| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
From the
The p65 subunit of the inducible transcription factor NF-
Gene expression in higher eucaryotes is governed by
transactivating polypeptides which bind in a sequence-specific manner
to cis-regulatory elements of DNA. A paradigm for an inducible
transcriptional activator is nuclear factor
p65, RelB, and c-Rel are
transcriptionally active members of the NF-
In contrast to our detailed knowledge of DNA
binding domains, our current understanding of the structure of TADs is
relatively poor. TADs have been classified according to their
predominant amino acid composition as proline-rich, glutamine-rich, or
acidic (reviewed by Mitchell and Tjian(1989)). The structure of acidic
TADs is a matter of debate. They have been suggested to exist as
unstructured ``acidic blobs'' (Sigler, 1988)), amphipathic
TADs have also been
classified with respect to several functional criteria. For instance,
it was found that different classes of TADs behave differently in
histone H1 anti-repression assays (Croston et al., 1991). A
further functional difference is based on the observation that acidic
activation domains and those rich in serines and threonines are
functional in yeast cells, whereas glutamine-rich domains are not
(Berger et al., 1992). Activation domains can also be grouped
according to their ability to activate transcription from remote or
proximal promoter positions (Seipel et al., 1992). Biochemical
studies showed that different TADs contact different target molecules.
For example, it was shown that the glutamine-rich TAD of Sp1
specifically contacts TAF
In this
study, we show by squelching assays that both TADs of NF-
On-line formulae not verified for accuracy SEQUENCE I
This oligonucleotide has been labeled with
[
Cells from one of the two dishes
containing nonlabeled cells was also stimulated with PMA. For Western
blotting, the proteins were transferred from the SDS gel onto a
polyvinylidene difluoride membrane (Bio-Rad) in a semidry blot
apparatus (Schleicher und Schüll) according to the instructions of
the manufacturers. The detection of Gal4 proteins was performed by
first washing the membrane twice in TBST (10 mM Tris/HCl, pH
8, 150 mM NaCl, 0.05% Tween 20) and a subsequent incubation in
TBST containing 5% non-fat dry milk powder for 1 h. The membrane was
then incubated in a small volume of TBST, containing a 1:500 dilution
of the
Squelching experiments showed that
TA
Structure prediction programs and the absence of
any
The repeat
units are of functional importance because multimerized idealized
repeats fused to Gal4 allowed the design of artificial activators. This
finding suggests that the heptad repeats in TA
The
multimer construct containing 6 consensus repeats failed to
transactivate significantly stronger than that containing 4 repeats.
Comparable results have also been obtained after multimerization
experiments with the transactivation core domain of VP16 (Emanmi and
Carey, 1992). Using a template bearing multiple Gal4 binding sites,
this study showed that a synthetic activator containing four VP16
activation domains fused to Gal4 was transcriptionally not more active
than Gal4 containing two copies of the VP16 core domain. This finding
can be explained simply by the exhaustion of a limiting target factor.
Alternatively it could be possible that the spacing between the
respective repeats does not optimally fit the spatial requirements of
the target molecule. A previous study on TA
The
increased activity of the p65 TA
TA
Modulation of p65 activity by kinases and phosphatases which act on
TA
We thank Heike Klein and Susanne Kunz for excellent
technical assistance, Dr. Peter Angel (Karlsruhe) for the GHF
expression vector, Dr. Georg Arnold (Martinsried) for synthesis of
oligonucleotides, Dr. Brigitte Obermeier (Martinsried) for operating
the automatic DNA sequencing device, Dr. Flavio Meggio (Padua) for
purified CKII, Dr. Mark Ptashne (Cambridge, MA) for
Volume 270,
Number 26,
Issue of June 30, pp. 15576-15584, 1995
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
) of p65 NF-
B
SIMILARITY TO TA
AND PHORBOL ESTER-STIMULATED ACTIVITY
AND PHOSPHORYLATION IN INTACT CELLS (*)
ABSTRACT
INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
B
contains at least two strong transactivation domains (TADs) within its
C terminus. The first domain, TA, is contained within the
last 30 amino acids of p65, whereas TA
comprises the
adjacent 90 amino acids. In this study, squelching experiments revealed
that both TADs of p65, as well as the related subunit c-Rel, compete
for the same cofactor(s) mediating transactivation. Both TADs of p65
share a common sequence motif, which is evolutionarily conserved and
displays a remarkable degree of spatial organization when aligned on an
-helical surface. The functional importance of the common sequence
motif was confirmed by deletion analysis of TA. Within the
conserved sequence motif, a 7-amino-acid repeat was noted. Idealized
heptad repeats fused to the DNA binding domain of Gal4 were
transcriptionally active, but only as multimers. Phosphorylation and
transcriptional activity of a defined region within the TA domain was found to be stimulated by phorbol ester treatment of
cells. In contrast, TA was constitutively phosphorylated,
and its activity did not significantly respond to phorbol ester
stimulation. The stimulatory effect of phorbol ester on transcription
of the TA domain was completely blocked by the protein
kinase C inhibitor. These data suggest that protein kinase C has a dual
effect on NF-B activity. It not only causes removal of
IB- from cytoplasmic NF-B but also augments the
transactivation potential of activated nuclear NF-B.
