| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 275, Issue 26, 19461-19468, June 30, 2000
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,
From the Departament de Biologia Molecular i Cel·lular, Institut
de Biologia Molecular de Barcelona, Centre d'Investigació i
Desenvolupament-Consejo Superior de Investigaciones
Científicas, Jordi Girona 18
26, 08034 Barcelona, Spain
Received for publication, February 7, 2000, and in revised form, March 20, 2000
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
GAGA is a nuclear protein encoded by the
Trithorax-like gene in Drosophila that is
expressed in at least two isoforms generated by alternative splicing.
By means of its specific interaction with DNA, GAGA has been involved
in several nuclear transactions including regulation of gene
expression. Here we have studied the GAGA519 isoform as a
transcription factor. In vitro, the transactivation domain
has been assigned to the 93 C-terminal residues that correspond to a
glutamine-rich domain (Q-domain). It presents an internal modular
structure and acts independently of the rest of the protein. In
vivo, in Drosophila SL2 cells, Q-domain can
transactivate reporter genes either in the form of GAGA or Gal4BD-Q
fusions, whereas a GAGA mutant deleted of the Q-domain cannot. Our
results give support to the notion that GAGA can function as a
transcription activating factor.
Transcription of the homeotic and segmentation genes is a highly
regulated process in Drosophila in which many different
factors exert positive and negative effects. Some of those genes,
including engrailed (en), Ultrabithorax (Ubx), even-skipped
(eve), fushi tarazu (ftz), and Krüppel (Kr),
characteristically present multiple binding sites clustered in their
promoters (with a consensus sequence GAGAG) for a factor named GAGA.
The expression of Ubx, en, and ftz has
been shown to be clearly regulated by GAGA in vivo (1, 2).
This is also likely for the other genes listed above. The promoters of
some of the heat shock protein gene family also contain GAGA binding
sites (some of them remarkably long) and are also expected to be under
GAGA factor regulation. GAGA is a sequence-specific transcription
factor organized in several domains. At the N terminus there is a
BTB/POZ domain (POZ), which is required for oligomerization (3, 4). In
a central position, a single Zn2+ finger surrounded by
three short basic regions conforms the DNA binding region
(DBD)1 that specifically
recognizes the GAGA binding sites (5, 6). At the C terminus there is a
glutamine-rich domain (herein referred as Q-domain), still without a
defined function. It was speculated to be a transactivation domain
because its composition resembles that of the glutamine-rich family of
transactivation domains (7). More recently, however, functions on DNA
distortion and protein multimerization have been attributed to the
Q-domain (8, 9).
GAGA is encoded by the essential Trithorax-like
(Trl) gene and is of maternal effect (1). Expression of
Trl gives rise to at least two different proteins generated
by alternative splicing. These two isoforms (GAGA519 and
GAGA581) share the initial 381 residues and only diverge at
the C terminus, where both proteins still present a similar
glutamine-rich domain but of different lengths (10). At the genetic
level, Trl mutants down-regulate the expression of homeotic
genes, as for instance Ubx (1), and some of the mutants
display an enhancement in position effect variegation. On the other
hand, transient transfection experiments demonstrated a stimulation of
transcription only from reporters containing GAGA binding sites
in vivo (10, 11). In vitro, GAGA was shown to
stimulate transcription from several promoters only when bearing GAGA
binding sites (12, 13). However, the addition of GAGA to some fly
embryo extracts did not result in an increase in transcriptional
activity. This fact, along with the lack of net activation observed
with hsp promoters, prompted the suggestion that GAGA was
acting as an antirepressor (14, 15).
GAGA was also shown to have an effect in remodeling the chromatin
structure of the hsp70, hsp26, and ftz
promoters (16-18). However, this nucleosome remodeling does not appear
to be tightly associated to the presence of GAGA, since it can take
place with Gal4BD alone on synthetic promoters carrying Gal4 binding
sites. Moreover, if Gal4BD is fused to a transactivation domain, then stimulation can occur, thus separating both processes (19).
Here we report the biochemical characterization of the GAGA domains
functional in transcription. Our results indicate that deletion of the
glutamine-rich C-terminal domain (Q-domain) abolishes the
transcriptional activity in vitro and in vivo.
The Q-domain presents a modular structure in which the glutamine
residues are dispensable for the stimulatory activity. The
transcriptional activity of GAGA is reduced but not abolished by
deletion of the POZ domain in vitro.
DNA Constructions and Protein Expression and
Purification--
Constructs for recombinant protein expression in
Escherichia coli were cloned as His6-tagged
fusions into pET14b expression vector (Novagen). Constructs expressing
full-length GAGA,
Gal4BD was subcloned into pET14b using a polymerase chain reaction.
Gal4BD-Q was obtained by inserting an EcoRV-EcoRV
fragment from pET14b-GAGA at the SmaI site of the
pET14b-Gal4BD construct. Constructs Gal4BD-Q
Expression and purification of His6-tagged recombinant
proteins was carried out in E. coli BL21(DE3) strains
essentially as described before (20, 21). The Gal4-VP16 fusion was
expressed and purified as described (22).
