|
|
|
|||||||
|
|||||||||
(Received for publication, February 7, 1994; and in revised form, October 7, 1994) From the
In primary rodent cells transformed by the E1A region of the
highly oncogenic adenovirus type 12, repression of transcription
mediated by the far upstream TATA-like element was observed only in
conjunction with either possible juxtaposition of a CAA repeated
element in the presence of E1A and was dependent upon the relative
arrangement of both the TATA-like and CAA repeated motifs in both
homologous and heterologous promoter constructs. A gel shift
competition study demonstrated that the TATA-binding protein (TBP) or a
TBP-like protein can bind to both the upstream TATA-like sequence and
the regular TATA box on the H-2K Primary rodent cells transformed by the E1A region of the highly
oncogenic adenovirus type 12 (Ad12) but not by the same region of the
non-oncogenic virus Ad5 (or Ad2) (1) express the reduced level
of the products of major histocompatibility complex (MHC) ( Control of
the initiation of transcription of the mouse H-2K We showed previously that a distal 5`-element
(-1,521 to -1,837 relative to the start site of
transcription) that contributes to E1A-mediated negative regulation is
included in the promoter region of the H-2K Plasmids-The deletion and substitution CAT
mutants, namely, del-4ml, del-4m2, del-5ml, and del-5m2 were
constructed as follows. The BspHI/RsaI DNA fragment
of a distal 5`-flanking sequence of the H-2K
Figure 4:
Relative promoter activity of
pH-2K
Figure 5:
Effects of combinations of various
TATA-like sequences in the upstream region and various proximal
promoters in Ad12-E1A- or Ad5-E1A-transformed cells. Average results of
five independent experiments are shown in the histogram as CAT
activities relative to activity associated with
pH-2K
Separate plasmids containing the H-2K
Figure 1:
Deletion
study of the E1A-responsive negative element in the upstream region of
the promoter of the H-2K
Figure 2:
Relative promoter activities of H-2K
Figure 3:
Effects of the nature of the upstream
TATA-like sequence in the promoter region of the H-2K
Previous studies have
demonstrated that monomers and dimers of the CAA repeats do not
function as negative elements in Ad12-E1A-mediated repression of the H-2K
Figure 6:
Gel shift studies. A, competition
in gel shift assays with oligonucleotide DNA probes that correspond to
the upstream TATA-like sequence (-1,755 to -1,776) and the
proximal TATA-box sequence (-8 to -28), using nuclear
extracts from CYpAdC3 cells. Lane 1, DNA probe labeled with T4
kinase; [
Figure 7:
Immunoselection and CASTing studies.
Cloning of the binding sites in the H-2K
Figure 8:
Nucleotide sequence analysis of DNA clones
isolated by immunoselection and CASTing. A, the major or minor
binding sites indicate the nucleotide sequences of the putative cis-element of the DNA fragments cloned in pBluescript. The
frequency of the clones containing insert DNAs of the major binding
sites is more than 25% in total recombinant clones obtained. A number indicates the distance from the cap site of the H-2K
We previously identified the CAA repeats in the far upstream
region of the mouse MHC class I H-2K To determine the way in
which the repression of the MHC class I gene functions in the context
of heterologous basal promoters, a DNA fragment containing one CAA
repeated sequence was linked, in separate constructs, to the SV40
early, herpes simplex virus type 1, and chicken conalbumin basal
promoters. We did not detect any major difference in the suppressive
activity on these heterologous promoters as compared with the H-2 basal
promoter-enhancer combination in Ad12-E1A transformants (Fig. 3B). Our results are consistent with the previous
report by Ge et al.(30) . Thus, the H-2 TATA-box
sequence per se may not be important since its conversion to
other TATA-box sequences in the promoter failed to affect transcription
of the H-2 gene in Ad12-E1A-transformed cells. However, the suppressive
activity of these elements on heterologous promoters was less
significant in Ad5-E1A-transformed cells (Fig. 3B). We next examined the nature of the optimum TATA-like sequence in the
upstream region for E1A-mediated negative regulation. Examples of
endogenous cellular and viral genes that are activated as a result of
the E1A trans-activation process include the heat-shock gene
for hsp70(46, 47) , the gene for hsp89(48) ,
the gene for What is the
nature of the binding proteins that associate with the DNA sequence of
the far upstream CAA repeats and the TATA-like element, and how do
these proteins interact with TBP (or a TBP-like protein) to achieve
negative regulation of the H-2K The Ad12-E1A-dependent negative repression was independent
of the nature of the heterologous promoters tested but was
preferentially observed with the specific common sequence TAT(A/T)AA.
Recent studies by Brou et al.(59) demonstrated that
distinct TFIID complexes mediate the effects of different types of
transcriptional activation. Similarly, different complexes that
included TBP may be formed and may bind to the upstream TATA-like
sequence or to the proximal TATA sequence in different ways. Further
characterization of the genes that encode the proteins that bind to the
TATA-like sequence and to the basal TATA-box should help us to
elucidate the mechanism of regulation of transcription of the MHC class
I H-2K The nucleotide
sequence(s) reported in this paper has been submitted to the
GenBank(TM)/EMBL Data Bank with accession number(s)
X15104[GenBank].
Volume 270,
Number 5,
Issue of February 3, 1995 pp. 2327-2336
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
Class I Gene (*)
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
basal
promoter. Moreover, employing immunoselection and cyclic amplification
and selection of targets (CASTing) methods with nuclear extracts
derived from Ad12-E1A transformants, we have identified a high affinity
binding site in the H-2K
class I promoter
for E1A-associated DNA-binding proteins. The sequences of the binding
sites were identified and were found to contain both the upstream
TATA-like motif and the CAA repeated motifs. Our results suggest that
the TATA-like sequence in the far upstream region of the H-2K
gene is one of the elements that is
required for Ad12-E1A-mediated negative repression.
