| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
B Mediates TNF-
Inhibitory Effect on 2(I) Collagen (COL1A2) Gene Transcription in Human Dermal Fibroblasts1
,
,
Departments of
*
Dermatology and Cutaneous Biology and
Biochemistry and Molecular Pharmacology, Jefferson Medical College, and
Jefferson Institute of Molecular Medicine, Thomas Jefferson University, Philadelphia, PA 19107; and
§
Department of Dermatology, Yonsei University College of Medicine, Seoul, Korea
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
has potent regulatory control on extracellular matrix
production and degradation. Earlier studies have documented that
TNF- inhibits type I collagen gene (COL1A2) expression at the
transcriptional level, but the characterization of the transcription
factors involved has been elusive. In the present study, using
transient cell transfection of human dermal fibroblasts with a battery
of 5' end deletion/chloramphenicol acetyltransferase (CAT) reporter
gene constructs, we have characterized the TNF- response element of
the COL1A2 promoter. The TNF- response element was attributed to a
specific region that comprises noncanonical activator protein-1 (AP-1)
(CGAGTCA) and NF-
B (AGAGTTTCCC) binding sites. TNF- effect was
eliminated by a 2-bp substitution mutation in the NF-B1 binding half
site of the NF-B cis element. Electrophoretic
mobility shift assays (EMSA) showed that recombinant human NF-B
heterodimers as well as NF-B1 and RelA homodimers, but not AP-1,
were capable of binding this element. Further, EMSA with human
fibroblast nuclear extracts demonstrated enhanced binding of a single,
specific complex within 5 min of TNF- stimulation, which reached a
plateau by 1 h and was not affected by preincubation of cells with
cycloheximide. Gel supershift assays identified the complex as the
NF-B (p50/p65) heterodimer, whereas Abs to nuclear factor of
activated T cells (NF-AT) and Jun family members failed to recognize
the complex. These data suggest that in fibroblasts TNF- activates
and initiates the nuclear translocation of NF-B that binds a
divergent NF-B element and plays a critical role in the observed
inhibition of 2(I) collagen gene
transcription. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
inhibits,
the transcription of collagens in a coordinated fashion within the
wound bed. Recently, several attempts have been made to elucidate the
mechanisms by which inflammatory cytokines, including TGF-ß and
TNF-, regulate transcription of the human 2(I) collagen gene
(COL1A2).
Initial observations demonstrated that a 135-bp region of the
COL1A2 promoter within 330 bp of the transcription start site could
confer responsiveness to both TGF-ß (1) and TNF- (2). This rather
large region was shown to contain two smaller regions, the first
referred to as box A, spanning nucleotides -313 to -286 and
containing two distinct nuclear protein binding sites called 5A and 3A,
and the other referred to as box B from nucleotides -271 to -255. The
upstream element, box A, was shown to bind Sp1 and to confer high basal
promoter activity (1, 3, 4). Box B harbors both a putative activator
protein-1 (AP-1) binding element and a noncanonical NF-B binding
site (1, 2). Although no specific transcription factor could be
identified by Inagaki et al. (1, 2), they postulated that unknown
factors interacting with both box B and the upstream Sp1 binding sites
present in box A were necessary for both TGF-ß and TNF- responses
(1, 2). Recently, we have demonstrated that the region between
nucleotides -271 and -235 is critical and sufficient for enhanced
COL1A2 transcription upon TGF-ß stimulation (4) and does not require
the upstream Sp1 binding sites that only play a role in the basal
activity of the promoter. Also, we showed that this short promoter
segment, containing both AP-1 and NF-B-like binding sites, permits
antagonist activity of TNF- against TGF-ß, but the exact
mechanisms and the responsive sequence elements have not been
elucidated in further detail.
In this report, we have identified the specific TNF- response
element (TaRE) that allows COL1A2 transcriptional inhibitory response
to TNF- in human dermal fibroblasts. Specifically, the TaRE was
found to reside between nucleotides -271 and -235 relative to the
transcription initiation site, a region previously shown to confer
TGF-ß responsiveness. Electrophoretic mobility supershift assays
(EMSA) identified NF-B1 and RelA NF-B family members as
transcription factors binding the TaRE and mediating TNF- repression
of COL1A2 promoter activity. Precise nucleotide requirement for NF-B
binding to the TaRE was determined in nucleotide substitution
experiments and clearly indicated that TNF- and TGF-ß responses
involve closely located, but clearly distinct, elements within the
-271/-235 segment of COL1A2 promoter. Furthermore, our results
indicate that NF-B, in the context of the proximal COL1A2 promoter,
may act as a potent transcription inhibitor.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Human dermal fibroblast cultures, established by explanting
tissue specimens from neonatal foreskins, were used in passages 3–6.
The cultures were maintained in DMEM supplemented with 10% FCS, 2 mM
glutamine, 100 U/ml of penicillin, and 50 µg/ml of streptomycin.
Human recombinant TNF- was purchased from Boehringer Mannheim
(Indianapolis, IN).
