|
|
|
|||||||
|
|||||||||
(Received for publication, October 19, 1995; and in revised form, December 15, 1995) From the
The promoter and its upstream regulatory region of the mouse
cellular retinoic acid-binding protein I (crabp-I) gene were
examined in transgenic mouse embryos, a mouse embryonal carcinoma cell
line P19, and a mouse embryonic fibroblast cell line 3T6. In transgenic
mouse embryos, a
Retinoic acid (RA) ( Although crabp-I
deficient mice displayed no apparent
phenotypes(14, 15) , previous studies in transgenic
mice (16) and embryonal carcinoma cells (17) showed an
association of elevated crabp-I expression with abnormal
cellular differentiation and RA-regulated gene expression. Studies in
embryonic palate cells demonstrated that expression of RA
receptor- Consistent with the observation of weak crabp-I expression in most adult tissues, its expression is
also very weak in most cell lines examined, except in a mouse embryonic
fibroblast cell line 3T6 (11, 20) . Significant
induction of this gene has only been observed in embryonal carcinoma
cell lines, such as P19 and F9, treated with RA (20) . The
study of the mouse crabp-I genomic structure has revealed
several interesting features within a 3-kb upstream sequence, such as a
GC content of greater than 70%, 9 pairs of inverted repeats, 5 copies
of GC boxes (Sp-1 sites, GGGCGG), and several potential binding sites
for transcription factors(11, 12) . Recently, using
pharmacological treatments, we showed that RA induction of this gene
could be enhanced by 5-azacytidine (21) and
DC-erythro-dihydrosphingosine (sphinganine)(22) . The
effect of sphinganine was associated with an 870-bp DNA fragment in the
most 5`-end of the upstream region containing a putative AP-1 binding
site (TGACTCA). The effect of 5-azacytidine was examined by analyzing
the methylation status of the 3-kb upstream sequence, which revealed
hypermethylation of this region in cells where crabp-I
expression was low. Demethylation was associated with up-regulation of
this gene expression(21) . The biological activity of the 3-kb
upstream sequence was demonstrated in transgenic mouse embryos using an Escherichia coli In this study, to address both
the promoter and the upstream regulatory activities of the mouse crabp-I gene, we constructed a series of lacZ reporter fusion
genes by inserting a lacZ structural gene fragment, in frame, into the
fifth amino acid codon of the mouse crabp-I gene. The
biological activity of the full-length fusion gene was tested in both
cultured cells and transgenic mouse embryos, and systematic deletion
mutants were made to dissect minimal promoter and cell type-specific
regulatory regions. Gel retardation assays were conducted to
demonstrate specific protein-DNA interactions between the regulatory
DNA fragments and nuclear extracts of P19 and 3T6 cells.
Figure 1:
Reporter fusion gene constructs for
5`-deletion analysis of mouse crabp-I upstream region. The BamHI fragment of E. coli lacZ structure gene
(pMC1871, Pharmacia) was filled-in with Klenow enzyme and fused, in
frame, into blunt-ended KpnI site of the mouse crabp-I EcoRI genomic fragment containing exon I (12) . A fragment containing SV 40 poly(A) fragment was added
to the 3`-end of this fusion. This generated the parental vector,
designated as CRABP-lacZ, which contained 3.2 kb of crabp-I
genomic sequence including 3 kb in the upstream region. The constructs
Figure 2:
Spatial and temporal specific CRABP-lacZ
transgene expression in transgenic mouse embryos. Transgenic mouse
embryos were dissected at gestation dates of E9.5 (A), E10.5 (B), E11.5 (C), and E12.5 (D) and analyzed
for lacZ expression in whole mount embryos as described(12) .
The stained (lacZ positive) areas are indicated. m,
mesencephalon; r, rhombencephalon; s, spinal cord.