)class I genes(2, 3) . The lower levels of
MHC class I mRNA in Ad12-transformed cells reflect decreased rates of
transcription of the endogenous class I genes. This down-regulation of
transcription is mediated by the 1,266-amino acid 13S product of the
Ad12-E1A
region(4, 5, 6, 7, 8, 9, 10) .
The resultant decrease in cell-surface levels of MHC class I antigens
is reflected by the lower susceptibility of Ad12-transformed cells to
allogenic cytotoxic T cells and NK cells, suggesting a model by which
Ad12-transformed cells escape immune surveillance and develop into
proliferating and evasive
tumors(4, 5, 11, 12) .
MHC class
I gene has been studied extensively, and the regulatory region has been
shown to contain the common CCAAT and TATA regions as well as multiple cis-acting regulatory elements. The best characterized of
these elements are RI, RI`, and R2 in the CRE/IRS
region(3, 13, 14, 15) , which have
also been defined by a footprinting study in
vivo(16) . Several transcriptional factors have been shown
to bind to these
elements(17, 18, 19, 20, 21, 22, 23, 24, 25, 26) .
The level of binding to the R1 element has been shown to be higher in
the case of extracts from Ad5-transformed cells than in the case of
extracts from Ad12-transformed
cells(27, 28, 29) . The R2 binding activity
is significantly higher in extracts from Ad12-transformed cells than in
extracts from Ad5-transformed cells or parental
lines(7, 28, 30) . The poor enhancer activity
of the R1 site in Ad12-transformed cells is correlated with increases
in the extent of binding of nuclear factors to the R2 element,
suggesting the presence of a repressor whose effects are mediated via
the R2 site. Kralli et al.(31) demonstrated recently
that the putative R2-binding repressor protein, designated R2BF, is
similar to members of the family of thyroid hormone/retinoic acid
receptors. gene(9) . The CAA repeated sequences in regions
-1,736 to -1,689 and -1,616 to -1,535 are both
necessary for full negative regulation of the H-2K
gene by E1A. We report here that an additional element, a
TATA-like sequence, in the far upstream region of the 5`-flanking
sequence (-1,773 to -1,767) is also a key element in the
negative regulation of expression of the MHC class I H-2K
gene by E1A in conjunction with
either upstream or downstream CAA repeats. A gel shift competition
assay demonstrated that the factor that binds to the far upstream
TATA-like sequence can also bind to the TATA-box sequence in the basal
promoter. Using an immunoselection and cyclic amplification and
selection of targets (CASTing) method, we also obtained direct evidence
that Ad12-E1A can associate with the proteins that bind to the
TATA-like sequence and to the CAA repeats in Ad12-E1A-transformed
cells.
gene, including the element CTGTAAGCCAGACCC or the TATA-like
sequence TATTAAA, was mutated to either ATGTAAGAAAGAAAC or TAGCGAA,
respectively, and ligated to the RsaI/DdeI DNA
fragment (-1,735/-1,188) or the BstNI/Sau3A fragment of a distal 5`-flanking sequence
of the H-2K
gene
(-1,615/-1,534) with 4-6-bp BamHI or HindIII linkers of both ends of the DNA. The resultant
fragments were inserted into the BamHI site or the HindIII site of pH-2K
(367)CAT (32) to
generate appropriate CAT derivatives in the correct or reversed
orientation. The TATA-like sequence in the BspHI/RsaI
DNA fragment was converted to various TATA-like sequences found in the
promoters of the H-2K, adenovirus major
late promoter (AdML), hsp70, c-fos, histone,
-globin,
adenoviruses E3 and E4, and SV40 early genes (see (33) ) by
site-directed mutagenesis (using a kit from Amersham Japan, Tokyo,
Japan) and the appropriate primers. In each case, overhang was
generated by EcoRI, and the DNA fragment was either inserted
into the EcoRI site of pH-2K(367)CAT(CAAu) or
coinserted with the DNA fragment RsaI/DdeI
(-1,735/-1,688) into either the EcoRI site of
pSV2CAT (34) or the HindIII site of pBLCAT2 (35) to generate various CAT plasmids ( Fig. 4and Fig. 5).
CAT constructs with various TATA-like sequences and a
CAA repeated motif. The histogram represents the CAT activity of the
pH-2KCAT constructs in CYpAdC3 cells or Ad5-E1A-transformed
3Y1 cells, as indicated. All CAT values are the averages of results
from at least five transfections. Normalized CAT activity associated
with pH-2K(367)CAT was taken arbitrarily as 1.0. The
standard deviation for each result is indicated by a bar.
(367)CAT in CYpAdC3 cells, which was arbitrarily taken
as 1.0. P+E is the H-2K promoter/enhancer
region. The standard deviation for each result is shown by a bar.
minimum promoter (pH-2P) and the H-2
enhancer (pH-2E) joined to the basal promoters of the early SV40,
thymidine kinsase, and conalbumin genes were generated as described
elsewhere(10) . Sequences of constructs were verified by the
dideoxy chain termination method(36) . Plasmids pE1A,
pE1A12-12S, pE1A12-13S, pE1A5-13S,
pE1A5-12S(37, 38) ,
pH-2K
(2015)CAT(9) , and pRSV-gal (39) have been described previously.Cell Culture, DNA Transfection, and CAT
Assay
Rat 3Y1 cells were cultured as described
elsewhere(32, 40) . The primary cells, such as BALB/c
baby mouse kidney cells, baby hamster kidney cells, and baby rat kidney
cells, were infected by Ad12 or Ad5 virus and cultured as described
elsewhere(6) . Transfections for long term expression were
carried out as described elsewhere(13) . The activities of CAT
and -galactosidase were assayed as described
elsewhere(9, 13) . The ratio of the activity of CAT to
that of -galactosidase and the number of copies of the transfected
genes that had been integrated into genomic DNA were used for
normalization of results.Preparation of Nuclear Extracts and Gel Shift
Assay
Nuclear extracts were prepared from various cell
lines essentially as described by Yano et al.(26) .
