ABSTRACT
We detected a novel nuclear protein, MRF, that binds to multiple sites on the
modulator which is located upstream of the human cytomegalovirus major immediate early gene enhancer. The expression of MRF is differentiation specific; the DNA binding activity is present in nuclear extracts from undifferentiated Tera-2 and THP-1 cells, but significantly reduced after these cells are induced to
differentiate. In undifferentiated cells the enhancer activity is repressed by
the modulator and upon differentiation the enhancer becomes active. Competitive
binding assays demonstrate that MRF requires the presence of multiple A+T
stretches for binding to DNA, rather than binding to a specific DNA sequence.
Mutations of these stretches in the modulator reduce the binding activity of
MRF, as well as the repressing activity on the enhancer. These results suggest
that MRF may act as a repressor of enhancer function. We propose that MRF binds
over the entire modulator and exerts repressor activity.
Transcription of the major immediate early (M-IE) gene of human cytomegalovirus (HCMV) in Tera-2 cells, a human embryonal carcinoma cell line, provides a model
system for studying cellular factors which regulate gene expression during
differentiation. Tera-2 cells can be maintained in an undifferentiated state or can be induced
to differentiate by treatment with retinoic acid (
1
). Undifferentiated Tera-2 cells are not permissive for HCMV replication, but retinoic acid-treated Tera-2 cells are (
2
,
3
). The block to viral replication in undifferentiated Tera-2 cells has been demonstrated to be due to repression of M-IE gene expression (
4
). The M-IE gene product regulates expression of other viral genes (
5
-
8
), but M-IE gene expression itself is heavily dependent upon cellular factors (
2
). Previously described 5' regulatory elements of the M-IE gene include a very strong enhancer containing repetitive
palindromic sequence elements (
9
,
10
), a cluster of NF-1 binding sites (
11
,
12
) and a far upstream region termed the modulator, comprising nt -750 to -1145 relative to the transcription start site (
4
,
13
-
16
) (see Fig.
1
). Nelson
et al.
(
4
,
15
) have demonstrated that the modulator plays a crucial role in mediating the
cell-specific expression of the M-IE gene observed in undifferentiated and differentiated Tera-2 cells. They found that
in vivo
DNase I hypersensitive and chloracetaldehyde sensitive sites present only in
cell types permissive for HCMV M-IE gene expression map to the modulator. They also demonstrated by
transient transfection assays that the modulator functions as an up-regulator in permissive cells and as a down-regulator in non-permissive cells. A similar result was reported for transient
transfection assays using undifferentiated and differentiated THP-1 cells, where HCMV replication is also differentiation dependent (
17
). These observations support the role of the modulator of the M-IE gene as a mediator of cell-specific gene expression.
Tera-2 cells (subclone NT2D1;
1
) were obtained from P.W. Andrews of the Wistar Institute and grown in DME/high
glucose medium containing 10% fetal calf serum. For retinoic acid treatment
cells were plated at a density of 10
6
cells/T75 flask in medium containing 10
-5
M retinoic acid and harvested after retinoic acid treatment for the indicated
times. THP-1 cells were purchased from ATCC and grown in RPMI medium containing 10%
fetal calf serum. For differentiation cells were seeded at a density of 5 * 10
5
cells/ml in medium containing 1.6 x 10
-7
M TPA and harvested after 3 days (
21
).
A plasmid containing the entire control region of the M-IE gene, derived from HCMV strain AD169, was a gift from Dr H.Pande
(Beckman Research Institute, City of Hope). For the location of
cis
-acting suppressor elements plasmids containing the region from -1140 to +74 (p307CAT), -825 to +74 (p305CAT) and -600 to +74 (p303CAT) were constructed by cloning the
Pst
I-
Xma
III,
Taq
I-
Xma
III and
Hin
cII-
Xma
III fragments of the M-IE gene of HCMV into the
Hin
dIII site of pSVO-CAT. Plasmids with 5' deletions to -915 (p504CAT), -844 (p508CAT) and -800 (p507CAT) were constructed by Bal31
digestion of the p307CAT construct. A
Hin
dIII linker was ligated to the 5'-end of each deletion product and the DNA was cut with
Hin
dIII and
Nsi
I. The
Hin
dIII-
Nsi
I fragment was then cloned into the
Hin
dIII
and Nsi
I sites of p307CAT. The 5' deletion position of each plasmid was determined by sequencing. These
original plasmids were inconvenient for the construction of plasmids necessary
for subsequent experiments and therefore all inserts were subcloned into pCAT-Basic (Promega) by digesting with
Hin
dIII, filling in with Klenow enzyme and blunt-end ligating to the filled in
Xba
I site. For convenience the same nomenclature of plasmids was used after
subcloning. The HA probe was produced by digestion of p508CAT with
Xba
I and
Alu
I unless otherwise noted. pM
5
CAT were produced by mutation of TA to GC at the underlined places in the
palindromic sequence A
Nuclear extract preparation was adapted from the method of Dignam
et al.