In Vitro Transcription Assays--
In vitro
transcription assays using unfractionated HeLa cell nuclear extracts
contained 30-50 µg of protein extract, variable amounts of
recombinant proteins, and 200 ng of supercoiled DNA template. Templates
used contained either a CT22 sequence or 8 Gal4 binding
sites cloned immediately upstream of a minimal TATA box and fused to a
G-less cassette. Reactions were in a final volume of 25 µl and
contained 3.5 mM MgCl2, 0.8 mM
3'-O-methyl-GTP (Amersham Pharmacia Biotech), 0.4 mM ATP, 0.4 mM CTP, 20 µCi of [ Transient Cell Transfections--
Drosophila
expression plasmid Act5CPPA (kindly provided by G. Jiménez,
Institut de Biologia Molecular de Barcelona (IBMB)) was used to
subclone in its polylinker region GAGA, GAGA-
Drosophila S2 cells were grown in Schneider's insect medium
(Sigma) containing 10% fetal calf serum (Life Technologies, Inc.) and
gentamycin. Cells were kept at 1-8 × 106 cells/ml.
For transfection, 2-3 × 106 cells in 5 ml of medium
were plated onto 6-cm diameter tissue culture dishes and allowed to
stand for 24 h at 25 °C. Cells were then transfected using the
calcium phosphate technique as described (25) using 0-10 µg of test
constructs, 5 µg of reporter constructs, 7 µg of pUC19, and 0.5 µg of cytomegalovirus-luciferase construct. The final volume and the
total amount of DNA was kept constant at 20 µg by the addition of the
Act5PPA vector and water required in each case. After incubation with
precipitates for 48 h at 25 °C, cells were lysed, and
The Q-domain of GAGA Is Necessary for the Transcriptional Activity
in Vitro--
The study of GAGA as a transcription factor was
initially carried out in an heterologous system using nuclear extracts
from HeLa cells in vitro. The rationale behind this was to
avoid the presence of GAGA in crude fly embryo nuclear extracts. HeLa
cells provided a convenient system because the general transcription machinery is rather conserved between human and Drosophila
and also because GAGA is not expected to exist in human cells. In fact,
we have previously shown that there are no proteins in HeLa nuclear
extracts that can either footprint or stimulate transcription from
templates containing GAGA binding sites (3, 20). With this approach we
tested the activities of several deletion mutants, as outlined in Fig.
1A using a template DNA
containing a G-less cassette fused to a promoter that contained a
minimal TATA box (derived from the AdML promoter) and a
d(CT)22 sequence, which acted as a GAGA binding sequence,
inserted shortly upstream of the TATA box (at position Q-domain Is a Modular Transactivation Domain--
To study the
contribution of the Q-domain independently of the other domains of GAGA
protein, fusions to the Gal4 DNA binding domain were prepared. This
approach also allowed us to determine whether Q was a separable domain
that could transactivate from heterologous DNA binding sites, as
characteristically described for classical transactivation domains (22,
26). In vitro transcription experiments were carried out in
HeLa cell nuclear extracts and used a template DNA containing a
G-less cassette fused to an artificial promoter that contained a
minimal artificial TATA box and eight tandemly repeated binding sites
for Gal4 (inserted upstream of position
Similar results were also obtained when Gal4BD-Q was assayed in a
Drosophila embryonic nuclear extract using exactly the same template DNA (Fig. 2C). Titration of recombinant Gal4BD-Q in
Drosophila SNF extracts resulted in a ~12-fold maximal
increase in transcription (Fig. 2C, lane 3).
Gal4BD-VP16 used as a control stimulated transcription up to ~38-fold
(Fig. 2C, lanes 7-9). In these extracts,
however, titration of GAGA factor did not produce any increase in
transcription rates (data not shown), in agreement with previously
published results (14).
As it can be appreciated in Fig. 2A, Q-domain can be roughly
subdivided into three regions according to the presence of long runs of
glutamine residues. Mutants -Q
Similar results were obtained when equivalent mutants in the GAGA
context were assayed in HeLa nuclear extract using the template containing the d(CT)22 sequence described above. Fig.
3 shows that in the GAGA context,
deletion of most of the Q-domain except of the 18 N-terminal-most
residues (GAGA- GAGA Transcriptional Activity Depends on Q-domain in
Vivo--
GAGA was shown earlier to be able to stimulate transcription
in vivo in transiently transfected Drosophila SL2
cells (10, 11). Here, the requirement of the Q-domain for this
transcriptional activity was studied in vivo by transient
transfection of constructs expressing Gal4BD-Q, GAGA, and GAGA-
Initially, SL2 cells were transiently transfected with constructs
expressing Gal4BD-Q, and Gal4BD-VP16 and Gal4BD as positive and
negative controls, respectively, under the control of the constitutive
actin5 promoter, and their activities were assayed. The
results showed (Fig. 4A) that
Gal4BD-Q stimulated transcription from the 5×Gal4-
In a second set of experiments, GAGA and GAGA-
All these results indicate that the transcription stimulatory activity
of GAGA depends on the Q-domain, which can act independently of the
rest of the protein in vitro and in vivo.
From the knowledge accumulated to date, GAGA seems to be a
multi-purpose factor that is involved in many different nuclear processes, including gene regulation, chromatin remodeling, and nuclear
division (2). GAGA is transcribed from the Trl gene into at
least two different mRNAs generated by alternative splicing. They
give rise to GAGA519 and GAGA581 forms, which
only differ at the C-terminal domain. In both cases there is a
glutamine-rich domain, but the exact composition and length are different.
Initial studies showed that GAGA could stimulate transcription in
Drosophila both in vitro and in vivo
(11, 12, 14, 28). Here we have shown that this stimulatory activity
depends on the C-terminal Q-domain both in vitro and
in vivo.