B (NF-B).()This protein can rapidly activate numerous target
genes encoding proteins involved in inflammatory, immune, and acute
phase responses (reviewed by Baeuerle and Henkel(1994) and Thanos and
Maniatis(1995)). The DNA binding form of NF-B is composed of two
subunits. Molecular cloning revealed that the DNA-binding subunits of
NF-B in mammals comprise five proteins encoded by a novel
multigene family. The subunits contain a conserved region,
approximately 300 amino acids in length in their N terminus, which is
responsible for DNA binding, dimerization, and nuclear localization
(reviewed by Blank et al.(1992) and Schmitz and
Baeuerle(1995)). The most abundant form of NF-B consists of p50
and p65 heterodimers. NF-B is located in the cytoplasm of many
cell types in an inducible form, in which the heterodimer is complexed
to the inhibitory subunit IB (Baeuerle and Baltimore, 1988).
IB has been shown to mask the nuclear location signals of both
DNA-binding subunits, thus preventing their nuclear uptake (Beg et
al., 1992; Zabel et al., 1993). Stimulation of cells with
numerous pathogenic agents leads to rapid proteolytic degradation of
IB (Beg et al., 1993; Henkel et al., 1993; Sun et al., 1993). Within minutes, the released nucleophilic
heterodimer is then transported to the nucleus, binds to its cognate
DNA, and induces gene transcription.
B family, whereas p50
and p52 primarily serve as mere DNA-binding subunits (reviewed by Liou
and Baltimore(1993)). The respective TADs of p65, RelB, and c-Rel are
contained in their unique C-terminal portions. RelB has additional
transactivating sequences in its N terminus (Rysek et al.,
1992). p65 was shown to contain at least two independent TADs within
its C-terminal 120 amino acids (Schmitz and Baeuerle, 1991; Ballard et al., 1992; Fujita et al., 1992; Moore et
al., 1993; Ruben et al., 1992). One p65 activation
domain, TA, is confined to the terminal 30 amino acids. The
second domain, TA, is contained within the N-terminally
adjacent 90 amino acids (Schmitz and Baeuerle, 1991). The TAD of murine
c-Rel has been localized between amino acids 403 and 568 (Bull et
al., 1990).
-helices (Giniger and Ptashne, 1987), and 
-sheets (Leuther et al., 1993; Van Hoy et al., 1993). It seems that
only under certain experimental conditions, acidic TADs adopt a defined
secondary structure. As revealed by CD spectroscopy, an -helix is
induced in the case of the AH domain (Van Hoy et al., 1993),
VP16 (Donaldson and Capone, 1992), p65 TA (Schmitz et
al., 1994), and a -sheet in the case of GCN4 (Leuther et
al., 1993; Van Hoy et al., 1993).
110 (Hoey et al., 1993),
whereas the acidic activation domain of the herpes virus VP16 protein
was shown to contact TBP (Stringer et al., 1990) and TFIIB
(Lin et al., 1991), as well as TAF
40 (Goodrich et al., 1993). Further evidence for functional differences of
TADs was obtained in so-called squelching experiments, where the
activity of a given TAD is influenced by the simultaneous
overexpression of a second TAD (reviewed by Ptashne(1988)).
B p65
mutually interfere with their activating function, suggesting that they
belong to the same class of acidic TADs. This observation led to the
discovery of an evolutionarily conserved sequence motif present in both
TA and TA of p65. This motif is shown to be
necessary, but not sufficient for proper TA function.
Within the conserved motif, a heptapeptide repeat was found. An
idealized heptad repeat sequence was able to activate transcription,
but only as a multimer. Both TADs of p65 were found to be
phosphorylated in intact cells. Both the activity of TA and
its phosphorylation status were increased upon stimulation of cells
with phorbol 12-myristate 13-acetate (PMA). In contrast, TA did not significantly respond to PMA. This suggests that the
activity of NF-B can be further modulated by phosphorylation of
its active nuclear form.
Cell Culture, Transfections, and CAT
Assays
Monkey COS-7 cells were grown at 37 °C in
Dulbecco's modified Eagle's medium, supplied with 1%
penicillin/streptomycin and 10% fetal calf serum (all from Life
Technologies, Inc.). For transfection of COS cells, approximately 5
10
of exponentially growing cells were transfected
with 2 pmol of the reporter construct and 0.5 pmol of the expression
plasmid. All plasmids used for transfection experiments were purified
twice on CsCl gradients. Cells were transfected in suspension (Lopata et al., 1984) and then plated on 10-cm dishes. After another
48 h of growth, cells were harvested and extracts were prepared. The
protein concentration was determined by the method of Bradford
(Bradford, 1976), and equal amounts of protein were assayed for CAT
activity as described (Pierce et al., 1988). The acetylated
and nonacetylated forms of [
C]chloramphenicol
were separated by thin layer chromatography, and the incubation
conditions were chosen to result in conversion of
[C]chloramphenicol not exceeding 60%.
