|
|
|
|||||||
|
|||||||||
(Received for publication, April 12,
1994; and in revised form, November 7, 1994) From the
Scatter factor/hepatocyte growth factor (SF/HGF) and its
receptor c-Met represent a paracrine signaling system involved in
mesenchymal-epithelial interactions during development and during tumor
progression. We have examined the promoters of the mouse and human
SF/HGF genes by deletion mapping followed by CAT assays as well as by
gel retardation and footprinting analysis. The promoter sequences are
highly conserved (89.5% identity) up to position -453 from the
major transcription start site but diverged considerably further
upstream. Both promoters are active in mesenchymal but not epithelial
cells thus reflecting the expression pattern of the SF/HGF gene in
cells in vitro and in vivo. We have here identified
two regulatory sequences in the SF/HGF promoter: a negative element at
positions -239 to -258 and a positive element near the
major transcription start site; specific deletions destroyed the
activities of these elements. We were not able to localize elements on
the SF/HGF promoter region that mediate the previously described
effects of transforming growth factor Mesenchymal-epithelial interactions are essential for organ
development and regenerative processes in vertebrates, and disturbances
of these interactions play a major role in various diseases including
cancer. The biology of mesenchymal-epithelial interactions have been
extensively studied, and it is recognized that a variety of mesenchymal
factors participate in the regulation of epithelial cell growth,
differentiation, and morphogenesis(1, 2) . Less is
known, however, about the molecular nature of the signals between
mesenchyme and epithelium; these may involve cell adhesion molecules,
components of the extracellular matrix, or secretory factors produced
by mesenchymal cells and acting on epithelia in a paracrine manner
(reviewed in (3) ). Among the latter, there exist several
ligands for epithelial receptor tyrosine kinases, e.g. scatter
factor/hepatocyte growth factor (SF/HGF), ( Scatter
factor/hepatocyte growth factor, the specific ligand for the c-met
receptor, is a 90-kDa secreted glycoprotein, which consists of
disulfide-linked heavy (H) and light (L) chains generated by
proteolytic cleavage from a single precursor molecule: the H-chain
contains an N-terminal hairpin structure and four kringle domains, the
L-chain is an inactive serine protease due to replacement mutations in
2 out of 3 resides of the catalytic
site(11, 12, 13, 14, 15, 16, 17) .
The structure of SF/HGF is thus similar to blood proteases such as
plasminogen but not to other known ligands for receptor type tyrosine
kinases. Two distinct activities, the ability to induce proliferation
and movement of epithelial cells, have been used to independently
isolate and molecularily characterize the
factor(5, 10, 11, 12, 13, 14, 15, 16, 17) .
SF/HGF can also increase invasiveness of epithelial cells and acts as a
cytostatic factor on certain other
cells(10, 18, 19) . In addition, SF/HGF is an
inducer of epithelial tubulogenesis in vitro(20) :
When Madin-Darby canine kidney epithelial cells are cultured in
collagen gels together with SF/HGF, they rapidly proliferate and form
complex networks of branching tubules. In situ hybridization
analysis demonstrated that during mouse development, the c-met receptor
is expressed in many epithelia whereas transcripts for the ligand
SF/HGF are preferentially found in nearby mesenchymal
cells(6) . SF/HGF also plays an important role in liver
regeneration since in animals, plasma levels of the factor and cellular
mRNA are elevated after partial hepatectomy or liver damage induced by
hepatotoxins(21, 22, 23, 24) . Thus,
SF/HGF and c-met constitute a paracrine signaling system, a concept
originally proposed by Stoker and
colleagues(5, 6, 25) , that is acting during
organ development and regeneration. SF/HGF is also expressed in
distinct embryonal muscle and brain cells(6) . Little is
known about the regulation of SF/HGF expression in mesenchymal cells.
It has been shown that in MRC5 fibroblasts, SF/HGF expression is
down-regulated by TGF
Fragments
spanning the human promoter were cloned after PCR amplification using
5`-nested primers (35) and human placenta DNA (donated by Dr.
