal, M., Valentine, M., Look, T., Harlow, E., Dyson, N., and Helin,
K. (1993). The retinoblastoma protein binds to a family of E2F transcription factors. Mol Cell
Biol 13, 7813-25.
Leonard, J. H., Williams, G., Walters, M. K., Nancarrow, D. J., and Rabbitts, P. H. (1996).
Deletion mapping of the short arm of chromosome 3 in Merkel cell carcinoma. Genes
Chromosomes Cancer 15, 102-7.
Leone, G., DeGregori, J., Yan, Z., Jakoi, L., Ishida, S., Williams, R. S., and Nevins, J. R.
(1998). E2F3 activity is regulated during the cell cycle and is required for the induction of S
phase. Genes Dev 12, 2120-30.
Lindeman, G. J., Dagnino, L., Gaubatz, S., Xu, Y., Bronson, R. T., Warren, H. B., and
Livingston, D. M. (1998). A specific, nonproliferative role for E2F-5 in choroid plexus function
revealed by gene targeting. Genes Dev 12, 1092-8.
Lukas, J., Petersen, B. 0., Holm, K., Bartek, J., and Helin, K. (1996). Deregulated expression
of E2F family members induces S-phase entry and overcomes p16INK4A-mediated growth
suppression. Mol Cell Biol 16, 1047-57.
Lundberg, A. S., and Weinberg, R. A. (1998). Functional inactivation of the retinoblastoma
protein requires sequential modification by at least two distinct cyclin-cdk complexes. Mol Cell
Biol 18, 753-61.
Martin, K., Trouche, D., Hagemeier, C., Sorensen, T. S., La Thangue, N. B., and Kouzarides,
T. (1995). Stimulation of E2Fl/DP1 transcriptional activity by MDM2 oncoprotein. Nature 375,
691-4.
McMorrow, I., Bastos, R., Horton, H., and Burke, B. (1994). Sequence analysis of a cDNA
encoding a human nuclear pore complex protein, hnup153. Biochim Biophys Acta 1217, 21923.
Moberg, K., Starz, M. A., and Lees, J. A. (1996). E2F-4 switches from p130 to p107 and pRB
in response to cell cycle reentry. Mol Cell Biol 16, 1436-49.
Morkel, M., Wenkel, J., Bannister, A. J., Kouzarides, T., and Hagemeier, C. (1997). An E2Flike repressor of transcription [letter]. Nature 390, 567-8.
Oliner, J. D., Pietenpol, J. A., Thiagalingam, S., Gyuris, J., Kinzler, K. W., and Vogelstein, B.
(1993). Oncoprotein MDM2 conceals the activation domain of tumour suppressor p53. Nature
362, 857-60.
Ormondroyd, E., de la Luna, S., and La Thangue, N. B. (1995). A new member of the DP
family, DP-3, with distinct protein products suggests a regulatory role for alternative splicing in the
cell cycle transcription factor DRTF1/E2F. Oncogene 11, 1437-46.
158
Pierce, A. M., Schneider-Broussard, R., Philhower, J. L., and Johnson, D. G. (1998).
Differential activities of E2F family members: unique functions in regulating transcription. Mol
Carcinog 22, 190-8.
Qin, X. Q., Livingston, D. M., Kaelin, W. G., Jr., and Adams, P. D. (1994). Deregulated
transcription factor E2F-1 expression leads to S-phase entry and p53-mediated apoptosis. Proc
Natl Acad Sci U S A 91, 10918-22.
Rader, J. S., Gerhard, D. S., O'Sullivan, M. J., Li, Y., Li, L., Liapis, H., and Huettner, P. C.
(1998). Cervical intraepithelial neoplasia III shows frequent allelic loss in 3p and 6p. Genes
Chromosomes Cancer 22, 57-65.
Rogers, K. T., Higgins, P. D., Milla, M. M., Phillips, R. S., and Horowitz, J. M. (1996). DP2, a heterodimeric partner of E2F: identification and characterization of DP-2 proteins expressed in
vivo. Proc Natl Acad Sci U S A 93, 7594-9.
Sardet, C., Vidal, M., Cobrinik, D., Geng, Y., Onufryk, C., Chen, A., and Weinberg, R. A.
(1995). E2F-4 and E2F-5, two members of the E2F family, are expressed in the early phases of
the cell cycle. Proc Natl Acad Sci U S A 92, 2403-7.
Schwarz, J. K., Bassing, C. H., Kovesdi, I., Datto, M. B., Blazing, M., George, S., Wang, X.
F., and Nevins, J. R. (1995). Expression of the E2F1 transcription factor overcomes type beta
transforming growth factor-mediated growth suppression. Proc Natl Acad Sci U S A 92, 483-7.
Shan, B., Farmer, A. A., and Lee, W. H. (1996). The molecular basis of E2F-1/DP-1-induced
S-phase entry and apoptosis. Cell Growth Differ 7, 689-97.
Shan, B., and Lee, W. H. (1994). Deregulated expression of E2F-I induces S-phase entry and
leads to apoptosis. Mol Cell Biol 14, 8166-73.