Plasmid constructs
Transient transfection experiments were performed with several
5' deletion constructs derived from pMS3.5CAT (kindly provided by Dr.
Francesco Ramirez, Mt. Sinai School of Medicine, New York, NY), a
plasmid containing 
3.5 kb of human COL1A2 promoter linked to the
chloramphenicol acetyltransferase (CAT) reporter gene (5). Several
additional deletion constructs were generated by PCR as described
previously (4). The pRSV-ß-galactosidase control vector was used as a
standard for control of transfection efficiency (Promega, Madison, WI).
pRSV-NF-B1 (p50) and pRSV-RelA (p65) expression vectors were
obtained through the AIDS Research and Reference Reagent Program (
National Institute of Allergy and Infectious Diseases, National
Institutes of Health) from Dr. Gary Nabel and Dr. Neil Perkins (6).
Site-directed mutagenesis of the TaRE
Point mutations were introduced into the putative AP-1 and
NF-B binding elements of construct -271COL1A2/CAT, containing the
region -271 to +58 of the COL1A2 promoter cloned upstream of the CAT
gene, using PCR with 5'-end/BamHI overhang mutagenic primers
containing two point mutations (bold) in the putative AP-1, NF-B1,
or RelA binding elements (AP-1, CGAGTCA 
CCAGTGA; NF-B1, AGAGTTTCC
ATCGTTTCC; RelA, AGAGTTTCC AGAGTTGAC) and a
3'-end primer close to the transcription initiation site. The PCR
products were digested with BamHI/XmaI, purified
on a 2% agarose gel, and cloned into similarly digested
WT-271COL1A2/CAT, generating constructs containing mutated AP-1,
NF-B1 and RelA binding sites, -271 mAP1/CAT, -271 mK1/CAT, and
-271 mK2/CAT, respectively. Also, the mK1 mutation was introduced into
-376COL1A2/CAT, using the "Quick-Change" site-directed mutagenesis
kit according to the manufacturer’s protocol (Stratagene, La Jolla,
CA). Sequence integrity of all constructs was verified by automated
sequencing (Applied Biosystems, Foster City, CA).
Transient transfections and CAT assays
Transient transfections of human foreskin fibroblasts were
performed by the calcium phosphate/DNA coprecipitation method, as
described previously (7). Briefly, the cells were transfected with 10
µg of reporter DNA mixed with 5 µg of the RSV-ß-galactosidase
plasmid DNA to monitor transfection efficiencies. After glycerol shock,
the cells were placed in DMEM containing 10% FCS for 2 h before
the addition of TNF-. After 40 h of incubation, the cells were
rinsed twice with PBS, harvested by scraping, and lysed in reporter
lysis buffer (Promega). The ß-galactosidase activities were measured
according to standard protocols (8). Aliquots corresponding to
identical ß-galactosidase activity were used for each CAT assay with
[14C]chloramphenicol as substrate (9) using thin layer
chromatography. After autoradiography, the plates were cut and counted
by liquid scintillation to quantify the acetylated
[14C]chloramphenicol. Percent acetylation was determined
as the average quotient of acetylated products and total (unacetylated
and acetylated chloramphenicol) ± SD.
EMSA
Nuclear extracts were prepared from human dermal fibroblast
cultures according to the method of Andrews and Faller (10). The
following oligonucleotides were used: a 37-bp double-stranded
oligonucleotide corresponding to the -271 to -235 region of the
COL1A2 promoter, wild-type (WT) TaRE,
5'-GAGGTATGCAGACAACGAGTCAGAGTTTCCCCTTGAA-3'; a similar 37-bp
oligonucleotide in which two point mutations were made (bold), mK1
TaRE, 5'-GAGGTATGCAGACAACGAGTCATCGTTTCCCCTTGAA-3'; a
20-bp oligonucleotide spanning the promoter region from residues -258
to -239, WT short TaRE, 5'-ACGAGTCAGAGTTTCCCCT-3'; and a similar
oligonucleotide with two (bold) point mutations, mK1 short TaRE,
5'-ACGAGTCATCGTTTCCCCT-3'. In some experiments, a 22-bp
oligonucleotide containing the consensus collagenase AP-1 binding site
(underlined), 5'-CTAGTGATGAGTCAGCCGGATC-3', or a 22-bp
oligonucleotide containing a consensus NF-B binding element
(underlined) found in the Ig- light chain gene,
5'-GATCGAGGGGACTTTCCCTAGC-3' (11) was used as either
probe or unlabeled competitor. The oligonucleotides were end labeled
with [
-32P]dATP and 4 x 104 cpm were
incubated with 6 µg of nuclear extract for 1 h on ice in 20 µl
binding reaction buffer (12 mM HEPES/KOH (pH 7.9), 4 mM Tris (pH 7.9),
60 mM KCl, 1 mM EDTA, and 12% glycerol) in the presence of 1 µg
poly[dI-dC] (Boehringer Mannheim), as described previously (12).