Ilse Wieland, University of Essen); 5` primer (h-991/-10179):
CTCCTGCAGGATTTCCGGTGAAAGTCAGTCCTAACC; 5` primer (h-345/-372):
CTCCTGCAGCTGCCTGTGCCTTGATTTAGCCATTGG; 3` primer (h+32/+58):
CCAGGCATCTCCTCCAGAGGGATCCGCTCTAGACTC. After digestion with the
restriction enzymes XbaI and PstI, the fragments were
cloned into Bluescript SK+ and pCAT-basic (Promega). The correct
sequence was confirmed by sequencing with T7 DNA polymerase (Pharmacia
Biotech Inc.). The human promoter CAT constructs had a common 3`-end at
position +58. Relative promoter activities of the various
constructs were estimated by comparison with the promoter activity of a
CAT plasmid containing the simian virus 40 promoter/enhancer CAT gene
(Promega). CAT assays were performed after transient transfection using
the calcium-phosphate method, and activities were quantified as
described(37) . A control plasmid containing the Rous sarcoma
virus promoter and the Escherichia coli lacZ gene was
cotransfected in each experiment. The amount of cell extracts used in
CAT assays was adjusted according to the
Figure 1:
Sequences of the mouse (m) and human
(h) SF/HGF promoter. Genomic sequences are aligned in the region
between nt -514 and +113 relative to the most proximal
transcription start site of the mouse gene. The ATG codon marked by dots represents the translation start site; nucleotides marked
by stars represent the identity between the two sequences. Arrows mark the major transcription start site of the human
gene (33) and the identified transcription start site of the
mouse gene, which corresponds to the major and minor start sites of the
rat gene(34) .
Figure 2:
Relative activities of human (h)
and mouse (m) SF/HGF promoter CAT constructs. Specific
transcript initiation in fibroblasts. CAT assays are shown for ras3T3
fibroblasts (A) and MCF7 breast epithelial (carcinoma) cells (B). The activities of various deletion constructs were
compared with those of the SV40 promoter/enhancer construct and of pCAT
basic. Numbers indicate the position of the 5`-end of the
promoter fragment used, relative to the major transcription start site.
The m-755 construct of the mouse promoter was also tested in the
reverse orientation (m-755r). C, scheme of the
experimental design for the RNase protection assay. The HindIII-PvuII fragment from the chimeric
-150/+34 SF/HGF-CAT fusion gene was cloned into the HindIII-Sma sites of Bluescribe M13+ and
linearized with BglII as indicated to generate the specific
run-off transcript. Relative positions of transcriptional start sites
are marked with upward-pointing arrows. Hybridization of the
specific antisense probe to RNA of transient transfectants yielded
protected fragments of 270, 220, and 200 nt. Fragments of 113 nt
correspond to properly initiated transcripts of the cotransfected SV40
promoter/enhancer plasmid. D, RNase protection assay of
various transfected constructs in ras3T3 or MCF7 cells as indicated
above the slots. The size marker M is pBR digested with MSPII.
The input lane (I) shows the T7 antisense probe. The protected
fragments corresponding to transcripts initiated at either the SF/HGF
or the SV40 promoter are indicated by arrows.
In order to confirm that the observed differences in CAT activity
result from a transcriptional effect, we mapped by RNase protection
chimeric CAT-mRNA transcripts in fibroblasts (ras3T3) and epithelial
cells (MCF7) transiently transfected with the mouse SF/HGF promoter.
The antisense riboprobe that we used recognizes the N-terminal part of
the CAT gene and mouse SF/HGF promoter sequences up to position
-70 (Fig. 2C). Cells were transfected with the
SV40 promoter/enhancer-driven CAT gene alone or in combination with the
SF/HGF promoter construct m-365. We detected three protected fragments
in ras3T3 fibroblasts but none in epithelial cells (Fig. 2D). This indicates tissue-specific transcription
from the SF/HGF promoter in agreement with the results obtained in the
CAT assays. The major protected fragments of 200, 220, and 270
nucleotides in length correspond to transcript initiation sites that
have been mapped for the rat SF/HGF gene(34, 42) .
Figure 3:
Activity of progressive 5` deletions of
the mouse SF/HGF promoter. On the left, the transfected chimeric
deletion CAT constructs are displayed; the numbers indicate
the length of the constructs with respect to the major transcription
start site. On the right, relative CAT-activities of the constructs in
fibroblasts (ras3T3) and epithelial cells (MCF7) in comparison to the
SV40 promoter/enhancer are shown. Major and minor transcription start
sites are indicated by the two arrows.