The magnification is 30, 20, 15, and 15
Figure 3:
Specific reporter activity of 5`-deletion
analysis. Promoter activity of each construct (shown in Fig. 1)
was determined as described under ``Experimental Procedures''
and represented as A
Figure 4:
Relative reporter activity of internal
deletion analysis. A, the deletion of various sequences in the
region between nt 2100 and 2600 was made by restriction enzyme
digestion from the parental construct CRABP-lacZ. The constructs
Figure 5:
Regulatory activity of region 2100/2600 on
the ras promoter. A, the 2100/2600, 2400/2600, and
2140/2400 constructs were made by ligating HindIII-SmaI, PstI-SmaI, and XhoI-PstI fragments (restriction sites shown in Fig. 1), respectively, to the 5`-end of a lacZ reporter
containing the ras promoter in the sense orientation.
Antisense constructs for each region were also made and designated as
Figure 6:
Regulatory activity of regions 2100/2600
and 2140/2400 on the thymidine kinase promoter. A, the
2100/2600 and 2140/2600 constructs were made by ligating HindIII-SmaI and XhoI-PstI
fragments (restriction sites shown in Fig. 1), respectively, to
the 5`-end of a luciferase reporter containing the thymidine kinase
promoter in the sense orientation. The antisense constructs were also
made and designated as
Figure 7:
Gel
retardation. The HindIII-PstI fragment (restriction
sites shown in Fig. 1) was labeled with
We have demonstrated that the mouse crabp-I gene
promoter, including approximately 3 kb of its upstream region, is able
to direct a lacZ reporter expression in transgenic mouse embryos. The
expression pattern agrees with results generated from several in
situ hybridization
studies(7, 8, 9, 10) , indicating
that this region contains spatial and temporal information for crabp-I
gene expression. By systematic deletion analysis in both high
expressing (3T6) and low expressing (P19) cells, the minimal promoter
is located between nt 2990 and 3020, approximately 120-150 bp
upstream from the transcription initiation site. In transient reporter
assays, a 390-bp fragment (nt 2600-2990) immediately upstream
from the transcription initiation site (nt 3140) encodes the maximal
promoter activity in the high expressing cell line 3T6, whereas a
complete 3-kb upstream sequence is needed to obtain the maximal
promoter activity in the low expressing cell line P19. Additional
sequence in the further upstream region of the 3-kb sequence appears to
have no effect on this promoter in either cell line (data not shown).
The 870-bp fragment in the 5`-end of this 3-kb fragment is important
for the maximal activity in P19 cells. This agrees with our previous
study(21) , showing the requirement of this 870-bp fragment for
optimal crabp-I expression in P19 cells treated with RA and
sphinganine, a compound known to increase AP-1 activity(30) . Common to both P19 and 3T6 cells, a strong negative effect is
observed for deletion to nt 2100, approximately 1 kb upstream from the
transcription initiation site. A further deletion of 40 bp (deletion to
nt 2140) abolishes this negative effect in both cell types. Results
from these 5`-deletion studies would suggest that the 40-bp sequence
(2100-2140) contains negative regulatory information for crabp-I gene expression. However, studies of internal
deletions (Fig. 4) show that the sequence 2100-2140 is
critical for a high level of reporter expression in both P19 and 3T6
cells, as deletion of 40 bp results in greater than 94% decrease of the
full promoter activity in both cell types. This would argue against a
negative activity of this 40-bp sequence because deletion of this
presumably negative element should have either little effect or have
abolished its negative effect. It is possible that the 2100/2140
sequence is a portion of a complex regulatory unit, which could operate
very differently depending upon the sequence in its vicinity and the
combination of available regulatory proteins. The regulatory
mechanism under physiological conditions could be much more complicated
considering the complexity of DNA structure, modification of DNA, and
multiple protein interactions in the cells. Based upon studies of
internal deletions (Fig. 4) and heterologous promoters ( Fig. 5and Fig. 6), it is suggested that fragment
2100/2600 may be a complex regulatory unit that can be affected by many
factors. It is clear that the 40-bp sequence is critical for the full crabp-I gene promoter activity in the context of the natural crabp-I gene regulatory region (as demonstrated by internal
deletion analysis in Fig. 4), and yet, in conjunction with its
3`-flanking sequence, this region becomes a negative regulatory element (Fig. 4Fig. 5Fig. 6). It is interesting that the
40-bp sequence 2100/2140 contains a putative DR5-type RARE. However, by
itself, this sequence has little effect on heterologous promoters (data
not shown). It would be interesting to determine the protein factors
bound to these sequences and how they interact with each other. Based upon studies using heterologous promoters, the sequences
derived from 2140/2400 and 2100/2600 are able to function as strong
negative regulatory elements in both cell types. However, 5`-deletion
analysis reveals positive activity of 2140/2400 when it is fused to its
natural 3`-flanking sequence (the construct P19 cells, in undifferentiated
states, express endogenous crabp-I at a very low level. The
expression can be specifically induced by RA, which is prohibited by
cyclohexamide, a protein synthesis inhibitor(19) . In contrast,
3T6 cells express endogenous crabp-I constitutively at a much
higher level, yet RA has little effect on its
expression(11, 20) . Based upon data collected from
this and other studies, a model is proposed for the regulatory elements
controlling crabp-I gene expression as shown in Fig. 8.