Gel shift and competition assays were carried out as described
elsewhere(41) . The probes were the AccI/RsaI
DNA fragment (-1,837 to -1,736) and the AccI/DdeI DNA fragment (-1,837 to -1,689)
of the H-2K promoter. Effects on antibodies
against TBP, Ad12-E1A, or Ad5-E1A were analyzed in the gel shift
assays.Gel Shift Assays with Products of Translation in
Vitro
The transcription and translation of the gene for TBP
were carried out in vitro using the expression plasmid
referred to as human TBP/pGEM1 (42) as recommended by the
manufacturer (Promega, Madison, WI). Gel shift assays with products of
translation in vitro were carried out as
described(20) . The DNA-protein complex and free probe were
separated by polyacrylamide gel electrophoresis on a 5% gel in 0.5
Tris-borate buffer (1 TBE buffer: 0.089 M Tris, 0.089 M boric acid, 2 mM EDTA). The gel
was dried and exposed to two sheets of x-ray film. The film closest to
the gel recorded signals from both S and
P
and the second film recorded only signals from
P.
Immunoselection and
CASTing
Immunoselection and CASTing were performed by a
combination of the methods of Kinzler et al.(43) and
Perkins et al.(44) with slight modifications. A DNA
fragment containing the promoter region of H-2K (HindIII-NruI) was sheared to an average size
of 200-400 bp (mass average) by sonication on ice. The ends of
the DNA fragments were blunted with the Klenow fragment of DNA
polymerase I from Escherichia coli and then ligated to
``catch linkers'' (5`-GAGTAGAATTCTAATATCTC-3`) using T4
ligase. To separate catch linkers that had become ligated to
themselves, the ligated linkers were digested with XhoI. The
reaction mixture for the first cycle of immunoselection and CASTing
included 84 µg of DNA fragments and 40 µl of a nuclear extract
of 3Y1 cells (400 µg of protein) in a total volume of 1.0 ml
of buffer (100 mM NaCl, 20 mM Hepes (pH 7.5), 1.5
mM MgCl
, 10 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 10 µg/ml pepstatin, 10
µg/ml leupeptin, 0.1% Triton X-100, 20% glycerol). After a 25-min
incubation at 20 °C, 10 µl of a solution of rat IgG were added,
and then the mixture was slowly agitated at 20 °C for 1 h after the
addition of 100 µl of a suspension of protein A-Sepharose (Sigma).
The mixture was centrifuged, and the supernatant was further incubated
with 40 µl of a nuclear extract of CYpAdC3 cells or
Ad5-E1A-transformed 3Y1 cells (400 µg of protein) for another
25 min at 20 °C. Then 10 µl of a solution of monoclonal
antibodies specific for Ad12-E1A or Ad5-E1A and 100 µl of the
suspension of protein A-Sepharose were added to the reaction mixture,
and the final mixture was treated according to the same protocol as
described above. The beads were washed five times with 1.0 ml of
isotonic saline that contained 0.1% Triton X-100 and 0.1 mg of bovine
serum albumin per ml and then twice with 1.0 ml of immunoprecipitation
buffer (50 mM Hepes (pH 7.5), 150 mM KCl, 5 mM MgCl
, 10 mM ZnSO
, 1% Triton
X-100, 0.05% SDS). DNA was freed from complexes by incubation of beads
for 50 min at 50 °C in dissociation buffer (500 mM Tris
(pH 9.0), 20 mM EDTA, 10 mM NaCl, 0.2% SDS). Protein
A-Sepharose was removed by centrifugation, the supernatant containing
DNA was extracted with phenol and chloroform, and the DNA was
precipitated with ethanol. Various amounts of recovered DNA (1/2, 1/20,
and 1/200 dilutions) were amplified using the catch linkers as primers
(25 cycles; 94 °C for 60 s, 55 °C for 60 s, and 72 °C for
120 s) as described elsewhere(43, 44) . Amplified
products were analyzed by agarose gel electrophoresis (1.2% agarose, 1
TBE), and 10 µl of a solution of the amplified material
(typically containing 30-50 ng of DNA) were used to initiate the
second and subsequent rounds of CASTing. After six cycles of CASTing,
the amplified DNA was radiolabeled by the Klenow fragment of DNA
polymerase I with [
-P]dTT(45) .
Preparative gel shift assays were performed as described
elsewhere(45) . Each shifted band, representing a protein-DNA
complex, was excised from the gel and eluted from the gel by overnight
agitation at 37 °C in 0.6 ml of a solution of 0.5 M ammonium acetate, 1 mM EDTA, 0.1% SDS, and 10 mM Tris (pH 7.5). DNA was extracted with phenol and chloroform,
precipitated in ethanol. The DNA pellet was resuspended in 10 µl of
10 mM Tris (pH 8.0), 1 mM EDTA. PCR was then
performed under the conditions described above. After the fourth cycle
of gel shifting and PCR, the resultant DNAs were digested with EcoRI and ligated into pBluescript KS(
)
(Stratagene, La Jolla, CA). The products were used to transform
competent XL-1 Blue E. coli cells.