It has previously been shown that SF/HGF mRNA levels in
fibroblasts are modulated after treatment with TGF
Figure 4:
Nuclear factor binding to the SF/HGF
promoter by footprint analysis. Fragments of the mouse SF/HGF promoter
were labeled at position -239 (left picture, non-coding
strand) and -70 (right picture, coding strand) and
subjected to DNaseI footprint analysis using nuclear extracts of ras3T3
fibroblasts and MCF7 epithelial cells (see ``Materials and
Methods''). The lanes marked(-) indicate digestion
in the absence of nuclear extract. The lanes G and G
+ A are Maxam-Gilbert sequence reaction products. The region
specifically protected in fibroblasts (+14 to -7 and -21 to -70) and the
region protected in both cell lines (-229 to -258) are schematically displayed on the left and right
side, respectively.
The
region around the major transcription start site was also examined by
gel retardation analysis with nuclear extracts from various cell lines (Fig. 5). Using an oligonucleotide spanning positions -7
to +34 and nuclear extracts of fibroblasts (ras3T3 and NIH3T3),
three major retarded complexes were detected (arrowheads). The
formation of these complexes was competed by the unlabeled
oligonucleotide but not by an oligonucleotide from positions +14
to +34 or by an unrelated oligonucleotide (E-Pal). Interestingly,
extracts of MCF7 epithelial and neuro 2A cells did not form the complex
of intermediate size (large arrowhead).
Figure 5:
Nuclear factor binding to the region
covering the transcription start site of the mouse SF/HGF promoter by
gel retardation analysis: Difference between SF/HGF-producing and
non-producing cell lines. A, schematic representation of the
radiolabeled oligonucleotide probe (-7 to -34) used for gel
retardation assays. B, gel retardation assay using nuclear
extracts from ras3T3 fibroblasts and MCF7 epithelial cells. The
specific competitor was the unlabeled -7 to +34 oligonucleotide, a second oligonucleotide was from position +14 to +34, and a nonspecific competitor
was from the E-cadherin promoter (E-Pal, cf. (37) ). Unlabeled oligonucleotides were used at 50-fold molar
excess. C, gel retardation assay using nuclear extracts from
3T3 fibroblasts and neuro 2A (neuroblastoma) cells. Conditions were as
in B.
Figure 6:
Deletion of the promoter regions presumed
to be important for positive and negative regulation. A, the
region -258 to -239 specifically
protected in footprint analysis (cf.Fig. 4) was
deleted from the m-291 construct, and B, the region -66 to +34 was deleted from the mouse
-755 construct. C and D, CAT assays showing the
effects of these two deletions. CAT activities of the constructs m-755,
m-150, and pCAT basic are shown for
comparison.
In the present investigation we examined, by promoter
analysis, the regulation of the scatter factor/hepatocyte growth factor
gene which is, in vivo, expressed in mesenchymal (and some
neuronal) but not in epithelial cells(6) . The target of SF/HGF
is the Met receptor tyrosine kinase which is predominantly produced by
epithelial and endothelial but not mesenchymal cells. We show here that
the activity of our SF/HGF promoter constructs is restricted to
mesenchymal cells, as shown by CAT and RNase protection assays, and we
have identified positive and negative regulatory elements in this
promoter fragment (Fig. 7). The positive regulatory elements are
located around the major transcription start site and upstream of
position -291, a negative regulatory element is located at
positions -239 to -258. Our work has failed to provide
evidence for a role of the IL6 and TGF
Figure 7:
Functionally important sequences in the
SF/HGF promoter. Large (shadowed) boxes,
positive and negative regulatory regions identified by deletion and
footprint analysis. Arrows indicate major and minor
transcription start sites. HLH, NF1, hAPF1,
and AP-1 are consensus binding sites for helix-loop-helix
transcription factors, nuclear factor-1, human interleukin 6-dependent
transcription factor, and AP-1 transcription factor. TGF
The negative regulatory element in the SF/HGF promoter was
uncovered due to a 3-4-fold drop of promoter activity in a 5`
deletion experiment and was also detected in our footprint analysis as
a region of factor binding (positions -258 to -229).