DNA/nucleoprotein complex specificity was determined by coincubation of
nuclear extracts with unlabeled homologous or nonhomologous competitor
DNA. For competition experiments, 60-fold molar excess of unlabeled
oligonucleotide was included in the binding reaction. DNA-protein
complexes were separated from unbound oligonucleotide on 4% acrylamide
gel in 0.5x TBE. The gels were fixed for 30 min in 30% methanol and
10% acetic acid, vacuum dried, and exposed to x-ray film at -70°C.
The following Abs were used in supershift experiments: rabbit
polyclonal Abs against NF-B1 and RelA (13), NF-B2 (p52) (14),
c-Rel (p75) (15), RelB (p68) (our unpublished data) raised
against a synthetic 17-amino acid peptide covering the C terminus of
human RelB, REAAFGGGLLPGPEAT, a pan-NF-AT (nuclear factor of
activated T cells) Ab, raised against an internal peptide of human
NF-AT common to all members of the NF-AT family (16), and a rabbit
polyclonal anti-c-Jun (Santa Cruz Biotechnology, Santa Clara, CA).
All NF-B and NF-AT Abs were generous gifts from Dr. Nancy Rice
(National Cancer Institute, Frederick, MD). Briefly, Abs were added to
the nuclear extracts (0.1 µg polyclonal antiserum per 6 µg protein)
and incubated on ice for 2 h before the binding reaction. Also, in
some experiments, recombinant human NF-B, NF-B1, or RelA protein
(100 ng/binding reaction; a generous gift of Dr. Timothy Coleman, Human
Genome Sciences, Rockville, MD) was used with 1.0 µg of poly[dI-dC]
and 10 µg of BSA as a carrier.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Previous work from our laboratory localized a TNF--response
region within the COL1A2 promoter between residues -271 and -235 (4).
However, because the major focus of that study was to investigate the
regulatory regions involved in TGF-ß up-regulation of COL1A2 gene
expression, the experiments were performed in low serum conditions that
kept collagen promoter activity at a low level, unfavorable for a
precise evaluation of the inhibitory action of TNF-. Therefore, to
ascertain the localization of the TNF--responsive element within the
COL1A2 promoter, we repeated the 5' deletion studies by performing the
transfection experiments in medium containing 10% FCS, which we
previously showed to provide high basal promoter activity (4). All 5'
deletion constructs containing >271 bp of promoter demonstrate a
significantly greater basal activity than the -271COL1A2/CAT
construct, previously attributed to three Sp1 elements between -313
and -286 (3, 4). The loss of these elements upstream from nucleotide
-271 predictably lowers basal activity (Fig. 1
A). To maintain CAT activity
for larger deletions within the linear range of the assay, small
volumes of lysate were used. Therefore, activity of the -271
COL1A2/CAT promoter deletion is hardly detectable on the autoradiogram
shown. However, for quantitation purposes, the samples from cells
transfected with short constructs were assayed with more extract,
providing accurate and reproducible determination of their activity.
|
challenge (50–75% inhibition);
however, the loss of elements between -271 and -235 entirely
abolished the inhibitory response, and even a significant reversal of
the effect of TNF- was noted (Fig. 1
). This stimulation in response
to TNF- was maintained in further deletions to -197. However,
further promoter deletions past -186 resulted in very low basal
promoter activity accompanied with minor, statistically insignificant,
inhibitory response to TNF- (data not shown). These data narrow down
the essential TaRE, originally described by Inagaki et al. (4) between
residues -378 and -235 of the human COL1A2 promoter to a 36-bp
fragment located between nucleotides -271 and -235.
Further characterization of the TaRE was accomplished by analyzing the
effect of various point mutations within the promoter region from -271
to -235 (Fig. 2). In these experiments,
CAT activity is significantly higher than that shown in Fig. 1A because correction was not relative to longer constructs
with higher basal activity. Computer analysis of the sequences
comprising this region identified two putative cis elements:
an AP-1-like element and a noncanonical NF-B binding site, adjacent
and downstream of AP-1, as schematically represented in Fig. 2A. The TaRE was previously identified as spanning the
region between -340 and -235 (2), and several boxes with regulatory
functions were identified (2). We identified the AP-1 element as
critical for the TGF-ß response (4). The TaRE region also allows
antagonism of the TGF-ß effect by TNF- (2); however, the exact
mechanisms have not been elucidated before this study. Our results
indicate that a double point mutation in the putative NF-B1 binding
site (mK1) completely abolishes the response to TNF- (5% inhibition
with the construct mK1/CAT vs 80% inhibition with WT -271COL1A2/CAT
construct; Fig. 2, B and C). In contrast,
mutations in both putative AP-1 and RelA binding regions (construct
mAP-1 and mK2, respectively) had little if any effect on the inhibition
of COL1A2 promoter activity by TNF-, as compared with the
WT-271COL1A2/CAT construct (Fig. 2, B and C).
|
inhibitory effect, the -376COL1A2/CAT construct was
mutated similarly, and its responsiveness to TNF- was compared with
that of the parent construct. Similar to the data presented above with
construct -271, the mK1 mutation essentially abolished TNF-
inhibitory activity (Fig. 3). These
results contrast significantly with mutational analysis of the TGF-ß
response element (4) that identified the AP-1 binding site as essential
for TGF-ß-induced up-regulation of COL1A2 promoter activity and
indicate that the two cytokines use different, although very closely
located, regulatory elements.
|
induces nuclear proteins that bind both TaRE and NF-B
elements
Previous studies have documented that numerous cytokines,
including TNF-, can rapidly activate latent NF-B in the
cytoplasm, leading to its rapid nuclear translocation and subsequent
gene activation (17). In addition, several members of the AP-1 gene
family, including c-jun, jun-B and
c-fos, are activated in fibroblasts by TNF- (18).
Therefore, given the putative cis elements recognized within
the TNF- responsive region, it was important to establish whether
TNF- activation of either AP-1 or NF-B was involved as
transcription factors binding to the TaRE. For this purpose, confluent
dermal fibroblast cultures were incubated for 24 h with or without
TNF-, and EMSAs were performed to compare the binding of nuclear
extracts to WT TaRE with their binding pattern to consensus NF-B and
AP-1 probes. As shown in Fig. 4, TNF-
dramatically enhanced the binding of nuclear proteins to both TaRE and
NF-B probes. The relative migration rate of the shifted TaRE
complexes (Fig. 4, lane 3) was similar to that of NF-B
complexes (lane 6). No differential complex formation
was observed with the consensus AP-1 probe. This lack of enhancement of
AP-1 binding is likely to be due to the fact that AP-1 activation by
TNF-, and subsequent DNA binding activity, are a transient
phenomenon that occurs within a few h after stimulation of the cells
and can no longer be observed at the 24 h time point studied.
|
on the COL1A2 promoter is
mediated through a similar rapid nuclear translocation as is the case
of NF-B, the time course of nuclear factor binding to the TaRE was
examined in EMSA experiments. Appearance of the TNF-specific complex
was observed as early as 5 min after TNF- addition. A plateau was
reached by 15 min and maintained for at least 24 h of stimulation
(data not shown). Activation of TaRE binding is independent of de novo protein synthesis
Although the appearance of TaRE-specific binding proteins in
nuclear fractions was very rapid, we wanted to ascertain the
involvement, if any, of de novo protein synthesis in this event. For
this purpose, dermal fibroblast cultures were incubated with or without
100 µg/ml of cycloheximide, a concentration that effectively blocks
fibroblast protein synthesis, for 1 h before TNF- addition, and
nuclear extracts were subsequently prepared. As shown in Fig. 5, and as expected from the above data,
TNF- alone induced strong nuclear protein binding to the TaRE as
compared with that of unstimulated fibroblast extracts (Fig. 5, lane 3 vs lane 2). Addition of cycloheximide
alone had no effect on TaRE binding and did not alter TNF--induced
binding (Fig. 5, lanes 4 and 5,
respectively), indicating that de novo protein synthesis is not
required for TNF- effect.
|
EMSA results presented in Fig. 4 demonstrated a strong correlation
between TNF--induced TaRE binding protein and NF-B binding to its
consensus sequence. Kinetic studies also correlate strongly with the
rapid induction of NF-B by TNF- (19, 20). To determine whether a
complex of NF-B family members was, in fact, the TaRE binding
protein, EMSAs were performed to demonstrate complex specificity
through comparative competitions (Fig. 6A). Nuclear extracts were
incubated without (Fig. 6A, lanes 1–3) or
with (lanes 4–7) unlabeled competitor
oligonucleotides. A 20- and 60-fold molar excess of homologous
competitor DNA strongly reduced the TNF--induced complex (Fig. 6A, lane 3) in a dose-dependent manner
(lanes 4 and 5 vs lane 3).
However, competition with 60-fold molar excess of unlabeled AP-1
oligonucleotide failed to eliminate complex formation
(lane 6). In contrast, 60-fold molar excess of
NF-B oligonucleotide competed successfully for the shifted complex
(lane 7), indicating that complexes that bind the TaRE also
recognize the consensus NF-B site.
|
B may indeed be the
TaRE binding factor and to characterize the possible Rel subunit
combinations that could comprise this factor, we used Abs directed
against five of the known Rel/NF-B family members: NF-B1,
NF-B2, RelA, RelB, and c-rel in gel supershift assays.
Also, antisera cross-reactive with either all NF-AT family members and
against c-Jun were included in the experiments. Results are shown in
Fig. 6B. Notably, only Abs against either NF-B1 and RelA
recognized the complex (Fig. 6B, lanes 4 and
6, respectively), whereas no supershift was observed with
Abs directed against the other members of the Rel family
(lanes 5, 7, and 8) or NF-AT
(lane 9) or c-Jun (lane 10).
These results suggest that the protein binding the TaRE in the COL1A2
promoter is the NF-B heterodimer.