Shan, B., Zhu, X., Chen, P. L., Durfee, T., Yang, Y., Sharp, D., and Lee, W. H. (1992).
Molecular cloning of cellular genes encoding retinoblastoma-associated proteins: identification of a
gene with properties of the transcription factor E2F. Mol Cell Biol 12, 5620-31.
Sherr, C. J. (1994). GI phase progression: cycling on cue [see comments]. Cell 79, 551-5.
Sukegawa, J., and Blobel, G. (1993). A nuclear pore complex protein that contains zinc finger
motifs, binds DNA, and faces the nucleoplasm. Cell 72, 29-38.
Trimarchi, J. M., Fairchild, B., Verona, R., Moberg, K., Andon, N., and Lees, J. A. (1998).
E2F-6, a member of the E2F family that can behave as a transcriptional repressor. Proc Natl Acad
Sci USA 95, 2850-5.
Tsai, K. Y., Hu, Y., Macleod, K. F., Crowley, D., Yamasaki, L., and Jacks, T. (1998).
Mutation of E2f-1 suppresses apoptosis and inappropriate S phase entry and extends survival of
Rb-deficient mouse embryos [In Process Citation]. Mol Cell 2, 293-304.
Venugopalan, M., Wood, T. F., Wilczynski, S. P., Sen, S., Peters, J., Ma, G. C., Evans, G.
A., and Srivatsan, E. S. (1998). Loss of heterozygosity in squamous cell carcinomas of the head
and neck defines a tumor suppressor gene region on 1 1q13. Cancer Genet Cytogenet 104, 12432.
Weinberg, R. A. (1996). E2F and cell proliferation: a world turned upside down. Cell 85, 4579.
159
Weintraub, S. J., Chow, K. N., Luo, R. X., Zhang, S. H., He, S., and Dean, D. C. (1995).
Mechanism of active transcriptional repression by the retinoblastoma protein. Nature 375, 812-5.
Weintraub, S. J., Prater, C. A., and Dean, D. C. (1992). Retinoblastoma protein switches the
E2F site from positive to negative element. Nature 358, 259-61.
Wu, C. L., Classon, M., Dyson, N., and Harlow, E. (1996). Expression of dominant-negative
mutant DP-1 blocks cell cycle progression in G1. Mol Cell Biol 16, 3698-706.
Wu, C. L., Zukerberg, L. R., Ngwu, C., Harlow, E., and Lees, J. A. (1995a). In vivo
association of E2F and DP family proteins. Mol Cell Biol 15, 2536-46.
Wu, J., Matunis, M. J., Kraemer, D., Blobel, G., and Coutavas, E. (1995b). Nup358, a
cytoplasmically exposed nucleoporin with peptide repeats, Ran- GTP binding sites, zinc fingers, a
cyclophilin A homologous domain, and a leucine-rich region. J Biol Chem 270, 14209-13.
Xu, M., Sheppard, K. A., Peng, C. Y., Yee, A. S., and Piwnica-Worms, H. (1994). 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, 8420-31.
Yamasaki, L., Bronson, R., Williams, B. 0., Dyson, N. J., Harlow, E., and Jacks, T. (1998).
Loss of E2F-1 reduces tumorigenesis and extends the lifespan of Rbl(+/- )mice. Nat Genet 18,
360-4.
Yokoyama, N., Hayashi, N., Seki, T., Pante, N., Ohba, T., Nishii, K., Kuma, K., Hayashida,
T., Miyata, T., Aebi, U., and et al. (1995). A giant nucleopore protein that binds Ran/TC4.
Nature 376, 184-8.
Yokoyama, S., Yamakawa, K., Tsuchiya, E., Murata, M., Sakiyama, S., and Nakamura, Y.
(1992). Deletion mapping on the short arm of chromosome 3 in squamous cell carcinoma and
adenocarcinoma of the lung. Cancer Res 52, 873-7.
Zamanian, M., and La Thangue, N. B. (1993). Transcriptional repression by the Rb-related
protein p107. Mol Biol Cell 4, 389-96.
Zhang, Y., and Chellappan, S. P. (1995). Cloning and characterization of human DP2, a novel
dimerization partner of E2F. Oncogene 10, 2085-93.
160
CHAPTER V
Conclusion
A.
Cloning and characterization of novel E2F genes
A wealth of data indicate that the E2F transcription factors mediate the cell-cycle regulated
transcription of a number of cellular genes at the GUS transition, and that the proper regulation of
these factors is necessary for accurate and timely passage through this transition and the
subsequent S phase (reviewed in CHAPTER I). Consequences of their deregulation can also
include apoptosis, transformation, and tumorigenesis. Both overexpression studies and genetic
studies indicate that the six known E2F genes have some independent functions in addition to any
shared ones. This evidence that the E2F genes are non-equivalent has provided much of the
impetus for the research herein described. If the E2F genes have independent functions, only the
identification and characterization of each of the E2F genes can, in this case, provide an accurate
indication of the biological activities of E2F function in the cell. Similarly, the known E2F genes
have been reported to differ in their regulation at the levels of mRNA abundance, protein stability,
subcellular localization, or DNA-binding activity, again underlining the importance of studying
each of these genes individually.