Nucleotide Sequencing
Nucleotide sequencing was
performed by the dideoxy chain termination method as described
elsewhere(36) .
The Upstream TATA-like Sequence and the CAA Repeated
Motif
Ad12-E1A-transformed CYpAdC3 cells and baby mouse kidney
primary cultured cells, infected with either Ad12 or Ad5 virus, were
used for the promoter-screening assay. A distal 5'-end element
between -1,521 and -1,837 that contributes to negative
regulation mediated by the E1A gene products was previously defined
within the H-2K gene(9) . In an attempt to
define the precise controlling sequence in the 316-bp DNA fragment, we
constructed a set of fusions of partially deleted and substituted
regulatory regions with the pH-2K(367)CAT construct, which
contained the promoter and enhancer regions of the mouse H-2K gene, as described previously(32) .
To our surprise, the TATA-like sequence (-1,773 to 1,767) in the
316-bp far upstream region was found to contribute to the E1A-mediated
repression in Ad12-E1A-transformed cells. The plasmid del-3CAT (lane 2 (Fig. 1, lane 2)), which contained
only the TATA-like sequence, did not show this negative activity.
However, when the DNA fragment from -1,809 to -1,735 was
ligated with either an upstream CAA repeated element (-1,735 to
-1,688) or a downstream CAA repeated element (-1,615 to
-1,534), repression was detected at significant levels in
Ad12-E1A-transformed cells, but not in Ad5-infected cells (lanes 3-6). The results obtained with substitution mutations of
the TATA-like sequence in this region showed that this negative
regulation was caused by the TATA-like sequence (lanes 7 and 8). Similar results were obtained in adenovirus-infected baby
hamster kidney and baby rat kidney primary cells (data not shown).
Collectively, the results indicate that the far upstream TATA-like
sequence is one of the negative elements involved in the E1A-mediated
control of expression of the H-2K MHC class I
gene. The same results were obtained in co-transfection experiments
using 3Y1 cells and NIH3T3 cells, with CAT reporter genes and an E1A
expression vector (data not shown).
gene. A
schematic representation of the CAT constructs derived from
pH-2K
(367)CAT (32) and their relative CAT
activities is shown. Dotted lines represent internal
deletions. Relative CAT activities of mutants were measured in
Ad12-E1A-transformed CYpAdC3 cells(9) , in Ad12-infected or
Ad5-infected baby mouse kidney cells, and in uninfected baby mouse
kidney cells. Normalized CAT activity associated with
pH-2K(367)CAT was taken arbitrarily as 1.0. The boxes represent TATA-like sequences and the upstream and downstream CAA
repeated sequences (TATTAAA, CAAu, CAAd). Shaded boxes represent a mutated TATA-like sequence (TAGCGAA). Empty
oval, upstream element, 5`-CTGTAAGCCAGACCC-3` (-1,788 to
-1,774); striped oval, mutated upstream element,
5`-ATGTAAGAAAGAAAC-3` (-1,788 to -1,774). The construction
of CAT plasmids is described under ``Materials and
Methods.''
Specificity in Terms of the Type of Adenovirus and the Type of
mRNA
We studied the effects of the upstream TATA-like element
together with the CAA repeated motif in the presence or absence of
either pE1A12S or pE1A13S cDNA constructs from adenovirus type 12 and
type 5 (Fig. 2). Only the 13S mRNA from Ad12 had the ability to
regulate negatively the expression of the MHC class I promoter (panel B). Thus, the negative effect exerted by the TATA-like
sequence together with either an upstream or a downstream CAA repeated
element appears to be specific to the mRNA for Ad12-13S E1A.
-CAT constructs with the upstream
TATA-like sequence and a version of the CAA repeated motif in 3Y1 cells (panel A) and E1A-transformants obtained with Ad12-13S
E1A (panel B), Ad12-12S E1A (panel C),
Ad5-13S E1A (panel D), and Ad5-12S E1A cDNA (panel E). Normalized CAT activity associated with
pH-2K
(367)CAT in 3Y1 was arbitrarily taken as 100. All CAT
activities are the averages of results from at least five
transfections, and the standard deviation for each is indicated. A, pH-2K(2015)CAT ((9) ); B,
pH-2K(367)CAT; C, del-4m1CAT; D,
del-5m1CAT.
Study of the Influence of the Spatial Relationship
between the Upstream TATA-like and CAA Repeated Sequences
The
effects of the position and the orientation of the upstream TATA-like
sequence were studied in transfection experiments (Fig. 3A). The TATA-like sequence was placed in front
of the upstream CAA repeated motif, upstream of
pH-2K(367)CAT (promoter/enhancer) in both the original and
the reverse orientation. The CAT plasmid, which had the original
orientation but in which the TATA-like sequence was separated from the H-2K promoter/enhancer by a CAA upstream sequence (lanes 4 and 5), displayed about the same level of
suppression as the wild-type CAT construct (see Fig. 1, lane
1). By contrast, the CAT plasmid with the opposite orientation of
the upstream TATA-like element was not associated with a negative
effect (lanes 6 and 7). When the upstream TATA-like
sequence was placed at the downstream BamHI site of the vector
in the original orientation, a significant decrease in CAT activity was
observed (lanes 8 and 10). However, when the upstream
TATA-like sequence was placed at the same BamHI site of the
plasmid in the opposite orientation, no suppression was detected (lanes 9 and 11). When the position of the upstream
TATA-like element was exchanged with that of the CAA repeated motif in
the original or the reverse orientation, no significant repression of H-2K promoter activity was detected (lanes 12 and 13). Thus, the TATA-like element was active at an
increased distance from its original position, but the repression of H-2K promoter activity was sensitive to the
orientation of the TATA-like element. These results were obtained in
Ad12-E1A-transformants and were not obtained at similar significant
levels in Ad5-E1A-transformed cells.