To determine whether the Rel family members binding to the TaRE were
representative of all Rel members induced by TNF-, or instead,
represent a COL1A2 promoter TaRE-specific subset, the NF-B element
of the Ig- light chain enhancer was used as a probe in similar gel
supershift assays (Fig. 6C). The Rel heterodimer complex
binding the Ig- NF-B consensus sequence upon TNF- stimulation
was supershifted by NF-B1 and RelA Abs only (Fig. 6C,
lanes 4 and 6, respectively), whereas antisera
directed to NF-B2, RelB, and c-Rel failed to recognize the complex
(lanes 5, 7, and 8, respectively). These
data demonstrate that sequence variation between the NF-B binding
sites of the two genes, COL1A2 and Ig- light chain, does not lead to
differential binding of Rel subunits.
Mutations within the TaRE eliminate its ability to bind NF-B
Our next step consisted of establishing the causal relationship
between NF-B binding to the TaRE and the inhibition of COL1A2
promoter activity by TNF-. For this purpose, the binding ability of
a 37-bp oligonucleotide, mK1 TaRE, containing the same mutation as the
one generated in the -271 mK1/CAT construct (rendering it unresponsive
to TNF-; see above, Fig. 2), to baculovirus-produced recombinant
NF-B1 and RelA, was compared with that of the WT TaRE probe. As
predicted, the TaRE was able to bind both NF-B1 and RelA homodimers,
as well as NF-B heterodimers (Fig. 7,
lanes 3–5). It should be noted that RelA homodimers
(*) were partially degraded and lacked a portion of their C terminus
region. Although it had no effect on their binding ability, it resulted
in a faster migration rate than that of NF-B heterodimers in EMSA
(21). Mutation of the TaRE probe in the NF-B1 binding half site (mK1
TaRE) totally eliminated recombinant NF-B/DNA complex formation
(Fig. 7, lanes 8–10), in agreement with results of
transient transfection experiments in which similarly mutated COL1A2
promoter constructs, -376 mK1/CAT and -271 mK1/CAT, showed no TNF-
responsiveness (see Figs. 2 and 3). In addition, mutations in the RelA
half-site (mK2 TaRE) had no effect on recombinant NF-B binding (data
not shown).
|
for 24 h,
to bind the mutated TaRE probes was examined and compared with binding
the WT TaRE (Fig. 8). The differential
complex formation between unstimulated and stimulated nuclear extracts
observed with the WT TaRE probe (Fig. 8, lane 3 vs.
lane 2) was eliminated by the 2-bp mutation generated in the
mK1 probe (lanes 5 and 6 vs
lanes 2 and 3). In contrast, the binding
pattern observed with the mK2 probe was similar to that observed with
the WT TaRE probe (Fig. 8, lanes 8 and 9 vs
lanes 2 and 3). Taken together, these
results attest to the specificity of the causal relationship between
NF-B binding to the TaRE and TNF- inhibitory response of the
COL1A2 promoter. Also, the ability of point mutations within the
NF-B1 binding region to abrogate not only NF-B heterodimer but
also recombinant RelA homodimer complex formation underscores the
strong binding affinity of this half of the cis element and
further emphasizes the key role of the NF-B1 site relative to the
RelA binding half site in mediating TNF- effect. These data
corroborate the results from transfection experiments in which both the
-271 mK2/CAT and -271WT/CAT constructs responded equally to TNF-,
whereas constructs harboring the mK1 mutation were not responsive (see
Figs. 2 and 3).
|
B-like site and responsible for TNF- inhibitory effect.
Overexpression of NF-B inhibits COL1A2 promoter activity
To establish a functional link between RelA binding to the COL1A2
TaRE and the inhibitory effect of TNF- on COL1A2 promoter
activity, cotransfection experiments were performed in which NF-B1
and RelA expression vectors were transfected together with several
COL1A2 promoter CAT constructs. As shown in Fig. 9, NF-B overexpression significantly
inhibited the activity of both -342COL1A2/CAT and -271COL1A2/CAT
constructs. In contrast, the activity of the -235COL1A2/CAT construct
was not altered. Similar results were obtained when RelA alone was
cotransfected with the CAT constructs (data not shown), indicating that
the inhibitory effect of RelA, is exerted through the NF-B binding
element identified as the TaRE and located between nucleotides -271
and -235.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, TGF-ß, and IFN-
orchestrate the balance between ECM destruction and regeneration, and
therefore, play an important role in reparative processes as well as in
the pathogenesis of fibrotic diseases (22).
Several studies have suggested that TNF- depresses type I collagen
gene expression, acting predominantly at the level of transcription
(23, 24, 25). In more recent work using footprinting analysis of the COL1A2
promoter, responsiveness to TNF- was localized to a region between
-378 and -235 relative to the transcriptional start site (2). Two
protected regions of promoter within this 143-bp element were
identified: box A and box B. When placed together in the context of a
heterologous promoter, these two boxes conferred responsiveness to
TNF- (2). However, the discrete sequences and the transcription
factors involved in TNF- response were not fully characterized.
Indeed, an Ab directed against RelA obtained from Santa Cruz
Biotechnology failed to supershift or abolish the TNF--induced
complex that they observed in EMSA with probes spanning the entire
region from -340 to -235 or box 5A (see Fig. 2A) in which
the authors identified a potential NF-B binding site (2). The 3'-end
of the TaRE that contains the functional NF-B site identified in our
study was not investigated in further detail by these authors. Based on
their results, Inagaki et al. (2) excluded NF-B from being part of
the transcription complex involved in the regulation of COL1A2 promoter
activity by TNF-.
Recently, we have shown that regions downstream of nucleotide -271 are
sufficient for antagonist activity of TNF- against TGF-ß-mediated
up-regulation of COL1A2 promoter activity (4). In the present study, we
have expanded upon these results by demonstrating that regions upstream
of nucleotide -271 are not required for TNF--mediated
transcriptional repression of COL1A2 promoter activity, and we have
demonstrated that the TaRE is indeed restricted to a short segment of
promoter between nucleotides -271 and -235. Further, we have
identified two nucleotides critical for transcription factor binding to
the TaRE, and we have shown that NF-B is the TaRE binding protein
complex that mediates TNF- inhibitory effect, as evidenced by the
fact that mutations, which abolish NF-B binding to the TaRE, also
prevent TNF--mediated repression of COL1A2 promoter activity. These
data exclude the involvement of upstream Sp1 binding sites between
residues -286 and -330 in growth factor regulation of COL1A2 promoter
activity as previously believed (1, 2) but are in total agreement with
the previous demonstration of the fundamental role of these sites
for high COL1A2 promoter basal activity (3, 4).
It should be noted that COL1A2 promoter response to TGF-ß requires
the AP-1 site slightly upstream of the TaRE described in this study
(4). These two growth factor response elements bind different
transcription factors, and mutations in either binding site only
prevent the effect of the corresponding cytokine. The close proximity
and partial physical overlap of these sites suggest that the antagonist
activity of TNF- and TGF-ß may be the result of a steric
interaction between the respective transcription factors. Specifically,
it is conceivable that a protein complex bound to either site will
prevent binding of transcription factor(s) to the adjacent site, or it
will be displaced depending on their respective binding affinity.
The mechanism described above differs significantly from the recently
described antagonism between TGF-ß and TNF- on
NF-1/CTF-1-mediated transcription (26). Specifically, in the
latter case, a single element, CTF-1, is the molecular target for
mutually antagonistic regulation by TGF-ß and TNF-. CTF-1/NF-1 is
involved in the activation of murine COL1A2 promoter by TGF-ß (27).
However, no data are currently available to support the possibility
that TNF- may exert an antagonistic activity through this element in
the case of the murine COL1A2 promoter. Regarding the human COL1A2
promoter, we and others have shown that CTF-1 is not involved in either
TGF-ß or TNF- regulation. Instead, Sp1 and AP-1 transcription
factors, as well as NF-B, have been implicated (1, 2, 3, 4). In the
present study, we have demonstrated the fundamental role played by
NF-B in TNF- down-regulation of the human COL1A2 promoter. It is
tempting to speculate that the antagonism between TGF-ß and TNF-
in the context of the human COL1A2 promoter may result from steric
interactions between NF-B and AP-1 proteins binding the
corresponding cis element immediately upstream of the
NF-B binding site (4).
As described in other NF-B elements, discrete variations within the
B element are critical in defining NF-B binding specificity (13, 28). Indeed, a double mutation within the NF-B1 binding half site of
the COL1A2 TaRE cis element abrogated both NF-B complex
formation and TNF- responsiveness, whereas mutation of the
downstream or RelA binding half site had little effect on either of
these parameters. These findings are not surprising considering the
roles played by each subunit in transcriptional regulation and their
relative binding affinities to the NF-B element. Whereas RelA
confers transcriptional modulation of the basal promoter machinery,
NF-B1 serves predominantly as a DNA binding anchor for the
heterodimer to its cognate DNA element (29, 30, 31, 32). These functional
differences are reflected in the relative binding affinities of the
NF-B1 and RelA homodimer complexes to the Ig- enhancer element
(Kd NF-B1/NF-B1 = 0.9 x
10-12 M; Kd RelA/RelA =
32.2 x 10-12 M). These differences in
Kd explain the relative insensitivity of NF-B
heterodimer binding to sites mutated in their lower affinity RelA half,
as shown for the Ig- element (33), or the COL1A2 TaRE
(present study).
