Oncogene (2008) 27, 4446–4455
& 2008 Macmillan Publishers Limited All rights reserved 0950-9232/08 $30.00
www.nature.com/onc
ORIGINAL ARTICLE
The transcription-repression domain of the adenovirus E1A oncoprotein
targets p300 at the promoter
M Green, NK Panesar and PM Loewenstein
Institute for Molecular Virology, Saint Louis University School of Medicine, St Louis, MO, USA
Extensive mutational/functional analysis of the transcriptionrepression domain encoded in the N-terminal 80 amino
acids of the adenovirus E1A 243R oncoprotein suggests a
model for the molecular mechanism of E1A repression:
E1A accesses transcriptional co-activators such as p300
on specific promoters and then interacts with TBP to
disrupt the TBP–TATA complex. In support of this
model, as reported here, a basal core promoter activated
by tethering p300 is repressible by E1A at the promoter
level as shown by chromatin immunoprecipitation
(ChIP) analysis. Sequestration of p300 by E1A does not
play a significant role, as indicated by dose-response
measurements. Furthermore, when the core promoter
is transcriptionally activated by tethering activation
domains of several transcription factors that can recruit
p300 (p65, MyoD, cMyb and TFE3), transcription is
repressible by E1A. However, when the core promoter is
activated by factors not known to recruit p300 (USF1 and
USF2), transcription is resistant to E1A repression.
Finally, tethering p300 to the non-repressible adenovirus
major late promoter (MLP) renders it repressible by E1A.
ChIP analysis shows that E1A occupies the repressed
MLP. These findings provide support for the hypothesis
that p300 can serve as a scaffold for the E1A repression
domain to access specific cellular gene promoters involved
in growth regulation.
Oncogene (2008) 27, 4446–4455; doi:10.1038/onc.2008.85;
published online 14 April 2008
Keywords: adenovirus; Ad E1A; transcription repression; p300; promoter interaction
Introduction
The E1A oncogene is the first viral gene expressed
during productive infection of cells with human
adenoviruses (Ads). Group C Ad E1A encodes two
major regulatory proteins of 243 and 289 amino-acid
Correspondence: Dr M Green, Institute for Molecular Virology, Saint
Louis University School of Medicine, EA Doisy Research Center, 1100
S Grand Boulevard, St Louis, MO 63104, USA.
E-mail: [email protected]
Received 10 September 2007; revised 11 January 2008; accepted 4
February 2008; published online 14 April 2008
residues (E1A 243R and E1A 289R) synthesized from
alternatively spliced RNA transcripts of 12S and 13S,
respectively (reviewed by Shenk, 2001). E1A proteins
encode multiple domains with diverse biochemical and
biological functions including transcriptional activation,
transcriptional repression, induction of cellular DNA
synthesis, cell immortalization, cell transformation, as
well as the inhibition of metastasis and cell differentiation (reviewed by Gallimore and Turnell, 2001;
Shenk, 2001; Frisch and Mymryk, 2002; Berk, 2005;
Chinnadurai, 2006). E1A 289R differs from E1A 243R
by conserved region 3 (CR3) (amino-acid residues
140–185), a 46 amino-acid domain unique to 289R. CR3
is essential and sufficient for transcriptional activation of
early Ad genes (Lillie et al., 1987; Green et al., 1988).
The Ad E1A 243R oncoprotein encodes two conserved
domains and a non-conserved N-terminus in exon 1,
which are essential for cell immortalization and cell
transformation and can induce S-phase DNA synthesis
and cell cycle progression by two pathways. The first,
the Rb-E2F pathway, involves E1A sequences within
CR1 (residues 41–80) and CR2 (residues 121–139),
which possess contact sites for Rb family proteins. The
second, the N-terminal pathway, is a major focus of
our laboratory and has been mapped within the E1A
N-terminal 80 amino acids (E1A 1–80). E1A 1–80
consists of CR1 and non-conserved residues 1–40
and takes on added importance because the growth
regulatory functions of E1A require sequences within
this region.
The protein domains of E1A have evolved to interact
with key cellular transcription regulators and promoters
to control cell cycle progression, cell differentiation and
chromatin remodeling. There is limited understanding
of the molecular mechanisms and the precise roles of
cellular regulatory factors in these processes. The
systematic study of E1A functional activities continues
to provide opportunities to explore important and
complex questions in cell growth regulation, oncogenesis and virus regulatory control.
An interesting biochemical function encoded in
the E1A N-terminus is the ability to transcriptionally
repress cellular genes involved in cellular proliferation
and cell differentiation. The E1A repression function
involves interaction with two pivotal cellular proteins—
the TATA-binding protein (TBP) component of
TFIID, which is important for transcription initiation
and the multifunctional co-activator and histone
E1A transcription repression targets promoter-bound p300
M Green et al
4447
acetyltransferase p300. Sequences within the E1A
N-terminus have been reported to bind several other
important cellular proteins, including PCAF, p400,
hGCN5, TRRAP and Tip60, which are likely involved
in chromatin remodeling and histone acetylation
(reviewed by Frisch and Mymryk, 2002).