It is hypothesized that both positive and negative regulatory
mechanisms are needed for the control of crabp-I gene
expression. For most cell types, crabp-I gene utilizes the
minimal promoter located between nt 2990 and nt 3020, which is
constantly demethylated (21) and active. The upstream region of
this promoter contains numerous regulatory DNA elements for both
positive and negative transcription factors and their associate
proteins. The region between nt 2600 and nt 2990 contains sequence for
some positive factors that are probably present more abundantly in some
highly expressing cells such as 3T6. The 870-bp fragment of the most
5`-end contains a sequence for positive factors that can be induced by
certain drugs (such as sphinganine) in some cells (like P19). In
contrast, the region 2100/2600 contains a sequence for negative
transcription factors (such as repressors). For most cell types, either
the lack of these positive factors for sequences 1/870 and 2600/2990 or
the presence of negative factors for sequence 2100/2600 prohibited its
optimal level of expression. In the presence of RA, this promoter
activity is enhanced in certain cell types such as embryonal carcinoma
because of diminishing levels of negative factors, induction of
positive factors, or a combination of both. A total of nine inverted
repeats are present within the 3-kb region, which have the potential to
form complex structures, thereby bringing these regulatory elements to
a close proximity. This model is being tested by asking specifically if
the dissected elements can be associated with known regulatory proteins
in terms of physical interaction and biological activity. With this
information, it would be possible to begin to address how crabp-I gene expression is regulated in specific cell types
and during developmental stages.
Figure 8:
A
model for mouse crabp-I gene regulatory elements. Nucleotide
number starts from the 5`-end to the 3`-end according to previous
sequence data (12) for consistency. Putative regulatory
elements in the 3-kb upstream region, such as AP-1, RARE, and Sp-1, are
indicated above the sequence. Question marks represent unknown
factors. Transcription initiation site is indicated with a horizontal arrow under the sequence. Relative regulatory
activity of each region in P19 and 3T6, as shown above the sequence, is
arbitrarily scaled from -4 to +4 according to relative
activity detected in transient transfection studies (Fig. 3Fig. 4Fig. 5Fig. 6). The negative signs for region 2140/2400 shown in the parentheses indicate negative activity of this region fused to
the heterologous promoters.
Volume 271,
Number 9,
Issue of March 1, 1996 pp. 5073-5078
©1996 by The American Society for Biochemistry and Molecular Biology, Inc.
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
-galactosidase reporter gene under the control of crabp-I promoter and its upstream regulatory region displayed
a very specific pattern of expression characteristic of crabp-I gene expression during developmental stages. In tissue
culture systems, the minimal promoter of this gene was identified, and
regions containing positive and negative regulatory activities were
dissected from the upstream 3-kilobase sequence using assays for
transient reporter activity. It is concluded that the minimal promoter
of the mouse crabp-I gene is located between 120 and 150 base
pairs upstream from the transcription initiation site. Several cell
type-specific positive and negative regulatory regions for this
promoter have been identified. A region encoding a common negative
regulatory activity in both P19 and 3T6 cells is also inhibitory to two
heterologous promoters, and specific protein-DNA interactions between
this DNA fragment and nuclear extracts of P19 and 3T6 are demonstrated
by gel retardation experiments.