On an empirical basis, we have defined Q-domain of GAGA519
as comprising residues 426 to 519, since its deletion causes a major drop in the transcriptional activity of GAGA- Q-domain has been shown to work independently of the rest of the GAGA
protein, since it also works very efficiently in Gal4BD fusions. In
this context, Q-domain has been dissected further and, resembling other
classical transactivation domains like those of VP16 or Sp1, for
instance, has been shown to possess an internal structure. We observe
the presence of three different regions that seem to work
synergistically, although none of them seems to be absolutely required.
This is indicated by the results obtained with the deletion mutants.
Those mutants were essentially designed to remove sequence blocks from
glutamine run to glutamine run. In this way, Homopolymeric stretches of glutamines or prolines fused to Gal4BD can
stimulate transcription when they include 10 or more uninterrupted
residues (29). The results obtained with mutant For the sequence analysis, and as mentioned above, Q-domain was
subdivided into three regions of unique sequence, proximal, intermediate, and distal, separated by glutamine runs. Sequence comparisons of these three regions by pairs detected very weak homologies. Similarly, none of those sequences seem to have any clear
homologue in other glutamine-rich transcription factors (not shown).
However, when the four GAGA factors reported so far (i.e.
Drosophila melanogaster 519 and 581 and
Drosophila virilis GAGA-A and -B) are compared,
reasonably conserved homologies for distal and proximal regions appear
but are less apparent for the intermediate region (Fig.
5). In addition, homologies to the distal region always appear at the C-terminal part of the domain, whereas homologies to the proximal region always appear to the N-terminal part
of the domain. This colinearity is lost for the intermediate region.
All these GAGA proteins show a strong sequence conservation at the
N-terminal portion, and although they differ at the glutamine-rich C-terminal domain, they also show a similar overall structure and
composition (10, 30). The fact that the variable region has partially
conserved the relevant motifs described above suggests that the other
members of this group could also retain some transcriptional activation
function.
INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
MATERIALS AND METHODS
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
POZ, and DBDGAGA have been recently
described (3, 20). Construct pET14b-GAGA-Q lacking the Q-rich
C-terminal domain was obtained by inserting an
NdeI-EcoRV fragment from the pET14b-GAGA
into NdeI and BamHI (filled-in with Klenow
enzyme) sites of pET14b. pET14b-GAGA-3 was obtained by replacing the
SfiI-HindIII fragment of pET14b-GAGA by the
SfiI-HindIII fragment from pET14b-Gal4Q. pET14b-GAGA-4 was constructed by inserting at the EcoRV
site in the GAGA sequence a blunt-ended DNA fragment generated by
polymerase chain reaction and spanning Pro-445 to Gln-519
residues of GAGA. pET14b-GAGAQ-VP16 was constructed by ligation of
the VP16 transactivation domain obtained by polymerase chain reaction
from pJL2 to the pET14b-GAGA-Q.
1, -Q2, -Q3, and
-Q4 were obtained by polymerase chain reaction from the Gal4BD-Q
construct and are described in Fig. 3A. All constructs were
verified by DNA sequencing. Plasmid pJL2 expressing Gal4-VP16 fusion
was kindly provided by M. Carey and M. Ptashne (Harvard University).
Constructs AdML50[180], (CT)22-AdML50[180], WT[180],
and 8×Gal4-WT[180] have been previously described (20, 21).
-32P]UTP (800 Ci/mmol) adjusted to 1 µM
with cold UTP, 3 units of RNase T1, 1.5 mM dithiothreitol,
2% polyethylene glycol 8000, 0.2 mM EGTA, 33 mM HEPES-KOH, pH 7.9; nuclear extract was added up to 10 µl in D-buffer (0.1 M KCl, 0.2 mM EDTA, 0.5 mM dithiothreitol, 0.1 mM phenylmethylsulfonyl
fluoride, 20% glycerol, 20 mM HEPES, pH 7.9). Reactions
were typically incubated for 60 min at 30 °C and terminated by SDS
addition, phenol extraction, and ethanol precipitation as described
before (23). For the study of GAGA deletion mutants and since we
observed some variability from experiment to experiment, all
experiments were run alongside with GAGA titrations to correct for
potential deviations. Drosophila embryo nuclear extracts
(SNF) prepared as in Kamakaka et al. (24) were a generous gift from Peter Becker (EMBL). In vitro transcription assays
were performed as above except that reactions were allowed to proceed for 35 min at 26 °C. All the in vitro transcription
results were quantified from the corresponding autoradiographs using a
Molecular Dynamics laser microdensitometer.
Q, GAGA-POZ, GAGAQ-VP16, Gal4BD, Gal4BD-Q, and Gal4BD-VP16 constructions reported above. To study the effects of the Gal4-fusions, reporters
5×Gal4-hsp70TATA-
gal and hsp70TATA-gal were constructed by
inserting either 5×Gal4 binding sites-hsp70 TATA box (derived from
pUAST construct) and a hsp70 TATA box (derived from pWHL construct) in
the pgal-basic vector (CLONTECH), respectively.
pUAST and pWHL were kindly provided by S. González (IBMB). To
study the effects of GAGA mutants, fragments containing a
d(GA·TC)22 sequence were inserted in both orientations in
the hsp70 TATA box-gal construct described above, giving rise to the
plasmids named CT22-hsp70TATA-gal and
GA22-hsp70TATA-gal. A cytomegalovirus-luciferase
reporter plasmid was always used in co-transfection assays as an
internal control for transfection efficiency.
-galactosidase activity was measured using the -galactosidase
gene reporter kit (Roche Molecular Biochemicals). As an internal
control, luciferase activities were assayed using the luciferase
activity assay kit (Promega). -Galactosidase activities were
corrected with respect to luciferase activities to normalize for
transfection efficiency.