A recombinant protein containing only the
N-terminal 80 amino acids strongly represses
transcription of E1A-repressible promoters in a manner
that recapitulates the repression activity of the
full-length E1A 243R protein in vitro and in vivo (Song
et al., 1995a–c, 1997; Boyd et al., 2002; Loewenstein
et al., 2006). A detailed mutational/functional analysis
has identified two regions or sub-domains that are
critical for E1A repression: amino acids approximately
1–30 and approximately 48–60. Key amino acids in the
first sub-domain include (i) residues 2–6 with 6Cys being
especially important, and (ii) residue 20Leu. All of these
residues are essential for the transcription-repression
function and for disruption of a TBP–TATA complex,
but only amino-acid residue 6 appears to be critical for
binding p300 under in vitro conditions (Boyd et al.,
2002). Of interest, the E1A N-terminus can interact with
the DNA methyltransferase, Dnmt1, and this reaction is
abolished by a point mutation at Leu 20 of E1A
(Burgers et al., 2007), which is critical for E1A
transcription repression (Boyd et al., 2002). In contrast,
amino acids 53Ala, 54Pro, 55Glu and 56Asp within
the second sub-domain are important for the E1A
repression function and for binding of E1A’s cellular
partner p300, but not for binding TBP or for disruption
of a TBP–TATA complex (Loewenstein et al., 2006).
These combined findings suggest a two-step model as a
molecular mechanism for E1A repression (Figure 1).
Briefly, E1A gains access to repressible promoters by
interaction with cellular partners such as p300 as
molecular scaffolds and, thereby, gains access to TBP
where it disrupts the basal transcription machinery.
Consistent with this model, TBP is recruited to a p300E1A 1–80 complex as shown by an in vitro assembly
assay (Loewenstein et al., 2006). To further explore the
first step of this model, we have carried out studies using
model promoters containing tethered p300 and promoters containing tethered transcription activators that
either recruit or do not recruit p300. The results of these
transcription-repression experiments support a direct
role for p300 in the E1A repression function consistent
with the proposed model.
Results
Transcription repression by E1A 243R in vivo
is promoter-specific and targets the basal transcription
machinery
Several lines of evidence show that E1A repression
targets the TBP component of the basal transcription
machinery (Song et al., 1995a, 1997; Boyd et al., 2002):
(i) TFIID or recombinant TBP reverses transcription
repression by E1A in vitro; (ii) TBP restores the ability
of an E1A 1–80 affinity-depleted nuclear extract to
Figure 1 Schematic of two-step model for E1A transcription
repression.
Figure 2 Transcription-repression analysis of the E1A repressible
HIV promoter (Benn-CAT) and the non-repressible major late
promoter (MLP-CAT). A cartoon of the reporter plasmids is
shown at the top of the figure. C33a cells in 6-well plates were
transfected with pBenn-CAT (70 ng) or pMLP-CAT (200 ng) using
a 5:1 ratio of Fugene (Roche) to DNA. Coexpression of 500 ng
of the E1A 243R-expressing plasmid pcDNA3-E1A 243R
effectively represses transcription from pBenn-CAT but not from
pMLP-CAT, as measured by CAT assay and quantitated on a
Phosphorimager.
support transcription and (iii) TBP–TATA interaction
in vitro is disrupted by E1A 243R or E1A 1–80. If E1A
targets the basal transcription machinery, why aren’t all
Oncogene
E1A transcription repression targets promoter-bound p300
M Green et al
4448
promoters repressed by E1A? As shown in Figure 2,
transcription from the reporter plasmid pBenn-CAT
(the CAT gene driven by the upstream 533 nt of HIV
LTR) is efficiently repressed by expression of E1A 243R
in transient expression experiments. In contrast, transcription from the reporter plasmid pMLP-CAT
(the CAT gene driven by the upstream 262 nt of MLP)
is not repressed under the same conditions (and in fact is
to an extent activated). Either the MLP contains an
element(s) that protects it from repression by E1A or the
HIV LTR contains an element(s) that facilitates interaction with E1A. With regard to the E1A repression
model (Figure 1), it is noted that HIV LTR can interact
with p300 via NF-kB (Parker et al., 1997), whereas MLP
is not known to interact with p300 but instead interacts
with USF transcription factors (Carthew et al., 1985;
Sawadogo and Roeder, 1985).
A minimal core promoter activated by tethering p300
is repressible by E1A at the promoter level
The Ad E1B minimal core promoter, fused to
DNA-binding sites for the yeast Gal4 DNA-binding
protein and the bacterial Lex A DNA-binding protein
(2Lex2Gal-E1BTATA-CAT), was co-transfected with
the vector Gal4-p300 (1737–2414) (Yuan et al., 1996).
Gal4-p300 (1737–2414) expresses the Gal4 DNA-binding domain fused to p300 residues containing a
transcription-activation region and the E1A-binding
site on p300 (residues 1763–1811). Basal activity from
the core promoter is negligible, but is strongly activated
by expression of Gal4-p300 (1737–2414) (Figure 3).
Significantly, coexpression of E1A 243R efficiently
represses Gal4-p300-activated transcription by greater
than 90% (Figure 3). For comparison, the well-studied
repressor ZEB as a fusion protein with LexA
(L-RD-ZEB) also efficiently represses p300-activated
transcription (Figure 3). ZEB represses by a mechanism
that appears to be different from that of E1A (Postigo
and Dean, 1999a).