Indeed, internal deletion of the protected sequences fully released the
inhibitory effect of the element. This sequence, therefore, is a
negative regulatory element for which some preliminary evidence may
have been presented by others(41) . Interestingly, the sequence
contains putative binding sites for helix-loop-helix transcription
factors and nuclear factor-1 (42) as well as a previously
identified palindrome, P2(33) . However, this element appears
to be largely promoter context dependent since it had no negative
influence on the heterologous SV40 and TK promoters; to our surprise,
it stimulated the activity of the E-cadherin promoter when inserted at
position -78 (data not shown, cf.(37) ). A
promoter context-dependent element with similar characteristics has
also been described in the human erbB-2 promoter(43) .
Footprint analysis also uncovered nuclear factor binding sites to
regions close to the transcriptional start sites, which are specific
for extracts of mesenchymal cells. The 5` deletion analysis revealed
that the region between positions -7 and +14 is of
particular importance, and gel retardation experiments with an
oligonucleotide from this region showed the formation of a specific
protein Much effort has here been undertaken
to identify elements in the SF/HGF promoter which reflect the effects
of various modulators of SF/HGF expression in vivo and in cell
culture, such as TGF Future experiments in transgenic animals will
show whether the SF/HGF promoter elements identified in the present
study are sufficient to generate the expression pattern as seen with
the endogenous gene. In particular, it will be interesting to examine
whether the SF/HGF cDNA driven by this promoter fragment can rescue the
lethal phenotype of mice with homozygous deletion of the SF/HGF gene. ( The nucleotide sequence(s) reported in this paper has been submitted
to the GenBank(TM)/EMBL Data Bank with accession number(s)
X81630[GenBank].
Volume 270,
Number 2,
Issue of January 13, 1995 pp. 830-836
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
POSITIVE AND NEGATIVE REGULATORY ELEMENTS DIRECT GENE EXPRESSION TO
MESENCHYMAL CELLS (*)
e 10, 13125 Berlin, Germany
and ICRF, Cell Interaction Laboratory, Cambridge University
Medical School, Cambridge CB2 2QH, United Kingdom,
12-O-tetradecanoylphorbol-13-acetate, and coculture of
epithelial cells on SF/HGF gene expression. This study represents a
first step toward understanding the intricately regulated and cell
type-specific expression of the paracrine acting SF/HGF.
)neuregulin (also
called neu differentiation factor, or heregulin) or keratinocyte growth
factor. All these factors are produced by mesenchymal cells, bind to
membrane receptors expressed in mainly epithelial cells (c-met, the
c-erbBs, and the keratinocyte growth factor receptor, 4-8), and
are potent effectors of epithelial growth, movement, and
differentiation in
vitro(5, 8, 9, 10) .1 or by coculture with epithelial
cells(26, 27, 28) . In contrast, interleukin
1
, tumor necrosis factor-, and the newly identified factor
injurin increased SF/HGF expression(29, 30) . In
primary human fibroblasts and HL60 leukemia cells, SF/HGF expression
was also stimulated by phorbol esters(31, 32) . The
sequences of the human and rat SF/HGF gene promoters have been
determined and the major transcription start sites were
mapped(33, 34) , but no activity studies have been
reported. In the present investigation, we show that sequences around
the major transcription start site of the mouse SF/HGF promoter are
sufficient to direct expression in fibroblasts but not in epithelial
cells and that a negative regulatory element is located between
positions -239 and -258 of the promoter.
DNA Clones and Promoter CAT Assays
A mouse
genomic library (E 14 TG2A, from mouse strain 129) in 2001 was
screened for the SF/HGF promoter region with
P-labeled
fragments spanning 5` sequences of the mouse SF/HGF cDNA. A 10-kilobase XhoI fragment containing 5`-flanking sequences, the first
exon, and part of the first intron was subcloned from a particular
lambda clone into Blue Script KS- (Stratagene, La Jolla) and used to
generate the chimeric promoter CAT constructs m-3000 and m-755 by
ligating appropriate restriction fragments into the promoter-less
plasmid pCAT-basic (Promega). The -365, -291, -239,
-150, and -70 mouse promoter CAT constructs were prepared
by BAL-31 nuclease (Boehringer Mannheim) digestion of the m-755
fragment. The -7 and +14 mouse promoter CAT constructs were
generated by cloning respective oligonucleotides into pCAT-basic. All
deletion fragments had a common 3`-end at position +34 which
represents a natural PstI site in the mouse SF/HGF promoter.