(
22
), except that the homogenization step of isolated nuclei was eliminated because
the yield of MRF was significantly reduced. Washed cell pellets were suspended
in 5 vol buffer A, which contained 10 mM HEPES, pH 7.5, 1.5 mM MgCl
2
, 10 mM KCl, 0.5 mM DTT and 1 mM PMSF. The suspension was incubated on ice for 10 min and centrifuged at 900
g
for 5 min at 4oC. The pellet was resuspended in 4 vol buffer A, homogenized with a Dounce homogenizer using 20 strokes with a `B' pestle and then centrifuged at 325
g
for 10 min. The pellet was resuspended in 0.8 ml buffer A and centrifuged at
325
g
for 10 min. The supernatant was removed and the pellet was suspended in 4 vol
buffer C (20 mM HEPES, pH 7.4, 25% glycerol, 0.42 M NaCl, 1.5 mM MgCl
2
, 0.2 mM EDTA, 1 mM PMSF and 0.5 mM DTT) and placed on a tube rotator at 4oC for 30 min. The tubes were then centrifuged at 8000
g
for 30 min at 4oC. The supernatant was removed and dialyzed against buffer D (20 mM HEPES,
pH 7.4, 20% glycerol, 0.1 M KCl, 0.2 mM EDTA, 0.5 mM PMSF and 0.5 mM DTT).
After dialysis for 5 h the sample was centrifuged at 8000
g
for 30 min at 4oC and the supernatant aliquoted and stored at -80oC.
Calcium phosphate transfections were performed as described (
23
). For the series of nested deletion mutants Tera-2 cells (0.5 * 10
6
) were co-transfected with 0.05 [mu]g
CAT
vector, 1 [mu]g human growth hormone (hGH) expression vector (
24
) and 4 [mu]g Bluescript SKII+ (Stratagene) plasmid as carrier DNA. Levels of bacterial
CAT were assayed by thin layer chromatography after 48 h transfection (
23
). Secreted human growth hormone was assayed with a radioimmunoassay kit
(Nichols Institute Diagnostics, San Juan Capistrano, CA), following the
manufacturer's recommended protocol. For the base mutation studies cells were
co-transfected with 0.1 [mu]g
CAT
vector and 10 [mu]g Bluescript. CAT activities were measured by direct scintillation counting
(
25
). THP-1 cells (2 * 10
6
) in 2 ml RPMI without serum were transfected with 20 [mu]l DEAE-dextran (10 mg/ml) containing 3 [mu]g
CAT
vector for 20 min. Cells were collected and re-suspended in 4 ml RPMI plus 10% fetal calf serum. After 48 h CAT
activities were measured as described above.
Up to 10 [mu]l of nuclear extract was added to a tube containing 4 [mu]l H
2
O, 2 [mu]l 10* binding buffer (500 mM NaCl, 100 mM Tris, pH 7.5, 10 mM DTT, 10 mM
EDTA and 50% glycerol), 1 [mu]g poly(dI-dC), 1 [mu]l 200 mM MgCl
2
and 1 ng end-labeled DNA probe. Competitor DNA, if added, was dissolved in 4 [mu]l water. The volume was brought to 20 [mu]l with nuclear extract in buffer D described above. The mixture was incubated at room temperature
for 30 min to allow complex formation and was analyzed on a 3.5% polyacrylamide
gel using a low ionic strength buffer system (6.7 mM Tris, pH 7.5, 3 mM sodium
acetate) at room temperature for 60-90 min at 100 V.
The sequences of the plus strands of the synthetic oligonucleotide duplexes used
in gel mobility shift assays are as follows:
TAQ, TCGCAAATATCGCAGTTTCGATATAGGTGA (-845 to -816);
-800, GACAGACGATATGAGGCTATATCGCCGATA (-817 to -788);
MBF-1, GATTTTTGGGCATACGCGATATCTG (-938 to -914);
NF-1, CGCCCTCTCTCGCCCTTGCCAAGTTGGAAG.