)exerts pleiotropic effects in
animals, and the effects are mediated through various cellular
components. The RA receptors and retinoic acid X receptors are
transcription factors that regulate gene expression in response to RA
(For review, see (1) and (2) ), whereas a group of
cellular retinoic acid-binding proteins (CRABPs) are believed to be
involved in metabolic pathways of RA (For review, see (3) and (4) ). crabp-I is ubiquitously expressed in adult
tissues at a very low basal level and is highly expressed in several
RA-sensitive tissues such as the eye and the
testis(5, 6) . In embryos, strong expression of this
gene is also spatially and temporally specific to tissues that are most
sensitive to RA, especially the central nervous
system(7, 8, 9, 10) . Based upon the
promoter sequence, the mouse crabp-I gene has been
characterized as a house keeping gene(11) . However, its
upstream region contains numerous inverted repeat sequences and
putative binding sites for transcription factors, suggesting that a
complex regulatory mechanism may be involved in its cell- and
stage-specific expression (12) . The bovine crabp-I
gene has also been characterized (13) , and it appears that
both the exon/intron junctions and the promoter region of this gene are
highly conserved among animal species., TGF-3, and tenascin was altered as a result of
introduction of anti-crabp-I oligonucleotides(18) .
Recent biochemical studies provided more evidence for a role of crabp-I in RA catabolism(19) . It is suggested that
the level of crabp-I expression must be tightly controlled
because abnormally high level of expression may disturb RA
concentration, thereby affecting gene expression in specific cells at a
critical time(16) .-galactosidase (lacZ)
reporter(12) . However, the transgene expression pattern
differs slightly from the endogenous crabp-I expression
pattern detected by in situ hybridization(8) ,
possibly due to the use of a heterologous DNA fragment, the mouse
Hox1.3(23) , in the fusion.
Techniques for Cell Cultures
P19 cells were
maintained in 
-minimal essential medium supplemented with 2.5%
fetal calf serum, and 7.5% calf serum and 3T6 cells were maintained in
Dulbecco's modified Eagle's medium supplemented with 10%
calf serum as described (11, 20) .lacZ Reporter Constructs
A reporter construct was
made by ligating an E. coli lacZ structural gene containing an
SV40 poly(A) site (22) to a 3.2-kb crabp-I genomic DNA
fragment at the KpnI site within exon I(12) . This
fusion gene was first tested in P19 cells as well as transgenic mouse
embryos and used as the parental vector for making fusion genes
systematically deleted in specific upstream sequences.Cell Transformation and Quantitation of lacZ and
Luciferase Reporter Activity
Cells were plated in 24-well plates
(5 
10
cells/well) and transformed with reporter
plasmid DNA using the calcium phosphate precipitation method. Lac
activity was determined between 24 and 40 h, using
orthonitrophenyl--D-galactopyranoside (Sigma) as
substrate as described previously(22, 24) . Protein
concentrations were determined using a Bio-Rad protein assay kit.
Transient lacZ activity was represented as A/µg total protein. For internal deletion
and heterologous promoter fusion, the activity of each fusion gene was
represented as a percentage of the parental vector activity. Luciferase
activity was determined with a Promega luciferase assay system and
represented as relative luciferase unit by normalizing to a
cotransfected lacZ control activity in each transfection. For all the
assays, triplicate cultures were used in each experiment, and three
independent experiments were conducted to obtain the means and standard
errors of the mean (S.E.).