Transfections were performed at least in duplicate, and the results
were quantified by liquid scintillation counting. Each construct was
tested in at least 3 independent experiments. For squelching
experiments, between 5 and 10 10
COS cells were
transfected as described above with 4 pmol of reporter plasmid, 0.25
pmol of the Gal4 fusion construct, and routinely with 3 pmol of either
the squelching plasmid or control vector (pRc/CMV, Invitrogen). After
transfection, cells were plated onto 6-cm dishes and allowed to grow
for another 36-48 h. In the PMA induction experiments, 1
10
COS cells were transfected with 4 pmol of reporter
plasmid, 1 pmol of the Gal4-p65 fusion construct, and 1 pmol of a
RSV-lacZ construct. After transfection, the cells were split
onto two 6-cm dishes and incubated for another 10 h. One of the dishes
was treated with 20 ng/ml PMA (Sigma; dissolved in dimethyl sulfoxide),
and cells were incubated for another 16-20 h. Subsequently, the
cells were harvested, and an aliquot was assayed for
-galactosidase activity. Equal amounts of cell extracts normalized
for -galactosidase activity were then assayed for CAT activity and
compared directly. The protein kinase C inhibitor bisindolylmaleimide
(Boehringer Mannheim) was dissolved in dimethyl sulfoxide and added in
concentrations between 1 and 5 µM simultaneously with PMA
to the cells.
Constructs
Details about the construction of the
clones presented in this paper can be obtained from the authors upon
request. The constructs RSV-GHF-1 (Theill et al., 1989),
Gal4-VP16 (Schmitz and Baeuerle, 1991), and the Gal4- (Baniahmad et
al., 1992) and NF-B-dependent reporter plasmids (Pierce et al., 1988) have been described previously. All clones were
verified by restriction analysis and sequencing on the automated
sequencer Genesis 2000 (DuPont NEN). Expression and DNA binding
activity of each Gal4 fusion construct was ensured by monitoring the
migration of the nuclear Gal4 fusion proteins in band-shift experiments
using a P-labeled oligonucleotide containing a Gal4
binding site.
Electrophoretic Mobility Shift Assays
5
10
COS cells were transfected with the respective Gal4
fusion constructs and harvested 48 h later. Cells were washed with
phosphate-buffered saline and transferred to a precooled Eppendorf tube
after being scraped off the plate with a rubber policeman. Cells were
pelleted by centrifugation at 4 °C for 2 min at 400 g. The pellet was resuspended in 5
EB buffer (100
mM Tris/HCl, pH 7.5, 500 mM KCl, 25 mM MgCl
, 35% (v/v) glycerol, 5 mM dithiothreitol, 0.5 phenylmethylsulfonyl fluoride, and 1%
aprotinin). Cells were lysed by 2 cycles of freeze-thawing and
subsequently centrifuged for 20 min at 4 °C at 100,000 g. Five to ten µg of protein was incubated with 1-2
µg of poly(dI-dC) (Sigma) and incubated with 10,000 cpm of a
labeled oligonucleotide for 20 min on ice. The free and protein-bound
oligonucleotides were separated on a 4% polyacrylamide gel containing
2% glycerol. Gel and running buffer were identical and contained 25
mM Tris, 25 mM boric acid, 0.5 mM EDTA, and
1 mM MgCl
. The gel was dried after electrophoresis
and exposed to a Kodak XAR5 film. The oligonucleotide used for
electrophoretic mobility shift assays (EMSAs) contains a single Gal4
binding site, which is shown underlined. -
P]ATP using T4 polynucleotide kinase
(Boehringer Mannheim).
Immunoprecipitations and Western Blotting
COS
cells were transfected in batch with 2 pmol of expression plasmid,
plated onto 4 Petri dishes, and grown for 24 h. Two out of four dishes
were then washed and incubated in phosphate-free medium. The medium was
supplied with 1% penicillin/streptavidin and 5% fetal calf serum (Life
Technologies, Inc.) dialyzed against 20 mM Tris/HCl, pH 7.9,
and 100 mM KCl. Subsequently, the cells were grown for 4 h in
phosphate-free medium containing 0.25 mCi of
[P]orthophosphate/ml (Amersham). After another
hour of incubation with or without 50 ng/ml PMA, cells were harvested
with a rubber policeman after they had been washed twice in
phosphate-buffered saline. After centrifugation at 4 °C with 400
g, the pellet was dissolved in 100 µl of IP buffer
(20 mM Tris/HCl, pH 8, 150 mM NaCl, 0.5% Nonidet
P-40, 0.1 mM phenylmethylsulfonyl fluoride, 10 mM sodium pyrocarbonate, 100 mM NaF, and 10 mM glucose 6-phosphate). The cells were subjected to two cycles of
freeze-thawing and centrifuged for 15 min at 4 °C at 14,000
g. The supernatant was mixed with 3 µl of
-Gal4
antibody (rabbit polyclonal antibodies, a kind gift of Mark Ptashne)
and incubated on a rotating spinning wheel at 4 °C for 3 h.
Subsequently, 80 µl of Protein A-Sepharose beads (Pharmacia) were
added that have been preswollen for 20 min in 1 IP buffer and
10 µg of bovine serum albumin (Sigma). This mixture was again
incubated on a spinning wheel at 4 °C for 1 h. The beads were
washed 6 times with 1
IP buffer. Finally, the beads were boiled
for 5 min in 80 µl of 1
SDS sample buffer, and proteins
were separated on a SDS-gel. The gel was then dried and exposed to
x-ray film at -80 °C.