These data does not distinguish between two
possibilities: (i) E1A represses the activated promoter
by interacting with Gal4-p300 at the promoter, or
(ii) E1A sequesters intracellular Gal4-p300, thereby
squelching activation of the promoter. To address this
issue, the same construct fused to a Luc reporter
(2Lex2Gal-E1BTATA-Luc) was co-transfected with
fixed amounts of E1A 243R (50 or 100 ng) and
escalating amounts of Gal4-p300 (1737–2414). If E1A
were to repress by sequestering p300, one would expect
that some level of Gal4-p300 would overcome, at least in
part, repression by E1A. As shown, when either 100 ng
of E1A 243R (Figure 4a) or 50 ng (Figure 4b) was
transfected, no amount of co-transfected Gal4-p300 was
found to overcome E1A repression; the p300-activated
promoter was repressed over 90% in all cases by E1A
243R. Table 1 summarizes four experiments that
cover from 1- to 200-fold molar ratios of Gal4-p300
(1737–2414) to E1A 243R DNAs. As seen in Table 1,
Gal4-p300 (1737–2414) at 50, 100 and 200 ng exhibits a
linear dose response, that is, a doubling of promoter
Oncogene
Figure 3 Activated transcription from a core promoter by
tethering p300 is strongly repressed by coexpression of E1A
243R. The core promoter-reporter 2Lex2Gal-E1BTATA-CAT
(300 ng) was transfected with Gal4-p300 (1737–2414; 500 ng)
with and without pcDNA3-E1A 243R (250 ng) or L-RD-ZEB
(250 ng) and analysed as described in the legend to Figure 2. The
p300-activated core promoter is efficiently repressed by E1A 243R
or L-RD-ZEB. A schematic illustrating the interpretation of these
results is shown.
activation occurs with each doubling of Gal4-p300
concentration. E1A 243R DNA at 2 ng is below
saturating levels as it represses only at approximately
60 % at molar ratios of 50–200 of Gal4-p300
(1737–2414) to E1A 243R. E1A 243R DNA at
5–100 ng efficiently represses transcription up to 98%
at high ratios of p300 to E1A under conditions where
Gal4-p300 (1737–2414) nearly saturates the promoter
(200–800 ng) as determined by luciferase assay. The data
in Figure 4c demonstrate that Gal4-p300 (1737–2414)
and E1A 243R are efficiently expressed in transfection
experiments. Antibodies against Gal4 and E1A detects
the expression of transfected Gal4-p300 (1737–2414)
and E1A 243R in a dose-dependent manner at levels
where E1A efficiently represses the p300-activated basal
promoter. Thus, these combined results show that
Gal4-p300 (1737–2414) is expressed in a dose-dependent
manner both in terms of promoter activation and
protein levels expressed.
Finally, to demonstrate that E1A does repress by
interaction with p300 on the promoter and not merely
by sequestering transfected Gal4-p300, transient
chromatin immunoprecipitation (ChIP) analysis
(Lavrrar and Farnham, 2004) was performed with
antibodies specific for Gal4 and for E1A on the
5Gal-E1BTATA-Luc core promoter activated by
Gal4-p300 (1737–2414) and repressed by E1A 243R.
Real-time PCR analysis was carried out using primers
specific for the E1B promoter. As shown in Figure 4d,
Gal4 protein is greatly enriched on the core E1B
promoter as expected. Importantly, E1A protein is also
substantially enriched on the promoter, indicating that
E1A transcription repression targets promoter-bound p300
M Green et al
4449
Figure 4 (a–d) Increasing amounts of Gal4-p300 are unable to decrease repression by E1A 243R. C33a cells in 24-well plates were
transfected, using a 3 to 1 ratio of jetPEI (Polyplus) to DNA, with 100 ng amounts of 2Lex2Gal-E1BTATA-Luc and escalating
amounts of Gal4-p300 (1737–2414) and challenged with either 100 ng (a) or 50 ng (b) of pcDNA3-E1A 243R. Gal4-p300 (1737–2414)
was unable to decrease transcription repression by E1A. Analysis was performed by luciferase assay. (c) Protein expression levels of
E1A 243R and Gal4-p300 (1747–2414) in co-transfected cells. Shown in the first panel are immunoblot analysis of the protein
expressed from 20 and 100 ng of Gal4-p300 (1737–2414). The relative amounts of protein expressed as measured by quantitative
immunoblot analysis is indicated (Materials and methods). At these levels, there is a clear dose response with increasing amounts of
Gal4-p300 (1737–2414). The second panel shows analysis of protein expressed from 5 and 20 ng of E1A 243R expression plasmid and
the degree of transcription repression (luciferase). Also shown is the relative amounts of E1A 243R expressed as determined by
immunoblot analysis. (d) Chromatin immunoprecipitation (ChIP) analysis indicates that E1A directly associates with an E1A 243R
repressed promoter. Transient ChIP analysis was performed using 10 cm2 cultures of C33a cells transfected with 5Gal-E1BTATA-Luc
(1 mg), activated by Gal4-p300 (3 mg), and repressed by E1A 243R (2 mg) (see schematic). Real-time PCR analysis demonstrates that
p300 (Gal4) and E1A occupy the promoter. Shown is a plot of fluorescence versus PCR cycle number generated by the Opticon 2
monitor software. In parentheses are given the average Ct value (the cycle number at which fluorescence crosses the threshold value) of
triplicate samples of input chromatin, chromatin precipitated by Gal4 antibody, chromatin precipitated by E1A antibody and
background (no antibody (No-Ab) control). A difference of one Ct between an antibody specific immunoprecipitate and the No-Ab
control represents about a twofold enrichment of the 5Gal-E1BTATA-luc reporter in the sample.
under these conditions of transcription repression, E1A
interacts with the promoter containing p300. E1A 243R
is not enriched on the promoter in the absence of
co-transfected Gal4-p300 (1737–2414) (data not shown).
The minimal core promoter activated by transcription
factors that can recruit p300 is repressible by E1A
Under natural conditions, endogenous p300 is recruited
to a promoter by interaction with promoter-bound
DNA-binding transcription factors. A number of
transcription factors have been reported to possess
binding sites on p300, many of which have been mapped
on the 2414 amino acid p300 molecule (reviewed by
Chan and La Thangue, 2001). To more accurately depict
the physiological state, 2Lex2Gal-E1BTATA-CAT was
activated by transfection with activation domains of
several transcription factors fused to either Gal4 or
LexA DNA-binding domains, and analysed for ability
to be repressed by co-transfected E1A 243R.