Deletion construct
-66/+34 was generated by digesting clone
m-755 with restriction enzymes BglII and PstI. Blunt
ends were created with Klenow DNA polymerase (Boehringer Mannheim), and
the fragment was religated using T4 DNA ligase (Biolabs, Beverly, Ma).
Internal deletion construct -239/-258 was generated via PCR (35) using the respective plasmid clones as the starting
templates and Vent polymerase (Biolabs, Beverly, MA). Primer pairs had
the following sequence: a, CACAGGAAACAGCTATGACC; b,
TTACAAAGCAAAGGTCACCTTTTGGAAGCTGGAGCTCCAGATCC; a`,
TTTAGCTTCCTTAGCTCCTG; b`, AGGTGACCTTTGCTTTGTAAG.-galactosidase
activities. Each transfection experiment was carried out twice with
double values and with two different preparations of the same plasmid.
To assess the effect of TGF1 on promoter activity, transfected
cells were treated 14 h post-transfection with 5 ng/ml of the factor
(Boehringer Mannheim), and cell extracts were prepared 24 h later. To
analyze the effect of coculture with epithelial cells, equal numbers of
either mitomycin-treated Madin-Darby canine kidney cells or, as a
contol, mitomycin-treated fibroblasts were immediately seeded on top of
the transfectants. Cocultures were continued for 40 h after which cell
extracts were prepared. Mitomycin treatment was as described (27) .RNA Preparation and RNase Protection
Analysis
Total RNA from cells transiently transfected with
promoter CAT constructs was prepared by guanidinium isothiocyanate
extraction 16 h after the removal of the precipitates and treated with
DNaseI (Boehringer) in the presence of RNasin ribonuclease inhibitor
(Boehringer). RNase protection experiments were performed as
described(6) . A subclone of deletion construct m-150 covering
SF/HGF promoter sequences and the N-terminal part of the CAT gene was
generated after digestion with HindIII and PvuII and insertion into Bluescribe M13+ between HindIII and SmaII sites. The
antisense probe was synthesized as T7 run-off transcript after
linearization with BglII. For RNase protection,
50,000 counts/min of labeled transcript were hybridized for 12 h to 50
µg of RNA, digested with 6 units of RNase T2 (Sigma) for 1 h, and
analyzed on denaturing polyacrylamide gels.Gel Retardation Assay and DNase
Footprinting
Nuclear extracts were prepared and gel retardation
assays performed as described(37) ; each sample contained 1
µg of poly(dI-dC) (Boehringer Mannheim) as nonspecific competitor.
Double-stranded oligonucleotides span the positions -7 to
+34 of the mouse SF/HGF promoter; competitor oligonucleotides span
nucleotides -7 to + 34 or +14 to +34. For DNaseI
footprinting of the promoter region covering the transcription start
sites, clone m-150 was radioactively labeled at the 3`-end of the
noncoding strand (with Klenow after linearization with HindIII). An end-labeled fragment of 200 nt in length was
recovered after digestion with XbaI. To analyze factor binding
in the region between position -70 to -450, the m-755 clone
was labeled at the 3`-end of the coding strand after linearization wirh
either BSTE II or BglII. Digestion with NsiI or HindIII resulted in end-labeled fragments of 330 and 520
nucleotides in length. Footprinting assays were performed as described (37) with 100-150 µg of protein from crude nuclear
extracts, 1 µg of poly(dI-dC), and 1-2 ng of labeled probe at
a specific activity of 2 
10
counts/min/µg.Cell Lines
Two sources of NIH3T3 mouse fibroblasts
were used: the clone D4 and ras transfectants of the same
clone(5) . The other cell lines were as described in Refs. 10,
13, and 37.