RESULTS
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
50 with
respect to the transcription start site). All experiments were done by
titrating the amount of recombinant protein added to a fixed amount of
nuclear extract and template DNA. Maximal activation rates observed
after normalization to recovery controls were considered. The maximal
activation rates obtained with full-length GAGA (GAGA in Fig. 1,
A and B) were around 15-fold with respect to the
control, which was taken as 100% activation; all other numbers are
referred to it and represent the average of at least three independent
experiments. In the presence of high amounts of GAGA, a significant
drop in transactivation was observed (Fig. 1B, compare
lanes 2-5 in the upper gel, and lanes 10-14 in
the lower gel). With this approach we assayed the effect of GAGA
mutants deleted in some domains. As controls, DBD(GAGA) and
POZ(GAGA), which only retained the binding domain and a
long version of the POZ domain, respectively, showed no activation at
all (Fig. 1, A and B). Deletion of POZ domain
generated a moderate reduction that resulted in a 76% of the maximal
activity (Fig. 1, A and B,
GAGA-POZ). A clear drop in transcription was
also observed at amounts of GAGA-POZ higher than GAGA and may
reflect squelching in both cases. Deletion of the Q-domain resulted in a much larger reduction but not a complete inactivation, and a residual
23% of maximal activity was obtained (Fig. 1, A and
B, GAGA-Q). DNase I footprinting
analysis showed that GAGA and GAGA-Q interact with similar intensity
and identical specificity to the d(GA·TC)22 sequence,
thus ruling out any potential impairment in its binding to the template
DNA (results not shown). Therefore, the clear drop in activity can only
be attributed to a lack in the transcriptional activity of the
GAGA-Q mutant. GAGA-POZ capabilities to bind the template DNA
were previously studied and shown to be of the same specificity and
affinity as intact GAGA (3). From these results, it can be concluded
that Q-domain has a major contribution to the observed transcriptional
activity.
View larger version (31K):
[in a new window]
Fig. 1.
The C-terminal glutamine-rich domain is the
transactivation domain of GAGA. A, scheme of the GAGA
mutants assayed (on the left) and results of activation (on the right)
expressed as a fraction of the maximal activation obtained with
wild-type GAGA. B, transcriptional activity of the GAGA
mutants in an in vitro transcription assay using nuclear
HeLa cell extracts. Upper panel: lanes 1,
6, and 12 are controls for basal transcription;
lanes 2-5 received 11.25, 22.5, 45, and 90 ng of GAGA,
respectively; lanes 7-11 received 4, 8, 32, 56, and 80 ng
of POZ(GAGA), respectively; lanes 13-16
received 2.5, 5, 20, and 35 ng of DBD(GAGA), respectively.
Lower panel: lane 1 is a control for basal
transcription; lanes 2-5 received 6.25, 12.5, 25, and 43.75 ng of GAGA- Q, respectively; lanes 6-9 received 20, 40, 80, and 140 ng of GAGA-POZ, respectively; lanes 10-14
received 2.25, 11.25, 22.5, 45, and 90 ng of GAGA, respectively. The
arrows indicate full-length transcripts; the
asterisks denotes a recovery control.
50). Reactions were
performed as described under "Material and Methods" and are shown
in Fig. 2. Gal4BD-Q showed a remarkably strong activity in these assays, reaching 20-25-fold activation from
promoters containing Gal4 binding sites (Fig. 2B,
Gal4BD-Q, compare lanes 1 in all three
panels with lanes 8 in the upper and
middle panels and lane 3 in the lower
panel); there was no activation at all when the same promoter
lacking the Gal4 binding sites was used (not shown). As a control
Gal4BD (comprising residues 1-147) was titrated and showed a marginal
activation as noted before (26).
View larger version (26K):
[in a new window]
Fig. 2.
The Q-domain of GAGA transactivates
efficiently when fused to Gal4BD and shows an internal modular
structure. A, fusion proteins to the Gal4BD (residues
1-147) assayed in B are indicated on the left. On the
right, transcriptional rates are expressed as the fraction of the
maximal activation obtained with the entire Q-domain. B,
increasing amounts of Gal4BD (10, 20, and 60 ng), Gal4BD-Q (10, 30, 50, and 100 ng in the upper and central panels, and
50, 100, and 150 ng in the lower panel), Gal4BD-Q 1 (4, 10, 20, 40, and 60 ng), Gal4BD-Q2 (2, 4, 10, 20, and 60 ng),
Gal4BD-Q3 (2, 4, 8, 20, and 40 ng), and Gal4BD-Q4 (3, 6, 12, 15, 21, and 36 ng) were assayed in HeLa nuclear extract as indicated.
Lanes 1 and 5 in all three panels did
not receive any of these proteins and represent basal, non-activated
transcription. C, increasing amounts of Gal4BD-Q (20, 50, 75, 100, and 300 ng (lanes 2-6, respectively)) and
Gal4BD-VP16 (2.5, 5, and 15 ng (lanes 7-9, respectively))
were assayed in a SNF Drosophila embryo extract. Lane
1 is a control for basal transcription and received no Gal4
protein. The arrows indicate full-length transcripts, the
single asterisk denotes a recovery control, and double
asterisk denotes an internal translabeled band coming from the
embryo extract.