TAQ-800 corresponds to nucleotides -845 to -788. Concatenated oligonucleotide duplexes (a mixture of dimer to octamer) were prepared by self-ligation of kinased oligonucleotide duplexes with T4 DNA ligase. After phenol extraction and
ethanol precipitation the ligation reaction mixtures were analyzed on
polyacrylamide gels by electrophoresis and used as competitors without further
purification. Dimer oligonucleotides were separated by gel electrophoresis,
extracted from the gel by electroelution and cloned into the Bluescript
plasmid. TAQ dimer was cloned into the blunt-ended
Hin
dIII site such that only the tandem repeat of the dimer regenerated a
Hin
dIII site. Both the -800 dimer and TAQ-800 were cloned into blunt-ended
Hin
dIII and
Hin
cII sites. All DNA sequences were confirmed by sequencing. The DNA fragments
from the modulator were prepared by digestion with appropriate restriction
enzymes. Bluescript SKII+ was digested with
Hae
III to obtain two DNA fragments of 102 (509-611) and 79 bp (661-740). All these DNA fragments were used as competitors for gel
mobility shift assays after gel purification.
Previous work indicated that the modulator is responsible for repression of HCMV
M-IE gene transcription in undifferentiated Tera-2 cells (
13
,
15
,
16
). With the intention of locating negative regulatory elements in the modulator,
we constructed a series of nested deletion mutants in the modulator that is
linked to the enhancer and promoter of the M-IE gene fused to the bacterial
CAT
gene. We co-transfected these reporter constructs with a human growth hormone (hGH)
expression vector into undifferentiated Tera-2 cells using the calcium phosphate method and the levels of transiently
expressed CAT enzyme and secreted hGH were measured. The ratio of CAT activity
to secreted hGH for each construct was plotted versus the 5' deletion positions (Fig.
1
A). The relative CAT activities of the constructs gradually increased as the
modulator sequence was deleted, except for deletion of the nucleotides between
positions -860 and -845. Deletion of the 20 nt between positions -845 and -825 resulted in a significant increase in relative
CAT activity. These results suggest that there are several
cis
-acting negative regulatory elements in the modulator. The increase in
expression seen upon deletion of the 20 nt from -845 to -825 led us to concentrate our initial work on a putative negative
regulatory element that might be present near or within this sequence. However,
it should be noted that the observed increase in transcription might be caused
by losing an activating element or creating an artificial repressing element at
the new junction in the -845 construct. The increase in CAT activity due to the deletion from -845 to -825 was roughly equal to the decrease due to deletion from -860 to -845.
In order to identify DNA binding factors that recognize the negative regulatory
element we labeled the HA probe (-845 to -774 bp; Fig.
1
B) and used it for gel mobility shift assays with nuclear extracts from
undifferentiated and retinoic acid-treated Tera-2 cells. With the nuclear extract from untreated Tera-2 cells a slow moving DNA-protein complex was detected (Fig.
2
, left panel, lane 2). Formation of the complex was significantly reduced with
the nuclear extracts prepared from Tera-2 cells after 2 days of retinoic acid treatment (lane 3) and barely
detectable after 5 days (lane 4). On the other hand, complex formation of the
oligonucleotide probe with an NF-1 recognition sequence was not significantly affected by retinoic acid
treatment (Fig.
2
, right panel). To examine whether formation of the slow moving DNA-protein complex is specific to the HA probe we carried out competitive
gel mobility shift assays. Complex formation was extinguished by competition
with an excess amount of the non-radioactive HA probe (10-fold), suggesting that binding of the nuclear factor(s) to the probe
was specific (Fig.
3
B, lanes 3 and 4). We refer to this DNA binding factor as the modulator
recognition factor (MRF), to distinguish it from the modulator binding factors
(MBF) described by Shelbourn. The latter recognizes sequences within the dyad
symmetry element (-963 to -912 bp;
18
). Occasionally we observed formation of other shifted bands with higher
mobility than the slow moving complex in gel mobility shift assays. However,
formation of those bands was not inhibited in the presence of cold HA probe
and, therefore, the complexes may be formed by non-specific factors. Sinclair
et al.
demonstrated by transient transfection assays that the monocytic cell line THP-1 is non-permissive for HCMV replication due to a block in expression of the
M-IE gene promoter (
17
). Differentiation of THP-1 cells into macrophages lifts this block and permits M-IE gene expression. To investigate whether binding activity of MRF
to the HA probe is present in THP-1 cells and controlled by differentiation of THP-1 cells we carried out gel mobility shift assays with nuclear
extracts from untreated and TPA-treated THP-1 cells. Formation of a DNA-protein complex similar to the MRF-HA complex was observed with untreated THP-1 cells and significantly reduced with the
differentiated cells (Fig.