-Gal4 antibody. After a 4-h incubation at room temperature,
the membrane was washed 8 times in TBST and incubated for another hour
in TBST containing a 1:3000 dilution of the second anti-rabbit antibody
coupled to horseradish peroxidase (Bio-Rad). After extensive washing,
the bound antibodies were detected using the ECL system (Amersham),
according to the manufacturer's instructions.
The TADs of NF-
Squelching experiments were performed in order to
assess the effect of overexpressing various TADs on p65-dependent
transcription. An expression vector encoding the fusion protein
Gal4-p65B p65 Rely on the Same
Coactivator(s), containing the complete C terminus of
p65 linked to the DNA binding domain of Gal4 (amino acids 1-147),
was co-transfected with a Gal4-dependent reporter plasmid and an excess
of various TAD expression vectors into COS cells. Co-expression of a
12-fold excess of RelA
DNA, a p65 mutant incapable of binding to
DNA due to a lack of amino acids 79-95 (Schmitz et al.,
1994), reduced transcription of Gal4-p65 to
approximately one-fourth of the original activity (Fig. 1A, compare columns 1 and 2).
Furthermore, two derivatives of RelA
DNA,
RelADNATA, and RelADNATA were tested which lack the indicated TAD, but retain the
respective second TAD. These two constructs were used in order to
investigate whether both p65 TADs contact the same or different target
molecules. Both co-expression of RelADNATA and
RelADNATA interfered with RelA-mediated
transcription (Fig. 1A, columns 3 and 4, respectively), suggesting that TA and TA contact the same target molecule(s). Overexpression of c-Rel also
impaired transactivation by Gal4-p65 (Fig. 1A, column 5). Co-expression of the
control plasmid RelA
DNATA, a
RelA
DNA derivative lacking both TADs, did not significantly affect
p65-mediated transactivation (Fig. 1A, column
6). The activity of Gal4-p65 also remained
unaltered when the human growth hormone transcription factor GHF was
overexpressed (Fig. 1A, column 7). The finding
that both TADs of p65 contact the same molecule was confirmed by
additional squelching assays. Experiments employing either
Gal4-p65
(encompassing only the TA
domain) or Gal4-p65 (encompassing only
the TA
domain) as activators and co-transfecting the
identical squelching plasmids used in Fig. 1resulted in a
comparable expression pattern. More or less identical results were
obtained when the acidic activator Gal4-VP16 was used. Overexpression
of both p65 TADs reduced its activity, indicating that they both belong
to the same class of acidic activators as VP16 (Fig. 1B).
Figure 1:
Squelching experiments. A,
squelching of p65. In each column shown, the construct
Gal-p65 was transfected together with a
Gal4-dependent CAT reporter gene and one of the indicated squelching
plasmids into COS cells. Column 1 received as a control an equal amount
of control vector (pRc/CMV). A p65 mutant incapable of binding to DNA
was co-transfected in column 2. Also, two derivatives of that construct
lacking either TA
(column 3) or TA (column 4)
were tested. The effect of co-expression of an expression vector
containing c-Rel is shown in column 5. The effect of the p65 mutant
lacking both TADs is displayed in column 6, and the effect of
co-expression of the transcription factor GHF in column 7. The
different effectors were co-transfected in a 12-fold excess in one
typical series of experiments shown here. B, squelching of
VP16. In each column shown, the construct Gal4-VP16 was used as
activator plasmid. The experimental conditions were the same as
described for part A of this
figure.
The observed squelching effects were
dependent on the dose of the added squelching plasmid. A maximal
inhibitory effect was reached at a 12-fold excess of the squelching
plasmid over the activator plasmid. The amount of DNA-protein complex
formed between the Gal4 fusion proteins and a P-labeled
oligonucleotide containing a Gal4 binding site in mobility shift assays
was not significantly affected by the amount of co-transfected
squelching plasmids (data not shown).
A Common Sequence Motif in p65 TA
The finding that both TADs of p65 behaved
similarly in squelching experiments prompted us to search for sequence
homologies between the two domains. By inspection, we noted the
homology shown in Fig. 2A. The region within TAand
TA which was homologous to TA is referred to as
TA`. As is evident from Fig. 2B, both
homology regions are conserved between human (Ruben et al.,
1991), mouse (Nolan et al., 1991), and Xenopus p65
(Kao and Hopwood, 1991), suggesting the functional importance of this
evolutionarily conserved region. A recent study analyzing point mutants
of TA showed that the homology region is important for
TA function (Schmitz et al., 1994). Furthermore,
this region was shown to form an -helix in a hydrophobic solvent,
as revealed by CD spectroscopy. On the assumption that the TA` region can also form an -helical structure, both domains are
shown displayed on a helical wheel (Fig. 2C). Only the
regions free of helix-breaking amino acids are shown over a stretch of
18 amino acids. The only exception is the Xenopus TA region starting with a single glycine at position 1 of the
helical wheel, but proceeding over the following 17 amino acids without
any helix-breaking amino acids.
Figure 2:
TA contains a domain
homologous to TA. A, positions of two homology
regions within p65. The DNA binding and dimerization domains in the
N-terminal region of p65 are represented by the boxed areas.