1 to -Q3 were generated by
progressive deletion of the sequences between glutamine tracks as
depicted in Fig. 2A, and their transcriptional efficiencies were determined in transcription assays in vitro in HeLa
nuclear extracts using the same template DNA containing Gal4 binding
sequences. As above, quantification of the activity represents the
average of at least three independent experiments and is referred to
Gal4BD-Q maximal activity taken as 100%. The results showed that
Gal4BD-Q3 mutant, which only retained 18 residues from the Q-rich
domain and none of the long glutamine runs, was still remarkably active (60% of maximal). Shorter deletions (-Q2 and -Q1) showed
increased levels of activity, suggesting that several regions in the
Q-domain may contribute to stimulate transcription. This interpretation was further supported by the fact that the internal deletion of the 18 N-terminal residues of the Q-domain (-Q4) did not abolish transactivation but only reduced it to a 78% with respect to the entire Q-domain.
3) still retained more than 50% of the maximal
transcriptional activity, whereas the internal deletion of this
18-amino acid region (GAGA-4) retained more than 85% maximal
activation. Long glutamine runs do not seem to be directly required.
All these results indicated that Q-domain is an independent,
transportable transactivation domain with a modular architecture in
which the different regions cooperate to reach maximal activity.
View larger version (38K):
[in a new window]
Fig. 3.
GAGA Q-domain is an independent domain and
shows the same transactivation potential and internal modular
structure. A, scheme of the GAGA mutants assayed (on
the left) and results of activation (on the right), expressed as a
fraction of the maximal activation obtained with wild-type GAGA.
B, transcriptional activity of the GAGA mutants in an
in vitro transcription assay using nuclear HeLa cell
extracts. Upper panel, increasing amounts of GAGA (11.25, 22.5, and 45 ng, lanes 2-4, respectively) and GAGA- 3
(0.7, 1.25, 2.5, 5, 10, and 20 ng, lanes 6-11,
respectively) were assayed; lanes 1 and 5 did not
receive GAGA proteins and represent basal transcription. Lower
panel, increasing amounts of GAGA (5.63, 11.25, 22.5, and 45 ng,
lanes 2-5, respectively) and GAGA-4 (1, 2.5, 5, 10, 20, and 40 ng, lanes 7-12, respectively) were assayed;
lanes 1 and 6 did not receive GAGA proteins and
represent basal transcription. The arrows indicate
full-length transcripts; the asterisk denotes a recovery
control.
Q
proteins in SL2 cells.
galactosidase
reporter up to 8-fold (striped boxes), in good agreement
with the in vitro results presented above. Gal4BD-VP16
reached up to ~571-fold stimulation (cross-hatched boxes)
and appeared as a really strong activator in vivo, as
described (27). On the other hand, Gal4BD alone did not stimulate
transcription at all (gray boxes).
View larger version (25K):
[in a new window]
Fig. 4.
Q-domain can transactivate in vivo
in transiently transfected Drosophila SL2
cells. A, Gal4BD-Q stimulates transcription in
vivo. SL2 cells were transiently transfected with increasing
amounts of constructions expressing Gal4BD (dotted boxes),
Gal4BD-Q (hatched boxes), or Gal4BD-VP16
(cross-hatched boxes). After normalization to the luciferase
internal control, activation is plotted versus the amounts
of transfected DNA. A scheme of the -galactosidase reporter carrying
five Gal4 binding sites immediately upstream of a hsp70 TATA box is
shown below. B, GAGA deleted of the Q-domain cannot
transactivate in vivo. Increasing amounts of constructs
expressing GAGA (hatched boxes), GAGA-Q (stippled
boxes), and GAGA-Q-VP16 (cross-hatched boxes) were
assayed. After normalization to the luciferase internal control,
activation is plotted versus the amounts of transfected DNA.
A scheme of the -galactosidase reporter carrying a
(CT)22 sequence as GAGA binding site immediately upstream
of a hsp70 TATA box is shown below.
Q were similarly
assayed using a -galactosidase reporter bearing a (CT)22 sequence just upstream of the minimal hsp70 promoter. In
these experiments, a high background level, likely due to the presence of endogenous GAGA, was observed. The results showed that GAGA overexpression stimulated transcription up to ~2-fold (Fig.
4B, striped boxes). This stimulatory effect is
lower than the observed by Kornberg and co-workers (11) but similar to
the reported by Elgin and co-workers (10). In any case, GAGA-Q did
not stimulate transcription at all; instead, a repression of the
background level was observed (Fig. 4B, stippled
boxes). As a control for this negative result, a construct
expressing GAGAQ-VP16 was used and showed up to a ~10-fold
stimulation of transcription (Fig. 4B, cross-hatched
boxes). This result ruled out the possibility that the lack of
activity of GAGA-Q could be due to any incorrect intracellular
localization, since the VP16 domain used did not include any nuclear
localization signal. Additionally, these results also indicated that
the folding of the GAGA-Q moiety seemed to be correct in
vivo. Finally, GAGA stimulation of transcription was dependent on
the presence of GAGA binding sequences in the promoter and was
insensitive to heat-shock treatment of the transfected cells (results
not shown).
DISCUSSION
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
Q, and its fusion to
the Gal4 binding domain results in an active chimeric protein in
vitro and in vivo. GAGA-Q shows a residual 23% of
activity that might be due to the uncovering of a cryptic and weak
transactivation region. A similar effect has been observed for example
with the yeast Gal4 transcription factor (26). An extended C-terminal deletion up to residue 390 did not reduce it further (not shown). Deletion of the POZ domain has a relatively minor effect on
transcription in vitro. POZ domain has been previously shown
to be responsible for GAGA oligomerization in vitro (3, 4).