Cloning of Mouse and Human SF/HGF Promoter
Sequences
In order to carry out functional analyses of the
SF/HGF promoter, sequences upstream of the major transcription start
site of the human gene (33) were amplified by PCR using 5`
nested primers (see ``Materials and Methods''). The isolated
fragments span 1017 and 372 base pairs upstream of the major
transcription start site (33) and contained identical 3`-ends
at position +58. The mouse SF/HGF promoter was isolated from a
genomic library of embryonal stem cells (strain 129); the fragments
employed in this study span 3,000 and 755 base pairs of upstream
sequences and again had identical 3` -ends at position +34.
Alignment of the SF/HGF promoter sequences of the two species (Fig. 1) revealed 89.5% identity up to position -453 and
no homology further upstream (shown up to position -514, but see
also Genbank data base). Alignment of the mouse sequence and the
previously reported rat SF/HGF promoter sequence (34) showed
96.4% identity up to position -450 and 88% from positions
-450 to -758. Characteristically, the human promoter
contains two stretches of 13 and 16 nt that are absent in the mouse
sequence. The major transcription start site of the human gene (large downward pointing arrow, 33) is shown in Fig. 1,
and so are the mapped start sites of the mouse promoter (upward
pointing arrows; see also below). Identical sites were previously
determined for the rat promoter sequence(34) .
The SF/HGF Promoters Are Functional in Mesenchymal But
Not Epithelial Cells
Fragments of both the mouse and human
promoters were cloned into a CAT expression vector (pCAT basic, see (37) ) and examined by transfection into fibroblasts and
epithelial cells (Fig. 2). In ras3T3 fibroblasts, shorter
fragments of both promoters resulted in slightly stronger activity than
larger fragments (compare human-1017 with h-372 and mouse-3,000 with
m-755 or m-70). A promoter fragment in the reverse orientation (m-755
r) was inactive. None of the constructs tested showed significant
activity in the epithelial cell line MCF7. Similar differences in
promoter activities were observed in other fibroblast (3T3 and MRC5)
and epithelial (CSG 120/7 and MDA MB 435) cell lines (data not shown).
Deletion Analysis of the Mouse SF/HGF
Promoter
Progressive 5` deletions of the mouse SF/HGF promoter,
either generated by BAL-31 digestion of clone m-755 or by inserting
synthetic oligonucleotides, were also tested in the CAT assays (Fig. 3): In ras3T3 fibroblasts, promoter activity increased
with progressive deletion to position -365 and decreased sharply
after further deletion to -291. Removal of additional 52 nt to
-239 increased the promoter activity again, which remained
essentially constant when further sequences down to the major
transcription start site were deleted (position -7). Removal of
the sequence containing the major transcription start site (nt -7
to +14) resulted in complete loss of activity. The most active
promoter fragment (nt -365 to +34) had approximately 20% of
the activity of the SV40 promoter/enhancer. Similar results were
obtained with 3T3 fibroblasts (not shown). In MCF7 epithelial cells,
all constructs showed only background activity. These combined results
indicate that the SF/HGF promoter is composed of positive and negative
regulatory elements which restrict the activity of the promoter to
mesenchymal cells. Positive elements are located upstream of position
-291 and around the major transcription start site, and a
negative element is located between positions -291 and
-239.
1, TPA, and
coculture with epithelial
cells(26, 27, 32, 38, 39, 40) .
We subjected ras3T3 fibroblasts, which were transiently transfected
with the various promoter deletion constructs, to treatment with
TGF1, TPA, and coculture with Madin-Darby canine kidney epithelial
cells. These treatments did not lead to significant changes in the
amounts of promoter CAT activities and in the profile seen when
different deletion constructs were compared (data not shown).Localization of Nuclear Factor Binding Sites by
Footprinting and Gel Retardation Analysis
Footprint analyses of
the SF/HGF promoter with nuclear extracts of fibroblasts and epithelial
cells were performed in the region which is highly conserved between
species, i.e. between the transcription start sites and
position -450 (Fig. 4). With nuclear extracts of ras3T3
cells, specific DNaseI protection was observed between positions
+14 and -7 as well as between positions -21 and
-70. A site protected in both fibroblasts and epithelial cells
(MCF7) was detected between positions -229 and -258.