The locations of TA (filled box) and TA are shown in the C-terminal part of the molecule. The homology
region within TA is designated TA` and is shown
by the shaded box. The position of a potential leucine zipper
is indicated by three repeats of L. The two sequences are
aligned in the lower part of the figure. Identical positions
are shown by solid bars, the conserved hydrophobic amino acids
by dashed lines. The asterisk marks the C-terminal
end of the p65 protein. B, TA and TA` are highly conserved in evolution. The respective homology
regions were compared between human, mouse, and Xenopus p65
proteins. The sequence positions of the amino acids are indicated. The
derived consensus sequence is shown at the bottom of the
figure. Helix-breaking proline and glycine residues are underlined. The Y stands for any hydrophobic amino
acid. C, helical wheel plot of TA and TA` sequences. The amino acids displayed in the innermost circle are from the TA region of human and mouse. The next circles show the TA sequence from Xenopus, the TA` sequence from human and mouse,
and the outermost circle the TA` sequence from Xenopus. Identical or conserved positions are marked. Serines,
hydrophobic, acidic, and other amino acids are highlighted by different
shadings.
The wheel plot revealed a striking
degree of similarity with respect to the spatial organization of amino
acid residues (Fig. 2C). Five positions were strictly
hydrophobic. Two of them had a highly conserved phenylalanine and
leucine residue, respectively. The hydrophobic positions are
interrupted by hydrophilic residues, many of which are acidic. Only one
position in the plot was strictly acidic and would, in the primary
structure, directly precede the highly conserved phenylalanine. We have
previously described that Asp/Glu-Phe dipeptides are a key feature of
acidic activation domains (Schmitz et al., 1994). Basic amino
acids are not present. Another striking feature was the clustering of
serine residues opposite the helix surface with the mixed acidic and
hydrophobic positions. These structural characteristics support the
notion that an -helical conformation of TA and
TA (TA`) sequences is of functional importance.
TA`
The functional role of the
TA`Is Necessary but Not Sufficient for
the Activity of TA sequence for the activity of TA was
investigated by constructing a series of p65-mutants in which the
C-terminal portion of p65 was linked to the DNA binding domain of Gal4
(see Fig. 3). These constructs were co-transfected with a
Gal4-dependent reporter construct into COS cells and tested for their
ability to stimulate transcription. As shown in Fig. 3, the
construct Gal4-p65, which contains the entire
TA
region, resulted in a 96-fold increased rate of
transcription compared to Gal4 alone. This construct was nearly as
active as the entire C terminus of p65, which is represented by the
construct Gal4-p65 (see Fig. 3). Any
deletion of sequences from the C terminus affecting the integrity of
TA`
led to a drastic decrease in the ability of TA to stimulate transcription. A comparison between mutants
Gal4-p65 and Gal4-p65
showed that only the longer
version which retains TA
had a significant transactivating
capacity. However, construct Gal4-p65, which
contains the TA`
without any flanking sequences, showed no
significant transactivation potential, suggesting it was not as
independently active as TA. The addition of either C- or
N-terminally adjacent sequences to the TA` region in the
constructs Gal4-p65 and
Gal4-p65
increased the transcriptional
potential of the TA
region, but failed to fully restore the
activity of TA. Only the construct
Gal4-p65, containing C- as well as N-terminal
flanking sequences, restored the full transcriptional activity of
TA
. These data show that TA` is necessary, but
not sufficient for proper TA function.
Figure 3:
Deletion analysis of the TA domain. The indicated sequences of p65 were fused to the DNA
binding region of Gal4 (amino acids 1-147), which is shown as a striped box. The symbols used in this figure are
identical with those used in Fig. 2A. The sequence positions
of the fused p65 sequences, which are shown as boxed areas,
are given on the left. The constructs were co-transfected with
a Gal4-dependent reporter gene into COS cells and tested for their
ability to stimulate transcription. The level of activation of the
Gal4-p65 fusion protein was directly compared to the activity of Gal4
alone, which was given the arbitrary value of 1. The values are
averaged from more than four independent experiments. The standard
deviations did not exceed 15%. All constructs were verified by
sequencing and found to be correctly expressed, as revealed by EMSA
analysis of nuclear extracts.
TA
The finding that TAand TA`Contain Two
Copies of a Heptad Repeat with Copy-dependent Transcriptional
Activity
as well as
TA` are required for the transactivating activity of p65
prompted us to dissect their primary structure in more detail. As
depicted in Fig. 4A, both TA and TA` contain an internally repeated sequence as shown for human
TA (Fig. 4A, top). TA contains two copies of a repeat element with the consensus
sequence DXDFSX (
=
hydrophobic, X = any amino acid). The internal repeats
from the homology regions of TA
and TA` of
human, mouse, and Xenopus p65 proteins are listed in Fig. 4A by decreasing fulfillment of the consensus
sequence DFSSLLS. In TA and TA`, the repeats
are separated by one nonconserved amino acid residue. The ability of
the sequence motif DFSSLLS to activate transcription was tested by
fusing it in various copy numbers with a serine residue as spacer to
the DNA binding domain of Gal4. Such constructs were tested for their
transcriptional activity upon co-transfection with a Gal4-dependent CAT
reporter gene in COS cells. As shown in Fig. 4B, the
fusion of the sequence SSDFS to Gal4 gave rise to a transactivation not
exceeding that of Gal4 alone. Likewise, the sequence LLSSDFSSLLSSDFSSL,
which comprises one complete repeat, failed to transactivate
significantly. However, 4 copies fused to Gal4 gave a fusion protein
with strong transcriptional activity. A construct bearing 6 copies was
not significantly stronger in transcriptional activation. These results
show that the repeats have the potential to form strong transactivators
when multimerized, suggesting that they act in synergy.