4
A, lanes 6 and 7) in a similar manner to that described above for Tera-2 cells.
In order to identify the DNA sequence recognized by MRF we carried out
methylation interference and DNase I footprinting assays with the HA probe, but
failed to determine the recognition sequence. Methylation interference assays
with the crude nuclear extract from Tera-2 cells did not identify any guanine bases at which methylation
significantly disturbed DNA-MRF complex formation (data not shown). DNase I footprinting assays in
the presence of nuclear extracts from Tera-2 cells gave unusual results. Instead of distinct footprints we observed
simultaneously enhanced and suppressed DNase I digestion over the entire HA
probe in the presence of the nuclear extract (data not shown).
Examination of the DNA sequence of the HA fragment revealed the presence of one
inverted repeat motif, 5'-ATATCG N
6
CGATAT-3' at nucleotide positions -839 to -822 (Fig.
1
B) and the same inverted repeat in the opposite orientation at positions -811 to -794. Thus we synthesized two oligonucleotides (TAQ and -800) which contained the respective inverted repeat motif
(Fig.
1
B) and investigated whether either or both of the oligonucleotides were involved
in binding of MRF. We performed gel mobility shift assays, using the HA
fragment as a labeled probe and the double-stranded oligonucleotides as cold competitors (Fig.
3
A). The results showed that these single oligonucleotides could not inhibit
formation of the specific high molecular weight complex, even when present at
100-fold molar excess (lanes 3-6). However, the complex disappeared completely in the presence of a 10-fold excess of the concatenated TAQ and -800 oligonucleotides (lanes 9-12), but not in the presence of concatenated NF-1 (
26
,
27
) oligonucleotide (lanes 15-16). The finding that concatenated probes competed more efficiently than
monomer probes suggests that formation of the high molecular weight DNA-protein complex requires the presence of the two juxtaposed
oligonucleotides (TAQ and -800) and that each oligonucleotide plays a similar role in complex
formation. In order to investigate this, tandemly repeated dimers of TAQ and -800 oligonucleotides (TAQ
2
and -800
2
) and TAQ-800 oligonucleotide were prepared and examined for their ability to
inhibit formation of the HA-MRF complex. As shown in Figure
3
B, both the homodimers (lanes 5-8) and the TAQ-800 oligonucleotide (lanes 9 and 10) abolished formation of the
complex. We conclude that the two juxtaposed oligonucleotides are required for
binding to MRF.
The above result suggests that ATATCG could be a part of the recognition
sequence for MRF, since inspection of the oligonucleotide sequences did not
reveal any homology other than ATATCG tandem repeats. However, our methylation
interference experiments suggested that guanine bases do not play a significant
role in formation of the HA-MRF complex (not shown). Therefore, we speculated that four short A+T
stretches (4-6 bp) with the common ATAT sequence in the tandem repeats may play an
important role in MRF binding.
In order to test this speculation and examine whether the ability to form a
complex with MRF correlates with repressor activity we produced two mutants of
the plasmid p508CAT, which carries 845 nt of the 5' regulatory sequence of the M-IE gene (Fig.
1
A). In mutant M5 both of the inverted repeats ATATCG in the TAQ sequence were
mutated to A
GC
TCG. M6 incorporated both of the mutations in the inverted repeats of the -800 oligonucleotides. Base modified HA fragments were prepared from these
mutated plasmids as described in Materials and Methods and used as cold
competitors in gel mobility shift assays with the labeled wild-type HA probe (Fig.
4
A). Radioactivity of the residual HA-MRF complex in each lane showed that the wild-type probe competed better (lane 3, Tera-2 cells and lane 8, THP-1 cells) than the two mutants (lanes 4-5, Tera-2 cells and lanes 9-10, THP-1 cells). This result
suggests that the A+T stretches in the HA probe play an important role in MRF
binding. In order to investigate the effect of the mutations on repressing
activity we carried out transient transfection assays in uninduced Tera-2 and THP-1 cells with plasmids carrying these mutations (Fig.
4
B). We observed that the plasmid carrying the full-length modulator (p307CAT) had ~25% of the CAT activity of the plasmid carrying only the enhancer
(p303CAT). As shown above (Fig.