gene. A, effects of the
relative positions and the orientations of the upstream TATA-like
sequence and the CAA repeated motif on the activity of the H-2K
promoter. The orientation of each
upstream TATA-like sequence and CAA repeated motif is indicated by an arrow. Average results of four independent experiments are
shown in the histogram as CAT activities relative to that expressed by
pH-2K
(367)CAT. Normalized CAT activities of
pH-2K(367)CAT in CYpAdC3 cells and Ad5-E1A-transformed 3Y1
cells (lane 1) were taken arbitrarily as 1.0. B,
activity of the H-2 enhancer/silencer linked to heterologous promoters
or to the homologous basal promoter in Ad12-E1A- or Ad5-E1A-transformed
cells. All CAT values are the averages of results from at least four
transfections. Normalized CAT activity associated with
pH-2K(367)CAT in CYpAdC3 cells was taken arbitrarily as
1.0. The orientations of the upstream TATA-like sequence and the CAA
repeats are indicated by arrows. TATA, upstream
TATA-like sequences (BspHI/RsaI); CAAu,
upstream CAA repeated motif (RsaI/DdeI); CAAd, downstream CAA repeated motif (BstNI/Sau3A). H-2E, H-2 enhancer; H-2P, H-2 promoter; H-2P/E, H-2 promoter/enhancer; TK, thymidine kinase promoter; SV, SV40 promoter; Cona, conalbumin promoter. The standard deviation for each
result is indicated by a bar.
gene(10) . Therefore, we next
examined whether the upstream TATA-like sequence might affect the H-2K promoter activity of CAT plasmids with
oligomerized CAA repeated motifs. To our surprise, the effect of the
TATA-like sequence on the H-2K promoter activity
was apparent in conjunction with monomeric or dimeric CAA motifs (data
not shown). This effect was also sensitive to the relative orientations
of the TATA-like sequence and the CAA motifs but not to the relative
positional effects of the two elements (data not shown).The Negative Effect of the TATA-like Element Is
Independent of the Promoter
We analyzed the influence of the
promoter on the negative effect of the upstream TATA-like element (Fig. 3B). The repression of three different constructs
driven by SV40, thymidine kinase (TK), and conalbumin (Cona) promoters was clearly observed (5.0-7.0-fold
repression), similar to the repression associated with the homologous
promoter (10-fold repression), although the extent of repression was
variable (lanes 3, 6, 9, and 12).
Thus, this negative effect was not dependent upon the specific promoter
tested in the case of Ad12-E1A-transformed cells. By contrast, the
upstream TATA-like element had only a moderate negative effect
(1.5-2.0-fold repression) in Ad5-E1A transformants.Effects of Alterations in the Upstream TATA-like
Sequence
To determine the relative abilities of various
TATA-like sequences to repress the E1A-dependent activity of the H-2K promoter, we constructed a series of CAT
plasmids with various TATA-like sequences, such as those of the
adenoviruses E3 and E4, Ad-ML, hsp70, SV40 early genes, histone H1,
c-fos, and -globin, by site-directed mutagenesis (Fig. 4). When the upstream TATA-like sequence (TATTAAA) was
changed to the basic TATA sequence of promoters (TATAAA), we observed
significant repression (lanes A and B). When the
upstream TATA-like sequence was converted to the Ad-ML (TATAAAA), hsp70
(TATAAAA), or c-fos (TATAAAA) sequences, significant
repression was detected (lane C). The conversion to E4
(TATATATA) and SV40 early (TATTTAT) sequences, did not result in
negative repression (lanes G and H). Conversion to
the histone H1 (TATATAA), the -globin (CATATAA), or the E3
(TATAACT) sequences resulted in moderate negative repression (lanes
D, E, and F). These results imply that full
negative activity of this upstream element requires the sequence TAT
(A/T) AA. The sequence specificity of the TATA-like element was seen
only in the case of Ad12-E1A-transformed cells. No significant similar
result was obtained in Ad5-E1A-transformed cells.Effects of Combinations between Upstream TATA-like
Sequences and the Basal TATA-box
We next examined the combined
effects of the upstream TATA-like sequence and the TATA-box of the
basal promoter on the E1A-dependent negative repression of the MHC
class I H-2K promoter. Various upstream TATA-like
sequences were generated by site-directed mutagenesis, as indicated
above, and were introduced into either Ad12-E1A- or Ad5-E1A-transformed
cells. We constructed CAT plasmids driven by the SV40 early (TATTTAT)
promoter or the thymidine kinase promoter (TATTAA), which are known to
be negatively and positively affected by Ad5-E1A-mediated
transactivation, respectively ((33) ; Fig. 5). In the
case of H-2K promoter/enhancer constructs, as
already shown in Fig. 4, the negative effects of the H-2K upstream TATA-like motif (lane 3),
the TATA-box (lane 2), and the TATA-box of the Ad-ML promoter
sequences (lane 4) were significant. The TATA sequence of the
SV40 early promoter did not cause any significant reduction in the
promoter activity (lane 5). In the case of the SV40 early
promoter/enhancer or the thymidine kinase promoter constructs, the same
pattern was observed (lanes 6-17). It was noteworthy
that the repression observed with SV40 and thymidine kinase reporter
constructs was specific to Ad12-E1A-transformed cells as compared with
Ad5-E1A-transformed cells (8-10-fold repression versus 1.5-2.0-fold repression) even though it has been shown that
the SV40 promoter is negatively regulated by E1A without serotype
specificity(33) . It is clear that the extent of negative
repression in Ad12-E1A transformants is much greater than that observed
in Ad5-E1A-transformed cells. Taken together, the results suggest that
the trans-repression activity is dependent upon a sequence
preference that is associated with the upstream TATA-like element and
not upon the sequences of the TATA-box close to the basal promoter
together with the CAA repeated motif in the presence of Ad12-E1A.Mutual Competition between Binding Proteins
In an
attempt to determine whether similar proteins might bind to the