Figure 4:
Analysis of a redundant repeat motif
within TA and TA` of p65. A, alignment
of heptad repeats from TA and TA`. The top
part of the figure highlights the two repeated motifs I and II in
human p65 TA by bars. The lower part of the figure
aligns heptad repeats from TA and TA` domains
of different species. The repeats are ordered by decreasing similarity.
A first order homology is in a dark gray box, a second order
homology in a light gray box. Repeat II of Xenopus TA showed the best fulfillment of the consensus
sequence and was chosen for multimerization experiments. B,
multimers of a consensus heptapeptide repeat display strong
transactivating activity. The indicated Gal4 effector plasmids were
transfected into COS cells together with a Gal4-dependent reporter
plasmid and assayed for their ability to stimulate transcription. The
fused sequences of the effector plasmids are shown. Transactivation
seen with Gal4 alone and with the respective fusion proteins is
represented by the columns. Transcriptional activity is given as
percent conversion of [C]chloramphenicol. The
standard deviation is indicated by a bar and was obtained from
5 independent experiments.
The Activity of TA
We tested whether the activity of the transactivating C
terminus of p65 was responsive to stimulation with PMA. COS cells were
co-transfected with the construct Gal4-p65Can Be Stimulated by
PMA and
a Gal4-dependent reporter plasmid and treated with PMA. In
PMA-stimulated cells, the transcriptional activity of the fusion
protein was increased by a factor of 5 compared to control cells (Fig. 5A, lane 2). In order to delineate the
region within the transactivating C terminus of p65 responsive to PMA,
various other constructs were tested in COS cells. The activity of the
TA
domain could be stimulated weakly, not exceeding a
factor of 2 (Fig. 5A, lane 3), while the
construct Gal4-p65 (Fig. 5A, lane 4) was more strongly responsive to stimulation of cells
with PMA. The transcriptional activity of Gal4 alone (Fig. 5A, lane 1), Gal4-p65
(data not shown), and Gal4-p65
(Fig. 5A, lane 5) were completely
unaltered after stimulation with phorbol ester. These results define
the region between amino acids 442 and 470 of p65 as necessary for the
strong stimulatory effect of PMA on p65-dependent transcriptional
activity. This region contains part of the mini-leucine zipper and
repeat element I from the TA`
region. EMSAs showed that PMA
treatment did not alter the DNA binding activity of
Gal4-p65 and Gal4-p65
(Fig. 5B), indicating that PMA treatment affected
transactivation by p65 but not the DNA binding activity of the Gal4
fusion proteins. The amount of complexes formed between Gal4 protein
and a
P-labeled oligonucleotide containing a Gal4 binding
site was very similar after treatment of cells with PMA.
Figure 5:
The activity of TA can be
stimulated by PMA. A, mapping of the PMA-responsive region.
Different plasmids encoding various Gal4-p65 fusion proteins were
co-transfected with a Gal4-dependent reporter plasmid into COS cells.
Cells were split after transfection onto two dishes, one of which was
stimulated with PMA. The left panel shows the results of
representative CAT assays for the constructs displayed in the right
half of this figure. Treatment of cells with PMA is indicated with
+ and -. The positions of nonacetylated and acetylated forms
of [C]chloramphenicol are shown. The amount of
protein assayed for CAT activity was chosen individually for each
tested construct in order to obtain CAT conversion rates within the
linear range. B, analysis of the DNA binding activity of two
Gal4 fusion proteins. Plasmids encoding either
Gal4-p65 or Gal4-p65
were transfected into COS cells. The cells were further treated
as described for the experiments in A of this figure and
assayed for DNA binding activity. EMSAs were performed by incubating
equal amounts of protein with a
P-labeled oligonucleotide
containing a Gal4 binding site. The arrows point to specific
DNA
protein complexes, the open triangle to the unbound
DNA probe. C, a protein kinase C inhibitor abrogates the
stimulatory effect of PMA on transactivation by
Gal4-p65. COS cells were co-transfected with a
Gal4-dependent reporter construct and Gal4-p65
.
After transfection, the cells were stimulated with PMA in the presence
of the indicated amounts of the specific protein kinase C inhibitor
bisindolylmaleimide. The results from a representative CAT assay are
shown.
Since it is
known that PMA stimulates the activity of protein kinase C, the effects
of the protein kinase C inhibitor bisindolylmaleimide on PMA-activated
transcription by Gal4-p65 were tested. COS
cells co-transfected with a Gal4-dependent reporter construct and
Gal4-p65
were treated simultaneously with 20
ng/ml PMA and various concentrations of bisindolylmaleimide. As evident
from Fig. 5C, the inhibition of protein kinase C
activity with 3 µM bisindolylmaleimide completely
inhibited the stimulatory effect of PMA on transcription.