Transgenic Mice Production and Enzyme Histochemical
Analysis of lacZ Expression in Transgenic Mouse
Embryos
Microinjection was performed according to a standard
procedure (25) in the transgenic mouse center at the University
of Minnesota. Transgenic mice were identified by Southern blot analysis
of tail DNA. In situ detection of lacZ reporter gene
expression was conducted as described (12) . Stained embryos
were examined and photographed using a Nikon stereoscope SMZ-10 photo
system.Gel Retardation Assay
Gel retardation experiments
were modified from an established protocol(26) . Nuclear
extracts from P19 and 3T6 cells were prepared using the method of
Standke et al.(27) . Briefly, the cell pellet was
resuspended in a hypertonic solution (20 mM HEPES, pH 7.6, 10
mM KCl, 1 mM MgCl
, 0.5 mM dithiothreitol, 0.1% Triton X-100, 20% glycerol) and lysed by 50
strokes in a glass Dounce homogenizer. The homogenate was centrifuged
at 2,000 g for 5 min, and the pelleted nuclei were
extracted with 0.4 M NaCl by gently inverting at 4 °C for
30 min. Nuclear extract was preincubated in a solution containing 10
mM Tris (pH 7.5), 10 mM NaCl, 1 mM dithiothreitol, 1 mM EDTA, 1 µg of poly(dI-dC), 1
µg of salmon sperm DNA, and 5% glycerol at room temperature for 10
min. Cold competitor oligonucleotides, when included, were added during
this preincubation period. Oligonucleotides were labeled with
[-P]dCTP using Klenow fragment; a total
radioactivity of 2
10
cpm (prepared from
approximately 0.2 ng of DNA fragment) was added to each reaction, and
the mixture was incubated at room temperature for 20 min. The
protein-DNA complexes were analyzed on a 5% polyacrylamide gel
containing 5% glycerol in a low ionic strength buffer (6.7 mM Tris, pH 7.9, 3.3 mM sodium acetate, 1 mM EDTA).
Reporter Constructs for Generating Systematic Deletion
in Mouse crabp-I Gene Upstream Region
The lacZ fusion gene
containing the complete 3-kb upstream sequence of the mouse crabp-I gene was made as described under ``Experimental
Procedures'' and designated as CRABP-lacZ. From this parental
vector, systematic deletions were made using available restriction
sites or by polymerase chain reactions as shown in Fig. 1, where
the numbering system has been adopted from our previous study (12) for consistency. Within the 3-kb upstream sequence
(sequence reported in Fig. 1of (12) ), several
potential sequences for regulatory protein binding have been identified
by sequence comparison to a transcription factor data base, including
an AP-1 site between nucleotide (nt) 758 and 766 (TGACTCA), an
imperfect RA response element (RARE) of DR5 type between nt 2113 and
2131 (CCATGAAGGAAAAGTGA), 5 copies of Sp-1 binding sites between nt
2902 and 3075 (GGGCGG), and 9 inverted repeat sequences scattered in a
region between nt 85 and 2920(12) . Construct 870 is
deleted in the 5`-end 870 bp, deleting the putative AP-1 binding site
and the first four repeats. The
1960 construct is deleted for
three more inverted repeats, and the
2100 construct is further
deleted for one repeat. The
2140 construct is further deleted for
a 40-bp fragment containing a putative RARE. The
2400 construct is
further deleted for approximately 200 bp, and the
2600 construct
contains only the last inverted repeat and 5 copies of Sp-1 sites. The
2990,
3020, and
3110 constructs retain only a minimal
DNA fragment, each containing 4, 3, and 0 copies of Sp-1 binding sites,
respectively.
870,
1960,
2990,
3020, and
3110 were made
using PCR-amplified fragments, and the constructs of
2100,
2140,
2400, and
2600 were made by HindIII (H), XhoI (X), PstI (P),
and SmaI (S) digestion, respectively. The nucleotide
position was numbered from 5` to 3`-end, according to the published
sequence ( Fig. 1in (12) ) for consistency. A filled
triangle indicates the putative AP-1 site, and an arrow under the CRABP-lacZ construct indicates the putative RARE. Above
the constructs, detailed features of the promoter and its immediate
5`-flanking region are shown. Five vertical bars indicate the
five Sp-I sites, and a horizontal arrow indicates the position
of transcription initiation. Translation initiation codon (ATG) is
indicated at nt 3233. -, CRABP-I region;
, CRABP-I
coding region;
, lacZ; &cjs2110;, SV40 poly(A)
site.