NF-kB proteins are involved in inflammatory reactions, signal pathways and development. The transcription factor p65 member of this family binds p300 at the
KIX domain, which is distant from the CH3 domain
that binds E1A (Perkins et al., 1997). As shown in
Figure 5a, Gal4-p65 greatly activates the core promoter.
When E1A 243R is coexpressed from pcDNA3-E1A
243R, transcription is repressed to essentially
Oncogene
E1A transcription repression targets promoter-bound p300
M Green et al
4450
Table 1 Transcription-repression of p300-activated 2Lex2GalE1BTATA-Luc at different levels of E1A and p300
pcDNA3-E1A
243R (ng)
2
2
2
5
5
5
10
10
20
20
20
20
20
50
50
50
100
100
100
Gal4-p300
(1737–2410) (ng)
Molar
ratio
p300/E1A
%
repression
100
200
400
100
200
400
100
200
100
200
400
600
800
50
100
200
200
400
800
50
100
200
20
40
80
10
20
5
10
20
30
40
1
2
4
2
4
8
57
61
59
68
69
64
73
80
81
83
86
87
90
92
92
87
97
98
96
Tabulation of four independent experiments representing the averages
of multiple transfections. Transfections were performed in 24-well
plates with the indicated amount of plasmid DNA per well, 100 ng of
the 2Lex2Gal-E1BTATA-Luc reporter and a 3:1 ratio of jetPEI to
DNA. The percentage repression was measured by luciferase assay.
background levels. Coexpression with the empty
plasmid pcDNA3 does not have any significant effect
on Gal4-p65 activation. As expected, the repressor
L-RD-ZEB also represses transcription from the
activated core promoter.
This result was extended to other transcription factors
reported to interact with p300. E1A can repress the
expression of muscle-specific genes and block differentiation of muscle cells (Webster et al., 1988) in a
process regulated by the MyoD family of transactivators, which include the helix-loop-helix protein MyoD.
MyoD binds to the same general region of p300 as does
E1A (Yuan et al., 1996). As shown in Figure 5b,
Gal4-MyoD greatly activates the promoter, and
coexpression of E1A 243R dramatically represses
transcription from the activated promoter.
The transcription factor cMyb functions as an
oncogene when transduced into retrovirus genomes.
Like p65, cMyb binds to the KIX domain of p300. As
shown in Figure 5c, co-transfection of Gal4-cMyb with
2Lex2Gal-E1BTATA-CAT strongly activates the promoter. Just as with the promoter activated by MyoD
and by p65, the cMyb-activated promoter is efficiently
repressed by coexpression of E1A 243R.
TFE3 is a basic helix-loop-helix transcription factor,
which is a part of the MiT transcription factor family
that is involved, among other things, in the regulation of
osteoclast development and renal cell carcinoma.
Phosphorylation of TFE3 has been reported to trigger
recruitment of p300 in cultured osteoclasts (Weilbaecher
et al., 2001). As shown in Figure 5d, the fusion protein
Gal4-TFE3 (Postigo and Dean, 1999a) greatly activates
expression from the core promoter and activated
Oncogene
transcription is strongly repressed by E1A 243R. Thus,
activation by several transcription factors reported to
recruit p300 is efficiently repressed by E1A 243R.
These results are consistent with the transcriptionrepression model that E1A can access promoters
through interaction with p300.
The minimal core promoter activated by transcription
factors that do not interact with p300 is resistant to E1A
repression
It is important to examine transcription factors that are
not known to recruit p300 to the promoter. USF was
initially identified as a transcription factor for the MLP
promoter (Carthew et al., 1985; Sawadogo and Roeder,
1985). The USF transcription factors are basic helixloop-helix proteins. USF1 and USF2 form dimers that
recognize specific E-box elements with a similar
consensus sequence to that of cMyc. USF has been
reported to antagonize the ability of cMyc and E1A to
stimulate cell division (Luo and Sawadogo, 1996). As
demonstrated in Figure 5e, both Gal4-USF1 and Gal4USF2 activate transcription from 2Lex2Gal-E1BTATA.
However, in both cases, coexpression of E1A 243R has
no significant repression activity on the activated
promoter. By contrast, the repressor L-RD-ZEB, which
represses transcription by a mechanism different from
that of E1A, is able to repress transcription from the
promoter activated by either USF1 or USF2, demonstrating that USF-activated transcription can be
repressed by a different mechanism. The inability of
E1A to repress USF-activated transcription is consistent
with the inability of E1A 243R to repress the native
MLP promoter that can interact with USF but not p300.
Tethering p300 to the non-repressible promoter MLP
converts it to an E1A-repressible promoter
The results described above support the conclusion that
the E1A repression domain can use promoter-bound
p300 to facilitate access to the promoter. This suggests
that one feature of an E1A repressible promoter is the
presence of p300 or another co-activator on the
promoter that E1A can access. This is turn suggests
that the resistance of a non-repressible promoter to E1A
may be due, in part, to the lack of a suitable promoterbound co-activator. To test this hypothesis directly, we
co-transfected Gal4-p300 (1737–2414) with a reporter
plasmid containing the MLP promoter fused to the
CAT gene and five Gal4 DNA-binding sites (5GalMLP-CAT), followed by expression of E1A 243R from
an Ad vector. Transcription from the MLP is greatly
enhanced by Gal4-p300 (1737–2414) (Figure 6a). Significantly, MLP-enhanced transcription is substantially
repressed by E1A (Figure 6a). Thus, promoter-tethered
p300 can serve not only to activate transcription from
the MLP but also to facilitate access of E1A 243R to
the promoter.