PMA Enhances Phosphorylation of TA
We investigated whether the phorbol
ester-inducible activation of TAin
Intact Cells correlates with an
inducible phosphorylation of TA in response to PMA. COS
cells were transiently transfected with a plasmid encoding either the
DNA binding domain of Gal4 alone or fusion proteins between Gal4 and
TA. Immunoprecipitation of Gal4-p65 revealed that this protein was phosphorylated in intact cells
irrespective of PMA treatment of cells (Fig. 6A, compare lanes 1 and 2). The Gal4 protein alone showed no
significant phosphorylation on its own, suggesting that phosphate was
incorporated exclusively into the TA
portion of the fusion
protein. The correct expression of Gal4 and Gal4-p65 was verified by Western blotting (Fig. 6A, lanes 5-8). Similar experiments were performed with
Gal4-p65
. As is apparent in Fig. 6B (lanes 1 and 2), the constitutive
phosphorylation of the Gal4-p65
fusion protein
was enhanced after treatment of cells with PMA. Western blotting
demonstrated the correct expression of Gal4-p65
to similar levels (Fig. 6B, lanes 3 and 4).
Figure 6:
Both
TADs of p65 are phosphorylated in vivo. A, TA is constitutively phosphorylated in vivo. An
autoradiogram of immunoprecipitates using -Gal4 antibody is shown
in lanes 1-4. Lanes 1 and 2 show the
precipitation of Gal4-p65, lanes 3 and 4 the precipitation of Gal4. The presence of PMA is indicated
with + or -. Arrows mark the positions of specific
bands. Lanes 5-8 show a Western blot after incubation of
the filters with an
-Gal4 antibody and detection of bands with the
ECL system. Lanes 5 and 6 show the expression of
Gal4-p65; lanes 7 and 8 show
the expression of Gal4. The molecular masses of prestained standards
are shown in kilodaltons. For experimental details, see text. B, the PMA-responsive region of the TA
domain is
inducibly phosphorylated in vivo by phorbol ester. The
autoradiogram of the immunoprecipitations (lanes 1 and 2) shows the bands obtained by precipitation of
Gal4-p65. The Western blot (lanes 3 and 4) shows the correct expression of the proteins. The
details of the figure legends are as explained in A of this
figure.
The Structure of TA
In this study,
we have further analyzed the transactivating C-terminal portion of the
p65 (RelA) DNA-binding subunit of transcription factor NF-B. While
the sequence requirements of TA in the 30 C-terminal
residues of p65 have been studied in some detail (Schmitz et
al., 1994; Blair et al., 1994), it was not known how the
remaining transactivating sequences in p65 (referred to as
TA) are structured and how they are related to
TA. Here we show by a fine mapping analysis that TA cannot be further subdivided into autonomous transactivation
domains without significant loss of activity. These results are not
consistent with a previous study where the TA domain could
be subdivided into a mini-leucine zipper region and a more C-terminal
region, both of which displayed independent transcriptional activity
(Moore et al., 1993). However, it remains to be clarified
whether the different data are really conflicting, since the previous
study did not numerically quantify the results of the transactivation
assays. As shown here, Gal4 fusion proteins containing the region
between the leucine zipper and TA are not highly active on
their own. Only the entire TA domain has a strong
transactivating potential similar to that of TA. In
conclusion, the transactivating C terminus of p65 is highly redundant.
It harbors two strong and fully independent TA domains. Deletion of
either TA or TA leaves a strongly
transactivating p65 C terminus.
, TA and the viral activator VP16 compete for
the same co-activator molecules. This suggests that both TA and TA belong to the same class of acidic TA domains
as VP16. We have previously reported (Schmitz et al., 1994)
that TA and VP16 share sequence homology apart from
containing a high percentage of acidic and hydrophobic amino acid
residues. This finding prompted us to investigate also whether TA has a subdomain with sequence homology to TA (and
VP16). In fact, such a homology region was present in TA and is referred to as TA`. Comparison of p65 proteins
from different vertebrate species showed that TA` as well
as TA domains are highly conserved allowing us to decipher
a minimal consensus motif. The TA` subdomain in TA was essential for the activity of TA since its
deletion strongly reduced the transactivating potential of
TA. In contrast to TA, however, it was not
independently active when fused to Gal4 but required flanking sequences
for full activity.
-helix breaking proline and glycine residues suggested that
the TA and TA` sequences form -helices. A
CD and NMR analysis of TA showed a random coil structure
(Schmitz et al., 1994). In the presence of hydrophobic
solvents, TA adopted the expected -helical content.
For the basic region of leucine zipper proteins, it was shown that
-helices could only form if bound to DNA (Weiss et al.,
1990). In analogy, we assume that acidic sequences may form an
-helix only upon contacting target molecules by an ``induced
fit'' mechanism. If TA and TA` sequences
from different species are plotted on a helical wheel, a striking
similarity is observed. Highly conserved positions with hydrophobic and
acidic residues are found opposite of a helical surface which is rich
in serine residues. This finding provides the basis for a further
mutational analysis of these domains. An intriguing finding was that
TA and TA` are each composed of two heptad
repeat subdomains. These subdomains have recently been suggested to be
necessary for proper functioning of the TA domain (Blair et al., 1994; Schmitz et al., 1994).
and
TA` acted synergistically, which is supported by a point
mutation analysis of the TA repeats (Schmitz et
al., 1994). We anticipate that each of the consensus repeats
contacts a specific site on a target molecule. Only duplication makes
the interaction between the repeats and their target sites strong
enough to promote transcription. This hypothesis fits the
``simultaneous contact model'' proposed by Lin et al. (1990) and Carey et al.(1990). This model suggests that
the repeats contact multiple, redundant sites within one molecule.