Although NF-B sites are generally considered enhancer elements, we
have found that NF-B heterodimer binding to the COL1A2 TaRE is
responsible for TNF--mediated repression of COL1A2 promoter
activity. In fact, we have shown that NF-B overexpression
significantly reduces the activity of both -342 and -271COL1A2/CAT
constructs, whereas a shorter construct lacking the NF-B binding
site was unresponsive (see Fig. 9). Repression of gene transcription by
Rel proteins has been previously described in other systems, and
additional levels of control may be imposed upon their ability to
transactivate gene transcription (34, 35, 36, 37, 38, 39, 40). For example, Rel subunits
can pair outside their family with other transcription factors through
the Rel homology domain, subsequently modulating transcriptional
activity (34). Further, the interaction of activated NF-B complexes
with other nucleoplasmic DNA binding factors as well as nuclear matrix
proteins has been shown to modulate the binding characteristics of
NF-B, hence its transactivation properties (35). In the case of the
Drosophila protein, Dorsal, a member of the Rel family, a
high mobility group (HMG)-like protein termed dorsal switch protein
(DSP1) can, through its interaction with a negative regulatory element
flanking the IFN-ß NF-B element, convert Dorsal from
transcriptional activator to repressor (36, 37). DSP1 could also exert
the same effect on human NF-B, suggesting that a mammalian
counterpart to DSP1 may exist (38, 39). Together, these data indicate
that in the presence of different nuclear factors, NF-B may act
positively or negatively on transcription. This theory is indeed
demonstrated in mice inactivated for c-rel, in which Rel can
act either as a transcriptional activator or as a repressor of TNF-,
IL-6, granulocyte-macrophage CSF, and granulocyte CSF genes in
different mature macrophage populations (40). In the present study,
although direct evidence for a functional inhibitory role of RelA was
demonstrated, it cannot totally be excluded that other factors,
including c-rel, may be involved in this phenomenon.
In summary, we have clarified the COL1A2 promoter TNF--responsive
element as a short region immediately downstream and independent from
three tandem Sp1 elements that regulate COL1A2 basal activity but are
not involved in growth factor modulation. Within this TaRE, a solitary
NF-B site rapidly binds NF-B in response to TNF- challenge,
resulting in a pronounced down-regulation of 2(I) collagen gene
transcription.
|
Acknowledgments |
|---|
B proteins, Dr. Nancy Rice (National Cancer Institute, Frederick,
MD) for NF-B and NF-AT Abs, Dr. Hansjuerg Alder (Kimmel Cancer
Center, Thomas Jefferson University) for technical advice, and Ying-Jee
Song for expert technical assistance. This work was performed in
partial fulfillment of a Ph.D. degree for D.J.K. |
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Alain Mauviel, Department of Dermatology and Cutaneous Biology, Jefferson Medical College, 233 South 10th Street, BLSB Room 430, Philadelphia, PA 19107. E-mail address:
3 Abbreviations used in this paper: ECM, extracellular matrix; TaRE, TNF- response element; CAT, chloramphenicol acetyltransferase; NF-AT, nuclear factor of activated T cells; EMSA, electrophoretic mobility supershift assays; TK, thymidine kinase; WT, wild type; AP-1, activator protein-1; CTF-1, CCAAT-box binding transcription factor-1.
Received for publication April 23, 1998. Accepted for publication December 28, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2(I) collagen gene expression through a cis-acting element that contains an Sp1-binding site. J. Biol. Chem. 269:14828. and transforming growth factor-ß on 2(I) collagen gene transcription. J. Biol. Chem. 270:3353.2(I) collagen gene. J. Biol. Chem. 271:26717.2(I) collagen (COL1A2) promoter activity by transforming growth factor-ß. J. Biol. Chem. 271:3272.B2 with RelA (p65) regulates DNA binding, transcriptional activation, and inhibition by an IBa (MAD3). Mol. Cell. Biol. 13:1315. B: a pleiotropic mediator of inducible and tissue-specific gene control. Cell 58:227.[Medline]
B p50 has I-B-like functions. Cell 71:243.[Medline]
B2 (lyt-10) and c-rel: a mechanism for HTLV-1 Tax-mediated trans-activation via the NF-B. Oncogene 9:841.[Medline]
B site-dependent activation of the interleukin-2 receptor chain gene promoter by human c-rel. Mol. Cell. Biol. 12:4067.B. Annu. Rev. Cell Biol. 10:405.
B in HL60 cells by tumor necrosis factors and ß. Mol. Cell. Biol. 11:2315.B activity is dependent on protein synthesis and the continuous presence of external stimuli. Mol. Cell. Biol. 11:259.B from insect cells using a baculovirus vector. Protein Express. Purif. 9:40.[Medline]
on collagen production and corresponding procollagen mRNA levels in human dermal fibroblasts. J. Invest. Dermatol. 96:243.[Medline]
and interferon- suppress the activation of human type I collagen gene expression by transforming growth factor-ß1. J. Clin. Invest. 86:1489.
inhibits collagen gene transcription and collagen synthesis in cultured human fibroblasts. J. Biol. Chem. 263:5841.. J. Biol. Chem. 271:29672.B subunits determine the specificity of transcriptional activation. Proc. Natl. Acad. Sci. USA 89:1529.B is required for inhibition of NF-B by IB-. Genes Dev. 3:1689.B. EMBO J. 10:3805.[Medline]
B is a receptor for IB- and a modulator of DNA-binding specificity. Genes Dev. 4:1975.B: structure-function relationship of its protein subunits. Biochem. J. 290:297.