1
A), deletion of the nucleotides from -845 to -820 produced a sharp decrease in repressing activity. The
mutations in the ATATCG inverted repeats in this region produced a comparable
decrease in repressing activity: both p508-M5CAT (mutations in TAQ) and p508-M6CAT (mutations in -800) showed a consistent 2-fold increase in CAT activity compared with p508CAT in
both Tera-2 and THP-1 cells. Therefore, mutations in the TAQ and -800 sequences that disrupt binding of MRF also decrease the
cis
-repressing activity of the modulator. This suggests that MRF is a
trans
-acting regulatory factor that represses the M-IE enhancer activity.
The results of transient transfection assays using nested deletion mutants
suggested that there are several
cis
-acting negative elements within the modulator (Fig.
1
A). In order to investigate whether MRF binds to other DNA fragments in the
modulator we conducted competitive gel mobility shift assays. We prepared
several restriction fragments from the modulator and used them as non-radioactive competitors for binding of the HA probe to MRF in Tera-2 nuclear extracts. As shown in Figure
5
, all the modulator-derived fragments, including
Pst
I-
Cla
I (-1137 to -1020, lane 4),
Cla
I-
Hin
fI (-1020 to -856, lane 5),
Hin
fI-
Hin
cII (-856 to -599, lane 6), inhibited binding as efficiently as unlabeled HA
itself (lane 3). Two DNA fragments from the Bluescript SKII+ plasmid did not,
however, compete as well (lanes 7-8). These results suggest that MRF could bind to the modulator at
multiple sites, which may be necessary for full repression of M-IE gene transcription. All of the modulator fragments are A+T rich (
Pst
I-
Cla
I, 60% A+T;
Cla
I-
Hin
fI, 55%;
Hin
fI-
Hin
cII, 66%), but the Bluescript fragments are not (45 and 41%). The modulator
fragments all have uninterrupted stretches of A+T and some, but not all, have
the ATAT sequence; the Bluescript fragments have neither of these sequences.
These results are consistent with our belief that A+T stretches are required
for MRF binding.
The results of transient transfection assays by Nelson
et al.
indicated that the modulator of the M-IE gene represses its expression in cells that are non-permissive for HCMV (
15
). Shelbourn
et al.
have demonstrated that deletion of the imperfect dyad symmetry element within
the modulator (-963 to -912) releases some repressor activity in undifferentiated Tera-2 cells (
18
). However, it was apparent from further deletion analysis that sites other than
the dyad symmetry element repress expression of the M-IE gene in undifferentiated Tera-2 cells (
20
,
28
). Our transient transfection assays using a series of deletion mutants also
suggest that there are several repressor elements in the region (Fig.
1
A). We have focused our initial attention on a
cis
-acting element within the HA probe (nt -845 to -774), since we consistently observed an increase in CAT
activity upon deletion of the sequence.
Gel mobility shift assays demonstrated that the HA probe binds specifically to a
DNA binding factor (MRF) that is present in undifferentiated Tera-2 and THP-1 cells but substantially reduced in differentiated cells.
Unfortunately, all attempts to identify the DNA recognition sequence by
methylation interference and DNase I footprinting assays have failed. From the
results of gel mobility shift (Fig.
3
A and B) and methylation interference assays we speculated that A+T stretches in
ATATCG repeats are involved in binding of the HA probe to MRF. In order to test
this speculation various mutations in the A+T stretches present in the repeats
were produced. Competition experiments showed that HA fragments with TA -> GC mutations in the A
TA
TCG repeats exhibit decreased MRF binding activity and thus support our
speculation that A+T stretches with the ATAT sequence in the repeats are
required for binding to MRF (Fig.
4
A). The finding that formation of the HA-MRF complex is specifically inhibited in the presence of DNA fragments
from the modulator (Fig.
5
) indicates that binding sites for MRF are located throughout the modulator.
Examination of the sequences of these DNA fragments revealed that the
Cla
I-
Hin
fI fragment has four ATATCG repeats, the
Pst
I-
Cla
I fragment has one and the
Hin
fI-
Hin
cII fragment has none. These facts suggest that juxtaposed palindromes
containing the ATATCG repeats, present only in the HA probe, are not absolutely
required for recognition by MRF. The competing DNA fragments have numerous
contiguous A+T stretches, some of which carry the ATAT sequence, and they are
relatively A+T rich (53-66%). On the other hand, the non-competitive DNA fragments from the plasmid have fewer A+T stretches
and no ATAT sequences and are not A+T rich overall (41-45%). Our preliminary data of carbethoxylation interference assays (
29
) using nuclear extract from THP-1 cells and the HA probe demonstrated that A+T stretches are major contact
sites for binding to MRF (not shown). A+T-rich regions are known to adopt unusual DNA conformations, such as single-stranded (
30
) and bent (
31
). Therefore, we propose that specificity of binding is related to A+T stretches
in DNA which may have certain structural features, but not to a specific DNA
sequence, and that these features could be adopted preferentially by multiple
ATAT sequences.
Plasmids (p508-M5 and p508-M6) carrying mutations in the ATAT sequences of HA have lower MRF
binding activities and higher CAT activities than the wild-type plasmid (Fig.
4
A and B). MRF is highly abundant in undifferentiated cells such as Tera-2 and THP-1 and is reduced significantly upon differentiation (Figs
2
and
4
A). In undifferentiated cells enhancer activity is repressed by the modulator
located upstream of the enhancer and the repressing activity falls off with
progressive deletion from the 5'-end of the modulator (Fig.
1
A). In differentiated cells the repressing activity is released and enhancer
activity is restored (
15
,
17
; our unpublished results) with the decreasing amount of MRF (Figs
2
and
4
A). We conclude that MRF is a transcription factor involved in repression of the
enhancer activity in undifferentiated Tera-2 and THP-1 cells.
In vivo
footprinting assays of the M-IE gene showed no DNase I hypersensitive sites in the modulator in
undifferentiated Tera-2 cells. Upon differentiation DNase I hypersensitive sites are induced in
the modulator. In contrast, DNase I hypersensitive sites in the enhancer are
constitutive (
4
). These assays indicate that differentiation of the cellular host drastically
alters the structure of viral chromatin in the modulator, but not in the
enhancer. Taking these results together we propose that MRF binds over the
entire modulator and exerts a repressing activity which is dependent on the
amount of MRF bound. It is a very interesting problem to examine whether a
mutated virus carrying deletion of the modulator expresses the M-IE gene in undifferentiated cells (non-permissive for wild-type HCMV replication). These experiments are in progress in
other laboratories (private communications from Drs Mocarski and Stinski).
Recently numerous YY-1 binding sites have been identified in the modulator (
19
; our unpublished results) and co-transfection assays show that YY-1 is a
trans-
acting repressor of the M-IE gene in Tera-2 cells (
19
). YY-1 is a ubiquitously expressed zinc finger DNA binding protein which may
function as a positive or negative regulator of transcription, depending upon
the context of the binding site (
32
-
35
). The HA probe does not appear to have a YY-1 recognition site and formation of the HA-MRF complex is not inhibited in the presence of oligonucleotides
carrying either single or multiple YY binding sites (not shown). We conclude
that MRF is not identical to YY1. It might be possible that YY-1 cooperates with MRF to exert the repressing activity. The monomer of the
3'-half of the dyad symmetry sequence (
18
) which binds to MBF-1 does not compete with formation of the HA-MRF complex (Fig.
3
A, lanes 7-8), but the concatenated oligonucleotide does (lanes 13-14). We speculate that concatenation of the oligonucleotide
fortuitously creates a MRF binding site, since the oligonucleotide carries two
A+T stretches, ATTTTT and ATAT (see the sequence in Materials and Methods).
Thus we believe that MRF is a novel DNA binding protein that negatively
regulates transcription of the M-IE gene. In order to characterize the protein and study mechanisms
involved in the transcriptional regulation and interactions with DNA we are in
the process of cloning and purifying MRF.
During the time this manuscript was being considered for publication Thrower and co-workers (Thrower
et al.
, 1996,
J. Virol
.,
70
, 91-100) reported that a repeated sequence of TATCG or TGTCG found in the
modulator and in the R1 region of the US3 promoter down-regulates transcription from either the US3 promoter or a heterologous
promoter. In contrast to our results, the repressor binding to the repeated
sequence was found in both undifferentiated and differentiated cells. They
propose that the DNA binding activity is controlled by phosphorylation.
We would like to thank Dr P.W.Andrews for his generous gift of Tera-2 cells and Dr J.Sinclair for providing transfection conditions of THP-1 cells. We also thank Rosalina M.Hernandez for assistance with the
preparation of plasmid DNA for this research and Faith Sorensen for preparation
of this manuscript. This research was supported by the IDEN Foundation.
REFERENCES
Return