Alternatively, each repeat could contact a separate molecule.
revealed that
the repeats have to be properly spaced in order to optimally synergize
(Schmitz et al., 1994). Insertion or deletion of a single
amino acid in between the two repeats reduced the activity of TA by 90%. Likewise, introduction of a proline residue strongly
interfered with TA activity, suggesting secondary structure
constraints. The findings on TA, TA`, and VP16
indicate that acidic activators are not only characterized by a
specific amino acid composition but also by a defined primary and
secondary structure. Frequently, acidic activators contain Asp/Glu-Phe
dipeptides flanked by additional acidic and hydrophobic residues
(Schmitz et al., 1994; Tjian and Maniatis, 1994). In the case
of TA and TA`, a high percentage of serine
residues is seen. It appears that this hydroxyamino acid can
functionally substitute for acidic residues. This is also evident from
the artificial activators containing 4 and 6 copies of an idealized
heptad repeat: they contain only one negatively charged amino acid per
repeat. It is possible that in vivo additional negative charge
is introduced by phosphorylation of serine residues which would allow
us to modulate the transactivating potential of serine-rich acidic
domains by protein kinases.
Phorbol Ester-dependent Activity of
TA
Numerous transcription factors whose activity is
regulated by external stimuli were found to contain inducible TADs.
Enhancement of transcription involves in many cases phosphorylation of
the TADs (reviewed by Jackson(1992) and Karin(1994)). A well-studied
example is the phorbol ester-inducible activation domain in the N
terminus of c-Jun. The enhanced activity of this TAD was found to
correlate with PMA-induced phosphorylation of serines 63 and 73
(Franklin et al., 1992). These residues were shown to be
phosphorylated by JNK1, a UV-inducible protein kinase (Dérijard et al., 1994). Another well defined example is the
transcription factor CREB, which is involved in the cAMP-dependent
signaling. Here, transcriptional activity is stimulated by
cAMP-dependent protein kinase A which phosphorylates Ser-133 or,
alternatively, by Ca-calmodulin-dependent protein
kinases I and II (reviewed by Jackson, 1992). A positive
transcriptional control by TAD phosphorylation is also seen with the
SRF accessory protein Elk-1 (Marais et al., 1993) and
transcription factor CREM (deGroot et al., 1993).
domain after stimulation
with PMA correlates well with an increased phosphorylation of this
region, suggesting that PMA-dependent phosphorylation enhances the
transactivating potential of TA. The PMA-inducible region
between amino acids 442 and 470 contains several potential
phosphorylation sites (serines 457 and 468 and tyrosines 458 and 464).
Future studies have to reveal whether PMA-activated protein kinase C is
directly phosphorylating TA on serine or is initiating a
signaling cascade involving other kinases. Another candidate kinase is
casein kinase II (CKII), because there are several recognition
sequences for this kinase in TA and TA.
However, in vitro phosphorylation experiments with purified
CKII and a bacterially expressed protein containing the C-terminal 123
amino acids of p65 failed to show phosphorylation by CKII. The p65
fragment containing both TADs was only phosphorylated up to 5%, whereas
a control protein was quantitatively phosphorylated (data not shown).
Another candidate is an uncharacterized serine/threonine kinase, which
has been found associated with NF-B p65 (Ostrowski et
al., 1991). This kinase might also be responsible for the
increased phosphorylation of p65 seen after stimulation of HeLa cells
with tumor necrosis factor- (Naumann and Scheidereit, 1994).
activity was only weakly responsive to phorbol ester
stimulation, but was more highly phosphorylated than TA under control conditions. Because TA is devoid of
threonine and tyrosine residues, it is most likely phosphorylated on
serine residues within the conserved repeat domain. The conserved
clustering of serines on one side of the potential -helix would
restrict the increase in negative charge by phosphorylation to one
helical surface. Since TA belongs to the class of acidic
activators, additional negative charge introduced by phosphate groups
may increase its transcriptional activity. Likewise, a controlled
dephosphorylation could decrease the activity of TA.
and TA superimposes a second level of
regulation. The first one being binding of NF-B to its inhibitor
IB in the cytoplasm. Phosphorylation of TA is likely
to occur in the nucleus because it is observed with the Gal4 fusion
protein, which does not undergo a phase of cytoplasmic storage as
NF-B. Future studies are required to identify the sites of
phosphorylation in TA and TA and the
responsible kinases and phosphatases.
B, nuclear factor B; TAD, transactivation domain;
PMA, phorbol 12-myristate 13-acetate; CAT, chloramphenicol
acetyltransferase; EMSA, electrophoretic mobility shift assay.
-Gal4
antibodies, and Dr. Kathy Tamai for helpful comments on the manuscript.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Choudhary, M. Lu, R. Cui, and A. R. Brasier Involvement of a Novel Rac/RhoA Guanosine Triphosphatase-Nuclear Factor-{kappa}B Inducing Kinase Signaling Pathway Mediating Angiotensin II-Induced RelA Transactivation Mol. Endocrinol., September 1, 2007; 21(9): 2203 - 2217. |