upstream TATA-like sequence and the basal TATA-box in
Ad12-E1A-transformed cells, we performed a band shift assay using
31-mer oligonucleotides that corresponded to the upstream TATA-like
sequence and the proximal TATA-box sequence in the promoter of H-2K. As shown in Fig. 6A, both
DNA probes generated three major bands with extracts derived from
CYpAdC3 cells (lanes 2 and 3). Cross-competition
resulted in the disappearance of the three bands (lanes
4-7). Thus, similar or identical nuclear proteins seemed to
associate with the TATA-like sequence and the TATA-box sequence in the
promoter region of H-2K. To determine whether TBP
is directly associated with the binding proteins, a band shift assay,
using the 149-bp AccI/DdeI DNA fragment as a probe,
was carried out in the presence of TBP-specific antiserum. As shown in Fig. 6B, the two retarded bands (B1 and B2) were shifted still further, and the one retarded band (B3) disappeared upon addition of rabbit antibody against
human TBP (lanes 3-5) and mouse antibody against human
TBP (lanes 6-8) but not upon addition of the control
preimmune rabbit serum (lanes 11 and 12). Antisera
that had been preincubated with purified human TBP protein from HeLa
cells had no effect on the retarded bands (lanes 9 and 10). To confirm this result, we used a recombinant protein,
encoded by the TBP gene, that had been translated in vitro from the TBP gene in a reticulocyte lysate system (Fig. 6C, lanes 2 and 3) to
characterize the protein(s) that binds to the upstream and proximal
TATA sequences. The DNA probe specific for the upstream TATA-like
sequence gave rise to a shifted band, which could be competed out by an
excess of unlabeled oligonucleotides specific for the downstream
TATA-box (Fig. 6D, lanes 3 and 4) but
not by an excess of mutated competitors (lane 5). Taken
together, the results indicate that similar or identical proteins,
including TBP, associate with the upstream TATA-like sequence and the
proximal TATA-box sequence.
P]TATAu,
5`-ggacgctggaTATAAAgtccacgcagcccgc-3`; lanes 2, 4,
and 5, [
P]TATAd DNA probe,
5`-agccagacccTATTAAAtgtctccctttaga-3`; lanes 3, 6,
and 7, [
P]TATAu DNA probe; lanes 4 and 6, 500 ng of TATAu competitor oligomer; lanes 5 and 7, 500 ng of TATAd competitor oligomer. An arrowhead indicates the protein-DNA complex. B,
effects of TBP-specific antiserum on the shifted complex in the gel
shift assay. Lane 1, free DNA probe (AccI/DdeI); lane 2, nuclear extract from
CYpAdC3 cells; lanes 3-5, 10
,
10
, and 10
dilutions of rabbit
antiserum against human TBP (referred as anti-24-28); lanes
6-8, 10
, 10
, and
10
dilutions of mouse antiserum against human TBP; lanes 9 and 10, anti-24-28 (10
and 10
dilutions, respectively) that had been
preincubated with TBP protein (24 µg), which had been purified as
described elsewhere(42) ; lanes 11 and 12,
preimmune rabbit control serum (10
and
10
dilutions, respectively). Arrowheads indicate protein-DNA complexes (B1, B2, and B3). C, in vitro translation of
[
S]methionine-labeled TBP in a rabbit
reticulocyte lysate. Lane 1, pGEM1 without the rTBP sequence; lane 2, 500 ng of rTBP/pGEM1; lane 3, 1 µg of
rTBP/pGEM1. The arrowhead on the right indicates the
recombinant TBP. D, binding properties of in vitro translated rTBP. Gel shift analysis used
P-radiolabeled oligonucleotides (0.1 pmol) that
corresponded to the upstream TATA-like sequence in the H-2K
promoter, as described
elsewhere(42) . Approximately 20 fmol of
[
S]methionine-labeled TBP (lanes
2-5) were used in gel shift assays in the presence or
absence of a 500-fold molar excess of synthetic oligonucleotide
duplexes that corresponded to TATAu, TATAd, and a mutated TATA sequence
(TATAm; 5`-agccagacccTAAATTAtgtctccctttaga-3`). An arrowhead indicates the rTBP-DNA complex.
Immunoaffinity Selection and Amplification by PCR of the
Binding Sites of Ad12-E1A-associated Nuclear
Proteins
Immunoaffinity selection and CASTing experiments were
carried out to identify the binding site on the DNA at which the
Ad12-E1A- or Ad5-E1A-associated proteins bind. After a preparative gel
shift assay using DNA fragments isolated by immunoselection with
antiserum specific for Ad12-E1A or Ad5-E1A, we isolated the material in
the shifted bands obtained with extracts of Ad12-E1A or
Ad5-E1A-transformed cells (Fig. 7A). We purified the
DNA fragments extracted from the shifted bands and amplified the
fragments by PCR using both primers of the catch linkers. After
amplification of the DNA fragments by PCR, the DNA fragments were again
recycled for incubation in a reaction mixture that contained an extract
of Ad12-E1A or Ad5-E1A transformants. The amplified DNA fragments
obtained by immunoaffinity and CASTing experiments (Fig. 7B) were subcloned into pBluescript for DNA
sequencing. The mean size of DNAs cloned into the vector was
approximately 300 bp, which is the same as that of the DNAs used as
starting materials. Among 280 clones isolated by this CASTing method,
195 clones (70%) were shown to include DNA sequences such as CAA
repeated sequences and the upstream TATA-like sequence by use of an
extract of Ad12-E1A transformants (Fig. 8, A and B). By contrast when we started with an extract of Ad5-E1A
transformants, we found that most of the clones (178 clones; 64%)
included DNA sequences such as the TATA box of the H-2 promoter and H-2
enhancer elements (CRE/IRS) (Fig. 8B). To examine
whether these isolated clones represented DNA fragments specific for
the binding protein(s) present in extracts of Ad12-E1A-transformed
cells, we performed a gel shift assay using the DNA fragments, isolated
from this CASTing experiment, as DNA probes. As shown in Fig. 7C, when five representative DNA fragments that
included the upstream TATA-like motif and both CAA repeated motifs were
used as probes, all exhibited binding capacity specific for proteins
from Ad12-E1A transformants (lanes 6-10), but not for
proteins from Ad5-E1A-transformed cells (lanes 11-15).