Analysis of the Negative and Positive Regulatory Elements
of the SF/HGF Promoter by Specific Deletion
Two regions of the
promoter which appear to be important for negative and positive
regulation were deleted and analyzed in transient transfection
experiments (Fig. 6). Removal of the nucleotides between
-239 and -258 which were protected in footprint analysis
resulted in a release of the inhibition of promoter activity seen with
the mouse -291 construct. We inserted multiple copies of an
oligonucleotide containing nt -229 to -258 into
heterologous promoters (SV40, TK minimal promoter, and
E-cadherin(37) ). No inhibitory effect was detected; the
E-cadherin promoter was stimulated 2-3-fold (data not shown).
Deletion of the -66 to +34 region from the m-755 construct,
which harbors the transcription start sites and was found to be
involved in nuclear factor binding, resulted in complete loss of
promoter activity.
response elements
previously identified in the human and rat promoter (cf. Refs.
33, 34).
/TIE, TGF inhibitory element (black
box); IL6RE, interleukin 6-responsive elements (
); NFIL6, interleukin 6-dependent nuclear factor (). P1, P2, and P3 are palindromic sequences.
DNA complex which is characteristic for SF/HGF-expressing
cells. The sequence -7 to +14 of the SF/HGF promoter does
not fit the initiator sequence 5`-CTCANTCT-3` described for well
studied TATA-less promoters(44) . A noncanonical TATA element
(AATAAA) at position -24 might be responsible for transcription
initiation at multiple sites.(26, 38) , TPA(32) ,
or factors produced by cocultured epithelial cells(27) . None
of these factors significantly influenced promoter activity in our
transient transfection assays. This suggests that additional
cis-elements outside the promoter region we have studied here are
involved in the response to these agents or that the effect of
TGF, TPA, and coculture with epithelial cells is translational
rather than transcriptional. Nuclear run-off transcription reactions
and measurements of the half-life of the SF/HGF transcript in
fibroblast cultures exposed to TGF, TPA, or IL6 should clarify
this point. After completion of this work, a paper appeared (45) which describes aspects of the SF/HGF promoter.
Surprisingly, these authors found activity also in the carcinoma
(epithelial) cell line RL 95-2. No factor binding by footprinting
or bandshift analysis is shown. However, IL6 treatment stimulated
promoter activity 2.5-fold in a stably transfected cell clone. No
stimulation of SF/HGF expression by IL6 was reported by
others(29) .
)Furthermore, we will compare promoter fragments with and
without the negatively acting element in order to see in which tissues
this element may suppress SF/HGF expression in vivo.
)1, transforming growth factor 1; TPA,
12-O-tetradecanoylphorbol-13-acetate; CAT, chloramphenicol
acetyltransferase; PCR, polymerase chain reaction; nt, nucleotide(s);
IL6, interleukin-6.)
We thank Jörg
Hülsken for the help with the computer graphics,
Dr. Carmen Birchmeier (Cologne) for critically reading the manuscript,
and I. Wiznerowicz for excellent secretarial work.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
ABSTRACT
This article has been cited by other articles:
|
|
 |
|
  K. Nasu, T. Sugano, N. Matsui, H. Narahara, Y. Kawano, and I. Miyakawa Expression of Hepatocyte Growth Factor in Cultured Human Endometrial Stromal Cells Is Induced through a Protein Kinase C-Dependent Pathway Biol Reprod, May 1, 1999; 60(5): 1183 - 1187. [Abstract] [Full Text]
|
|
|
|
 |
|
 A. W. Bell, J.-G. Jiang, Q. Chen, Y. Liu, and R. Zarnegar The Upstream Regulatory Regions of the Hepatocyte Growth Factor Gene Promoter Are Essential for Its Expression in Transgenic Mice J. Biol. Chem., March 20, 1998; 273(12): 6900 - 6908. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.-G. Jiang, A. Bell, Y. Liu, and R. Zarnegar Transcriptional Regulation of the Hepatocyte Growth Factor Gene by the Nuclear Receptors Chicken Ovalbumin Upstream Promoter Transcription Factor and Estrogen Receptor J. Biol. Chem., February 14, 1997; 272(7): 3928 - 3934. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.-G. Jiang, C. Johnson, and R. Zarnegar Peroxisome Proliferator-activated Receptor gamma -mediated Transcriptional Up-regulation of the Hepatocyte Growth Factor Gene Promoter via a Novel Composite cis-Acting Element J. Biol. Chem., June 29, 2001; 276(27): 25049 - 25056. [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 |