Chromatin immunoprecipitation analysis was
performed with antibody against E1A to demonstrate
that repression involves occupancy of the promoter by
E1A. Real-time PCR analysis using primers specific for
E1A transcription repression targets promoter-bound p300
M Green et al
4451
Figure 5 Transcription of a core promoter activated by transcription factors that recruit p300 is repressible by E1A 243. C33a cells
were transfected and analysed as described in the legend to Figure 2 with (a) 2Lex2Gal-E1BTATA-CAT (200 ng), Gal4-p65 (200 ng)
and the indicated amounts of E1A 243R or L-RD-ZEB. (b) 2Lex2Gal-E1BTATA-CAT (750 ng), Gal4-MyoD (320 ng) and the
indicated amounts of E1A 243R. (c) 2Lex2Gal-E1BTATA-CAT (750 ng), Gal4-cMyb (200 ng) and pcDNA3-E1A 243R (500 ng).
(d) 2Lex2Gal-E1BTATA-CAT (750 ng), Gal4-TFE3 (320 ng) and the indicated amounts of pcDNA3-E1A 243R. (e) 2Lex2GalE1BTATA-CAT (750 ng), Gal4-USF1 (500 ng) or Gal4-USF2 (500 ng), pcDNA3-E1A 243R (500 ng) and L-RD-ZEB (500 ng). In all
four cases of transcription factors that recruit p300 (a–d) activated transcription is efficiently repressed by E1A 243R, whereas in the
two cases of transcription factors not known to use p300 activated transcription is resistant to repression by E1A 243R but is
susceptible to repression by L-RD-ZEB (e).
the MLP shows that the repressed MLP promoter
is highly enriched for E1A compared to the no-antibody
control (Figure 6b). Experiments using antibodies
against Gal4 and p300 show that Gal4-p300
(1737–2414) occupies the 5Gal-MLP-Luc promoter to
approximately the same extent as E1A (data not
shown). These findings are consistent with the
proposed model that E1A accesses the promoter by
interaction with a suitable promoter-bound co-activator
such as p300.
Oncogene
E1A transcription repression targets promoter-bound p300
M Green et al
4452
Discussion
p300 was originally discovered through its association
with E1A in co-immunoprecipitation experiments
(reviewed by Moran, 1993). The generalization has been
formulated that E1A transcription repression is caused
by sequestration of intracellular p300. However, this
explanation is far from clear (reviewed in Goodman and
Smolik, 2000; Gallimore and Turnell, 2001). We do not
exclude the possibility that p300 can interact with E1A
off the promoter, but suggest that such an interaction
may not be relevant to the mechanism of E1A repression
under physiological conditions. Interaction of p300 with
E1A off the promoter could represent an artifact of
overexpression of p300. In this regard, Eckner et al.,
1994, who first cloned and functionally analysed p300,
reported the presence of p300-specific speckles in
transfected cell nuclei , representing aggregates presumably due to p300 overexpression and its high cysteine
content. Transient expression of p300 with E1A only
marginally overcame E1A repression of the SV40
enhancer and only at very low levels of E1A. p300
mutants with deletion in the E1A-binding site overcame
(about 50%) repression of the SV40 enhancer (Eckner
et al., 1994). It was suggested that endogenous p300
was sequestered by transfected E1A and, thus, the
exogenous p300 mutants could then activate the SV40
enhancer. However, the observation that p300 mutants
defective in binding E1A were more effective in
overcoming E1A repression can be interpreted to
support our proposed model—p300 mutants lacking
the ability to bind E1A when bound to the promoter can
activate, but no longer serve as a scaffold that E1A must
use to access the promoter and repress transcription. In
several other reports, overexpression of CMV-p300
was found to overcome E1A repression of different
promoters to various degrees. In these studies, the role
of p300 as an activator for specific promoters was the
main focus and not the interaction of E1A with p300
(for example, Yuan et al., 1996; Parker et al., 1997).
The experiments described in Figures 4a–c and Table 1
are based on a different experimental approach. By use
of the basal promoter-reporter 2Lex2Gal-E1BTATALuc, complications due to endogenous p300 are
avoided. Expression of Gal4-p300 (1737–2414) delivers
the p300 E1A-binding site/activation domain directly to
the promoter-reporter where the effects of E1A are
measured. The highly transfectable C33a cell line in
combination with the efficient jetPEI transfection
reagent allows the use of low levels of E1A 243R vector
(for example, 2 ng) and truncated p300 vector (for
example, 50 ng), minimizing p300-E1A aggregate
formation. Previous reports in which E1A repression
was reduced by p300 generally used much higher
quantities of both E1A and full-length p300 vectors
(4–10 mg) where aggregate formation might compromise
the interpretation of the results. As discussed below, our
combined studies suggest a more direct and active role
for p300 in E1A transcription repression.