Promoter Activity of CRABP-lacZ in Transgenic Mouse
Embryos
The full-length fusion gene, CRABP-lacZ, was introduced
into transgenic mice. In situ lacZ analysis of two independent
lines was conducted systematically, which revealed an identical pattern
of reporter gene expression (Fig. 2). lacZ expression is
detected in the mesencephalon of embryos as early as E9.5 (A)
and extends to rhombencephalon and spinal cord at E10.5 (B).
By E11.5, the expression diffuses to a wider area in the mesencephalon,
rhombencephalon, and spinal cord, but the overall intensity of
expression is reduced (C). By E12.5, the expression decreases
dramatically with weak stain remained in the roof of the midbrain (D). This pattern agrees very well with the results of several
studies using in situ hybridization (7, 8, 9, 10) . Thus, it is
concluded that this fusion gene, driven by the mouse crabp-I
promoter and the complete 3-kb upstream sequence, contains information
needed for its spatially and temporally specific expression in mouse
embryos.
, for A, B, C, and D,
respectively.
Minimal Promoter Activity and Cell Type-specific
Regulatory Activity of crabp-I Gene Upstream Sequence
To
determine regulatory activities of various 5`-upstream sequences and to
shed light on factors responsible for different levels of expression in
high expressing and low expressing cells, specific lacZ activity of
each fusion gene was determined in P19 (a low expressing cell line, solid bars) and 3T6 (a high expressing cell line, open
bars) cells (11, 20) and represented as A/30 µg of protein as shown in Fig. 3. The construct
2990 represents the shortest fusion
gene active in both cell lines, whereas
3020 or
3110 has no
activity in either cell line. Thus, it is concluded that the same
minimal promoter is used in both high expressor (3T6) and low expressor
(P19), and the active promoter region is located between nt 2990 and
3020, approximately 120-150 bp upstream from the transcription
initiation site (nt 3140). By adding 390 bp containing one more Sp-1
site (the
2600 construct), the reporter activity increases
2-3-fold in P19 cells, and 8-9-fold in 3T6 cells. Thus,
sequence between nt 2600 and 2990 contains strong positive DNA elements
for crabp-I expression in the high expressor (3T6). The
2140 and
2400 constructs have a 4- and 2-fold, respectively,
higher activity over the construct
2600 in P19 cells, yet they are
both much weaker than the
2600 construct in 3T6 cells. Thus, this
region contains relatively strong positive regulatory activity in P19
cells yet slightly inhibitory activity in 3T6 cells. Surprisingly, a
dramatic decrease in reporter activity has been observed in both cell
lines for
2100, which contains an imperfect RARE of DR5 type
between nt 2100 and 2140. In 3T6 cells, this negative effect is
abolished when more 5`-sequence is added to the reporter (
1960,
870, and the full promoter constructs). In P19 cells,
1960
and
870 remain very weak, and a maximal promoter activity is
obtained only when an 870-bp fragment of the most 5`-end is added (the
full-length construct). Therefore, these 5`-deletion mutants
demonstrate both common and cell type-specific regulatory regions of crabp-I gene expression in transient assays. In both cell
lines, a common minimal promoter is utilized, between nt 2990 and 3020,
and a dramatic reduction in promoter activity is observed for the
construct
2100. In the high expressor (3T6), the region
responsible for the maximal transient promoter activity is located
between 150 and 540 bp upstream from the transcription initiation site
(between nt 2600 and 2990). In the low expressor P19, the same region
encodes approximately of the maximal transient promoter activity of
the full-length construct. Some positive activity is encoded in the
region between nt 870 and 2100 for 3T6 cells, whereas a strongly
positive element is located in the first 870 bp for P19 cells.
/30 µg of protein.
Triplicate cultures were used in each experiment, and three independent
experiments were conducted in P19 (solid bars) and 3T6 (open bars) cells to obtain the means (A
/30 µg protein) and S.E.
values.