B. Mol. Cell. Biol. 14:1939.B activity. Nucleic Acids Res. 24:1895.B: ten years after. Cell 87:13.[Medline]
B-dependent virus induction of the human IFN-ß gene. Cell 71:777.[Medline]
B: a lesson in family values. Cell 80:529.[Medline]
B-like transcription factors is a positive and negative regulator of macrophage gene expression: distinct roles for Rel in different macrophage populations. EMBO J. 15:7099.[Medline]
This article has been cited by other articles:
|
|
 |
|
  T. Miyake and R. Morishita Pharmacological treatment of abdominal aortic aneurysm Cardiovasc Res, August 1, 2009; 83(3): 436 - 443. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Miyake, M. Aoki, H. Masaki, T. Kawasaki, M. Oishi, K. Kataoka, T. Ogihara, Y. Kaneda, and R. Morishita Regression of Abdominal Aortic Aneurysms by Simultaneous Inhibition of Nuclear Factor {kappa}B and Ets in a Rabbit Model Circ. Res., November 26, 2007; 101(11): 1175 - 1184. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Miltyk, E. Karna, and J. A. Palka Prolidase-Independent Mechanism of Camptothecin-Induced Inhibition of Collagen Biosynthesis in Cultured Human Skin Fibroblasts J. Biochem., February 1, 2007; 141(2): 287 - 292. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. H. Durand, V. Flacher, A. Romeas, F. Carrouel, E. Colomb, C. Vincent, H. Magloire, M.-L. Couble, F. Bleicher, M.-J. Staquet, et al. Lipoteichoic Acid Increases TLR and Functional Chemokine Expression while Reducing Dentin Formation in In Vitro Differentiated Human Odontoblasts. J. Immunol., March 1, 2006; 176(5): 2880 - 2887. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. H. Sohn, C. B. Deming, D. C. Johns, H. C. Champion, C. Bian, K. Gardner, and J. J. Rade Regulation of endothelial thrombomodulin expression by inflammatory cytokines is mediated by activation of nuclear factor-kappa B Blood, May 15, 2005; 105(10): 3910 - 3917. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Tanaka, T. T. Antoniv, K. Liu, L. Wang, D. J. Wells, F. Ramirez, and G. Bou-Gharios Cooperativity between Far Upstream Enhancer and Proximal Promoter Elements of the Human {alpha}2(I) Collagen (COL1A2) Gene Instructs Tissue Specificity in Transgenic Mice J. Biol. Chem., December 31, 2004; 279(53): 56024 - 56031. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Novitskiy, J. J. Potter, L. Rennie-Tankersley, and E. Mezey Identification of a Novel NF-{kappa}B-binding Site with Regulation of the Murine {alpha}2(I) Collagen Promoter J. Biol. Chem., April 9, 2004; 279(15): 15639 - 15644. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Kubo, J. Czuwara-Ladykowska, O. Moussa, M. Markiewicz, E. Smith, R. M. Silver, S. Jablonska, M. Blaszczyk, D. K. Watson, and M. Trojanowska Persistent Down-Regulation of Fli1, a Suppressor of Collagen Transcription, in Fibrotic Scleroderma Skin Am. J. Pathol., August 1, 2003; 163(2): 571 - 581. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Higashi, Y. Inagaki, N. Suzuki, S. Mitsui, A. Mauviel, H. Kaneko, and I. Nakatsuka Y-box-binding Protein YB-1 Mediates Transcriptional Repression of Human alpha 2(I) Collagen Gene Expression by Interferon-gamma J. Biol. Chem., February 7, 2003; 278(7): 5156 - 5162. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. M. Luchetti, P. Paroncini, P. Majlingova, J. Frampton, M. Mucenski, S. S. Baroni, P. Sambo, J. Golay, M. Introna, and A. Gabrielli Characterization of the c-Myb-responsive Region and Regulation of the Human Type I Collagen alpha 2 Chain Gene by c-Myb J. Biol. Chem., January 10, 2003; 278(3): 1533 - 1541. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. K. Sengupta, J. Fargo, and B. D. Smith The RFX Family Interacts at the Collagen (COL1A2) Start Site and Represses Transcription J. Biol. Chem., July 5, 2002; 277(28): 24926 - 24937. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. K. Ghosh Factors Involved in the Regulation of Type I Collagen Gene Expression: Implication in Fibrosis Experimental Biology and Medicine, May 1, 2002; 227(5): 301 - 314. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Vermes, R. Chandrasekaran, J. J. Jacobs, J. O. Galante, K. A. Roebuck, and T. T. Glant The Effects of Particulate Wear Debris, Cytokines, and Growth Factors on the Functions of MG-63 Osteoblasts J. Bone Joint Surg. Am., February 1, 2001; 83(2): 201 - 201. [Abstract] [Full Text]
|
|
|
|
|
|
F. Verrecchia, M. Pessah, A. Atfi, and A. Mauviel Tumor Necrosis Factor-alpha Inhibits Transforming Growth Factor-beta /Smad Signaling in Human Dermal Fibroblasts via AP-1 Activation J. Biol. Chem., September 22, 2000; 275(39): 30226 - 30231. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Czuwara-Ladykowska, F. Shirasaki, P. Jackers, D. K. Watson, and M. Trojanowska Fli-1 Inhibits Collagen Type I Production in Dermal Fibroblasts via an Sp1-dependent Pathway J. Biol. Chem., June 8, 2001; 276(24): 20839 - 20848. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