Single amino-acid substitution analysis of E1A
repression sub-domains showed a strong correlation
Oncogene
Figure 6 (a) Tethering p300 to the non-repressible major late
promoter (MLP) converts it to an E1A repressible promoter. C33a
cells were transfected and analysed, as described in the legend to
Figure 2, with 5Gal-MLP-CAT (300 ng), Gal4-p300 (1737–2414)
(500 ng) and pcDNA3-E1A 243R (1000 ng). MLP is activated by
Gal4-p300 (1737–2414) and repressed by E1A 243R. (b) Chromatin
immunoprecipitation (ChIP) analysis shows that E1A 243R
occupies the repressed MLP. Transient ChIP analysis was
performed on 10 cm2 cultures of C33a cells transfected with 5GalMLP-Luc (1 mg), pcDNA3 (2 mg), Gal4-p300 (1737–2414) (3 mg)
and infected with 200 multiplicity of infection of an adenovirus
vector expressing E1A 243R (Ad/CMV-E1A 243R (Green et al.,
2008) at 6 h after transfection. Shown is a plot of fluorescence
versus cycle number for triplicate samples analysed by real-time
PCR. The average Ct values of the input chromatin, chromatin
precipitated by E1A antibody and background (No-Ab) are given
in parenthesis and demonstrate that E1A is highly enriched on the
promoter. The insert shows a linear plot of log concentration
versus Ct (generated by the Opticon 2 software) of triplicate
quantification standards, 0.5 and 5 pg of the reporter plasmid
5Gal-MLP-Luc.
E1A transcription repression targets promoter-bound p300
M Green et al
4453
between sequences required for binding p300 and those
essential for E1A transcription repression (Boyd et al.,
2002; Loewenstein et al., 2006) and for cell transformation (Rasti et al., 2005). These observations, coupled
with studies that implicate the basal transcription
machinery (the TBP component of TFIID) as a
functional target of E1A repression (Song et al.,
1995a, 1997; Boyd et al., 2002), suggest a two-step
model of E1A repression: (i) E1A accesses a specific
promoter by interaction with a DNA-bound transcription co-activator (such as p300) and (ii) E1A disrupts
TBP–TATA interaction at that promoter (Figure 1).
Consistent with this model, promoters activated by
p300-interacting transcription factors are strongly repressed by E1A 243R (Figures 5a–d), whereas activators
not known to interact with p300 are resistant to E1A
repression (Figure 5e). These experiments used the E1B
core promoter, which has negligible basal activity unless
activated by a transcription factor tethered to the
promoter. Thus, the interaction of E1A with specific
transcription activators via their interacting co-activator, that is p300, can be examined. The results indicate
that the E1B core promoter containing bound p300 can
be strongly repressed by E1A 243R. The conclusion
that E1A repression involves interaction with
promoter-bound p300 rather than intracellular p300
(sequestration or squelching) is supported by two lines
of evidence. First, dose-response measurements show
that increasing amounts of p300 do not significantly
decrease E1A repression (Figures 4a and b). Second,
ChIP analysis demonstrates that E1A 243R is present on
the E1A repressed promoter (Figure 4d).
Studies with MLP provide additional support for the
proposed model. MLP, which is not repressed by
E1A under natural conditions, can be converted to an
E1A-repressible promoter by tethering p300 (Figure 6a).
The conversion from a non-repressible to a repressible
promoter requires the occupancy of the promoter by
E1A, as shown by ChIP analysis (Figure 6b). This
compelling observation indicates that promoter-bound
p300 can be targeted by the E1A repression domain
to transcriptionally repress transcription from the
promoter.
Several caveats are noted. First, the fact that a
promoter does not use p300 does not necessarily mean
that it may not be repressible by E1A. E1A could target
other E1A N-terminus-interacting co-activators, for
example, CBP, p400, TRRAP, GCN5 and Tip60.
Second, the fact that a promoter uses p300 does not
necessarily mean that it will be repressible by E1A. The
critical issues may be (i) whether E1A can bind to the
specific promoter via interaction with a co-activator,
and (ii) whether p300-bound E1A can subsequently
interact with TBP. Promoters that do not facilitate these
two interactions would not be repressed. Different
promoters contain numerous interacting regulators,
including for example, upstream enhancers, activators,
repressors and TAFs (reviewed by Taatjes et al., 2004)
whose combinatorial interactions could sterically block
interaction of E1A with a promoter-bound co-activator
or co-activator-bound E1A with the TBP–TATA
complex. Further studies are needed to understand, in
molecular detail, the mechanism by which E1A accesses
and represses the complex endogenous promoters
involved in growth regulation.
Materials and methods
Plasmids
The construction of pcDNA3-E1A 243R is described in Boyd
et al., 2002 and Gal4-p300 (1737–2414) in Yuan et al., 1996.
Promoter-reporter plasmids 2Lex2Gal-E1BTATA-Luc and
5Gal-E1BTATA-luc are described in Dorris and Struhl,
2000; 5Gal-MLP-Luc, 2Lex2Gal-E1BTATA-CAT and
5Gal-MLP-CAT in Weintraub et al., 1995 and Postigo and
Dean, 1999b. pBenn-CAT was provided by the NIAIDS
repository. Gal4-activator plasmids were obtained from
several sources. Gal4-NF-kB-p65; Gal4-MyoD (amino acid
1–318); Gal4-cMyb; Gal4-TFE3 (amino acid 2–216) and
LexA-RD-ZEB are described in Weintraub et al., 1995 and
Postigo and Dean, 1999b; Gal4-USF1 and Gal4-USF2 in
Qyang et al., 1999. All plasmids used in transfection studies
were purified by double CsCl density gradient centrifugation to
facilitate reproducible transfection.
Cell culture and transfection
C33a cells were obtained from the American Type Culture
Collection and maintained in Dulbecco’s modified Eagle’s
medium and 10% fetal bovine serum. Cells were transfected as
described in the figure legends with either Fugene (Roche,
Indianapolis, IN, USA) or jetPEI (Bridge Bioscience Corp.,
Portsmouth, NH, USA) using the manufacturer’s guidelines.
The DNA content of transfection experiments was equalized
by addition of the empty vector pcDNA3. Co-transfected
plasmids expressing RSV-bGal (CAT assay) or Renilla
luciferase (luciferase assay) were used to normalize transfection
results. Transfections were repeated at least twice.