Effects of Deleting Specific DNA Sequence in the Region
between nt 2100 and 2600
The study shown in Fig. 3suggested that the sequence between nt 2100 and 2140 had a
profound negative regulatory activity. To study this region in more
detail, various mutants deleted internally in this region were
constructed (Fig. 4A) and tested in P19 and 3T6 cells (Fig. 4B). Specific lacZ activity of each internal
deletion was compared to the parental vector activity (CRABP-lacZ) and
represented as relative activity (%). In P19 cells (solid
bars), deletion of 2100/2140, 2100/2400, or 2100/2600 inhibits the
reporter activity dramatically (greater than 90%), and deletion of
2140/2400 or 2140/2600 slightly decreases the reporter activity
(approximately 40%). In 3T6 cells (open bars), deletion of
2100/2140 and 2100/2600 also inhibits the reporter activity
dramatically, yet all the other deletions show a slightly positive
effect. Thus, the two overlapping sequences, 2100/2140 and 2100/2600,
are critical for a strong promoter activity in both cell types, as
deletion of either region results in dramatic decrease in the reporter
activity. Yet, 5`-deletion analysis shows a profound negative effect of
the sequence between nt 2100 and 2140 when it is situated 5` to its
natural 3`-flanking sequence. This would suggest that this sequence
could not be a simple positive or negative element, and the sequence in
its vicinity is important for the regulatory activity.
2100/2140,
2100/2400,
2100/2600,
2140/2400, and
2140/2600 were made by using HindIII-XhoI, HindIII-PstI, HindIII-SmaI, XhoI-PstI, and XhoI-SmaI digestion
(restriction sites shown in Fig. 1), respectively. A filled
triangle indicates the putative RARE between nt 2100 and 2140. B, relative reporter activity of each construct was
represented as the percentage of the parental construct CRABP-lacZ
activity in P19 (solid bars) and 3T6 (open bars), and
three independent experiments were conducted to obtain the means and
S.E. values.
Positive and Negative Effects of Regulatory Sequence
between nt 2100 and 2600 on Heterologous Promoters
Deletion
studies shown in Fig. 3and Fig. 4suggested that the
most dramatic regulatory activity of CRABP-I promoter common to both
cell types was encoded in a fragment between nt 2100 and 2600. This
region was further examined using heterologous promoters. A human
Harvey ras promoter(28) , constitutively active in P19
and 3T6 cells, was first fused to the 40 bp (2100/2140). No significant
effect was observed in either P19 or 3T6 cells (data not shown).
Therefore, various fragments were dissected from region 2100/2600 and
fused, in both orientations, upstream to the ras promoter (Fig. 5A) and tested in P19 (solid bars) and
3T6 (open bars) as shown in Fig. 5B. It is
clear that 2140/2400 and 2100/2600 fragments are able to repress ras promoter, regardless the orientation of the fusion, in
both P19 and 3T6. In contrast, 2400/2600 sequence has a slight
enhancing effect on this promoter. To determine if the negative effects
on the ras promoter could be reproduced on other heterologous
promoters, a luciferase reporter driven by a thymidine kinase promoter (29) was used (Fig. 6A) and tested (Fig. 6B). It appears that both fragments, in either
sense or antisense orientation, are also inhibitory to the thymidine
kinase promoter in both P19 (solid bars) and 3T6 (open
bars). Therefore, it is concluded that the 2100/2600 fragment and
a minimal 2140/2400 sequence are inhibitory to heterologous promoters
regardless the orientation of the fusion.
2100/2600,
2400/2600, and
2140/2400. B,
relative reporter activity of each construct was represented as the
percentage of the parental vector ras-lacZ activity in P19 (solid
bars) and 3T6 (open bars), and three independent
experiments were conducted to obtain the means and S.E.
values.
2100/2600 and
2140/2400. B,
relative reporter activity of each construct was represented as the
percentage of the parental vector thymidine kinase-luciferase activity
in P19 (solid bars) and 3T6 (open bars), and three
independent experiments were conducted to obtain the means and S.E.
values.
Gel Retardation Experiment
To determine specific
DNA-protein interactions between the common negative regulatory region
2100/2600 and nuclear proteins, gel retardation experiments were
performed. Multiple bands were observed when the 2100/2600 fragment was
used (data not shown). Using the sequence 2100/2400 as the probe, a
single specifically retarded band was observed as shown in Fig. 7, where various amounts of unlabeled probe have been
included as the competitors. A specifically retarded band is observed
using nuclear extracts from both P19 (lane 2) and 3T6 cells (lane 6). This band, representing protein-bound fragment,
diminishes gradually when more competing probes are added (lanes
3-5 for P19 and lanes 7-9 for 3T6). Thus, the
2100/2400 sequence can be bound by specific nuclear proteins present in
both P19 and 3T6 nuclei.
P with
Klenow enzyme and tested for binding to specific proteins isolated from
P19 and 3T6 nuclei as described in the text. The sample order is as
follows: 1) no nuclear extract (probes alone), 2) 10 µg of P19
extract, 3) 10 µg of P19 extract + 10
unlabeled
fragment (cold competitor), 4) 10 µg of P19 extract +
100 unlabeled fragment, 5) 10 µg of P19 extract +
500 unlabeled fragment, 6) 7 µg of 3T6 extract, 7) 7 µg
of 3T6 extract + 10 unlabeled fragment, 8) 7 µg of 3T6
extract + 100 unlabeled fragment, and 9) 7 µg of 3T6
extract + 500 unlabeled fragment. Arrow head indicates the position of the specifically retarded
band.
2140) in P19 cells.
This also suggests that the sequence 2140/2400 constitutes a portion of
a complex regulatory unit. Interaction between this sequence and its
neighboring sequences determines the final regulatory activity of the
whole unit. Likewise, sequence 2400/2600 has little effect on
heterologous promoters, yet, when situated in its natural position, it
is able to increase and inhibit crabp-I promoter activity in
P19 and 3T6 cells, respectively (Fig. 3). Consistent with these
results, gel retardation experiment (Fig. 7) shows that fragment
2100-2400 can be bound by specific nuclear factors that are
present in both P19 and 3T6 cells, and the protein-bound fragments
migrate at the same position in both cases. In contrast, very different
and more complex patterns of band shift have been observed for
2100/2600 fragment (data not shown).
)
-We thank Dr. K. Roberts for critical reading of
this manuscript. We thank C.-H. Lee and Y. Lin for excellent technical
help.
©1996 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. Bi, X. Hu, H. H. Loh, and L.-N. Wei Regulation of Mouse {kappa} Opioid Receptor Gene Expression by Retinoids J. Neurosci., March 1, 2001; 21(5): 1590 - 1599. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. L. Means, J. R. Thompson, and L. J. Gudas Transcriptional Regulation of the Cellular Retinoic Acid Binding Protein I Gene in F9 Teratocarcinoma Cells Cell Growth Differ., February 1, 2000; 11(2): 71 - 82. [Abstract] [Full Text]
|
|
|
|
 |
|
 L.-N. Wei, X. Hu, J. Bi, and H. Loh Post-Transcriptional Regulation of Mouse kappa -Opioid Receptor Expression Mol. Pharmacol., February 1, 2000; 57(2): 401 - 408. [Abstract] [Full Text]
|
|
|
|
 |
|
 C.-H. Lee, C. Chinpaisal, and L.-N. Wei A Novel Nuclear Receptor Heterodimerization Pathway Mediated by Orphan Receptors TR2 and TR4 J. Biol. Chem., September 25, 1998; 273(39): 25209 - 25215. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Lu, H. H. Loh, and L.-N. Wei Studies of Dual Promoters of Mouse kappa -Opioid Receptor Gene Mol. Pharmacol., September 1, 1997; 52(3): 415 - 420. [Abstract] [Full Text]
|
|
|
|
|
|
L. Chang and L.-N. Wei Characterization of a Negative Response DNA Element in the Upstream Region of the Cellular Retinoic Acid-binding Protein-I Gene of the Mouse J. Biol. Chem., April 11, 1997; 272(15): 10144 - 10150. [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 |