Moreover, GAGA-POZ binding to DNA was shown to be progressive and
non-cooperative, as a result of its monomeric state, whereas GAGA bound
cooperatively and was present in an oligomeric form. A reduced maximal
activation of GAGA-POZ was also noticed in those experiments using
different templates. These results were interpreted as a lack of
cooperativity among the GAGA-POZ molecules, which could not properly
reorganize the promoter and render a fully active complex (3, 4).
1 deleted the
C-terminal Q-rich stretch, 2 deleted up to the central Q-rich
stretch plus the G-run, and 3 deleted up to the most N-terminal
Q-stretch, leaving a peptide only 18-residues long. All of them
retained transactivation potentials ranging from 90 to 60%,
respectively. Mutant 4, in which the 18-residue-long peptide was
deleted while leaving intact the rest of the domain, showed the same
transactivating potential as mutant 2 (around 75%). The conclusion
is that the three regions seem to be functional on their own and work
in synergy with each other. Remarkably, the two most relevant mutants,
2 and 4, showed a very similar behavior in the GAGA protein
context, strongly suggesting that Q-domain also works, irrespective of
the rest of the protein, as a transactivating domain in GAGA.
3 suggest that at
least the glutamine runs do not seem to be the only activation motif in
the Q-domain. In any case, an uninterrupted stretch of 7 glutamine
residues (or 12 if we accept two interruptions) is the longest Q-run
that can be found in the GAGA activation domain.
View larger version (44K):
[in a new window]
Fig. 5.
Sequence alignment of the proximal,
intermediate, and distal regions of the Q- domain. The sequences
present in the Q-domain of D. melanogaster
GAGA519 in between the Q stretches were used to search for
homologies in the D. melanogaster GAGA581 and
D. virilis GAGA forms A and B. The sequences of
GAGA519 used for the alignment were residues Ile-427 to
Gln-446 for the proximal region, residues His-460 to Gln-485 for the
intermediate region, and Ala-491 to Gln-512 for the distal region.
Homologies were studied to the entire region immediately C-terminal to
the DNA binding domain of the other three proteins by using the
programs BestFit and BoxShade of the GCG package and multalin (35).
Black-boxed residues indicate identical residues, and
shaded boxed residues indicate conserved residues. Consensus
sequences are shown for each region. All residue numbers refer to their
positions in GAGA519 only.
GAGA was put into question as a real activator because of its inability to stimulate transcription in Drosophila SNF extracts devoid of histone H1, whereas it could stimulate transcription when histone H1 was added back. This result, previously reported by Kadonaga and co-workers (14), was considered to be the indication that GAGA was acting as an antirepressor rather than a true activator (14). Histone H1 was identified in some crude extracts as the repressing factor GAGA was counteracting. The lack of GAGA activity in SNF-like extracts was then correlated with the absence of histone H1 in these extracts (14, 24, 28). This antirepression activity was subsequently extended to Gal4BD-VP16 (31), and in general, it can be assumed that all activators have antirepressor activities.
Here, we have used a HeLa crude nuclear extract that may contain some histone H1, and therefore, we cannot provide a definite answer to this question. Nevertheless, we have previously shown that the addition of plasmid DNA to compete for histone H1 binding did not reduce GAGA stimulation in HeLa extracts, suggesting that GAGA was not only acting as an antirepressor but also as a true activator (20). In any case, our results clearly show that the stimulatory activity of GAGA depends on the Q-domain. In addition, in using SNF extract from Drosophila embryos, we found that Gal4BD-Q could also stimulate transcription in vitro, whereas GAGA could not (not shown). Thus, it is conceivable that the lack of activity of added GAGA in these extracts may be due to some blockage of the Q or DNA binding domains by some factor(s) interacting with GAGA in regions other than Q-domain or to some factor(s) competing for binding to the same DNA sequences (e.g. endogenous GAGA). Some recent reports give support to possibilities other than the H1 repression mechanism and should be taken into consideration. In particular, the Drosophila gene pipsqueak encodes a POZ-domain-containing protein that is required in the oogenesis, is highly expressed in the early embryo, and binds GAGAG sequences with high specificity (32, 33). The presence of this factor in large amounts in the extracts may compete or at least interfere with GAGA activation. This possibility does not seem to exist in HeLa nuclear extracts, since we have not found any activity that could bind to GAGA sequences or transactivate reporters bearing GAGA binding sites (Ref. 20 and results not shown).
In summary, the properties described here seem to indicate that Q-domain acts as a bona fide transactivation domain in vitro because it (i) acts irrespective of the rest of the protein and is transportable, ii) presents an internal modular structure, and iii) is functional in SNF extracts in the form of Gal4BD-Q fusions. Additionally, GAGA appears to enhance transcription mainly by stimulating GTF recruitment and reinitiation,2 further confirming its properties as a real transactivation factor.
In vivo GAGA can stimulate transcription in SL2 cells (Refs.
10 and 11 and this work) in a Q-domain-dependent manner. Deletion of the Q-domain abolishes activation, and even more
transcription levels go below the background. This result can be
explained by a down-regulation of the endogenous GAGA factor activity
present in SL2 (as detected by Western blot analysis and also by
comparing the activities of templates bearing and lacking GAGA binding
sites, not shown). We have estimated that endogenous GAGA can
transactivate templates containing GAGA binding sites ~5-8-fold
relative to the same ones without them (not shown). Thus, expression of
GAGA-Q reduced the total levels of activation, probably because it
efficiently competes for binding to the GAGA sequences. The presence of
the POZ domain may also permit the formation of mixed oligomers of both
GAGA forms and, thus, contribute to reduce activation yields. These
interpretations are supported in part by the fact that expression of a
GAGA mutant in which the Q-domain was replaced by the VP16 transactivation domain stimulated transcription in vivo but
to a much lesser extent than observed with Gal4BD-VP16.