Quantitation of promoter activity
Depending upon the reporter, expression was measured either
by CAT assay or luciferase assay. CAT assay was performed
essentially as described by Gorman et al., 1982. Quantitation
of CAT activity was determined by Phosphorimage analysis of
thin layer plates on a Storm 840 Phosphorimager (GE
Healthcare, Piscataway, NJ, USA). Luciferase assay was
performed with the Dual-Luciferase assay system (Promega,
Madison, WI, USA) and quantitated using a Turner Design
luminometer (TD-20/20).
Immunoblot analysis
Western blot analysis was performed using primary antibody
against Gal4 (Santa Cruz, Santa Cruz, CA, USA, SC-577) to
detect expression of Gal4-p300 (1737–2414) or against E1A
(Santa Cruz SC-430) to detect expression of E1A 243R. Blots
were treated with secondary antibody labeled with horse radish
peroxidase (Pierce Biotechnology, Rockford, IL, USA) and
developed using Supersignal ECL chemistry (Pierce Biotechnology). After exposure to X-ray film, signals on blots were
quantitated on a Typhoon Trio (GE Healthcare) using
ImageQuant TL software.
ChIP assay and real-time PCR analysis
C33a cells growing in 10 cm2 dishes were transfected as
described in the figure legends, cross-linked with formaldehyde
and chromatin isolated and immunoprecipitated essentially as
Oncogene
E1A transcription repression targets promoter-bound p300
M Green et al
4454
described in Takahashi et al. (2000) (B Dynlacht, personal
communication). To determine whether E1A 243R was
present on the promoter, monoclonal antibody against
E1A (BD Biosciences, San Jose, CA, USA, 554155) was
used. To detect Gal4-p300 (1737–2414) on the promoter,
polyclonal antibodies against Gal4 (Santa Cruz SC-577X) and
p300 (Santa Cruz SC-585X) were used. No-antibody controls
were used in all transient transfection ChIP experiments.
Isotype-specific antibody controls (HA polyclonal, SC-805
or His-6 monoclonal, SC-8036) were used in many of
the transient transfection ChIP experiments and gave
essentially the same values as no-antibody controls. The
final purified immunoprecipitated DNA in 100 ml was
used for real-time PCR analysis in 96-well plates with the
Opticon 2 real-time detection system (Bio-Rad, Hercules,
CA, USA). Briefly, to triplicate 5 ml aliquots of chromatin
DNA samples were added 5 ml of 2.5 mM primer pair specific
for the promoter-reporter plasmids 2Gal2Lex-E1BTATA-Luc
or 5Gal-MLP-Luc, and 10 ml of a 2X SYBR Green
Quantitative RT–PCR (reverse transcription–PCR) reagent
(Sigma, St Louis, MO, USA). Forty cycles of PCR
were performed using a cycling protocol recommended by
Sigma with optimization of annealing temperature for each
primer pair.
Acknowledgements
We thank Steve Weintraub, Douglas Dean, Antonio Postigo,
Antonio Giordano, Michle Sawadogo and Kevin Struhl for
the plasmid stocks. We also thank Steve Weintraub and Ling
Zhao for critical reading of the manuscript and Carolyn
Mulhall for valuable editorial assistance. This work was
supported by Research Career Award AI-04739 and Public
Health Service Grant CA29561 to MG from the National
Institutes of Health.
References
Berk AJ. (2005). Recent lessons in gene expression, cell cycle control,
and cell biology from adenovirus. Oncogene 24: 7673–7685.
Boyd J, Loewenstein PM, Tang Q, Yu L, Green M. (2002). The
adenovirus 2 E1A N-terminal domain sequence requirements for
repression of transcription in vitro and in vivo correlate with those
required for E1A interference with TBP/TATA complex formation.
J Virol 76: 1461–1474.
Burgers WA, Blanchon L, Pradhan S, De Launoit Y, Kouzarides T,
Fuks F. (2007). Viral oncoproteins target the DNA methyltransferases. Oncogene 26: 1650–1655.
Carthew RW, Chodosh LA, Sharp P. (1985). An RNA polymerase II
transcription factor binds to an upstream element in the adenovirus
major late promoter. Cell 43: 439–448.
Chan HM, La Thangue NB. (2001). p300/CBP proteins: HATs for
transcriptional bridges and scaffolds. J Cell Sci 114: 2363–2373.
Chinnadurai G (ed). (2006). CtBP Family Proteins: Unique Transcriptional Regulators in the Nucleus with Diverse Cytosolic Functions.
Springer-Eurekah Bioscience, pp 1–17.
Dorris DR, Struhl K. (2000). Artificial recruitment of TFIID, but not
RNA polymerase II holoenzyme, activates transcription in mammalian cells. Mol Cell Biol 12: 4350–4358.
Eckner R, Ewen ME, Newsome D, Gerdes M, DeCaprio J, Lawrence
JB et al. (1994). Molecular cloning and functional analysis of the
adenovirus E1A-associated 300-kD protein (p300) reveals a
protein with properties of a transcriptional adaptor. Genes Dev 8:
869–884.
Frisch SM, Mymryk JS. (2002). Adenovirus-5 E1A: paradox and
paradigm. Nat Rev Mol Cell Biol 3: 441–452.
Gallimore PH, Turnell AS. (2001). Adenovirus E1A: remodelling the
host cell, a life or death experience. Oncogene 20: 7824–7835.
Goodman RH, Smolik S. (2000). CBP/p300 in cell growth, transformation, and development. Genes Dev 14: 1553–1557.
Gorman CM, Moffat LF, Howard BH. (1982). Recombinant genomes
which express chloramphenicol acetyltransferase in mammalian
cells. Mol Cell Biol 2: 1044–1051.
Green M, Loewenstein PM, Pusztai R, Symington JS. (1988). An
adenovirus E1A protein domain activates transcription in vivo and in
vitro in the absence of protein synthesis. Cell 53: 921–926.
Green M, Panesar NK, Loewenstein PM. (2008). Adenovirus E1A
proteins are closely associated with chromatin in productively
infected and transformed cells. Virology 371: 1–7.
Lavrrar JL, Farnham PJ. (2004). The use of transient chromatin
immunoprecipitation assays to test models for E2F1-specific
transcriptional activation. J Biol Chem 44: 46343–46349.
Lillie JW, Loewenstein PM, Green MR, Green M. (1987). Functional
domains of adenovirus type 5 E1a proteins. Cell 50: 1091–1100.
Loewenstein PM, Arackal S, Green M. (2006). Mutational and
functional analysis of an essential sub-domain of the adenovirus
Oncogene
E1A N-terminal transcription repression domain. Virology 351:
312–321.
Luo X, Sawadogo M. (1996). Antiproliferative properties of the USF
family of helix-loop-helix transcription factors. Proc Natl Acad Sci
USA 93: 1308–1313.
Moran E. (1993). DNA tumor virus transforming proteins and the cell
cycle. Curr Opin Genet Dev 3: 63–70.
Parker SF, Felzien LK, Perkins ND, Imperiale MJ, Nabel GJ. (1997).
Distinct domains of adenovirus E1A interact with specific cellular
factors to differentially modulate human immunodeficiency virus
transcription. J Virol 71: 2004–2012.
Perkins ND, Felzien LK, Betts JC, Leung K, Beach DH, Nabel GJ.
(1997). Regulation of NF-kB by cyclin-dependent kinases associated
with the p300 coactivator. Science 275: 523–527.
Postigo AA, Dean DC. (1999a). ZEB represses transcription through
interaction with corepressor CtBP. Proc Natl Acad Sc USA 96:
6683–6688.
Postigo AA, Dean DC. (1999b). Independent repressor domains in
ZEB regulate muscle and T-cell differentiation. Mol Cell Biol 19:
7961–7971.
Qyang Y, Luo X, Lu T, Ismail PM, Krylov D, Vinson C. et al. (1999).
Cell-type-dependent activity of the ubiquitous transcription factor
USF in cellular proliferation and transcriptional activation. Mol
Cell Biol 19: 1508–1517.
Rasti M, Grand RJ, Mymryk JS, Gallimore PH, Turnell AS. (2005).
Recruitment of CBP/p300, TATA-binding protein, and S8 to
distinct regions at the N terminus of adenovirus E1A. J Virol 79:
5594–5605.
Sawadogo M, Roeder RG. (1985). Interaction of a gene-specific
transcription factor with the adenovirus major late promoter
upstream of the TATA box region. Cell 43: 165–175.
Shenk T. (2001). Adenoviridae: The viruses and their replication. In:
Fundamental Virology 1053–1088. Publisher: Lippincott, Williams
and Wilkins: New York, NY.
Song C-Z, Loewenstein PM, Green M. (1995b). Repression in vitro, by
human adenovirus E1A protein domains, of basal or Tat-activated
transcription of the human immunodeficiency virus type 1 long
terminal repeat. J Virol 69: 2907–2911.
Song C-Z, Loewenstein PM, Toth K, Tang Q, Nishikawa A, Green M.
(1997). The adenovirus E1A repression domain disrupts the
interaction between the TATA binding protein and the TATA box
in a manner reversible by TFIIB. Mol Cell Biol 17: 2186–2193.
Song C-Z, Loewenstein PM, Toth K, Green M. (1995c). TFIID is a
direct functional target of the adenovirus E1A transcriptionrepression domain. Proc Natl Acad Sci USA 92: 10330–10333.
Song C-Z, Tierney CJ, Loewenstein PM, Pusztai R, Symington JS,
Tang Q et al. (1995a). Transcriptional repression by human
adenovirus E1A N-terminus/conserved domain 1 polypeptides
E1A transcription repression targets promoter-bound p300
M Green et al
4455
in vivo and in vitro in the absence of protein synthesis. J Biol Chem
40: 23263–23267.
Taatjes DJ, Marr MT, Tjian R. (2004). Regulatory diversity among
metazoan co-activator complexes. Nat Rev Mol Cell Biol 5: 403–410.
Takahashi Y, Rayman JB, Dynlacht BD. (2000). Analysis of promoter
binding by the E2F and pRB families in vivo: distinct E2F proteins
mediate activation and repression. Genes Dev 14: 804–816.
Webster KA, Muscat GE, Kedes L. (1988). Adenovirus E1A products
suppress myogenic differentiation and inhibit transcription from
muscle-specific promoters. Nature 332: 553–557.
Weilbaecher KN, Motyckova B, Huber WE, Takemoto CM,
Hemesath TJ, Xu Y et al. (2001). Linkage of M-CSF signaling to
Mitf, TFE3, and the osteoclast defect in Mitf (mi/mi) mice. Mol Cell
8: 749–758.
Weintraub SJ, Chow KN, Luo RX, Zhang SH, He S, Dean DC.
(1995). Mechanism of active transcriptional repression by the
retinoblastoma protein. Nature 375: 812–815.
Yuan W, Condorelli G, Caruso M, Felsani A, Giordano A. (1996).
Human p300 protein is a coactivator for the transcription factor
MyoD. J Biol Chem 271: 9009–9013.
Oncogene