Constructs mutated in the TATA-like element did not show any
significant binding (lanes 1-5). We next examined the
effect of the addition of antiserum specific for Ad12-E1A. As shown in Fig. 7D, when the DNA fragment that included the
upstream TATA-like sequence was used as probe, the shifted bands
disappeared after addition of the specific antibodies (lanes
3-6). By contrast, antibodies raised against Ad5-E1A did not
affect the mobility of bands (lanes 7 and 8).
Moreover, the antiserum against Ad12-E1A that had been preincubated
with Ad12-E1A protein did not affect the shifted bands on the gel (lane 9). Thus, we conclude that the far upstream TATA-like
sequence is one of the targets for binding by the Ad12-E1A-associated
complex, which is specific for E1A-mediated negative repression.
promoter region by immunoselection and CASTing is shown. A,
preparative gel shift assay using DNA fragments isolated by
immunoaffinity precipitation with Ad12-E1A-specific or Ad5-E1A-specific
antiserum and PCR cycling.
P-labeled DNA fragments (5
10 cpm) isolated from the sonicated DNA from the H-2K promoter region by immunoselection
and PCR cycling were incubated with an extract from CYpAdC3 cells or
Ad5-E1A-transformed 3Y1 cells. B1 and B2, DNA-protein
complexes; F, free DNA probe. Lane 1, extract of
CYpAdC3 cells; lane 2, extract of 3Y1 cells; lane 3,
extract of Ad5-E1A-transformed 3Y1 cells after passage through a
heparin column (see (10) ). B, agarose gel 1.2%
electrophoresis of the DNA fragments after immunoselection with
E1A-specific antibody and the CASTing assay. Lanes 1-6,
isolation of binding site DNAs; lane 7, HaeIII digest
of
174; lane 8, HindIII digest of
. C, gel shift assays of representative DNA fragments
isolated by immunoselection and CASTing. Lanes 1-5, DNA
fragments with mutations in the TATA-like sequence (TATTAAA) to TAGCGAA
as the gel shift probe and the nuclear extract from CYpAdC3 cells; lanes 6-10, representative examples of DNA fragments
used as DNA probes for gel shift assays with a nuclear extract from
CYpAdC3 cells; lanes 11-15, DNA probes used in lanes
6-10 as probe and an extract from Ad5-E1A-transformed 3Y1
cells. An arrowhead indicates the protein-DNA complex. D, effects of Ad12-E1A-specific antiserum on the shifted
complex. One of the representative DNA probes (5
10 cpm) isolated as described above was incubated with an extract of
CYpAdC3 cells in the presence of increasing levels of Ad12-E1A-specific
or Ad5-E1A-specific antiserum.Lane 1, free DNA
probe; lane 2, nuclear extract from CYpAdC3 cells; lanes
3-6, 10, 10
,
10
, and 10
dilutions of
Ad12-E1A-specific antiserum; lanes 7 and 8,
10
and 10
dilutions of
Ad5-E1A-specific antiserum (M73); lane 9, 10
dilution of Ad12-E1A-specific antiserum that had been
preincubated with Ad12 E1A protein (10
µg).
gene. B, size distribution
of the DNA fragments isolated by immunoselection and CASTing. The sizes
of DNA fragments cloned in pBluescript were analyzed statistically. Upper panel, for the extract from CYpAdC3 cells; lower
panel, for the extract from Ad5-E1A-transformed 3Y1 cells.
Putative cis-elements are represented in the
figure.
gene as
negative regulatory elements by using the product of E1A gene of
adenovirus type 12(9, 10) . We have now provided
evidence that the TATA-like sequence located between -1,773 and
-1,767 is also required for the E1A-dependent negative regulation
of the MHC class I H-2K gene in conjunction with
one of two CAA repeated elements. A TATA-like element with either an
upstream CAA repeat or a downstream CAA repeat is functional (see Fig. 1). However, whereas such a TATA-like element can function
independently of its relative position and distance from the promoter,
its function is dependent on its relative orientation with respect to
the CAA repeats and the TATA-like sequence. A TATA-like sequence in an
inverted orientation plus one of two CAA repeats is associated with
decreased E1A-dependent negative regulation (Fig. 3A).
We failed to observe such a negative function of a TATA-like element in
an orientation-dependent but position- and distance-independent manner
in normal cells and Ad5-E1A-transformed cells (Fig. 3A). Moreover, we found that the combined
negative activity of the upstream TATA-like sequence with CAA repeats
appeared to be specific to the 13S E1A from Ad12 and not to the 12S E1A
or the 13S E1A from Ad5 (Fig. 2).-tubulin (49) , the c-fos gene(50) , the viral E1B gene(51) , and the long
terminal repeat of human HIV(52) . A binding study in vitro demonstrated that Ad-E1A binds to TFIID(53, 54) .