The activation levels obtained by GAGA overexpression in SL2 cells are modest (2-fold on the average) compared with the levels reported by Kornberg and co-workers (11) but are similar to the levels reported by Elgin and co-workers (10) and significantly lower than obtained with Gal4BD-Q or Gal4BD-VP16. The reasons for the differences are not fully understood but might be related to differential responses due to the several promoters used (multimerized engrailed C site versus synthetic (CT)22 sequences) as recently pointed out (34).
From all the accumulated data, GAGA appears to be a rather complex
factor involved in a whole series of nuclear events. To carry out all
their functions in the fly, GAGA will probably be under tight
regulation by means of posttranslational modifications, association
with other factors, etc. Here we have described its potential to
stimulate transcription in a simple system. More work is required to
define the exact contribution of GAGA in all the processes where it is
involved and how it is regulated.
| |
ACKNOWLEDGEMENTS |
|---|
We thank Dr. Peter Becker for a generous gift of SNF extract and for advice, and we thank all other members of our group for fruitful discussions and comments.
| |
FOOTNOTES |
|---|
* This work was supported by Spanish Dirección General de Enzeñanza Superior Grant PB96-0812, European Union Grant FMRX-CT97-0109, and Comissió Interdepartamental de Recerca i InnovacióTecnològica (CIRIT) of the Generalitat de Catalunya Grant SGR97-55. This work was carried out in the context of the Centre de Referència en Biotecnologia of the CIRIT of the Generalitat de Catalunya.
A postdoctoral fellow of the Spanish Dirección General de
Enzeñanza Superior.
§
To whom correspondence should be addressed: Dept. Biologia
Molecular i Cel·lular, Inst. Biologia Molecular de Barcelona,
CID-CSIC, Jordi Girona, 1826, 08034 Barcelona, Spain. Tel.:
34-93-400 61 77; Fax: 34-93-204 59 04; E-mail:
jbmbmc@cid.csic.es.
Published, JBC Papers in Press, April 11, 2000, DOI 10.1074/jbc.M000967200
2 A. Vaquero, M. L. Espinás, F. Azorín, and J. Bernués, manuscript in preparation.
| |
ABBREVIATIONS |
|---|
The abbreviations used are: DBD, DNA binding region; Q-domain, glutamine-rich domain; SNF, soluble nuclear fraction.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
| 1. | Farkas, G., Gausz, J., Galloni, M., Reuter, G., Gyurkovics, H., and Karch, F. (1994) Nature 371, 806-808 |
| 2. | Bhat, K. M., Farkas, G., Karch, F., Gyurkovics, H., Gausz, J., and Schedl, P. (1996) Development 122, 1113-1124 |
| 3. | Espinás, M. L., Jiménez-García, E., Vaquero, A., Canudas, S., Bernués, J., and Azorín, F. (1999) J. Biol. Chem. 274, 16461-16469 |
| 4. | Katsani, K. R., Hajibagueri, M. A. N., and Verrijzer, C. P. (1999) EMBO J. 18, 698-708 |
| 5. | Pedone, P. V., Ghirlando, R., Clore, G. M., Gronenborn, A. M., Felsenfeld, G., and Omichinski, J. G. (1995) Proc. Natl. Acad. Sci. U. S. A. 93, 2822-2826 |
| 6. | Omichinski, J. G., Pedone, P. V., Felsenfeld, G., Gronenborn, A. M., and Clore, G. M. (1997) Nat. Struct. Biol. 4, 122-132 |
| 7. | Granok, H., Leibovitch, B. A., Shaffer, C. D., and Elgin, S. C. R. (1995) Curr. Biol. 5, 238-241 |
| 8. | Agianian, B., Leonard, K., Bonte, E., Van der Zandt, H., Becker, P. B., and Tucker, P. A. (1999) J. Mol. Biol. 285, 527-544 |
| 9. | Wilkins, R. C., and Lis, J. T. (1999) J. Mol. Biol. 285, 515-525 |
| 10. | Benyajati, C., Mueller, L., Xu, N., Pappano, M., Gao, J., Mosammaparast, M., Conklin, D., Granok, H., Craig, C., and Elgin, S. (1997) Nucleic Acids Res. 25, 3345-3353 |
| 11. | Soeller, W. C., Oh, C. E., and Kornberg, T. B. (1993) Mol. Cell. Biol. 13, 7961-7970 |
| 12. | Soeller, W. C., Poole, S. J., and Kornberg, T. (1988) Genes Dev. 2, 68-81 |
| 13. | Read, D., Nishigaki, T., and Manley, J. L. (1990) Mol. Cell. Biol. 10, 4334-4344 |
| 14. | Kerrigan, L. A., Croston, G. E., Lira, L. M., and Kadonaga, J. T. (1991) J. Biol. Chem. 266, 574-582 |
| 15. | Lee, H.-S., Kraus, K. H., Wolfer, M. F., and Lis, J. T. (1992) Genes Dev. 6, 284-295 |
| 16. | Okada, M., and Hirose, S. (1998) Mol. Cell. Biol. 18, 2455-2461 |
| 17. | Lu, Q., Wallrath, L. L., Granok, H., and Elgin, S. C. R. (1993) Mol. Cell Biol. 13, 2802-2814 |
| 18. | Tsukiyama, T., Becker, P. B., and Wu, C. (1994) Nature 367, 525-532 |
| 19. | Mizuguchi, G., Tsukiyama, T., Wisniewski, J., and Wu, C. (1997) Mol. Cell 1, 141-150 |
| 20. | Jiménez-García, E., Vaquero, A., Espinás, M. L., Soliva, R., Orozco, M., Bernués, J., and Azorín, F. (1998) J. Biol. Chem. 273, 24640-24648 |
| 21. | Sutrias-Grau, M., Bianchi, M. E., and Bernués, J. (1999) J. Biol. Chem. 274, 1628-1634 |
| 22. | Chasman, D. I., Leatherwood, J., Carey, M., Ptashne, M., and Kornberg, R. D. (1989) Mol. Cell. Biol. 9, 4746-4749 |
| 23. | Bernués, J., Simmen, K. A., Lewis, J. D., Gunderson, S. I., Polycarpou-Schwarz, M., Moncollin, V., Egly, J.-M., and Mattaj, I. W. (1993) EMBO J. 12, 3573-3585 |
| 24. | Kamakaka, R. T., Tyree, C. M., and Kadonaga, J. T. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 1024-1028 |
| 25. | Krasnow, M. A., Saffman, E. E., Kornfeld, K., and Hogness, D. S. (1989) Cell 57, 1031-1043 |
| 26. | Lin, Y.-S., Carey, M. F., Ptashne, M., and Green, M. R. (1988) Cell 54, 659-664 |
| 27. | Sadowski, I., Ma, J., Triezenberg, S., and Ptashne, M. (1988) Nature 335, 563-564 |
| 28. | Croston, G. E., Kerrigan, L. A., Lira, L. M., Marshak, D. R., and Kadonaga, J. T. (1991) Science 251, 643-649 |
| 29. | Gerber, H.-P., Seipel, K., Georgiev, O., Höfferer, M., Hug, M., Rusconi, S., and Schaffner, W. (1994) Science 263, 808-811 |
| 30. | Lintermann, K.-G., Roth, G. E., King-Jones, K., Korge, G., and Lehmann, M. (1998) Dev. Genes Evol. 208, 447-456 |
| 31. | Croston, G. E., Laybourn, P. J., Paranjape, S. M., and Kadonaga, J. T. (1992) Genes Dev. 6, 2270-2281 |
| 32. | Lehmann, M., Siegmund, T., Lintermann, K.-G., and Korge, G. (1998) J. Biol. Chem. 273, 28504-28509 |
| 33. | Horowitz, H., and Berg, C. A. (1996) Development 122, 1859-1871 |
| 34. | Pile, L. A., and Cartwright, I. L. (2000) J. Biol. Chem. 275, 1398-1404 |
| 35. | Corpet, F. (1988) Nucleic Acids Res. 16, 10881-10890 |
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Schweinsberg, K. Hagstrom, D. Gohl, P. Schedl, R. P. Kumar, R. Mishra, and F. Karch The Enhancer-Blocking Activity of the Fab-7 Boundary From the Drosophila Bithorax Complex Requires GAGA-Factor-Binding Sites Genetics, November 1, 2004; 168(3): 1371 - 1384. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Choi, J. H. Kim, and H. Kende Whole Genome Analysis of the OsGRF Gene Family Encoding Plant-Specific Putative Transcription Activators in Rice (Oryza sativa L.) Plant Cell Physiol., July 15, 2004; 45(7): 897 - 904. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Pagans, D. Pineyro, A. Kosoy, J. Bernues, and F. Azorin Repression by TTK69 of GAGA-mediated Activation Occurs in the Absence of TTK69 Binding to DNA and Solely Requires the Contribution of the POZ/BTB Domain of TTK69 J. Biol. Chem., March 12, 2004; 279(11): 9725 - 9732. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Salvaing, A. Lopez, A. Boivin, J. S. Deutsch, and F. Peronnet The Drosophila Corto protein interacts with Polycomb-group proteins and the GAGA factor Nucleic Acids Res., June 1, 2003; 31(11): 2873 - 2882. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Kosoy, S. Pagans, M. L. Espinas, F. Azorin, and J. Bernues GAGA Factor Down-regulates Its Own Promoter J. Biol. Chem., October 25, 2002; 277(44): 42280 - 42288. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Pagans, M. Ortiz-Lombardia, M. L. Espinas, J. Bernues, and F. Azorin The Drosophila transcription factor tramtrack (TTK) interacts with Trithorax-like (GAGA) and represses GAGA-mediated activation Nucleic Acids Res., October 15, 2002; 30(20): 4406 - 4413. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. A. Leibovitch, Q. Lu, L. R. Benjamin, Y. Liu, D. S. Gilmour, and S. C. R. Elgin GAGA Factor and the TFIID Complex Collaborate in Generating an Open Chromatin Structure at the Drosophila melanogaster hsp26 Promoter Mol. Cell. Biol., September 1, 2002; 22(17): 6148 - 6157. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. J. Greenberg and P. Schedl GAGA Factor Isoforms Have Distinct but Overlapping Functions In Vivo Mol. Cell. Biol., December 15, 2001; 21(24): 8565 - 8574. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. B. Kumar, P. Ramadoss, R. K. Reen, J. P. Vanden Heuvel, and G. H. Perdew The Q-rich Subdomain of the Human Ah Receptor Transactivation Domain Is Required for Dioxin-mediated Transcriptional Activity J. Biol. Chem., November 2, 2001; 276(45): 42302 - 42310. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