Al.
What are the biological functions of E2F-6?
CHAPTERS III and IV detail the identification of E2F-6 and its early characterization.
Although genetic studies with E2F-6 have not yet been carried out, overexpression studies suggest
that E2F-6 may have biological functions that differ from those of the other E2F genes and
suggests even more strongly that E2F-6 is likely to differ from other E2F genes in how it is
regulated. For example, E2F-6 does not interact with pRB or p107 when overexpressed, and is
therefore presumably independent of their regulatory effects; E2F-6 is the only E2F protein yet
characterized for which this appears to be true. The lack of a transcriptional activation domain in
161
E2F-6 further suggests that the biological functions of this protein are likely to be different from
those of other E2F family members. If E2F-6 functions as a transcriptional repressor as the
overexpression data seem to suggest, then its activity is indeed different from that of the other E2F
genes, conferring transcriptional repression in a manner independent of the pRB family. This
repression could involve the recruitment of other transcriptional repressors and/or histone
deacetylases. On the other hand, if E2F-6 can activate transcription, this function must be
mediated by an accessory protein providing a transcriptional activation domain; the requirement for
such a protein might provide an additional level of regulation of E2F-6 function.
The regulation of E2F-6 functions may vary among tissues or cell types, or among stages
of differentiation or development. There are perhaps innumerable research avenues that could be
explored to address this possibility. One approach would be to look for any variations in E2F-6
activity among cell types, tissues, or developmental stages. Although it is possible to monitor the
effects of overexpressed E2F-6 on an E2F-dependent reporter construct in a variety of cell lines,
and perhaps even to do so in cell lines that can be induced to differentiate, characterization of the
endogenous E2F-6 activity is crucial. It is possible, for example, to study levels of E2F-6 protein
in developing embryos by immunohistochemistry. Potentially more interesting would be a study
of E2F-6 DNA binding activity, but conditions permitting the detection of endogenous E2F-6 DNA
binding activity have not yet been reported. Whether this may point to a difference in the DNA
binding specificity of endogenous E2F-6 proteins, a difference in required accessory proteins, or
simply a difference in the assay conditions required has not yet been determined. The ability to
identify and follow the DNA binding activity of the protein would provide a much more powerful
tool to analyze the biochemistry of E2F-6 function, and may facilitate the study of its
transcriptional activity as well as the study of its regulation in various cells and tissues.
Much as the other E2F complexes are known to oscillate between roles as transcriptional
repressors and transcriptional activators, the transcriptional effects of E2F-6 could vary in different
cellular contexts, depending perhaps on the presence or absence of particular accessory factors.
An example of the identification and initial characterization of an E2F-6-interacting protein, FIP-1,
162
is described in CHAPTER IV. E2F-6-interacting proteins could conceivably provide or recruit
domains permitting E2F-6 either to activate or to repress transcription. Their presence or absence
in a particular cell or tissue could therefore determine the transcriptional effects of E2F-6 in that
specific environment. Alternatively, they could enhance or modify the DNA binding activity of
E2F-6. They could also regulate E2F-6 functions in other ways, such as at the level of protein
stability, or of subcellular localization. Characterizing the effects of E2F-6-interacting proteins on
the functions of E2F-6 may prove indispensable to a complete understanding of its biological
activities.
Determinationof E2F-6functions: a genetic approach
Although biochemical studies and the characterization of E2F-6-interacting proteins will
likely provide useful clues to the in vivo functions of E2F-6, a genetic study of E2F-6 will almost
certainly be required to fully identify and characterize those functions. By selectively eliminating
E2F-6 function in the mouse, for example, the contributions of E2F-6 to the regulation of E2Fresponsive genes can be accurately assessed, and the requirements for E2F-6 in the overall biology
of the organism can be determined. This approach has proven extremely useful in the assignment
of biological functions to other E2F family members: the specific requirements for E2F-4 in
hematopoesis and for E2F-5 in choroid plexus function had not been indicated before they were
studied in mice lacking these genes, and the phenotypes associated with E2F- 1 mutation strongly
support the overexpression studies indicating a pro-apoptotic function for this gene (reviewed in
CHAPTER I).
There is, of course, the possibility that the biological functions of E2F-6 may be shared
with other proteins; functional redundancy or functional compensation might therefore mask
functions of E2F-6 that might otherwise be revealed in an E2F-6-deficient animal. For other
genes, this difficulty has been partially circumvented by also mutagenizing other genes already
suspected to cooperate in the same biological function; this has proven fruitful in the intercrossing
of mice deficient for individual pRB family members, for example (Cobrinik et al., 1996; Lee et
163
al., 1996; Robanus-Maandag et al., 1998). In the case of E2F-6, it is much less clear with which
other proteins it might be cooperating, principally because of the questions that remain regarding
the in vivo function(s) of the protein. Despite any concerns that the study of E2F-6-deficient mice
might not reveal every function of the gene, this approach almost certainly has the greatest chance
to identify its in vivo function(s). This approach has often met with success in the study of other
E2F family members and regulators; furthermore, no obvious alternative exists that can study
effects in the whole animal and does not rely on overexpression. A genetic analysis may reveal
particular tissues with a heightened requirement for E2F-6 function. Such information could serve
as a focus for further research into E2F-6 activities. This approach will also provide cells lacking
E2F-6 that could be used to analyze or screen for E2F-6 transcriptional targets, or used to
determine if endogenous E2F-6 activity affects progression through the cell cycle.
A2.
Additional E2F genes must exist
The total E2F DNA-binding activity in the cell is more than that which can be accounted for
by the combined activities of E2F- 1, -2, -3, -4, and -5. This is true both for the overall levels of
E2F DNA-binding activity and for that subset that is associated with pRB. Although, when
overexpressed with a DP subunit, E2F-6 can bind to an E2F DNA-binding site, endogenous E2F6 does not appear to make a substantial contribution to E2F DNA-binding activity as determined by
this standard gel shift assay. It cannot necessarily be inferred from this that E2F-6 is not, in fact,
associated with E2F DNA sequences in vivo, merely that it does not appear to do so under the
conditions of this assay. It can, however, be inferred that that one or more other E2F species must
be responsible for the remaining E2F DNA-binding activity. Furthermore, since E2F-6 does not
associate with pRB even when overexpressed, E2F-6 presumably does not account for the
remaining pRB-associated E2F activity. Thus, there must exist at least one additional E2F species,
and of these, at least one must be pRB-associated. The same arguments that previously impelled a
search for E2F-6 therefore currently impel a search for the remaining E2F family members. The
identification of E2F-6 has provided an additional piece of the E2F puzzle, and its obvious
164
differences from other E2F genes underscores the likelihood that its contributions to E2F function
are considerably different from those of other family members. Nevertheless, the final piece or
pieces of the puzzle have yet to be uncovered.
165
B.
Toward an understanding of the mechanisms of E2F target specificity
As there is a need to identify and characterize E2F genes, there is a similar need to identify
and characterize the targets of E2F cellular function. Because E2F complexes have been shown to
have non-equivalent biological functions, much effort has recently been invested in determining
how these different functions can be accomplished by a family of proteins with similar DNAbinding properties and (except E2F-6) similar abilities to transactivate or repress target genes in a
manner dependent on the presence or absence of pRB family members. The differences in the
biological functions of the pRB-associated E2F family members and their differential ability to
affect individual E2F-responsive genes when overexpressed could, in principle, be mediated by a
number of mechanisms. One possibility, that individual E2F complexes might vary in their DNAbinding specificities, was explored in CHAPTER II. The sequences of high-affinity DNA-binding
sites for heterodimers of DP-1 with E2F-1, -2, -3, or -4 appear indistinguishable: they do not seem
to determine which E2F complexes can bind to or regulate transcription through a particular DNA
site. Another possibility is that specific E2F proteins may vary in their ability to interact with other
promoter elements, leading to the specific stabililization of particular E2F complexes at individual
promoters, depending on the other factors present at those promoters. E2F- 1, -2, and -3, for
example, have been proposed to interact with Sp-1 in vitro, whereas E2F-4 and -5 do not.
E2F
complexes obviously vary in their biological functions and in the transcriptional targets they can
regulate, but the mechanism(s) of this specificity remain largely unresolved.
Chimeric E2Fproteins as tools to decipher targetspecificity
Because specificity can be observed in overexpression assays, such as the specificity of
E2F-I-induced apoptosis, or the specificity of E2F target gene activation, it should be possible to
use these assays to identify the protein domain(s) conferring this specificity. Chimeric proteins
have already been generated combining various regions of E2F- 1 and E2F-4 or E2F-2 and E2F-4,
and these hybrid proteins are competent to bind to DNA and activate transcription (Muller et al.,
1997; Verona et al., 1997). Chimeric proteins combining regions of E2F-l with regions of
166
another E2F, such as E2F-3 or E2F-4, could presumably be used in overexpression assays to
determine which region of E2F- 1 mediates its proapoptotic function. If, for example, a version of
E2F-3 whose DNA-binding domain has been replaced with that of E2F- 1 can drive apoptosis
when overexpressed, then the manner in which E2F- 1 associates with DNA could be important for
this effect, or perhaps E2F- 1 associates through this domain with an additional protein affecting
promoter selection or transcriptional regulation. Regardless of which domain proves to mediate
this function, increasingly smaller regions of E2F- 1 and E2F-3 could be interchanged until the key
residues for the promotion of apoptosis are identified, which would hopefully lead to the rapid
identification of the relevant biochemical activity (protein-protein binding, protein-DNA binding,
etc.).
Much as overexpression assays could be used to determine the function(s) of E2F- 1
involved in the promotion of apoptosis, they could also be used to identify elements conferring
specificity of target gene activation. For example, E2F- 1 has been reported to be more efficient
than E2F-3 in the induction of ribonucleotide reductase 2 (RR2) mRNA levels, whereas E2F-3 is
the more efficient at inducing CDK2 (DeGregori et al., 1997). Overexpression of chimeric E2F-1I/
E2F-3 proteins should identify the residues of these proteins conferring this specificity. Similarly,
the CDK2 promoter could be altered to more closely resemble the RR2 promoter (location of E2F
site, sequence and orientation of E2F site, presence or absence of binding sites for other
transcription factors) to identify promoter elements conferring this specificity. The specificity of
target gene activation has only been reported to occur with respect to the expression of endogenous
target genes. The ability of these overexpression assays to maintain specificity in
extrachromosomal constructs has therefore not yet been determined, and the identification of the
promoter elements conferring specificity may require the modification of the endogenous loci.
This, although more laborious than studies involving reporter plasmids, should nevertheless be
feasible for a limited number of mutations, and might be very productive when used in conjunction
with chimeric protein experiments.
167
Coupling E2F biology to E2F target genes
Understanding how E2F regulates transcriptional targets and understanding how these
promoters are targetted specifically by individual E2F complexes are of interest primarily insofar as
they relate to cellular and organismal biology. Overexpression of a version of E2F-1 lacking the
residues required for its pro-apoptotic function, for example, could potentially be used to identify
the transcriptional target genes mediating this effect. The overexpression of wild-type E2F- 1
induces p19-ARF mRNA levels more efficiently than the overexpression of E2F-2, -3 or -4 does
(DeGregori et al., 1997), and p19-ARF has been proposed to mediate at least some of the proapoptotic effects of E2F-1 (Bates et al., 1998). It would be of interest to determine if the residues
required for these pro-apoptotic effects also mediate the specificity of pI9-ARF induction by E2F1. Other potential E2F- 1 target genes could also be assayed for their dependence on any proapoptotic domain of E2F- 1. It should even be possible to screen for novel genes whose
expression can be induced by wild-type E2F- 1 but not by a version lacking the pro-apoptotic
residues; such genes would be additional candidates for mediators of the pro-apoptotic functions of
E2F-1.
Although other E2F genes have also been shown to have specific functions, these have
generally been in loss-of-function assays (e.g. deletion, or antibody-mediated inactivation), and
may therefore require other approaches to match deregulated target genes to biological
consequences. E2F-3-deficient MEFs, for example, proliferate more slowly than wild-type cells
(P. Humbert, K. Rogers, S. Dandapani, and J. Lees, unpublished data). This is presumably
mediated by inappropriate levels of one or more E2F-3 target genes, resulting either from a lack of
E2F-3/ DP-mediated transcriptional activation or from a lack of pRB/ E2F-3/ DP-mediated
transcriptional repression; the overall level of expression could be inappropriate, or merely its
timing during the cell cycle. A careful study of the expression of E2F target genes in these cells
would therefore be predicted to identify one or more genes whose timing or levels of expression
differ from wild-type expression patterns; some subset of these E2F-3 target genes are presumably
involved in the reduced proliferation rate of E2F-3 deficient MEFs. In contrast, E2F-4-deficient
168
MEFs have an approximately wild-type proliferation rate (P. Humbert, K. Rogers, S. Dandapani,
and J. Lees, unpublished data). Any E2F target genes deregulated in both E2F-3-deficient MEFs
and in E2F-4-deficient MEFs are therefore less likely to be involved in the proliferative effects of
E2F-3 than are target genes only deregulated in E2F-3 deficient MEFs. In contrast, loss of E2F-4
function disrupts hematopoesis (P. Humbert, K. Rogers, S. Dandapani, and J. Lees, unpublished
data), and loss of E2F-5 deregulates cerebrospinal fluid production in the choroid plexus
(Lindeman et al., 1998); an analysis of E2F target gene expression in these tissues of mutant mice
may enhance the current understanding of these functions of E2F-4 and -5. Indeed, cells from
these mice could also be used in screens to identify novel E2F transcriptional targets that may be
specifically regulated by these E2F complexes.
169
C. Broadening our understanding of E2F: tumor genetics and genetic screens
C1.
E2F and cancer genetics
E2F-1 as a tumor suppressor
Of particular interest to cancer biologists are the roles played by the E2F genes in
tumorigenesis. A loss-of-function mutation of mouse E2F- 1 is associated with late-onset tumors
of the reproductive tract (Yamasaki et al., 1996). The causes of this tumorigenesis are unknown.
It could result from the failure to express an E2F- 1-responsive oncogene, or from a derepression
of a pRB/ E2F- 1/ DP-regulated tumor suppressor. Whether this tumorigenesis results from a cell
autonomous defect in these mice has not been reported, nor has it been reported if the tumors that
also arise in mice heterozygous for an E2F-1 mutation display a loss-of-heterozygosity at the
locus. The relevance of this tumorigenesis to human cancer, moreover, has not yet been
established; no tumor-specific alterations in the human E2F- 1 gene have been reported. This is
perhaps surprising, unless the defect is not cell autonomous, in which case its specific detection in
tumors would not be expected.
pRB inactivation, tumorigenesis,and E2F
Inactivation of pRB, in contrast, is known to be an extremely frequent event in human
cancers, either through loss-of-function mutations in the RB gene, mutations in the cyclins, CDKs,
and CDK inhibitors that modulate pRB phosphorylation, or expression of viral oncoproteins
directly inactivating pRB, p107, and p130. Understanding the effects of E2F deregulation in this
context could answer many important questions. First, such an understanding would address the
degree to which E2F deregulation is a (or perhaps the) critical functional target in pRB-suppressed
tumorigenesis. Second, it would determine whether derepression of pRB-regulated transcriptional
targets is sufficient for tumorigenesis, or if there is a further requirement for E2F-mediated
transcriptional activation of these targets. Third, it would identify the subset of E2F genes relevant
170
to tumorigenesis. Fourth, it would also, ideally, identify the transcriptional targets of this subset
that promote tumor formation and growth.
This work has already begun with studies done on E2F- 1 inactivation in the background of
SV40 expression or RB heterozygosity in the mouse. In the context of SV40-induced tumors of
the choroid plexus, E2F-l mutation simultaneously reduces apoptosis five-fold, presumably
accelerating tumor growth, and reduces cell cycle progression, leading to an unchanging overall
rate of tumor growth (Pan et al., 1998). In mice heterozygous for an RB loss-of-function
mutation, deletion of E2F- 1 prevents tumorigenesis in the thyroid (0% penetrance versus 53% in
the presence of E2F-1) and reduces its occurence in the pituitary (62% penetrance versus 95%)
(Tsai et al., 1998). Thus, in at least one tissue the induction of tumorigenesis by the loss of pRB
requires the deregulation of E2F-1, and in another tissue the efficiency of tumorigenesis requires
E2F- 1. Deletion of E2F- 1 in a background of RB heterozygosity also extends the mouse life span,
presumably as a direct result of decreased tumorigenesis (Tsai et al., 1998); the decreased
penetrances of the tumor phenotypes are therefore more striking, given that the tumors have had
additional time to arise. These results also indicate that the transcriptional activation function of
E2F- 1 is essential for the efficient tumorigenesis of RB heterozygotes and for efficient tumorassociated apoptosis and cell proliferation in the SV40 transgenic mice. Interestingly,
tumorigenesis in the thyroids of RB-heterozygous mice may also be dependent on the dosage of
E2F-1: tumor penentrance in this tissue is reduced from 53% to 6% by heterozygous mutation of
the E2F-1 gene (Tsai et al., 1998), arguing that only maximal transcriptional activation by E2F-I is
sufficient for efficient tumorigenesis in this tissue.
While these studies clearly argue that E2F activity is a critical target for tumorigenesis
induced by loss of pRB, other important questions remain unresolved. One is the degree of
involvement of other E2F genes in this pathway. The effects of E2F-I gene dosage in thyroid
tumorigenesis, for example, could be explained by a requirement for high levels of an E2F- 1specific target gene, or by a requirement for high levels of a function shared among the E2F
proteins, such as the regulation of S phase entry. Studies on the effects of mutations in other E2F
171
genes in the background of pRB inactivation may help to address this question. If mutation of
E2F-2, -3, or -4 phenocopies a mutation of E2F- 1, that may argue that a function shared among
the E2F genes is required for efficient tumorigenesis, and that the mutation of any of these genes
merely drops the overall E2F levels below a threshold requisite for tumorigenesis. The formal
possibility that a different function for each E2F gene, required in for tumorigenesis in the same
tissues and to the same degree as for E2F-1, would remain, but would seem a much less likely
explanation. If, on the other hand, mutation of the other E2F genes does not affect tumorigenesis,
or affects tumorigenesis in different tissues, this would indicate a very specific requirement for the
deregulation of particular E2F complexes in tumorigenesis.
Identifying the E2F genes that promote tumorigenesis willfocus future research
The identification of the E2F proteins relevant to tumorigenesis upon pRB inactivation is
important for at least two reasons. (i) If, for example, E2F- 1 has a role in the tumorigenesis of
some tissues that is not shared with other E2F genes, then the search for E2F transcriptional targets
relevant to tumorigenesis can be narrowed to E2F-1-specific targets in these tissues. (ii) Research
can be focused on this branch or branches of the E2F pathway in an attempt to identify
opportunities for clinical intervention. Given that the E2F proteins are regulated in different ways,
it might be possible to selectively inactivate those complexes required for tumorigenesis. A
generalized inhibition of E2F complexes may prove to have wide ranging effects in vivo: the
introduction of double-stranded DNA bearing E2F DNA binding sites into smooth muscle cells,
for example, inhibits their proliferation in rats (Morishita et al., 1995). In contrast, it has been
suggested that an in vitro selection scheme that has isolated an RNA ligand capable of inactivating
E2F proteins 1-5 and inhibiting Gl/S-progression could be modified to identify ligands selectively
inhibiting individual E2F complexes (Ishizaki et al., 1996). If so, such a ligand might someday be
used in cancer patients to inactivate only those E2F complexes that contribute to tumor growth,
while minimizing any side effects associated with the additional inactivation of other E2F
complexes.
172
C2.
Genetic screens could uncover additional E2F functions or regulators
"Reverse genetics", in which an animal with an engineered mutation is studied to further
the understanding of a particular gene, can be invaluable in assessing the functions of individual
E2F genes in biological pathways. Nevertheless, these studies have intrinsic limitations. One
such limitation is that although these studies can reveal biological functions for which the gene in
question is essential, they do not reveal functions for which the activity of the gene is simply
sufficient, but not essential. ElA, for example, is known to promote cellular proliferation both
through regions of the protein that inactivate pRB family members and through regions that bind
p300 and CBP; either region of the protein can be sufficient for this effect (Zerler et al., 1987;
Bellett et al., 1989; Howe et al., 1990). It is possible, by extension, that in some cellular contexts
E2F transcription might be sufficient but not required to promote entry into the cell cycle. Mouse
studies in which E2F genes have been inactivated would not reveal such a function.
The identification of biological activities that the E2F pathway may share with other
signalling pathways may require other approaches, such as genetic screens in Drosophilaor other
genetically tractable organisms. There may be differences between the functions of E2F in
Drosophilaand in mammals; any function suggested by studies in other organisms would need to
be confirmed in mammals. Nevertheless, overexpression studies and mutational studies indicate
that in both Drosophilaand mammals E2F activity is required for the GI/S-specific transcription of
genes involved in DNA synthesis and cell cycle control, that E2F activity promotes cell cycle
progression and that an inhibitor of E2F inhibits cell proliferation, and that the overexpression of
an E2F gene can induce apoptosis (reviewed in CHAPTER I). Thus, many functions of E2F have
been conserved among these species, and it seems reasonable to expect that any additional
functions attributed to the pathway might be similarly conserved. Furthermore, Drosophilahas
recently been shown to have at least two E2F genes that vary in their transcriptional activity
(Sawado et al., 1998), and E2F-responsive genes in Drosophilahave been reported to vary in the
173
extents to which their responsiveness results from E2F-mediated repression or E2F-mediated
activation, suggesting additional similarities to mammalian E2F regulation.
Genetic screens could reveal much about the functions and regulation of E2F. It is
possible, for example, to screen for mutations that are synthetically lethal with weakly
hypomorphic mutations in the E2F pathway, or for mutations accelerating the lethality of a dE2F or
dDP null mutation. Such a screen could uncover new biological functions of E2F in the regulation
of growth, proliferation, apoptosis, differentiation, or other aspects of development. At the same
time, it would reveal genes that regulate E2F function (such as pRB family members), genes that
participate in E2F function (such as additional E2F/ DP family members), critical targets of E2F
function (such as cyclin E), or genes in other signalling pathways that cooperate with E2F to
induce these biological effects. Other potential genetic screens to identify modifiers of E2F
function include screening for mutations affecting the expression pattern of E2F-responsive genes,
a technique that has already successfully identified mutations both in dDP and in other
complementation groups (Royzman et al., 1997); the proper development of the Drosophila eye is
very sensitive to the gene dosage of cell cycle regulators, including dE2F and RBF (Du et al.,
1996; Secombe et al., 1998) and could also be a basis for screening for mutations interacting
genetically with E2F.
C3.
Genetics, molecular biology and biochemistry in the study of E2F
The importance of E2F regulation was first suggested by the genetics of El A, the genetics
of RB, and the realization that ElA releases E2F from pRB. Biochemical and overexpression
studies of E2F function have revealed much about its roles in transcriptional activation and
repression, and have suggested functions in the regulation of cell biology, some of which have
been confirmed genetically. These biochemical and overexpression studies may yet give us more
insight into the ways E2F is regulated, and into the ways E2F regulates transcription and cell
biology. Molecular biology will likely reveal the missing participants in E2F activity. Ultimately,
however, the genetics will confirm or disprove the importance of E2F in tumorigenesis, apoptosis,
174
and other aspects of cell biology, and the genetics may prove the most efficient means to identify
additional regulators of E2F function, additional mediators of E2F function, and additional E2F
functions in transcriptional control, cell biology, and organismal biology.
175
D.
References
Bates, S., Phillips, A. C., Clark, P. A., Stott, F., Peters, G., Ludwig, R. L., and Vousden, K.
H. (1998). p14ARF links the tumour suppressors RB and p53 [letter]. Nature 395, 124-5.
Bellett, A. J., Jackson, P., David, E. T., Bennett, E. J., and Cronin, B. (1989). Functions of the
two adenovirus early ElA proteins and their conserved domains in cell cycle alteration, actin
reorganization, and gene activation in rat cells. J Virol 63, 303-10.
Cobrinik, D., Lee, M. H., Hannon, G., Mulligan, G., Bronson, R. T., Dyson, N., Harlow, E.,
Beach, D., Weinberg, R. A., and Jacks, T. (1996). Shared role of the pRB-related p130 and
p107 proteins in limb development. Genes Dev 10, 1633-44.
DeGregori, J., Leone, G., Miron, A., Jakoi, L., and Nevins, J. R. (1997). Distinct roles for
E2F proteins in cell growth control and apoptosis. Proc Natl Acad Sci U S A 94, 7245-50.
Du, W., Vidal, M., Xie, J. E., and Dyson, N. (1996). RBF, a novel RB-related gene that
regulates E2F activity and interacts with cyclin E in Drosophila. Genes Dev 10, 1206-18.
Howe, J. A., Mymryk, J. S., Egan, C., Branton, P. E., and Bayley, S. T. (1990).
Retinoblastoma growth suppressor and a 300-kDa protein appear to regulate cellular DNA
synthesis. Proc Natl Acad Sci U S A 87, 5883-7.
Ishizaki, J., Nevins, J. R., and Sullenger, B. A. (1996). Inhibition of cell proliferation by an
RNA ligand that selectively blocks E2F function. Nat Med 2, 1386-9.
Lee, M. H., Williams, B. 0., Mulligan, G., Mukai, S., Bronson, R. T., Dyson, N., Harlow, E.,
and Jacks, T. (1996). Targeted disruption of p107: functional overlap between p107 and Rb.
Genes Dev 10, 1621-32.
Lindeman, G. J., Dagnino, L., Gaubatz, S., Xu, Y., Bronson, R. T., Warren, H. B., and
Livingston, D. M. (1998). A specific, nonproliferative role for E2F-5 in choroid plexus function
revealed by gene targeting. Genes Dev 12, 1092-8.
Morishita, R., Gibbons, G. H., Horiuchi, M., Ellison, K. E., Nakama, M., Zhang, L., Kaneda,
Y., Ogihara, T., and Dzau, V. J. (1995). A gene therapy strategy using a transcription factor
decoy of the E2F binding site inhibits smooth muscle proliferation in vivo. Proc Natl Acad Sci U
S A 92, 5855-9.
Muller, H., Moroni, M. C., Vigo, E., Petersen, B. 0., Bartek, J., and Helin, K. (1997).
Induction of S-phase entry by E2F transcription factors depends on their nuclear localization. Mol
Cell Biol 17, 5508-20.
Pan, H., Yin, C., Dyson, N. J., Harlow, E., Yamasaki, L., and Dyke, T. V. (1998). Key roles
for E2F1 in signaling p53-dependent apoptosis and in cell division within developing tumors [In
Process Citation]. Mol Cell 2, 283-92.
Robanus-Maandag, E., Dekker, M., van der Valk, M., Carrozza, M. L., Jeanny, J. C.,
Dannenberg, J. H., Berns, A., and te Riele, H. (1998). p107 is a suppressor of retinoblastoma
development in pRb-deficient mice. Genes Dev 12, 1599-609.
Royzman, I., Whittaker, A. J., and Orr-Weaver, T. L. (1997). Mutations in Drosophila DP and
E2F distinguish Gl-S progression from an associated transcriptional program. Genes Dev 11,
1999-2011.
176
Sawado, T., Yamaguchi, M., Nishimoto, Y., Ohno, K., Sakaguchi, K., and Matsukage, A.
(1998). dE2F2, a novel E2F-family transcription factor in Drosophila melanogaster. Biochem
Biophys Res Commun 251, 409-15.
Secombe, J., Pispa, J., Saint, R., and Richardson, H. (1998). Analysis of a Drosophila cyclin E
hypomorphic mutation suggests a novel role for cyclin E in cell proliferation control during eye
imaginal disc development. Genetics 149, 1867-82.
Tsai, K. Y., Hu, Y., Macleod, K. F., Crowley, D., Yamasaki, L., and Jacks, T. (1998).
Mutation of E2f-l suppresses apoptosis and inappropriate S phase entry and extends survival of
Rb-deficient mouse embryos [In Process Citation]. Mol Cell 2, 293-304.
Verona, R., Moberg, K., Estes, S., Starz, M., Vernon, J. P., and Lees, J. A. (1997). E2F
activity is regulated by cell cycle-dependent changes in subcellular localization. Mol Cell Biol 17,
7268-82.
Yamasaki, L., Jacks, T., Bronson, R., Goillot, E., Harlow, E., and Dyson, N. J. (1996).
Tumor induction and tissue atrophy in mice lacking E2F-1. Cell 85, 537-48.
Zerler, B., Roberts, R. J., Mathews, M. B., and Moran, E. (1987). Different functional domains
of the adenovirus ElA gene are involved in regulation of host cell cycle products. Mol Cell Biol
7, 821-9.
177
ACKNOWLEDGEMENTS
Those who have helped me over the course of this work are many, very many, and I thank
you each and all. Jackie deserves a special note of thanks for accepting me into her new lab and
encouraging me to learn what it is to be a scientist. Special thanks, too, should go to all of the
members of the Lees lab who gave their support in the times of scientific bounty and of drought. I
thank Joel Pomerantz who gave me insights into "graduate studentdom" that I am still unraveling
and appreciating to this day. I would also like to thank all of the people of the fifth floor who made
lab a more interesting and enjoyable place to be. I doubt I would have made it this far without the
constant support of my wife Susan, to whom I am particularly grateful. All of you I have listed
here, and many more, had a big part in this work. It would not have happened without you.
Thank you.