The correlation between the much greater trans-activation by
the 13S E1A protein than by the 12S E1A protein (56, 57) and the greater affinity of TFIID for 13S-E1A (53, 54, 55, 56) suggests that the
E1A-TFIID interaction plays a fundamental role in E1A trans-activation in nononcogenic E1A-transformed cells.
Therefore, we attempted to change the far upstream TATA-like sequence
of the H-2 promoter to the TATA-box sequences from heterologous genes
to examine the effects of these elements on E1A-mediated negative
repression. As shown in Fig. 4, TATA-box sequences from H-2K and Ad-ML (as well as from hsp70 and c-fos genes) were all able to function as upstream TATA-like
elements in mediating the E1A-dependent repression of the H-2K promoter (lanes A, B, and C). Moreover, it is clear that the upstream TATA-like sequence
is preferentially functional in conjunction with either an upstream or
a downstream CAA element ( Fig. 1and Fig. 5). In
addition, the results of our immunoselection and CASTing studies
suggest that nuclear proteins present in Ad12-E1A-transformed cells can
bind to DNA fragments that contain the TATA-like sequence and the basal
TATA-box sequence in cooperation with E1A molecules ( Fig. 7and Fig. 8). To our surprise, we found that the majority of DNA
fragments isolated by the immunoaffinity and CASTing method by use of
an extract from Ad12-E1A transformants contained the upstream TATA-like
sequence and the CAA repeated motif. By contrast, most of the DNA
fragments isolated by use of an extract of Ad5-E1A transformants
contained the basal TATA-box and H-2 enhancer (for example, CRE/IRS
(R1/R2) sequences) (Fig. 8B). Addition of antiserum
against Ad12-E1A to the reaction mixture used for the gel shift assays
with a DNA fragment that contained the upstream TATA-like sequence as
probe resulted in the disappearance of shifted bands (Fig. 7D). Addition of antiserum against Ad5-E1A did
not have such an effect (Fig. 7D). gene in Ad12-E1A
transformants but not in Ad5-E1A transformants? Gel shift experiments
using competitive oligonucleotides designed on the basis of binding
sites, recombinant TBP, and TBP-specific antiserum revealed that
similar or identical molecules, possibly including TBP, bound to both
the upstream TATA-like sequence and the basal TATA-box sequence in
vitro and in vivo (Fig. 6). We speculate that the
similar or identical TBP (or TBP-like protein) can bind to both TATA
sequences in parent 3Y-1 cells and Ad5-E1A-transformed cells. Several
lines of experimental evidence have suggested that E1A-CR3 of
nononcogenic adenovirus binds in vitro specifically and stably
to the isolated TBP of
holo-TFIID(53, 54, 55, 56) . Recent
studies support this model and provide direct evidence that trans-activation is mediated through a direct physical
association between the E1A-CR3 domain and TBP in the holo-THIID
complex(55, 56) . We also observed that the affinity
of Ad12-13S E1A for TBP is much stronger than that of
Ad5-13S E1A protein for TBP in vitro (data not shown).
The affinity of TBP (or TBP-like protein) for the region of the
upstream TATA-like element and the CAA repeated motifs is weak in
Ad5-transformed cells, about 10-fold lower than that in Ad12-E1A
transformants (see (10) ). Recently we succeeded in
characterizing the recombinant proteins that bound to the CAA repeats
and were associated with Ad12-E1A protein in vitro (data not
shown). In the presence of Ad12-E1A, E1A associates with TBP on the
upstream TATA-like element and the TATA-box to form a protein complex
that includes TBP, proteins that bind to CAA repeats, and E1A.
Alternatively, E1A may modify the binding affinities within complexes
that include TBP to allow more efficient binding to the upstream
TATA-like sequence. In this way, the basal transcriptional activity of
the H-2K gene can be decreased. The formation of a
large protein complex including TBP and proteins that bind to CAA
repeats might be critical for association with E1A in Ad12-E1A
transformants but not in Ad5-E1A transformants. We also cannot exclude
the possibility that the distinct sequences of respective E1A proteins
may be involved in the binding to the TBP or to a TBP-like protein to
cause the conformational changes that alter the transcriptional
machinery of the H-2K gene. Preliminary studies by
glycerol gradient centrifugation indicate that a large complex of these
components derived from Ad12-E1A transformants but not from Ad5-E1A
transformants or parent cells is formed in vitro. (
)We speculate that the ability to form such a large
complex, which includes TBP, the proteins that bind to CAA repeats, and
E1A, might be critical for regulation of the H-2K gene. Kralli et al.(31) and Ge et al.(30) reported that the R2 class I enhancer element is
involved in Ad12-E1A-mediated and not in Ad5-E1A-mediated repression in
transformed cells. We cannot rule out such a possibility since we
detected moderate repression of the promoter activity of H-2 constructs
with the R1/R2 region in the presence of Ad12-E1A (see (10) ).
At this time we cannot explain the possible functional relationship
between R2BP(31) , COUP-TF, which has been identified more
recently by Liu et al.(58) , and the factors with
which we have been dealing here. It is possible that dual regulation,
involving proximal and far upstream elements, might be important for
the full negative control of the activity of the H-2 promoter by
Ad12-E1A. gene.
)))
The authors thank H. Guo for the excellent technical
assistance and A. Fujita and S. Shimizu for preparing 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:
|
|
 |
|
  J. Girdlestone Synergistic induction of HLA class I expression by RelA and CIITA Blood, June 15, 2000; 95(12): 3804 - 3808. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |